From 6219f432ec037317a77c40910da12a626c34af1c Mon Sep 17 00:00:00 2001 From: Bhuvanchandra DV Date: Mon, 19 Oct 2015 21:24:36 +0530 Subject: [PATCH 001/335] vf610_adc: Fix internal temperature calculation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Calculate ADCR_VTEMP25 using VTEMP25 at VREFH_ADC 3V3. Existing calculations consider the typical values provided in datasheet. Those typical values are valid for VREFH_ADC at 3.0V. VTEMP25 is different for different VREFH_ADC voltages. With VREFH_ADC at 3.3V, voltage at 25°C is 0.699V. Hence update the VTEMP25 to 0.699V which gives ADCR@Temp25 as 867. Formula for finding ADCR@Temp25: ADCR@Temp25 = (ADCR@Vdd * V@TEMP25 * 10) / VDDconv ADCR@Vdd for 12-Bit ADC = 4095 VDDconv = VREFH_ADC * 10 VREFH_ADC@3.3V ADCR@Temp25 = (4095 * .699 * 10) / 33 ADCR@Temp25 ~= 867 | VREFH_ADC | V@TEMP25 | VDDconv | ADCR@Temp25 | | 3.0V | 0.696mV | 30 | 950 | | 3.3V | 0.699mV | 33 | 867 | Signed-off-by: Bhuvanchandra DV Acked-by: Fugang Duan Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/adc/vf610_adc.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/iio/adc/vf610_adc.c b/drivers/iio/adc/vf610_adc.c index 6bf4c20eb231..42d7c87ea5be 100644 --- a/drivers/iio/adc/vf610_adc.c +++ b/drivers/iio/adc/vf610_adc.c @@ -103,6 +103,13 @@ #define DEFAULT_SAMPLE_TIME 1000 +/* V at 25°C of 696 mV */ +#define VF610_VTEMP25_3V0 950 +/* V at 25°C of 699 mV */ +#define VF610_VTEMP25_3V3 867 +/* Typical sensor slope coefficient at all temperatures */ +#define VF610_TEMP_SLOPE_COEFF 1840 + enum clk_sel { VF610_ADCIOC_BUSCLK_SET, VF610_ADCIOC_ALTCLK_SET, @@ -635,11 +642,13 @@ static int vf610_read_raw(struct iio_dev *indio_dev, break; case IIO_TEMP: /* - * Calculate in degree Celsius times 1000 - * Using sensor slope of 1.84 mV/°C and - * V at 25°C of 696 mV - */ - *val = 25000 - ((int)info->value - 864) * 1000000 / 1840; + * Calculate in degree Celsius times 1000 + * Using the typical sensor slope of 1.84 mV/°C + * and VREFH_ADC at 3.3V, V at 25°C of 699 mV + */ + *val = 25000 - ((int)info->value - VF610_VTEMP25_3V3) * + 1000000 / VF610_TEMP_SLOPE_COEFF; + break; default: mutex_unlock(&indio_dev->mlock); From aea545fa90812cf68b3bf1f09715c465ba942ab3 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 17 Oct 2015 00:29:55 +0200 Subject: [PATCH 002/335] staging: iio: select IRQ_WORK for IIO_DUMMY_EVGEN The iio dummy code was recently changed to use irq_work_queue, but that code is compiled into the kernel only if IRQ_WORK is set, so we can get a link error here: drivers/built-in.o: In function `iio_evgen_poke': (.text+0x208a04): undefined reference to `irq_work_queue' This changes the Kconfig file to match what other drivers do. Signed-off-by: Arnd Bergmann Fixes: fd2bb310ca3d ("Staging: iio: Move evgen interrupt generation to irq_work") Acked-by: Daniel Baluta Signed-off-by: Jonathan Cameron --- drivers/staging/iio/Kconfig | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/staging/iio/Kconfig b/drivers/staging/iio/Kconfig index 6d5b38d69578..9d7f0004d2d7 100644 --- a/drivers/staging/iio/Kconfig +++ b/drivers/staging/iio/Kconfig @@ -18,7 +18,8 @@ source "drivers/staging/iio/resolver/Kconfig" source "drivers/staging/iio/trigger/Kconfig" config IIO_DUMMY_EVGEN - tristate + tristate + select IRQ_WORK config IIO_SIMPLE_DUMMY tristate "An example driver with no hardware requirements" From 01bb70ae0b98d266fa3e860482c7ce22fa482a6e Mon Sep 17 00:00:00 2001 From: Vladimir Zapolskiy Date: Sat, 17 Oct 2015 21:44:38 +0300 Subject: [PATCH 003/335] iio: lpc32xx_adc: fix warnings caused by enabling unprepared clock If common clock framework is configured, the driver generates a warning, which is fixed by this change: root@devkit3250:~# cat /sys/bus/iio/devices/iio\:device0/in_voltage0_raw ------------[ cut here ]------------ WARNING: CPU: 0 PID: 724 at drivers/clk/clk.c:727 clk_core_enable+0x2c/0xa4() Modules linked in: sc16is7xx snd_soc_uda1380 CPU: 0 PID: 724 Comm: cat Not tainted 4.3.0-rc2+ #198 Hardware name: LPC32XX SoC (Flattened Device Tree) Backtrace: [<>] (dump_backtrace) from [<>] (show_stack+0x18/0x1c) [<>] (show_stack) from [<>] (dump_stack+0x20/0x28) [<>] (dump_stack) from [<>] (warn_slowpath_common+0x90/0xb8) [<>] (warn_slowpath_common) from [<>] (warn_slowpath_null+0x24/0x2c) [<>] (warn_slowpath_null) from [<>] (clk_core_enable+0x2c/0xa4) [<>] (clk_core_enable) from [<>] (clk_enable+0x24/0x38) [<>] (clk_enable) from [<>] (lpc32xx_read_raw+0x38/0x80) [<>] (lpc32xx_read_raw) from [<>] (iio_read_channel_info+0x70/0x94) [<>] (iio_read_channel_info) from [<>] (dev_attr_show+0x28/0x4c) [<>] (dev_attr_show) from [<>] (sysfs_kf_seq_show+0x8c/0xf0) [<>] (sysfs_kf_seq_show) from [<>] (kernfs_seq_show+0x2c/0x30) [<>] (kernfs_seq_show) from [<>] (seq_read+0x1c8/0x440) [<>] (seq_read) from [<>] (kernfs_fop_read+0x38/0x170) [<>] (kernfs_fop_read) from [<>] (do_readv_writev+0x16c/0x238) [<>] (do_readv_writev) from [<>] (vfs_readv+0x50/0x58) [<>] (vfs_readv) from [<>] (default_file_splice_read+0x1a4/0x308) [<>] (default_file_splice_read) from [<>] (do_splice_to+0x78/0x84) [<>] (do_splice_to) from [<>] (splice_direct_to_actor+0xc8/0x1cc) [<>] (splice_direct_to_actor) from [<>] (do_splice_direct+0xa0/0xb8) [<>] (do_splice_direct) from [<>] (do_sendfile+0x1a8/0x30c) [<>] (do_sendfile) from [<>] (SyS_sendfile64+0x104/0x10c) [<>] (SyS_sendfile64) from [<>] (ret_fast_syscall+0x0/0x38) Signed-off-by: Vladimir Zapolskiy Cc: Signed-off-by: Jonathan Cameron --- drivers/staging/iio/adc/lpc32xx_adc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/staging/iio/adc/lpc32xx_adc.c b/drivers/staging/iio/adc/lpc32xx_adc.c index 5331c442fcfc..ef6d0b0ccf55 100644 --- a/drivers/staging/iio/adc/lpc32xx_adc.c +++ b/drivers/staging/iio/adc/lpc32xx_adc.c @@ -76,7 +76,7 @@ static int lpc32xx_read_raw(struct iio_dev *indio_dev, if (mask == IIO_CHAN_INFO_RAW) { mutex_lock(&indio_dev->mlock); - clk_enable(info->clk); + clk_prepare_enable(info->clk); /* Measurement setup */ __raw_writel(AD_INTERNAL | (chan->address) | AD_REFp | AD_REFm, LPC32XX_ADC_SELECT(info->adc_base)); @@ -84,7 +84,7 @@ static int lpc32xx_read_raw(struct iio_dev *indio_dev, __raw_writel(AD_PDN_CTRL | AD_STROBE, LPC32XX_ADC_CTRL(info->adc_base)); wait_for_completion(&info->completion); /* set by ISR */ - clk_disable(info->clk); + clk_disable_unprepare(info->clk); *val = info->value; mutex_unlock(&indio_dev->mlock); From 03fe472ef33b7f31fbd11d300dbb3fdab9c00fd4 Mon Sep 17 00:00:00 2001 From: Michael Hennerich Date: Tue, 13 Oct 2015 18:15:37 +0200 Subject: [PATCH 004/335] iio:ad5064: Make sure ad5064_i2c_write() returns 0 on success i2c_master_send() returns the number of bytes transferred on success while the ad5064 driver expects that the write() callback returns 0 on success. Fix that by translating any non negative return value of i2c_master_send() to 0. Fixes: commit 6a17a0768f77 ("iio:dac:ad5064: Add support for the ad5629r and ad5669r") Signed-off-by: Michael Hennerich Signed-off-by: Lars-Peter Clausen Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/dac/ad5064.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/iio/dac/ad5064.c b/drivers/iio/dac/ad5064.c index c067e6821496..7e7ebf30e954 100644 --- a/drivers/iio/dac/ad5064.c +++ b/drivers/iio/dac/ad5064.c @@ -598,10 +598,16 @@ static int ad5064_i2c_write(struct ad5064_state *st, unsigned int cmd, unsigned int addr, unsigned int val) { struct i2c_client *i2c = to_i2c_client(st->dev); + int ret; st->data.i2c[0] = (cmd << 4) | addr; put_unaligned_be16(val, &st->data.i2c[1]); - return i2c_master_send(i2c, st->data.i2c, 3); + + ret = i2c_master_send(i2c, st->data.i2c, 3); + if (ret < 0) + return ret; + + return 0; } static int ad5064_i2c_probe(struct i2c_client *i2c, From 5dcbe97bedd6ba4b0f574a96cc2e293d26f3d857 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Tue, 13 Oct 2015 18:15:38 +0200 Subject: [PATCH 005/335] iio: ad5064: Fix ad5629/ad5669 shift The ad5629/ad5669 are the I2C variant of the ad5628/ad5668, which has a SPI interface. They are mostly identical with the exception that the shift factor is different. Currently the driver does not take care of this difference which leads to incorrect DAC output values. Fix this by introducing a custom channel spec for the ad5629/ad5669 with the correct shift factor. Fixes: commit 6a17a0768f77 ("iio:dac:ad5064: Add support for the ad5629r and ad5669r") Signed-off-by: Lars-Peter Clausen Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/dac/ad5064.c | 83 +++++++++++++++++++++++++++------------- 1 file changed, 57 insertions(+), 26 deletions(-) diff --git a/drivers/iio/dac/ad5064.c b/drivers/iio/dac/ad5064.c index 7e7ebf30e954..978f130ca678 100644 --- a/drivers/iio/dac/ad5064.c +++ b/drivers/iio/dac/ad5064.c @@ -113,12 +113,16 @@ enum ad5064_type { ID_AD5065, ID_AD5628_1, ID_AD5628_2, + ID_AD5629_1, + ID_AD5629_2, ID_AD5648_1, ID_AD5648_2, ID_AD5666_1, ID_AD5666_2, ID_AD5668_1, ID_AD5668_2, + ID_AD5669_1, + ID_AD5669_2, }; static int ad5064_write(struct ad5064_state *st, unsigned int cmd, @@ -291,7 +295,7 @@ static const struct iio_chan_spec_ext_info ad5064_ext_info[] = { { }, }; -#define AD5064_CHANNEL(chan, addr, bits) { \ +#define AD5064_CHANNEL(chan, addr, bits, _shift) { \ .type = IIO_VOLTAGE, \ .indexed = 1, \ .output = 1, \ @@ -303,36 +307,39 @@ static const struct iio_chan_spec_ext_info ad5064_ext_info[] = { .sign = 'u', \ .realbits = (bits), \ .storagebits = 16, \ - .shift = 20 - bits, \ + .shift = (_shift), \ }, \ .ext_info = ad5064_ext_info, \ } -#define DECLARE_AD5064_CHANNELS(name, bits) \ +#define DECLARE_AD5064_CHANNELS(name, bits, shift) \ const struct iio_chan_spec name[] = { \ - AD5064_CHANNEL(0, 0, bits), \ - AD5064_CHANNEL(1, 1, bits), \ - AD5064_CHANNEL(2, 2, bits), \ - AD5064_CHANNEL(3, 3, bits), \ - AD5064_CHANNEL(4, 4, bits), \ - AD5064_CHANNEL(5, 5, bits), \ - AD5064_CHANNEL(6, 6, bits), \ - AD5064_CHANNEL(7, 7, bits), \ + AD5064_CHANNEL(0, 0, bits, shift), \ + AD5064_CHANNEL(1, 1, bits, shift), \ + AD5064_CHANNEL(2, 2, bits, shift), \ + AD5064_CHANNEL(3, 3, bits, shift), \ + AD5064_CHANNEL(4, 4, bits, shift), \ + AD5064_CHANNEL(5, 5, bits, shift), \ + AD5064_CHANNEL(6, 6, bits, shift), \ + AD5064_CHANNEL(7, 7, bits, shift), \ } -#define DECLARE_AD5065_CHANNELS(name, bits) \ +#define DECLARE_AD5065_CHANNELS(name, bits, shift) \ const struct iio_chan_spec name[] = { \ - AD5064_CHANNEL(0, 0, bits), \ - AD5064_CHANNEL(1, 3, bits), \ + AD5064_CHANNEL(0, 0, bits, shift), \ + AD5064_CHANNEL(1, 3, bits, shift), \ } -static DECLARE_AD5064_CHANNELS(ad5024_channels, 12); -static DECLARE_AD5064_CHANNELS(ad5044_channels, 14); -static DECLARE_AD5064_CHANNELS(ad5064_channels, 16); +static DECLARE_AD5064_CHANNELS(ad5024_channels, 12, 8); +static DECLARE_AD5064_CHANNELS(ad5044_channels, 14, 6); +static DECLARE_AD5064_CHANNELS(ad5064_channels, 16, 4); -static DECLARE_AD5065_CHANNELS(ad5025_channels, 12); -static DECLARE_AD5065_CHANNELS(ad5045_channels, 14); -static DECLARE_AD5065_CHANNELS(ad5065_channels, 16); +static DECLARE_AD5065_CHANNELS(ad5025_channels, 12, 8); +static DECLARE_AD5065_CHANNELS(ad5045_channels, 14, 6); +static DECLARE_AD5065_CHANNELS(ad5065_channels, 16, 4); + +static DECLARE_AD5064_CHANNELS(ad5629_channels, 12, 4); +static DECLARE_AD5064_CHANNELS(ad5669_channels, 16, 0); static const struct ad5064_chip_info ad5064_chip_info_tbl[] = { [ID_AD5024] = { @@ -382,6 +389,18 @@ static const struct ad5064_chip_info ad5064_chip_info_tbl[] = { .channels = ad5024_channels, .num_channels = 8, }, + [ID_AD5629_1] = { + .shared_vref = true, + .internal_vref = 2500000, + .channels = ad5629_channels, + .num_channels = 8, + }, + [ID_AD5629_2] = { + .shared_vref = true, + .internal_vref = 5000000, + .channels = ad5629_channels, + .num_channels = 8, + }, [ID_AD5648_1] = { .shared_vref = true, .internal_vref = 2500000, @@ -418,6 +437,18 @@ static const struct ad5064_chip_info ad5064_chip_info_tbl[] = { .channels = ad5064_channels, .num_channels = 8, }, + [ID_AD5669_1] = { + .shared_vref = true, + .internal_vref = 2500000, + .channels = ad5669_channels, + .num_channels = 8, + }, + [ID_AD5669_2] = { + .shared_vref = true, + .internal_vref = 5000000, + .channels = ad5669_channels, + .num_channels = 8, + }, }; static inline unsigned int ad5064_num_vref(struct ad5064_state *st) @@ -623,12 +654,12 @@ static int ad5064_i2c_remove(struct i2c_client *i2c) } static const struct i2c_device_id ad5064_i2c_ids[] = { - {"ad5629-1", ID_AD5628_1}, - {"ad5629-2", ID_AD5628_2}, - {"ad5629-3", ID_AD5628_2}, /* similar enough to ad5629-2 */ - {"ad5669-1", ID_AD5668_1}, - {"ad5669-2", ID_AD5668_2}, - {"ad5669-3", ID_AD5668_2}, /* similar enough to ad5669-2 */ + {"ad5629-1", ID_AD5629_1}, + {"ad5629-2", ID_AD5629_2}, + {"ad5629-3", ID_AD5629_2}, /* similar enough to ad5629-2 */ + {"ad5669-1", ID_AD5669_1}, + {"ad5669-2", ID_AD5669_2}, + {"ad5669-3", ID_AD5669_2}, /* similar enough to ad5669-2 */ {} }; MODULE_DEVICE_TABLE(i2c, ad5064_i2c_ids); From 785171fd6cd7dcd7ada5a733b6a2d44ec566c3a0 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Mon, 12 Oct 2015 14:56:28 +0200 Subject: [PATCH 006/335] iio:ad7793: Fix ad7785 product ID While the datasheet for the AD7785 lists 0xXB as the product ID the actual product ID is 0xX3. Fix the product ID otherwise the driver will reject the device due to non matching IDs. Fixes: e786cc26dcc5 ("staging:iio:ad7793: Implement stricter id checking") Signed-off-by: Lars-Peter Clausen Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/adc/ad7793.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/iio/adc/ad7793.c b/drivers/iio/adc/ad7793.c index b84922a4b32e..710aee8377bb 100644 --- a/drivers/iio/adc/ad7793.c +++ b/drivers/iio/adc/ad7793.c @@ -101,7 +101,7 @@ #define AD7795_CH_AIN1M_AIN1M 8 /* AIN1(-) - AIN1(-) */ /* ID Register Bit Designations (AD7793_REG_ID) */ -#define AD7785_ID 0xB +#define AD7785_ID 0x3 #define AD7792_ID 0xA #define AD7793_ID 0xB #define AD7794_ID 0xF From 61b590b9ee4221173ad6990a1150c5c9db73564e Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Fri, 23 Oct 2015 12:43:18 +0200 Subject: [PATCH 007/335] netfilter: ingress: don't use nf_hook_list_active nf_hook_list_active() always returns true once at least one device has NF_INGRESS hook enabled. Thus, don't use this function. Instead, inverse the test and use the static key to elide list_empty test if no NF_INGRESS hooks are active. Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter_ingress.h | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/include/linux/netfilter_ingress.h b/include/linux/netfilter_ingress.h index 187feabe557c..ba7ce8805fe3 100644 --- a/include/linux/netfilter_ingress.h +++ b/include/linux/netfilter_ingress.h @@ -5,10 +5,13 @@ #include #ifdef CONFIG_NETFILTER_INGRESS -static inline int nf_hook_ingress_active(struct sk_buff *skb) +static inline bool nf_hook_ingress_active(const struct sk_buff *skb) { - return nf_hook_list_active(&skb->dev->nf_hooks_ingress, - NFPROTO_NETDEV, NF_NETDEV_INGRESS); +#ifdef HAVE_JUMP_LABEL + if (!static_key_false(&nf_hooks_needed[NFPROTO_NETDEV][NF_NETDEV_INGRESS])) + return false; +#endif + return !list_empty(&skb->dev->nf_hooks_ingress); } static inline int nf_hook_ingress(struct sk_buff *skb) From b4865988eab598e56e6e628b9b32441acd142b28 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 6 Nov 2015 18:35:57 +0100 Subject: [PATCH 008/335] netfilter: ingress: fix wrong input interface on hook The input and output interfaces in nf_hook_state_init() are flipped. This fixes iif matching on nftables. Reported-by: Patrick McHardy Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter_ingress.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/linux/netfilter_ingress.h b/include/linux/netfilter_ingress.h index ba7ce8805fe3..5fcd375ef175 100644 --- a/include/linux/netfilter_ingress.h +++ b/include/linux/netfilter_ingress.h @@ -19,8 +19,8 @@ static inline int nf_hook_ingress(struct sk_buff *skb) struct nf_hook_state state; nf_hook_state_init(&state, &skb->dev->nf_hooks_ingress, - NF_NETDEV_INGRESS, INT_MIN, NFPROTO_NETDEV, NULL, - skb->dev, NULL, dev_net(skb->dev), NULL); + NF_NETDEV_INGRESS, INT_MIN, NFPROTO_NETDEV, + skb->dev, NULL, NULL, dev_net(skb->dev), NULL); return nf_hook_slow(skb, &state); } From 95ad1f4a9358dff1dcf84bf5c9cc84caa9215f7f Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Sat, 7 Nov 2015 11:21:47 +0100 Subject: [PATCH 009/335] netfilter: ipset: Fix extension alignment The data extensions in ipset lacked the proper memory alignment and thus could lead to kernel crash on several architectures. Therefore the structures have been reorganized and alignment attributes added where needed. The patch was tested on armv7h by Gerhard Wiesinger and on x86_64, sparc64 by Jozsef Kadlecsik. Reported-by: Gerhard Wiesinger Tested-by: Gerhard Wiesinger Tested-by: Jozsef Kadlecsik Signed-off-by: Jozsef Kadlecsik --- include/linux/netfilter/ipset/ip_set.h | 2 +- net/netfilter/ipset/ip_set_bitmap_gen.h | 17 +++--- net/netfilter/ipset/ip_set_bitmap_ip.c | 14 ++--- net/netfilter/ipset/ip_set_bitmap_ipmac.c | 64 ++++++++++------------- net/netfilter/ipset/ip_set_bitmap_port.c | 18 +++---- net/netfilter/ipset/ip_set_core.c | 14 ++--- net/netfilter/ipset/ip_set_hash_gen.h | 11 ++-- net/netfilter/ipset/ip_set_list_set.c | 5 +- 8 files changed, 65 insertions(+), 80 deletions(-) diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h index 48bb01edcf30..0e1f433cc4b7 100644 --- a/include/linux/netfilter/ipset/ip_set.h +++ b/include/linux/netfilter/ipset/ip_set.h @@ -421,7 +421,7 @@ extern void ip_set_free(void *members); extern int ip_set_get_ipaddr4(struct nlattr *nla, __be32 *ipaddr); extern int ip_set_get_ipaddr6(struct nlattr *nla, union nf_inet_addr *ipaddr); extern size_t ip_set_elem_len(struct ip_set *set, struct nlattr *tb[], - size_t len); + size_t len, size_t align); extern int ip_set_get_extensions(struct ip_set *set, struct nlattr *tb[], struct ip_set_ext *ext); diff --git a/net/netfilter/ipset/ip_set_bitmap_gen.h b/net/netfilter/ipset/ip_set_bitmap_gen.h index d05e759ed0fa..b0bc475f641e 100644 --- a/net/netfilter/ipset/ip_set_bitmap_gen.h +++ b/net/netfilter/ipset/ip_set_bitmap_gen.h @@ -33,7 +33,7 @@ #define mtype_gc IPSET_TOKEN(MTYPE, _gc) #define mtype MTYPE -#define get_ext(set, map, id) ((map)->extensions + (set)->dsize * (id)) +#define get_ext(set, map, id) ((map)->extensions + ((set)->dsize * (id))) static void mtype_gc_init(struct ip_set *set, void (*gc)(unsigned long ul_set)) @@ -67,12 +67,9 @@ mtype_destroy(struct ip_set *set) del_timer_sync(&map->gc); ip_set_free(map->members); - if (set->dsize) { - if (set->extensions & IPSET_EXT_DESTROY) - mtype_ext_cleanup(set); - ip_set_free(map->extensions); - } - kfree(map); + if (set->dsize && set->extensions & IPSET_EXT_DESTROY) + mtype_ext_cleanup(set); + ip_set_free(map); set->data = NULL; } @@ -92,16 +89,14 @@ mtype_head(struct ip_set *set, struct sk_buff *skb) { const struct mtype *map = set->data; struct nlattr *nested; + size_t memsize = sizeof(*map) + map->memsize; nested = ipset_nest_start(skb, IPSET_ATTR_DATA); if (!nested) goto nla_put_failure; if (mtype_do_head(skb, map) || nla_put_net32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)) || - nla_put_net32(skb, IPSET_ATTR_MEMSIZE, - htonl(sizeof(*map) + - map->memsize + - set->dsize * map->elements))) + nla_put_net32(skb, IPSET_ATTR_MEMSIZE, htonl(memsize))) goto nla_put_failure; if (unlikely(ip_set_put_flags(skb, set))) goto nla_put_failure; diff --git a/net/netfilter/ipset/ip_set_bitmap_ip.c b/net/netfilter/ipset/ip_set_bitmap_ip.c index 64a564334418..4783efff0bde 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ip.c +++ b/net/netfilter/ipset/ip_set_bitmap_ip.c @@ -41,7 +41,6 @@ MODULE_ALIAS("ip_set_bitmap:ip"); /* Type structure */ struct bitmap_ip { void *members; /* the set members */ - void *extensions; /* data extensions */ u32 first_ip; /* host byte order, included in range */ u32 last_ip; /* host byte order, included in range */ u32 elements; /* number of max elements in the set */ @@ -49,6 +48,8 @@ struct bitmap_ip { size_t memsize; /* members size */ u8 netmask; /* subnet netmask */ struct timer_list gc; /* garbage collection */ + unsigned char extensions[0] /* data extensions */ + __aligned(__alignof__(u64)); }; /* ADT structure for generic function args */ @@ -224,13 +225,6 @@ init_map_ip(struct ip_set *set, struct bitmap_ip *map, map->members = ip_set_alloc(map->memsize); if (!map->members) return false; - if (set->dsize) { - map->extensions = ip_set_alloc(set->dsize * elements); - if (!map->extensions) { - kfree(map->members); - return false; - } - } map->first_ip = first_ip; map->last_ip = last_ip; map->elements = elements; @@ -316,13 +310,13 @@ bitmap_ip_create(struct net *net, struct ip_set *set, struct nlattr *tb[], pr_debug("hosts %u, elements %llu\n", hosts, (unsigned long long)elements); - map = kzalloc(sizeof(*map), GFP_KERNEL); + set->dsize = ip_set_elem_len(set, tb, 0, 0); + map = ip_set_alloc(sizeof(*map) + elements * set->dsize); if (!map) return -ENOMEM; map->memsize = bitmap_bytes(0, elements - 1); set->variant = &bitmap_ip; - set->dsize = ip_set_elem_len(set, tb, 0); if (!init_map_ip(set, map, first_ip, last_ip, elements, hosts, netmask)) { kfree(map); diff --git a/net/netfilter/ipset/ip_set_bitmap_ipmac.c b/net/netfilter/ipset/ip_set_bitmap_ipmac.c index 1430535118fb..29dde208381d 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ipmac.c +++ b/net/netfilter/ipset/ip_set_bitmap_ipmac.c @@ -47,24 +47,26 @@ enum { /* Type structure */ struct bitmap_ipmac { void *members; /* the set members */ - void *extensions; /* MAC + data extensions */ u32 first_ip; /* host byte order, included in range */ u32 last_ip; /* host byte order, included in range */ u32 elements; /* number of max elements in the set */ size_t memsize; /* members size */ struct timer_list gc; /* garbage collector */ + unsigned char extensions[0] /* MAC + data extensions */ + __aligned(__alignof__(u64)); }; /* ADT structure for generic function args */ struct bitmap_ipmac_adt_elem { + unsigned char ether[ETH_ALEN] __aligned(2); u16 id; - unsigned char *ether; + u16 add_mac; }; struct bitmap_ipmac_elem { unsigned char ether[ETH_ALEN]; unsigned char filled; -} __attribute__ ((aligned)); +} __aligned(__alignof__(u64)); static inline u32 ip_to_id(const struct bitmap_ipmac *m, u32 ip) @@ -72,11 +74,11 @@ ip_to_id(const struct bitmap_ipmac *m, u32 ip) return ip - m->first_ip; } -static inline struct bitmap_ipmac_elem * -get_elem(void *extensions, u16 id, size_t dsize) -{ - return (struct bitmap_ipmac_elem *)(extensions + id * dsize); -} +#define get_elem(extensions, id, dsize) \ + (struct bitmap_ipmac_elem *)(extensions + (id) * (dsize)) + +#define get_const_elem(extensions, id, dsize) \ + (const struct bitmap_ipmac_elem *)(extensions + (id) * (dsize)) /* Common functions */ @@ -88,10 +90,9 @@ bitmap_ipmac_do_test(const struct bitmap_ipmac_adt_elem *e, if (!test_bit(e->id, map->members)) return 0; - elem = get_elem(map->extensions, e->id, dsize); - if (elem->filled == MAC_FILLED) - return !e->ether || - ether_addr_equal(e->ether, elem->ether); + elem = get_const_elem(map->extensions, e->id, dsize); + if (e->add_mac && elem->filled == MAC_FILLED) + return ether_addr_equal(e->ether, elem->ether); /* Trigger kernel to fill out the ethernet address */ return -EAGAIN; } @@ -103,7 +104,7 @@ bitmap_ipmac_gc_test(u16 id, const struct bitmap_ipmac *map, size_t dsize) if (!test_bit(id, map->members)) return 0; - elem = get_elem(map->extensions, id, dsize); + elem = get_const_elem(map->extensions, id, dsize); /* Timer not started for the incomplete elements */ return elem->filled == MAC_FILLED; } @@ -133,7 +134,7 @@ bitmap_ipmac_add_timeout(unsigned long *timeout, * and we can reuse it later when MAC is filled out, * possibly by the kernel */ - if (e->ether) + if (e->add_mac) ip_set_timeout_set(timeout, t); else *timeout = t; @@ -150,7 +151,7 @@ bitmap_ipmac_do_add(const struct bitmap_ipmac_adt_elem *e, elem = get_elem(map->extensions, e->id, dsize); if (test_bit(e->id, map->members)) { if (elem->filled == MAC_FILLED) { - if (e->ether && + if (e->add_mac && (flags & IPSET_FLAG_EXIST) && !ether_addr_equal(e->ether, elem->ether)) { /* memcpy isn't atomic */ @@ -159,7 +160,7 @@ bitmap_ipmac_do_add(const struct bitmap_ipmac_adt_elem *e, ether_addr_copy(elem->ether, e->ether); } return IPSET_ADD_FAILED; - } else if (!e->ether) + } else if (!e->add_mac) /* Already added without ethernet address */ return IPSET_ADD_FAILED; /* Fill the MAC address and trigger the timer activation */ @@ -168,7 +169,7 @@ bitmap_ipmac_do_add(const struct bitmap_ipmac_adt_elem *e, ether_addr_copy(elem->ether, e->ether); elem->filled = MAC_FILLED; return IPSET_ADD_START_STORED_TIMEOUT; - } else if (e->ether) { + } else if (e->add_mac) { /* We can store MAC too */ ether_addr_copy(elem->ether, e->ether); elem->filled = MAC_FILLED; @@ -191,7 +192,7 @@ bitmap_ipmac_do_list(struct sk_buff *skb, const struct bitmap_ipmac *map, u32 id, size_t dsize) { const struct bitmap_ipmac_elem *elem = - get_elem(map->extensions, id, dsize); + get_const_elem(map->extensions, id, dsize); return nla_put_ipaddr4(skb, IPSET_ATTR_IP, htonl(map->first_ip + id)) || @@ -213,7 +214,7 @@ bitmap_ipmac_kadt(struct ip_set *set, const struct sk_buff *skb, { struct bitmap_ipmac *map = set->data; ipset_adtfn adtfn = set->variant->adt[adt]; - struct bitmap_ipmac_adt_elem e = { .id = 0 }; + struct bitmap_ipmac_adt_elem e = { .id = 0, .add_mac = 1 }; struct ip_set_ext ext = IP_SET_INIT_KEXT(skb, opt, set); u32 ip; @@ -231,7 +232,7 @@ bitmap_ipmac_kadt(struct ip_set *set, const struct sk_buff *skb, return -EINVAL; e.id = ip_to_id(map, ip); - e.ether = eth_hdr(skb)->h_source; + memcpy(e.ether, eth_hdr(skb)->h_source, ETH_ALEN); return adtfn(set, &e, &ext, &opt->ext, opt->cmdflags); } @@ -265,11 +266,10 @@ bitmap_ipmac_uadt(struct ip_set *set, struct nlattr *tb[], return -IPSET_ERR_BITMAP_RANGE; e.id = ip_to_id(map, ip); - if (tb[IPSET_ATTR_ETHER]) - e.ether = nla_data(tb[IPSET_ATTR_ETHER]); - else - e.ether = NULL; - + if (tb[IPSET_ATTR_ETHER]) { + memcpy(e.ether, nla_data(tb[IPSET_ATTR_ETHER]), ETH_ALEN); + e.add_mac = 1; + } ret = adtfn(set, &e, &ext, &ext, flags); return ip_set_eexist(ret, flags) ? 0 : ret; @@ -300,13 +300,6 @@ init_map_ipmac(struct ip_set *set, struct bitmap_ipmac *map, map->members = ip_set_alloc(map->memsize); if (!map->members) return false; - if (set->dsize) { - map->extensions = ip_set_alloc(set->dsize * elements); - if (!map->extensions) { - kfree(map->members); - return false; - } - } map->first_ip = first_ip; map->last_ip = last_ip; map->elements = elements; @@ -361,14 +354,15 @@ bitmap_ipmac_create(struct net *net, struct ip_set *set, struct nlattr *tb[], if (elements > IPSET_BITMAP_MAX_RANGE + 1) return -IPSET_ERR_BITMAP_RANGE_SIZE; - map = kzalloc(sizeof(*map), GFP_KERNEL); + set->dsize = ip_set_elem_len(set, tb, + sizeof(struct bitmap_ipmac_elem), + __alignof__(struct bitmap_ipmac_elem)); + map = ip_set_alloc(sizeof(*map) + elements * set->dsize); if (!map) return -ENOMEM; map->memsize = bitmap_bytes(0, elements - 1); set->variant = &bitmap_ipmac; - set->dsize = ip_set_elem_len(set, tb, - sizeof(struct bitmap_ipmac_elem)); if (!init_map_ipmac(set, map, first_ip, last_ip, elements)) { kfree(map); return -ENOMEM; diff --git a/net/netfilter/ipset/ip_set_bitmap_port.c b/net/netfilter/ipset/ip_set_bitmap_port.c index 5338ccd5da46..7f0c733358a4 100644 --- a/net/netfilter/ipset/ip_set_bitmap_port.c +++ b/net/netfilter/ipset/ip_set_bitmap_port.c @@ -35,12 +35,13 @@ MODULE_ALIAS("ip_set_bitmap:port"); /* Type structure */ struct bitmap_port { void *members; /* the set members */ - void *extensions; /* data extensions */ u16 first_port; /* host byte order, included in range */ u16 last_port; /* host byte order, included in range */ u32 elements; /* number of max elements in the set */ size_t memsize; /* members size */ struct timer_list gc; /* garbage collection */ + unsigned char extensions[0] /* data extensions */ + __aligned(__alignof__(u64)); }; /* ADT structure for generic function args */ @@ -209,13 +210,6 @@ init_map_port(struct ip_set *set, struct bitmap_port *map, map->members = ip_set_alloc(map->memsize); if (!map->members) return false; - if (set->dsize) { - map->extensions = ip_set_alloc(set->dsize * map->elements); - if (!map->extensions) { - kfree(map->members); - return false; - } - } map->first_port = first_port; map->last_port = last_port; set->timeout = IPSET_NO_TIMEOUT; @@ -232,6 +226,7 @@ bitmap_port_create(struct net *net, struct ip_set *set, struct nlattr *tb[], { struct bitmap_port *map; u16 first_port, last_port; + u32 elements; if (unlikely(!ip_set_attr_netorder(tb, IPSET_ATTR_PORT) || !ip_set_attr_netorder(tb, IPSET_ATTR_PORT_TO) || @@ -248,14 +243,15 @@ bitmap_port_create(struct net *net, struct ip_set *set, struct nlattr *tb[], last_port = tmp; } - map = kzalloc(sizeof(*map), GFP_KERNEL); + elements = last_port - first_port + 1; + set->dsize = ip_set_elem_len(set, tb, 0, 0); + map = ip_set_alloc(sizeof(*map) + elements * set->dsize); if (!map) return -ENOMEM; - map->elements = last_port - first_port + 1; + map->elements = elements; map->memsize = bitmap_bytes(0, map->elements); set->variant = &bitmap_port; - set->dsize = ip_set_elem_len(set, tb, 0); if (!init_map_port(set, map, first_port, last_port)) { kfree(map); return -ENOMEM; diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index 69ab9c2634e1..54f3d7cb23e6 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -364,25 +364,27 @@ add_extension(enum ip_set_ext_id id, u32 flags, struct nlattr *tb[]) } size_t -ip_set_elem_len(struct ip_set *set, struct nlattr *tb[], size_t len) +ip_set_elem_len(struct ip_set *set, struct nlattr *tb[], size_t len, + size_t align) { enum ip_set_ext_id id; - size_t offset = len; u32 cadt_flags = 0; if (tb[IPSET_ATTR_CADT_FLAGS]) cadt_flags = ip_set_get_h32(tb[IPSET_ATTR_CADT_FLAGS]); if (cadt_flags & IPSET_FLAG_WITH_FORCEADD) set->flags |= IPSET_CREATE_FLAG_FORCEADD; + if (!align) + align = 1; for (id = 0; id < IPSET_EXT_ID_MAX; id++) { if (!add_extension(id, cadt_flags, tb)) continue; - offset = ALIGN(offset, ip_set_extensions[id].align); - set->offset[id] = offset; + len = ALIGN(len, ip_set_extensions[id].align); + set->offset[id] = len; set->extensions |= ip_set_extensions[id].type; - offset += ip_set_extensions[id].len; + len += ip_set_extensions[id].len; } - return offset; + return ALIGN(len, align); } EXPORT_SYMBOL_GPL(ip_set_elem_len); diff --git a/net/netfilter/ipset/ip_set_hash_gen.h b/net/netfilter/ipset/ip_set_hash_gen.h index 691b54fcaf2a..4ff22194ce55 100644 --- a/net/netfilter/ipset/ip_set_hash_gen.h +++ b/net/netfilter/ipset/ip_set_hash_gen.h @@ -72,8 +72,9 @@ struct hbucket { DECLARE_BITMAP(used, AHASH_MAX_TUNED); u8 size; /* size of the array */ u8 pos; /* position of the first free entry */ - unsigned char value[0]; /* the array of the values */ -} __attribute__ ((aligned)); + unsigned char value[0] /* the array of the values */ + __aligned(__alignof__(u64)); +}; /* The hash table: the table size stored here in order to make resizing easy */ struct htable { @@ -1323,12 +1324,14 @@ IPSET_TOKEN(HTYPE, _create)(struct net *net, struct ip_set *set, #endif set->variant = &IPSET_TOKEN(HTYPE, 4_variant); set->dsize = ip_set_elem_len(set, tb, - sizeof(struct IPSET_TOKEN(HTYPE, 4_elem))); + sizeof(struct IPSET_TOKEN(HTYPE, 4_elem)), + __alignof__(struct IPSET_TOKEN(HTYPE, 4_elem))); #ifndef IP_SET_PROTO_UNDEF } else { set->variant = &IPSET_TOKEN(HTYPE, 6_variant); set->dsize = ip_set_elem_len(set, tb, - sizeof(struct IPSET_TOKEN(HTYPE, 6_elem))); + sizeof(struct IPSET_TOKEN(HTYPE, 6_elem)), + __alignof__(struct IPSET_TOKEN(HTYPE, 6_elem))); } #endif if (tb[IPSET_ATTR_TIMEOUT]) { diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index 5a30ce6e8c90..bbede95c9f68 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -31,7 +31,7 @@ struct set_elem { struct rcu_head rcu; struct list_head list; ip_set_id_t id; -}; +} __aligned(__alignof__(u64)); struct set_adt_elem { ip_set_id_t id; @@ -618,7 +618,8 @@ list_set_create(struct net *net, struct ip_set *set, struct nlattr *tb[], size = IP_SET_LIST_MIN_SIZE; set->variant = &set_variant; - set->dsize = ip_set_elem_len(set, tb, sizeof(struct set_elem)); + set->dsize = ip_set_elem_len(set, tb, sizeof(struct set_elem), + __alignof__(struct set_elem)); if (!init_list_set(net, set, size)) return -ENOMEM; if (tb[IPSET_ATTR_TIMEOUT]) { From e9dfdc052d018268926ab769d5b7598226713d5a Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Sat, 7 Nov 2015 11:23:34 +0100 Subject: [PATCH 010/335] netfilter: ipset: Fix hash:* type expiration Incorrect index was used when the data blob was shrinked at expiration, which could lead to falsely expired entries and memory leak when the comment extension was used too. Signed-off-by: Jozsef Kadlecsik --- net/netfilter/ipset/ip_set_hash_gen.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/ipset/ip_set_hash_gen.h b/net/netfilter/ipset/ip_set_hash_gen.h index 4ff22194ce55..fa4f6374bb73 100644 --- a/net/netfilter/ipset/ip_set_hash_gen.h +++ b/net/netfilter/ipset/ip_set_hash_gen.h @@ -523,7 +523,7 @@ mtype_expire(struct ip_set *set, struct htype *h, u8 nets_length, size_t dsize) continue; data = ahash_data(n, j, dsize); memcpy(tmp->value + d * dsize, data, dsize); - set_bit(j, tmp->used); + set_bit(d, tmp->used); d++; } tmp->pos = d; From 0aae24eb409fc429f54ca3809f904f1b91e295e0 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Sat, 7 Nov 2015 11:24:51 +0100 Subject: [PATCH 011/335] netfilter: ipset: Fix hash type expire: release empty hash bucket block When all entries are expired/all slots are empty, release the bucket. Signed-off-by: Jozsef Kadlecsik --- net/netfilter/ipset/ip_set_hash_gen.h | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/net/netfilter/ipset/ip_set_hash_gen.h b/net/netfilter/ipset/ip_set_hash_gen.h index fa4f6374bb73..e5336ab36d67 100644 --- a/net/netfilter/ipset/ip_set_hash_gen.h +++ b/net/netfilter/ipset/ip_set_hash_gen.h @@ -476,7 +476,7 @@ static void mtype_expire(struct ip_set *set, struct htype *h, u8 nets_length, size_t dsize) { struct htable *t; - struct hbucket *n; + struct hbucket *n, *tmp; struct mtype_elem *data; u32 i, j, d; #ifdef IP_SET_HASH_WITH_NETS @@ -511,9 +511,14 @@ mtype_expire(struct ip_set *set, struct htype *h, u8 nets_length, size_t dsize) } } if (d >= AHASH_INIT_SIZE) { - struct hbucket *tmp = kzalloc(sizeof(*tmp) + - (n->size - AHASH_INIT_SIZE) * dsize, - GFP_ATOMIC); + if (d >= n->size) { + rcu_assign_pointer(hbucket(t, i), NULL); + kfree_rcu(n, rcu); + continue; + } + tmp = kzalloc(sizeof(*tmp) + + (n->size - AHASH_INIT_SIZE) * dsize, + GFP_ATOMIC); if (!tmp) /* Still try to delete expired elements */ continue; From 8546d2e5b15e454d2d492a1e9625732980e8eb26 Mon Sep 17 00:00:00 2001 From: Sanchayan Maity Date: Mon, 19 Oct 2015 13:13:52 +0530 Subject: [PATCH 012/335] iio: adc: vf610_adc: Fix division by zero error In case the fsl,adck-max-frequency property is not present in the device tree, a division by zero error results during the probe call on kernel boot (see below). This patch fixes it and also restores device tree compatibility in case kernels are booting with old device trees without this property specified. [ 1.063229] Division by zero in kernel. [ 1.067152] CPU: 0 PID: 1 Comm: swapper Not tainted 4.3.0-rc5-00212-gcc88cef #37 [ 1.074650] Hardware name: Freescale Vybrid VF5xx/VF6xx (Device Tree) [ 1.081135] Backtrace: [ 1.083694] [<800134a4>] (dump_backtrace) from [<8001369c>] (show_stack+0x18/0x1c) [ 1.091340] r7:00000008 r6:8e0ae210 r5:00000000 r4:8e299800 [ 1.097146] [<80013684>] (show_stack) from [<80297b1c>] (dump_stack+0x24/0x28) [ 1.104483] [<80297af8>] (dump_stack) from [<80013608>] (__div0+0x1c/0x20) [ 1.111421] [<800135ec>] (__div0) from [<802968b4>] (Ldiv0+0x8/0x10) [ 1.117865] [<80424350>] (vf610_adc_probe) from [<803153b4>] (platform_drv_probe+0x4c/0xac) [ 1.126311] r10:00000000 r9:8076a5ec r8:00000000 r7:fffffdfb r6:807cc67c r5:8e0ae210 [ 1.134319] r4:807f6c54 [ 1.136915] [<80315368>] (platform_drv_probe) from [<803138bc>] (driver_probe_device+0x20c/0x2f8) [ 1.145882] r7:807cc67c r6:00000000 r5:8e0ae210 r4:807f6c54 [ 1.151657] [<803136b0>] (driver_probe_device) from [<80313a3c>] (__driver_attach+0x94/0x98) [ 1.160190] r9:8076a5ec r8:00000098 r7:00000000 r6:8e0ae244 r5:807cc67c r4:8e0ae210 [ 1.168112] [<803139a8>] (__driver_attach) from [<80311cb8>] (bus_for_each_dev+0x70/0xa4) [ 1.176383] r7:00000000 r6:803139a8 r5:807cc67c r4:00000000 [ 1.182159] [<80311c48>] (bus_for_each_dev) from [<80313318>] (driver_attach+0x24/0x28) [ 1.190260] r6:807bb568 r5:8e2a5b00 r4:807cc67c [ 1.194996] [<803132f4>] (driver_attach) from [<80312f50>] (bus_add_driver+0x1a4/0x21c) [ 1.203113] [<80312dac>] (bus_add_driver) from [<803142a8>] (driver_register+0x80/0x100) [ 1.211275] r7:8e2a7dc0 r6:807a8160 r5:80789e14 r4:807cc67c [ 1.217075] [<80314228>] (driver_register) from [<803152f8>] (__platform_driver_register+0x5c/0x64) [ 1.226216] r5:80789e14 r4:807a8160 [ 1.229877] [<8031529c>] (__platform_driver_register) from [<80789e30>] (vf610_adc_driver_init+0x1c/0x20) [ 1.239556] [<80789e14>] (vf610_adc_driver_init) from [<800095f8>] (do_one_initcall+0x94/0x1dc) [ 1.248365] [<80009564>] (do_one_initcall) from [<8076ae34>] (kernel_init_freeable+0x13c/0x1e0) [ 1.257155] r10:80794830 r9:8076a5ec r8:00000098 r7:807d5780 r6:807d5780 r5:00000006 [ 1.265153] r4:807a0ee8 [ 1.267753] [<8076acf8>] (kernel_init_freeable) from [<80590ef0>] (kernel_init+0x18/0xf0) [ 1.276021] r10:00000000 r9:00000000 r8:00000000 r7:00000000 r6:00000000 r5:80590ed8 [ 1.284015] r4:807d5780 [ 1.286615] [<80590ed8>] (kernel_init) from [<8000f878>] (ret_from_fork+0x14/0x3c) [ 1.294278] r5:80590ed8 r4:00000000 Signed-off-by: Sanchayan Maity Acked-by: Fugang Duan Acked-by: Stefan Agner Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/adc/vf610_adc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/iio/adc/vf610_adc.c b/drivers/iio/adc/vf610_adc.c index 42d7c87ea5be..d937df4f7f2b 100644 --- a/drivers/iio/adc/vf610_adc.c +++ b/drivers/iio/adc/vf610_adc.c @@ -200,6 +200,8 @@ static inline void vf610_adc_calculate_rates(struct vf610_adc *info) adc_feature->clk_div = 8; } + adck_rate = ipg_rate / adc_feature->clk_div; + /* * Determine the long sample time adder value to be used based * on the default minimum sample time provided. @@ -224,7 +226,6 @@ static inline void vf610_adc_calculate_rates(struct vf610_adc *info) * BCT (Base Conversion Time): fixed to 25 ADCK cycles for 12 bit mode * LSTAdder(Long Sample Time): 3, 5, 7, 9, 13, 17, 21, 25 ADCK cycles */ - adck_rate = ipg_rate / info->adc_feature.clk_div; for (i = 0; i < ARRAY_SIZE(vf610_hw_avgs); i++) info->sample_freq_avail[i] = adck_rate / (6 + vf610_hw_avgs[i] * From c6eafbf9903c4283ba146098ad54240a24ffbeb1 Mon Sep 17 00:00:00 2001 From: Sascha Silbe Date: Wed, 4 Nov 2015 13:32:08 +0100 Subject: [PATCH 013/335] s390/head: fix error message on unsupported hardware startup calls the C function _sclp_print_early() if the machine we're running on is not supported by the kernel. sclp.c is getting built with -m64, so _sclp_print_early() expects the zSeries ELF ABI to be used. We previously called _sclp_print_early() using the S/390 ELF ABI, with a stack frame size of 96 bytes and while being in 31-bit address mode. This caused _sclp_wait_int() (called indirectly from _sclp_print_early()) to jump to an undefined address. While _sclp_wait_int() contained some code to deal with being called in 31-bit addressing mode, it didn't quite work. While fixing this is possible, the code would still only work by chance and could break any time. Ensure compliance with the zSeries ELF ABI by switching to 64-bit addressing mode early and using a minimum stack frame size of 160 bytes. Signed-off-by: Sascha Silbe Acked-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/head.S | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index 1255c6c5353e..1b0f624cebd7 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -26,6 +26,7 @@ #include #include #include +#include #define ARCH_OFFSET 4 @@ -364,7 +365,7 @@ ENTRY(startup_kdump) bras %r13,0f .fill 16,4,0x0 0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs - sam31 # switch to 31 bit addressing mode + sam64 # switch to 64 bit addressing mode basr %r13,0 # get base .LPG0: xc 0x200(256),0x200 # partially clear lowcore @@ -395,7 +396,7 @@ ENTRY(startup_kdump) jnz 1b j 4f 2: l %r15,.Lstack-.LPG0(%r13) - ahi %r15,-96 + ahi %r15,-STACK_FRAME_OVERHEAD la %r2,.Lals_string-.LPG0(%r13) l %r3,.Lsclp_print-.LPG0(%r13) basr %r14,%r3 @@ -429,8 +430,7 @@ ENTRY(startup_kdump) .long 1, 0xc0000000 #endif 4: - /* Continue with 64bit startup code in head64.S */ - sam64 # switch to 64 bit mode + /* Continue with startup code in head64.S */ jg startup_continue .align 8 From 230ccb370f8f95b2600a1fce90ceb8ee70a15dbc Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Thu, 5 Nov 2015 13:50:04 +0100 Subject: [PATCH 014/335] s390/diag: add a s390 prefix to the diagnose trace point Documentation/trace/tracepoints.txt states that the naming scheme for tracepoints is "subsys_event" to avoid collisions. Rename the 'diagnose' tracepoint to 's390_diagnose'. Reported-by: Peter Oberparleiter Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/trace/diag.h | 6 +++--- arch/s390/kernel/diag.c | 4 ++-- arch/s390/kernel/trace.c | 6 +++--- 3 files changed, 8 insertions(+), 8 deletions(-) diff --git a/arch/s390/include/asm/trace/diag.h b/arch/s390/include/asm/trace/diag.h index 776f307960cc..cc6cfe7889da 100644 --- a/arch/s390/include/asm/trace/diag.h +++ b/arch/s390/include/asm/trace/diag.h @@ -19,7 +19,7 @@ #define TRACE_INCLUDE_PATH asm/trace #define TRACE_INCLUDE_FILE diag -TRACE_EVENT(diagnose, +TRACE_EVENT(s390_diagnose, TP_PROTO(unsigned short nr), TP_ARGS(nr), TP_STRUCT__entry( @@ -32,9 +32,9 @@ TRACE_EVENT(diagnose, ); #ifdef CONFIG_TRACEPOINTS -void trace_diagnose_norecursion(int diag_nr); +void trace_s390_diagnose_norecursion(int diag_nr); #else -static inline void trace_diagnose_norecursion(int diag_nr) { } +static inline void trace_s390_diagnose_norecursion(int diag_nr) { } #endif #endif /* _TRACE_S390_DIAG_H */ diff --git a/arch/s390/kernel/diag.c b/arch/s390/kernel/diag.c index f98766ede4e1..48b37b8357e6 100644 --- a/arch/s390/kernel/diag.c +++ b/arch/s390/kernel/diag.c @@ -121,14 +121,14 @@ device_initcall(show_diag_stat_init); void diag_stat_inc(enum diag_stat_enum nr) { this_cpu_inc(diag_stat.counter[nr]); - trace_diagnose(diag_map[nr].code); + trace_s390_diagnose(diag_map[nr].code); } EXPORT_SYMBOL(diag_stat_inc); void diag_stat_inc_norecursion(enum diag_stat_enum nr) { this_cpu_inc(diag_stat.counter[nr]); - trace_diagnose_norecursion(diag_map[nr].code); + trace_s390_diagnose_norecursion(diag_map[nr].code); } EXPORT_SYMBOL(diag_stat_inc_norecursion); diff --git a/arch/s390/kernel/trace.c b/arch/s390/kernel/trace.c index 73239bb576c4..21a5df99552b 100644 --- a/arch/s390/kernel/trace.c +++ b/arch/s390/kernel/trace.c @@ -9,11 +9,11 @@ #define CREATE_TRACE_POINTS #include -EXPORT_TRACEPOINT_SYMBOL(diagnose); +EXPORT_TRACEPOINT_SYMBOL(s390_diagnose); static DEFINE_PER_CPU(unsigned int, diagnose_trace_depth); -void trace_diagnose_norecursion(int diag_nr) +void trace_s390_diagnose_norecursion(int diag_nr) { unsigned long flags; unsigned int *depth; @@ -22,7 +22,7 @@ void trace_diagnose_norecursion(int diag_nr) depth = this_cpu_ptr(&diagnose_trace_depth); if (*depth == 0) { (*depth)++; - trace_diagnose(diag_nr); + trace_s390_diagnose(diag_nr); (*depth)--; } local_irq_restore(flags); From cd5dead9d32423d63c802328c0563d7ff29647cd Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Thu, 16 Jul 2015 22:40:27 +0200 Subject: [PATCH 015/335] s390/cio: simplify css_generate_pgid Simplify the css_generate_pgid code by using stap() independent of CONFIG_SMP. For !CONFIG_SMP builds stap() will deliver the address of the cpu we IPL'ed from (which can be != 0). Note: the ifdef was likely added to be compatible with _very_ old machines which we don't support anyway. Reviewed-by: Peter Oberparleiter Signed-off-by: Sebastian Ott Signed-off-by: Martin Schwidefsky --- drivers/s390/cio/css.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c index 2ee3053bdc12..489e703dc82d 100644 --- a/drivers/s390/cio/css.c +++ b/drivers/s390/cio/css.c @@ -702,17 +702,12 @@ css_generate_pgid(struct channel_subsystem *css, u32 tod_high) css->global_pgid.pgid_high.ext_cssid.version = 0x80; css->global_pgid.pgid_high.ext_cssid.cssid = css->cssid; } else { -#ifdef CONFIG_SMP css->global_pgid.pgid_high.cpu_addr = stap(); -#else - css->global_pgid.pgid_high.cpu_addr = 0; -#endif } get_cpu_id(&cpu_id); css->global_pgid.cpu_id = cpu_id.ident; css->global_pgid.cpu_model = cpu_id.machine; css->global_pgid.tod_high = tod_high; - } static void From 86b68c3873a82144033a82707781829e3f7b3859 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 6 Nov 2015 13:50:25 +0100 Subject: [PATCH 016/335] s390/syscalls: remove system call number calculation Explicitly write the system call number for each define instead of calculating it. This makes it easier to parse the file when generating system call tables for various tools and libraries. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/uapi/asm/unistd.h | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/arch/s390/include/uapi/asm/unistd.h b/arch/s390/include/uapi/asm/unistd.h index a848adba1504..c55de84ab1ad 100644 --- a/arch/s390/include/uapi/asm/unistd.h +++ b/arch/s390/include/uapi/asm/unistd.h @@ -192,14 +192,14 @@ #define __NR_set_tid_address 252 #define __NR_fadvise64 253 #define __NR_timer_create 254 -#define __NR_timer_settime (__NR_timer_create+1) -#define __NR_timer_gettime (__NR_timer_create+2) -#define __NR_timer_getoverrun (__NR_timer_create+3) -#define __NR_timer_delete (__NR_timer_create+4) -#define __NR_clock_settime (__NR_timer_create+5) -#define __NR_clock_gettime (__NR_timer_create+6) -#define __NR_clock_getres (__NR_timer_create+7) -#define __NR_clock_nanosleep (__NR_timer_create+8) +#define __NR_timer_settime 255 +#define __NR_timer_gettime 256 +#define __NR_timer_getoverrun 257 +#define __NR_timer_delete 258 +#define __NR_clock_settime 259 +#define __NR_clock_gettime 260 +#define __NR_clock_getres 261 +#define __NR_clock_nanosleep 262 /* Number 263 is reserved for vserver */ #define __NR_statfs64 265 #define __NR_fstatfs64 266 From 4d5a6b72959601d6c12e7e1ef3aa4132f0a62523 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 26 Oct 2015 11:15:28 +0100 Subject: [PATCH 017/335] s390/pci_dma: unify label of invalid translation table entries Newly allocated translation table entries are flagged as invalid and protected. If an existing translation table entry is invalidated, the protection flag is left unchanged. If a page (with invalid and protection flag set) is accessed it's undefined which type of exception we'll receive. Make sure to always set the invalid flag only. Signed-off-by: Sebastian Ott Reviewed-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/pci/pci_dma.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c index 37d10f74425a..e4a3a31fd59a 100644 --- a/arch/s390/pci/pci_dma.c +++ b/arch/s390/pci/pci_dma.c @@ -33,7 +33,7 @@ unsigned long *dma_alloc_cpu_table(void) return NULL; for (entry = table; entry < table + ZPCI_TABLE_ENTRIES; entry++) - *entry = ZPCI_TABLE_INVALID | ZPCI_TABLE_PROTECTED; + *entry = ZPCI_TABLE_INVALID; return table; } @@ -51,7 +51,7 @@ static unsigned long *dma_alloc_page_table(void) return NULL; for (entry = table; entry < table + ZPCI_PT_ENTRIES; entry++) - *entry = ZPCI_PTE_INVALID | ZPCI_TABLE_PROTECTED; + *entry = ZPCI_PTE_INVALID; return table; } @@ -127,7 +127,6 @@ void dma_update_cpu_trans(unsigned long *dma_table, void *page_addr, if (flags & ZPCI_PTE_INVALID) { invalidate_pt_entry(entry); - return; } else { set_pt_pfaa(entry, page_addr); validate_pt_entry(entry); @@ -311,7 +310,7 @@ static void s390_dma_unmap_pages(struct device *dev, dma_addr_t dma_addr, npages = iommu_num_pages(dma_addr, size, PAGE_SIZE); dma_addr = dma_addr & PAGE_MASK; if (dma_update_trans(zdev, 0, dma_addr, npages * PAGE_SIZE, - ZPCI_TABLE_PROTECTED | ZPCI_PTE_INVALID)) { + ZPCI_PTE_INVALID)) { zpci_err("unmap error:\n"); zpci_err_hex(&dma_addr, sizeof(dma_addr)); } From 66728eeea6d80060e4b9df55c7845c838ff2799f Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 26 Oct 2015 11:19:13 +0100 Subject: [PATCH 018/335] s390/pci_dma: handle dma table failures We use lazy allocation for translation table entries but don't handle allocation (and other) failures during translation table updates. Handle these failures and undo translation table updates when it's meaningful. Signed-off-by: Sebastian Ott Reviewed-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/pci_dma.h | 4 +++- arch/s390/pci/pci_dma.c | 38 +++++++++++++++++++++------------ drivers/iommu/s390-iommu.c | 23 ++++++++++++++++++-- 3 files changed, 48 insertions(+), 17 deletions(-) diff --git a/arch/s390/include/asm/pci_dma.h b/arch/s390/include/asm/pci_dma.h index 7a7abf1a5537..1aac41e83ea1 100644 --- a/arch/s390/include/asm/pci_dma.h +++ b/arch/s390/include/asm/pci_dma.h @@ -195,5 +195,7 @@ void zpci_dma_exit_device(struct zpci_dev *); void dma_free_seg_table(unsigned long); unsigned long *dma_alloc_cpu_table(void); void dma_cleanup_tables(unsigned long *); -void dma_update_cpu_trans(unsigned long *, void *, dma_addr_t, int); +unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr); +void dma_update_cpu_trans(unsigned long *entry, void *page_addr, int flags); + #endif diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c index e4a3a31fd59a..f137949c9abf 100644 --- a/arch/s390/pci/pci_dma.c +++ b/arch/s390/pci/pci_dma.c @@ -95,7 +95,7 @@ static unsigned long *dma_get_page_table_origin(unsigned long *entry) return pto; } -static unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr) +unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr) { unsigned long *sto, *pto; unsigned int rtx, sx, px; @@ -114,17 +114,8 @@ static unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr return &pto[px]; } -void dma_update_cpu_trans(unsigned long *dma_table, void *page_addr, - dma_addr_t dma_addr, int flags) +void dma_update_cpu_trans(unsigned long *entry, void *page_addr, int flags) { - unsigned long *entry; - - entry = dma_walk_cpu_trans(dma_table, dma_addr); - if (!entry) { - WARN_ON_ONCE(1); - return; - } - if (flags & ZPCI_PTE_INVALID) { invalidate_pt_entry(entry); } else { @@ -145,18 +136,25 @@ static int dma_update_trans(struct zpci_dev *zdev, unsigned long pa, u8 *page_addr = (u8 *) (pa & PAGE_MASK); dma_addr_t start_dma_addr = dma_addr; unsigned long irq_flags; + unsigned long *entry; int i, rc = 0; if (!nr_pages) return -EINVAL; spin_lock_irqsave(&zdev->dma_table_lock, irq_flags); - if (!zdev->dma_table) + if (!zdev->dma_table) { + rc = -EINVAL; goto no_refresh; + } for (i = 0; i < nr_pages; i++) { - dma_update_cpu_trans(zdev->dma_table, page_addr, dma_addr, - flags); + entry = dma_walk_cpu_trans(zdev->dma_table, dma_addr); + if (!entry) { + rc = -ENOMEM; + goto undo_cpu_trans; + } + dma_update_cpu_trans(entry, page_addr, flags); page_addr += PAGE_SIZE; dma_addr += PAGE_SIZE; } @@ -175,6 +173,18 @@ static int dma_update_trans(struct zpci_dev *zdev, unsigned long pa, rc = zpci_refresh_trans((u64) zdev->fh << 32, start_dma_addr, nr_pages * PAGE_SIZE); +undo_cpu_trans: + if (rc && ((flags & ZPCI_PTE_VALID_MASK) == ZPCI_PTE_VALID)) { + flags = ZPCI_PTE_INVALID; + while (i-- > 0) { + page_addr -= PAGE_SIZE; + dma_addr -= PAGE_SIZE; + entry = dma_walk_cpu_trans(zdev->dma_table, dma_addr); + if (!entry) + break; + dma_update_cpu_trans(entry, page_addr, flags); + } + } no_refresh: spin_unlock_irqrestore(&zdev->dma_table_lock, irq_flags); diff --git a/drivers/iommu/s390-iommu.c b/drivers/iommu/s390-iommu.c index cbe198cb3699..471ee36b9c6e 100644 --- a/drivers/iommu/s390-iommu.c +++ b/drivers/iommu/s390-iommu.c @@ -216,6 +216,7 @@ static int s390_iommu_update_trans(struct s390_domain *s390_domain, u8 *page_addr = (u8 *) (pa & PAGE_MASK); dma_addr_t start_dma_addr = dma_addr; unsigned long irq_flags, nr_pages, i; + unsigned long *entry; int rc = 0; if (dma_addr < s390_domain->domain.geometry.aperture_start || @@ -228,8 +229,12 @@ static int s390_iommu_update_trans(struct s390_domain *s390_domain, spin_lock_irqsave(&s390_domain->dma_table_lock, irq_flags); for (i = 0; i < nr_pages; i++) { - dma_update_cpu_trans(s390_domain->dma_table, page_addr, - dma_addr, flags); + entry = dma_walk_cpu_trans(s390_domain->dma_table, dma_addr); + if (!entry) { + rc = -ENOMEM; + goto undo_cpu_trans; + } + dma_update_cpu_trans(entry, page_addr, flags); page_addr += PAGE_SIZE; dma_addr += PAGE_SIZE; } @@ -242,6 +247,20 @@ static int s390_iommu_update_trans(struct s390_domain *s390_domain, break; } spin_unlock(&s390_domain->list_lock); + +undo_cpu_trans: + if (rc && ((flags & ZPCI_PTE_VALID_MASK) == ZPCI_PTE_VALID)) { + flags = ZPCI_PTE_INVALID; + while (i-- > 0) { + page_addr -= PAGE_SIZE; + dma_addr -= PAGE_SIZE; + entry = dma_walk_cpu_trans(s390_domain->dma_table, + dma_addr); + if (!entry) + break; + dma_update_cpu_trans(entry, page_addr, flags); + } + } spin_unlock_irqrestore(&s390_domain->dma_table_lock, irq_flags); return rc; From 52d43d8184b1840c7cf6136724223585f51a1074 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 26 Oct 2015 11:20:44 +0100 Subject: [PATCH 019/335] s390/pci_dma: improve debugging of errors during dma map Improve debugging to find out what went wrong during a failed dma map/unmap operation. Signed-off-by: Sebastian Ott Reviewed-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/pci/pci_dma.c | 41 ++++++++++++++++++++++++++++++----------- 1 file changed, 30 insertions(+), 11 deletions(-) diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c index f137949c9abf..d348f2c09a1e 100644 --- a/arch/s390/pci/pci_dma.c +++ b/arch/s390/pci/pci_dma.c @@ -269,6 +269,16 @@ out: spin_unlock_irqrestore(&zdev->iommu_bitmap_lock, flags); } +static inline void zpci_err_dma(unsigned long rc, unsigned long addr) +{ + struct { + unsigned long rc; + unsigned long addr; + } __packed data = {rc, addr}; + + zpci_err_hex(&data, sizeof(data)); +} + static dma_addr_t s390_dma_map_pages(struct device *dev, struct page *page, unsigned long offset, size_t size, enum dma_data_direction direction, @@ -279,33 +289,40 @@ static dma_addr_t s390_dma_map_pages(struct device *dev, struct page *page, unsigned long pa = page_to_phys(page) + offset; int flags = ZPCI_PTE_VALID; dma_addr_t dma_addr; + int ret; /* This rounds up number of pages based on size and offset */ nr_pages = iommu_num_pages(pa, size, PAGE_SIZE); iommu_page_index = dma_alloc_iommu(zdev, nr_pages); - if (iommu_page_index == -1) + if (iommu_page_index == -1) { + ret = -ENOSPC; goto out_err; + } /* Use rounded up size */ size = nr_pages * PAGE_SIZE; dma_addr = zdev->start_dma + iommu_page_index * PAGE_SIZE; - if (dma_addr + size > zdev->end_dma) + if (dma_addr + size > zdev->end_dma) { + ret = -ERANGE; goto out_free; + } if (direction == DMA_NONE || direction == DMA_TO_DEVICE) flags |= ZPCI_TABLE_PROTECTED; - if (!dma_update_trans(zdev, pa, dma_addr, size, flags)) { - atomic64_add(nr_pages, &zdev->mapped_pages); - return dma_addr + (offset & ~PAGE_MASK); - } + ret = dma_update_trans(zdev, pa, dma_addr, size, flags); + if (ret) + goto out_free; + + atomic64_add(nr_pages, &zdev->mapped_pages); + return dma_addr + (offset & ~PAGE_MASK); out_free: dma_free_iommu(zdev, iommu_page_index, nr_pages); out_err: zpci_err("map error:\n"); - zpci_err_hex(&pa, sizeof(pa)); + zpci_err_dma(ret, pa); return DMA_ERROR_CODE; } @@ -315,14 +332,16 @@ static void s390_dma_unmap_pages(struct device *dev, dma_addr_t dma_addr, { struct zpci_dev *zdev = to_zpci(to_pci_dev(dev)); unsigned long iommu_page_index; - int npages; + int npages, ret; npages = iommu_num_pages(dma_addr, size, PAGE_SIZE); dma_addr = dma_addr & PAGE_MASK; - if (dma_update_trans(zdev, 0, dma_addr, npages * PAGE_SIZE, - ZPCI_PTE_INVALID)) { + ret = dma_update_trans(zdev, 0, dma_addr, npages * PAGE_SIZE, + ZPCI_PTE_INVALID); + if (ret) { zpci_err("unmap error:\n"); - zpci_err_hex(&dma_addr, sizeof(dma_addr)); + zpci_err_dma(ret, dma_addr); + return; } atomic64_add(npages, &zdev->unmapped_pages); From 26d49fe7195385f2f1e406feddb01c16b53e77b6 Mon Sep 17 00:00:00 2001 From: Robert Jarzmik Date: Thu, 5 Nov 2015 20:46:53 +0100 Subject: [PATCH 020/335] mmc: pxamci: fix read-only gpio detection polarity The commit converting pxamci to slot-gpio API inverted the logic of the read-only gpio. Fix it by inverting the logic again. Fixes: fd546ee6a7dc ("mmc: pxamci: fix card detect with slot-gpio API") Signed-off-by: Robert Jarzmik Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson --- drivers/mmc/host/pxamci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mmc/host/pxamci.c b/drivers/mmc/host/pxamci.c index 8cadd74e8407..ce08896b9d69 100644 --- a/drivers/mmc/host/pxamci.c +++ b/drivers/mmc/host/pxamci.c @@ -805,7 +805,7 @@ static int pxamci_probe(struct platform_device *pdev) goto out; } else { mmc->caps |= host->pdata->gpio_card_ro_invert ? - MMC_CAP2_RO_ACTIVE_HIGH : 0; + 0 : MMC_CAP2_RO_ACTIVE_HIGH; } if (gpio_is_valid(gpio_cd)) From 1815e61b1a7efe81017a883e817292daf7d2f922 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 28 Oct 2015 14:25:40 +0200 Subject: [PATCH 021/335] mmc: mmc: Improve reliability of mmc_select_hs200() Currently mmc_select_hs200() uses __mmc_switch() which checks the success of the switch to HS200 mode using CMD13 (SEND_STATUS). The problem is that it does that using the timing settings of legacy mode. That is prone to error, not least because the timing parameters for legacy mode are tighter than the timing parameters for HS200 mode. In the case when CMD13 polling is used (i.e. not MMC_CAP_WAIT_WHILE_BUSY) with the switch command, it must be assumed that using different modes on the card and host must work. However in the case when CMD13 polling is not used (i.e. MMC_CAP_WAIT_WHILE_BUSY) mmc_select_hs200() can be made more reliable by setting the host to the correct timing before sending CMD13. This patch does that. A complication is that the caller, mmc_select_timing(), will ignore a switch error (indicated by -EBADMSG), assume the old mode is valid and continue, so the old timing must be restored in that case. Signed-off-by: Adrian Hunter Cc: # 4.2+ Tested-by: Alim Akhtar Signed-off-by: Ulf Hansson --- drivers/mmc/core/mmc.c | 25 ++++++++++++++++++++++--- 1 file changed, 22 insertions(+), 3 deletions(-) diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c index c793fda27321..2cef40ce9851 100644 --- a/drivers/mmc/core/mmc.c +++ b/drivers/mmc/core/mmc.c @@ -1219,6 +1219,8 @@ static void mmc_select_driver_type(struct mmc_card *card) static int mmc_select_hs200(struct mmc_card *card) { struct mmc_host *host = card->host; + bool send_status = true; + unsigned int old_timing; int err = -EINVAL; u8 val; @@ -1234,6 +1236,9 @@ static int mmc_select_hs200(struct mmc_card *card) mmc_select_driver_type(card); + if (host->caps & MMC_CAP_WAIT_WHILE_BUSY) + send_status = false; + /* * Set the bus width(4 or 8) with host's support and * switch to HS200 mode if bus width is set successfully. @@ -1245,11 +1250,25 @@ static int mmc_select_hs200(struct mmc_card *card) err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_HS_TIMING, val, card->ext_csd.generic_cmd6_time, - true, true, true); - if (!err) - mmc_set_timing(host, MMC_TIMING_MMC_HS200); + true, send_status, true); + if (err) + goto err; + old_timing = host->ios.timing; + mmc_set_timing(host, MMC_TIMING_MMC_HS200); + if (!send_status) { + err = mmc_switch_status(card); + /* + * mmc_select_timing() assumes timing has not changed if + * it is a switch error. + */ + if (err == -EBADMSG) + mmc_set_timing(host, old_timing); + } } err: + if (err) + pr_err("%s: %s failed, error %d\n", mmc_hostname(card->host), + __func__, err); return err; } From 51b12f7764fa8bb464cbd0f7bbd3a408d21ade16 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 28 Oct 2015 14:25:41 +0200 Subject: [PATCH 022/335] mmc: mmc: Fix HS setting in mmc_select_hs400() mmc_select_hs400() begins with the card and host in HS200 mode. Therefore, any commands sent to the card should use HS200 timing. It is incorrect to set the host to High Speed (HS) timing before sending the switch command. Doing so is unreliable because the timing parameters for HS mode are tighter than the timing parameters for HS200 mode. Thus the HS timings should be set only after the card has switched mode. However, it is not unreasonable first to reduce the frequency to the HS mode frequency, which should make the switch command and subsequent CMD13 commands more reliable. This patch does that. Signed-off-by: Adrian Hunter Cc: # 4.2+ Tested-by: Alim Akhtar Signed-off-by: Ulf Hansson --- drivers/mmc/core/mmc.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c index 2cef40ce9851..14fb767ee8dd 100644 --- a/drivers/mmc/core/mmc.c +++ b/drivers/mmc/core/mmc.c @@ -1043,6 +1043,7 @@ static int mmc_select_hs_ddr(struct mmc_card *card) static int mmc_select_hs400(struct mmc_card *card) { struct mmc_host *host = card->host; + unsigned int max_dtr; int err = 0; u8 val; @@ -1053,13 +1054,11 @@ static int mmc_select_hs400(struct mmc_card *card) host->ios.bus_width == MMC_BUS_WIDTH_8)) return 0; - /* - * Before switching to dual data rate operation for HS400, - * it is required to convert from HS200 mode to HS mode. - */ - mmc_set_timing(card->host, MMC_TIMING_MMC_HS); - mmc_set_bus_speed(card); + /* Reduce frequency to HS frequency */ + max_dtr = card->ext_csd.hs_max_dtr; + mmc_set_clock(host, max_dtr); + /* Switch card to HS mode */ val = EXT_CSD_TIMING_HS | card->drive_strength << EXT_CSD_DRV_STR_SHIFT; err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, @@ -1072,6 +1071,9 @@ static int mmc_select_hs400(struct mmc_card *card) return err; } + /* Set host controller to HS timing */ + mmc_set_timing(card->host, MMC_TIMING_MMC_HS); + err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_BUS_WIDTH, EXT_CSD_DDR_BUS_WIDTH_8, From 974007aaf240aa195b31c34cfdb013524a2dcfca Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 28 Oct 2015 14:25:42 +0200 Subject: [PATCH 023/335] mmc: mmc: Move mmc_switch_status() Move the mmc_switch_status() function in preparation for calling it in mmc_select_hs400(). Signed-off-by: Adrian Hunter Cc: # 4.2+ Tested-by: Alim Akhtar Signed-off-by: Ulf Hansson --- drivers/mmc/core/mmc.c | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c index 14fb767ee8dd..5884c79346da 100644 --- a/drivers/mmc/core/mmc.c +++ b/drivers/mmc/core/mmc.c @@ -1040,6 +1040,19 @@ static int mmc_select_hs_ddr(struct mmc_card *card) return err; } +/* Caller must hold re-tuning */ +static int mmc_switch_status(struct mmc_card *card) +{ + u32 status; + int err; + + err = mmc_send_status(card, &status); + if (err) + return err; + + return mmc_switch_status_error(card->host, status); +} + static int mmc_select_hs400(struct mmc_card *card) { struct mmc_host *host = card->host; @@ -1107,19 +1120,6 @@ int mmc_hs200_to_hs400(struct mmc_card *card) return mmc_select_hs400(card); } -/* Caller must hold re-tuning */ -static int mmc_switch_status(struct mmc_card *card) -{ - u32 status; - int err; - - err = mmc_send_status(card, &status); - if (err) - return err; - - return mmc_switch_status_error(card->host, status); -} - int mmc_hs400_to_hs200(struct mmc_card *card) { struct mmc_host *host = card->host; From d23029332c3d51fb5ac117ba5cde4dc0a3ec3fa6 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 28 Oct 2015 14:25:43 +0200 Subject: [PATCH 024/335] mmc: mmc: Improve reliability of mmc_select_hs400() mmc_select_hs400() calls __mmc_switch() which checks the switch is successful using CMD13 (SEND_STATUS). The problem is that it does that using the timing settings of the previous mode. That is prone to error, especially when switching from HS to HS400 because the timing parameters for HS mode are tighter than the timing parameters for HS400 mode. In the case when CMD13 polling is used (i.e. not MMC_CAP_WAIT_WHILE_BUSY) with the switch command, it must be assumed that using different modes on the card and host must work. However in the case when CMD13 polling is not used (i.e. MMC_CAP_WAIT_WHILE_BUSY) mmc_select_hs400() can be made more reliable by setting the host to the correct timing before sending CMD13. This patch does that. Signed-off-by: Adrian Hunter Cc: # 4.2+ Tested-by: Alim Akhtar Signed-off-by: Ulf Hansson --- drivers/mmc/core/mmc.c | 28 ++++++++++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c index 5884c79346da..3a9a79ec4343 100644 --- a/drivers/mmc/core/mmc.c +++ b/drivers/mmc/core/mmc.c @@ -1056,6 +1056,7 @@ static int mmc_switch_status(struct mmc_card *card) static int mmc_select_hs400(struct mmc_card *card) { struct mmc_host *host = card->host; + bool send_status = true; unsigned int max_dtr; int err = 0; u8 val; @@ -1067,6 +1068,9 @@ static int mmc_select_hs400(struct mmc_card *card) host->ios.bus_width == MMC_BUS_WIDTH_8)) return 0; + if (host->caps & MMC_CAP_WAIT_WHILE_BUSY) + send_status = false; + /* Reduce frequency to HS frequency */ max_dtr = card->ext_csd.hs_max_dtr; mmc_set_clock(host, max_dtr); @@ -1077,7 +1081,7 @@ static int mmc_select_hs400(struct mmc_card *card) err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_HS_TIMING, val, card->ext_csd.generic_cmd6_time, - true, true, true); + true, send_status, true); if (err) { pr_err("%s: switch to high-speed from hs200 failed, err:%d\n", mmc_hostname(host), err); @@ -1087,6 +1091,13 @@ static int mmc_select_hs400(struct mmc_card *card) /* Set host controller to HS timing */ mmc_set_timing(card->host, MMC_TIMING_MMC_HS); + if (!send_status) { + err = mmc_switch_status(card); + if (err) + goto out_err; + } + + /* Switch card to DDR */ err = mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_BUS_WIDTH, EXT_CSD_DDR_BUS_WIDTH_8, @@ -1097,22 +1108,35 @@ static int mmc_select_hs400(struct mmc_card *card) return err; } + /* Switch card to HS400 */ val = EXT_CSD_TIMING_HS400 | card->drive_strength << EXT_CSD_DRV_STR_SHIFT; err = __mmc_switch(card, EXT_CSD_CMD_SET_NORMAL, EXT_CSD_HS_TIMING, val, card->ext_csd.generic_cmd6_time, - true, true, true); + true, send_status, true); if (err) { pr_err("%s: switch to hs400 failed, err:%d\n", mmc_hostname(host), err); return err; } + /* Set host controller to HS400 timing and frequency */ mmc_set_timing(host, MMC_TIMING_MMC_HS400); mmc_set_bus_speed(card); + if (!send_status) { + err = mmc_switch_status(card); + if (err) + goto out_err; + } + return 0; + +out_err: + pr_err("%s: %s failed, error %d\n", mmc_hostname(card->host), + __func__, err); + return err; } int mmc_hs200_to_hs400(struct mmc_card *card) From ae3fbdd1becdfe4488e43bc023ea672bf2c90660 Mon Sep 17 00:00:00 2001 From: Ulf Hansson Date: Fri, 6 Nov 2015 12:21:33 +0100 Subject: [PATCH 025/335] MAINTAINERS: mmc: Remove Seungwon Jeon from dw_mmc Unfortunate, Seungwon's email has been bouncing for some time and it seems like he has moved on to other duties. So, I would like to thank him for his efforts, but it's now time to remove him as the co-maintainer for the dw_mmc driver. Signed-off-by: Ulf Hansson Acked-by: Jaehoon Chung --- MAINTAINERS | 1 - 1 file changed, 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 7af7f4a01f0b..bb53cb612f89 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9290,7 +9290,6 @@ F: drivers/i2c/busses/i2c-designware-* F: include/linux/platform_data/i2c-designware.h SYNOPSYS DESIGNWARE MMC/SD/SDIO DRIVER -M: Seungwon Jeon M: Jaehoon Chung L: linux-mmc@vger.kernel.org S: Maintained From 62d494ca2773563e333e670cd18378705dad32d4 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Fri, 6 Nov 2015 12:22:08 +0100 Subject: [PATCH 026/335] mmc: mediatek: Preinitialize delay_phase in get_best_delay() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit drivers/mmc/host/mtk-sd.c: In function ‘get_best_delay’: drivers/mmc/host/mtk-sd.c:1284: warning: ‘delay_phase.start’ is used uninitialized in this function drivers/mmc/host/mtk-sd.c:1284: warning: ‘delay_phase.maxlen’ is used uninitialized in this function If delay is zero, these fields are indeed not initialized. Let the compiler preinitialize the whole struct to fix this. Signed-off-by: Geert Uytterhoeven Signed-off-by: Ulf Hansson --- drivers/mmc/host/mtk-sd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c index 39568cc29a2a..33dfd7e72516 100644 --- a/drivers/mmc/host/mtk-sd.c +++ b/drivers/mmc/host/mtk-sd.c @@ -1276,7 +1276,7 @@ static struct msdc_delay_phase get_best_delay(struct msdc_host *host, u32 delay) int start = 0, len = 0; int start_final = 0, len_final = 0; u8 final_phase = 0xff; - struct msdc_delay_phase delay_phase; + struct msdc_delay_phase delay_phase = { 0, }; if (delay == 0) { dev_err(host->dev, "phase error: [map:%x]\n", delay); From 3d35877294005911da54c36e2ab4d7b72d5e9331 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Mon, 9 Nov 2015 10:36:01 +0100 Subject: [PATCH 027/335] mmc: MMC_GOLDFISH should depend on HAS_DMA If NO_DMA=y: ERROR: dma_unmap_sg [drivers/mmc/host/android-goldfish.ko] undefined! ERROR: dma_alloc_coherent [drivers/mmc/host/android-goldfish.ko] undefined! ERROR: dma_map_sg [drivers/mmc/host/android-goldfish.ko] undefined! ERROR: dma_free_coherent [drivers/mmc/host/android-goldfish.ko] undefined! Add a dependency on HAS_DMA to fix this. Signed-off-by: Geert Uytterhoeven Signed-off-by: Ulf Hansson --- drivers/mmc/host/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig index af71de5fda3b..1dee533634c9 100644 --- a/drivers/mmc/host/Kconfig +++ b/drivers/mmc/host/Kconfig @@ -473,6 +473,7 @@ config MMC_DAVINCI config MMC_GOLDFISH tristate "goldfish qemu Multimedia Card Interface support" + depends on HAS_DMA depends on GOLDFISH || COMPILE_TEST help This selects the Goldfish Multimedia card Interface emulation From c255cb2ed3c7960b2c68f45de1dc0ac2197c8f78 Mon Sep 17 00:00:00 2001 From: Anthony Lineham Date: Thu, 22 Oct 2015 11:17:03 +1300 Subject: [PATCH 028/335] netfilter: Fix removal of GRE expectation entries created by PPTP The uninitialized tuple structure caused incorrect hash calculation and the lookup failed. Link: https://bugzilla.kernel.org/show_bug.cgi?id=106441 Signed-off-by: Anthony Lineham Signed-off-by: Pablo Neira Ayuso --- net/ipv4/netfilter/nf_nat_pptp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/netfilter/nf_nat_pptp.c b/net/ipv4/netfilter/nf_nat_pptp.c index 657d2307f031..b3ca21b2ba9b 100644 --- a/net/ipv4/netfilter/nf_nat_pptp.c +++ b/net/ipv4/netfilter/nf_nat_pptp.c @@ -45,7 +45,7 @@ static void pptp_nat_expected(struct nf_conn *ct, struct net *net = nf_ct_net(ct); const struct nf_conn *master = ct->master; struct nf_conntrack_expect *other_exp; - struct nf_conntrack_tuple t; + struct nf_conntrack_tuple t = {}; const struct nf_ct_pptp_master *ct_pptp_info; const struct nf_nat_pptp *nat_pptp_info; struct nf_nat_range range; From d3df0465db00cf4ed9f90d0bfc3b827d32b9c796 Mon Sep 17 00:00:00 2001 From: Luca Porzio Date: Fri, 6 Nov 2015 15:12:26 +0000 Subject: [PATCH 029/335] mmc: remove bondage between REQ_META and reliable write Anytime a write operation is performed with Reliable Write flag enabled, the eMMC device is enforced to bypass the cache and do a write to the underling NVM device by Jedec specification; this causes a performance penalty since write operations can't be optimized by the device cache. In our tests, we replayed a typical mobile daily trace pattern and found ~9% overall time reduction in trace replay by using this patch. Also the write ops within 4KB~64KB chunk size range get a 40~60% performance improvement by using the patch (as this range of write chunks are the ones affected by REQ_META). This patch has been discussed in the Mobile & Embedded Linux Storage Forum and it's the results of feedbacks from many people. We also checked with fsdevl and f2fs mailing list developers that this change in the usage of REQ_META is not affecting FS behavior and we got positive feedbacks. Reporting here the feedbacks: http://comments.gmane.org/gmane.linux.file-systems/97219 http://thread.gmane.org/gmane.linux.file-systems.f2fs/3178/focus=3183 Signed-off-by: Bruce Ford Signed-off-by: Luca Porzio Fixes: ce39f9d17c14 ("mmc: support packed write command for eMMC4.5 devices") Signed-off-by: Ulf Hansson --- drivers/mmc/card/block.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c index 23b6c8e8701c..d8486168415a 100644 --- a/drivers/mmc/card/block.c +++ b/drivers/mmc/card/block.c @@ -65,8 +65,7 @@ MODULE_ALIAS("mmc:block"); #define MMC_SANITIZE_REQ_TIMEOUT 240000 #define MMC_EXTRACT_INDEX_FROM_ARG(x) ((x & 0x00FF0000) >> 16) -#define mmc_req_rel_wr(req) (((req->cmd_flags & REQ_FUA) || \ - (req->cmd_flags & REQ_META)) && \ +#define mmc_req_rel_wr(req) ((req->cmd_flags & REQ_FUA) && \ (rq_data_dir(req) == WRITE)) #define PACKED_CMD_VER 0x01 #define PACKED_CMD_WR 0x02 @@ -1467,13 +1466,9 @@ static void mmc_blk_rw_rq_prep(struct mmc_queue_req *mqrq, /* * Reliable writes are used to implement Forced Unit Access and - * REQ_META accesses, and are supported only on MMCs. - * - * XXX: this really needs a good explanation of why REQ_META - * is treated special. + * are supported only on MMCs. */ - bool do_rel_wr = ((req->cmd_flags & REQ_FUA) || - (req->cmd_flags & REQ_META)) && + bool do_rel_wr = (req->cmd_flags & REQ_FUA) && (rq_data_dir(req) == WRITE) && (md->flags & MMC_BLK_REL_WR); From c872a2d9e3627829591736ddd8e8710a0afb2f95 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 10 Nov 2015 13:08:15 +0100 Subject: [PATCH 030/335] netfilter: nfnetlink_log: work around uninitialized variable warning After a recent (correct) change, gcc started warning about the use of the 'flags' variable in nfulnl_recv_config() net/netfilter/nfnetlink_log.c: In function 'nfulnl_recv_config': net/netfilter/nfnetlink_log.c:320:14: warning: 'flags' may be used uninitialized in this function [-Wmaybe-uninitialized] net/netfilter/nfnetlink_log.c:828:6: note: 'flags' was declared here The warning first shows up in ARM s3c2410_defconfig with gcc-4.3 or higher (including 5.2.1, which is the latest version I checked) I tried working around it by rearranging the code but had no success with that. As a last resort, this initializes the variable to zero, which shuts up the warning, but means that we don't get a warning if the code is ever changed in a way that actually causes the variable to be used without first being written. Signed-off-by: Arnd Bergmann Fixes: 8cbc870829ec ("netfilter: nfnetlink_log: validate dependencies to avoid breaking atomicity") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nfnetlink_log.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 06eb48fceb42..740cce4685ac 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -825,7 +825,7 @@ nfulnl_recv_config(struct sock *ctnl, struct sk_buff *skb, struct net *net = sock_net(ctnl); struct nfnl_log_net *log = nfnl_log_pernet(net); int ret = 0; - u16 flags; + u16 flags = 0; if (nfula[NFULA_CFG_CMD]) { u_int8_t pf = nfmsg->nfgen_family; From 74ec4d55c4d243330d93fc52e23e37d2e76548ba Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 10 Nov 2015 13:22:15 +0100 Subject: [PATCH 031/335] netfilter: fix xt_TEE and xt_TPROXY dependencies Kconfig is too smart for its own good: a Kconfig line that states select NF_DEFRAG_IPV6 if IP6_NF_IPTABLES means that if IP6_NF_IPTABLES is set to 'm', then NF_DEFRAG_IPV6 will also be set to 'm', regardless of the state of the symbol from which it is selected. When the xt_TEE driver is built-in and nothing else forces NF_DEFRAG_IPV6 to be built-in, this causes a link-time error: net/built-in.o: In function `tee_tg6': net/netfilter/xt_TEE.c:46: undefined reference to `nf_dup_ipv6' This works around that behavior by changing the dependency to 'if IP6_NF_IPTABLES != n', which is interpreted as boolean expression rather than a tristate and causes the NF_DEFRAG_IPV6 symbol to be built-in as well. The bug only occurs once in thousands of 'randconfig' builds and does not really impact real users. From inspecting the other surrounding Kconfig symbols, I am guessing that NETFILTER_XT_TARGET_TPROXY and NETFILTER_XT_MATCH_SOCKET have the same issue. If not, this change should still be harmless. Signed-off-by: Arnd Bergmann Signed-off-by: Pablo Neira Ayuso --- net/netfilter/Kconfig | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig index e22349ea7256..4692782b5280 100644 --- a/net/netfilter/Kconfig +++ b/net/netfilter/Kconfig @@ -869,7 +869,7 @@ config NETFILTER_XT_TARGET_TEE depends on IPV6 || IPV6=n depends on !NF_CONNTRACK || NF_CONNTRACK select NF_DUP_IPV4 - select NF_DUP_IPV6 if IP6_NF_IPTABLES + select NF_DUP_IPV6 if IP6_NF_IPTABLES != n ---help--- This option adds a "TEE" target with which a packet can be cloned and this clone be rerouted to another nexthop. @@ -882,7 +882,7 @@ config NETFILTER_XT_TARGET_TPROXY depends on IP6_NF_IPTABLES || IP6_NF_IPTABLES=n depends on IP_NF_MANGLE select NF_DEFRAG_IPV4 - select NF_DEFRAG_IPV6 if IP6_NF_IPTABLES + select NF_DEFRAG_IPV6 if IP6_NF_IPTABLES != n help This option adds a `TPROXY' target, which is somewhat similar to REDIRECT. It can only be used in the mangle table and is useful @@ -1375,7 +1375,7 @@ config NETFILTER_XT_MATCH_SOCKET depends on IPV6 || IPV6=n depends on IP6_NF_IPTABLES || IP6_NF_IPTABLES=n select NF_DEFRAG_IPV4 - select NF_DEFRAG_IPV6 if IP6_NF_IPTABLES + select NF_DEFRAG_IPV6 if IP6_NF_IPTABLES != n help This option adds a `socket' match, which can be used to match packets for which a TCP or UDP socket lookup finds a valid socket. From aabc92bbe3cfe4c545f8ccdaaeeea012a46f0abf Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 10 Nov 2015 14:31:18 +0100 Subject: [PATCH 032/335] net: add __netdev_alloc_pcpu_stats() to indicate gfp flags nf_tables may create percpu counters from the packet path through its dynamic set instantiation infrastructure, so we need a way to allocate this through GFP_ATOMIC. Signed-off-by: Pablo Neira Ayuso Acked-by: David S. Miller --- include/linux/netdevice.h | 27 +++++++++++++++------------ 1 file changed, 15 insertions(+), 12 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 2c00772bd136..e9d0c8a75380 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2068,20 +2068,23 @@ struct pcpu_sw_netstats { struct u64_stats_sync syncp; }; -#define netdev_alloc_pcpu_stats(type) \ -({ \ - typeof(type) __percpu *pcpu_stats = alloc_percpu(type); \ - if (pcpu_stats) { \ - int __cpu; \ - for_each_possible_cpu(__cpu) { \ - typeof(type) *stat; \ - stat = per_cpu_ptr(pcpu_stats, __cpu); \ - u64_stats_init(&stat->syncp); \ - } \ - } \ - pcpu_stats; \ +#define __netdev_alloc_pcpu_stats(type, gfp) \ +({ \ + typeof(type) __percpu *pcpu_stats = alloc_percpu_gfp(type, gfp);\ + if (pcpu_stats) { \ + int __cpu; \ + for_each_possible_cpu(__cpu) { \ + typeof(type) *stat; \ + stat = per_cpu_ptr(pcpu_stats, __cpu); \ + u64_stats_init(&stat->syncp); \ + } \ + } \ + pcpu_stats; \ }) +#define netdev_alloc_pcpu_stats(type) \ + __netdev_alloc_pcpu_stats(type, GFP_KERNEL); + #include /* netdevice notifier chain. Please remember to update the rtnetlink From 086f332167d64b645d37405854f049b9ad7371ab Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 10 Nov 2015 13:39:42 +0100 Subject: [PATCH 033/335] netfilter: nf_tables: add clone interface to expression operations With the conversion of the counter expressions to make it percpu, we need to clone the percpu memory area, otherwise we crash when using counters from flow tables. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 16 ++++++++-- net/netfilter/nft_counter.c | 49 ++++++++++++++++++++++++++----- net/netfilter/nft_dynset.c | 5 ++-- 3 files changed, 58 insertions(+), 12 deletions(-) diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index c9149cc0a02d..4bd7508bedc9 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -618,6 +618,8 @@ struct nft_expr_ops { void (*eval)(const struct nft_expr *expr, struct nft_regs *regs, const struct nft_pktinfo *pkt); + int (*clone)(struct nft_expr *dst, + const struct nft_expr *src); unsigned int size; int (*init)(const struct nft_ctx *ctx, @@ -660,10 +662,20 @@ void nft_expr_destroy(const struct nft_ctx *ctx, struct nft_expr *expr); int nft_expr_dump(struct sk_buff *skb, unsigned int attr, const struct nft_expr *expr); -static inline void nft_expr_clone(struct nft_expr *dst, struct nft_expr *src) +static inline int nft_expr_clone(struct nft_expr *dst, struct nft_expr *src) { + int err; + __module_get(src->ops->type->owner); - memcpy(dst, src, src->ops->size); + if (src->ops->clone) { + dst->ops = src->ops; + err = src->ops->clone(dst, src); + if (err < 0) + return err; + } else { + memcpy(dst, src, src->ops->size); + } + return 0; } /** diff --git a/net/netfilter/nft_counter.c b/net/netfilter/nft_counter.c index 1067fb4c1ffa..c7808fc19719 100644 --- a/net/netfilter/nft_counter.c +++ b/net/netfilter/nft_counter.c @@ -47,27 +47,34 @@ static void nft_counter_eval(const struct nft_expr *expr, local_bh_enable(); } -static int nft_counter_dump(struct sk_buff *skb, const struct nft_expr *expr) +static void nft_counter_fetch(const struct nft_counter_percpu __percpu *counter, + struct nft_counter *total) { - struct nft_counter_percpu_priv *priv = nft_expr_priv(expr); - struct nft_counter_percpu *cpu_stats; - struct nft_counter total; + const struct nft_counter_percpu *cpu_stats; u64 bytes, packets; unsigned int seq; int cpu; - memset(&total, 0, sizeof(total)); + memset(total, 0, sizeof(*total)); for_each_possible_cpu(cpu) { - cpu_stats = per_cpu_ptr(priv->counter, cpu); + cpu_stats = per_cpu_ptr(counter, cpu); do { seq = u64_stats_fetch_begin_irq(&cpu_stats->syncp); bytes = cpu_stats->counter.bytes; packets = cpu_stats->counter.packets; } while (u64_stats_fetch_retry_irq(&cpu_stats->syncp, seq)); - total.packets += packets; - total.bytes += bytes; + total->packets += packets; + total->bytes += bytes; } +} + +static int nft_counter_dump(struct sk_buff *skb, const struct nft_expr *expr) +{ + struct nft_counter_percpu_priv *priv = nft_expr_priv(expr); + struct nft_counter total; + + nft_counter_fetch(priv->counter, &total); if (nla_put_be64(skb, NFTA_COUNTER_BYTES, cpu_to_be64(total.bytes)) || nla_put_be64(skb, NFTA_COUNTER_PACKETS, cpu_to_be64(total.packets))) @@ -118,6 +125,31 @@ static void nft_counter_destroy(const struct nft_ctx *ctx, free_percpu(priv->counter); } +static int nft_counter_clone(struct nft_expr *dst, const struct nft_expr *src) +{ + struct nft_counter_percpu_priv *priv = nft_expr_priv(src); + struct nft_counter_percpu_priv *priv_clone = nft_expr_priv(dst); + struct nft_counter_percpu __percpu *cpu_stats; + struct nft_counter_percpu *this_cpu; + struct nft_counter total; + + nft_counter_fetch(priv->counter, &total); + + cpu_stats = __netdev_alloc_pcpu_stats(struct nft_counter_percpu, + GFP_ATOMIC); + if (cpu_stats == NULL) + return ENOMEM; + + preempt_disable(); + this_cpu = this_cpu_ptr(cpu_stats); + this_cpu->counter.packets = total.packets; + this_cpu->counter.bytes = total.bytes; + preempt_enable(); + + priv_clone->counter = cpu_stats; + return 0; +} + static struct nft_expr_type nft_counter_type; static const struct nft_expr_ops nft_counter_ops = { .type = &nft_counter_type, @@ -126,6 +158,7 @@ static const struct nft_expr_ops nft_counter_ops = { .init = nft_counter_init, .destroy = nft_counter_destroy, .dump = nft_counter_dump, + .clone = nft_counter_clone, }; static struct nft_expr_type nft_counter_type __read_mostly = { diff --git a/net/netfilter/nft_dynset.c b/net/netfilter/nft_dynset.c index 513a8ef60a59..9dec3bd1b63c 100644 --- a/net/netfilter/nft_dynset.c +++ b/net/netfilter/nft_dynset.c @@ -50,8 +50,9 @@ static void *nft_dynset_new(struct nft_set *set, const struct nft_expr *expr, } ext = nft_set_elem_ext(set, elem); - if (priv->expr != NULL) - nft_expr_clone(nft_set_ext_expr(ext), priv->expr); + if (priv->expr != NULL && + nft_expr_clone(nft_set_ext_expr(ext), priv->expr) < 0) + return NULL; return elem; } From e0bedada3a497d0640dd5db93e7ad0735f487492 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Fri, 6 Nov 2015 09:58:06 +0100 Subject: [PATCH 034/335] s390/ipl: fix out of bounds access in scpdata_write The input buffer in reipl_fcp_scpdata_write is accessed out of bounds when an offset is specified. The problem is that the offset refers to the data we should write to and not to the buffer we read from. So instead of memcpy(scp_data, buf + off, count); we could just do memcpy(scp_data + off, buf, count); However we not only modify the data but also store its length. For this to work we'd need to remember a state per open FH. Since that's not possible with sysfs callbacks let's just fail when an offset is specified. Signed-off-by: Sebastian Ott Acked-by: Hendrik Brueckner Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/ipl.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index f6d8acd7e136..9f80f65c3d97 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -687,21 +687,14 @@ static ssize_t reipl_fcp_scpdata_write(struct file *filp, struct kobject *kobj, struct bin_attribute *attr, char *buf, loff_t off, size_t count) { + size_t scpdata_len = count; size_t padding; - size_t scpdata_len; - if (off < 0) + + if (off) return -EINVAL; - if (off >= DIAG308_SCPDATA_SIZE) - return -ENOSPC; - - if (count > DIAG308_SCPDATA_SIZE - off) - count = DIAG308_SCPDATA_SIZE - off; - - memcpy(reipl_block_fcp->ipl_info.fcp.scp_data, buf + off, count); - scpdata_len = off + count; - + memcpy(reipl_block_fcp->ipl_info.fcp.scp_data, buf, count); if (scpdata_len % 8) { padding = 8 - (scpdata_len % 8); memset(reipl_block_fcp->ipl_info.fcp.scp_data + scpdata_len, @@ -717,7 +710,7 @@ static ssize_t reipl_fcp_scpdata_write(struct file *filp, struct kobject *kobj, } static struct bin_attribute sys_reipl_fcp_scp_data_attr = __BIN_ATTR(scp_data, (S_IRUGO | S_IWUSR), reipl_fcp_scpdata_read, - reipl_fcp_scpdata_write, PAGE_SIZE); + reipl_fcp_scpdata_write, DIAG308_SCPDATA_SIZE); static struct bin_attribute *reipl_fcp_bin_attrs[] = { &sys_reipl_fcp_scp_data_attr, From 18e22a1772260045648e49e2048ecc7193422a80 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 29 Jun 2015 18:39:54 +0200 Subject: [PATCH 035/335] s390: add support for ipl devices in subchannel sets > 0 Allow to ipl from CCW based devices residing in any subchannel set. Reviewed-by: Michael Holzheu Signed-off-by: Sebastian Ott Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/cio.h | 1 + arch/s390/include/asm/ipl.h | 3 ++- arch/s390/kernel/ipl.c | 48 ++++++++++++++++++++++++++++--------- drivers/s390/cio/chsc.c | 37 +++++++++++++--------------- drivers/s390/cio/chsc.h | 15 ++++++++++++ drivers/s390/cio/cio.c | 14 +++++++++++ 6 files changed, 86 insertions(+), 32 deletions(-) diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h index 0c5d8ee657f0..d1e7b0a0feeb 100644 --- a/arch/s390/include/asm/cio.h +++ b/arch/s390/include/asm/cio.h @@ -312,6 +312,7 @@ extern void css_schedule_reprobe(void); extern void reipl_ccw_dev(struct ccw_dev_id *id); struct cio_iplinfo { + u8 ssid; u16 devno; int is_qdio; }; diff --git a/arch/s390/include/asm/ipl.h b/arch/s390/include/asm/ipl.h index 39ae6a359747..86634e71b69f 100644 --- a/arch/s390/include/asm/ipl.h +++ b/arch/s390/include/asm/ipl.h @@ -64,7 +64,8 @@ struct ipl_block_fcp { struct ipl_block_ccw { u8 reserved1[84]; - u8 reserved2[2]; + u16 reserved2 : 13; + u8 ssid : 3; u16 devno; u8 vm_flags; u8 reserved3[3]; diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 9f80f65c3d97..b1f0a90f933b 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -121,6 +121,7 @@ static char *dump_type_str(enum dump_type type) * Must be in data section since the bss section * is not cleared when these are accessed. */ +static u8 ipl_ssid __attribute__((__section__(".data"))) = 0; static u16 ipl_devno __attribute__((__section__(".data"))) = 0; u32 ipl_flags __attribute__((__section__(".data"))) = 0; @@ -197,6 +198,33 @@ static ssize_t sys_##_prefix##_##_name##_show(struct kobject *kobj, \ return snprintf(page, PAGE_SIZE, _format, ##args); \ } +#define IPL_ATTR_CCW_STORE_FN(_prefix, _name, _ipl_blk) \ +static ssize_t sys_##_prefix##_##_name##_store(struct kobject *kobj, \ + struct kobj_attribute *attr, \ + const char *buf, size_t len) \ +{ \ + unsigned long long ssid, devno; \ + \ + if (sscanf(buf, "0.%llx.%llx\n", &ssid, &devno) != 2) \ + return -EINVAL; \ + \ + if (ssid > __MAX_SSID || devno > __MAX_SUBCHANNEL) \ + return -EINVAL; \ + \ + _ipl_blk.ssid = ssid; \ + _ipl_blk.devno = devno; \ + return len; \ +} + +#define DEFINE_IPL_CCW_ATTR_RW(_prefix, _name, _ipl_blk) \ +IPL_ATTR_SHOW_FN(_prefix, _name, "0.%x.%04x\n", \ + _ipl_blk.ssid, _ipl_blk.devno); \ +IPL_ATTR_CCW_STORE_FN(_prefix, _name, _ipl_blk); \ +static struct kobj_attribute sys_##_prefix##_##_name##_attr = \ + __ATTR(_name, (S_IRUGO | S_IWUSR), \ + sys_##_prefix##_##_name##_show, \ + sys_##_prefix##_##_name##_store) \ + #define DEFINE_IPL_ATTR_RO(_prefix, _name, _format, _value) \ IPL_ATTR_SHOW_FN(_prefix, _name, _format, _value) \ static struct kobj_attribute sys_##_prefix##_##_name##_attr = \ @@ -395,7 +423,7 @@ static ssize_t sys_ipl_device_show(struct kobject *kobj, switch (ipl_info.type) { case IPL_TYPE_CCW: - return sprintf(page, "0.0.%04x\n", ipl_devno); + return sprintf(page, "0.%x.%04x\n", ipl_ssid, ipl_devno); case IPL_TYPE_FCP: case IPL_TYPE_FCP_DUMP: return sprintf(page, "0.0.%04x\n", ipl->ipl_info.fcp.devno); @@ -807,9 +835,7 @@ static struct attribute_group reipl_fcp_attr_group = { }; /* CCW reipl device attributes */ - -DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", - reipl_block_ccw->ipl_info.ccw.devno); +DEFINE_IPL_CCW_ATTR_RW(reipl_ccw, device, reipl_block_ccw->ipl_info.ccw); /* NSS wrapper */ static ssize_t reipl_nss_loadparm_show(struct kobject *kobj, @@ -1049,8 +1075,8 @@ static void __reipl_run(void *unused) switch (reipl_method) { case REIPL_METHOD_CCW_CIO: + devid.ssid = reipl_block_ccw->ipl_info.ccw.ssid; devid.devno = reipl_block_ccw->ipl_info.ccw.devno; - devid.ssid = 0; reipl_ccw_dev(&devid); break; case REIPL_METHOD_CCW_VM: @@ -1185,6 +1211,7 @@ static int __init reipl_ccw_init(void) reipl_block_ccw_init(reipl_block_ccw); if (ipl_info.type == IPL_TYPE_CCW) { + reipl_block_ccw->ipl_info.ccw.ssid = ipl_ssid; reipl_block_ccw->ipl_info.ccw.devno = ipl_devno; reipl_block_ccw_fill_parms(reipl_block_ccw); } @@ -1329,9 +1356,7 @@ static struct attribute_group dump_fcp_attr_group = { }; /* CCW dump device attributes */ - -DEFINE_IPL_ATTR_RW(dump_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", - dump_block_ccw->ipl_info.ccw.devno); +DEFINE_IPL_CCW_ATTR_RW(dump_ccw, device, dump_block_ccw->ipl_info.ccw); static struct attribute *dump_ccw_attrs[] = { &sys_dump_ccw_device_attr.attr, @@ -1411,8 +1436,8 @@ static void __dump_run(void *unused) switch (dump_method) { case DUMP_METHOD_CCW_CIO: + devid.ssid = dump_block_ccw->ipl_info.ccw.ssid; devid.devno = dump_block_ccw->ipl_info.ccw.devno; - devid.ssid = 0; reipl_ccw_dev(&devid); break; case DUMP_METHOD_CCW_VM: @@ -1932,14 +1957,14 @@ void __init setup_ipl(void) ipl_info.type = get_ipl_type(); switch (ipl_info.type) { case IPL_TYPE_CCW: + ipl_info.data.ccw.dev_id.ssid = ipl_ssid; ipl_info.data.ccw.dev_id.devno = ipl_devno; - ipl_info.data.ccw.dev_id.ssid = 0; break; case IPL_TYPE_FCP: case IPL_TYPE_FCP_DUMP: + ipl_info.data.fcp.dev_id.ssid = 0; ipl_info.data.fcp.dev_id.devno = IPL_PARMBLOCK_START->ipl_info.fcp.devno; - ipl_info.data.fcp.dev_id.ssid = 0; ipl_info.data.fcp.wwpn = IPL_PARMBLOCK_START->ipl_info.fcp.wwpn; ipl_info.data.fcp.lun = IPL_PARMBLOCK_START->ipl_info.fcp.lun; break; @@ -1971,6 +1996,7 @@ void __init ipl_save_parameters(void) if (cio_get_iplinfo(&iplinfo)) return; + ipl_ssid = iplinfo.ssid; ipl_devno = iplinfo.devno; ipl_flags |= IPL_DEVNO_VALID; if (!iplinfo.is_qdio) diff --git a/drivers/s390/cio/chsc.c b/drivers/s390/cio/chsc.c index 548a18916a31..a831d18596a5 100644 --- a/drivers/s390/cio/chsc.c +++ b/drivers/s390/cio/chsc.c @@ -1080,28 +1080,10 @@ void __init chsc_init_cleanup(void) free_page((unsigned long)sei_page); } -int chsc_enable_facility(int operation_code) +int __chsc_enable_facility(struct chsc_sda_area *sda_area, int operation_code) { - unsigned long flags; int ret; - struct { - struct chsc_header request; - u8 reserved1:4; - u8 format:4; - u8 reserved2; - u16 operation_code; - u32 reserved3; - u32 reserved4; - u32 operation_data_area[252]; - struct chsc_header response; - u32 reserved5:4; - u32 format2:4; - u32 reserved6:24; - } __attribute__ ((packed)) *sda_area; - spin_lock_irqsave(&chsc_page_lock, flags); - memset(chsc_page, 0, PAGE_SIZE); - sda_area = chsc_page; sda_area->request.length = 0x0400; sda_area->request.code = 0x0031; sda_area->operation_code = operation_code; @@ -1119,10 +1101,25 @@ int chsc_enable_facility(int operation_code) default: ret = chsc_error_from_response(sda_area->response.code); } +out: + return ret; +} + +int chsc_enable_facility(int operation_code) +{ + struct chsc_sda_area *sda_area; + unsigned long flags; + int ret; + + spin_lock_irqsave(&chsc_page_lock, flags); + memset(chsc_page, 0, PAGE_SIZE); + sda_area = chsc_page; + + ret = __chsc_enable_facility(sda_area, operation_code); if (ret != 0) CIO_CRW_EVENT(2, "chsc: sda (oc=%x) failed (rc=%04x)\n", operation_code, sda_area->response.code); -out: + spin_unlock_irqrestore(&chsc_page_lock, flags); return ret; } diff --git a/drivers/s390/cio/chsc.h b/drivers/s390/cio/chsc.h index 76c9b50700b2..0de134c3a204 100644 --- a/drivers/s390/cio/chsc.h +++ b/drivers/s390/cio/chsc.h @@ -115,6 +115,20 @@ struct chsc_scpd { u8 data[PAGE_SIZE - 20]; } __attribute__ ((packed)); +struct chsc_sda_area { + struct chsc_header request; + u8 :4; + u8 format:4; + u8 :8; + u16 operation_code; + u32 :32; + u32 :32; + u32 operation_data_area[252]; + struct chsc_header response; + u32 :4; + u32 format2:4; + u32 :24; +} __packed __aligned(PAGE_SIZE); extern int chsc_get_ssd_info(struct subchannel_id schid, struct chsc_ssd_info *ssd); @@ -122,6 +136,7 @@ extern int chsc_determine_css_characteristics(void); extern int chsc_init(void); extern void chsc_init_cleanup(void); +int __chsc_enable_facility(struct chsc_sda_area *sda_area, int operation_code); extern int chsc_enable_facility(int); struct channel_subsystem; extern int chsc_secm(struct channel_subsystem *, int); diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index b5620e818d6b..690b8547e828 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -925,18 +925,32 @@ void reipl_ccw_dev(struct ccw_dev_id *devid) int __init cio_get_iplinfo(struct cio_iplinfo *iplinfo) { + static struct chsc_sda_area sda_area __initdata; struct subchannel_id schid; struct schib schib; schid = *(struct subchannel_id *)&S390_lowcore.subchannel_id; if (!schid.one) return -ENODEV; + + if (schid.ssid) { + /* + * Firmware should have already enabled MSS but whoever started + * the kernel might have initiated a channel subsystem reset. + * Ensure that MSS is enabled. + */ + memset(&sda_area, 0, sizeof(sda_area)); + if (__chsc_enable_facility(&sda_area, CHSC_SDA_OC_MSS)) + return -ENODEV; + } if (stsch_err(schid, &schib)) return -ENODEV; if (schib.pmcw.st != SUBCHANNEL_TYPE_IO) return -ENODEV; if (!schib.pmcw.dnv) return -ENODEV; + + iplinfo->ssid = schid.ssid; iplinfo->devno = schib.pmcw.dev; iplinfo->is_qdio = schib.pmcw.qf; return 0; From e387753c5fb4ae40393ba3f00b7958744670cac2 Mon Sep 17 00:00:00 2001 From: Sascha Silbe Date: Tue, 27 Oct 2015 18:29:52 +0100 Subject: [PATCH 036/335] s390/zcrypt: Fix kernel crash on systems without AP bus support On systems without AP bus (e.g. KVM) the kernel crashes during init calls when zcrypt is built-in: kernel BUG at drivers/base/driver.c:153! illegal operation: 0001 ilc:1 [#1] SMP Modules linked in: CPU: 1 PID: 1 Comm: swapper/0 Not tainted 4.2.0+ #221 task: 0000000010a40000 ti: 0000000010a48000 task.ti:0000000010a48000 Krnl PSW : 0704c00180000000 0000000000592bd6(driver_register+0x106/0x140) R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:0 PM:0 EA:3 0000000000000012 0000000000000000 0000000000c45328 0000000000c44e30 00000000009ef63c 000000000067f598 0000000000cf3c58 0000000000000000 000000000000007b 0000000000cb1030 0000000000000002 0000000000000000 0000000000ca8580 0000000010306700 00000000001001d8 0000000010a4bd88 Krnl Code: 0000000000592bc6: f0b00004ebcf srp 4(12,%r0),3023(%r14),0 0000000000592bcc: f0a0000407f4 srp 4(11,%r0),2036,0 #0000000000592bd2: a7f40001 brc 15,592bd4 >0000000000592bd6: e330d0000004 lg %r3,0(%r13) 0000000000592bdc: c0200021edfd larl %r2,9d07d6 0000000000592be2: c0e500126d8f brasl %r14,7e0700 0000000000592be8: e330d0080004 lg %r3,8(%r13) 0000000000592bee: a7f4ffab brc 15,592b44 Call Trace: ([<00000000001001c8>] do_one_initcall+0x90/0x1d0) [<0000000000c6dd34>] kernel_init_freeable+0x1e4/0x2a0 [<00000000007db53a>] kernel_init+0x2a/0x120 [<00000000007e8ece>] kernel_thread_starter+0x6/0xc [<00000000007e8ec8>] kernel_thread_starter+0x0/0xc Last Breaking-Event-Address: [<0000000000592bd2>] driver_register+0x102/0x140 When zcrypt is built as a module, the module loader ensures that the driver modules cannot be loaded if the AP bus module returns an error during initialisation. But if zcrypt and the driver are built-in, the driver is getting initialised even if the AP bus initialisation failed. The driver invokes ap_driver_register() during initialisation, which then causes operations on uninitialised data structures to be performed. Explicitly protect ap_driver_register() by introducing an "initialised" flag that gets set iff the AP bus initialisation was successful. When the AP bus initialisation failed, ap_driver_register() will error out with -ENODEV, causing the driver initialisation to fail as well. Test results: 1. Inside KVM (no AP bus), zcrypt built-in Boots. /sys/bus/ap not present (expected). 2. Inside KVM (no AP bus), zcrypt as module Boots. Loading zcrypt_cex4 fails because loading ap_bus fails (expected). 3. On LPAR with CEX5, zcrypt built-in Boots. /sys/bus/ap/devices/card* present but .../card*/type missing (i.e. zcrypt_device_register() fails, unrelated issue). 4. On LPAR with CEX5, zcrypt as module Boots. Loading zcrypt_cex4 successful, /sys/bus/ap/devices/card*/type present. No further testing (user-space functionality) was done. Signed-off-by: Sascha Silbe Signed-off-by: Martin Schwidefsky --- drivers/s390/crypto/ap_bus.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/s390/crypto/ap_bus.c b/drivers/s390/crypto/ap_bus.c index 9cb3dfbcaddb..61f768518a34 100644 --- a/drivers/s390/crypto/ap_bus.c +++ b/drivers/s390/crypto/ap_bus.c @@ -74,6 +74,7 @@ static struct device *ap_root_device = NULL; static struct ap_config_info *ap_configuration; static DEFINE_SPINLOCK(ap_device_list_lock); static LIST_HEAD(ap_device_list); +static bool initialised; /* * Workqueue timer for bus rescan. @@ -1384,6 +1385,9 @@ int ap_driver_register(struct ap_driver *ap_drv, struct module *owner, { struct device_driver *drv = &ap_drv->driver; + if (!initialised) + return -ENODEV; + drv->bus = &ap_bus_type; drv->probe = ap_device_probe; drv->remove = ap_device_remove; @@ -1808,6 +1812,7 @@ int __init ap_module_init(void) goto out_pm; queue_work(system_long_wq, &ap_scan_work); + initialised = true; return 0; @@ -1837,6 +1842,7 @@ void ap_module_exit(void) { int i; + initialised = false; ap_reset_domain(); ap_poll_thread_stop(); del_timer_sync(&ap_config_timer); From 121a868d05500b9d7e5108cc52474dafbf60e285 Mon Sep 17 00:00:00 2001 From: Sascha Silbe Date: Wed, 28 Oct 2015 11:06:08 +0100 Subject: [PATCH 037/335] s390/zcrypt: Fix initialisation when zcrypt is built-in ap_bus and zcrypt_api assumed module information to always be present and initialisation to be done in module loading order (symbol dependencies). These assumptions don't hold if zcrypt is built-in; THIS_MODULE will be NULL in this case and init call order is linker order, i.e. Makefile order. Fix initialisation order by ordering the object files in the Makefile according to their dependencies, like the module loader would do. Fix message type registration by using a dedicated "name" field rather than piggy-backing on the module ("owner") information. There's no change to the requirement that module name and msgtype name are identical. The existing name macros are used. We don't need any special code for dealing with the drivers being built-in; the generic module support code already does the right thing. Test results: 1. CONFIG_MODULES=y, CONFIG_ZCRYPT=y KVM: boots, no /sys/bus/ap (expected) LPAR with CEX5: boots, /sys/bus/ap/devices/card*/type present 2. CONFIG_MODULES=y, CONFIG_ZCRYPT=m=: KVM: boots, loading zcrypt_cex4 (and ap) fails (expected) LPAR with CEX5: boots, loading =zcrypt_cex4= succeeds, /sys/bus/ap/devices/card*/type present after explicit module loading 3. CONFIG_MODULES unset, CONFIG_ZCRYPT=y: KVM: boots, no /sys/bus/ap (expected) LPAR with CEX5: boots, /sys/bus/ap/devices/card*/type present No further testing (user-space functionality) was done. Fixes: 3b6245fd303f ("s390/zcrypt: Separate msgtype implementation from card modules.") Signed-off-by: Sascha Silbe Signed-off-by: Martin Schwidefsky --- drivers/s390/crypto/Makefile | 7 +++++-- drivers/s390/crypto/zcrypt_api.c | 10 ++++------ drivers/s390/crypto/zcrypt_api.h | 1 + drivers/s390/crypto/zcrypt_msgtype50.c | 1 + drivers/s390/crypto/zcrypt_msgtype6.c | 3 +++ 5 files changed, 14 insertions(+), 8 deletions(-) diff --git a/drivers/s390/crypto/Makefile b/drivers/s390/crypto/Makefile index 57f710b3c8a4..b8ab18676e69 100644 --- a/drivers/s390/crypto/Makefile +++ b/drivers/s390/crypto/Makefile @@ -3,6 +3,9 @@ # ap-objs := ap_bus.o -obj-$(CONFIG_ZCRYPT) += ap.o zcrypt_api.o zcrypt_pcixcc.o -obj-$(CONFIG_ZCRYPT) += zcrypt_cex2a.o zcrypt_cex4.o +# zcrypt_api depends on ap +obj-$(CONFIG_ZCRYPT) += ap.o zcrypt_api.o +# msgtype* depend on zcrypt_api obj-$(CONFIG_ZCRYPT) += zcrypt_msgtype6.o zcrypt_msgtype50.o +# adapter drivers depend on ap, zcrypt_api and msgtype* +obj-$(CONFIG_ZCRYPT) += zcrypt_pcixcc.o zcrypt_cex2a.o zcrypt_cex4.o diff --git a/drivers/s390/crypto/zcrypt_api.c b/drivers/s390/crypto/zcrypt_api.c index a9603ebbc1f8..9f8fa42c062c 100644 --- a/drivers/s390/crypto/zcrypt_api.c +++ b/drivers/s390/crypto/zcrypt_api.c @@ -317,11 +317,9 @@ EXPORT_SYMBOL(zcrypt_device_unregister); void zcrypt_msgtype_register(struct zcrypt_ops *zops) { - if (zops->owner) { - spin_lock_bh(&zcrypt_ops_list_lock); - list_add_tail(&zops->list, &zcrypt_ops_list); - spin_unlock_bh(&zcrypt_ops_list_lock); - } + spin_lock_bh(&zcrypt_ops_list_lock); + list_add_tail(&zops->list, &zcrypt_ops_list); + spin_unlock_bh(&zcrypt_ops_list_lock); } EXPORT_SYMBOL(zcrypt_msgtype_register); @@ -342,7 +340,7 @@ struct zcrypt_ops *__ops_lookup(unsigned char *name, int variant) spin_lock_bh(&zcrypt_ops_list_lock); list_for_each_entry(zops, &zcrypt_ops_list, list) { if ((zops->variant == variant) && - (!strncmp(zops->owner->name, name, MODULE_NAME_LEN))) { + (!strncmp(zops->name, name, sizeof(zops->name)))) { found = 1; break; } diff --git a/drivers/s390/crypto/zcrypt_api.h b/drivers/s390/crypto/zcrypt_api.h index 750876891931..38618f05ad92 100644 --- a/drivers/s390/crypto/zcrypt_api.h +++ b/drivers/s390/crypto/zcrypt_api.h @@ -96,6 +96,7 @@ struct zcrypt_ops { struct list_head list; /* zcrypt ops list. */ struct module *owner; int variant; + char name[128]; }; struct zcrypt_device { diff --git a/drivers/s390/crypto/zcrypt_msgtype50.c b/drivers/s390/crypto/zcrypt_msgtype50.c index 71ceee9137a8..74edf2934e7c 100644 --- a/drivers/s390/crypto/zcrypt_msgtype50.c +++ b/drivers/s390/crypto/zcrypt_msgtype50.c @@ -513,6 +513,7 @@ static struct zcrypt_ops zcrypt_msgtype50_ops = { .rsa_modexpo = zcrypt_cex2a_modexpo, .rsa_modexpo_crt = zcrypt_cex2a_modexpo_crt, .owner = THIS_MODULE, + .name = MSGTYPE50_NAME, .variant = MSGTYPE50_VARIANT_DEFAULT, }; diff --git a/drivers/s390/crypto/zcrypt_msgtype6.c b/drivers/s390/crypto/zcrypt_msgtype6.c index 74762214193b..9a2dd472c1cc 100644 --- a/drivers/s390/crypto/zcrypt_msgtype6.c +++ b/drivers/s390/crypto/zcrypt_msgtype6.c @@ -1119,6 +1119,7 @@ static long zcrypt_msgtype6_rng(struct zcrypt_device *zdev, */ static struct zcrypt_ops zcrypt_msgtype6_norng_ops = { .owner = THIS_MODULE, + .name = MSGTYPE06_NAME, .variant = MSGTYPE06_VARIANT_NORNG, .rsa_modexpo = zcrypt_msgtype6_modexpo, .rsa_modexpo_crt = zcrypt_msgtype6_modexpo_crt, @@ -1127,6 +1128,7 @@ static struct zcrypt_ops zcrypt_msgtype6_norng_ops = { static struct zcrypt_ops zcrypt_msgtype6_ops = { .owner = THIS_MODULE, + .name = MSGTYPE06_NAME, .variant = MSGTYPE06_VARIANT_DEFAULT, .rsa_modexpo = zcrypt_msgtype6_modexpo, .rsa_modexpo_crt = zcrypt_msgtype6_modexpo_crt, @@ -1136,6 +1138,7 @@ static struct zcrypt_ops zcrypt_msgtype6_ops = { static struct zcrypt_ops zcrypt_msgtype6_ep11_ops = { .owner = THIS_MODULE, + .name = MSGTYPE06_NAME, .variant = MSGTYPE06_VARIANT_EP11, .rsa_modexpo = NULL, .rsa_modexpo_crt = NULL, From e41b104c7dba92443e594e6bc86e4b0bf1cdf573 Mon Sep 17 00:00:00 2001 From: Zhou Chengming Date: Fri, 6 Nov 2015 14:25:00 +0800 Subject: [PATCH 038/335] livepatch: x86: fix relocation computation with kASLR With kASLR enabled, old_addr provided by patch module is being shifted accrodingly so that the symbol lookups work. To have module relocations handled properly as well, the same transformation needs to be perfomed on relocation address information. [jkosina@suse.cz: extended / reworded changelog a bit] Reported-by: Cyril B. Signed-off-by: Zhou Chengming Acked-by: Josh Poimboeuf Signed-off-by: Jiri Kosina --- kernel/livepatch/core.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 6e5344112419..db545cbcdb89 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -294,6 +294,12 @@ static int klp_write_object_relocations(struct module *pmod, for (reloc = obj->relocs; reloc->name; reloc++) { if (!klp_is_module(obj)) { + +#if defined(CONFIG_RANDOMIZE_BASE) + /* If KASLR has been enabled, adjust old value accordingly */ + if (kaslr_enabled()) + reloc->val += kaslr_offset(); +#endif ret = klp_verify_vmlinux_symbol(reloc->name, reloc->val); if (ret) From 46561c3959d6307d22139c24cd0bf196162e5681 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Wed, 11 Nov 2015 10:19:31 -0800 Subject: [PATCH 039/335] x86/mpx: Do proper get_user() when running 32-bit binaries on 64-bit kernels When you call get_user(foo, bar), you effectively do a copy_from_user(&foo, bar, sizeof(*bar)); Note that the sizeof() is implicit. When we reach out to userspace to try to zap an entire "bounds table" we need to go read a "bounds directory entry" in order to locate the table's address. The size of a "directory entry" depends on the binary being run and is always the size of a pointer. But, when we have a 64-bit kernel and a 32-bit application, the directory entry is still only 32-bits long, but we fetch it with a 64-bit pointer which makes get_user() does a 64-bit fetch. Reading 4 extra bytes isn't harmful, unless we are at the end of and run off the table. It might also cause the zero page to get faulted in unnecessarily even if you are not at the end. Fix it up by doing a special 32-bit get_user() via a cast when we have 32-bit userspace. Signed-off-by: Dave Hansen Cc: Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Dave Hansen Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20151111181931.3ACF6822@viggo.jf.intel.com Signed-off-by: Ingo Molnar --- arch/x86/mm/mpx.c | 25 ++++++++++++++++++++++++- 1 file changed, 24 insertions(+), 1 deletion(-) diff --git a/arch/x86/mm/mpx.c b/arch/x86/mm/mpx.c index b0ae85f90f10..6127c5e3a19b 100644 --- a/arch/x86/mm/mpx.c +++ b/arch/x86/mm/mpx.c @@ -585,6 +585,29 @@ static unsigned long mpx_bd_entry_to_bt_addr(struct mm_struct *mm, return bt_addr; } +/* + * We only want to do a 4-byte get_user() on 32-bit. Otherwise, + * we might run off the end of the bounds table if we are on + * a 64-bit kernel and try to get 8 bytes. + */ +int get_user_bd_entry(struct mm_struct *mm, unsigned long *bd_entry_ret, + long __user *bd_entry_ptr) +{ + u32 bd_entry_32; + int ret; + + if (is_64bit_mm(mm)) + return get_user(*bd_entry_ret, bd_entry_ptr); + + /* + * Note that get_user() uses the type of the *pointer* to + * establish the size of the get, not the destination. + */ + ret = get_user(bd_entry_32, (u32 __user *)bd_entry_ptr); + *bd_entry_ret = bd_entry_32; + return ret; +} + /* * Get the base of bounds tables pointed by specific bounds * directory entry. @@ -605,7 +628,7 @@ static int get_bt_addr(struct mm_struct *mm, int need_write = 0; pagefault_disable(); - ret = get_user(bd_entry, bd_entry_ptr); + ret = get_user_bd_entry(mm, &bd_entry, bd_entry_ptr); pagefault_enable(); if (!ret) break; From f3119b830264d89d216bfb378ab65065dffa02d9 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Wed, 11 Nov 2015 10:19:34 -0800 Subject: [PATCH 040/335] x86/mpx: Fix 32-bit address space calculation I received a bug report that running 32-bit MPX binaries on 64-bit kernels was broken. I traced it down to this little code snippet. We were switching our "number of bounds directory entries" calculation correctly. But, we didn't switch the other side of the calculation: the virtual space size. This meant that we were calculating an absurd size for bd_entry_virt_space() on 32-bit because we used the 64-bit virt_space. This was _also_ broken for 32-bit kernels running on 64-bit hardware since boot_cpu_data.x86_virt_bits=48 even when running in 32-bit mode. Correct that and properly handle all 3 possible cases: 1. 32-bit binary on 64-bit kernel 2. 64-bit binary on 64-bit kernel 3. 32-bit binary on 32-bit kernel This manifested in having bounds tables not properly unmapped. It "leaked" memory but had no functional impact otherwise. Signed-off-by: Dave Hansen Cc: Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Dave Hansen Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20151111181934.FA7FAC34@viggo.jf.intel.com Signed-off-by: Ingo Molnar --- arch/x86/mm/mpx.c | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) diff --git a/arch/x86/mm/mpx.c b/arch/x86/mm/mpx.c index 6127c5e3a19b..1202d5ca2fb5 100644 --- a/arch/x86/mm/mpx.c +++ b/arch/x86/mm/mpx.c @@ -723,11 +723,23 @@ static unsigned long mpx_get_bt_entry_offset_bytes(struct mm_struct *mm, */ static inline unsigned long bd_entry_virt_space(struct mm_struct *mm) { - unsigned long long virt_space = (1ULL << boot_cpu_data.x86_virt_bits); - if (is_64bit_mm(mm)) - return virt_space / MPX_BD_NR_ENTRIES_64; - else - return virt_space / MPX_BD_NR_ENTRIES_32; + unsigned long long virt_space; + unsigned long long GB = (1ULL << 30); + + /* + * This covers 32-bit emulation as well as 32-bit kernels + * running on 64-bit harware. + */ + if (!is_64bit_mm(mm)) + return (4ULL * GB) / MPX_BD_NR_ENTRIES_32; + + /* + * 'x86_virt_bits' returns what the hardware is capable + * of, and returns the full >32-bit adddress space when + * running 32-bit kernels on 64-bit hardware. + */ + virt_space = (1ULL << boot_cpu_data.x86_virt_bits); + return virt_space / MPX_BD_NR_ENTRIES_64; } /* From ab6b52947545a5355154f64f449f97af9d05845f Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Tue, 10 Nov 2015 16:23:54 -0800 Subject: [PATCH 041/335] x86/fpu: Fix 32-bit signal frame handling (This should have gone to LKML originally. Sorry for the extra noise, folks on the cc.) Background: Signal frames on x86 have two formats: 1. For 32-bit executables (whether on a real 32-bit kernel or under 32-bit emulation on a 64-bit kernel) we have a 'fpregset_t' that includes the "FSAVE" registers. 2. For 64-bit executables (on 64-bit kernels obviously), the 'fpregset_t' is smaller and does not contain the "FSAVE" state. When creating the signal frame, we have to be aware of whether we are running a 32 or 64-bit executable so we create the correct format signal frame. Problem: save_xstate_epilog() uses 'fx_sw_reserved_ia32' whenever it is called for a 32-bit executable. This is for real 32-bit and ia32 emulation. But, fpu__init_prepare_fx_sw_frame() only initializes 'fx_sw_reserved_ia32' when emulation is enabled, *NOT* for real 32-bit kernels. This leads to really wierd situations where 32-bit programs lose their extended state when returning from a signal handler. The kernel copies the uninitialized (zero) 'fx_sw_reserved_ia32' out to userspace in save_xstate_epilog(). But when returning from the signal, the kernel errors out in check_for_xstate() when it does not see FP_XSTATE_MAGIC1 present (because it was zeroed). This leads to the FPU/XSAVE state being initialized. For MPX, this leads to the most permissive state and means we silently lose bounds violations. I think this would also mean that we could lose *ANY* FPU/SSE/AVX state. I'm not sure why no one has spotted this bug. I believe this was broken by: 72a671ced66d ("x86, fpu: Unify signal handling code paths for x86 and x86_64 kernels") way back in 2012. Signed-off-by: Dave Hansen Cc: Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Brian Gerst Cc: Denys Vlasenko Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: dave@sr71.net Cc: fenghua.yu@intel.com Cc: yu-cheng.yu@intel.com Link: http://lkml.kernel.org/r/20151111002354.A0799571@viggo.jf.intel.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/fpu/signal.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c index ef29b742cea7..31c6a60505e6 100644 --- a/arch/x86/kernel/fpu/signal.c +++ b/arch/x86/kernel/fpu/signal.c @@ -385,20 +385,19 @@ fpu__alloc_mathframe(unsigned long sp, int ia32_frame, */ void fpu__init_prepare_fx_sw_frame(void) { - int fsave_header_size = sizeof(struct fregs_state); int size = xstate_size + FP_XSTATE_MAGIC2_SIZE; - if (config_enabled(CONFIG_X86_32)) - size += fsave_header_size; - fx_sw_reserved.magic1 = FP_XSTATE_MAGIC1; fx_sw_reserved.extended_size = size; fx_sw_reserved.xfeatures = xfeatures_mask; fx_sw_reserved.xstate_size = xstate_size; - if (config_enabled(CONFIG_IA32_EMULATION)) { + if (config_enabled(CONFIG_IA32_EMULATION) || + config_enabled(CONFIG_X86_32)) { + int fsave_header_size = sizeof(struct fregs_state); + fx_sw_reserved_ia32 = fx_sw_reserved; - fx_sw_reserved_ia32.extended_size += fsave_header_size; + fx_sw_reserved_ia32.extended_size = size + fsave_header_size; } } From a05917b6ba9dc9a95fc42bdcbe3a875e8ad83935 Mon Sep 17 00:00:00 2001 From: Huaitong Han Date: Fri, 6 Nov 2015 17:00:23 +0800 Subject: [PATCH 042/335] x86/fpu: Fix get_xsave_addr() behavior under virtualization KVM uses the get_xsave_addr() function in a different fashion from the native kernel, in that the 'xsave' parameter belongs to guest vcpu, not the currently running task. But 'xsave' is replaced with current task's (host) xsave structure, so get_xsave_addr() will incorrectly return the bad xsave address to KVM. Fix it so that the passed in 'xsave' address is used - as intended originally. Signed-off-by: Huaitong Han Reviewed-by: Dave Hansen Cc: Cc: Andy Lutomirski Cc: Paolo Bonzini Cc: Borislav Petkov Cc: Fenghua Yu Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Quentin Casasnovas Cc: Thomas Gleixner Cc: dave.hansen@intel.com Link: http://lkml.kernel.org/r/1446800423-21622-1-git-send-email-huaitong.han@intel.com [ Tidied up the changelog. ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/fpu/xstate.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index 6454f2731b56..70fc312221fc 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -694,7 +694,6 @@ void *get_xsave_addr(struct xregs_state *xsave, int xstate_feature) if (!boot_cpu_has(X86_FEATURE_XSAVE)) return NULL; - xsave = ¤t->thread.fpu.state.xsave; /* * We should not ever be requesting features that we * have not enabled. Remember that pcntxt_mask is From f07f21b3e20c11017cea17ec841f0150a62aac53 Mon Sep 17 00:00:00 2001 From: Sascha Silbe Date: Wed, 4 Nov 2015 14:16:57 +0100 Subject: [PATCH 043/335] s390/sclp: _sclp_wait_int(): retain full PSW mask There's no reason to clear all PSW mask bits other than the addressing mode bits. Just use the previous PSW mask as-is. Signed-off-by: Sascha Silbe Acked-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/sclp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/s390/kernel/sclp.c b/arch/s390/kernel/sclp.c index fa0bdff1d413..9fe7781a45cd 100644 --- a/arch/s390/kernel/sclp.c +++ b/arch/s390/kernel/sclp.c @@ -21,7 +21,7 @@ static void _sclp_wait_int(void) __ctl_load(cr0_new, 0, 0); psw_ext_save = S390_lowcore.external_new_psw; - psw_mask = __extract_psw() & (PSW_MASK_EA | PSW_MASK_BA); + psw_mask = __extract_psw(); S390_lowcore.external_new_psw.mask = psw_mask; psw_wait.mask = psw_mask | PSW_MASK_EXT | PSW_MASK_WAIT; S390_lowcore.ext_int_code = 0; From 836d525baada26e1a7462e87805337b9a9b729f6 Mon Sep 17 00:00:00 2001 From: Kamal Mostafa Date: Wed, 11 Nov 2015 14:25:33 -0800 Subject: [PATCH 044/335] tools: Actually install tmon in the install rule Signed-off-by: Kamal Mostafa Cc: Jiri Olsa Cc: Pali Rohar Cc: Roberta Dobrescu Link: http://lkml.kernel.org/r/1447280736-2161-1-git-send-email-kamal@canonical.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/Makefile b/tools/Makefile index d6f307dfb1a3..fa9d0a614d97 100644 --- a/tools/Makefile +++ b/tools/Makefile @@ -101,7 +101,7 @@ freefall_install: install: acpi_install cgroup_install cpupower_install hv_install firewire_install lguest_install \ perf_install selftests_install turbostat_install usb_install \ virtio_install vm_install net_install x86_energy_perf_policy_install \ - tmon freefall_install + tmon_install freefall_install acpi_clean: $(call descend,power/acpi,clean) From 39174291d8e8acfd1113214a943263aaa03c57c8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?fran=C3=A7ois=20romieu?= Date: Wed, 11 Nov 2015 23:35:18 +0100 Subject: [PATCH 045/335] r8169: fix kasan reported skb use-after-free. Signed-off-by: Francois Romieu Reported-by: Dave Jones Fixes: d7d2d89d4b0af ("r8169: Add software counter for multicast packages") Acked-by: Eric Dumazet Acked-by: Corinna Vinschen Signed-off-by: David S. Miller --- drivers/net/ethernet/realtek/r8169.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c index b4f21232019a..79ef799f88ab 100644 --- a/drivers/net/ethernet/realtek/r8169.c +++ b/drivers/net/ethernet/realtek/r8169.c @@ -7429,15 +7429,15 @@ process_pkt: rtl8169_rx_vlan_tag(desc, skb); + if (skb->pkt_type == PACKET_MULTICAST) + dev->stats.multicast++; + napi_gro_receive(&tp->napi, skb); u64_stats_update_begin(&tp->rx_stats.syncp); tp->rx_stats.packets++; tp->rx_stats.bytes += pkt_size; u64_stats_update_end(&tp->rx_stats.syncp); - - if (skb->pkt_type == PACKET_MULTICAST) - dev->stats.multicast++; } release_descriptor: desc->opts2 = 0; From 49e4a2293035b420e807e739999d59c8ec1488e9 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 12 Nov 2015 22:03:40 +0100 Subject: [PATCH 046/335] stmmac: avoid ipq806x constant overflow warning Building dwmac-ipq806x on a 64-bit architecture produces a harmless warning from gcc: stmmac/dwmac-ipq806x.c: In function 'ipq806x_gmac_probe': include/linux/bitops.h:6:19: warning: overflow in implicit constant conversion [-Woverflow] val = QSGMII_PHY_CDR_EN | stmmac/dwmac-ipq806x.c:333:8: note: in expansion of macro 'QSGMII_PHY_CDR_EN' #define QSGMII_PHY_CDR_EN BIT(0) #define BIT(nr) (1UL << (nr)) This is a result of the type conversion rules in C, when we take the logical OR of multiple different types. In particular, we have and unsigned long QSGMII_PHY_CDR_EN == BIT(0) == (1ul << 0) == 0x0000000000000001ul and a signed int 0xC << QSGMII_PHY_TX_DRV_AMP_OFFSET == 0xc0000000 which together gives a signed long value 0xffffffffc0000001l and when this is passed into a function that takes an unsigned int type, gcc warns about the signed overflow and the loss of the upper 32-bits that are all ones. This patch adds 'ul' type modifiers to the literal numbers passed in here, so now the expression remains an 'unsigned long' with the upper bits all zero, and that avoids the signed overflow and the warning. Signed-off-by: Arnd Bergmann Fixes: b1c17215d718 ("stmmac: add ipq806x glue layer") Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c index 9d89bdbf029f..82de68b1a452 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c @@ -337,11 +337,11 @@ static int ipq806x_gmac_probe(struct platform_device *pdev) QSGMII_PHY_RX_SIGNAL_DETECT_EN | QSGMII_PHY_TX_DRIVER_EN | QSGMII_PHY_QSGMII_EN | - 0x4 << QSGMII_PHY_PHASE_LOOP_GAIN_OFFSET | - 0x3 << QSGMII_PHY_RX_DC_BIAS_OFFSET | - 0x1 << QSGMII_PHY_RX_INPUT_EQU_OFFSET | - 0x2 << QSGMII_PHY_CDR_PI_SLEW_OFFSET | - 0xC << QSGMII_PHY_TX_DRV_AMP_OFFSET); + 0x4ul << QSGMII_PHY_PHASE_LOOP_GAIN_OFFSET | + 0x3ul << QSGMII_PHY_RX_DC_BIAS_OFFSET | + 0x1ul << QSGMII_PHY_RX_INPUT_EQU_OFFSET | + 0x2ul << QSGMII_PHY_CDR_PI_SLEW_OFFSET | + 0xCul << QSGMII_PHY_TX_DRV_AMP_OFFSET); } plat_dat->has_gmac = true; From f6ba98c5dc78708cb7fd29950c4a50c4c7e88f95 Mon Sep 17 00:00:00 2001 From: Kamal Mostafa Date: Wed, 11 Nov 2015 14:25:34 -0800 Subject: [PATCH 047/335] tools: Add a "make all" rule Signed-off-by: Kamal Mostafa Acked-by: Pavel Machek Cc: Jiri Olsa Cc: Jonathan Cameron Cc: Pali Rohar Cc: Roberta Dobrescu Link: http://lkml.kernel.org/r/1447280736-2161-2-git-send-email-kamal@canonical.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/Makefile | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/tools/Makefile b/tools/Makefile index fa9d0a614d97..7dc820a8c1f1 100644 --- a/tools/Makefile +++ b/tools/Makefile @@ -32,6 +32,10 @@ help: @echo ' from the kernel command line to build and install one of' @echo ' the tools above' @echo '' + @echo ' $$ make tools/all' + @echo '' + @echo ' builds all tools.' + @echo '' @echo ' $$ make tools/install' @echo '' @echo ' installs all tools.' @@ -77,6 +81,11 @@ tmon: FORCE freefall: FORCE $(call descend,laptop/$@) +all: acpi cgroup cpupower hv firewire lguest \ + perf selftests turbostat usb \ + virtio vm net x86_energy_perf_policy \ + tmon freefall + acpi_install: $(call descend,power/$(@:_install=),install) From 866548dd6e22c3795ae5146a9746a5cf659698f1 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 6 Nov 2015 15:59:29 +0200 Subject: [PATCH 048/335] perf symbols: Rebuild rbtree when adjusting symbols for kcore Normally symbols are read from the DSO and adjusted, if need be, so that the symbol start matches the file offset in the DSO file (we want the file offset because that is what we know from MMAP events). That is done by dso__load_sym() which inserts the symbols *after* adjusting them. In the case of kcore, the symbols have been read from kallsyms and the symbol start is the memory address. The symbols have to be adjusted to match the kcore file offsets. dso__split_kallsyms_for_kcore() does that, but now the adjustment is being done *after* the symbols have been inserted. It appears dso__split_kallsyms_for_kcore() was assuming that changing the symbol start would not change the order in the rbtree - which is, of course, not guaranteed. Signed-off-by: Adrian Hunter Tested-by: Wang Nan Cc: Jiri Olsa Cc: Masami Hiramatsu Cc: Namhyung Kim Cc: Zefan Li Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/563CB241.2090701@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol.c | 30 ++++++++++++++---------------- 1 file changed, 14 insertions(+), 16 deletions(-) diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index b4cc7662677e..09343a880c0b 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -654,19 +654,24 @@ static int dso__split_kallsyms_for_kcore(struct dso *dso, struct map *map, struct map_groups *kmaps = map__kmaps(map); struct map *curr_map; struct symbol *pos; - int count = 0, moved = 0; + int count = 0; + struct rb_root old_root = dso->symbols[map->type]; struct rb_root *root = &dso->symbols[map->type]; struct rb_node *next = rb_first(root); if (!kmaps) return -1; + *root = RB_ROOT; + while (next) { char *module; pos = rb_entry(next, struct symbol, rb_node); next = rb_next(&pos->rb_node); + rb_erase_init(&pos->rb_node, &old_root); + module = strchr(pos->name, '\t'); if (module) *module = '\0'; @@ -674,28 +679,21 @@ static int dso__split_kallsyms_for_kcore(struct dso *dso, struct map *map, curr_map = map_groups__find(kmaps, map->type, pos->start); if (!curr_map || (filter && filter(curr_map, pos))) { - rb_erase_init(&pos->rb_node, root); symbol__delete(pos); - } else { - pos->start -= curr_map->start - curr_map->pgoff; - if (pos->end) - pos->end -= curr_map->start - curr_map->pgoff; - if (curr_map->dso != map->dso) { - rb_erase_init(&pos->rb_node, root); - symbols__insert( - &curr_map->dso->symbols[curr_map->type], - pos); - ++moved; - } else { - ++count; - } + continue; } + + pos->start -= curr_map->start - curr_map->pgoff; + if (pos->end) + pos->end -= curr_map->start - curr_map->pgoff; + symbols__insert(&curr_map->dso->symbols[curr_map->type], pos); + ++count; } /* Symbols have been adjusted */ dso->adjust_symbols = 1; - return count + moved; + return count; } /* From fae00650f7910da7064f45d49c95df88fab816f6 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Nov 2015 15:59:26 -0300 Subject: [PATCH 049/335] perf hists browser: Fix 'd' hotkey action to filter by DSO When pressing 'd' the expected action is to filter all entries by the DSO in the current entry, but for that the action->map needs to be set, and only action->dso was being set, fix it. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Fixes: 045b80dd0340 ("perf hists browser: Use the map to determine if a DSO is being used as a kernel") Link: http://lkml.kernel.org/n/tip-xqhfzgoblq49lk5h5u82atro@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index e5afb8936040..a7ce0cf117e7 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -1875,6 +1875,7 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, continue; case 'd': actions->dso = dso; + actions->ms.map = map; do_zoom_dso(browser, actions); continue; case 'V': From b7f294b58adb02d928dccb04bd39cb789db09e1f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Nov 2015 16:01:56 -0300 Subject: [PATCH 050/335] perf hists browser: The dso can be obtained from popup_action->ms.map->dso So no need to have a 'dso' member in 'popup_action', remove it as no code is using it, already. Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-76a6s0007slug0op0wkl6o8b@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index a7ce0cf117e7..fa9eb92c9e24 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -1430,7 +1430,6 @@ close_file_and_continue: struct popup_action { struct thread *thread; - struct dso *dso; struct map_symbol ms; int socket; @@ -1565,7 +1564,6 @@ add_dso_opt(struct hist_browser *browser, struct popup_action *act, return 0; act->ms.map = map; - act->dso = map->dso; act->fn = do_zoom_dso; return 1; } @@ -1827,7 +1825,6 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, while (1) { struct thread *thread = NULL; - struct dso *dso = NULL; struct map *map = NULL; int choice = 0; int socked_id = -1; @@ -1839,8 +1836,6 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, if (browser->he_selection != NULL) { thread = hist_browser__selected_thread(browser); map = browser->selection->map; - if (map) - dso = map->dso; socked_id = browser->he_selection->socket; } switch (key) { @@ -1874,7 +1869,6 @@ static int perf_evsel__hists_browse(struct perf_evsel *evsel, int nr_events, hist_browser__dump(browser); continue; case 'd': - actions->dso = dso; actions->ms.map = map; do_zoom_dso(browser, actions); continue; From 2059fc7a5a9e667797b8ec503bfb4685afee48d8 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 12 Nov 2015 16:50:13 -0300 Subject: [PATCH 051/335] perf symbols: Allow forcing reading of non-root owned files by root When the root user tries to read a file owned by some other user we get: # ls -la perf.data -rw-------. 1 acme acme 20032 Nov 12 15:50 perf.data # perf report File perf.data not owned by current user or root (use -f to override) # perf report -f | grep -v ^# | head -2 30.96% ls [kernel.vmlinux] [k] do_set_pte 28.24% ls libc-2.20.so [.] intel_check_word # That wasn't happening when the symbol code tried to read a JIT map, where the same check was done but no forcing was possible, fix it. Reported-by: Brendan Gregg Tested-by: Brendan Gregg Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://permalink.gmane.org/gmane.linux.kernel.perf.user/2380 Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-report.c | 6 +++--- tools/perf/util/symbol.c | 4 ++-- tools/perf/util/symbol.h | 1 + 3 files changed, 6 insertions(+), 5 deletions(-) diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 2853ad2bd435..f256fac1e722 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -44,7 +44,7 @@ struct report { struct perf_tool tool; struct perf_session *session; - bool force, use_tui, use_gtk, use_stdio; + bool use_tui, use_gtk, use_stdio; bool hide_unresolved; bool dont_use_callchains; bool show_full_info; @@ -678,7 +678,7 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused) "file", "vmlinux pathname"), OPT_STRING(0, "kallsyms", &symbol_conf.kallsyms_name, "file", "kallsyms pathname"), - OPT_BOOLEAN('f', "force", &report.force, "don't complain, do it"), + OPT_BOOLEAN('f', "force", &symbol_conf.force, "don't complain, do it"), OPT_BOOLEAN('m', "modules", &symbol_conf.use_modules, "load module symbols - WARNING: use only with -k and LIVE kernel"), OPT_BOOLEAN('n', "show-nr-samples", &symbol_conf.show_nr_samples, @@ -832,7 +832,7 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused) } file.path = input_name; - file.force = report.force; + file.force = symbol_conf.force; repeat: session = perf_session__new(&file, false, &report.tool); diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 09343a880c0b..cd08027a6d2c 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1436,9 +1436,9 @@ int dso__load(struct dso *dso, struct map *map, symbol_filter_t filter) if (lstat(dso->name, &st) < 0) goto out; - if (st.st_uid && (st.st_uid != geteuid())) { + if (!symbol_conf.force && st.st_uid && (st.st_uid != geteuid())) { pr_warning("File %s not owned by current user or root, " - "ignoring it.\n", dso->name); + "ignoring it (use -f to override).\n", dso->name); goto out; } diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 40073c60b83d..dcd786e364f2 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -84,6 +84,7 @@ struct symbol_conf { unsigned short priv_size; unsigned short nr_events; bool try_vmlinux_path, + force, ignore_vmlinux, ignore_vmlinux_buildid, show_kernel_path, From 176155dac13f528e0a58c14dc322623219365d91 Mon Sep 17 00:00:00 2001 From: Tadeusz Struk Date: Wed, 21 Oct 2015 14:57:09 -0700 Subject: [PATCH 052/335] crypto: qat - don't use userspace pointer Bugfix - don't dereference userspace pointer. Cc: stable@vger.kernel.org Signed-off-by: Tadeusz Struk Signed-off-by: Herbert Xu --- drivers/crypto/qat/qat_common/adf_ctl_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/crypto/qat/qat_common/adf_ctl_drv.c b/drivers/crypto/qat/qat_common/adf_ctl_drv.c index 03856ad280b9..473d36d91644 100644 --- a/drivers/crypto/qat/qat_common/adf_ctl_drv.c +++ b/drivers/crypto/qat/qat_common/adf_ctl_drv.c @@ -198,7 +198,7 @@ static int adf_copy_key_value_data(struct adf_accel_dev *accel_dev, goto out_err; } - params_head = section_head->params; + params_head = section.params; while (params_head) { if (copy_from_user(&key_val, (void __user *)params_head, From ff38e82931aeaa339d6bf4a9980e011cbfd14927 Mon Sep 17 00:00:00 2001 From: Ping Cheng Date: Thu, 12 Nov 2015 17:21:14 -0800 Subject: [PATCH 053/335] HID: wacom: Add outbounding area for DTU1141 DTU1141 has an extra 1 mm tablet active outbounding area on each side of the display, same as those recently released DTUs. Signed-off-by: Ping Cheng Signed-off-by: Jiri Kosina --- drivers/hid/wacom_wac.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c index 8b29949507d1..0ff22dbe334c 100644 --- a/drivers/hid/wacom_wac.c +++ b/drivers/hid/wacom_wac.c @@ -3213,7 +3213,8 @@ static const struct wacom_features wacom_features_0x32F = WACOM_DTU_OFFSET, WACOM_DTU_OFFSET }; static const struct wacom_features wacom_features_0x336 = { "Wacom DTU1141", 23472, 13203, 1023, 0, - DTUS, WACOM_INTUOS_RES, WACOM_INTUOS_RES, 4 }; + DTUS, WACOM_INTUOS_RES, WACOM_INTUOS_RES, 4, + WACOM_DTU_OFFSET, WACOM_DTU_OFFSET }; static const struct wacom_features wacom_features_0x57 = { "Wacom DTK2241", 95640, 54060, 2047, 63, DTK, WACOM_INTUOS3_RES, WACOM_INTUOS3_RES, 6, From e266a753bf51b2c3b46d0d230349662c35ac5629 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 13 Nov 2015 11:48:30 +0200 Subject: [PATCH 054/335] perf symbols: Fix dso lookup by long name and missing buildids Commit 4598a0a6d22f ("perf symbols: Improve DSO long names lookup speed with rbtree") Added a tree to lookup dsos by long name. That tree gets corrupted whenever a dso long name is changed because the tree is not updated. One effect of that is buildid-list does not work with the 'with-hits' option because dso lookup fails and results in two structs for the same dso. The first has the buildid but no hits, the second has hits but no buildid. e.g. Before: $ tools/perf/perf record ls arch certs CREDITS Documentation firmware include ipc Kconfig lib Makefile net REPORTING-BUGS scripts sound usr block COPYING crypto drivers fs init Kbuild kernel MAINTAINERS mm README samples security tools virt [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.012 MB perf.data (11 samples) ] $ tools/perf/perf buildid-list 574da826c66538a8d9060d393a8866289bd06005 [kernel.kallsyms] 30c94dc66a1fe95180c3d68d2b89e576d5ae213c /lib/x86_64-linux-gnu/libc-2.19.so $ tools/perf/perf buildid-list -H 574da826c66538a8d9060d393a8866289bd06005 [kernel.kallsyms] 0000000000000000000000000000000000000000 /lib/x86_64-linux-gnu/libc-2.19.so After: $ tools/perf/perf buildid-list -H 574da826c66538a8d9060d393a8866289bd06005 [kernel.kallsyms] 30c94dc66a1fe95180c3d68d2b89e576d5ae213c /lib/x86_64-linux-gnu/libc-2.19.so The fix is to record the root of the tree on the dso so that dso__set_long_name() can update the tree when the long name changes. Signed-off-by: Adrian Hunter Tested-by: Arnaldo Carvalho de Melo Cc: Adrian Hunter Cc: Don Zickus Cc: Douglas Hatch Cc: Jiri Olsa Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Scott J Norton Cc: Waiman Long Fixes: 4598a0a6d22f ("perf symbols: Improve DSO long names lookup speed with rbtree") Link: http://lkml.kernel.org/r/1447408112-1920-2-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/dso.c | 17 +++++++++++++++++ tools/perf/util/dso.h | 1 + tools/perf/util/machine.c | 1 + 3 files changed, 19 insertions(+) diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c index 7c0c08386a1d..425df5c86c9c 100644 --- a/tools/perf/util/dso.c +++ b/tools/perf/util/dso.c @@ -933,6 +933,7 @@ static struct dso *__dso__findlink_by_longname(struct rb_root *root, /* Add new node and rebalance tree */ rb_link_node(&dso->rb_node, parent, p); rb_insert_color(&dso->rb_node, root); + dso->root = root; } return NULL; } @@ -945,15 +946,30 @@ static inline struct dso *__dso__find_by_longname(struct rb_root *root, void dso__set_long_name(struct dso *dso, const char *name, bool name_allocated) { + struct rb_root *root = dso->root; + if (name == NULL) return; if (dso->long_name_allocated) free((char *)dso->long_name); + if (root) { + rb_erase(&dso->rb_node, root); + /* + * __dso__findlink_by_longname() isn't guaranteed to add it + * back, so a clean removal is required here. + */ + RB_CLEAR_NODE(&dso->rb_node); + dso->root = NULL; + } + dso->long_name = name; dso->long_name_len = strlen(name); dso->long_name_allocated = name_allocated; + + if (root) + __dso__findlink_by_longname(root, dso, NULL); } void dso__set_short_name(struct dso *dso, const char *name, bool name_allocated) @@ -1046,6 +1062,7 @@ struct dso *dso__new(const char *name) dso->kernel = DSO_TYPE_USER; dso->needs_swap = DSO_SWAP__UNSET; RB_CLEAR_NODE(&dso->rb_node); + dso->root = NULL; INIT_LIST_HEAD(&dso->node); INIT_LIST_HEAD(&dso->data.open_entry); pthread_mutex_init(&dso->lock, NULL); diff --git a/tools/perf/util/dso.h b/tools/perf/util/dso.h index fc8db9c764ac..45ec4d0a50ed 100644 --- a/tools/perf/util/dso.h +++ b/tools/perf/util/dso.h @@ -135,6 +135,7 @@ struct dso { pthread_mutex_t lock; struct list_head node; struct rb_node rb_node; /* rbtree node sorted by long name */ + struct rb_root *root; /* root of rbtree that rb_node is in */ struct rb_root symbols[MAP__NR_TYPES]; struct rb_root symbol_names[MAP__NR_TYPES]; struct { diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 5ef90be2a249..8b303ff20289 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -91,6 +91,7 @@ static void dsos__purge(struct dsos *dsos) list_for_each_entry_safe(pos, n, &dsos->head, node) { RB_CLEAR_NODE(&pos->rb_node); + pos->root = NULL; list_del_init(&pos->node); dso__put(pos); } From 1216b65c502e0f130cc9984dfd5f9e1968c1eb46 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 13 Nov 2015 11:48:31 +0200 Subject: [PATCH 055/335] perf buildid-list: Requires ordered events 'perf buildid-list' processes events to determine hits (i.e. with-hits option). That may not work if events are not sorted in order. i.e. MMAP events must be processed before the samples that depend on them so that sample processing can 'hit' the DSO to which the MMAP refers. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Link: http://lkml.kernel.org/r/1447408112-1920-3-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/build-id.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c index d909459fb54c..217b5a60e2ab 100644 --- a/tools/perf/util/build-id.c +++ b/tools/perf/util/build-id.c @@ -76,6 +76,7 @@ struct perf_tool build_id__mark_dso_hit_ops = { .exit = perf_event__exit_del_thread, .attr = perf_event__process_attr, .build_id = perf_event__process_build_id, + .ordered_events = true, }; int build_id__sprintf(const u8 *build_id, int len, char *bf) From d8145b3e30a24280c396d88c8703c50a1ea0aa3a Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 13 Nov 2015 11:48:32 +0200 Subject: [PATCH 056/335] perf inject: Also re-pipe lost_samples event perf inject must re-pipe all events otherwise they get dropped from the output file. Signed-off-by: Adrian Hunter Cc: Jiri Olsa Link: http://lkml.kernel.org/r/1447408112-1920-4-git-send-email-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-inject.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/builtin-inject.c b/tools/perf/builtin-inject.c index 0a945d2e8ca5..99d127fe9c35 100644 --- a/tools/perf/builtin-inject.c +++ b/tools/perf/builtin-inject.c @@ -675,6 +675,7 @@ int cmd_inject(int argc, const char **argv, const char *prefix __maybe_unused) .fork = perf_event__repipe, .exit = perf_event__repipe, .lost = perf_event__repipe, + .lost_samples = perf_event__repipe, .aux = perf_event__repipe, .itrace_start = perf_event__repipe, .context_switch = perf_event__repipe, From 0196e787ceb58cdfea822482ec70019bc16cbd51 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 13 Nov 2015 12:29:10 +0000 Subject: [PATCH 057/335] perf probe: Fix memory leaking on failure by clearing all probe_trace_events Fix memory leaking on the debuginfo__find_trace_events() failure path which frees an array of probe_trace_events but doesn't clears all the allocated sub-structures and strings. So, before doing zfree(tevs), clear all the array elements which may have allocated resources. Reported-by: Wang Nan Signed-off-by: Masami Hiramatsu Cc: Alexei Starovoitov Cc: Zefan Li Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447417761-156094-2-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/probe-finder.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index bd8f03de5e40..63993d7e0fac 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -1246,7 +1246,7 @@ int debuginfo__find_trace_events(struct debuginfo *dbg, struct trace_event_finder tf = { .pf = {.pev = pev, .callback = add_probe_trace_event}, .max_tevs = probe_conf.max_probes, .mod = dbg->mod}; - int ret; + int ret, i; /* Allocate result tevs array */ *tevs = zalloc(sizeof(struct probe_trace_event) * tf.max_tevs); @@ -1258,6 +1258,8 @@ int debuginfo__find_trace_events(struct debuginfo *dbg, ret = debuginfo__find_probes(dbg, &tf.pf); if (ret < 0) { + for (i = 0; i < tf.ntevs; i++) + clear_probe_trace_event(&tf.tevs[i]); zfree(tevs); return ret; } From 092b1f0b5f9f797812da0de927c3aa26acbe8762 Mon Sep 17 00:00:00 2001 From: Wang Nan Date: Fri, 13 Nov 2015 12:29:11 +0000 Subject: [PATCH 058/335] perf probe: Clear probe_trace_event when add_probe_trace_event() fails When probing with a glob, errors in add_probe_trace_event() won't be passed to debuginfo__find_trace_events() because it would be modified by probe_point_search_cb(). It causes a segfault if perf fails to find an argument for a probe point matched by the glob. For example: # ./perf probe -v -n 'SyS_dup? oldfd' probe-definition(0): SyS_dup? oldfd symbol:SyS_dup? file:(null) line:0 offset:0 return:0 lazy:(null) parsing arg: oldfd into oldfd 1 arguments Looking at the vmlinux_path (7 entries long) Using /lib/modules/4.3.0-rc4+/build/vmlinux for symbols Open Debuginfo file: /lib/modules/4.3.0-rc4+/build/vmlinux Try to find probe point from debuginfo. Matched function: SyS_dup3 found inline addr: 0xffffffff812095c0 Probe point found: SyS_dup3+0 Searching 'oldfd' variable in context. Converting variable oldfd into trace event. oldfd type is long int. found inline addr: 0xffffffff812096d4 Probe point found: SyS_dup2+36 Searching 'oldfd' variable in context. Failed to find 'oldfd' in this function. Matched function: SyS_dup3 Probe point found: SyS_dup3+0 Searching 'oldfd' variable in context. Converting variable oldfd into trace event. oldfd type is long int. Matched function: SyS_dup2 Probe point found: SyS_dup2+0 Searching 'oldfd' variable in context. Converting variable oldfd into trace event. oldfd type is long int. Found 4 probe_trace_events. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:probe/SyS_dup3 _text+2135488 oldfd=%di:s64 Segmentation fault (core dumped) # This patch ensures that add_probe_trace_event() doesn't touches tf->ntevs and tf->tevs if those functions fail. After the patch: # perf probe 'SyS_dup? oldfd' Failed to find 'oldfd' in this function. Added new events: probe:SyS_dup3 (on SyS_dup? with oldfd) probe:SyS_dup3_1 (on SyS_dup? with oldfd) probe:SyS_dup2 (on SyS_dup? with oldfd) You can now use it in all perf tools, such as: perf record -e probe:SyS_dup2 -aR sleep 1 Signed-off-by: Wang Nan Tested-by: Arnaldo Carvalho de Melo Cc: Alexei Starovoitov Cc: Masami Hiramatsu Cc: Zefan Li Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447417761-156094-3-git-send-email-wangnan0@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/probe-finder.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 63993d7e0fac..05012bb178d7 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -1183,7 +1183,7 @@ static int add_probe_trace_event(Dwarf_Die *sc_die, struct probe_finder *pf) container_of(pf, struct trace_event_finder, pf); struct perf_probe_point *pp = &pf->pev->point; struct probe_trace_event *tev; - struct perf_probe_arg *args; + struct perf_probe_arg *args = NULL; int ret, i; /* Check number of tevs */ @@ -1198,19 +1198,23 @@ static int add_probe_trace_event(Dwarf_Die *sc_die, struct probe_finder *pf) ret = convert_to_trace_point(&pf->sp_die, tf->mod, pf->addr, pp->retprobe, pp->function, &tev->point); if (ret < 0) - return ret; + goto end; tev->point.realname = strdup(dwarf_diename(sc_die)); - if (!tev->point.realname) - return -ENOMEM; + if (!tev->point.realname) { + ret = -ENOMEM; + goto end; + } pr_debug("Probe point found: %s+%lu\n", tev->point.symbol, tev->point.offset); /* Expand special probe argument if exist */ args = zalloc(sizeof(struct perf_probe_arg) * MAX_PROBE_ARGS); - if (args == NULL) - return -ENOMEM; + if (args == NULL) { + ret = -ENOMEM; + goto end; + } ret = expand_probe_args(sc_die, pf, args); if (ret < 0) @@ -1234,6 +1238,10 @@ static int add_probe_trace_event(Dwarf_Die *sc_die, struct probe_finder *pf) } end: + if (ret) { + clear_probe_trace_event(tev); + tf->ntevs--; + } free(args); return ret; } From ec0adc539b8bf59b7c00db0748671f6594b77843 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Thu, 12 Nov 2015 02:42:31 -0500 Subject: [PATCH 059/335] tools/power turbostat: use new name for MSR_PLATFORM_INFO MSR_PLATFORM_INFO is the new name for MSR_NHM_PLATFORM_INFO no functional change Signed-off-by: Len Brown Signed-off-by: Rafael J. Wysocki --- tools/power/x86/turbostat/turbostat.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index d8e4b20b6d54..0dac7e05a6ac 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -1173,9 +1173,9 @@ dump_nhm_platform_info(void) unsigned long long msr; unsigned int ratio; - get_msr(base_cpu, MSR_NHM_PLATFORM_INFO, &msr); + get_msr(base_cpu, MSR_PLATFORM_INFO, &msr); - fprintf(stderr, "cpu%d: MSR_NHM_PLATFORM_INFO: 0x%08llx\n", base_cpu, msr); + fprintf(stderr, "cpu%d: MSR_PLATFORM_INFO: 0x%08llx\n", base_cpu, msr); ratio = (msr >> 40) & 0xFF; fprintf(stderr, "%d * %.0f = %.0f MHz max efficiency frequency\n", @@ -1807,7 +1807,7 @@ void check_permissions() * * MSR_SMI_COUNT 0x00000034 * - * MSR_NHM_PLATFORM_INFO 0x000000ce + * MSR_PLATFORM_INFO 0x000000ce * MSR_NHM_SNB_PKG_CST_CFG_CTL 0x000000e2 * * MSR_PKG_C3_RESIDENCY 0x000003f8 @@ -1876,7 +1876,7 @@ int probe_nhm_msrs(unsigned int family, unsigned int model) get_msr(base_cpu, MSR_NHM_SNB_PKG_CST_CFG_CTL, &msr); pkg_cstate_limit = pkg_cstate_limits[msr & 0xF]; - get_msr(base_cpu, MSR_NHM_PLATFORM_INFO, &msr); + get_msr(base_cpu, MSR_PLATFORM_INFO, &msr); base_ratio = (msr >> 8) & 0xFF; base_hz = base_ratio * bclk * 1000000; From 5369a21e3f26ef9d2bf6ea1b322d6899a4ed08e0 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Thu, 12 Nov 2015 02:42:32 -0500 Subject: [PATCH 060/335] x86: remove unused definition of MSR_NHM_PLATFORM_INFO MSR_NHM_PLATFORM_INFO has been replaced by... MSR_PLATFORM_INFO Signed-off-by: Len Brown Signed-off-by: Rafael J. Wysocki --- arch/x86/include/asm/msr-index.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index b8c14bb7fc8f..705c4082b7f1 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -35,7 +35,7 @@ #define MSR_IA32_PERFCTR0 0x000000c1 #define MSR_IA32_PERFCTR1 0x000000c2 #define MSR_FSB_FREQ 0x000000cd -#define MSR_NHM_PLATFORM_INFO 0x000000ce +#define MSR_PLATFORM_INFO 0x000000ce #define MSR_NHM_SNB_PKG_CST_CFG_CTL 0x000000e2 #define NHM_C3_AUTO_DEMOTE (1UL << 25) @@ -44,7 +44,6 @@ #define SNB_C1_AUTO_UNDEMOTE (1UL << 27) #define SNB_C3_AUTO_UNDEMOTE (1UL << 28) -#define MSR_PLATFORM_INFO 0x000000ce #define MSR_MTRRcap 0x000000fe #define MSR_IA32_BBL_CR_CTL 0x00000119 #define MSR_IA32_BBL_CR_CTL3 0x0000011e From d29d67357db8dad08f97162bbeb51e4975ee67fd Mon Sep 17 00:00:00 2001 From: Ashwin Chaugule Date: Thu, 12 Nov 2015 19:52:30 -0500 Subject: [PATCH 061/335] ACPI / CPPC: Use h/w reduced version of the PCCT structure CPPC is enabled only on platforms which support the h/w reduced ACPI specification, so use the h/w reduced version of the PCCT consistently when deferencing PCCT contents. Fixes: 337aadff8e45 (ACPI: Introduce CPU performance controls using CPPC) Signed-off-by: Ashwin Chaugule Signed-off-by: Rafael J. Wysocki --- drivers/acpi/cppc_acpi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c index 0bbf84bdcdc4..9361966f1b61 100644 --- a/drivers/acpi/cppc_acpi.c +++ b/drivers/acpi/cppc_acpi.c @@ -304,7 +304,7 @@ EXPORT_SYMBOL_GPL(acpi_get_psd_map); static int register_pcc_channel(unsigned pcc_subspace_idx) { - struct acpi_pcct_subspace *cppc_ss; + struct acpi_pcct_hw_reduced *cppc_ss; unsigned int len; if (pcc_subspace_idx >= 0) { From 0d2f6fd3ffd5e7a447233a57112246bc00064752 Mon Sep 17 00:00:00 2001 From: Chris Lesiak Date: Mon, 2 Nov 2015 15:45:03 -0600 Subject: [PATCH 062/335] iio: si7020: Swap data byte order The Silicon Labs Si7013, Si7020, and Si7021 family of I2C humidity and temperature sensors deliver 16 bit data high byte first. See the datasheet available at: https://www.silabs.com/Support%20Documents%2fTechnicalDocs%2fSi7020-A20.pdf But as documented in Documentation/i2c/smbus-protocol, i2c_smbus_read_word_data() expects the low byte first. Change the driver to use i2c_smbus_read_word_swapped to get correct byte order. Signed-off-by: Chris Lesiak Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/humidity/si7020.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/iio/humidity/si7020.c b/drivers/iio/humidity/si7020.c index fa3b809aff5e..1b6935d99100 100644 --- a/drivers/iio/humidity/si7020.c +++ b/drivers/iio/humidity/si7020.c @@ -50,10 +50,10 @@ static int si7020_read_raw(struct iio_dev *indio_dev, switch (mask) { case IIO_CHAN_INFO_RAW: - ret = i2c_smbus_read_word_data(*client, - chan->type == IIO_TEMP ? - SI7020CMD_TEMP_HOLD : - SI7020CMD_RH_HOLD); + ret = i2c_smbus_read_word_swapped(*client, + chan->type == IIO_TEMP ? + SI7020CMD_TEMP_HOLD : + SI7020CMD_RH_HOLD); if (ret < 0) return ret; *val = ret >> 2; From a57f8dac46fbac5ab0e31aef1a98b3f6eb30c079 Mon Sep 17 00:00:00 2001 From: Thomas Betker Date: Wed, 11 Nov 2015 21:24:38 +0100 Subject: [PATCH 063/335] iio: adc: xilinx: Fix VREFN scale The scaling factor for VREFN is 3.0/4096 (not 1.0/4096), just as for VREFP. This is not immediately obvious from the specification (Xilinx UG480), but has been confirmed by Xilinx support. Suggested-by: Hartmut Knaack Signed-off-by: Thomas Betker Cc: Signed-off-by: Jonathan Cameron --- drivers/iio/adc/xilinx-xadc-core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/iio/adc/xilinx-xadc-core.c b/drivers/iio/adc/xilinx-xadc-core.c index ce93bd8e3f68..95c0b3d9890c 100644 --- a/drivers/iio/adc/xilinx-xadc-core.c +++ b/drivers/iio/adc/xilinx-xadc-core.c @@ -857,6 +857,7 @@ static int xadc_read_raw(struct iio_dev *indio_dev, case XADC_REG_VCCINT: case XADC_REG_VCCAUX: case XADC_REG_VREFP: + case XADC_REG_VREFN: case XADC_REG_VCCBRAM: case XADC_REG_VCCPINT: case XADC_REG_VCCPAUX: From 73ed5d25dce0354ea381d6dc93005c3085fae03d Mon Sep 17 00:00:00 2001 From: Hannes Frederic Sowa Date: Tue, 10 Nov 2015 16:23:15 +0100 Subject: [PATCH 064/335] af-unix: fix use-after-free with concurrent readers while splicing During splicing an af-unix socket to a pipe we have to drop all af-unix socket locks. While doing so we allow another reader to enter unix_stream_read_generic which can read, copy and finally free another skb. If exactly this skb is just in process of being spliced we get a use-after-free report by kasan. First, we must make sure to not have a free while the skb is used during the splice operation. We simply increment its use counter before unlocking the reader lock. Stream sockets have the nice characteristic that we don't care about zero length writes and they never reach the peer socket's queue. That said, we can take the UNIXCB.consumed field as the indicator if the skb was already freed from the socket's receive queue. If the skb was fully consumed after we locked the reader side again we know it has been dropped by a second reader. We indicate a short read to user space and abort the current splice operation. This bug has been found with syzkaller (http://github.com/google/syzkaller) by Dmitry Vyukov. Fixes: 2b514574f7e8 ("net: af_unix: implement splice for stream af_unix sockets") Reported-by: Dmitry Vyukov Cc: Dmitry Vyukov Cc: Eric Dumazet Acked-by: Eric Dumazet Signed-off-by: Hannes Frederic Sowa Signed-off-by: David S. Miller --- net/unix/af_unix.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index aaa0b58d6aba..12b886f07982 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -441,6 +441,7 @@ static void unix_release_sock(struct sock *sk, int embrion) if (state == TCP_LISTEN) unix_release_sock(skb->sk, 1); /* passed fds are erased in the kfree_skb hook */ + UNIXCB(skb).consumed = skb->len; kfree_skb(skb); } @@ -2072,6 +2073,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state) do { int chunk; + bool drop_skb; struct sk_buff *skb, *last; unix_state_lock(sk); @@ -2152,7 +2154,11 @@ unlock: } chunk = min_t(unsigned int, unix_skb_len(skb) - skip, size); + skb_get(skb); chunk = state->recv_actor(skb, skip, chunk, state); + drop_skb = !unix_skb_len(skb); + /* skb is only safe to use if !drop_skb */ + consume_skb(skb); if (chunk < 0) { if (copied == 0) copied = -EFAULT; @@ -2161,6 +2167,18 @@ unlock: copied += chunk; size -= chunk; + if (drop_skb) { + /* the skb was touched by a concurrent reader; + * we should not expect anything from this skb + * anymore and assume it invalid - we can be + * sure it was dropped from the socket queue + * + * let's report a short read + */ + err = 0; + break; + } + /* Mark read part of skb as used */ if (!(flags & MSG_PEEK)) { UNIXCB(skb).consumed += chunk; From 9001d94dfd558f175bede9f847276f691cd9c164 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Wed, 11 Nov 2015 15:22:40 +0000 Subject: [PATCH 065/335] fjes: fix inconsistent indenting minor change, indenting is one tab out. Signed-off-by: Colin Ian King Signed-off-by: Taku Izumi Signed-off-by: David S. Miller --- drivers/net/fjes/fjes_hw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/fjes/fjes_hw.c b/drivers/net/fjes/fjes_hw.c index bb8b5304d851..b103adb8d62e 100644 --- a/drivers/net/fjes/fjes_hw.c +++ b/drivers/net/fjes/fjes_hw.c @@ -599,7 +599,7 @@ int fjes_hw_unregister_buff_addr(struct fjes_hw *hw, int dest_epid) FJES_CMD_REQ_RES_CODE_BUSY) && (timeout > 0)) { msleep(200 + hw->my_epid * 20); - timeout -= (200 + hw->my_epid * 20); + timeout -= (200 + hw->my_epid * 20); res_buf->unshare_buffer.length = 0; res_buf->unshare_buffer.code = 0; From 0d3f6d297bfb7af24d0508460fdb3d1ec4903fa3 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Wed, 11 Nov 2015 11:51:06 -0800 Subject: [PATCH 066/335] ipv6: Avoid creating RTF_CACHE from a rt that is not managed by fib6 tree The original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1272571 The setup has a IPv4 GRE tunnel running in a IPSec. The bug happens when ndisc starts sending router solicitation at the gre interface. The simplified oops stack is like: __lock_acquire+0x1b2/0x1c30 lock_acquire+0xb9/0x140 _raw_write_lock_bh+0x3f/0x50 __ip6_ins_rt+0x2e/0x60 ip6_ins_rt+0x49/0x50 ~~~~~~~~ __ip6_rt_update_pmtu.part.54+0x145/0x250 ip6_rt_update_pmtu+0x2e/0x40 ~~~~~~~~ ip_tunnel_xmit+0x1f1/0xf40 __gre_xmit+0x7a/0x90 ipgre_xmit+0x15a/0x220 dev_hard_start_xmit+0x2bd/0x480 __dev_queue_xmit+0x696/0x730 dev_queue_xmit+0x10/0x20 neigh_direct_output+0x11/0x20 ip6_finish_output2+0x21f/0x770 ip6_finish_output+0xa7/0x1d0 ip6_output+0x56/0x190 ~~~~~~~~ ndisc_send_skb+0x1d9/0x400 ndisc_send_rs+0x88/0xc0 ~~~~~~~~ The rt passed to ip6_rt_update_pmtu() is created by icmp6_dst_alloc() and it is not managed by the fib6 tree, so its rt6i_table == NULL. When __ip6_rt_update_pmtu() creates a RTF_CACHE clone, the newly created clone also has rt6i_table == NULL and it causes the ip6_ins_rt() oops. During pmtu update, we only want to create a RTF_CACHE clone from a rt which is currently managed (or owned) by the fib6 tree. It means either rt->rt6i_node != NULL or rt is a RTF_PCPU clone. It is worth to note that rt6i_table may not be NULL even it is not (yet) managed by the fib6 tree (e.g. addrconf_dst_alloc()). Hence, rt6i_node is a better check instead of rt6i_table. Fixes: 45e4fd26683c ("ipv6: Only create RTF_CACHE routes after encountering pmtu") Signed-off-by: Martin KaFai Lau Reported-by: Chris Siebenmann Cc: Chris Siebenmann Cc: Hannes Frederic Sowa Signed-off-by: David S. Miller --- net/ipv6/route.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index c8bc9b4ac328..74907c58a5c7 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1322,6 +1322,12 @@ static void rt6_do_update_pmtu(struct rt6_info *rt, u32 mtu) rt6_update_expires(rt, net->ipv6.sysctl.ip6_rt_mtu_expires); } +static bool rt6_cache_allowed_for_pmtu(const struct rt6_info *rt) +{ + return !(rt->rt6i_flags & RTF_CACHE) && + (rt->rt6i_flags & RTF_PCPU || rt->rt6i_node); +} + static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk, const struct ipv6hdr *iph, u32 mtu) { @@ -1335,7 +1341,7 @@ static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk, if (mtu >= dst_mtu(dst)) return; - if (rt6->rt6i_flags & RTF_CACHE) { + if (!rt6_cache_allowed_for_pmtu(rt6)) { rt6_do_update_pmtu(rt6, mtu); } else { const struct in6_addr *daddr, *saddr; From 5973fb1e245086071bf71994c8b54d99526ded03 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Wed, 11 Nov 2015 11:51:07 -0800 Subject: [PATCH 067/335] ipv6: Check expire on DST_NOCACHE route Since the expires of the DST_NOCACHE rt can be set during the ip6_rt_update_pmtu(), we also need to consider the expires value when doing ip6_dst_check(). This patches creates __rt6_check_expired() to only check the expire value (if one exists) of the current rt. In rt6_dst_from_check(), it adds __rt6_check_expired() as one of the condition check. Signed-off-by: Martin KaFai Lau Cc: Hannes Frederic Sowa Signed-off-by: David S. Miller --- net/ipv6/route.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 74907c58a5c7..3754cf9287a0 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -404,6 +404,14 @@ static void ip6_dst_ifdown(struct dst_entry *dst, struct net_device *dev, } } +static bool __rt6_check_expired(const struct rt6_info *rt) +{ + if (rt->rt6i_flags & RTF_EXPIRES) + return time_after(jiffies, rt->dst.expires); + else + return false; +} + static bool rt6_check_expired(const struct rt6_info *rt) { if (rt->rt6i_flags & RTF_EXPIRES) { @@ -1252,7 +1260,8 @@ static struct dst_entry *rt6_check(struct rt6_info *rt, u32 cookie) static struct dst_entry *rt6_dst_from_check(struct rt6_info *rt, u32 cookie) { - if (rt->dst.obsolete == DST_OBSOLETE_FORCE_CHK && + if (!__rt6_check_expired(rt) && + rt->dst.obsolete == DST_OBSOLETE_FORCE_CHK && rt6_check((struct rt6_info *)(rt->dst.from), cookie)) return &rt->dst; else From 02bcf4e082e4dc634409a6a6cb7def8806d6e5e6 Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Wed, 11 Nov 2015 11:51:08 -0800 Subject: [PATCH 068/335] ipv6: Check rt->dst.from for the DST_NOCACHE route All DST_NOCACHE rt6_info used to have rt->dst.from set to its parent. After commit 8e3d5be73681 ("ipv6: Avoid double dst_free"), DST_NOCACHE is also set to rt6_info which does not have a parent (i.e. rt->dst.from is NULL). This patch catches the rt->dst.from == NULL case. Fixes: 8e3d5be73681 ("ipv6: Avoid double dst_free") Signed-off-by: Martin KaFai Lau Cc: Hannes Frederic Sowa Signed-off-by: David S. Miller --- include/net/ip6_fib.h | 3 ++- net/ipv6/route.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h index aaf9700fc9e5..fb961a576abe 100644 --- a/include/net/ip6_fib.h +++ b/include/net/ip6_fib.h @@ -167,7 +167,8 @@ static inline void rt6_update_expires(struct rt6_info *rt0, int timeout) static inline u32 rt6_get_cookie(const struct rt6_info *rt) { - if (rt->rt6i_flags & RTF_PCPU || unlikely(rt->dst.flags & DST_NOCACHE)) + if (rt->rt6i_flags & RTF_PCPU || + (unlikely(rt->dst.flags & DST_NOCACHE) && rt->dst.from)) rt = (struct rt6_info *)(rt->dst.from); return rt->rt6i_node ? rt->rt6i_node->fn_sernum : 0; diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 3754cf9287a0..6f01fe122abd 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1281,7 +1281,8 @@ static struct dst_entry *ip6_dst_check(struct dst_entry *dst, u32 cookie) rt6_dst_from_metrics_check(rt); - if ((rt->rt6i_flags & RTF_PCPU) || unlikely(dst->flags & DST_NOCACHE)) + if (rt->rt6i_flags & RTF_PCPU || + (unlikely(dst->flags & DST_NOCACHE) && rt->dst.from)) return rt6_dst_from_check(rt, cookie); else return rt6_check(rt, cookie); From d7475de58575c904818efa369c82e88c6648ce2e Mon Sep 17 00:00:00 2001 From: Kamal Mostafa Date: Wed, 11 Nov 2015 14:24:27 -0800 Subject: [PATCH 069/335] tools/net: Use include/uapi with __EXPORTED_HEADERS__ Use the local uapi headers to keep in sync with "recently" added #define's (e.g. SKF_AD_VLAN_TPID). Refactored CFLAGS, and bpf_asm doesn't need -I. Fixes: 3f356385e8a4 ("filter: bpf_asm: add minimal bpf asm tool") Signed-off-by: Kamal Mostafa Acked-by: Daniel Borkmann Signed-off-by: David S. Miller --- tools/net/Makefile | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/tools/net/Makefile b/tools/net/Makefile index ee577ea03ba5..ddf888010652 100644 --- a/tools/net/Makefile +++ b/tools/net/Makefile @@ -4,6 +4,9 @@ CC = gcc LEX = flex YACC = bison +CFLAGS += -Wall -O2 +CFLAGS += -D__EXPORTED_HEADERS__ -I../../include/uapi -I../../include + %.yacc.c: %.y $(YACC) -o $@ -d $< @@ -12,15 +15,13 @@ YACC = bison all : bpf_jit_disasm bpf_dbg bpf_asm -bpf_jit_disasm : CFLAGS = -Wall -O2 -DPACKAGE='bpf_jit_disasm' +bpf_jit_disasm : CFLAGS += -DPACKAGE='bpf_jit_disasm' bpf_jit_disasm : LDLIBS = -lopcodes -lbfd -ldl bpf_jit_disasm : bpf_jit_disasm.o -bpf_dbg : CFLAGS = -Wall -O2 bpf_dbg : LDLIBS = -lreadline bpf_dbg : bpf_dbg.o -bpf_asm : CFLAGS = -Wall -O2 -I. bpf_asm : LDLIBS = bpf_asm : bpf_asm.o bpf_exp.yacc.o bpf_exp.lex.o bpf_exp.lex.o : bpf_exp.yacc.c From efdfa2f7848f64517008136fb41f53c4a1faf93a Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 11 Nov 2015 23:25:40 +0100 Subject: [PATCH 070/335] packet: do skb_probe_transport_header when we actually have data In tpacket_fill_skb() commit c1aad275b029 ("packet: set transport header before doing xmit") and later on 40893fd0fd4e ("net: switch to use skb_probe_transport_header()") was probing for a transport header on the skb from a ring buffer slot, but at a time, where the skb has _not even_ been filled with data yet. So that call into the flow dissector is pretty useless. Lets do it after we've set up the skb frags. Fixes: c1aad275b029 ("packet: set transport header before doing xmit") Reported-by: Eric Dumazet Signed-off-by: Daniel Borkmann Acked-by: Jason Wang Signed-off-by: David S. Miller --- net/packet/af_packet.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index af399cac5205..80c36c0867d3 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2368,8 +2368,6 @@ static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, skb_reserve(skb, hlen); skb_reset_network_header(skb); - if (!packet_use_direct_xmit(po)) - skb_probe_transport_header(skb, 0); if (unlikely(po->tp_tx_has_off)) { int off_min, off_max, off; off_min = po->tp_hdrlen - sizeof(struct sockaddr_ll); @@ -2449,6 +2447,9 @@ static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, len = ((to_write > len_max) ? len_max : to_write); } + if (!packet_use_direct_xmit(po)) + skb_probe_transport_header(skb, 0); + return tp_len; } From 8fd6c80d9dd938ca338c70698533a7e304752846 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 11 Nov 2015 23:25:41 +0100 Subject: [PATCH 071/335] packet: always probe for transport header We concluded that the skb_probe_transport_header() should better be called unconditionally. Avoiding the call into the flow dissector has also not really much to do with the direct xmit mode. While it seems that only virtio_net code makes use of GSO from non RX/TX ring packet socket paths, we should probe for a transport header nevertheless before they hit devices. Reference: http://thread.gmane.org/gmane.linux.network/386173/ Signed-off-by: Daniel Borkmann Acked-by: Jason Wang Signed-off-by: David S. Miller --- net/packet/af_packet.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 80c36c0867d3..bdecf17a15bb 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2447,8 +2447,7 @@ static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, len = ((to_write > len_max) ? len_max : to_write); } - if (!packet_use_direct_xmit(po)) - skb_probe_transport_header(skb, 0); + skb_probe_transport_header(skb, 0); return tp_len; } @@ -2808,8 +2807,8 @@ static int packet_snd(struct socket *sock, struct msghdr *msg, size_t len) len += vnet_hdr_len; } - if (!packet_use_direct_xmit(po)) - skb_probe_transport_header(skb, reserve); + skb_probe_transport_header(skb, reserve); + if (unlikely(extra_len == 4)) skb->no_fcs = 1; From 3c70c132488794e2489ab045559b0ce0afcf17de Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 11 Nov 2015 23:25:42 +0100 Subject: [PATCH 072/335] packet: only allow extra vlan len on ethernet devices Packet sockets can be used by various net devices and are not really restricted to ARPHRD_ETHER device types. However, when currently checking for the extra 4 bytes that can be transmitted in VLAN case, our assumption is that we generally probe on ARPHRD_ETHER devices. Therefore, before looking into Ethernet header, check the device type first. This also fixes the issue where non-ARPHRD_ETHER devices could have no dev->hard_header_len in TX_RING SOCK_RAW case, and thus the check would test unfilled linear part of the skb (instead of non-linear). Fixes: 57f89bfa2140 ("network: Allow af_packet to transmit +4 bytes for VLAN packets.") Fixes: 52f1454f629f ("packet: allow to transmit +4 byte in TX_RING slot for VLAN case") Signed-off-by: Daniel Borkmann Acked-by: Willem de Bruijn Signed-off-by: David S. Miller --- net/packet/af_packet.c | 60 ++++++++++++++++++------------------------ 1 file changed, 25 insertions(+), 35 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index bdecf17a15bb..8795b0fb1ed1 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -1741,6 +1741,20 @@ static void fanout_release(struct sock *sk) kfree_rcu(po->rollover, rcu); } +static bool packet_extra_vlan_len_allowed(const struct net_device *dev, + struct sk_buff *skb) +{ + /* Earlier code assumed this would be a VLAN pkt, double-check + * this now that we have the actual packet in hand. We can only + * do this check on Ethernet devices. + */ + if (unlikely(dev->type != ARPHRD_ETHER)) + return false; + + skb_reset_mac_header(skb); + return likely(eth_hdr(skb)->h_proto == htons(ETH_P_8021Q)); +} + static const struct proto_ops packet_ops; static const struct proto_ops packet_ops_spkt; @@ -1902,18 +1916,10 @@ retry: goto retry; } - if (len > (dev->mtu + dev->hard_header_len + extra_len)) { - /* Earlier code assumed this would be a VLAN pkt, - * double-check this now that we have the actual - * packet in hand. - */ - struct ethhdr *ehdr; - skb_reset_mac_header(skb); - ehdr = eth_hdr(skb); - if (ehdr->h_proto != htons(ETH_P_8021Q)) { - err = -EMSGSIZE; - goto out_unlock; - } + if (len > (dev->mtu + dev->hard_header_len + extra_len) && + !packet_extra_vlan_len_allowed(dev, skb)) { + err = -EMSGSIZE; + goto out_unlock; } skb->protocol = proto; @@ -2525,18 +2531,10 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) tp_len = tpacket_fill_skb(po, skb, ph, dev, size_max, proto, addr, hlen); if (likely(tp_len >= 0) && - tp_len > dev->mtu + dev->hard_header_len) { - struct ethhdr *ehdr; - /* Earlier code assumed this would be a VLAN pkt, - * double-check this now that we have the actual - * packet in hand. - */ + tp_len > dev->mtu + dev->hard_header_len && + !packet_extra_vlan_len_allowed(dev, skb)) + tp_len = -EMSGSIZE; - skb_reset_mac_header(skb); - ehdr = eth_hdr(skb); - if (ehdr->h_proto != htons(ETH_P_8021Q)) - tp_len = -EMSGSIZE; - } if (unlikely(tp_len < 0)) { if (po->tp_loss) { __packet_set_status(po, ph, @@ -2765,18 +2763,10 @@ static int packet_snd(struct socket *sock, struct msghdr *msg, size_t len) sock_tx_timestamp(sk, &skb_shinfo(skb)->tx_flags); - if (!gso_type && (len > dev->mtu + reserve + extra_len)) { - /* Earlier code assumed this would be a VLAN pkt, - * double-check this now that we have the actual - * packet in hand. - */ - struct ethhdr *ehdr; - skb_reset_mac_header(skb); - ehdr = eth_hdr(skb); - if (ehdr->h_proto != htons(ETH_P_8021Q)) { - err = -EMSGSIZE; - goto out_free; - } + if (!gso_type && (len > dev->mtu + reserve + extra_len) && + !packet_extra_vlan_len_allowed(dev, skb)) { + err = -EMSGSIZE; + goto out_free; } skb->protocol = proto; From c72219b75fde768efccf7666342282fab7f9e4e7 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 11 Nov 2015 23:25:43 +0100 Subject: [PATCH 073/335] packet: infer protocol from ethernet header if unset In case no struct sockaddr_ll has been passed to packet socket's sendmsg() when doing a TX_RING flush run, then skb->protocol is set to po->num instead, which is the protocol passed via socket(2)/bind(2). Applications only xmitting can go the path of allocating the socket as socket(PF_PACKET, , 0) and do a bind(2) on the TX_RING with sll_protocol of 0. That way, register_prot_hook() is neither called on creation nor on bind time, which saves cycles when there's no interest in capturing anyway. That leaves us however with po->num 0 instead and therefore the TX_RING flush run sets skb->protocol to 0 as well. Eric reported that this leads to problems when using tools like trafgen over bonding device. I.e. the bonding's hash function could invoke the kernel's flow dissector, which depends on skb->protocol being properly set. In the current situation, all the traffic is then directed to a single slave. Fix it up by inferring skb->protocol from the Ethernet header when not set and we have ARPHRD_ETHER device type. This is only done in case of SOCK_RAW and where we have a dev->hard_header_len length. In case of ARPHRD_ETHER devices, this is guaranteed to cover ETH_HLEN, and therefore being accessed on the skb after the skb_store_bits(). Reported-by: Eric Dumazet Signed-off-by: Daniel Borkmann Acked-by: Willem de Bruijn Signed-off-by: David S. Miller --- net/packet/af_packet.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 8795b0fb1ed1..0066da2b8e44 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2338,6 +2338,15 @@ static bool ll_header_truncated(const struct net_device *dev, int len) return false; } +static void tpacket_set_protocol(const struct net_device *dev, + struct sk_buff *skb) +{ + if (dev->type == ARPHRD_ETHER) { + skb_reset_mac_header(skb); + skb->protocol = eth_hdr(skb)->h_proto; + } +} + static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, void *frame, struct net_device *dev, int size_max, __be16 proto, unsigned char *addr, int hlen) @@ -2419,6 +2428,8 @@ static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, dev->hard_header_len); if (unlikely(err)) return err; + if (!skb->protocol) + tpacket_set_protocol(dev, skb); data += dev->hard_header_len; to_write -= dev->hard_header_len; From 5cfb4c8d05b4409c4044cb9c05b19705c1d9818b Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 11 Nov 2015 23:25:44 +0100 Subject: [PATCH 074/335] packet: fix tpacket_snd max frame len Since it's introduction in commit 69e3c75f4d54 ("net: TX_RING and packet mmap"), TX_RING could be used from SOCK_DGRAM and SOCK_RAW side. When used with SOCK_DGRAM only, the size_max > dev->mtu + reserve check should have reserve as 0, but currently, this is unconditionally set (in it's original form as dev->hard_header_len). I think this is not correct since tpacket_fill_skb() would then take dev->mtu and dev->hard_header_len into account for SOCK_DGRAM, the extra VLAN_HLEN could be possible in both cases. Presumably, the reserve code was copied from packet_snd(), but later on missed the check. Make it similar as we have it in packet_snd(). Fixes: 69e3c75f4d54 ("net: TX_RING and packet mmap") Signed-off-by: Daniel Borkmann Acked-by: Willem de Bruijn Signed-off-by: David S. Miller --- net/packet/af_packet.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 0066da2b8e44..242bce1cf0f3 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2510,12 +2510,13 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) if (unlikely(!(dev->flags & IFF_UP))) goto out_put; - reserve = dev->hard_header_len + VLAN_HLEN; + if (po->sk.sk_socket->type == SOCK_RAW) + reserve = dev->hard_header_len; size_max = po->tx_ring.frame_size - (po->tp_hdrlen - sizeof(struct sockaddr_ll)); - if (size_max > dev->mtu + reserve) - size_max = dev->mtu + reserve; + if (size_max > dev->mtu + reserve + VLAN_HLEN) + size_max = dev->mtu + reserve + VLAN_HLEN; do { ph = packet_current_frame(po, &po->tx_ring, @@ -2542,7 +2543,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) tp_len = tpacket_fill_skb(po, skb, ph, dev, size_max, proto, addr, hlen); if (likely(tp_len >= 0) && - tp_len > dev->mtu + dev->hard_header_len && + tp_len > dev->mtu + reserve && !packet_extra_vlan_len_allowed(dev, skb)) tp_len = -EMSGSIZE; From ed5a377d87dc4c87fb3e1f7f698cba38cd893103 Mon Sep 17 00:00:00 2001 From: lucien Date: Thu, 12 Nov 2015 13:07:07 +0800 Subject: [PATCH 075/335] sctp: translate host order to network order when setting a hmacid now sctp auth cannot work well when setting a hmacid manually, which is caused by that we didn't use the network order for hmacid, so fix it by adding the transformation in sctp_auth_ep_set_hmacs. even we set hmacid with the network order in userspace, it still can't work, because of this condition in sctp_auth_ep_set_hmacs(): if (id > SCTP_AUTH_HMAC_ID_MAX) return -EOPNOTSUPP; so this wasn't working before and thus it won't break compatibility. Fixes: 65b07e5d0d09 ("[SCTP]: API updates to suport SCTP-AUTH extensions.") Signed-off-by: Xin Long Signed-off-by: Marcelo Ricardo Leitner Acked-by: Neil Horman Acked-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/auth.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/sctp/auth.c b/net/sctp/auth.c index 4f15b7d730e1..1543e39f47c3 100644 --- a/net/sctp/auth.c +++ b/net/sctp/auth.c @@ -809,8 +809,8 @@ int sctp_auth_ep_set_hmacs(struct sctp_endpoint *ep, if (!has_sha1) return -EINVAL; - memcpy(ep->auth_hmacs_list->hmac_ids, &hmacs->shmac_idents[0], - hmacs->shmac_num_idents * sizeof(__u16)); + for (i = 0; i < hmacs->shmac_num_idents; i++) + ep->auth_hmacs_list->hmac_ids[i] = htons(hmacs->shmac_idents[i]); ep->auth_hmacs_list->param_hdr.length = htons(sizeof(sctp_paramhdr_t) + hmacs->shmac_num_idents * sizeof(__u16)); return 0; From 5883d9c6d7e680bcdc7a8a9ed2509cd10dd98206 Mon Sep 17 00:00:00 2001 From: Pavel Fedin Date: Thu, 12 Nov 2015 14:55:18 +0300 Subject: [PATCH 076/335] net: thunder: Fix crash upon shutdown after failed probe If device probe fails, driver remains bound to the PCI device. However, driver data has been reset to NULL. This causes crash upon dereferencing it in nicvf_remove() Signed-off-by: Pavel Fedin Signed-off-by: David S. Miller --- drivers/net/ethernet/cavium/thunder/nicvf_main.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_main.c b/drivers/net/ethernet/cavium/thunder/nicvf_main.c index a9377727c11c..372c39e5bcbd 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@ -1600,6 +1600,9 @@ static void nicvf_remove(struct pci_dev *pdev) static void nicvf_shutdown(struct pci_dev *pdev) { + if (!pci_get_drvdata(pdev)) + return; + nicvf_remove(pdev); } From 00fd38d938db3f1ab1c486549afc450cb7e751b1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 12 Nov 2015 08:43:18 -0800 Subject: [PATCH 077/335] tcp: ensure proper barriers in lockless contexts Some functions access TCP sockets without holding a lock and might output non consistent data, depending on compiler and or architecture. tcp_diag_get_info(), tcp_get_info(), tcp_poll(), get_tcp4_sock() ... Introduce sk_state_load() and sk_state_store() to fix the issues, and more clearly document where this lack of locking is happening. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/sock.h | 25 +++++++++++++++++++++++++ net/ipv4/inet_connection_sock.c | 4 ++-- net/ipv4/tcp.c | 21 +++++++++++---------- net/ipv4/tcp_diag.c | 2 +- net/ipv4/tcp_ipv4.c | 14 ++++++++------ net/ipv6/tcp_ipv6.c | 19 +++++++++++++++---- 6 files changed, 62 insertions(+), 23 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index bbf7c2cf15b4..7f89e4ba18d1 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2226,6 +2226,31 @@ static inline bool sk_listener(const struct sock *sk) return (1 << sk->sk_state) & (TCPF_LISTEN | TCPF_NEW_SYN_RECV); } +/** + * sk_state_load - read sk->sk_state for lockless contexts + * @sk: socket pointer + * + * Paired with sk_state_store(). Used in places we do not hold socket lock : + * tcp_diag_get_info(), tcp_get_info(), tcp_poll(), get_tcp4_sock() ... + */ +static inline int sk_state_load(const struct sock *sk) +{ + return smp_load_acquire(&sk->sk_state); +} + +/** + * sk_state_store - update sk->sk_state + * @sk: socket pointer + * @newstate: new state + * + * Paired with sk_state_load(). Should be used in contexts where + * state change might impact lockless readers. + */ +static inline void sk_state_store(struct sock *sk, int newstate) +{ + smp_store_release(&sk->sk_state, newstate); +} + void sock_enable_timestamp(struct sock *sk, int flag); int sock_get_timestamp(struct sock *, struct timeval __user *); int sock_get_timestampns(struct sock *, struct timespec __user *); diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 1feb15f23de8..46b9c887bede 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -563,7 +563,7 @@ static void reqsk_timer_handler(unsigned long data) int max_retries, thresh; u8 defer_accept; - if (sk_listener->sk_state != TCP_LISTEN) + if (sk_state_load(sk_listener) != TCP_LISTEN) goto drop; max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; @@ -749,7 +749,7 @@ int inet_csk_listen_start(struct sock *sk, int backlog) * It is OK, because this socket enters to hash table only * after validation is complete. */ - sk->sk_state = TCP_LISTEN; + sk_state_store(sk, TCP_LISTEN); if (!sk->sk_prot->get_port(sk, inet->inet_num)) { inet->inet_sport = htons(inet->inet_num); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 0cfa7c0c1e80..c1728771cf89 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -451,11 +451,14 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) unsigned int mask; struct sock *sk = sock->sk; const struct tcp_sock *tp = tcp_sk(sk); + int state; sock_rps_record_flow(sk); sock_poll_wait(file, sk_sleep(sk), wait); - if (sk->sk_state == TCP_LISTEN) + + state = sk_state_load(sk); + if (state == TCP_LISTEN) return inet_csk_listen_poll(sk); /* Socket is not locked. We are protected from async events @@ -492,14 +495,14 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) * NOTE. Check for TCP_CLOSE is added. The goal is to prevent * blocking on fresh not-connected or disconnected socket. --ANK */ - if (sk->sk_shutdown == SHUTDOWN_MASK || sk->sk_state == TCP_CLOSE) + if (sk->sk_shutdown == SHUTDOWN_MASK || state == TCP_CLOSE) mask |= POLLHUP; if (sk->sk_shutdown & RCV_SHUTDOWN) mask |= POLLIN | POLLRDNORM | POLLRDHUP; /* Connected or passive Fast Open socket? */ - if (sk->sk_state != TCP_SYN_SENT && - (sk->sk_state != TCP_SYN_RECV || tp->fastopen_rsk)) { + if (state != TCP_SYN_SENT && + (state != TCP_SYN_RECV || tp->fastopen_rsk)) { int target = sock_rcvlowat(sk, 0, INT_MAX); if (tp->urg_seq == tp->copied_seq && @@ -507,9 +510,6 @@ unsigned int tcp_poll(struct file *file, struct socket *sock, poll_table *wait) tp->urg_data) target++; - /* Potential race condition. If read of tp below will - * escape above sk->sk_state, we can be illegally awaken - * in SYN_* states. */ if (tp->rcv_nxt - tp->copied_seq >= target) mask |= POLLIN | POLLRDNORM; @@ -1934,7 +1934,7 @@ void tcp_set_state(struct sock *sk, int state) /* Change state AFTER socket is unhashed to avoid closed * socket sitting in hash tables. */ - sk->sk_state = state; + sk_state_store(sk, state); #ifdef STATE_TRACE SOCK_DEBUG(sk, "TCP sk=%p, State %s -> %s\n", sk, statename[oldstate], statename[state]); @@ -2644,7 +2644,8 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) if (sk->sk_type != SOCK_STREAM) return; - info->tcpi_state = sk->sk_state; + info->tcpi_state = sk_state_load(sk); + info->tcpi_ca_state = icsk->icsk_ca_state; info->tcpi_retransmits = icsk->icsk_retransmits; info->tcpi_probes = icsk->icsk_probes_out; @@ -2672,7 +2673,7 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info) info->tcpi_snd_mss = tp->mss_cache; info->tcpi_rcv_mss = icsk->icsk_ack.rcv_mss; - if (sk->sk_state == TCP_LISTEN) { + if (info->tcpi_state == TCP_LISTEN) { info->tcpi_unacked = sk->sk_ack_backlog; info->tcpi_sacked = sk->sk_max_ack_backlog; } else { diff --git a/net/ipv4/tcp_diag.c b/net/ipv4/tcp_diag.c index 479f34946177..b31604086edd 100644 --- a/net/ipv4/tcp_diag.c +++ b/net/ipv4/tcp_diag.c @@ -21,7 +21,7 @@ static void tcp_diag_get_info(struct sock *sk, struct inet_diag_msg *r, { struct tcp_info *info = _info; - if (sk->sk_state == TCP_LISTEN) { + if (sk_state_load(sk) == TCP_LISTEN) { r->idiag_rqueue = sk->sk_ack_backlog; r->idiag_wqueue = sk->sk_max_ack_backlog; } else if (sk->sk_type == SOCK_STREAM) { diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 950e28c0cdf2..ba09016d1bfd 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2158,6 +2158,7 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i) __u16 destp = ntohs(inet->inet_dport); __u16 srcp = ntohs(inet->inet_sport); int rx_queue; + int state; if (icsk->icsk_pending == ICSK_TIME_RETRANS || icsk->icsk_pending == ICSK_TIME_EARLY_RETRANS || @@ -2175,17 +2176,18 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i) timer_expires = jiffies; } - if (sk->sk_state == TCP_LISTEN) + state = sk_state_load(sk); + if (state == TCP_LISTEN) rx_queue = sk->sk_ack_backlog; else - /* - * because we dont lock socket, we might find a transient negative value + /* Because we don't lock the socket, + * we might find a transient negative value. */ rx_queue = max_t(int, tp->rcv_nxt - tp->copied_seq, 0); seq_printf(f, "%4d: %08X:%04X %08X:%04X %02X %08X:%08X %02X:%08lX " "%08X %5u %8d %lu %d %pK %lu %lu %u %u %d", - i, src, srcp, dest, destp, sk->sk_state, + i, src, srcp, dest, destp, state, tp->write_seq - tp->snd_una, rx_queue, timer_active, @@ -2199,8 +2201,8 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i) jiffies_to_clock_t(icsk->icsk_ack.ato), (icsk->icsk_ack.quick << 1) | icsk->icsk_ack.pingpong, tp->snd_cwnd, - sk->sk_state == TCP_LISTEN ? - (fastopenq ? fastopenq->max_qlen : 0) : + state == TCP_LISTEN ? + fastopenq->max_qlen : (tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh)); } diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 5baa8e754e41..c5429a636f1a 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1690,6 +1690,8 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) const struct tcp_sock *tp = tcp_sk(sp); const struct inet_connection_sock *icsk = inet_csk(sp); const struct fastopen_queue *fastopenq = &icsk->icsk_accept_queue.fastopenq; + int rx_queue; + int state; dest = &sp->sk_v6_daddr; src = &sp->sk_v6_rcv_saddr; @@ -1710,6 +1712,15 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) timer_expires = jiffies; } + state = sk_state_load(sp); + if (state == TCP_LISTEN) + rx_queue = sp->sk_ack_backlog; + else + /* Because we don't lock the socket, + * we might find a transient negative value. + */ + rx_queue = max_t(int, tp->rcv_nxt - tp->copied_seq, 0); + seq_printf(seq, "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X " "%02X %08X:%08X %02X:%08lX %08X %5u %8d %lu %d %pK %lu %lu %u %u %d\n", @@ -1718,9 +1729,9 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) src->s6_addr32[2], src->s6_addr32[3], srcp, dest->s6_addr32[0], dest->s6_addr32[1], dest->s6_addr32[2], dest->s6_addr32[3], destp, - sp->sk_state, - tp->write_seq-tp->snd_una, - (sp->sk_state == TCP_LISTEN) ? sp->sk_ack_backlog : (tp->rcv_nxt - tp->copied_seq), + state, + tp->write_seq - tp->snd_una, + rx_queue, timer_active, jiffies_delta_to_clock_t(timer_expires - jiffies), icsk->icsk_retransmits, @@ -1732,7 +1743,7 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) jiffies_to_clock_t(icsk->icsk_ack.ato), (icsk->icsk_ack.quick << 1) | icsk->icsk_ack.pingpong, tp->snd_cwnd, - sp->sk_state == TCP_LISTEN ? + state == TCP_LISTEN ? fastopenq->max_qlen : (tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh) ); From 340c78e5906264c42f9415005c7ba0a4efcfe735 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 12 Nov 2015 09:14:12 -0800 Subject: [PATCH 078/335] ipvs: use skb_to_full_sk() helper SYNACK packets might be attached to request sockets. Use skb_to_full_sk() helper to avoid illegal accesses to inet_sk(skb->sk) Fixes: ca6fb0651883 ("tcp: attach SYNACK messages to request sockets instead of listener") Signed-off-by: Eric Dumazet Reported-by: Sander Eikelenboom Acked-by: Julian Anastasov Acked-by: Simon Horman Signed-off-by: David S. Miller --- net/netfilter/ipvs/ip_vs_core.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 1e24fff53e4b..f57b4dcdb233 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -1176,6 +1176,7 @@ ip_vs_out(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, in struct ip_vs_protocol *pp; struct ip_vs_proto_data *pd; struct ip_vs_conn *cp; + struct sock *sk; EnterFunction(11); @@ -1183,13 +1184,12 @@ ip_vs_out(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, in if (skb->ipvs_property) return NF_ACCEPT; + sk = skb_to_full_sk(skb); /* Bad... Do not break raw sockets */ - if (unlikely(skb->sk != NULL && hooknum == NF_INET_LOCAL_OUT && + if (unlikely(sk && hooknum == NF_INET_LOCAL_OUT && af == AF_INET)) { - struct sock *sk = skb->sk; - struct inet_sock *inet = inet_sk(skb->sk); - if (inet && sk->sk_family == PF_INET && inet->nodefrag) + if (sk->sk_family == PF_INET && inet_sk(sk)->nodefrag) return NF_ACCEPT; } @@ -1681,6 +1681,7 @@ ip_vs_in(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, int struct ip_vs_conn *cp; int ret, pkts; int conn_reuse_mode; + struct sock *sk; /* Already marked as IPVS request or reply? */ if (skb->ipvs_property) @@ -1708,12 +1709,11 @@ ip_vs_in(struct netns_ipvs *ipvs, unsigned int hooknum, struct sk_buff *skb, int ip_vs_fill_iph_skb(af, skb, false, &iph); /* Bad... Do not break raw sockets */ - if (unlikely(skb->sk != NULL && hooknum == NF_INET_LOCAL_OUT && + sk = skb_to_full_sk(skb); + if (unlikely(sk && hooknum == NF_INET_LOCAL_OUT && af == AF_INET)) { - struct sock *sk = skb->sk; - struct inet_sock *inet = inet_sk(skb->sk); - if (inet && sk->sk_family == PF_INET && inet->nodefrag) + if (sk->sk_family == PF_INET && inet_sk(sk)->nodefrag) return NF_ACCEPT; } From ba6c4c094470dc83a7275000bac2fbd46bd5ab69 Mon Sep 17 00:00:00 2001 From: Saeed Mahameed Date: Thu, 12 Nov 2015 19:35:25 +0200 Subject: [PATCH 079/335] net/mlx5e: Fix inline header size calculation mlx5e_get_inline_hdr_size didn't take into account the vlan insertion into the inline WQE segment. This could lead to max inline violation in cases where skb_headlen(skb) + VLAN_HLEN >= sq->max_inline. Fixes: 3ea4891db8d0 ("net/mlx5e: Fix LSO vlan insertion") Signed-off-by: Saeed Mahameed Signed-off-by: Achiad Shochat Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx5/core/en_tx.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c index cd8f85a251d7..f687ebf20d9c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c @@ -118,8 +118,15 @@ static inline u16 mlx5e_get_inline_hdr_size(struct mlx5e_sq *sq, */ #define MLX5E_MIN_INLINE ETH_HLEN - if (bf && (skb_headlen(skb) <= sq->max_inline)) - return skb_headlen(skb); + if (bf) { + u16 ihs = skb_headlen(skb); + + if (skb_vlan_tag_present(skb)) + ihs += VLAN_HLEN; + + if (ihs <= sq->max_inline) + return skb_headlen(skb); + } return MLX5E_MIN_INLINE; } From 66189961e986e53ae39822898fc2ce88f44c61bb Mon Sep 17 00:00:00 2001 From: Tariq Toukan Date: Thu, 12 Nov 2015 19:35:26 +0200 Subject: [PATCH 080/335] net/mlx5e: Added self loopback prevention Prevent outgoing multicast frames from looping back to the RX queue. By introducing new HW capability self_lb_en_modifiable, which indicates the support to modify self_lb_en bit in modify_tir command. When this capability is set we can prevent TIRs from sending back loopback multicast traffic to their own RQs, by "refreshing TIRs" with modify_tir command, on every time new channels (SQs/RQs) are created at device open. This is needed since TIRs are static and only allocated once on driver load, and the loopback decision is under their responsibility. Fixes issues of the kind: "IPv6: eth2: IPv6 duplicate address fe80::e61d:2dff:fe5c:f2e9 detected!" The issue is seen since the IPv6 solicitations multicast messages are loopedback and the network stack thinks they are coming from another host. Fixes: 5c50368f3831 ("net/mlx5e: Light-weight netdev open/stop") Signed-off-by: Tariq Toukan Signed-off-by: Saeed Mahameed Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- .../net/ethernet/mellanox/mlx5/core/en_main.c | 48 +++++++++++++++++++ include/linux/mlx5/mlx5_ifc.h | 24 ++++++---- 2 files changed, 62 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 5fc4d2d78cdf..df001754bcd1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -1332,6 +1332,42 @@ static int mlx5e_modify_tir_lro(struct mlx5e_priv *priv, int tt) return err; } +static int mlx5e_refresh_tir_self_loopback_enable(struct mlx5_core_dev *mdev, + u32 tirn) +{ + void *in; + int inlen; + int err; + + inlen = MLX5_ST_SZ_BYTES(modify_tir_in); + in = mlx5_vzalloc(inlen); + if (!in) + return -ENOMEM; + + MLX5_SET(modify_tir_in, in, bitmask.self_lb_en, 1); + + err = mlx5_core_modify_tir(mdev, tirn, in, inlen); + + kvfree(in); + + return err; +} + +static int mlx5e_refresh_tirs_self_loopback_enable(struct mlx5e_priv *priv) +{ + int err; + int i; + + for (i = 0; i < MLX5E_NUM_TT; i++) { + err = mlx5e_refresh_tir_self_loopback_enable(priv->mdev, + priv->tirn[i]); + if (err) + return err; + } + + return 0; +} + static int mlx5e_set_dev_port_mtu(struct net_device *netdev) { struct mlx5e_priv *priv = netdev_priv(netdev); @@ -1376,6 +1412,13 @@ int mlx5e_open_locked(struct net_device *netdev) goto err_clear_state_opened_flag; } + err = mlx5e_refresh_tirs_self_loopback_enable(priv); + if (err) { + netdev_err(netdev, "%s: mlx5e_refresh_tirs_self_loopback_enable failed, %d\n", + __func__, err); + goto err_close_channels; + } + mlx5e_update_carrier(priv); mlx5e_redirect_rqts(priv); @@ -1383,6 +1426,8 @@ int mlx5e_open_locked(struct net_device *netdev) return 0; +err_close_channels: + mlx5e_close_channels(priv); err_clear_state_opened_flag: clear_bit(MLX5E_STATE_OPENED, &priv->state); return err; @@ -1909,6 +1954,9 @@ static int mlx5e_check_required_hca_cap(struct mlx5_core_dev *mdev) "Not creating net device, some required device capabilities are missing\n"); return -ENOTSUPP; } + if (!MLX5_CAP_ETH(mdev, self_lb_en_modifiable)) + mlx5_core_warn(mdev, "Self loop back prevention is not supported\n"); + return 0; } diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h index dd2097455a2e..1565324eb620 100644 --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -453,26 +453,28 @@ struct mlx5_ifc_per_protocol_networking_offload_caps_bits { u8 lro_cap[0x1]; u8 lro_psh_flag[0x1]; u8 lro_time_stamp[0x1]; - u8 reserved_0[0x6]; + u8 reserved_0[0x3]; + u8 self_lb_en_modifiable[0x1]; + u8 reserved_1[0x2]; u8 max_lso_cap[0x5]; - u8 reserved_1[0x4]; + u8 reserved_2[0x4]; u8 rss_ind_tbl_cap[0x4]; - u8 reserved_2[0x3]; + u8 reserved_3[0x3]; u8 tunnel_lso_const_out_ip_id[0x1]; - u8 reserved_3[0x2]; + u8 reserved_4[0x2]; u8 tunnel_statless_gre[0x1]; u8 tunnel_stateless_vxlan[0x1]; - u8 reserved_4[0x20]; + u8 reserved_5[0x20]; - u8 reserved_5[0x10]; + u8 reserved_6[0x10]; u8 lro_min_mss_size[0x10]; - u8 reserved_6[0x120]; + u8 reserved_7[0x120]; u8 lro_timer_supported_periods[4][0x20]; - u8 reserved_7[0x600]; + u8 reserved_8[0x600]; }; struct mlx5_ifc_roce_cap_bits { @@ -4051,9 +4053,11 @@ struct mlx5_ifc_modify_tis_in_bits { }; struct mlx5_ifc_modify_tir_bitmask_bits { - u8 reserved[0x20]; + u8 reserved_0[0x20]; - u8 reserved1[0x1f]; + u8 reserved_1[0x1b]; + u8 self_lb_en[0x1]; + u8 reserved_2[0x3]; u8 lro[0x1]; }; From 50a9eea694ab8e0779069e0a4e0b12e145521468 Mon Sep 17 00:00:00 2001 From: Doron Tsur Date: Thu, 12 Nov 2015 19:35:27 +0200 Subject: [PATCH 081/335] net/mlx5e: Max mtu comparison fix On change mtu the driver compares between hardware queried mtu and software requested mtu. We need to compare between software representation of the queried mtu and the requested mtu. Fixes: facc9699f0fe ('net/mlx5e: Fix HW MTU settings') Signed-off-by: Doron Tsur Signed-off-by: Saeed Mahameed Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index df001754bcd1..1e52db32c73d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -1901,6 +1901,8 @@ static int mlx5e_change_mtu(struct net_device *netdev, int new_mtu) mlx5_query_port_max_mtu(mdev, &max_mtu, 1); + max_mtu = MLX5E_HW2SW_MTU(max_mtu); + if (new_mtu > max_mtu) { netdev_err(netdev, "%s: Bad MTU (%d) > (%d) Max\n", From d4e28cbd24c8cb004960ddb8b22124953f6c220c Mon Sep 17 00:00:00 2001 From: Achiad Shochat Date: Thu, 12 Nov 2015 19:35:28 +0200 Subject: [PATCH 082/335] net/mlx5e: Use the right DMA free function on TX path On xmit path we use skb_frag_dma_map() which is using dma_map_page(), while upon completion we dma-unmap the skb fragments using dma_unmap_single() rather than dma_unmap_page(). To fix this, we now save the dma map type on xmit path and use this info to call the right dma unmap method upon TX completion. Signed-off-by: Achiad Shochat Signed-off-by: Saeed Mahameed Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 10 ++- .../net/ethernet/mellanox/mlx5/core/en_tx.c | 69 ++++++++++--------- 2 files changed, 45 insertions(+), 34 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index f2ae62dd8c09..22e72bf1ae48 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -334,9 +334,15 @@ struct mlx5e_tx_skb_cb { #define MLX5E_TX_SKB_CB(__skb) ((struct mlx5e_tx_skb_cb *)__skb->cb) +enum mlx5e_dma_map_type { + MLX5E_DMA_MAP_SINGLE, + MLX5E_DMA_MAP_PAGE +}; + struct mlx5e_sq_dma { - dma_addr_t addr; - u32 size; + dma_addr_t addr; + u32 size; + enum mlx5e_dma_map_type type; }; enum { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c index f687ebf20d9c..1341b1d3c421 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c @@ -61,41 +61,49 @@ void mlx5e_send_nop(struct mlx5e_sq *sq, bool notify_hw) } } -static void mlx5e_dma_pop_last_pushed(struct mlx5e_sq *sq, dma_addr_t *addr, - u32 *size) +static inline void mlx5e_tx_dma_unmap(struct device *pdev, + struct mlx5e_sq_dma *dma) { - sq->dma_fifo_pc--; - *addr = sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].addr; - *size = sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].size; + switch (dma->type) { + case MLX5E_DMA_MAP_SINGLE: + dma_unmap_single(pdev, dma->addr, dma->size, DMA_TO_DEVICE); + break; + case MLX5E_DMA_MAP_PAGE: + dma_unmap_page(pdev, dma->addr, dma->size, DMA_TO_DEVICE); + break; + default: + WARN_ONCE(true, "mlx5e_tx_dma_unmap unknown DMA type!\n"); + } +} + +static inline void mlx5e_dma_push(struct mlx5e_sq *sq, + dma_addr_t addr, + u32 size, + enum mlx5e_dma_map_type map_type) +{ + sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].addr = addr; + sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].size = size; + sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].type = map_type; + sq->dma_fifo_pc++; +} + +static inline struct mlx5e_sq_dma *mlx5e_dma_get(struct mlx5e_sq *sq, u32 i) +{ + return &sq->dma_fifo[i & sq->dma_fifo_mask]; } static void mlx5e_dma_unmap_wqe_err(struct mlx5e_sq *sq, struct sk_buff *skb) { - dma_addr_t addr; - u32 size; int i; for (i = 0; i < MLX5E_TX_SKB_CB(skb)->num_dma; i++) { - mlx5e_dma_pop_last_pushed(sq, &addr, &size); - dma_unmap_single(sq->pdev, addr, size, DMA_TO_DEVICE); + struct mlx5e_sq_dma *last_pushed_dma = + mlx5e_dma_get(sq, --sq->dma_fifo_pc); + + mlx5e_tx_dma_unmap(sq->pdev, last_pushed_dma); } } -static inline void mlx5e_dma_push(struct mlx5e_sq *sq, dma_addr_t addr, - u32 size) -{ - sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].addr = addr; - sq->dma_fifo[sq->dma_fifo_pc & sq->dma_fifo_mask].size = size; - sq->dma_fifo_pc++; -} - -static inline void mlx5e_dma_get(struct mlx5e_sq *sq, u32 i, dma_addr_t *addr, - u32 *size) -{ - *addr = sq->dma_fifo[i & sq->dma_fifo_mask].addr; - *size = sq->dma_fifo[i & sq->dma_fifo_mask].size; -} - u16 mlx5e_select_queue(struct net_device *dev, struct sk_buff *skb, void *accel_priv, select_queue_fallback_t fallback) { @@ -225,7 +233,7 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_sq *sq, struct sk_buff *skb) dseg->lkey = sq->mkey_be; dseg->byte_count = cpu_to_be32(headlen); - mlx5e_dma_push(sq, dma_addr, headlen); + mlx5e_dma_push(sq, dma_addr, headlen, MLX5E_DMA_MAP_SINGLE); MLX5E_TX_SKB_CB(skb)->num_dma++; dseg++; @@ -244,7 +252,7 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_sq *sq, struct sk_buff *skb) dseg->lkey = sq->mkey_be; dseg->byte_count = cpu_to_be32(fsz); - mlx5e_dma_push(sq, dma_addr, fsz); + mlx5e_dma_push(sq, dma_addr, fsz, MLX5E_DMA_MAP_PAGE); MLX5E_TX_SKB_CB(skb)->num_dma++; dseg++; @@ -360,13 +368,10 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq) } for (j = 0; j < MLX5E_TX_SKB_CB(skb)->num_dma; j++) { - dma_addr_t addr; - u32 size; + struct mlx5e_sq_dma *dma = + mlx5e_dma_get(sq, dma_fifo_cc++); - mlx5e_dma_get(sq, dma_fifo_cc, &addr, &size); - dma_fifo_cc++; - dma_unmap_single(sq->pdev, addr, size, - DMA_TO_DEVICE); + mlx5e_tx_dma_unmap(sq->pdev, dma); } npkts++; From f5adbfee72282bb1f456d52b04adacd4fe6ac502 Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Thu, 12 Nov 2015 19:35:29 +0200 Subject: [PATCH 083/335] net/mlx4_core: Fix sleeping while holding spinlock at rem_slave_counters When cleaning slave's counter resources, we hold a spinlock that protects the slave's counters list. As part of the clean, we call __mlx4_clear_if_stat which calls mlx4_alloc_cmd_mailbox which is a sleepable function. In order to fix this issue, hold the spinlock, and copy all counter indices into a temporary array, and release the spinlock. Afterwards, iterate over this array and free every counter. Repeat this scenario until the original list is empty (a new counter might have been added while releasing the counters from the temporary array). Fixes: b72ca7e96acf ("net/mlx4_core: Reset counters data when freed") Reported-by: Moni Shoua Tested-by: Moni Shoua Signed-off-by: Jack Morgenstein Signed-off-by: Eran Ben Elisha Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- .../ethernet/mellanox/mlx4/resource_tracker.c | 39 +++++++++++++------ 1 file changed, 27 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index 9813d34f3e5b..6fec3e993d02 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -4952,26 +4952,41 @@ static void rem_slave_counters(struct mlx4_dev *dev, int slave) struct res_counter *counter; struct res_counter *tmp; int err; - int index; + int *counters_arr = NULL; + int i, j; err = move_all_busy(dev, slave, RES_COUNTER); if (err) mlx4_warn(dev, "rem_slave_counters: Could not move all counters - too busy for slave %d\n", slave); - spin_lock_irq(mlx4_tlock(dev)); - list_for_each_entry_safe(counter, tmp, counter_list, com.list) { - if (counter->com.owner == slave) { - index = counter->com.res_id; - rb_erase(&counter->com.node, - &tracker->res_tree[RES_COUNTER]); - list_del(&counter->com.list); - kfree(counter); - __mlx4_counter_free(dev, index); + counters_arr = kmalloc_array(dev->caps.max_counters, + sizeof(*counters_arr), GFP_KERNEL); + if (!counters_arr) + return; + + do { + i = 0; + j = 0; + spin_lock_irq(mlx4_tlock(dev)); + list_for_each_entry_safe(counter, tmp, counter_list, com.list) { + if (counter->com.owner == slave) { + counters_arr[i++] = counter->com.res_id; + rb_erase(&counter->com.node, + &tracker->res_tree[RES_COUNTER]); + list_del(&counter->com.list); + kfree(counter); + } + } + spin_unlock_irq(mlx4_tlock(dev)); + + while (j < i) { + __mlx4_counter_free(dev, counters_arr[j++]); mlx4_release_resource(dev, slave, RES_COUNTER, 1, 0); } - } - spin_unlock_irq(mlx4_tlock(dev)); + } while (i); + + kfree(counters_arr); } static void rem_slave_xrcdns(struct mlx4_dev *dev, int slave) From d49c2197fd70c37d57982804465268440a33183a Mon Sep 17 00:00:00 2001 From: Noa Osherovich Date: Thu, 12 Nov 2015 19:35:30 +0200 Subject: [PATCH 084/335] net/mlx4_core: Avoid returning success in case of an error flow The err variable wasn't set with the correct error value in some cases. Fixes: 47605df95398 ('mlx4: Modify proxy/tunnel QP mechanism [..]') Signed-off-by: Noa Osherovich Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/main.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c index 85f1b1e7e505..31c491e02e69 100644 --- a/drivers/net/ethernet/mellanox/mlx4/main.c +++ b/drivers/net/ethernet/mellanox/mlx4/main.c @@ -892,9 +892,10 @@ static int mlx4_slave_cap(struct mlx4_dev *dev) dev->caps.qp1_proxy[i - 1] = func_cap.qp1_proxy_qpn; dev->caps.port_mask[i] = dev->caps.port_type[i]; dev->caps.phys_port_id[i] = func_cap.phys_port_id; - if (mlx4_get_slave_pkey_gid_tbl_len(dev, i, - &dev->caps.gid_table_len[i], - &dev->caps.pkey_table_len[i])) + err = mlx4_get_slave_pkey_gid_tbl_len(dev, i, + &dev->caps.gid_table_len[i], + &dev->caps.pkey_table_len[i]); + if (err) goto err_mem; } @@ -906,6 +907,7 @@ static int mlx4_slave_cap(struct mlx4_dev *dev) dev->caps.uar_page_size * dev->caps.num_uars, (unsigned long long) pci_resource_len(dev->persist->pdev, 2)); + err = -ENOMEM; goto err_mem; } From dc1e7eb846ae6677426c2dcdfe5ea323cf9036e9 Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:14 +0100 Subject: [PATCH 085/335] net: dsa: mv88e6060: remove poll_link callback As of mv88e6xxx remove the poll_link callback since the link state change polling is now handled by the phylib. Tested on a mv88e6060 B0 device with a TI DM816X SoC. Suggested-by: Andrew Lunn Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 49 ------------------------------------- 1 file changed, 49 deletions(-) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index 9093577755f6..6885ef5678bc 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -225,54 +225,6 @@ mv88e6060_phy_write(struct dsa_switch *ds, int port, int regnum, u16 val) return reg_write(ds, addr, regnum, val); } -static void mv88e6060_poll_link(struct dsa_switch *ds) -{ - int i; - - for (i = 0; i < DSA_MAX_PORTS; i++) { - struct net_device *dev; - int uninitialized_var(port_status); - int link; - int speed; - int duplex; - int fc; - - dev = ds->ports[i]; - if (dev == NULL) - continue; - - link = 0; - if (dev->flags & IFF_UP) { - port_status = reg_read(ds, REG_PORT(i), 0x00); - if (port_status < 0) - continue; - - link = !!(port_status & 0x1000); - } - - if (!link) { - if (netif_carrier_ok(dev)) { - netdev_info(dev, "link down\n"); - netif_carrier_off(dev); - } - continue; - } - - speed = (port_status & 0x0100) ? 100 : 10; - duplex = (port_status & 0x0200) ? 1 : 0; - fc = ((port_status & 0xc000) == 0xc000) ? 1 : 0; - - if (!netif_carrier_ok(dev)) { - netdev_info(dev, - "link up, %d Mb/s, %s duplex, flow control %sabled\n", - speed, - duplex ? "full" : "half", - fc ? "en" : "dis"); - netif_carrier_on(dev); - } - } -} - static struct dsa_switch_driver mv88e6060_switch_driver = { .tag_protocol = DSA_TAG_PROTO_TRAILER, .probe = mv88e6060_probe, @@ -280,7 +232,6 @@ static struct dsa_switch_driver mv88e6060_switch_driver = { .set_addr = mv88e6060_set_addr, .phy_read = mv88e6060_phy_read, .phy_write = mv88e6060_phy_write, - .poll_link = mv88e6060_poll_link, }; static int __init mv88e6060_init(void) From c37909b4bb24320a59f697439bc77c700ca354fe Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:19 +0100 Subject: [PATCH 086/335] net: dsa: mv88e6060: use the correct InitReady bit According to the mv88e6060 datasheet, the InitReady bit position is 11 and the polarity is inverted. Use the bit correctly to detect the end of initialization. Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Acked-by: Vivien Didelot Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index 6885ef5678bc..eff5e18079a7 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -102,7 +102,7 @@ static int mv88e6060_switch_reset(struct dsa_switch *ds) timeout = jiffies + 1 * HZ; while (time_before(jiffies, timeout)) { ret = REG_READ(REG_GLOBAL, 0x00); - if ((ret & 0x8000) == 0x0000) + if (ret & 0x800) break; usleep_range(1000, 2000); From b73c774e08cf1e8877b000585e8598ecee890c7f Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:24 +0100 Subject: [PATCH 087/335] net: dsa: mv88e6060: use the correct MaxFrameSize bit According to the mv88e6060 datasheet, the MaxFrameSize bit position is 10 instead of 11 which is reserved. Use the bit correctly to setup max frame size to 1536. Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index eff5e18079a7..10647ade2f6d 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -119,7 +119,7 @@ static int mv88e6060_setup_global(struct dsa_switch *ds) * set the maximum frame size to 1536 bytes, and mask all * interrupt sources. */ - REG_WRITE(REG_GLOBAL, 0x04, 0x0800); + REG_WRITE(REG_GLOBAL, 0x04, 0x400); /* Enable automatic address learning, set the address * database size to 1024 entries, and set the default aging From 83ea0f4cb344089b6f240d3793d8c522a8501037 Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:32 +0100 Subject: [PATCH 088/335] net: dsa: mv88e6060: use the correct bit shift for mac0 According to the mv88e6060 datasheet, the first mac byte must be at position 9 instead of 8 since the bit 8 is used to select if the mac address must differ for each port for Pause frames. Use the correct shift and set the same mac address for all port. Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index 10647ade2f6d..cd08079e84ad 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -188,7 +188,8 @@ static int mv88e6060_setup(struct dsa_switch *ds) static int mv88e6060_set_addr(struct dsa_switch *ds, u8 *addr) { - REG_WRITE(REG_GLOBAL, 0x01, (addr[0] << 8) | addr[1]); + /* Use the same MAC Address as FD Pause frames for all ports */ + REG_WRITE(REG_GLOBAL, 0x01, (addr[0] << 9) | addr[1]); REG_WRITE(REG_GLOBAL, 0x02, (addr[2] << 8) | addr[3]); REG_WRITE(REG_GLOBAL, 0x03, (addr[4] << 8) | addr[5]); From f7e3931181595cd15e22e199d1bbabb0468d5a93 Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:42 +0100 Subject: [PATCH 089/335] net: dsa: mv88e6060: add register defines header file To align with the mv88e6xxx code, add a similar header file with all the register defines. The file is based on the mv88e6xxx header for coherency. Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Acked-by: Vivien Didelot Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.h | 111 ++++++++++++++++++++++++++++++++++++ 1 file changed, 111 insertions(+) create mode 100644 drivers/net/dsa/mv88e6060.h diff --git a/drivers/net/dsa/mv88e6060.h b/drivers/net/dsa/mv88e6060.h new file mode 100644 index 000000000000..cc9b2ed4aff4 --- /dev/null +++ b/drivers/net/dsa/mv88e6060.h @@ -0,0 +1,111 @@ +/* + * drivers/net/dsa/mv88e6060.h - Marvell 88e6060 switch chip support + * Copyright (c) 2015 Neil Armstrong + * + * Based on mv88e6xxx.h + * Copyright (c) 2008 Marvell Semiconductor + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#ifndef __MV88E6060_H +#define __MV88E6060_H + +#define MV88E6060_PORTS 6 + +#define REG_PORT(p) (0x8 + (p)) +#define PORT_STATUS 0x00 +#define PORT_STATUS_PAUSE_EN BIT(15) +#define PORT_STATUS_MY_PAUSE BIT(14) +#define PORT_STATUS_FC (PORT_STATUS_MY_PAUSE | PORT_STATUS_PAUSE_EN) +#define PORT_STATUS_RESOLVED BIT(13) +#define PORT_STATUS_LINK BIT(12) +#define PORT_STATUS_PORTMODE BIT(11) +#define PORT_STATUS_PHYMODE BIT(10) +#define PORT_STATUS_DUPLEX BIT(9) +#define PORT_STATUS_SPEED BIT(8) +#define PORT_SWITCH_ID 0x03 +#define PORT_SWITCH_ID_6060 0x0600 +#define PORT_SWITCH_ID_6060_MASK 0xfff0 +#define PORT_SWITCH_ID_6060_R1 0x0601 +#define PORT_SWITCH_ID_6060_R2 0x0602 +#define PORT_CONTROL 0x04 +#define PORT_CONTROL_FORCE_FLOW_CTRL BIT(15) +#define PORT_CONTROL_TRAILER BIT(14) +#define PORT_CONTROL_HEADER BIT(11) +#define PORT_CONTROL_INGRESS_MODE BIT(8) +#define PORT_CONTROL_VLAN_TUNNEL BIT(7) +#define PORT_CONTROL_STATE_MASK 0x03 +#define PORT_CONTROL_STATE_DISABLED 0x00 +#define PORT_CONTROL_STATE_BLOCKING 0x01 +#define PORT_CONTROL_STATE_LEARNING 0x02 +#define PORT_CONTROL_STATE_FORWARDING 0x03 +#define PORT_VLAN_MAP 0x06 +#define PORT_VLAN_MAP_DBNUM_SHIFT 12 +#define PORT_VLAN_MAP_TABLE_MASK 0x1f +#define PORT_ASSOC_VECTOR 0x0b +#define PORT_ASSOC_VECTOR_MONITOR BIT(15) +#define PORT_ASSOC_VECTOR_PAV_MASK 0x1f +#define PORT_RX_CNTR 0x10 +#define PORT_TX_CNTR 0x11 + +#define REG_GLOBAL 0x0f +#define GLOBAL_STATUS 0x00 +#define GLOBAL_STATUS_SW_MODE_MASK (0x3 << 12) +#define GLOBAL_STATUS_SW_MODE_0 (0x0 << 12) +#define GLOBAL_STATUS_SW_MODE_1 (0x1 << 12) +#define GLOBAL_STATUS_SW_MODE_2 (0x2 << 12) +#define GLOBAL_STATUS_SW_MODE_3 (0x3 << 12) +#define GLOBAL_STATUS_INIT_READY BIT(11) +#define GLOBAL_STATUS_ATU_FULL BIT(3) +#define GLOBAL_STATUS_ATU_DONE BIT(2) +#define GLOBAL_STATUS_PHY_INT BIT(1) +#define GLOBAL_STATUS_EEINT BIT(0) +#define GLOBAL_MAC_01 0x01 +#define GLOBAL_MAC_01_DIFF_ADDR BIT(8) +#define GLOBAL_MAC_23 0x02 +#define GLOBAL_MAC_45 0x03 +#define GLOBAL_CONTROL 0x04 +#define GLOBAL_CONTROL_DISCARD_EXCESS BIT(13) +#define GLOBAL_CONTROL_MAX_FRAME_1536 BIT(10) +#define GLOBAL_CONTROL_RELOAD_EEPROM BIT(9) +#define GLOBAL_CONTROL_CTRMODE BIT(8) +#define GLOBAL_CONTROL_ATU_FULL_EN BIT(3) +#define GLOBAL_CONTROL_ATU_DONE_EN BIT(2) +#define GLOBAL_CONTROL_PHYINT_EN BIT(1) +#define GLOBAL_CONTROL_EEPROM_DONE_EN BIT(0) +#define GLOBAL_ATU_CONTROL 0x0a +#define GLOBAL_ATU_CONTROL_SWRESET BIT(15) +#define GLOBAL_ATU_CONTROL_LEARNDIS BIT(14) +#define GLOBAL_ATU_CONTROL_ATUSIZE_256 (0x0 << 12) +#define GLOBAL_ATU_CONTROL_ATUSIZE_512 (0x1 << 12) +#define GLOBAL_ATU_CONTROL_ATUSIZE_1024 (0x2 << 12) +#define GLOBAL_ATU_CONTROL_ATE_AGE_SHIFT 4 +#define GLOBAL_ATU_CONTROL_ATE_AGE_MASK (0xff << 4) +#define GLOBAL_ATU_CONTROL_ATE_AGE_5MIN (0x13 << 4) +#define GLOBAL_ATU_OP 0x0b +#define GLOBAL_ATU_OP_BUSY BIT(15) +#define GLOBAL_ATU_OP_NOP (0 << 12) +#define GLOBAL_ATU_OP_FLUSH_ALL ((1 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_OP_FLUSH_UNLOCKED ((2 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_OP_LOAD_DB ((3 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_OP_GET_NEXT_DB ((4 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_OP_FLUSH_DB ((5 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_OP_FLUSH_UNLOCKED_DB ((6 << 12) | GLOBAL_ATU_OP_BUSY) +#define GLOBAL_ATU_DATA 0x0c +#define GLOBAL_ATU_DATA_PORT_VECTOR_MASK 0x3f0 +#define GLOBAL_ATU_DATA_PORT_VECTOR_SHIFT 4 +#define GLOBAL_ATU_DATA_STATE_MASK 0x0f +#define GLOBAL_ATU_DATA_STATE_UNUSED 0x00 +#define GLOBAL_ATU_DATA_STATE_UC_STATIC 0x0e +#define GLOBAL_ATU_DATA_STATE_UC_LOCKED 0x0f +#define GLOBAL_ATU_DATA_STATE_MC_STATIC 0x07 +#define GLOBAL_ATU_DATA_STATE_MC_LOCKED 0x0e +#define GLOBAL_ATU_MAC_01 0x0d +#define GLOBAL_ATU_MAC_23 0x0e +#define GLOBAL_ATU_MAC_45 0x0f + +#endif From 6a4b2980d18164a09734c2069ddb4ad4a3b69c71 Mon Sep 17 00:00:00 2001 From: Neil Armstrong Date: Tue, 10 Nov 2015 16:51:36 +0100 Subject: [PATCH 090/335] net: dsa: mv88e6060: replace magic values with register defines To align with the mv88e6xxx code, use the register defines to access all the register addresses and bit fields. Acked-by: Andrew Lunn Signed-off-by: Neil Armstrong Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 64 +++++++++++++++++++++---------------- 1 file changed, 37 insertions(+), 27 deletions(-) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index cd08079e84ad..0527f485c3dc 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -15,9 +15,7 @@ #include #include #include - -#define REG_PORT(p) (8 + (p)) -#define REG_GLOBAL 0x0f +#include "mv88e6060.h" static int reg_read(struct dsa_switch *ds, int addr, int reg) { @@ -67,13 +65,14 @@ static char *mv88e6060_probe(struct device *host_dev, int sw_addr) if (bus == NULL) return NULL; - ret = mdiobus_read(bus, sw_addr + REG_PORT(0), 0x03); + ret = mdiobus_read(bus, sw_addr + REG_PORT(0), PORT_SWITCH_ID); if (ret >= 0) { - if (ret == 0x0600) + if (ret == PORT_SWITCH_ID_6060) return "Marvell 88E6060 (A0)"; - if (ret == 0x0601 || ret == 0x0602) + if (ret == PORT_SWITCH_ID_6060_R1 || + ret == PORT_SWITCH_ID_6060_R2) return "Marvell 88E6060 (B0)"; - if ((ret & 0xfff0) == 0x0600) + if ((ret & PORT_SWITCH_ID_6060_MASK) == PORT_SWITCH_ID_6060) return "Marvell 88E6060"; } @@ -87,22 +86,26 @@ static int mv88e6060_switch_reset(struct dsa_switch *ds) unsigned long timeout; /* Set all ports to the disabled state. */ - for (i = 0; i < 6; i++) { - ret = REG_READ(REG_PORT(i), 0x04); - REG_WRITE(REG_PORT(i), 0x04, ret & 0xfffc); + for (i = 0; i < MV88E6060_PORTS; i++) { + ret = REG_READ(REG_PORT(i), PORT_CONTROL); + REG_WRITE(REG_PORT(i), PORT_CONTROL, + ret & ~PORT_CONTROL_STATE_MASK); } /* Wait for transmit queues to drain. */ usleep_range(2000, 4000); /* Reset the switch. */ - REG_WRITE(REG_GLOBAL, 0x0a, 0xa130); + REG_WRITE(REG_GLOBAL, GLOBAL_ATU_CONTROL, + GLOBAL_ATU_CONTROL_SWRESET | + GLOBAL_ATU_CONTROL_ATUSIZE_1024 | + GLOBAL_ATU_CONTROL_ATE_AGE_5MIN); /* Wait up to one second for reset to complete. */ timeout = jiffies + 1 * HZ; while (time_before(jiffies, timeout)) { - ret = REG_READ(REG_GLOBAL, 0x00); - if (ret & 0x800) + ret = REG_READ(REG_GLOBAL, GLOBAL_STATUS); + if (ret & GLOBAL_STATUS_INIT_READY) break; usleep_range(1000, 2000); @@ -119,13 +122,15 @@ static int mv88e6060_setup_global(struct dsa_switch *ds) * set the maximum frame size to 1536 bytes, and mask all * interrupt sources. */ - REG_WRITE(REG_GLOBAL, 0x04, 0x400); + REG_WRITE(REG_GLOBAL, GLOBAL_CONTROL, GLOBAL_CONTROL_MAX_FRAME_1536); /* Enable automatic address learning, set the address * database size to 1024 entries, and set the default aging * time to 5 minutes. */ - REG_WRITE(REG_GLOBAL, 0x0a, 0x2130); + REG_WRITE(REG_GLOBAL, GLOBAL_ATU_CONTROL, + GLOBAL_ATU_CONTROL_ATUSIZE_1024 | + GLOBAL_ATU_CONTROL_ATE_AGE_5MIN); return 0; } @@ -139,25 +144,30 @@ static int mv88e6060_setup_port(struct dsa_switch *ds, int p) * state to Forwarding. Additionally, if this is the CPU * port, enable Ingress and Egress Trailer tagging mode. */ - REG_WRITE(addr, 0x04, dsa_is_cpu_port(ds, p) ? 0x4103 : 0x0003); + REG_WRITE(addr, PORT_CONTROL, + dsa_is_cpu_port(ds, p) ? + PORT_CONTROL_TRAILER | + PORT_CONTROL_INGRESS_MODE | + PORT_CONTROL_STATE_FORWARDING : + PORT_CONTROL_STATE_FORWARDING); /* Port based VLAN map: give each port its own address * database, allow the CPU port to talk to each of the 'real' * ports, and allow each of the 'real' ports to only talk to * the CPU port. */ - REG_WRITE(addr, 0x06, - ((p & 0xf) << 12) | - (dsa_is_cpu_port(ds, p) ? - ds->phys_port_mask : - (1 << ds->dst->cpu_port))); + REG_WRITE(addr, PORT_VLAN_MAP, + ((p & 0xf) << PORT_VLAN_MAP_DBNUM_SHIFT) | + (dsa_is_cpu_port(ds, p) ? + ds->phys_port_mask : + BIT(ds->dst->cpu_port))); /* Port Association Vector: when learning source addresses * of packets, add the address to the address database using * a port bitmap that has only the bit for this port set and * the other bits clear. */ - REG_WRITE(addr, 0x0b, 1 << p); + REG_WRITE(addr, PORT_ASSOC_VECTOR, BIT(p)); return 0; } @@ -177,7 +187,7 @@ static int mv88e6060_setup(struct dsa_switch *ds) if (ret < 0) return ret; - for (i = 0; i < 6; i++) { + for (i = 0; i < MV88E6060_PORTS; i++) { ret = mv88e6060_setup_port(ds, i); if (ret < 0) return ret; @@ -189,16 +199,16 @@ static int mv88e6060_setup(struct dsa_switch *ds) static int mv88e6060_set_addr(struct dsa_switch *ds, u8 *addr) { /* Use the same MAC Address as FD Pause frames for all ports */ - REG_WRITE(REG_GLOBAL, 0x01, (addr[0] << 9) | addr[1]); - REG_WRITE(REG_GLOBAL, 0x02, (addr[2] << 8) | addr[3]); - REG_WRITE(REG_GLOBAL, 0x03, (addr[4] << 8) | addr[5]); + REG_WRITE(REG_GLOBAL, GLOBAL_MAC_01, (addr[0] << 9) | addr[1]); + REG_WRITE(REG_GLOBAL, GLOBAL_MAC_23, (addr[2] << 8) | addr[3]); + REG_WRITE(REG_GLOBAL, GLOBAL_MAC_45, (addr[4] << 8) | addr[5]); return 0; } static int mv88e6060_port_to_phy_addr(int port) { - if (port >= 0 && port <= 5) + if (port >= 0 && port < MV88E6060_PORTS) return port; return -1; } From 0cfec916e86d881e209de4b4ae9959a6271e6660 Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Sat, 5 Sep 2015 17:44:13 -0500 Subject: [PATCH 091/335] ipmi: Start the timer and thread on internal msgs The timer and thread were not being started for internal messages, so in interrupt mode if something hung the timer would never go off and clean things up. Factor out the internal message sending and start the timer for those messages, too. Signed-off-by: Corey Minyard Tested-by: Gouji, Masayuki Cc: stable@vger.kernel.org --- drivers/char/ipmi/ipmi_si_intf.c | 73 +++++++++++++++++++------------- 1 file changed, 44 insertions(+), 29 deletions(-) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 654f6f36a071..20c3d7b97602 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -412,18 +412,42 @@ static enum si_sm_result start_next_msg(struct smi_info *smi_info) return rv; } -static void start_check_enables(struct smi_info *smi_info) +static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val) +{ + smi_info->last_timeout_jiffies = jiffies; + mod_timer(&smi_info->si_timer, new_val); + smi_info->timer_running = true; +} + +/* + * Start a new message and (re)start the timer and thread. + */ +static void start_new_msg(struct smi_info *smi_info, unsigned char *msg, + unsigned int size) +{ + smi_mod_timer(smi_info, jiffies + SI_TIMEOUT_JIFFIES); + + if (smi_info->thread) + wake_up_process(smi_info->thread); + + smi_info->handlers->start_transaction(smi_info->si_sm, msg, size); +} + +static void start_check_enables(struct smi_info *smi_info, bool start_timer) { unsigned char msg[2]; msg[0] = (IPMI_NETFN_APP_REQUEST << 2); msg[1] = IPMI_GET_BMC_GLOBAL_ENABLES_CMD; - smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2); + if (start_timer) + start_new_msg(smi_info, msg, 2); + else + smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2); smi_info->si_state = SI_CHECKING_ENABLES; } -static void start_clear_flags(struct smi_info *smi_info) +static void start_clear_flags(struct smi_info *smi_info, bool start_timer) { unsigned char msg[3]; @@ -432,7 +456,10 @@ static void start_clear_flags(struct smi_info *smi_info) msg[1] = IPMI_CLEAR_MSG_FLAGS_CMD; msg[2] = WDT_PRE_TIMEOUT_INT; - smi_info->handlers->start_transaction(smi_info->si_sm, msg, 3); + if (start_timer) + start_new_msg(smi_info, msg, 3); + else + smi_info->handlers->start_transaction(smi_info->si_sm, msg, 3); smi_info->si_state = SI_CLEARING_FLAGS; } @@ -442,10 +469,8 @@ static void start_getting_msg_queue(struct smi_info *smi_info) smi_info->curr_msg->data[1] = IPMI_GET_MSG_CMD; smi_info->curr_msg->data_size = 2; - smi_info->handlers->start_transaction( - smi_info->si_sm, - smi_info->curr_msg->data, - smi_info->curr_msg->data_size); + start_new_msg(smi_info, smi_info->curr_msg->data, + smi_info->curr_msg->data_size); smi_info->si_state = SI_GETTING_MESSAGES; } @@ -455,20 +480,11 @@ static void start_getting_events(struct smi_info *smi_info) smi_info->curr_msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD; smi_info->curr_msg->data_size = 2; - smi_info->handlers->start_transaction( - smi_info->si_sm, - smi_info->curr_msg->data, - smi_info->curr_msg->data_size); + start_new_msg(smi_info, smi_info->curr_msg->data, + smi_info->curr_msg->data_size); smi_info->si_state = SI_GETTING_EVENTS; } -static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val) -{ - smi_info->last_timeout_jiffies = jiffies; - mod_timer(&smi_info->si_timer, new_val); - smi_info->timer_running = true; -} - /* * When we have a situtaion where we run out of memory and cannot * allocate messages, we just leave them in the BMC and run the system @@ -478,11 +494,11 @@ static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val) * Note that we cannot just use disable_irq(), since the interrupt may * be shared. */ -static inline bool disable_si_irq(struct smi_info *smi_info) +static inline bool disable_si_irq(struct smi_info *smi_info, bool start_timer) { if ((smi_info->irq) && (!smi_info->interrupt_disabled)) { smi_info->interrupt_disabled = true; - start_check_enables(smi_info); + start_check_enables(smi_info, start_timer); return true; } return false; @@ -492,7 +508,7 @@ static inline bool enable_si_irq(struct smi_info *smi_info) { if ((smi_info->irq) && (smi_info->interrupt_disabled)) { smi_info->interrupt_disabled = false; - start_check_enables(smi_info); + start_check_enables(smi_info, true); return true; } return false; @@ -510,7 +526,7 @@ static struct ipmi_smi_msg *alloc_msg_handle_irq(struct smi_info *smi_info) msg = ipmi_alloc_smi_msg(); if (!msg) { - if (!disable_si_irq(smi_info)) + if (!disable_si_irq(smi_info, true)) smi_info->si_state = SI_NORMAL; } else if (enable_si_irq(smi_info)) { ipmi_free_smi_msg(msg); @@ -526,7 +542,7 @@ static void handle_flags(struct smi_info *smi_info) /* Watchdog pre-timeout */ smi_inc_stat(smi_info, watchdog_pretimeouts); - start_clear_flags(smi_info); + start_clear_flags(smi_info, true); smi_info->msg_flags &= ~WDT_PRE_TIMEOUT_INT; if (smi_info->intf) ipmi_smi_watchdog_pretimeout(smi_info->intf); @@ -879,8 +895,7 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info, msg[0] = (IPMI_NETFN_APP_REQUEST << 2); msg[1] = IPMI_GET_MSG_FLAGS_CMD; - smi_info->handlers->start_transaction( - smi_info->si_sm, msg, 2); + start_new_msg(smi_info, msg, 2); smi_info->si_state = SI_GETTING_FLAGS; goto restart; } @@ -910,7 +925,7 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info, * disable and messages disabled. */ if (smi_info->supports_event_msg_buff || smi_info->irq) { - start_check_enables(smi_info); + start_check_enables(smi_info, true); } else { smi_info->curr_msg = alloc_msg_handle_irq(smi_info); if (!smi_info->curr_msg) @@ -3613,7 +3628,7 @@ static int try_smi_init(struct smi_info *new_smi) * Start clearing the flags before we enable interrupts or the * timer to avoid racing with the timer. */ - start_clear_flags(new_smi); + start_clear_flags(new_smi, false); /* * IRQ is defined to be set when non-zero. req_events will @@ -3908,7 +3923,7 @@ static void cleanup_one_si(struct smi_info *to_clean) poll(to_clean); schedule_timeout_uninterruptible(1); } - disable_si_irq(to_clean); + disable_si_irq(to_clean, false); while (to_clean->curr_msg || (to_clean->si_state != SI_NORMAL)) { poll(to_clean); schedule_timeout_uninterruptible(1); From 314ef52fe67f8f03453b69169f954e2d04679bbd Mon Sep 17 00:00:00 2001 From: Corey Minyard Date: Sat, 5 Sep 2015 17:58:13 -0500 Subject: [PATCH 092/335] ipmi: Stop the timer immediately if idle The IPMI driver would let the final timeout just happen, but it could easily just stop the timer. If the timer stop fails that's ok, that should be rare. Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_si_intf.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 20c3d7b97602..71b59cb6ae0c 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -935,6 +935,13 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info, } goto restart; } + + if (si_sm_result == SI_SM_IDLE && smi_info->timer_running) { + /* Ok it if fails, the timer will just go off. */ + if (del_timer(&smi_info->si_timer)) + smi_info->timer_running = false; + } + out: return si_sm_result; } From 66f44018300c5e6f53c9d30d6920332cf0e6a8f9 Mon Sep 17 00:00:00 2001 From: Luis de Bethencourt Date: Sat, 19 Sep 2015 16:43:23 +0100 Subject: [PATCH 093/335] char: ipmi: Move MODULE_DEVICE_TABLE() to follow struct The policy for drivers is to have MODULE_DEVICE_TABLE() just after the struct used in it. For clarity. Suggested-by: Corey Minyard Signed-off-by: Luis de Bethencourt Signed-off-by: Corey Minyard --- drivers/char/ipmi/ipmi_si_intf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 71b59cb6ae0c..55fe9020459f 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2582,6 +2582,7 @@ static const struct of_device_id of_ipmi_match[] = { .data = (void *)(unsigned long) SI_BT }, {}, }; +MODULE_DEVICE_TABLE(of, of_ipmi_match); static int of_ipmi_probe(struct platform_device *dev) { @@ -2668,7 +2669,6 @@ static int of_ipmi_probe(struct platform_device *dev) } return 0; } -MODULE_DEVICE_TABLE(of, of_ipmi_match); #else #define of_ipmi_match NULL static int of_ipmi_probe(struct platform_device *dev) From 36fa4a530b7798aa85789953b08d94c03fb09fa5 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Sun, 25 Oct 2015 11:09:33 +0100 Subject: [PATCH 094/335] dmaengine: sh: usb-dmac: Fix crash on runtime suspend If CONFIG_PREEMPT=y: Unable to handle kernel NULL pointer dereference at virtual address 00000014 pgd = c0003000 [00000014] *pgd=80000040004003, *pmd=00000000 Internal error: Oops: 206 [#1] PREEMPT SMP ARM Modules linked in: CPU: 0 PID: 17 Comm: kworker/0:1 Tainted: G W 4.3.0-rc3-koelsch-022 71-g705498fc5e6a5da8-dirty #1789 Hardware name: Generic R8A7791 (Flattened Device Tree) Workqueue: pm pm_runtime_work task: ef578e40 ti: ef57a000 task.ti: ef57a000 PC is at usb_dmac_chan_halt+0xc/0xc0 LR is at usb_dmac_runtime_suspend+0x28/0x38 pc : [] lr : [] psr: 80000113 sp : ef57bdf8 ip : 00000008 fp : 00000003 r10: 00000008 r9 : c06ab928 r8 : ef49e810 r7 : 00000000 r6 : 000000ac r5 : ef770010 r4 : 00000000 r3 : 00000000 r2 : 8ffc2b84 r1 : 00000000 r0 : ef770010 Flags: Nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel Control: 30c5307d Table: 40003000 DAC: fffffffd Process kworker/0:1 (pid: 17, stack limit = 0xef57a210) Stack: (0xef57bdf8 to 0xef57c000) [... [] (usb_dmac_chan_halt) from [] (usb_dmac_runtime_suspend+0x28/0x38) [] (usb_dmac_runtime_suspend) from [] (pm_genpd_runtime_suspend+0x74/0x23c) This happens because usb_dmac_probe() calls pm_runtime_put() before usb_dmac_chan_probe(), leading to the device being suspended before the DMA channels are initialized, causing a NULL pointer dereference. Move the call to pm_runtime_put() to the end of usb_dmac_probe() to fix this. Add a check to usb_dmac_runtime_suspend() to prevent the crash from happening in the error path. Reported-by: Sergei Shtylyov Signed-off-by: Geert Uytterhoeven Tested-by: Yoshihiro Shimoda Signed-off-by: Vinod Koul --- drivers/dma/sh/usb-dmac.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/dma/sh/usb-dmac.c b/drivers/dma/sh/usb-dmac.c index ebd8a5f398b0..f305738b5adf 100644 --- a/drivers/dma/sh/usb-dmac.c +++ b/drivers/dma/sh/usb-dmac.c @@ -679,8 +679,11 @@ static int usb_dmac_runtime_suspend(struct device *dev) struct usb_dmac *dmac = dev_get_drvdata(dev); int i; - for (i = 0; i < dmac->n_channels; ++i) + for (i = 0; i < dmac->n_channels; ++i) { + if (!dmac->channels[i].iomem) + break; usb_dmac_chan_halt(&dmac->channels[i]); + } return 0; } @@ -803,7 +806,6 @@ static int usb_dmac_probe(struct platform_device *pdev) } ret = usb_dmac_init(dmac); - pm_runtime_put(&pdev->dev); if (ret) { dev_err(&pdev->dev, "failed to reset device\n"); @@ -851,10 +853,12 @@ static int usb_dmac_probe(struct platform_device *pdev) if (ret < 0) goto error; + pm_runtime_put(&pdev->dev); return 0; error: of_dma_controller_free(pdev->dev.of_node); + pm_runtime_put(&pdev->dev); pm_runtime_disable(&pdev->dev); return ret; } From bf55555baaf80cdf2cc4176fee02545a07a8ff4a Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Sun, 25 Oct 2015 11:09:34 +0100 Subject: [PATCH 095/335] dmaengine: sh: usb-dmac: Fix pm_runtime_{enable,disable}() imbalance If the call to pm_runtime_get_sync() failed, Runtime PM was left enabled. Signed-off-by: Geert Uytterhoeven Signed-off-by: Vinod Koul --- drivers/dma/sh/usb-dmac.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/dma/sh/usb-dmac.c b/drivers/dma/sh/usb-dmac.c index f305738b5adf..f1bcc2a163b3 100644 --- a/drivers/dma/sh/usb-dmac.c +++ b/drivers/dma/sh/usb-dmac.c @@ -802,7 +802,7 @@ static int usb_dmac_probe(struct platform_device *pdev) ret = pm_runtime_get_sync(&pdev->dev); if (ret < 0) { dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret); - return ret; + goto error_pm; } ret = usb_dmac_init(dmac); @@ -859,6 +859,7 @@ static int usb_dmac_probe(struct platform_device *pdev) error: of_dma_controller_free(pdev->dev.of_node); pm_runtime_put(&pdev->dev); +error_pm: pm_runtime_disable(&pdev->dev); return ret; } From 500404ebcbd074ca11aa0c3fd9a268aa4054fd8b Mon Sep 17 00:00:00 2001 From: Peter Ujfalusi Date: Tue, 3 Nov 2015 12:28:10 +0200 Subject: [PATCH 096/335] dmaengine: of_dma: Correct return code for of_dma_request_slave_channel in case !CONFIG_OF of_dma_request_slave_channel should return either pointer for valid dma_chan or ERR_PTR() error code, NULL is not expected to be returned. Signed-off-by: Peter Ujfalusi Acked-by: Arnd Bergmann Signed-off-by: Vinod Koul --- include/linux/of_dma.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/of_dma.h b/include/linux/of_dma.h index 36112cdd665a..b90d8ec57c1f 100644 --- a/include/linux/of_dma.h +++ b/include/linux/of_dma.h @@ -80,7 +80,7 @@ static inline int of_dma_router_register(struct device_node *np, static inline struct dma_chan *of_dma_request_slave_channel(struct device_node *np, const char *name) { - return NULL; + return ERR_PTR(-ENODEV); } static inline struct dma_chan *of_dma_simple_xlate(struct of_phandle_args *dma_spec, From 638bdc8ce845285e1112f4870aded1638187fc98 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 3 Nov 2015 15:00:57 +0100 Subject: [PATCH 097/335] dmaengine: edma: fix build without CONFIG_OF During the edma rework, a build error was introduced for the case that CONFIG_OF is disabled: drivers/built-in.o: In function `edma_tc_set_pm_state': :(.text+0x43bf0): undefined reference to `of_find_device_by_node' As the edma_tc_set_pm_state() function does nothing in case we are running without OF, this adds an IS_ENABLED() check that turns the function into an empty stub then and avoids the link error. Signed-off-by: Arnd Bergmann Fixes: ca304fa9bb76 ("ARM/dmaengine: edma: Public API to use private struct pointer") Acked-by: Peter Ujfalusi Signed-off-by: Vinod Koul --- drivers/dma/edma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c index 6b03e4e84e6b..c69d22c0b9ba 100644 --- a/drivers/dma/edma.c +++ b/drivers/dma/edma.c @@ -1565,7 +1565,7 @@ static void edma_tc_set_pm_state(struct edma_tc *tc, bool enable) struct platform_device *tc_pdev; int ret; - if (!tc) + if (!IS_ENABLED(CONFIG_OF) || !tc) return; tc_pdev = of_find_device_by_node(tc->node); From f5ea7ad252452f0b461e04a0d0e468f1f6ad8b3b Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 4 Nov 2015 16:38:31 +0300 Subject: [PATCH 098/335] dmaengine: edma: predecence bug in GET_NUM_QDMACH() The current code uses bits 0-2 instead of 4-6 as the comment says. Fixes: 633e42b8c546 ('dmaengine: edma: Get qDMA channel information from HW also') Signed-off-by: Dan Carpenter Acked-by: Peter Ujfalusi Signed-off-by: Vinod Koul --- drivers/dma/edma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c index c69d22c0b9ba..0675e268d577 100644 --- a/drivers/dma/edma.c +++ b/drivers/dma/edma.c @@ -107,7 +107,7 @@ /* CCCFG register */ #define GET_NUM_DMACH(x) (x & 0x7) /* bits 0-2 */ -#define GET_NUM_QDMACH(x) (x & 0x70 >> 4) /* bits 4-6 */ +#define GET_NUM_QDMACH(x) ((x & 0x70) >> 4) /* bits 4-6 */ #define GET_NUM_PAENTRY(x) ((x & 0x7000) >> 12) /* bits 12-14 */ #define GET_NUM_EVQUE(x) ((x & 0x70000) >> 16) /* bits 16-18 */ #define GET_NUM_REGN(x) ((x & 0x300000) >> 20) /* bits 20-21 */ From 29f493dafa6113d28763e9efb853e6972b41ade5 Mon Sep 17 00:00:00 2001 From: Jason Liu Date: Wed, 11 Nov 2015 17:20:49 +0800 Subject: [PATCH 099/335] dmaengine: imx-sdma: remove __init annotation on sdma_event_remap The sdma_probe function will call sdma_event_remap, but sdma_event_remap marked with the __init annotation which make the kbuild complains as the following log: WARNING: drivers/dma/built-in.o(.text+0x56fc): Section mismatch in reference from the function sdma_probe() to the function .init.text:sdma_event_remap() The function sdma_probe() references the function __init sdma_event_remap(). This is often because sdma_probe lacks a __init annotation or the annotation of sdma_event_remap is wrong. Remove the __init annotation on sdma_event_remap to kill this build warning Signed-off-by: Jason Liu Acked-by: Arnd Bergmann Signed-off-by: Vinod Koul --- drivers/dma/imx-sdma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index 7058d58ba588..0f6fd42f55ca 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -1462,7 +1462,7 @@ err_firmware: #define EVENT_REMAP_CELLS 3 -static int __init sdma_event_remap(struct sdma_engine *sdma) +static int sdma_event_remap(struct sdma_engine *sdma) { struct device_node *np = sdma->dev->of_node; struct device_node *gpr_np = of_parse_phandle(np, "gpr", 0); From 268914f4e7a013a4798b5544cce9d9584de99889 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 12 Nov 2015 15:16:53 +0100 Subject: [PATCH 100/335] dmaengine: at_xdmac: use %pad format string for dma_addr_t dma_addr_t may be defined as 32 or 64 bit depending on configuration, so it cannot be printed using the normal format strings, as gcc correctly warns: drivers/dma/at_xdmac.c: In function 'at_xdmac_interleaved_queue_desc': drivers/dma/at_xdmac.c:922:51: warning: format '%x' expects argument of type 'unsigned int', but argument 5 has type 'dma_addr_t {aka long long unsigned int}' [-Wformat=] This changes the format strings to use the special "%pad" format string that prints a dma_addr_t, and changes the arguments so we pass the address by reference as required. Signed-off-by: Arnd Bergmann Signed-off-by: Vinod Koul --- drivers/dma/at_xdmac.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c index b5e132d4bae5..7f039de143f0 100644 --- a/drivers/dma/at_xdmac.c +++ b/drivers/dma/at_xdmac.c @@ -920,8 +920,8 @@ at_xdmac_interleaved_queue_desc(struct dma_chan *chan, desc->lld.mbr_cfg = chan_cc; dev_dbg(chan2dev(chan), - "%s: lld: mbr_sa=0x%08x, mbr_da=0x%08x, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n", - __func__, desc->lld.mbr_sa, desc->lld.mbr_da, + "%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n", + __func__, &desc->lld.mbr_sa, &desc->lld.mbr_da, desc->lld.mbr_ubc, desc->lld.mbr_cfg); /* Chain lld. */ @@ -953,8 +953,8 @@ at_xdmac_prep_interleaved(struct dma_chan *chan, if ((xt->numf > 1) && (xt->frame_size > 1)) return NULL; - dev_dbg(chan2dev(chan), "%s: src=0x%08x, dest=0x%08x, numf=%d, frame_size=%d, flags=0x%lx\n", - __func__, xt->src_start, xt->dst_start, xt->numf, + dev_dbg(chan2dev(chan), "%s: src=%pad, dest=%pad, numf=%d, frame_size=%d, flags=0x%lx\n", + __func__, &xt->src_start, &xt->dst_start, xt->numf, xt->frame_size, flags); src_addr = xt->src_start; @@ -1179,8 +1179,8 @@ static struct at_xdmac_desc *at_xdmac_memset_create_desc(struct dma_chan *chan, desc->lld.mbr_cfg = chan_cc; dev_dbg(chan2dev(chan), - "%s: lld: mbr_da=0x%08x, mbr_ds=0x%08x, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n", - __func__, desc->lld.mbr_da, desc->lld.mbr_ds, desc->lld.mbr_ubc, + "%s: lld: mbr_da=%pad, mbr_ds=%pad, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n", + __func__, &desc->lld.mbr_da, &desc->lld.mbr_ds, desc->lld.mbr_ubc, desc->lld.mbr_cfg); return desc; @@ -1193,8 +1193,8 @@ at_xdmac_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value, struct at_xdmac_chan *atchan = to_at_xdmac_chan(chan); struct at_xdmac_desc *desc; - dev_dbg(chan2dev(chan), "%s: dest=0x%08x, len=%d, pattern=0x%x, flags=0x%lx\n", - __func__, dest, len, value, flags); + dev_dbg(chan2dev(chan), "%s: dest=%pad, len=%d, pattern=0x%x, flags=0x%lx\n", + __func__, &dest, len, value, flags); if (unlikely(!len)) return NULL; @@ -1229,8 +1229,8 @@ at_xdmac_prep_dma_memset_sg(struct dma_chan *chan, struct scatterlist *sgl, /* Prepare descriptors. */ for_each_sg(sgl, sg, sg_len, i) { - dev_dbg(chan2dev(chan), "%s: dest=0x%08x, len=%d, pattern=0x%x, flags=0x%lx\n", - __func__, sg_dma_address(sg), sg_dma_len(sg), + dev_dbg(chan2dev(chan), "%s: dest=%pad, len=%d, pattern=0x%x, flags=0x%lx\n", + __func__, &sg_dma_address(sg), sg_dma_len(sg), value, flags); desc = at_xdmac_memset_create_desc(chan, atchan, sg_dma_address(sg), From 2c5d7407e012721f02741f1adae2b1bdf6ef6449 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 12 Nov 2015 15:18:22 +0100 Subject: [PATCH 101/335] dmaengine: at_hdmac: use %pad format string for dma_addr_t dma_addr_t may be defined as 32 or 64 bit depending on configuration, so it cannot be printed using the normal format strings, as gcc correctly warns: drivers/dma/at_hdmac.c: In function 'atc_prep_dma_interleaved': drivers/dma/at_hdmac.c:731:28: warning: format '%x' expects argument of type 'unsigned int', but argument 4 has type 'dma_addr_t {aka long long unsigned int}' [-Wformat=] This changes the format strings to use the special "%pad" format string that prints a dma_addr_t, and changes the arguments so we pass the address by reference as required. Signed-off-by: Arnd Bergmann Acked-by: Nicolas Ferre Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 20 ++++++++++---------- drivers/dma/at_hdmac_regs.h | 6 +++--- 2 files changed, 13 insertions(+), 13 deletions(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 4e55239c7a30..53d22eb73b56 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -729,8 +729,8 @@ atc_prep_dma_interleaved(struct dma_chan *chan, return NULL; dev_info(chan2dev(chan), - "%s: src=0x%08x, dest=0x%08x, numf=%d, frame_size=%d, flags=0x%lx\n", - __func__, xt->src_start, xt->dst_start, xt->numf, + "%s: src=%pad, dest=%pad, numf=%d, frame_size=%d, flags=0x%lx\n", + __func__, &xt->src_start, &xt->dst_start, xt->numf, xt->frame_size, flags); /* @@ -824,8 +824,8 @@ atc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src, u32 ctrla; u32 ctrlb; - dev_vdbg(chan2dev(chan), "prep_dma_memcpy: d0x%x s0x%x l0x%zx f0x%lx\n", - dest, src, len, flags); + dev_vdbg(chan2dev(chan), "prep_dma_memcpy: d%pad s%pad l0x%zx f0x%lx\n", + &dest, &src, len, flags); if (unlikely(!len)) { dev_dbg(chan2dev(chan), "prep_dma_memcpy: length is zero!\n"); @@ -938,8 +938,8 @@ atc_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value, void __iomem *vaddr; dma_addr_t paddr; - dev_vdbg(chan2dev(chan), "%s: d0x%x v0x%x l0x%zx f0x%lx\n", __func__, - dest, value, len, flags); + dev_vdbg(chan2dev(chan), "%s: d%pad v0x%x l0x%zx f0x%lx\n", __func__, + &dest, value, len, flags); if (unlikely(!len)) { dev_dbg(chan2dev(chan), "%s: length is zero!\n", __func__); @@ -1022,8 +1022,8 @@ atc_prep_dma_memset_sg(struct dma_chan *chan, dma_addr_t dest = sg_dma_address(sg); size_t len = sg_dma_len(sg); - dev_vdbg(chan2dev(chan), "%s: d0x%08x, l0x%zx\n", - __func__, dest, len); + dev_vdbg(chan2dev(chan), "%s: d%pad, l0x%zx\n", + __func__, &dest, len); if (!is_dma_fill_aligned(chan->device, dest, 0, len)) { dev_err(chan2dev(chan), "%s: buffer is not aligned\n", @@ -1439,9 +1439,9 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, unsigned int periods = buf_len / period_len; unsigned int i; - dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@0x%08x - %d (%d/%d)\n", + dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@%pad - %d (%d/%d)\n", direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE", - buf_addr, + &buf_addr, periods, buf_len, period_len); if (unlikely(!atslave || !buf_len || !period_len)) { diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h index d1cfc8c876f9..7f58f06157f6 100644 --- a/drivers/dma/at_hdmac_regs.h +++ b/drivers/dma/at_hdmac_regs.h @@ -385,9 +385,9 @@ static void vdbg_dump_regs(struct at_dma_chan *atchan) {} static void atc_dump_lli(struct at_dma_chan *atchan, struct at_lli *lli) { dev_crit(chan2dev(&atchan->chan_common), - " desc: s0x%x d0x%x ctrl0x%x:0x%x l0x%x\n", - lli->saddr, lli->daddr, - lli->ctrla, lli->ctrlb, lli->dscr); + " desc: s%pad d%pad ctrl0x%x:0x%x l0x%pad\n", + &lli->saddr, &lli->daddr, + lli->ctrla, lli->ctrlb, &lli->dscr); } From 92e11f002cac5f6d60bc7498d8d33e6a5fb9cb6f Mon Sep 17 00:00:00 2001 From: Li Yang Date: Thu, 5 Nov 2015 14:18:17 -0600 Subject: [PATCH 102/335] hwmon: (ina2xx) Fix build issue by selecting REGMAP_I2C Since a0de56c81fcf ("hwmon: (ina2xx) convert driver to using regmap") the driver requires REGMAP_I2C to build. Select it by default in Kconfig. Reported-by: Guo Chunrong Cc: Marc Titinger Signed-off-by: Li Yang Fixes: a0de56c81fcf ("hwmon: (ina2xx) convert driver to using regmap") Signed-off-by: Guenter Roeck --- drivers/hwmon/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig index 842b0043ad94..6e4e0464e776 100644 --- a/drivers/hwmon/Kconfig +++ b/drivers/hwmon/Kconfig @@ -1471,6 +1471,7 @@ config SENSORS_INA209 config SENSORS_INA2XX tristate "Texas Instruments INA219 and compatibles" depends on I2C + select REGMAP_I2C help If you say yes here you get support for INA219, INA220, INA226, INA230, and INA231 power monitor chips. From 5e0a0ee4d397665e5a509ed03ed9c41727c5f228 Mon Sep 17 00:00:00 2001 From: Shuah Khan Date: Thu, 5 Nov 2015 15:39:27 -0700 Subject: [PATCH 103/335] hwmon : (applesmc) Fix uninitialized variables warnings MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the following "maybe used uninitialized" warnings by initializing the variables to keep the compiler quiet. There is no "used uninitialized" in this case. CC [M] drivers/hwmon/applesmc.o drivers/hwmon/applesmc.c: In function ‘applesmc_init_smcreg’: drivers/hwmon/applesmc.c:595:43: warning: ‘right_light_sensor’ may be used uninitialized in this function [-Wmaybe-uninitialized] s->num_light_sensors = left_light_sensor + right_light_sensor; ^ drivers/hwmon/applesmc.c:540:26: note: ‘right_light_sensor’ was declared here bool left_light_sensor, right_light_sensor; ^ drivers/hwmon/applesmc.c:595:43: warning: ‘left_light_sensor’ may be used uninitialized in this function [-Wmaybe-uninitialized] s->num_light_sensors = left_light_sensor + right_light_sensor; ^ drivers/hwmon/applesmc.c:540:7: note: ‘left_light_sensor’ was declared here bool left_light_sensor, right_light_sensor; ^ Signed-off-by: Shuah Khan Signed-off-by: Guenter Roeck --- drivers/hwmon/applesmc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hwmon/applesmc.c b/drivers/hwmon/applesmc.c index 1f5e956941b1..0af7fd311979 100644 --- a/drivers/hwmon/applesmc.c +++ b/drivers/hwmon/applesmc.c @@ -537,7 +537,7 @@ static int applesmc_init_index(struct applesmc_registers *s) static int applesmc_init_smcreg_try(void) { struct applesmc_registers *s = &smcreg; - bool left_light_sensor, right_light_sensor; + bool left_light_sensor = 0, right_light_sensor = 0; unsigned int count; u8 tmp[1]; int ret; From 1451ad03fac3e86948728e0b11f63dee73d3106c Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Wed, 11 Nov 2015 21:54:26 +1100 Subject: [PATCH 104/335] powerpc: Wire up sys_mlock2() The selftest passes on 64-bit LE and 32-bit BE. Signed-off-by: Michael Ellerman --- arch/powerpc/include/asm/systbl.h | 1 + arch/powerpc/include/asm/unistd.h | 2 +- arch/powerpc/include/uapi/asm/unistd.h | 1 + 3 files changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/include/asm/systbl.h b/arch/powerpc/include/asm/systbl.h index c9e26cb264f4..f2b0b1b0c72a 100644 --- a/arch/powerpc/include/asm/systbl.h +++ b/arch/powerpc/include/asm/systbl.h @@ -382,3 +382,4 @@ COMPAT_SYS(shmat) SYSCALL(shmdt) SYSCALL(shmget) COMPAT_SYS(shmctl) +SYSCALL(mlock2) diff --git a/arch/powerpc/include/asm/unistd.h b/arch/powerpc/include/asm/unistd.h index 6d8f8023ac27..4b6b8ace18e0 100644 --- a/arch/powerpc/include/asm/unistd.h +++ b/arch/powerpc/include/asm/unistd.h @@ -12,7 +12,7 @@ #include -#define __NR_syscalls 378 +#define __NR_syscalls 379 #define __NR__exit __NR_exit #define NR_syscalls __NR_syscalls diff --git a/arch/powerpc/include/uapi/asm/unistd.h b/arch/powerpc/include/uapi/asm/unistd.h index 81579e93c659..1effea5193d6 100644 --- a/arch/powerpc/include/uapi/asm/unistd.h +++ b/arch/powerpc/include/uapi/asm/unistd.h @@ -400,5 +400,6 @@ #define __NR_shmdt 375 #define __NR_shmget 376 #define __NR_shmctl 377 +#define __NR_mlock2 378 #endif /* _UAPI_ASM_POWERPC_UNISTD_H_ */ From bd1c6ff74ce0bbd8cda6eb7763fa0e2625dfcc8b Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Wed, 4 Nov 2015 13:23:52 +0000 Subject: [PATCH 105/335] arm64/dma-mapping: Fix sizes in __iommu_{alloc,free}_attrs The iommu-dma layer does its own size-alignment for coherent DMA allocations based on IOMMU page sizes, but we still need to consider CPU page sizes for the cases where a non-cacheable CPU mapping is created. Whilst everything on the alloc/map path seems to implicitly align things enough to make it work, some functions used by the corresponding unmap/free path do not, which leads to problems freeing odd-sized allocations. Either way it's something we really should be handling explicitly, so do that to make both paths suitably robust. Reported-by: Yong Wu Signed-off-by: Robin Murphy Signed-off-by: Catalin Marinas --- arch/arm64/mm/dma-mapping.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/arch/arm64/mm/dma-mapping.c b/arch/arm64/mm/dma-mapping.c index 131a199114b4..97fd714035f9 100644 --- a/arch/arm64/mm/dma-mapping.c +++ b/arch/arm64/mm/dma-mapping.c @@ -552,10 +552,14 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size, { bool coherent = is_device_dma_coherent(dev); int ioprot = dma_direction_to_prot(DMA_BIDIRECTIONAL, coherent); + size_t iosize = size; void *addr; if (WARN(!dev, "cannot create IOMMU mapping for unknown device\n")) return NULL; + + size = PAGE_ALIGN(size); + /* * Some drivers rely on this, and we probably don't want the * possibility of stale kernel data being read by devices anyway. @@ -566,7 +570,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size, struct page **pages; pgprot_t prot = __get_dma_pgprot(attrs, PAGE_KERNEL, coherent); - pages = iommu_dma_alloc(dev, size, gfp, ioprot, handle, + pages = iommu_dma_alloc(dev, iosize, gfp, ioprot, handle, flush_page); if (!pages) return NULL; @@ -574,7 +578,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size, addr = dma_common_pages_remap(pages, size, VM_USERMAP, prot, __builtin_return_address(0)); if (!addr) - iommu_dma_free(dev, pages, size, handle); + iommu_dma_free(dev, pages, iosize, handle); } else { struct page *page; /* @@ -591,7 +595,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size, if (!addr) return NULL; - *handle = iommu_dma_map_page(dev, page, 0, size, ioprot); + *handle = iommu_dma_map_page(dev, page, 0, iosize, ioprot); if (iommu_dma_mapping_error(dev, *handle)) { if (coherent) __free_pages(page, get_order(size)); @@ -606,6 +610,9 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size, static void __iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr, dma_addr_t handle, struct dma_attrs *attrs) { + size_t iosize = size; + + size = PAGE_ALIGN(size); /* * @cpu_addr will be one of 3 things depending on how it was allocated: * - A remapped array of pages from iommu_dma_alloc(), for all @@ -617,17 +624,17 @@ static void __iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr, * Hence how dodgy the below logic looks... */ if (__in_atomic_pool(cpu_addr, size)) { - iommu_dma_unmap_page(dev, handle, size, 0, NULL); + iommu_dma_unmap_page(dev, handle, iosize, 0, NULL); __free_from_pool(cpu_addr, size); } else if (is_vmalloc_addr(cpu_addr)){ struct vm_struct *area = find_vm_area(cpu_addr); if (WARN_ON(!area || !area->pages)) return; - iommu_dma_free(dev, area->pages, size, &handle); + iommu_dma_free(dev, area->pages, iosize, &handle); dma_common_free_remap(cpu_addr, size, VM_USERMAP); } else { - iommu_dma_unmap_page(dev, handle, size, 0, NULL); + iommu_dma_unmap_page(dev, handle, iosize, 0, NULL); __free_pages(virt_to_page(cpu_addr), get_order(size)); } } From c7e8b2c21c6a6fd88022ae64f997ebc574036067 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 10 Nov 2015 12:30:28 +0100 Subject: [PATCH 106/335] s390: avoid cache aliasing under z/VM and KVM commit 1f6b83e5e4d3 ("s390: avoid z13 cache aliasing") checks for the machine type to optimize address space randomization and zero page allocation to avoid cache aliases. This check might fail under a hypervisor with migration support. z/VMs "Single System Image and Live Guest Relocation" facility will "fake" the machine type of the oldest system in the group. For example in a group of zEC12 and Z13 the guest appears to run on a zEC12 (architecture fencing within the relocation domain) Remove the machine type detection and always use cache aliasing rules that are known to work for all machines. These are the z13 aliasing rules. Suggested-by: Christian Borntraeger Reviewed-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/elf.h | 13 ++++++-- arch/s390/kernel/process.c | 6 +--- arch/s390/mm/init.c | 30 ++----------------- arch/s390/mm/mmap.c | 60 ++++++------------------------------- 4 files changed, 23 insertions(+), 86 deletions(-) diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 3ad48f22de78..bab6739a1154 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -206,9 +206,16 @@ do { \ } while (0) #endif /* CONFIG_COMPAT */ -extern unsigned long mmap_rnd_mask; - -#define STACK_RND_MASK (test_thread_flag(TIF_31BIT) ? 0x7ff : mmap_rnd_mask) +/* + * Cache aliasing on the latest machines calls for a mapping granularity + * of 512KB. For 64-bit processes use a 512KB alignment and a randomization + * of up to 1GB. For 31-bit processes the virtual address space is limited, + * use no alignment and limit the randomization to 8MB. + */ +#define BRK_RND_MASK (is_32bit_task() ? 0x7ffUL : 0x3ffffUL) +#define MMAP_RND_MASK (is_32bit_task() ? 0x7ffUL : 0x3ff80UL) +#define MMAP_ALIGN_MASK (is_32bit_task() ? 0 : 0x7fUL) +#define STACK_RND_MASK MMAP_RND_MASK #define ARCH_DLINFO \ do { \ diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index 688a3aad9c79..114ee8b96f17 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -243,11 +243,7 @@ unsigned long arch_align_stack(unsigned long sp) static inline unsigned long brk_rnd(void) { - /* 8MB for 32bit, 1GB for 64bit */ - if (is_32bit_task()) - return (get_random_int() & 0x7ffUL) << PAGE_SHIFT; - else - return (get_random_int() & 0x3ffffUL) << PAGE_SHIFT; + return (get_random_int() & BRK_RND_MASK) << PAGE_SHIFT; } unsigned long arch_randomize_brk(struct mm_struct *mm) diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index c3c07d3505ba..c722400c7697 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -48,37 +48,13 @@ EXPORT_SYMBOL(zero_page_mask); static void __init setup_zero_pages(void) { - struct cpuid cpu_id; unsigned int order; struct page *page; int i; - get_cpu_id(&cpu_id); - switch (cpu_id.machine) { - case 0x9672: /* g5 */ - case 0x2064: /* z900 */ - case 0x2066: /* z900 */ - case 0x2084: /* z990 */ - case 0x2086: /* z990 */ - case 0x2094: /* z9-109 */ - case 0x2096: /* z9-109 */ - order = 0; - break; - case 0x2097: /* z10 */ - case 0x2098: /* z10 */ - case 0x2817: /* z196 */ - case 0x2818: /* z196 */ - order = 2; - break; - case 0x2827: /* zEC12 */ - case 0x2828: /* zEC12 */ - order = 5; - break; - case 0x2964: /* z13 */ - default: - order = 7; - break; - } + /* Latest machines require a mapping granularity of 512KB */ + order = 7; + /* Limit number of empty zero pages for small memory sizes */ while (order > 2 && (totalram_pages >> 10) < (1UL << order)) order--; diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c index 6e552af08c76..ea01477b4aa6 100644 --- a/arch/s390/mm/mmap.c +++ b/arch/s390/mm/mmap.c @@ -31,9 +31,6 @@ #include #include -unsigned long mmap_rnd_mask; -static unsigned long mmap_align_mask; - static unsigned long stack_maxrandom_size(void) { if (!(current->flags & PF_RANDOMIZE)) @@ -62,10 +59,7 @@ static inline int mmap_is_legacy(void) unsigned long arch_mmap_rnd(void) { - if (is_32bit_task()) - return (get_random_int() & 0x7ff) << PAGE_SHIFT; - else - return (get_random_int() & mmap_rnd_mask) << PAGE_SHIFT; + return (get_random_int() & MMAP_RND_MASK) << PAGE_SHIFT; } static unsigned long mmap_base_legacy(unsigned long rnd) @@ -92,7 +86,6 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, struct mm_struct *mm = current->mm; struct vm_area_struct *vma; struct vm_unmapped_area_info info; - int do_color_align; if (len > TASK_SIZE - mmap_min_addr) return -ENOMEM; @@ -108,15 +101,14 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, return addr; } - do_color_align = 0; - if (filp || (flags & MAP_SHARED)) - do_color_align = !is_32bit_task(); - info.flags = 0; info.length = len; info.low_limit = mm->mmap_base; info.high_limit = TASK_SIZE; - info.align_mask = do_color_align ? (mmap_align_mask << PAGE_SHIFT) : 0; + if (filp || (flags & MAP_SHARED)) + info.align_mask = MMAP_ALIGN_MASK << PAGE_SHIFT; + else + info.align_mask = 0; info.align_offset = pgoff << PAGE_SHIFT; return vm_unmapped_area(&info); } @@ -130,7 +122,6 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, struct mm_struct *mm = current->mm; unsigned long addr = addr0; struct vm_unmapped_area_info info; - int do_color_align; /* requested length too big for entire address space */ if (len > TASK_SIZE - mmap_min_addr) @@ -148,15 +139,14 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, return addr; } - do_color_align = 0; - if (filp || (flags & MAP_SHARED)) - do_color_align = !is_32bit_task(); - info.flags = VM_UNMAPPED_AREA_TOPDOWN; info.length = len; info.low_limit = max(PAGE_SIZE, mmap_min_addr); info.high_limit = mm->mmap_base; - info.align_mask = do_color_align ? (mmap_align_mask << PAGE_SHIFT) : 0; + if (filp || (flags & MAP_SHARED)) + info.align_mask = MMAP_ALIGN_MASK << PAGE_SHIFT; + else + info.align_mask = 0; info.align_offset = pgoff << PAGE_SHIFT; addr = vm_unmapped_area(&info); @@ -254,35 +244,3 @@ void arch_pick_mmap_layout(struct mm_struct *mm) mm->get_unmapped_area = s390_get_unmapped_area_topdown; } } - -static int __init setup_mmap_rnd(void) -{ - struct cpuid cpu_id; - - get_cpu_id(&cpu_id); - switch (cpu_id.machine) { - case 0x9672: - case 0x2064: - case 0x2066: - case 0x2084: - case 0x2086: - case 0x2094: - case 0x2096: - case 0x2097: - case 0x2098: - case 0x2817: - case 0x2818: - case 0x2827: - case 0x2828: - mmap_rnd_mask = 0x7ffUL; - mmap_align_mask = 0UL; - break; - case 0x2964: /* z13 */ - default: - mmap_rnd_mask = 0x3ff80UL; - mmap_align_mask = 0x7fUL; - break; - } - return 0; -} -early_initcall(setup_mmap_rnd); From ddfd4a054b91def32cd1fe214f0a4d5506bb553b Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 13 Nov 2015 12:45:12 +0100 Subject: [PATCH 107/335] s390: remove g5 elf platform support Remove dead code, since this could only happen on a 31 bit machine where the kernel wouldn't IPL. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/setup.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index ce0cbd6ba7ca..c837bcacf218 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -764,9 +764,6 @@ static int __init setup_hwcaps(void) get_cpu_id(&cpu_id); add_device_randomness(&cpu_id, sizeof(cpu_id)); switch (cpu_id.machine) { - case 0x9672: - strcpy(elf_platform, "g5"); - break; case 0x2064: case 0x2066: default: /* Use "z900" as default for 64 bit kernels. */ From 95486e4979e56e7da2fbb4fd32eb54d672b1e074 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Mon, 16 Nov 2015 12:07:10 +0100 Subject: [PATCH 108/335] MIPS: Fix flood of warnings about comparsion being always true. ./arch/mips/include/asm/page.h:204:13: warning: comparison of unsigned expression >= 0 is always true [-Wtype-limits] The default value of ARCH_PFN_OFFSET is 0 thus triggering this warning for all platforms using the default value. Signed-off-by: Ralf Baechle --- arch/mips/include/asm/page.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/mips/include/asm/page.h b/arch/mips/include/asm/page.h index ad1fccdb8d13..2046c0230224 100644 --- a/arch/mips/include/asm/page.h +++ b/arch/mips/include/asm/page.h @@ -200,8 +200,9 @@ static inline int pfn_valid(unsigned long pfn) { /* avoid include hell */ extern unsigned long max_mapnr; + unsigned long pfn_offset = ARCH_PFN_OFFSET; - return pfn >= ARCH_PFN_OFFSET && pfn < max_mapnr; + return pfn >= pfn_offset && pfn < max_mapnr; } #elif defined(CONFIG_SPARSEMEM) From 932f608193cdbd3a275a4aefb94760dfd4a1f736 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 16 Nov 2015 12:31:33 +0100 Subject: [PATCH 109/335] s390: wire up mlock2 system call Passes mlock2-tests test case in 64 bit and compat mode. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/uapi/asm/unistd.h | 3 ++- arch/s390/kernel/compat_wrapper.c | 1 + arch/s390/kernel/syscalls.S | 1 + 3 files changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/s390/include/uapi/asm/unistd.h b/arch/s390/include/uapi/asm/unistd.h index c55de84ab1ad..34ec202472c6 100644 --- a/arch/s390/include/uapi/asm/unistd.h +++ b/arch/s390/include/uapi/asm/unistd.h @@ -309,7 +309,8 @@ #define __NR_recvfrom 371 #define __NR_recvmsg 372 #define __NR_shutdown 373 -#define NR_syscalls 374 +#define __NR_mlock2 374 +#define NR_syscalls 375 /* * There are some system calls that are not present on 64 bit, some diff --git a/arch/s390/kernel/compat_wrapper.c b/arch/s390/kernel/compat_wrapper.c index 09f194052df3..fac4eeddef91 100644 --- a/arch/s390/kernel/compat_wrapper.c +++ b/arch/s390/kernel/compat_wrapper.c @@ -176,3 +176,4 @@ COMPAT_SYSCALL_WRAP4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr, COMPAT_SYSCALL_WRAP3(getsockname, int, fd, struct sockaddr __user *, usockaddr, int __user *, usockaddr_len); COMPAT_SYSCALL_WRAP3(getpeername, int, fd, struct sockaddr __user *, usockaddr, int __user *, usockaddr_len); COMPAT_SYSCALL_WRAP6(sendto, int, fd, void __user *, buff, size_t, len, unsigned int, flags, struct sockaddr __user *, addr, int, addr_len); +COMPAT_SYSCALL_WRAP3(mlock2, unsigned long, start, size_t, len, int, flags); diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S index 8c56929c8d82..5378c3ea1b98 100644 --- a/arch/s390/kernel/syscalls.S +++ b/arch/s390/kernel/syscalls.S @@ -382,3 +382,4 @@ SYSCALL(sys_sendmsg,compat_sys_sendmsg) /* 370 */ SYSCALL(sys_recvfrom,compat_sys_recvfrom) SYSCALL(sys_recvmsg,compat_sys_recvmsg) SYSCALL(sys_shutdown,sys_shutdown) +SYSCALL(sys_mlock2,compat_sys_mlock2) From f52c74fee95f1f4dd2bc1c75e016d849150eb2de Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 13 Nov 2015 14:17:14 +0100 Subject: [PATCH 110/335] s390: remove SALIPL loader There is no known user, therefore remove the code. Acked-by: Rob Van Der Heij Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/head.S | 87 ++++------------------------------------- 1 file changed, 8 insertions(+), 79 deletions(-) diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index 1b0f624cebd7..301ee9c70688 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -60,19 +60,6 @@ __HEAD .long 0x020006e0,0x20000050 .org 0x200 -# -# subroutine to set architecture mode -# -.Lsetmode: - mvi __LC_AR_MODE_ID,1 # set esame flag - slr %r0,%r0 # set cpuid to zero - lhi %r1,2 # mode 2 = esame (dump) - sigp %r1,%r0,0x12 # switch to esame mode - bras %r13,0f - .fill 16,4,0x0 -0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs - sam31 # switch to 31 bit addressing mode - br %r14 # # subroutine to wait for end I/O @@ -160,7 +147,14 @@ __HEAD .long 0x02200050,0x00000000 iplstart: - bas %r14,.Lsetmode # Immediately switch to 64 bit mode + mvi __LC_AR_MODE_ID,1 # set esame flag + slr %r0,%r0 # set cpuid to zero + lhi %r1,2 # mode 2 = esame (dump) + sigp %r1,%r0,0x12 # switch to esame mode + bras %r13,0f + .fill 16,4,0x0 +0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs + sam31 # switch to 31 bit addressing mode lh %r1,0xb8 # test if subchannel number bct %r1,.Lnoload # is valid l %r1,0xb8 # load ipl subchannel number @@ -269,71 +263,6 @@ iplstart: .align 8 .Lcpuid:.fill 8,1,0 -# -# SALIPL loader support. Based on a patch by Rob van der Heij. -# This entry point is called directly from the SALIPL loader and -# doesn't need a builtin ipl record. -# - .org 0x800 -ENTRY(start) - stm %r0,%r15,0x07b0 # store registers - bas %r14,.Lsetmode # Immediately switch to 64 bit mode - basr %r12,%r0 -.base: - l %r11,.parm - l %r8,.cmd # pointer to command buffer - - ltr %r9,%r9 # do we have SALIPL parameters? - bp .sk8x8 - - mvc 0(64,%r8),0x00b0 # copy saved registers - xc 64(240-64,%r8),0(%r8) # remainder of buffer - tr 0(64,%r8),.lowcase - b .gotr -.sk8x8: - mvc 0(240,%r8),0(%r9) # copy iplparms into buffer -.gotr: - slr %r0,%r0 - st %r0,INITRD_SIZE+ARCH_OFFSET-PARMAREA(%r11) - st %r0,INITRD_START+ARCH_OFFSET-PARMAREA(%r11) - j startup # continue with startup -.cmd: .long COMMAND_LINE # address of command line buffer -.parm: .long PARMAREA -.lowcase: - .byte 0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07 - .byte 0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f - .byte 0x10,0x11,0x12,0x13,0x14,0x15,0x16,0x17 - .byte 0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f - .byte 0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27 - .byte 0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f - .byte 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37 - .byte 0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f - .byte 0x40,0x41,0x42,0x43,0x44,0x45,0x46,0x47 - .byte 0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f - .byte 0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57 - .byte 0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f - .byte 0x60,0x61,0x62,0x63,0x64,0x65,0x66,0x67 - .byte 0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f - .byte 0x70,0x71,0x72,0x73,0x74,0x75,0x76,0x77 - .byte 0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f - - .byte 0x80,0x81,0x82,0x83,0x84,0x85,0x86,0x87 - .byte 0x88,0x89,0x8a,0x8b,0x8c,0x8d,0x8e,0x8f - .byte 0x90,0x91,0x92,0x93,0x94,0x95,0x96,0x97 - .byte 0x98,0x99,0x9a,0x9b,0x9c,0x9d,0x9e,0x9f - .byte 0xa0,0xa1,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7 - .byte 0xa8,0xa9,0xaa,0xab,0xac,0xad,0xae,0xaf - .byte 0xb0,0xb1,0xb2,0xb3,0xb4,0xb5,0xb6,0xb7 - .byte 0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf - .byte 0xc0,0x81,0x82,0x83,0x84,0x85,0x86,0x87 # .abcdefg - .byte 0x88,0x89,0xca,0xcb,0xcc,0xcd,0xce,0xcf # hi - .byte 0xd0,0x91,0x92,0x93,0x94,0x95,0x96,0x97 # .jklmnop - .byte 0x98,0x99,0xda,0xdb,0xdc,0xdd,0xde,0xdf # qr - .byte 0xe0,0xe1,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7 # ..stuvwx - .byte 0xa8,0xa9,0xea,0xeb,0xec,0xed,0xee,0xef # yz - .byte 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7 - .byte 0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff - # # startup-code at 0x10000, running in absolute addressing mode # this is called either by the ipl loader or directly by PSW restart From c7f42c63901b964833eb23a9bda873b799e7f308 Mon Sep 17 00:00:00 2001 From: Jean-Yves Faye Date: Tue, 29 Sep 2015 11:39:19 +0200 Subject: [PATCH 111/335] ipmi watchdog : add panic_wdt_timeout parameter In order to allow panic actions to be processed, the ipmi watchdog driver sets a new timeout value on panic. The 255s timeout was designed to allow kdump and others actions on panic, as in http://lkml.iu.edu/hypermail/linux/kernel/0711.3/0258.html This is counter-intuitive for a end-user who sets watchdog timeout value to something like 30s and who expects BMC to reset the system within 30s of a panic. This commit allows user to configure the timeout on panic. Signed-off-by: Jean-Yves Faye Signed-off-by: Corey Minyard --- Documentation/IPMI.txt | 7 +++++-- drivers/char/ipmi/ipmi_watchdog.c | 8 +++++++- 2 files changed, 12 insertions(+), 3 deletions(-) diff --git a/Documentation/IPMI.txt b/Documentation/IPMI.txt index 31d1d658827f..c0d8788e75d3 100644 --- a/Documentation/IPMI.txt +++ b/Documentation/IPMI.txt @@ -587,7 +587,7 @@ used to control it: modprobe ipmi_watchdog timeout= pretimeout= action= preaction= preop= start_now=x - nowayout=x ifnum_to_use=n + nowayout=x ifnum_to_use=n panic_wdt_timeout= ifnum_to_use specifies which interface the watchdog timer should use. The default is -1, which means to pick the first one registered. @@ -597,7 +597,9 @@ is the amount of seconds before the reset that the pre-timeout panic will occur (if pretimeout is zero, then pretimeout will not be enabled). Note that the pretimeout is the time before the final timeout. So if the timeout is 50 seconds and the pretimeout is 10 seconds, then the pretimeout -will occur in 40 second (10 seconds before the timeout). +will occur in 40 second (10 seconds before the timeout). The panic_wdt_timeout +is the value of timeout which is set on kernel panic, in order to let actions +such as kdump to occur during panic. The action may be "reset", "power_cycle", or "power_off", and specifies what to do when the timer times out, and defaults to @@ -634,6 +636,7 @@ for configuring the watchdog: ipmi_watchdog.preop= ipmi_watchdog.start_now=x ipmi_watchdog.nowayout=x + ipmi_watchdog.panic_wdt_timeout= The options are the same as the module parameter options. diff --git a/drivers/char/ipmi/ipmi_watchdog.c b/drivers/char/ipmi/ipmi_watchdog.c index 0ac3bd1a5497..096f0cef4da1 100644 --- a/drivers/char/ipmi/ipmi_watchdog.c +++ b/drivers/char/ipmi/ipmi_watchdog.c @@ -153,6 +153,9 @@ static int timeout = 10; /* The pre-timeout is disabled by default. */ static int pretimeout; +/* Default timeout to set on panic */ +static int panic_wdt_timeout = 255; + /* Default action is to reset the board on a timeout. */ static unsigned char action_val = WDOG_TIMEOUT_RESET; @@ -293,6 +296,9 @@ MODULE_PARM_DESC(timeout, "Timeout value in seconds."); module_param(pretimeout, timeout, 0644); MODULE_PARM_DESC(pretimeout, "Pretimeout value in seconds."); +module_param(panic_wdt_timeout, timeout, 0644); +MODULE_PARM_DESC(timeout, "Timeout value on kernel panic in seconds."); + module_param_cb(action, ¶m_ops_str, action_op, 0644); MODULE_PARM_DESC(action, "Timeout action. One of: " "reset, none, power_cycle, power_off."); @@ -1189,7 +1195,7 @@ static int wdog_panic_handler(struct notifier_block *this, /* Make sure we do this only once. */ panic_event_handled = 1; - timeout = 255; + timeout = panic_wdt_timeout; pretimeout = 0; panic_halt_ipmi_set_timeout(); } From f44e26593aea9b920e892fc490ffd2a1a0b9123f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Fri, 13 Nov 2015 19:16:13 +0200 Subject: [PATCH 112/335] drm/i915: Fix crtc_y assignment in intel_find_initial_plane_obj() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Let's set crtc_y to 0 instead of setting src_y twice. Multiple assignments in one statement is a good way to hide bugs. Please don't do that. Cc: Maarten Lankhorst Fixes: be5651f2d581 ("drm/i915: Update missing properties in find_initial_plane_obj") Signed-off-by: Ville Syrjälä Reviewed-by: Maarten Lankhorst Link: http://patchwork.freedesktop.org/patch/msgid/1447434973-12369-1-git-send-email-ville.syrjala@linux.intel.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/intel_display.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f62ffc04c21d..b0f23692346e 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2646,11 +2646,13 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc, return; valid_fb: - plane_state->src_x = plane_state->src_y = 0; + plane_state->src_x = 0; + plane_state->src_y = 0; plane_state->src_w = fb->width << 16; plane_state->src_h = fb->height << 16; - plane_state->crtc_x = plane_state->src_y = 0; + plane_state->crtc_x = 0; + plane_state->crtc_y = 0; plane_state->crtc_w = fb->width; plane_state->crtc_h = fb->height; From 1b9448b071caa7d10bb2569fabe3020a2c25ae59 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 5 Nov 2015 11:49:59 +0200 Subject: [PATCH 113/335] drm/i915: quirk backlight present on Macbook 4, 1 Unsurprisingly macbooks have backlights, just the VBT doesn't seem to know it in this case. Reported-and-tested-by: Daniel Nicoletti Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88325 Fixes: c675949ec58c ("drm/i915: do not setup backlight if not available according to VBT") Cc: stable@vger.kernel.org # v3.15+ Reviewed-by: Ander Conselvan de Oliveira Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1446716999-1796-1-git-send-email-jani.nikula@intel.com --- drivers/gpu/drm/i915/intel_display.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index b0f23692346e..bcf227356f78 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14707,6 +14707,9 @@ static struct intel_quirk intel_quirks[] = { /* Apple Macbook 2,1 (Core 2 T7400) */ { 0x27a2, 0x8086, 0x7270, quirk_backlight_present }, + /* Apple Macbook 4,1 */ + { 0x2a02, 0x106b, 0x00a1, quirk_backlight_present }, + /* Toshiba CB35 Chromebook (Celeron 2955U) */ { 0x0a06, 0x1179, 0x0a88, quirk_backlight_present }, From 9d5b5ed796d7afd7e8d2ac4b4fb77c6a49463f4b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Petr=20=C5=A0tetiar?= Date: Tue, 3 Nov 2015 11:25:28 +0100 Subject: [PATCH 114/335] USB: qcserial: Add support for Quectel EC20 Mini PCIe module MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It seems like this device has same vendor and product IDs as G2K devices, but it has different number of interfaces(4 vs 5) and also different interface layout which makes it currently unusable: usbcore: registered new interface driver qcserial usbserial: USB Serial support registered for Qualcomm USB modem usb 2-1.2: unknown number of interfaces: 5 lsusb output: Bus 002 Device 003: ID 05c6:9215 Qualcomm, Inc. Acer Gobi 2000 Wireless Device Descriptor: bLength 18 bDescriptorType 1 bcdUSB 2.00 bDeviceClass 0 (Defined at Interface level) bDeviceSubClass 0 bDeviceProtocol 0 bMaxPacketSize0 64 idVendor 0x05c6 Qualcomm, Inc. idProduct 0x9215 Acer Gobi 2000 Wireless Modem bcdDevice 2.32 iManufacturer 1 Quectel iProduct 2 Quectel LTE Module iSerial 0 bNumConfigurations 1 Configuration Descriptor: bLength 9 bDescriptorType 2 wTotalLength 209 bNumInterfaces 5 bConfigurationValue 1 iConfiguration 0 bmAttributes 0xa0 (Bus Powered) Remote Wakeup MaxPower 500mA Signed-off-by: Petr Štetiar Cc: stable [johan: rename define and add comment ] Signed-off-by: Johan Hovold --- drivers/usb/serial/qcserial.c | 40 +++++++++++++++++++++++++++++++++++ 1 file changed, 40 insertions(+) diff --git a/drivers/usb/serial/qcserial.c b/drivers/usb/serial/qcserial.c index 5022fcfa0260..0e46af0d0cb4 100644 --- a/drivers/usb/serial/qcserial.c +++ b/drivers/usb/serial/qcserial.c @@ -22,6 +22,8 @@ #define DRIVER_AUTHOR "Qualcomm Inc" #define DRIVER_DESC "Qualcomm USB Serial driver" +#define QUECTEL_EC20_PID 0x9215 + /* standard device layouts supported by this driver */ enum qcserial_layouts { QCSERIAL_G2K = 0, /* Gobi 2000 */ @@ -171,6 +173,38 @@ static const struct usb_device_id id_table[] = { }; MODULE_DEVICE_TABLE(usb, id_table); +static int handle_quectel_ec20(struct device *dev, int ifnum) +{ + int altsetting = 0; + + /* + * Quectel EC20 Mini PCIe LTE module layout: + * 0: DM/DIAG (use libqcdm from ModemManager for communication) + * 1: NMEA + * 2: AT-capable modem port + * 3: Modem interface + * 4: NDIS + */ + switch (ifnum) { + case 0: + dev_dbg(dev, "Quectel EC20 DM/DIAG interface found\n"); + break; + case 1: + dev_dbg(dev, "Quectel EC20 NMEA GPS interface found\n"); + break; + case 2: + case 3: + dev_dbg(dev, "Quectel EC20 Modem port found\n"); + break; + case 4: + /* Don't claim the QMI/net interface */ + altsetting = -1; + break; + } + + return altsetting; +} + static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) { struct usb_host_interface *intf = serial->interface->cur_altsetting; @@ -240,6 +274,12 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) altsetting = -1; break; case QCSERIAL_G2K: + /* handle non-standard layouts */ + if (nintf == 5 && id->idProduct == QUECTEL_EC20_PID) { + altsetting = handle_quectel_ec20(dev, ifnum); + goto done; + } + /* * Gobi 2K+ USB layout: * 0: QMI/net From e07af133c3e2716db25e3e1e1d9f10c2088e9c1a Mon Sep 17 00:00:00 2001 From: Aleksander Morgado Date: Wed, 11 Nov 2015 19:51:40 +0100 Subject: [PATCH 115/335] USB: serial: option: add support for Novatel MiFi USB620L MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Also known as Verizon U620L. The device is modeswitched from 1410:9020 to 1410:9022 by selecting the 4th USB configuration: $ sudo usb_modeswitch –v 0x1410 –p 0x9020 –u 4 This configuration provides a ECM interface as well as TTYs ('Enterprise Mode' according to the U620 Linux integration guide). Signed-off-by: Aleksander Morgado Cc: stable Signed-off-by: Johan Hovold --- drivers/usb/serial/option.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 685fef71d3d1..2ab2a33e00c4 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -161,6 +161,7 @@ static void option_instat_callback(struct urb *urb); #define NOVATELWIRELESS_PRODUCT_HSPA_EMBEDDED_HIGHSPEED 0x9001 #define NOVATELWIRELESS_PRODUCT_E362 0x9010 #define NOVATELWIRELESS_PRODUCT_E371 0x9011 +#define NOVATELWIRELESS_PRODUCT_U620L 0x9022 #define NOVATELWIRELESS_PRODUCT_G2 0xA010 #define NOVATELWIRELESS_PRODUCT_MC551 0xB001 @@ -1052,6 +1053,7 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE_AND_INTERFACE_INFO(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_MC551, 0xff, 0xff, 0xff) }, { USB_DEVICE_AND_INTERFACE_INFO(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_E362, 0xff, 0xff, 0xff) }, { USB_DEVICE_AND_INTERFACE_INFO(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_E371, 0xff, 0xff, 0xff) }, + { USB_DEVICE_AND_INTERFACE_INFO(NOVATELWIRELESS_VENDOR_ID, NOVATELWIRELESS_PRODUCT_U620L, 0xff, 0x00, 0x00) }, { USB_DEVICE(AMOI_VENDOR_ID, AMOI_PRODUCT_H01) }, { USB_DEVICE(AMOI_VENDOR_ID, AMOI_PRODUCT_H01A) }, From 1bcb49e663f88bccee35b8688e6a3da2bea31fd4 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Sat, 14 Nov 2015 16:49:30 +0000 Subject: [PATCH 116/335] USB: ti_usb_3410_5052: Add Honeywell HGI80 ID The Honeywell HGI80 is a wireless interface to the evohome connected thermostat. It uses a TI 3410 USB-serial port. Signed-off-by: David Woodhouse Cc: stable@vger.kernel.org Signed-off-by: Johan Hovold --- drivers/usb/serial/ti_usb_3410_5052.c | 2 ++ drivers/usb/serial/ti_usb_3410_5052.h | 4 ++++ 2 files changed, 6 insertions(+) diff --git a/drivers/usb/serial/ti_usb_3410_5052.c b/drivers/usb/serial/ti_usb_3410_5052.c index e9da41d9fe7f..2694df2f4559 100644 --- a/drivers/usb/serial/ti_usb_3410_5052.c +++ b/drivers/usb/serial/ti_usb_3410_5052.c @@ -159,6 +159,7 @@ static const struct usb_device_id ti_id_table_3410[] = { { USB_DEVICE(ABBOTT_VENDOR_ID, ABBOTT_STEREO_PLUG_ID) }, { USB_DEVICE(ABBOTT_VENDOR_ID, ABBOTT_STRIP_PORT_ID) }, { USB_DEVICE(TI_VENDOR_ID, FRI2_PRODUCT_ID) }, + { USB_DEVICE(HONEYWELL_VENDOR_ID, HONEYWELL_HGI80_PRODUCT_ID) }, { } /* terminator */ }; @@ -191,6 +192,7 @@ static const struct usb_device_id ti_id_table_combined[] = { { USB_DEVICE(ABBOTT_VENDOR_ID, ABBOTT_PRODUCT_ID) }, { USB_DEVICE(ABBOTT_VENDOR_ID, ABBOTT_STRIP_PORT_ID) }, { USB_DEVICE(TI_VENDOR_ID, FRI2_PRODUCT_ID) }, + { USB_DEVICE(HONEYWELL_VENDOR_ID, HONEYWELL_HGI80_PRODUCT_ID) }, { } /* terminator */ }; diff --git a/drivers/usb/serial/ti_usb_3410_5052.h b/drivers/usb/serial/ti_usb_3410_5052.h index 4a2423e84d55..98f35c656c02 100644 --- a/drivers/usb/serial/ti_usb_3410_5052.h +++ b/drivers/usb/serial/ti_usb_3410_5052.h @@ -56,6 +56,10 @@ #define ABBOTT_PRODUCT_ID ABBOTT_STEREO_PLUG_ID #define ABBOTT_STRIP_PORT_ID 0x3420 +/* Honeywell vendor and product IDs */ +#define HONEYWELL_VENDOR_ID 0x10ac +#define HONEYWELL_HGI80_PRODUCT_ID 0x0102 /* Honeywell HGI80 */ + /* Commands */ #define TI_GET_VERSION 0x01 #define TI_GET_PORT_STATUS 0x02 From 500a3d2eb3883b71350036e15aad286cc6e5df21 Mon Sep 17 00:00:00 2001 From: Mika Kuoppala Date: Fri, 13 Nov 2015 19:29:41 +0200 Subject: [PATCH 117/335] drm/i915: Fix GT frequency rounding MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When we set and later readback a frequency value through sysfs interface, igt/pm_rpm assumes that we get same value back if it matches hw granularity. On bxt we have found out that this is not always the case. Currently frequency - hw ratio - frequency conversions round down, with few exceptions on platforms that have more specific conversions. On bxt the supported range can be for example from 100Mhz to 650Mhz. Midpoint is then calculated by test to be 375 which pm_rps uses to find a closest hw supported frequency. That is 366 (ratio 22), which it then writes back. But as the rounding down kicks in, driver actually sets 350 instead of 366, as 366 is 2/3 below 22 * 50/3. Fix this by rounding to closest instead of rounding down in freq-ratio-freq conversions. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92768 Testcase: igt/pm_rps/basic-api Tested-by: Bob Paauwe Cc: Bob Paauwe Signed-off-by: Imre Deak Signed-off-by: Mika Kuoppala Reviewed-by: Ville Syrjälä Reviewed-by: Bob Paauwe Link: http://patchwork.freedesktop.org/patch/msgid/1447435781-23416-1-git-send-email-mika.kuoppala@intel.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/intel_pm.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d52a15df6917..7b47da805685 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -7255,7 +7255,8 @@ static int chv_freq_opcode(struct drm_i915_private *dev_priv, int val) int intel_gpu_freq(struct drm_i915_private *dev_priv, int val) { if (IS_GEN9(dev_priv->dev)) - return (val * GT_FREQUENCY_MULTIPLIER) / GEN9_FREQ_SCALER; + return DIV_ROUND_CLOSEST(val * GT_FREQUENCY_MULTIPLIER, + GEN9_FREQ_SCALER); else if (IS_CHERRYVIEW(dev_priv->dev)) return chv_gpu_freq(dev_priv, val); else if (IS_VALLEYVIEW(dev_priv->dev)) @@ -7267,13 +7268,14 @@ int intel_gpu_freq(struct drm_i915_private *dev_priv, int val) int intel_freq_opcode(struct drm_i915_private *dev_priv, int val) { if (IS_GEN9(dev_priv->dev)) - return (val * GEN9_FREQ_SCALER) / GT_FREQUENCY_MULTIPLIER; + return DIV_ROUND_CLOSEST(val * GEN9_FREQ_SCALER, + GT_FREQUENCY_MULTIPLIER); else if (IS_CHERRYVIEW(dev_priv->dev)) return chv_freq_opcode(dev_priv, val); else if (IS_VALLEYVIEW(dev_priv->dev)) return byt_freq_opcode(dev_priv, val); else - return val / GT_FREQUENCY_MULTIPLIER; + return DIV_ROUND_CLOSEST(val, GT_FREQUENCY_MULTIPLIER); } struct request_boost { From fd0fe6acf1dd88aabfbf383f7e4c16315387a7b7 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Wed, 4 Nov 2015 21:25:32 +0200 Subject: [PATCH 118/335] drm/i915: get runtime PM reference around GEM set_caching IOCTL After Damien's D3 fix I started to get runtime suspend residency for the first time and that revealed a breakage on the set_caching IOCTL path that accesses the HW but doesn't take an RPM ref. Fix this up. Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Cc: stable@vger.kernel.org Link: http://patchwork.freedesktop.org/patch/msgid/1446665132-22491-1-git-send-email-imre.deak@intel.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_gem.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 5cf4a1998273..91bb1fc27420 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3809,6 +3809,7 @@ int i915_gem_get_caching_ioctl(struct drm_device *dev, void *data, int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { + struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_caching *args = data; struct drm_i915_gem_object *obj; enum i915_cache_level level; @@ -3837,9 +3838,11 @@ int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data, return -EINVAL; } + intel_runtime_pm_get(dev_priv); + ret = i915_mutex_lock_interruptible(dev); if (ret) - return ret; + goto rpm_put; obj = to_intel_bo(drm_gem_object_lookup(dev, file, args->handle)); if (&obj->base == NULL) { @@ -3852,6 +3855,9 @@ int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data, drm_gem_object_unreference(&obj->base); unlock: mutex_unlock(&dev->struct_mutex); +rpm_put: + intel_runtime_pm_put(dev_priv); + return ret; } From 96ea47c0ec8c012509116bee8c57414281428fc4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Thu, 5 Nov 2015 17:25:26 +0900 Subject: [PATCH 119/335] drm/radeon: Disable uncacheable CPU mappings of GTT with RV6xx MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit They reportedly cause random GPU hangs. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91268 Reviewed-by: Christian König Cc: stable@vger.kernel.org Signed-off-by: Michel Dänzer Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_object.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index d3024883b844..a35f5af77c6a 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -221,6 +221,12 @@ int radeon_bo_create(struct radeon_device *rdev, if (!(rdev->flags & RADEON_IS_PCIE)) bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC); + /* Write-combined CPU mappings of GTT cause GPU hangs with RV6xx + * See https://bugs.freedesktop.org/show_bug.cgi?id=91268 + */ + if (rdev->family >= CHIP_RV610 && rdev->family <= CHIP_RV635) + bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC); + #ifdef CONFIG_X86_32 /* XXX: Write-combined CPU mappings of GTT seem broken on 32-bit * See https://bugs.freedesktop.org/show_bug.cgi?id=84627 From a28bbd5824d4a2af98de45b300ab8d8fb39739fc Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Thu, 5 Nov 2015 17:25:27 +0900 Subject: [PATCH 120/335] drm/radeon: Always disable RADEON_GEM_GTT_UC along with RADEON_GEM_GTT_WC MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Write-combining is a CPU feature. From the GPU POV, these both simply mean no GPU<->CPU cache coherency. Reviewed-by: Christian König Signed-off-by: Michel Dänzer Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_object.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index a35f5af77c6a..dda2ecfbf056 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -231,7 +231,7 @@ int radeon_bo_create(struct radeon_device *rdev, /* XXX: Write-combined CPU mappings of GTT seem broken on 32-bit * See https://bugs.freedesktop.org/show_bug.cgi?id=84627 */ - bo->flags &= ~RADEON_GEM_GTT_WC; + bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC); #elif defined(CONFIG_X86) && !defined(CONFIG_X86_PAT) /* Don't try to enable write-combining when it can't work, or things * may be slow @@ -243,7 +243,7 @@ int radeon_bo_create(struct radeon_device *rdev, DRM_INFO_ONCE("Please enable CONFIG_MTRR and CONFIG_X86_PAT for " "better performance thanks to write-combining\n"); - bo->flags &= ~RADEON_GEM_GTT_WC; + bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC); #endif radeon_ttm_placement_from_domain(bo, domain); From 938204985c1c46c28a88cb558d9e38610044b67a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Thu, 5 Nov 2015 17:25:28 +0900 Subject: [PATCH 121/335] drm/radeon: Only prompt for enabling PAT when we'd allow write-combining MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit No use bothering users about this for whom we disable write-combining for other reasons anyway. Reviewed-by: Christian König Signed-off-by: Michel Dänzer Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_object.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index dda2ecfbf056..84d45633d28c 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -241,8 +241,9 @@ int radeon_bo_create(struct radeon_device *rdev, #warning Please enable CONFIG_MTRR and CONFIG_X86_PAT for better performance \ thanks to write-combining - DRM_INFO_ONCE("Please enable CONFIG_MTRR and CONFIG_X86_PAT for " - "better performance thanks to write-combining\n"); + if (bo->flags & RADEON_GEM_GTT_WC) + DRM_INFO_ONCE("Please enable CONFIG_MTRR and CONFIG_X86_PAT for " + "better performance thanks to write-combining\n"); bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC); #endif From fa6760482bed1ed752568fe09135c078e9768595 Mon Sep 17 00:00:00 2001 From: Flora Cui Date: Fri, 30 Oct 2015 18:29:46 +0800 Subject: [PATCH 122/335] drm/amdgpu: update Fiji's mmPA_SC_RASTER_CONFIG value Change-Id: I6d138306a878450e5bf8a77a2f1aacc380a39fe5 Signed-off-by: Flora Cui Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 6776cf756d40..4410e5735154 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -2971,10 +2971,13 @@ static int gfx_v8_0_cp_gfx_start(struct amdgpu_device *adev) amdgpu_ring_write(ring, mmPA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START); switch (adev->asic_type) { case CHIP_TONGA: - case CHIP_FIJI: amdgpu_ring_write(ring, 0x16000012); amdgpu_ring_write(ring, 0x0000002A); break; + case CHIP_FIJI: + amdgpu_ring_write(ring, 0x3a00161a); + amdgpu_ring_write(ring, 0x0000002e); + break; case CHIP_TOPAZ: case CHIP_CARRIZO: amdgpu_ring_write(ring, 0x00000002); From 935c186aaecc25258495351adaba34f7c507d298 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Tue, 3 Nov 2015 21:23:08 +0100 Subject: [PATCH 123/335] drm/amdgpu: remove fence trace points MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Mostly unused and replaced by the common trace points. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Acked-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 1 - drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h | 43 ----------------------- 2 files changed, 44 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 257d72205bb5..b92c1937543a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -132,7 +132,6 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner, amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr, (*fence)->seq, AMDGPU_FENCE_FLAG_INT); - trace_amdgpu_fence_emit(ring->adev->ddev, ring->idx, (*fence)->seq); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 76ecbaf72a2e..26e2d50a6f64 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -196,49 +196,6 @@ TRACE_EVENT(amdgpu_bo_list_set, TP_printk("list=%p, bo=%p", __entry->list, __entry->bo) ); -DECLARE_EVENT_CLASS(amdgpu_fence_request, - - TP_PROTO(struct drm_device *dev, int ring, u32 seqno), - - TP_ARGS(dev, ring, seqno), - - TP_STRUCT__entry( - __field(u32, dev) - __field(int, ring) - __field(u32, seqno) - ), - - TP_fast_assign( - __entry->dev = dev->primary->index; - __entry->ring = ring; - __entry->seqno = seqno; - ), - - TP_printk("dev=%u, ring=%d, seqno=%u", - __entry->dev, __entry->ring, __entry->seqno) -); - -DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_emit, - - TP_PROTO(struct drm_device *dev, int ring, u32 seqno), - - TP_ARGS(dev, ring, seqno) -); - -DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_wait_begin, - - TP_PROTO(struct drm_device *dev, int ring, u32 seqno), - - TP_ARGS(dev, ring, seqno) -); - -DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_wait_end, - - TP_PROTO(struct drm_device *dev, int ring, u32 seqno), - - TP_ARGS(dev, ring, seqno) -); - DECLARE_EVENT_CLASS(amdgpu_semaphore_request, TP_PROTO(int ring, struct amdgpu_semaphore *sem), From c2776afe740db5598c4c457dcacb94d4427b13f9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Tue, 3 Nov 2015 13:27:39 +0100 Subject: [PATCH 124/335] drm/amdgpu: use a timer for fence fallback MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Less overhead than a work item and also adds proper cleanup handling. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Acked-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 3 +- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 80 ++++++++++------------- 2 files changed, 36 insertions(+), 47 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 615ce6d464fb..3630cb545d0d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -389,7 +389,6 @@ struct amdgpu_clock { * Fences. */ struct amdgpu_fence_driver { - struct amdgpu_ring *ring; uint64_t gpu_addr; volatile uint32_t *cpu_addr; /* sync_seq is protected by ring emission lock */ @@ -398,7 +397,7 @@ struct amdgpu_fence_driver { bool initialized; struct amdgpu_irq_src *irq_src; unsigned irq_type; - struct delayed_work lockup_work; + struct timer_list fallback_timer; wait_queue_head_t fence_queue; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index b92c1937543a..257fce356319 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -84,24 +84,6 @@ static u32 amdgpu_fence_read(struct amdgpu_ring *ring) return seq; } -/** - * amdgpu_fence_schedule_check - schedule lockup check - * - * @ring: pointer to struct amdgpu_ring - * - * Queues a delayed work item to check for lockups. - */ -static void amdgpu_fence_schedule_check(struct amdgpu_ring *ring) -{ - /* - * Do not reset the timer here with mod_delayed_work, - * this can livelock in an interaction with TTM delayed destroy. - */ - queue_delayed_work(system_power_efficient_wq, - &ring->fence_drv.lockup_work, - AMDGPU_FENCE_JIFFIES_TIMEOUT); -} - /** * amdgpu_fence_emit - emit a fence on the requested ring * @@ -135,6 +117,19 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner, return 0; } +/** + * amdgpu_fence_schedule_fallback - schedule fallback check + * + * @ring: pointer to struct amdgpu_ring + * + * Start a timer as fallback to our interrupts. + */ +static void amdgpu_fence_schedule_fallback(struct amdgpu_ring *ring) +{ + mod_timer(&ring->fence_drv.fallback_timer, + jiffies + AMDGPU_FENCE_JIFFIES_TIMEOUT); +} + /** * amdgpu_fence_activity - check for fence activity * @@ -201,32 +196,11 @@ static bool amdgpu_fence_activity(struct amdgpu_ring *ring) } while (atomic64_xchg(&ring->fence_drv.last_seq, seq) > seq); if (seq < last_emitted) - amdgpu_fence_schedule_check(ring); + amdgpu_fence_schedule_fallback(ring); return wake; } -/** - * amdgpu_fence_check_lockup - check for hardware lockup - * - * @work: delayed work item - * - * Checks for fence activity and if there is none probe - * the hardware if a lockup occured. - */ -static void amdgpu_fence_check_lockup(struct work_struct *work) -{ - struct amdgpu_fence_driver *fence_drv; - struct amdgpu_ring *ring; - - fence_drv = container_of(work, struct amdgpu_fence_driver, - lockup_work.work); - ring = fence_drv->ring; - - if (amdgpu_fence_activity(ring)) - wake_up_all(&ring->fence_drv.fence_queue); -} - /** * amdgpu_fence_process - process a fence * @@ -242,6 +216,20 @@ void amdgpu_fence_process(struct amdgpu_ring *ring) wake_up_all(&ring->fence_drv.fence_queue); } +/** + * amdgpu_fence_fallback - fallback for hardware interrupts + * + * @work: delayed work item + * + * Checks for fence activity. + */ +static void amdgpu_fence_fallback(unsigned long arg) +{ + struct amdgpu_ring *ring = (void *)arg; + + amdgpu_fence_process(ring); +} + /** * amdgpu_fence_seq_signaled - check if a fence sequence number has signaled * @@ -289,7 +277,7 @@ static int amdgpu_fence_ring_wait_seq(struct amdgpu_ring *ring, uint64_t seq) if (atomic64_read(&ring->fence_drv.last_seq) >= seq) return 0; - amdgpu_fence_schedule_check(ring); + amdgpu_fence_schedule_fallback(ring); wait_event(ring->fence_drv.fence_queue, ( (signaled = amdgpu_fence_seq_signaled(ring, seq)))); @@ -490,9 +478,8 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring) atomic64_set(&ring->fence_drv.last_seq, 0); ring->fence_drv.initialized = false; - INIT_DELAYED_WORK(&ring->fence_drv.lockup_work, - amdgpu_fence_check_lockup); - ring->fence_drv.ring = ring; + setup_timer(&ring->fence_drv.fallback_timer, amdgpu_fence_fallback, + (unsigned long)ring); init_waitqueue_head(&ring->fence_drv.fence_queue); @@ -556,6 +543,7 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev) mutex_lock(&adev->ring_lock); for (i = 0; i < AMDGPU_MAX_RINGS; i++) { struct amdgpu_ring *ring = adev->rings[i]; + if (!ring || !ring->fence_drv.initialized) continue; r = amdgpu_fence_wait_empty(ring); @@ -567,6 +555,7 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev) amdgpu_irq_put(adev, ring->fence_drv.irq_src, ring->fence_drv.irq_type); amd_sched_fini(&ring->sched); + del_timer_sync(&ring->fence_drv.fallback_timer); ring->fence_drv.initialized = false; } mutex_unlock(&adev->ring_lock); @@ -750,7 +739,8 @@ static bool amdgpu_fence_enable_signaling(struct fence *f) fence->fence_wake.func = amdgpu_fence_check_signaled; __add_wait_queue(&ring->fence_drv.fence_queue, &fence->fence_wake); fence_get(f); - amdgpu_fence_schedule_check(ring); + if (!timer_pending(&ring->fence_drv.fallback_timer)) + amdgpu_fence_schedule_fallback(ring); FENCE_TRACE(&fence->base, "armed on ring %i!\n", ring->idx); return true; } From b56c22853eabde22a93d77bbd5250c1655405d74 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 29 Oct 2015 17:01:19 +0100 Subject: [PATCH 125/335] drm/amdgpu: use fence_is_later() for vm_flush as well v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit v2: remove superfluous check Signed-off-by: Christian König Reviewed-by: Alex Deucher (v1) Reviewed-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 633a32a48560..8c80381b1ea9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -207,24 +207,21 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, uint64_t pd_addr = amdgpu_bo_gpu_offset(vm->page_directory); struct amdgpu_vm_id *vm_id = &vm->ids[ring->idx]; struct fence *flushed_updates = vm_id->flushed_updates; - bool is_earlier = false; + bool is_later; - if (flushed_updates && updates) { - BUG_ON(flushed_updates->context != updates->context); - is_earlier = (updates->seqno - flushed_updates->seqno <= - INT_MAX) ? true : false; - } - - if (pd_addr != vm_id->pd_gpu_addr || !flushed_updates || - is_earlier) { + if (!flushed_updates) + is_later = true; + else if (!updates) + is_later = false; + else + is_later = fence_is_later(updates, flushed_updates); + if (pd_addr != vm_id->pd_gpu_addr || is_later) { trace_amdgpu_vm_flush(pd_addr, ring->idx, vm_id->id); - if (is_earlier) { + if (is_later) { vm_id->flushed_updates = fence_get(updates); fence_put(flushed_updates); } - if (!flushed_updates) - vm_id->flushed_updates = fence_get(updates); vm_id->pd_gpu_addr = pd_addr; amdgpu_ring_emit_vm_flush(ring, vm_id->id, vm_id->pd_gpu_addr); } From 16ae42feb0ce6cc80473d35dd82a1cd07e2e7178 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Tue, 3 Nov 2015 14:53:28 +0100 Subject: [PATCH 126/335] drm/amdgpu: use common fence for amdgpu_vm_fence MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Just cleanup the function parameters. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 12 ++++++------ 3 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 3630cb545d0d..b801b6702812 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -2332,7 +2332,7 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, struct fence *updates); void amdgpu_vm_fence(struct amdgpu_device *adev, struct amdgpu_vm *vm, - struct amdgpu_fence *fence); + struct fence *fence); uint64_t amdgpu_vm_map_gart(struct amdgpu_device *adev, uint64_t addr); int amdgpu_vm_update_page_directory(struct amdgpu_device *adev, struct amdgpu_vm *vm); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index e65987743871..3902e7a9f8cc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -216,7 +216,7 @@ int amdgpu_ib_schedule(struct amdgpu_device *adev, unsigned num_ibs, } if (ib->vm) - amdgpu_vm_fence(adev, ib->vm, ib->fence); + amdgpu_vm_fence(adev, ib->vm, &ib->fence->base); amdgpu_ring_unlock_commit(ring); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 8c80381b1ea9..a9fcc995d27e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -241,16 +241,16 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, */ void amdgpu_vm_fence(struct amdgpu_device *adev, struct amdgpu_vm *vm, - struct amdgpu_fence *fence) + struct fence *fence) { - unsigned ridx = fence->ring->idx; - unsigned vm_id = vm->ids[ridx].id; + struct amdgpu_ring *ring = amdgpu_ring_from_fence(fence); + unsigned vm_id = vm->ids[ring->idx].id; fence_put(adev->vm_manager.active[vm_id]); - adev->vm_manager.active[vm_id] = fence_get(&fence->base); + adev->vm_manager.active[vm_id] = fence_get(fence); - fence_put(vm->ids[ridx].last_id_use); - vm->ids[ridx].last_id_use = fence_get(&fence->base); + fence_put(vm->ids[ring->idx].last_id_use); + vm->ids[ring->idx].last_id_use = fence_get(fence); } /** From 451f698bcac84c49dd4e25dd8e89dbd66796b4cd Mon Sep 17 00:00:00 2001 From: Flora Cui Date: Thu, 5 Nov 2015 15:24:46 +0800 Subject: [PATCH 127/335] drm/amdgpu: update fiji_mgcg_cgcg_init table Change-Id: If44b8057741c78208f1976f60f31b535c944d0bd Signed-off-by: Flora Cui Reviewed-by: Jammy Zhou --- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 4410e5735154..3de6a8894e2d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -268,7 +268,6 @@ static const u32 fiji_mgcg_cgcg_init[] = mmCGTT_CP_CLK_CTRL, 0xffffffff, 0x00000100, mmCGTT_CPC_CLK_CTRL, 0xffffffff, 0x00000100, mmCGTT_CPF_CLK_CTRL, 0xffffffff, 0x40000100, - mmCGTT_DRM_CLK_CTRL0, 0xffffffff, 0x00600100, mmCGTT_GDS_CLK_CTRL, 0xffffffff, 0x00000100, mmCGTT_IA_CLK_CTRL, 0xffffffff, 0x06000100, mmCGTT_PA_CLK_CTRL, 0xffffffff, 0x00000100, @@ -296,10 +295,6 @@ static const u32 fiji_mgcg_cgcg_init[] = mmCGTS_SM_CTRL_REG, 0xffffffff, 0x96e00200, mmCP_RB_WPTR_POLL_CNTL, 0xffffffff, 0x00900100, mmRLC_CGCG_CGLS_CTRL, 0xffffffff, 0x0020003c, - mmPCIE_INDEX, 0xffffffff, 0x0140001c, - mmPCIE_DATA, 0x000f0000, 0x00000000, - mmCGTT_DRM_CLK_CTRL0, 0xff000fff, 0x00000100, - mmHDP_XDP_CGTT_BLK_CTRL, 0xc0000fff, 0x00000104, mmCP_MEM_SLP_CNTL, 0x00000001, 0x00000001, }; From b49c84a5765cf68b8e2fdb1dc4eded9c208e6d58 Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Thu, 5 Nov 2015 11:28:28 +0800 Subject: [PATCH 128/335] drm/amdgpu: add kmem cache for amdgpu fence MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: I5ad8dd156ccf27a6f18004aa0a215a0925b6e67b Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 257fce356319..3671f9f220bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -47,6 +47,9 @@ * that the the relevant GPU caches have been flushed. */ +static struct kmem_cache *amdgpu_fence_slab; +static atomic_t amdgpu_fence_slab_ref = ATOMIC_INIT(0); + /** * amdgpu_fence_write - write a fence value * @@ -100,7 +103,7 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner, struct amdgpu_device *adev = ring->adev; /* we are protected by the ring emission mutex */ - *fence = kmalloc(sizeof(struct amdgpu_fence), GFP_KERNEL); + *fence = kmem_cache_alloc(amdgpu_fence_slab, GFP_KERNEL); if ((*fence) == NULL) { return -ENOMEM; } @@ -522,6 +525,13 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring) */ int amdgpu_fence_driver_init(struct amdgpu_device *adev) { + if (atomic_inc_return(&amdgpu_fence_slab_ref) == 1) { + amdgpu_fence_slab = kmem_cache_create( + "amdgpu_fence", sizeof(struct amdgpu_fence), 0, + SLAB_HWCACHE_ALIGN, NULL); + if (!amdgpu_fence_slab) + return -ENOMEM; + } if (amdgpu_debugfs_fence_init(adev)) dev_err(adev->dev, "fence debugfs file creation failed\n"); @@ -540,6 +550,8 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev) { int i, r; + if (atomic_dec_and_test(&amdgpu_fence_slab_ref)) + kmem_cache_destroy(amdgpu_fence_slab); mutex_lock(&adev->ring_lock); for (i = 0; i < AMDGPU_MAX_RINGS; i++) { struct amdgpu_ring *ring = adev->rings[i]; @@ -745,13 +757,19 @@ static bool amdgpu_fence_enable_signaling(struct fence *f) return true; } +static void amdgpu_fence_release(struct fence *f) +{ + struct amdgpu_fence *fence = to_amdgpu_fence(f); + kmem_cache_free(amdgpu_fence_slab, fence); +} + const struct fence_ops amdgpu_fence_ops = { .get_driver_name = amdgpu_fence_get_driver_name, .get_timeline_name = amdgpu_fence_get_timeline_name, .enable_signaling = amdgpu_fence_enable_signaling, .signaled = amdgpu_fence_is_signaled, .wait = fence_default_wait, - .release = NULL, + .release = amdgpu_fence_release, }; /* From f5617f9dde5ae2466560f7cb008c741e2b88adab Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Thu, 5 Nov 2015 11:41:50 +0800 Subject: [PATCH 129/335] drm/amd: add kmem cache for sched fence MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: I45bb8ff10ef05dc3b15e31a77fbcf31117705f11 Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 12 ++++++++++++ drivers/gpu/drm/amd/scheduler/gpu_scheduler.h | 3 +++ drivers/gpu/drm/amd/scheduler/sched_fence.c | 10 ++++++++-- 3 files changed, 23 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index 89619a5a4289..fe5b3c415f18 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -34,6 +34,9 @@ static struct amd_sched_job * amd_sched_entity_pop_job(struct amd_sched_entity *entity); static void amd_sched_wakeup(struct amd_gpu_scheduler *sched); +struct kmem_cache *sched_fence_slab; +atomic_t sched_fence_slab_ref = ATOMIC_INIT(0); + /* Initialize a given run queue struct */ static void amd_sched_rq_init(struct amd_sched_rq *rq) { @@ -450,6 +453,13 @@ int amd_sched_init(struct amd_gpu_scheduler *sched, init_waitqueue_head(&sched->wake_up_worker); init_waitqueue_head(&sched->job_scheduled); atomic_set(&sched->hw_rq_count, 0); + if (atomic_inc_return(&sched_fence_slab_ref) == 1) { + sched_fence_slab = kmem_cache_create( + "amd_sched_fence", sizeof(struct amd_sched_fence), 0, + SLAB_HWCACHE_ALIGN, NULL); + if (!sched_fence_slab) + return -ENOMEM; + } /* Each scheduler will run on a seperate kernel thread */ sched->thread = kthread_run(amd_sched_main, sched, sched->name); @@ -470,4 +480,6 @@ void amd_sched_fini(struct amd_gpu_scheduler *sched) { if (sched->thread) kthread_stop(sched->thread); + if (atomic_dec_and_test(&sched_fence_slab_ref)) + kmem_cache_destroy(sched_fence_slab); } diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h index 929e9aced041..4d05ca6fb057 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h @@ -30,6 +30,9 @@ struct amd_gpu_scheduler; struct amd_sched_rq; +extern struct kmem_cache *sched_fence_slab; +extern atomic_t sched_fence_slab_ref; + /** * A scheduler entity is a wrapper around a job queue or a group * of other entities. Entities take turns emitting jobs from their diff --git a/drivers/gpu/drm/amd/scheduler/sched_fence.c b/drivers/gpu/drm/amd/scheduler/sched_fence.c index d802638094f4..8d2130b9ff05 100644 --- a/drivers/gpu/drm/amd/scheduler/sched_fence.c +++ b/drivers/gpu/drm/amd/scheduler/sched_fence.c @@ -32,7 +32,7 @@ struct amd_sched_fence *amd_sched_fence_create(struct amd_sched_entity *s_entity struct amd_sched_fence *fence = NULL; unsigned seq; - fence = kzalloc(sizeof(struct amd_sched_fence), GFP_KERNEL); + fence = kmem_cache_zalloc(sched_fence_slab, GFP_KERNEL); if (fence == NULL) return NULL; fence->owner = owner; @@ -71,11 +71,17 @@ static bool amd_sched_fence_enable_signaling(struct fence *f) return true; } +static void amd_sched_fence_release(struct fence *f) +{ + struct amd_sched_fence *fence = to_amd_sched_fence(f); + kmem_cache_free(sched_fence_slab, fence); +} + const struct fence_ops amd_sched_fence_ops = { .get_driver_name = amd_sched_fence_get_driver_name, .get_timeline_name = amd_sched_fence_get_timeline_name, .enable_signaling = amd_sched_fence_enable_signaling, .signaled = NULL, .wait = fence_default_wait, - .release = NULL, + .release = amd_sched_fence_release, }; From e4a58a28b50f30e72292b6659d94410cbf7355ad Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 5 Nov 2015 17:00:25 +0100 Subject: [PATCH 130/335] drm/amdgpu: fix leaking the IBs on error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fixing a memory leak when the scheduler is enabled. Signed-off-by: Christian König Reviewed-by: Junwei Zhang Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 25 ++++++++++++------------- 1 file changed, 12 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index dfc4d02c7a38..ecc82dfe83f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -499,16 +499,12 @@ static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser) for (i = 0; i < parser->nchunks; i++) drm_free_large(parser->chunks[i].kdata); kfree(parser->chunks); - if (!amdgpu_enable_scheduler) - { - if (parser->ibs) - for (i = 0; i < parser->num_ibs; i++) - amdgpu_ib_free(parser->adev, &parser->ibs[i]); - kfree(parser->ibs); - if (parser->uf.bo) - drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base); - } - + if (parser->ibs) + for (i = 0; i < parser->num_ibs; i++) + amdgpu_ib_free(parser->adev, &parser->ibs[i]); + kfree(parser->ibs); + if (parser->uf.bo) + drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base); kfree(parser); } @@ -888,11 +884,14 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) job->base.owner = parser->filp; mutex_init(&job->job_lock); if (job->ibs[job->num_ibs - 1].user) { - memcpy(&job->uf, &parser->uf, - sizeof(struct amdgpu_user_fence)); + job->uf = parser->uf; job->ibs[job->num_ibs - 1].user = &job->uf; + parser->uf.bo = NULL; } + parser->ibs = NULL; + parser->num_ibs = 0; + job->free_job = amdgpu_cs_free_job; mutex_lock(&job->job_lock); r = amd_sched_entity_push_job(&job->base); @@ -905,7 +904,7 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) cs->out.handle = amdgpu_ctx_add_fence(parser->ctx, ring, &job->base.s_fence->base); - parser->ibs[parser->num_ibs - 1].sequence = cs->out.handle; + job->ibs[job->num_ibs - 1].sequence = cs->out.handle; list_sort(NULL, &parser->validated, cmp_size_smaller_first); ttm_eu_fence_buffer_objects(&parser->ticket, From 7e52a81c2f0326a85d3ebc005829bcd604731c6d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Wed, 4 Nov 2015 15:44:39 +0100 Subject: [PATCH 131/335] drm/amdgpu: cleanup amdgpu_cs_parser handling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit No need any more to allocate that structure dynamically, just put it on the stack. This is a start to cleanup some of the scheduler fallouts. Signed-off-by: Christian König Reviewed-by: Junwei Zhang Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 5 -- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 88 ++++++++++---------------- 2 files changed, 33 insertions(+), 60 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index b801b6702812..7b02e3455172 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -2256,11 +2256,6 @@ void amdgpu_pci_config_reset(struct amdgpu_device *adev); bool amdgpu_card_posted(struct amdgpu_device *adev); void amdgpu_update_display_priority(struct amdgpu_device *adev); bool amdgpu_boot_test_post_card(struct amdgpu_device *adev); -struct amdgpu_cs_parser *amdgpu_cs_parser_create(struct amdgpu_device *adev, - struct drm_file *filp, - struct amdgpu_ctx *ctx, - struct amdgpu_ib *ibs, - uint32_t num_ibs); int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data); int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index ecc82dfe83f8..bf32096b8eb7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -127,30 +127,6 @@ int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type, return 0; } -struct amdgpu_cs_parser *amdgpu_cs_parser_create(struct amdgpu_device *adev, - struct drm_file *filp, - struct amdgpu_ctx *ctx, - struct amdgpu_ib *ibs, - uint32_t num_ibs) -{ - struct amdgpu_cs_parser *parser; - int i; - - parser = kzalloc(sizeof(struct amdgpu_cs_parser), GFP_KERNEL); - if (!parser) - return NULL; - - parser->adev = adev; - parser->filp = filp; - parser->ctx = ctx; - parser->ibs = ibs; - parser->num_ibs = num_ibs; - for (i = 0; i < num_ibs; i++) - ibs[i].ctx = ctx; - - return parser; -} - int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) { union drm_amdgpu_cs *cs = data; @@ -490,6 +466,7 @@ static void amdgpu_cs_parser_fini_early(struct amdgpu_cs_parser *parser, int err static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser) { unsigned i; + if (parser->ctx) amdgpu_ctx_put(parser->ctx); if (parser->bo_list) @@ -505,7 +482,6 @@ static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser) kfree(parser->ibs); if (parser->uf.bo) drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base); - kfree(parser); } /** @@ -824,36 +800,36 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) union drm_amdgpu_cs *cs = data; struct amdgpu_fpriv *fpriv = filp->driver_priv; struct amdgpu_vm *vm = &fpriv->vm; - struct amdgpu_cs_parser *parser; + struct amdgpu_cs_parser parser = {}; bool reserved_buffers = false; int i, r; if (!adev->accel_working) return -EBUSY; - parser = amdgpu_cs_parser_create(adev, filp, NULL, NULL, 0); - if (!parser) - return -ENOMEM; - r = amdgpu_cs_parser_init(parser, data); + parser.adev = adev; + parser.filp = filp; + + r = amdgpu_cs_parser_init(&parser, data); if (r) { DRM_ERROR("Failed to initialize parser !\n"); - amdgpu_cs_parser_fini(parser, r, false); + amdgpu_cs_parser_fini(&parser, r, false); r = amdgpu_cs_handle_lockup(adev, r); return r; } mutex_lock(&vm->mutex); - r = amdgpu_cs_parser_relocs(parser); + r = amdgpu_cs_parser_relocs(&parser); if (r == -ENOMEM) DRM_ERROR("Not enough memory for command submission!\n"); else if (r && r != -ERESTARTSYS) DRM_ERROR("Failed to process the buffer list %d!\n", r); else if (!r) { reserved_buffers = true; - r = amdgpu_cs_ib_fill(adev, parser); + r = amdgpu_cs_ib_fill(adev, &parser); } if (!r) { - r = amdgpu_cs_dependencies(adev, parser); + r = amdgpu_cs_dependencies(adev, &parser); if (r) DRM_ERROR("Failed in the dependencies handling %d!\n", r); } @@ -861,36 +837,38 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) if (r) goto out; - for (i = 0; i < parser->num_ibs; i++) - trace_amdgpu_cs(parser, i); + for (i = 0; i < parser.num_ibs; i++) + trace_amdgpu_cs(&parser, i); - r = amdgpu_cs_ib_vm_chunk(adev, parser); + r = amdgpu_cs_ib_vm_chunk(adev, &parser); if (r) goto out; - if (amdgpu_enable_scheduler && parser->num_ibs) { + if (amdgpu_enable_scheduler && parser.num_ibs) { struct amdgpu_job *job; - struct amdgpu_ring * ring = parser->ibs->ring; + struct amdgpu_ring * ring = parser.ibs->ring; + job = kzalloc(sizeof(struct amdgpu_job), GFP_KERNEL); if (!job) { r = -ENOMEM; goto out; } + job->base.sched = &ring->sched; - job->base.s_entity = &parser->ctx->rings[ring->idx].entity; - job->adev = parser->adev; - job->ibs = parser->ibs; - job->num_ibs = parser->num_ibs; - job->base.owner = parser->filp; + job->base.s_entity = &parser.ctx->rings[ring->idx].entity; + job->adev = parser.adev; + job->ibs = parser.ibs; + job->num_ibs = parser.num_ibs; + job->base.owner = parser.filp; mutex_init(&job->job_lock); if (job->ibs[job->num_ibs - 1].user) { - job->uf = parser->uf; + job->uf = parser.uf; job->ibs[job->num_ibs - 1].user = &job->uf; - parser->uf.bo = NULL; + parser.uf.bo = NULL; } - parser->ibs = NULL; - parser->num_ibs = 0; + parser.ibs = NULL; + parser.num_ibs = 0; job->free_job = amdgpu_cs_free_job; mutex_lock(&job->job_lock); @@ -902,24 +880,24 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) goto out; } cs->out.handle = - amdgpu_ctx_add_fence(parser->ctx, ring, + amdgpu_ctx_add_fence(parser.ctx, ring, &job->base.s_fence->base); job->ibs[job->num_ibs - 1].sequence = cs->out.handle; - list_sort(NULL, &parser->validated, cmp_size_smaller_first); - ttm_eu_fence_buffer_objects(&parser->ticket, - &parser->validated, + list_sort(NULL, &parser.validated, cmp_size_smaller_first); + ttm_eu_fence_buffer_objects(&parser.ticket, + &parser.validated, &job->base.s_fence->base); mutex_unlock(&job->job_lock); - amdgpu_cs_parser_fini_late(parser); + amdgpu_cs_parser_fini_late(&parser); mutex_unlock(&vm->mutex); return 0; } - cs->out.handle = parser->ibs[parser->num_ibs - 1].sequence; + cs->out.handle = parser.ibs[parser.num_ibs - 1].sequence; out: - amdgpu_cs_parser_fini(parser, r, reserved_buffers); + amdgpu_cs_parser_fini(&parser, r, reserved_buffers); mutex_unlock(&vm->mutex); r = amdgpu_cs_handle_lockup(adev, r); return r; From 24dd2f64c5a877392925202321c7c2c46c2b0ddf Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 10 Nov 2015 13:01:35 -0500 Subject: [PATCH 132/335] drm/radeon: unconditionally set sysfs_initialized Avoids spew on resume for systems where sysfs may fail even on init. bug: https://bugzilla.kernel.org/show_bug.cgi?id=106851 Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/radeon/radeon_pm.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 6d80dde23400..f4f03dcc1530 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -1542,8 +1542,7 @@ int radeon_pm_late_init(struct radeon_device *rdev) ret = device_create_file(rdev->dev, &dev_attr_power_method); if (ret) DRM_ERROR("failed to create device file for power method\n"); - if (!ret) - rdev->pm.sysfs_initialized = true; + rdev->pm.sysfs_initialized = true; } mutex_lock(&rdev->pm.mutex); From a80b30476d240482d360a25a1b2e8c13036ed750 Mon Sep 17 00:00:00 2001 From: Jay Cornwall Date: Thu, 5 Nov 2015 13:06:15 -0600 Subject: [PATCH 133/335] drm/amdgpu: Fix default page access routing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The VM default page (used when a VM translation fails) is allocated in system memory. The VM is misconfigured to interpret the physical address as referencing a VRAM physical page. Route default page accesses to system memory. Reviewed-by: Christian König Signed-off-by: Jay Cornwall Cc: # v4.2+ Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c | 1 + drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 85bbcdc73fff..b93b649d9e36 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -501,6 +501,7 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE, 1); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, EFFECTIVE_L2_QUEUE_SIZE, 7); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY, 1); WREG32(mmVM_L2_CNTL, tmp); tmp = REG_SET_FIELD(0, VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS, 1); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL2, INVALIDATE_L2_CACHE, 1); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 1bcc4e74e3b4..c9209b45d901 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -629,6 +629,7 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE, 1); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, EFFECTIVE_L2_QUEUE_SIZE, 7); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, CONTEXT1_IDENTITY_ACCESS_MODE, 1); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_DEFAULT_PAGE_OUT_TO_SYSTEM_MEMORY, 1); WREG32(mmVM_L2_CNTL, tmp); tmp = RREG32(mmVM_L2_CNTL2); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS, 1); From 515c752dabee9945c1e8686c87f7cdeb3935eea4 Mon Sep 17 00:00:00 2001 From: Maxim Sheviakov Date: Tue, 10 Nov 2015 13:09:13 -0500 Subject: [PATCH 134/335] drm/radeon: fix quirk for MSI R7 370 Armor 2X There was a typo in the original. bug: https://bugs.freedesktop.org/show_bug.cgi?id=92865 Signed-off-by: Maxim Sheviakov Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/si_dpm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index e72bf46042e0..a82b891ae1fe 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -2927,7 +2927,7 @@ static struct si_dpm_quirk si_dpm_quirk_list[] = { { PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 }, { PCI_VENDOR_ID_ATI, 0x6811, 0x174b, 0xe271, 0, 120000 }, { PCI_VENDOR_ID_ATI, 0x6810, 0x174b, 0xe271, 85000, 90000 }, - { PCI_VENDOR_ID_ATI, 0x6811, 0x1762, 0x2015, 0, 120000 }, + { PCI_VENDOR_ID_ATI, 0x6811, 0x1462, 0x2015, 0, 120000 }, { PCI_VENDOR_ID_ATI, 0x6811, 0x1043, 0x2015, 0, 120000 }, { 0, 0, 0, 0 }, }; From e1b35f6103b37e0d81184b32906b7010170dda02 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 10 Nov 2015 13:17:55 +0100 Subject: [PATCH 135/335] drm/amdgpu: fix seq_printf format string The amdgpu driver has a debugfs interface that shows the amount of VRAM in use, but the newly added code causes a build error on all 32-bit architectures: drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:1076:17: warning: format '%lu' expects argument of type 'long unsigned int', but argument 4 has type 'long long int' [-Wformat=] This fixes the format string to use "%llu" for printing 64-bit numbers, which works everywhere, as long as we also cast to 'u64'. Unlike atomic64_t, u64 is defined as 'unsigned long long' on all architectures. Signed-off-by: Arnd Bergmann Fixes: a2ef8a974931 ("drm/amdgpu: add vram usage into debugfs") Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 81bb8e9fc26d..d4bac5f49939 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1073,10 +1073,10 @@ static int amdgpu_mm_dump_table(struct seq_file *m, void *data) ret = drm_mm_dump_table(m, mm); spin_unlock(&glob->lru_lock); if (ttm_pl == TTM_PL_VRAM) - seq_printf(m, "man size:%llu pages, ram usage:%luMB, vis usage:%luMB\n", + seq_printf(m, "man size:%llu pages, ram usage:%lluMB, vis usage:%lluMB\n", adev->mman.bdev.man[ttm_pl].size, - atomic64_read(&adev->vram_usage) >> 20, - atomic64_read(&adev->vram_vis_usage) >> 20); + (u64)atomic64_read(&adev->vram_usage) >> 20, + (u64)atomic64_read(&adev->vram_vis_usage) >> 20); return ret; } From c305fd5fffae5b826b9f8f7ef32fe72ee83354fc Mon Sep 17 00:00:00 2001 From: Rex Zhu Date: Tue, 13 Oct 2015 13:57:52 +0800 Subject: [PATCH 136/335] drm/amdgpu: fix bug that can't enter thermal interrupt for bonaire. Set reversed bit to enable/disable thermal interrupt. Signed-off-by: Rex Zhu Reviewed-by: Jammy Zhou --- drivers/gpu/drm/amd/amdgpu/ci_dpm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index a1a35a5df8e7..57a2e347f04d 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c @@ -6569,12 +6569,12 @@ static int ci_dpm_set_interrupt_state(struct amdgpu_device *adev, switch (state) { case AMDGPU_IRQ_STATE_DISABLE: cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT); - cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK; + cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK; WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int); break; case AMDGPU_IRQ_STATE_ENABLE: cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT); - cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK; + cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK; WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int); break; default: @@ -6586,12 +6586,12 @@ static int ci_dpm_set_interrupt_state(struct amdgpu_device *adev, switch (state) { case AMDGPU_IRQ_STATE_DISABLE: cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT); - cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK; + cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK; WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int); break; case AMDGPU_IRQ_STATE_ENABLE: cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT); - cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK; + cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK; WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int); break; default: From 5f2e816b2957fdde3c7d51d55db6751a980f4bea Mon Sep 17 00:00:00 2001 From: Flora Cui Date: Thu, 5 Nov 2015 09:50:21 +0800 Subject: [PATCH 137/335] drm/amdgpu: update Fiji's tiling mode table Change-Id: I925c15015390113f7e27746ec5751eaa6a92c2a7 Signed-off-by: Flora Cui Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 292 +++++++++++++++++++++++++- 1 file changed, 291 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 3de6a8894e2d..e1dcab98e249 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -995,7 +995,7 @@ static void gfx_v8_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.max_cu_per_sh = 16; adev->gfx.config.max_sh_per_se = 1; adev->gfx.config.max_backends_per_se = 4; - adev->gfx.config.max_texture_channel_caches = 8; + adev->gfx.config.max_texture_channel_caches = 16; adev->gfx.config.max_gprs = 256; adev->gfx.config.max_gs_threads = 32; adev->gfx.config.max_hw_contexts = 8; @@ -1608,6 +1608,296 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev) WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden); } case CHIP_FIJI: + for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) { + switch (reg_offset) { + case 0: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 1: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 2: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 3: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 4: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 5: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 6: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 7: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P4_16x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING)); + break; + case 8: + gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16)); + break; + case 9: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 10: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 11: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + case 12: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P4_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + case 13: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 14: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 15: + gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 16: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + case 17: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P4_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + case 18: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 19: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 20: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 21: + gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 22: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 23: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P4_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 24: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 25: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_XTHICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 26: + gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_XTHICK) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1)); + break; + case 27: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 28: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2)); + break; + case 29: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + case 30: + gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) | + PIPE_CONFIG(ADDR_SURF_P4_16x16) | + MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) | + SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8)); + break; + default: + gb_tile_moden = 0; + break; + } + adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden; + WREG32(mmGB_TILE_MODE0 + reg_offset, gb_tile_moden); + } + for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) { + switch (reg_offset) { + case 0: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 1: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 2: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 3: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 4: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 5: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 6: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 8: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 9: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 10: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 11: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 12: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 13: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) | + NUM_BANKS(ADDR_SURF_8_BANK)); + break; + case 14: + gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) | + NUM_BANKS(ADDR_SURF_4_BANK)); + break; + case 7: + /* unused idx */ + continue; + default: + gb_tile_moden = 0; + break; + } + adev->gfx.config.macrotile_mode_array[reg_offset] = gb_tile_moden; + WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden); + } + break; case CHIP_TONGA: for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) { switch (reg_offset) { From 7034decf6a5b1ff778d83ff9d7ce1f0b404804e4 Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Wed, 11 Nov 2015 14:56:00 +0800 Subject: [PATCH 138/335] drm/amdgpu: add command submission workflow tracepoint OGL needs these tracepoints to investigate performance issue. Change-Id: I5e58187d061253f7d665dfce8e4e163ba91d3e2b Signed-off-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c | 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h | 51 +++++++++++++++++++ .../gpu/drm/amd/scheduler/gpu_sched_trace.h | 24 +++++++-- drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 1 + 5 files changed, 76 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index bf32096b8eb7..2ae73d5232dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -888,7 +888,7 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) ttm_eu_fence_buffer_objects(&parser.ticket, &parser.validated, &job->base.s_fence->base); - + trace_amdgpu_cs_ioctl(job); mutex_unlock(&job->job_lock); amdgpu_cs_parser_fini_late(&parser); mutex_unlock(&vm->mutex); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c index dcf4a8aca680..67f778f6eedb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c @@ -26,6 +26,7 @@ #include #include #include "amdgpu.h" +#include "amdgpu_trace.h" static struct fence *amdgpu_sched_dependency(struct amd_sched_job *sched_job) { @@ -45,6 +46,7 @@ static struct fence *amdgpu_sched_run_job(struct amd_sched_job *sched_job) } job = to_amdgpu_job(sched_job); mutex_lock(&job->job_lock); + trace_amdgpu_sched_run_job(job); r = amdgpu_ib_schedule(job->adev, job->num_ibs, job->ibs, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 26e2d50a6f64..8f9834ab1bd5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -48,6 +48,57 @@ TRACE_EVENT(amdgpu_cs, __entry->fences) ); +TRACE_EVENT(amdgpu_cs_ioctl, + TP_PROTO(struct amdgpu_job *job), + TP_ARGS(job), + TP_STRUCT__entry( + __field(struct amdgpu_device *, adev) + __field(struct amd_sched_job *, sched_job) + __field(struct amdgpu_ib *, ib) + __field(struct fence *, fence) + __field(char *, ring_name) + __field(u32, num_ibs) + ), + + TP_fast_assign( + __entry->adev = job->adev; + __entry->sched_job = &job->base; + __entry->ib = job->ibs; + __entry->fence = &job->base.s_fence->base; + __entry->ring_name = job->ibs[0].ring->name; + __entry->num_ibs = job->num_ibs; + ), + TP_printk("adev=%p, sched_job=%p, first ib=%p, sched fence=%p, ring name:%s, num_ibs:%u", + __entry->adev, __entry->sched_job, __entry->ib, + __entry->fence, __entry->ring_name, __entry->num_ibs) +); + +TRACE_EVENT(amdgpu_sched_run_job, + TP_PROTO(struct amdgpu_job *job), + TP_ARGS(job), + TP_STRUCT__entry( + __field(struct amdgpu_device *, adev) + __field(struct amd_sched_job *, sched_job) + __field(struct amdgpu_ib *, ib) + __field(struct fence *, fence) + __field(char *, ring_name) + __field(u32, num_ibs) + ), + + TP_fast_assign( + __entry->adev = job->adev; + __entry->sched_job = &job->base; + __entry->ib = job->ibs; + __entry->fence = &job->base.s_fence->base; + __entry->ring_name = job->ibs[0].ring->name; + __entry->num_ibs = job->num_ibs; + ), + TP_printk("adev=%p, sched_job=%p, first ib=%p, sched fence=%p, ring name:%s, num_ibs:%u", + __entry->adev, __entry->sched_job, __entry->ib, + __entry->fence, __entry->ring_name, __entry->num_ibs) +); + + TRACE_EVENT(amdgpu_vm_grab_id, TP_PROTO(unsigned vmid, int ring), TP_ARGS(vmid, ring), diff --git a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h index 144f50acc971..c89dc777768f 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h @@ -16,6 +16,8 @@ TRACE_EVENT(amd_sched_job, TP_ARGS(sched_job), TP_STRUCT__entry( __field(struct amd_sched_entity *, entity) + __field(struct amd_sched_job *, sched_job) + __field(struct fence *, fence) __field(const char *, name) __field(u32, job_count) __field(int, hw_job_count) @@ -23,16 +25,32 @@ TRACE_EVENT(amd_sched_job, TP_fast_assign( __entry->entity = sched_job->s_entity; + __entry->sched_job = sched_job; + __entry->fence = &sched_job->s_fence->base; __entry->name = sched_job->sched->name; __entry->job_count = kfifo_len( &sched_job->s_entity->job_queue) / sizeof(sched_job); __entry->hw_job_count = atomic_read( &sched_job->sched->hw_rq_count); ), - TP_printk("entity=%p, ring=%s, job count:%u, hw job count:%d", - __entry->entity, __entry->name, __entry->job_count, - __entry->hw_job_count) + TP_printk("entity=%p, sched job=%p, fence=%p, ring=%s, job count:%u, hw job count:%d", + __entry->entity, __entry->sched_job, __entry->fence, __entry->name, + __entry->job_count, __entry->hw_job_count) ); + +TRACE_EVENT(amd_sched_process_job, + TP_PROTO(struct amd_sched_fence *fence), + TP_ARGS(fence), + TP_STRUCT__entry( + __field(struct fence *, fence) + ), + + TP_fast_assign( + __entry->fence = &fence->base; + ), + TP_printk("fence=%p signaled", __entry->fence) +); + #endif /* This part must be outside protection */ diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index fe5b3c415f18..b8925fea577c 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -346,6 +346,7 @@ static void amd_sched_process_job(struct fence *f, struct fence_cb *cb) list_del_init(&s_fence->list); spin_unlock_irqrestore(&sched->fence_list_lock, flags); } + trace_amd_sched_process_job(s_fence); fence_put(&s_fence->base); wake_up_interruptible(&sched->wake_up_worker); } From 4a562283376197722b295d27633134401bbc80f5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Fri, 6 Nov 2015 14:09:21 +0100 Subject: [PATCH 139/335] drm/amdgpu: cleanup scheduler fence get/put dance MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The code was correct, but getting two references when the ownership is linearly moved on is a bit awkward and just overhead. Signed: Christian König Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c | 1 - drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 1 - 2 files changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c index 67f778f6eedb..8ef9e4415fcc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c @@ -64,7 +64,6 @@ err: job->free_job(job); mutex_unlock(&job->job_lock); - fence_put(&job->base.s_fence->base); kfree(job); return fence ? &fence->base : NULL; } diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index b8925fea577c..ccb7c1554f5e 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -285,7 +285,6 @@ int amd_sched_entity_push_job(struct amd_sched_job *sched_job) if (!fence) return -ENOMEM; - fence_get(&fence->base); sched_job->s_fence = fence; wait_event(entity->sched->job_scheduled, From e284022163716ecf11c37fd1057c35d689ef2c11 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 5 Nov 2015 19:49:48 +0100 Subject: [PATCH 140/335] drm/amdgpu: fix incorrect mutex usage v3 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Before this patch the scheduler fence was created when we push the job into the queue, so we could only get the fence after pushing it. The mutex now was necessary to prevent the thread pushing the jobs to the hardware from running faster than the thread pushing the jobs into the queue. Otherwise the thread pushing jobs into the queue would have accessed possible freed up memory when it tries to get a reference to the fence. So what you get in the end is thread A: mutex_lock(&job->lock); ... Kick of thread B. ... mutex_unlock(&job->lock); And thread B: mutex_lock(&job->lock); .... mutex_unlock(&job->lock); kfree(job); I'm actually not sure if I'm still up to date on this, but this usage pattern used to be not allowed with mutexes. See here as well https://lwn.net/Articles/575460/. v2: remove unrelated changes, fix missing owner v3: rebased, add more commit message Signed-off-by: Christian König Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 43 +++++++++++-------- drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c | 27 +++++------- drivers/gpu/drm/amd/scheduler/gpu_scheduler.c | 10 +---- drivers/gpu/drm/amd/scheduler/gpu_scheduler.h | 3 +- 5 files changed, 37 insertions(+), 48 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 7b02e3455172..0f187027c753 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1225,7 +1225,7 @@ struct amdgpu_job { struct amdgpu_device *adev; struct amdgpu_ib *ibs; uint32_t num_ibs; - struct mutex job_lock; + void *owner; struct amdgpu_user_fence uf; int (*free_job)(struct amdgpu_job *job); }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 2ae73d5232dd..44cf977ae4f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -845,8 +845,9 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) goto out; if (amdgpu_enable_scheduler && parser.num_ibs) { - struct amdgpu_job *job; struct amdgpu_ring * ring = parser.ibs->ring; + struct amd_sched_fence *fence; + struct amdgpu_job *job; job = kzalloc(sizeof(struct amdgpu_job), GFP_KERNEL); if (!job) { @@ -859,37 +860,41 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) job->adev = parser.adev; job->ibs = parser.ibs; job->num_ibs = parser.num_ibs; - job->base.owner = parser.filp; - mutex_init(&job->job_lock); + job->owner = parser.filp; + job->free_job = amdgpu_cs_free_job; + if (job->ibs[job->num_ibs - 1].user) { job->uf = parser.uf; job->ibs[job->num_ibs - 1].user = &job->uf; parser.uf.bo = NULL; } - parser.ibs = NULL; - parser.num_ibs = 0; - - job->free_job = amdgpu_cs_free_job; - mutex_lock(&job->job_lock); - r = amd_sched_entity_push_job(&job->base); - if (r) { - mutex_unlock(&job->job_lock); + fence = amd_sched_fence_create(job->base.s_entity, + parser.filp); + if (!fence) { + r = -ENOMEM; amdgpu_cs_free_job(job); kfree(job); goto out; } - cs->out.handle = - amdgpu_ctx_add_fence(parser.ctx, ring, - &job->base.s_fence->base); + job->base.s_fence = fence; + fence_get(&fence->base); + + cs->out.handle = amdgpu_ctx_add_fence(parser.ctx, ring, + &fence->base); job->ibs[job->num_ibs - 1].sequence = cs->out.handle; - list_sort(NULL, &parser.validated, cmp_size_smaller_first); - ttm_eu_fence_buffer_objects(&parser.ticket, - &parser.validated, - &job->base.s_fence->base); + parser.ibs = NULL; + parser.num_ibs = 0; + trace_amdgpu_cs_ioctl(job); - mutex_unlock(&job->job_lock); + amd_sched_entity_push_job(&job->base); + + list_sort(NULL, &parser.validated, cmp_size_smaller_first); + ttm_eu_fence_buffer_objects(&parser.ticket, &parser.validated, + &fence->base); + fence_put(&fence->base); + amdgpu_cs_parser_fini_late(&parser); mutex_unlock(&vm->mutex); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c index 8ef9e4415fcc..438c05254695 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c @@ -45,12 +45,8 @@ static struct fence *amdgpu_sched_run_job(struct amd_sched_job *sched_job) return NULL; } job = to_amdgpu_job(sched_job); - mutex_lock(&job->job_lock); trace_amdgpu_sched_run_job(job); - r = amdgpu_ib_schedule(job->adev, - job->num_ibs, - job->ibs, - job->base.owner); + r = amdgpu_ib_schedule(job->adev, job->num_ibs, job->ibs, job->owner); if (r) { DRM_ERROR("Error scheduling IBs (%d)\n", r); goto err; @@ -63,7 +59,6 @@ err: if (job->free_job) job->free_job(job); - mutex_unlock(&job->job_lock); kfree(job); return fence ? &fence->base : NULL; } @@ -89,21 +84,19 @@ int amdgpu_sched_ib_submit_kernel_helper(struct amdgpu_device *adev, return -ENOMEM; job->base.sched = &ring->sched; job->base.s_entity = &adev->kernel_ctx.rings[ring->idx].entity; + job->base.s_fence = amd_sched_fence_create(job->base.s_entity, owner); + if (!job->base.s_fence) { + kfree(job); + return -ENOMEM; + } + *f = fence_get(&job->base.s_fence->base); + job->adev = adev; job->ibs = ibs; job->num_ibs = num_ibs; - job->base.owner = owner; - mutex_init(&job->job_lock); + job->owner = owner; job->free_job = free_job; - mutex_lock(&job->job_lock); - r = amd_sched_entity_push_job(&job->base); - if (r) { - mutex_unlock(&job->job_lock); - kfree(job); - return r; - } - *f = fence_get(&job->base.s_fence->base); - mutex_unlock(&job->job_lock); + amd_sched_entity_push_job(&job->base); } else { r = amdgpu_ib_schedule(adev, num_ibs, ibs, owner); if (r) diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index ccb7c1554f5e..ea30d6ad4c13 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -276,21 +276,13 @@ static bool amd_sched_entity_in(struct amd_sched_job *sched_job) * * Returns 0 for success, negative error code otherwise. */ -int amd_sched_entity_push_job(struct amd_sched_job *sched_job) +void amd_sched_entity_push_job(struct amd_sched_job *sched_job) { struct amd_sched_entity *entity = sched_job->s_entity; - struct amd_sched_fence *fence = amd_sched_fence_create( - entity, sched_job->owner); - - if (!fence) - return -ENOMEM; - - sched_job->s_fence = fence; wait_event(entity->sched->job_scheduled, amd_sched_entity_in(sched_job)); trace_amd_sched_job(sched_job); - return 0; } /** diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h index 4d05ca6fb057..939692b14f4b 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h @@ -79,7 +79,6 @@ struct amd_sched_job { struct amd_gpu_scheduler *sched; struct amd_sched_entity *s_entity; struct amd_sched_fence *s_fence; - void *owner; }; extern const struct fence_ops amd_sched_fence_ops; @@ -131,7 +130,7 @@ int amd_sched_entity_init(struct amd_gpu_scheduler *sched, uint32_t jobs); void amd_sched_entity_fini(struct amd_gpu_scheduler *sched, struct amd_sched_entity *entity); -int amd_sched_entity_push_job(struct amd_sched_job *sched_job); +void amd_sched_entity_push_job(struct amd_sched_job *sched_job); struct amd_sched_fence *amd_sched_fence_create( struct amd_sched_entity *s_entity, void *owner); From 5d82730af746abca2aa74e00de6370d338df7e95 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Fri, 13 Nov 2015 13:04:50 +0100 Subject: [PATCH 141/335] drm/amdgpu: fix handling order in scheduler CS MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We need to clear parser.ibs and num_ibs before amd_sched_fence_create, otherwise the IB could be freed twice if fence creates fails. Signed-off-by: Christian König Reviewed-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 44cf977ae4f6..6096effd6a56 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -858,11 +858,14 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) job->base.sched = &ring->sched; job->base.s_entity = &parser.ctx->rings[ring->idx].entity; job->adev = parser.adev; - job->ibs = parser.ibs; - job->num_ibs = parser.num_ibs; job->owner = parser.filp; job->free_job = amdgpu_cs_free_job; + job->ibs = parser.ibs; + job->num_ibs = parser.num_ibs; + parser.ibs = NULL; + parser.num_ibs = 0; + if (job->ibs[job->num_ibs - 1].user) { job->uf = parser.uf; job->ibs[job->num_ibs - 1].user = &job->uf; @@ -884,9 +887,6 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) &fence->base); job->ibs[job->num_ibs - 1].sequence = cs->out.handle; - parser.ibs = NULL; - parser.num_ibs = 0; - trace_amdgpu_cs_ioctl(job); amd_sched_entity_push_job(&job->base); From 43c27fb5c21eb238ce5ffc7766bc59ab773bfd7c Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Thu, 12 Nov 2015 15:33:09 +0800 Subject: [PATCH 142/335] drm/amdgpu: update pd while updating vm as well MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: I93a861cd6707f7d91672b9e19757cc50008cd7a2 Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 087332858853..16dca46314bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -483,6 +483,9 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, if (domain == AMDGPU_GEM_DOMAIN_CPU) goto error_unreserve; } + r = amdgpu_vm_update_page_directory(adev, bo_va->vm); + if (r) + goto error_unreserve; r = amdgpu_vm_clear_freed(adev, bo_va->vm); if (r) From b4c580a43d520b7812c0fd064fbab929ce2f1da0 Mon Sep 17 00:00:00 2001 From: Heikki Krogerus Date: Wed, 21 Oct 2015 14:37:04 +0300 Subject: [PATCH 143/335] usb: dwc3: pci: add support for Intel Broxton SOC PCI IDs for Broxton based platforms. Signed-off-by: Heikki Krogerus Signed-off-by: Felipe Balbi --- drivers/usb/dwc3/dwc3-pci.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/usb/dwc3/dwc3-pci.c b/drivers/usb/dwc3/dwc3-pci.c index 77a622cb48ab..009d83048c8c 100644 --- a/drivers/usb/dwc3/dwc3-pci.c +++ b/drivers/usb/dwc3/dwc3-pci.c @@ -34,6 +34,8 @@ #define PCI_DEVICE_ID_INTEL_BSW 0x22b7 #define PCI_DEVICE_ID_INTEL_SPTLP 0x9d30 #define PCI_DEVICE_ID_INTEL_SPTH 0xa130 +#define PCI_DEVICE_ID_INTEL_BXT 0x0aaa +#define PCI_DEVICE_ID_INTEL_APL 0x5aaa static const struct acpi_gpio_params reset_gpios = { 0, 0, false }; static const struct acpi_gpio_params cs_gpios = { 1, 0, false }; @@ -210,6 +212,8 @@ static const struct pci_device_id dwc3_pci_id_table[] = { { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_MRFLD), }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SPTLP), }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SPTH), }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_BXT), }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_APL), }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_NL_USB), }, { } /* Terminating Entry */ }; From 705e63d2b29c8bbf091119084544d353bda70393 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Fri, 23 Oct 2015 09:53:50 +0200 Subject: [PATCH 144/335] usb: musb: core: fix order of arguments to ulpi write callback MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is a bit of a mess in the order of arguments to the ulpi write callback. There is int ulpi_write(struct ulpi *ulpi, u8 addr, u8 val) in drivers/usb/common/ulpi.c; struct usb_phy_io_ops { ... int (*write)(struct usb_phy *x, u32 val, u32 reg); } in include/linux/usb/phy.h. The callback registered by the musb driver has to comply to the latter, but up to now had "offset" first which effectively made the function broken for correct users. So flip the order and while at it also switch to the parameter names of struct usb_phy_io_ops's write. Fixes: ffb865b1e460 ("usb: musb: add ulpi access operations") Signed-off-by: Uwe Kleine-König Signed-off-by: Felipe Balbi --- drivers/usb/musb/musb_core.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c index ba13529cbd52..3de9087ecad4 100644 --- a/drivers/usb/musb/musb_core.c +++ b/drivers/usb/musb/musb_core.c @@ -132,7 +132,7 @@ static inline struct musb *dev_to_musb(struct device *dev) /*-------------------------------------------------------------------------*/ #ifndef CONFIG_BLACKFIN -static int musb_ulpi_read(struct usb_phy *phy, u32 offset) +static int musb_ulpi_read(struct usb_phy *phy, u32 reg) { void __iomem *addr = phy->io_priv; int i = 0; @@ -151,7 +151,7 @@ static int musb_ulpi_read(struct usb_phy *phy, u32 offset) * ULPICarKitControlDisableUTMI after clearing POWER_SUSPENDM. */ - musb_writeb(addr, MUSB_ULPI_REG_ADDR, (u8)offset); + musb_writeb(addr, MUSB_ULPI_REG_ADDR, (u8)reg); musb_writeb(addr, MUSB_ULPI_REG_CONTROL, MUSB_ULPI_REG_REQ | MUSB_ULPI_RDN_WR); @@ -176,7 +176,7 @@ out: return ret; } -static int musb_ulpi_write(struct usb_phy *phy, u32 offset, u32 data) +static int musb_ulpi_write(struct usb_phy *phy, u32 val, u32 reg) { void __iomem *addr = phy->io_priv; int i = 0; @@ -191,8 +191,8 @@ static int musb_ulpi_write(struct usb_phy *phy, u32 offset, u32 data) power &= ~MUSB_POWER_SUSPENDM; musb_writeb(addr, MUSB_POWER, power); - musb_writeb(addr, MUSB_ULPI_REG_ADDR, (u8)offset); - musb_writeb(addr, MUSB_ULPI_REG_DATA, (u8)data); + musb_writeb(addr, MUSB_ULPI_REG_ADDR, (u8)reg); + musb_writeb(addr, MUSB_ULPI_REG_DATA, (u8)val); musb_writeb(addr, MUSB_ULPI_REG_CONTROL, MUSB_ULPI_REG_REQ); while (!(musb_readb(addr, MUSB_ULPI_REG_CONTROL) From 2c2025b41aeff57963f9ae2dd909fea704c625ab Mon Sep 17 00:00:00 2001 From: Aaro Koskinen Date: Tue, 27 Oct 2015 23:26:33 +0200 Subject: [PATCH 145/335] usb: phy: omap-otg: fix uninitialized pointer otg_dev->extcon was referenced before otg_dev was initialized. Fix. Cc: # v4.3 Fixes: a2fd2423240f ("usb: phy: omap-otg: Replace deprecated API of extcon") Reviewed-by: Chanwoo Choi Signed-off-by: Aaro Koskinen Signed-off-by: Felipe Balbi --- drivers/usb/phy/phy-omap-otg.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/phy/phy-omap-otg.c b/drivers/usb/phy/phy-omap-otg.c index 1270906ccb95..c4bf2de6d14e 100644 --- a/drivers/usb/phy/phy-omap-otg.c +++ b/drivers/usb/phy/phy-omap-otg.c @@ -105,7 +105,6 @@ static int omap_otg_probe(struct platform_device *pdev) extcon = extcon_get_extcon_dev(config->extcon); if (!extcon) return -EPROBE_DEFER; - otg_dev->extcon = extcon; otg_dev = devm_kzalloc(&pdev->dev, sizeof(*otg_dev), GFP_KERNEL); if (!otg_dev) @@ -115,6 +114,7 @@ static int omap_otg_probe(struct platform_device *pdev) if (IS_ERR(otg_dev->base)) return PTR_ERR(otg_dev->base); + otg_dev->extcon = extcon; otg_dev->id_nb.notifier_call = omap_otg_id_notifier; otg_dev->vbus_nb.notifier_call = omap_otg_vbus_notifier; From 0ba58de231066e47de87ccc4d61c5e396fe9bd27 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 27 Aug 2015 22:06:52 +0200 Subject: [PATCH 146/335] drivers: sh: Get rid of CONFIG_ARCH_SHMOBILE_MULTI Shmobile is all multiplatform these days, so get rid of the reference to CONFIG_ARCH_SHMOBILE_MULTI. Signed-off-by: Geert Uytterhoeven Signed-off-by: Simon Horman --- drivers/sh/pm_runtime.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/sh/pm_runtime.c b/drivers/sh/pm_runtime.c index 25abd4eb7d10..91a003011acf 100644 --- a/drivers/sh/pm_runtime.c +++ b/drivers/sh/pm_runtime.c @@ -34,7 +34,7 @@ static struct pm_clk_notifier_block platform_bus_notifier = { static int __init sh_pm_runtime_init(void) { - if (IS_ENABLED(CONFIG_ARCH_SHMOBILE_MULTI)) { + if (IS_ENABLED(CONFIG_ARCH_SHMOBILE)) { if (!of_find_compatible_node(NULL, NULL, "renesas,cpg-mstp-clocks")) return 0; From 59536da34513c594af2a6fd35ba65ea45b6960a1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Mon, 16 Nov 2015 13:15:46 +0100 Subject: [PATCH 147/335] USB: qcserial: Fix support for HP lt4112 LTE/HSPA+ Gobi 4G Modem MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The DEVICE_HWI type was added under the faulty assumption that Huawei devices based on Qualcomm chipsets and firmware use the static USB interface numbering known from Gobi devices. But this model does not apply to Huawei devices like the HP branded lt4112 (Huawei me906e). Huawei firmwares will dynamically assign interface numbers. Functions are renumbered when the firmware is reconfigured. Fix by changing the DEVICE_HWI type to use a simplified version of Huawei's subclass + protocol scheme: Blacklisting known network interface combinations and assuming the rest are serial. Reported-and-tested-by: Muri Nicanor Tested-by: Martin Hauke Cc: Fixes: e7181d005e84 ("USB: qcserial: Add support for HP lt4112 LTE/HSPA+ Gobi 4G Modem") Signed-off-by: Bjørn Mork Signed-off-by: Johan Hovold --- drivers/usb/serial/qcserial.c | 54 ++++++++++++++++++++++------------- 1 file changed, 34 insertions(+), 20 deletions(-) diff --git a/drivers/usb/serial/qcserial.c b/drivers/usb/serial/qcserial.c index 0e46af0d0cb4..9919d2a9faf2 100644 --- a/drivers/usb/serial/qcserial.c +++ b/drivers/usb/serial/qcserial.c @@ -215,6 +215,10 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) int altsetting = -1; bool sendsetup = false; + /* we only support vendor specific functions */ + if (intf->desc.bInterfaceClass != USB_CLASS_VENDOR_SPEC) + goto done; + nintf = serial->dev->actconfig->desc.bNumInterfaces; dev_dbg(dev, "Num Interfaces = %d\n", nintf); ifnum = intf->desc.bInterfaceNumber; @@ -341,29 +345,39 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) break; case QCSERIAL_HWI: /* - * Huawei layout: - * 0: AT-capable modem port - * 1: DM/DIAG - * 2: AT-capable modem port - * 3: CCID-compatible PCSC interface - * 4: QMI/net - * 5: NMEA + * Huawei devices map functions by subclass + protocol + * instead of interface numbers. The protocol identify + * a specific function, while the subclass indicate a + * specific firmware source + * + * This is a blacklist of functions known to be + * non-serial. The rest are assumed to be serial and + * will be handled by this driver */ - switch (ifnum) { - case 0: - case 2: - dev_dbg(dev, "Modem port found\n"); - break; - case 1: - dev_dbg(dev, "DM/DIAG interface found\n"); - break; - case 5: - dev_dbg(dev, "NMEA GPS interface found\n"); - break; - default: - /* don't claim any unsupported interface */ + switch (intf->desc.bInterfaceProtocol) { + /* QMI combined (qmi_wwan) */ + case 0x07: + case 0x37: + case 0x67: + /* QMI data (qmi_wwan) */ + case 0x08: + case 0x38: + case 0x68: + /* QMI control (qmi_wwan) */ + case 0x09: + case 0x39: + case 0x69: + /* NCM like (huawei_cdc_ncm) */ + case 0x16: + case 0x46: + case 0x76: altsetting = -1; break; + default: + dev_dbg(dev, "Huawei type serial port found (%02x/%02x/%02x)\n", + intf->desc.bInterfaceClass, + intf->desc.bInterfaceSubClass, + intf->desc.bInterfaceProtocol); } break; default: From ef83b6e8f40bb24b92ad73b5889732346e54a793 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 29 Sep 2015 15:48:11 -0400 Subject: [PATCH 148/335] ext2, ext4: warn when mounting with dax enabled Similar to XFS warn when mounting DAX while it is still considered under development. Also, aspects of the DAX implementation, for example synchronization against multiple faults and faults causing block allocation, depend on the correct implementation in the filesystem. The maturity of a given DAX implementation is filesystem specific. Cc: Cc: "Theodore Ts'o" Cc: Matthew Wilcox Cc: linux-ext4@vger.kernel.org Cc: Kirill A. Shutemov Reported-by: Dave Chinner Acked-by: Jan Kara Signed-off-by: Dan Williams --- fs/ext2/super.c | 2 ++ fs/ext4/super.c | 6 +++++- 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 3a71cea68420..748d35afc902 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -569,6 +569,8 @@ static int parse_options(char *options, struct super_block *sb) /* Fall through */ case Opt_dax: #ifdef CONFIG_FS_DAX + ext2_msg(sb, KERN_WARNING, + "DAX enabled. Warning: EXPERIMENTAL, use at your own risk"); set_opt(sbi->s_mount_opt, DAX); #else ext2_msg(sb, KERN_INFO, "dax option not supported"); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 753f4e68b820..c9ab67da6e5a 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1664,8 +1664,12 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, } sbi->s_jquota_fmt = m->mount_opt; #endif -#ifndef CONFIG_FS_DAX } else if (token == Opt_dax) { +#ifdef CONFIG_FS_DAX + ext4_msg(sb, KERN_WARNING, + "DAX enabled. Warning: EXPERIMENTAL, use at your own risk"); + sbi->s_mount_opt |= m->mount_opt; +#else ext4_msg(sb, KERN_INFO, "dax option not supported"); return -1; #endif From d42d5b6f729929f2bbd17d22a2b223b7138470da Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 17:56:39 +0100 Subject: [PATCH 149/335] hwmon: (scpi) add thermal-of dependency The newly added scpi thermal support is broken when the scpi driver is built-in but the thermal driver is a loadable module: drivers/built-in.o: In function `scpi_hwmon_probe': (.text+0x444d70): undefined reference to `thermal_zone_of_sensor_unregister' (.text+0x444d94): undefined reference to `thermal_zone_of_sensor_register' drivers/built-in.o: In function `scpi_hwmon_remove': (text+0x444e6c): undefined reference to `thermal_zone_of_sensor_unregister' This uses the same Kconfig trick that we have in a couple of other drivers already to ensure we can only select the driver in valid configurations when either THERMAL_OF is disabled, or when with a dependency on CONFIG_THERMAL that can force SCPI to be a loadable module in the case I was hitting. Signed-off-by: Arnd Bergmann Fixes: 68acc77a2d51 ("hwmon: Support thermal zones registration for SCP temperature sensors") Signed-off-by: Guenter Roeck --- drivers/hwmon/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig index 6e4e0464e776..8f59f057cdf4 100644 --- a/drivers/hwmon/Kconfig +++ b/drivers/hwmon/Kconfig @@ -324,6 +324,7 @@ config SENSORS_APPLESMC config SENSORS_ARM_SCPI tristate "ARM SCPI Sensors" depends on ARM_SCPI_PROTOCOL + depends on THERMAL || !THERMAL_OF help This driver provides support for temperature, voltage, current and power sensors available on ARM Ltd's SCP based platforms. The From cef03d7e66bd70c4baf61dd2431cf8de250834de Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Wed, 28 Oct 2015 17:17:31 +0000 Subject: [PATCH 150/335] hwmon: (scpi) skip unsupported sensors properly Currently it's assumed that firmware exports only the class of sensors supported by the driver. However with newer firmware or SCPI protocol revision, support for newer classes of sensors can be present. The driver fails to probe with the following warning if an unsupported class of sensor is encountered in the firmware. sysfs: cannot create duplicate filename '/devices/platform/scpi/scpi:sensors/hwmon/hwmon0/' ------------[ cut here ]------------ WARNING: at fs/sysfs/dir.c:31 Modules linked in: CPU: 0 PID: 6 Comm: kworker/u12:0 Not tainted 4.3.0-rc7 #137 Hardware name: ARM Juno development board (r0) (DT) Workqueue: deferwq deferred_probe_work_func PC is at sysfs_warn_dup+0x54/0x78 LR is at sysfs_warn_dup+0x54/0x78 This patch fixes the above issue by skipping through the unsupported class of SCPI sensors. Fixes: 68acc77a2d51 ("hwmon: Support thermal zones registration for SCP temperature sensors") Fixes: ea98b29a05e9 ("hwmon: Support sensors exported via ARM SCP interface") Cc: Guenter Roeck Reviewed-by: Punit Agrawal Signed-off-by: Sudeep Holla Signed-off-by: Guenter Roeck --- drivers/hwmon/scpi-hwmon.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-) diff --git a/drivers/hwmon/scpi-hwmon.c b/drivers/hwmon/scpi-hwmon.c index 2c1241bbf9af..7e20567bc369 100644 --- a/drivers/hwmon/scpi-hwmon.c +++ b/drivers/hwmon/scpi-hwmon.c @@ -117,7 +117,7 @@ static int scpi_hwmon_probe(struct platform_device *pdev) struct scpi_ops *scpi_ops; struct device *hwdev, *dev = &pdev->dev; struct scpi_sensors *scpi_sensors; - int ret; + int ret, idx; scpi_ops = get_scpi_ops(); if (!scpi_ops) @@ -146,8 +146,8 @@ static int scpi_hwmon_probe(struct platform_device *pdev) scpi_sensors->scpi_ops = scpi_ops; - for (i = 0; i < nr_sensors; i++) { - struct sensor_data *sensor = &scpi_sensors->data[i]; + for (i = 0, idx = 0; i < nr_sensors; i++) { + struct sensor_data *sensor = &scpi_sensors->data[idx]; ret = scpi_ops->sensor_get_info(i, &sensor->info); if (ret) @@ -183,7 +183,7 @@ static int scpi_hwmon_probe(struct platform_device *pdev) num_power++; break; default: - break; + continue; } sensor->dev_attr_input.attr.mode = S_IRUGO; @@ -194,11 +194,12 @@ static int scpi_hwmon_probe(struct platform_device *pdev) sensor->dev_attr_label.show = scpi_show_label; sensor->dev_attr_label.attr.name = sensor->label; - scpi_sensors->attrs[i << 1] = &sensor->dev_attr_input.attr; - scpi_sensors->attrs[(i << 1) + 1] = &sensor->dev_attr_label.attr; + scpi_sensors->attrs[idx << 1] = &sensor->dev_attr_input.attr; + scpi_sensors->attrs[(idx << 1) + 1] = &sensor->dev_attr_label.attr; - sysfs_attr_init(scpi_sensors->attrs[i << 1]); - sysfs_attr_init(scpi_sensors->attrs[(i << 1) + 1]); + sysfs_attr_init(scpi_sensors->attrs[idx << 1]); + sysfs_attr_init(scpi_sensors->attrs[(idx << 1) + 1]); + idx++; } scpi_sensors->group.attrs = scpi_sensors->attrs; @@ -236,8 +237,8 @@ static int scpi_hwmon_probe(struct platform_device *pdev) zone->sensor_id = i; zone->scpi_sensors = scpi_sensors; - zone->tzd = thermal_zone_of_sensor_register(dev, i, zone, - &scpi_sensor_ops); + zone->tzd = thermal_zone_of_sensor_register(dev, + sensor->info.sensor_id, zone, &scpi_sensor_ops); /* * The call to thermal_zone_of_sensor_register returns * an error for sensors that are not associated with From dde7632ed02382e4bac2b57c66ee2285764f2cd7 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 17:34:50 +0100 Subject: [PATCH 151/335] clocksource/fsl: Avoid harmless 64-bit warnings The ftm_clockevent_init passes the value of "~0UL" into a function that takes a 32-bit argument, which drops the upper 32 bits, as gcc warns about on ARM64: clocksource/fsl_ftm_timer.c: In function 'ftm_clockevent_init': clocksource/fsl_ftm_timer.c:206:13: warning: large integer implicitly truncated to unsigned type [-Woverflow] This was obviously unintended behavior, and is easily avoided by using '~0u' as the integer literal, because that is 32-bit wide on all architectures. Signed-off-by: Arnd Bergmann Cc: linux-arm-kernel@lists.infradead.org Cc: Xiubo Li Cc: Shawn Guo Cc: Sascha Hauer Cc: Stefan Agner Cc: Daniel Lezcano Link: http://lkml.kernel.org/r/3990834.xnjhm37Grs@wuerfel Signed-off-by: Thomas Gleixner --- drivers/clocksource/fsl_ftm_timer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/clocksource/fsl_ftm_timer.c b/drivers/clocksource/fsl_ftm_timer.c index 10202f1fdfd7..517e1c7624d4 100644 --- a/drivers/clocksource/fsl_ftm_timer.c +++ b/drivers/clocksource/fsl_ftm_timer.c @@ -203,7 +203,7 @@ static int __init ftm_clockevent_init(unsigned long freq, int irq) int err; ftm_writel(0x00, priv->clkevt_base + FTM_CNTIN); - ftm_writel(~0UL, priv->clkevt_base + FTM_MOD); + ftm_writel(~0u, priv->clkevt_base + FTM_MOD); ftm_reset_counter(priv->clkevt_base); @@ -230,7 +230,7 @@ static int __init ftm_clocksource_init(unsigned long freq) int err; ftm_writel(0x00, priv->clksrc_base + FTM_CNTIN); - ftm_writel(~0UL, priv->clksrc_base + FTM_MOD); + ftm_writel(~0u, priv->clksrc_base + FTM_MOD); ftm_reset_counter(priv->clksrc_base); From 3da6d49e847128378c30292848125cc3e207e5f7 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 17:51:17 +0100 Subject: [PATCH 152/335] clocksource: Disallow drivers for ARCH_USES_GETTIMEOFFSET We can now select clocksource drivers like ti-32k and CONFIG_OF on ancient machines that still use gettimeoffset, and the combination results in a link error. arch/arm/kernel/built-in.o: In function `time_init': (.init.text+0xc28): undefined reference to `clocksource_probe' The reason for this is that the Makefile is hidden behind CONFIG_ARCH_USES_GETTIMEOFFSET, but the Kconfig file is not, and it has shown up just now because the ti-32k driver was added and can be selected using COMPILE_TEST on all platforms. This patch hides the Kconfig menu in CONFIG_ARCH_USES_GETTIMEOFFSET as well. Fixes: dfedaf105d60 "clocksource: ti-32k: make it depend on GENERIC_CLOCKSOURCE" Signed-off-by: Arnd Bergmann Cc: linux-arm-kernel@lists.infradead.org Cc: Felipe Balbi Cc: Tony Lindgren Cc: Daniel Lezcano Link: http://lkml.kernel.org/r/7579471.4N90fYPQOK@wuerfel Signed-off-by: Thomas Gleixner --- drivers/clocksource/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig index 71cfdf7c9708..2eb5f0efae90 100644 --- a/drivers/clocksource/Kconfig +++ b/drivers/clocksource/Kconfig @@ -1,4 +1,5 @@ menu "Clock Source drivers" + depends on !ARCH_USES_GETTIMEOFFSET config CLKSRC_OF bool From 96dd922c198286681fbbc15100e196e0f629e2fb Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Wed, 11 Nov 2015 15:36:16 -0800 Subject: [PATCH 153/335] mtd: jz4740_nand: fix build on jz4740 after removing gpio.h Fallout from commit 832f5dacfa0b ("MIPS: Remove all the uses of custom gpio.h") We see errors like this: drivers/mtd/nand/jz4740_nand.c: In function 'jz_nand_detect_bank': drivers/mtd/nand/jz4740_nand.c:340:9: error: 'JZ_GPIO_MEM_CS0' undeclared (first use in this function) drivers/mtd/nand/jz4740_nand.c:340:9: note: each undeclared identifier is reported only once for each function it appears in drivers/mtd/nand/jz4740_nand.c:359:2: error: implicit declaration of function 'jz_gpio_set_function' [-Werror=implicit-function-declaration] drivers/mtd/nand/jz4740_nand.c:359:29: error: 'JZ_GPIO_FUNC_MEM_CS0' undeclared (first use in this function) drivers/mtd/nand/jz4740_nand.c:399:29: error: 'JZ_GPIO_FUNC_NONE' undeclared (first use in this function) drivers/mtd/nand/jz4740_nand.c: In function 'jz_nand_probe': drivers/mtd/nand/jz4740_nand.c:528:13: error: 'JZ_GPIO_MEM_CS0' undeclared (first use in this function) drivers/mtd/nand/jz4740_nand.c: In function 'jz_nand_remove': drivers/mtd/nand/jz4740_nand.c:555:14: error: 'JZ_GPIO_MEM_CS0' undeclared (first use in this function) Patched similarly to: https://patchwork.linux-mips.org/patch/11089/ Fixes: 832f5dacfa0b ("MIPS: Remove all the uses of custom gpio.h") Signed-off-by: Brian Norris --- drivers/mtd/nand/jz4740_nand.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mtd/nand/jz4740_nand.c b/drivers/mtd/nand/jz4740_nand.c index dc4e8446f1ff..5a99a93ed025 100644 --- a/drivers/mtd/nand/jz4740_nand.c +++ b/drivers/mtd/nand/jz4740_nand.c @@ -25,6 +25,7 @@ #include +#include #include #define JZ_REG_NAND_CTRL 0x50 From 9ca641b0f02a3a1eedbc8c296e695326da9bbaf9 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Mon, 9 Nov 2015 16:37:28 -0800 Subject: [PATCH 154/335] mtd: nand: fix shutdown/reboot for multi-chip systems If multiple NAND chips are registered to the same controller, then when rebooting the system, the first one will grab the controller lock, while the second will wait forever for the first one to release it. i.e., a classic deadlock. This problem was solved for a similar case (suspend/resume) back in commit 6b0d9a841249 ("mtd: nand: fix multi-chip suspend problem"), and the shutdown state really isn't much different for us, so rather than adding a new special case to nand_get_device(), we can just overload the FL_PM_SUSPENDED state. Now, multiple chips can "get" the same controller lock (preventing further I/O), while we still allow other chips to pass through nand_shutdown(). Original report: http://thread.gmane.org/gmane.linux.drivers.mtd/59726 http://lists.infradead.org/pipermail/linux-mtd/2015-July/059992.html Fixes: 72ea403669c7 ("mtd: nand: added nand_shutdown") Reported-by: Andrew E. Mileski Signed-off-by: Brian Norris Cc: Scott Branden Cc: Andrew E. Mileski Acked-by: Scott Branden Reviewed-by: Boris Brezillon --- drivers/mtd/nand/nand_base.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c index cc74142938b0..ece544efccc3 100644 --- a/drivers/mtd/nand/nand_base.c +++ b/drivers/mtd/nand/nand_base.c @@ -3110,7 +3110,7 @@ static void nand_resume(struct mtd_info *mtd) */ static void nand_shutdown(struct mtd_info *mtd) { - nand_get_device(mtd, FL_SHUTDOWN); + nand_get_device(mtd, FL_PM_SUSPENDED); } /* Set default functions */ From b4fe85f9c9146f60457e9512fb6055e69e6a7a65 Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Thu, 12 Nov 2015 17:35:58 +0100 Subject: [PATCH 155/335] ip_tunnel: disable preemption when updating per-cpu tstats Drivers like vxlan use the recently introduced udp_tunnel_xmit_skb/udp_tunnel6_xmit_skb APIs. udp_tunnel6_xmit_skb makes use of ip6tunnel_xmit, and ip6tunnel_xmit, after sending the packet, updates the struct stats using the usual u64_stats_update_begin/end calls on this_cpu_ptr(dev->tstats). udp_tunnel_xmit_skb makes use of iptunnel_xmit, which doesn't touch tstats, so drivers like vxlan, immediately after, call iptunnel_xmit_stats, which does the same thing - calls u64_stats_update_begin/end on this_cpu_ptr(dev->tstats). While vxlan is probably fine (I don't know?), calling a similar function from, say, an unbound workqueue, on a fully preemptable kernel causes real issues: [ 188.434537] BUG: using smp_processor_id() in preemptible [00000000] code: kworker/u8:0/6 [ 188.435579] caller is debug_smp_processor_id+0x17/0x20 [ 188.435583] CPU: 0 PID: 6 Comm: kworker/u8:0 Not tainted 4.2.6 #2 [ 188.435607] Call Trace: [ 188.435611] [] dump_stack+0x4f/0x7b [ 188.435615] [] check_preemption_disabled+0x19d/0x1c0 [ 188.435619] [] debug_smp_processor_id+0x17/0x20 The solution would be to protect the whole this_cpu_ptr(dev->tstats)/u64_stats_update_begin/end blocks with disabling preemption and then reenabling it. Signed-off-by: Jason A. Donenfeld Acked-by: Hannes Frederic Sowa Signed-off-by: David S. Miller --- include/net/ip6_tunnel.h | 3 ++- include/net/ip_tunnels.h | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/include/net/ip6_tunnel.h b/include/net/ip6_tunnel.h index aaee6fa02cf1..ff788b665277 100644 --- a/include/net/ip6_tunnel.h +++ b/include/net/ip6_tunnel.h @@ -90,11 +90,12 @@ static inline void ip6tunnel_xmit(struct sock *sk, struct sk_buff *skb, err = ip6_local_out(dev_net(skb_dst(skb)->dev), sk, skb); if (net_xmit_eval(err) == 0) { - struct pcpu_sw_netstats *tstats = this_cpu_ptr(dev->tstats); + struct pcpu_sw_netstats *tstats = get_cpu_ptr(dev->tstats); u64_stats_update_begin(&tstats->syncp); tstats->tx_bytes += pkt_len; tstats->tx_packets++; u64_stats_update_end(&tstats->syncp); + put_cpu_ptr(tstats); } else { stats->tx_errors++; stats->tx_aborted_errors++; diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h index f6dafec9102c..62a750a6a8f8 100644 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@ -287,12 +287,13 @@ static inline void iptunnel_xmit_stats(int err, struct pcpu_sw_netstats __percpu *stats) { if (err > 0) { - struct pcpu_sw_netstats *tstats = this_cpu_ptr(stats); + struct pcpu_sw_netstats *tstats = get_cpu_ptr(stats); u64_stats_update_begin(&tstats->syncp); tstats->tx_bytes += err; tstats->tx_packets++; u64_stats_update_end(&tstats->syncp); + put_cpu_ptr(tstats); } else if (err < 0) { err_stats->tx_errors++; err_stats->tx_aborted_errors++; From 0eae5982a3571dc15a9e7684c02e8774ba2fb1c7 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A5ns=20Rullg=C3=A5rd?= Date: Thu, 12 Nov 2015 17:40:20 +0000 Subject: [PATCH 156/335] net: phy: at803x: support interrupt on 8030 and 8035 Commit 77a993942 "phy/at8031: enable at8031 to work on interrupt mode" added interrupt support for the 8031 PHY but left out the other two chips supported by this driver. This patch sets the .ack_interrupt and .config_intr functions for the 8030 and 8035 drivers as well. Signed-off-by: Mans Rullgard Reviewed-by: Florian Fainelli Signed-off-by: David S. Miller --- drivers/net/phy/at803x.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/phy/at803x.c b/drivers/net/phy/at803x.c index fabf11d32d27..2d020a3ec0b5 100644 --- a/drivers/net/phy/at803x.c +++ b/drivers/net/phy/at803x.c @@ -308,6 +308,8 @@ static struct phy_driver at803x_driver[] = { .flags = PHY_HAS_INTERRUPT, .config_aneg = genphy_config_aneg, .read_status = genphy_read_status, + .ack_interrupt = at803x_ack_interrupt, + .config_intr = at803x_config_intr, .driver = { .owner = THIS_MODULE, }, @@ -327,6 +329,8 @@ static struct phy_driver at803x_driver[] = { .flags = PHY_HAS_INTERRUPT, .config_aneg = genphy_config_aneg, .read_status = genphy_read_status, + .ack_interrupt = at803x_ack_interrupt, + .config_intr = at803x_config_intr, .driver = { .owner = THIS_MODULE, }, From 7729b053814ac91af340f5055970afc87d7fee21 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A5ns=20Rullg=C3=A5rd?= Date: Thu, 12 Nov 2015 18:41:12 +0000 Subject: [PATCH 157/335] net: phy: vitesse: add support for VSC8601 This adds support for the Vitesse VSC8601 PHY. Generic functions are used for everything except interrupt handling. Signed-off-by: Mans Rullgard Reviewed-by: Florian Fainelli Signed-off-by: David S. Miller --- drivers/net/phy/vitesse.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/net/phy/vitesse.c b/drivers/net/phy/vitesse.c index 76cad712ddb2..dd295dbaa074 100644 --- a/drivers/net/phy/vitesse.c +++ b/drivers/net/phy/vitesse.c @@ -66,6 +66,7 @@ #define PHY_ID_VSC8244 0x000fc6c0 #define PHY_ID_VSC8514 0x00070670 #define PHY_ID_VSC8574 0x000704a0 +#define PHY_ID_VSC8601 0x00070420 #define PHY_ID_VSC8662 0x00070660 #define PHY_ID_VSC8221 0x000fc550 #define PHY_ID_VSC8211 0x000fc4b0 @@ -133,7 +134,8 @@ static int vsc82xx_config_intr(struct phy_device *phydev) (phydev->drv->phy_id == PHY_ID_VSC8234 || phydev->drv->phy_id == PHY_ID_VSC8244 || phydev->drv->phy_id == PHY_ID_VSC8514 || - phydev->drv->phy_id == PHY_ID_VSC8574) ? + phydev->drv->phy_id == PHY_ID_VSC8574 || + phydev->drv->phy_id == PHY_ID_VSC8601) ? MII_VSC8244_IMASK_MASK : MII_VSC8221_IMASK_MASK); else { @@ -271,6 +273,18 @@ static struct phy_driver vsc82xx_driver[] = { .ack_interrupt = &vsc824x_ack_interrupt, .config_intr = &vsc82xx_config_intr, .driver = { .owner = THIS_MODULE,}, +}, { + .phy_id = PHY_ID_VSC8601, + .name = "Vitesse VSC8601", + .phy_id_mask = 0x000ffff0, + .features = PHY_GBIT_FEATURES, + .flags = PHY_HAS_INTERRUPT, + .config_init = &genphy_config_init, + .config_aneg = &genphy_config_aneg, + .read_status = &genphy_read_status, + .ack_interrupt = &vsc824x_ack_interrupt, + .config_intr = &vsc82xx_config_intr, + .driver = { .owner = THIS_MODULE,}, }, { .phy_id = PHY_ID_VSC8662, .name = "Vitesse VSC8662", From 0fcd593b943bfcc21ad84d3321422401de071d8a Mon Sep 17 00:00:00 2001 From: Yang Shi Date: Thu, 12 Nov 2015 13:57:00 -0800 Subject: [PATCH 158/335] arm64: bpf: fix JIT frame pointer setup BPF fp should point to the top of the BPF prog stack. The original implementation made it point to the bottom incorrectly. Move A64_SP to fp before reserve BPF prog stack space. CC: Zi Shen Lim CC: Xi Wang Signed-off-by: Yang Shi Reviewed-by: Zi Shen Lim Signed-off-by: David S. Miller --- arch/arm64/net/bpf_jit_comp.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index a44e5293c6f5..ac8b548ce493 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -161,12 +161,12 @@ static void build_prologue(struct jit_ctx *ctx) if (ctx->tmp_used) emit(A64_PUSH(tmp1, tmp2, A64_SP), ctx); - /* Set up BPF stack */ - emit(A64_SUB_I(1, A64_SP, A64_SP, stack_size), ctx); - /* Set up frame pointer */ emit(A64_MOV(1, fp, A64_SP), ctx); + /* Set up BPF stack */ + emit(A64_SUB_I(1, A64_SP, A64_SP, stack_size), ctx); + /* Clear registers A and X */ emit_a64_mov_i64(ra, 0, ctx); emit_a64_mov_i64(rx, 0, ctx); From 30b50aa612018bd92f5a85534cc2668423e8c7e8 Mon Sep 17 00:00:00 2001 From: Yang Shi Date: Thu, 12 Nov 2015 14:07:46 -0800 Subject: [PATCH 159/335] bpf: samples: exclude asm/sysreg.h for arm64 commit 338d4f49d6f7114a017d294ccf7374df4f998edc ("arm64: kernel: Add support for Privileged Access Never") includes sysreg.h into futex.h and uaccess.h. But, the inline assembly used by asm/sysreg.h is incompatible with llvm so it will cause BPF samples build failure for ARM64. Since sysreg.h is useless for BPF samples, just exclude it from Makefile via defining __ASM_SYSREG_H. Signed-off-by: Yang Shi Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- samples/bpf/Makefile | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile index 79b4596b5f9a..edd638b5825f 100644 --- a/samples/bpf/Makefile +++ b/samples/bpf/Makefile @@ -67,10 +67,13 @@ HOSTLOADLIBES_lathist += -lelf # point this to your LLVM backend with bpf support LLC=$(srctree)/tools/bpf/llvm/bld/Debug+Asserts/bin/llc +# asm/sysreg.h inline assmbly used by it is incompatible with llvm. +# But, ehere is not easy way to fix it, so just exclude it since it is +# useless for BPF samples. $(obj)/%.o: $(src)/%.c clang $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \ - -D__KERNEL__ -Wno-unused-value -Wno-pointer-sign \ + -D__KERNEL__ -D__ASM_SYSREG_H -Wno-unused-value -Wno-pointer-sign \ -O2 -emit-llvm -c $< -o -| $(LLC) -march=bpf -filetype=obj -o $@ clang $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \ - -D__KERNEL__ -Wno-unused-value -Wno-pointer-sign \ + -D__KERNEL__ -D__ASM_SYSREG_H -Wno-unused-value -Wno-pointer-sign \ -O2 -emit-llvm -c $< -o -| $(LLC) -march=bpf -filetype=asm -o $@.s From ebaef649c26b44ff28114b452fd067a270ca7f02 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 14 Nov 2015 01:26:53 +0100 Subject: [PATCH 160/335] bpf, arm: start flushing icache range from header During review I noticed that the icache range we're flushing should start at header already and not at ctx.image. Reason is that after 55309dd3d4cd ("net: bpf: arm: address randomize and write protect JIT code"), we also want to make sure to flush the random-sized trap in front of the start of the actual program (analogous to x86). No operational differences from user side. Signed-off-by: Daniel Borkmann Tested-by: Nicolas Schichan Cc: Alexei Starovoitov Signed-off-by: David S. Miller --- arch/arm/net/bpf_jit_32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/net/bpf_jit_32.c b/arch/arm/net/bpf_jit_32.c index 2f4b14cfddb4..591f9db3bf40 100644 --- a/arch/arm/net/bpf_jit_32.c +++ b/arch/arm/net/bpf_jit_32.c @@ -1061,7 +1061,7 @@ void bpf_jit_compile(struct bpf_prog *fp) } build_epilogue(&ctx); - flush_icache_range((u32)ctx.target, (u32)(ctx.target + ctx.idx)); + flush_icache_range((u32)header, (u32)(ctx.target + ctx.idx)); #if __LINUX_ARM_ARCH__ < 7 if (ctx.imm_count) From c3d4c682c240595c3637c552dc1afa985bc6d382 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 14 Nov 2015 01:16:18 +0100 Subject: [PATCH 161/335] bpf, arm64: start flushing icache range from header While recently going over ARM64's BPF code, I noticed that the icache range we're flushing should start at header already and not at ctx.image. Reason is that after b569c1c622c5 ("net: bpf: arm64: address randomize and write protect JIT code"), we also want to make sure to flush the random-sized trap in front of the start of the actual program (analogous to x86). No operational differences from user side. Signed-off-by: Daniel Borkmann Acked-by: Zi Shen Lim Cc: Alexei Starovoitov Signed-off-by: David S. Miller --- arch/arm64/net/bpf_jit_comp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index ac8b548ce493..64a8bc12f4a2 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -740,7 +740,7 @@ void bpf_int_jit_compile(struct bpf_prog *prog) if (bpf_jit_enable > 1) bpf_jit_dump(prog->len, image_size, 2, ctx.image); - bpf_flush_icache(ctx.image, ctx.image + ctx.idx); + bpf_flush_icache(header, ctx.image + ctx.idx); set_memory_ro((unsigned long)header, header->pages); prog->bpf_func = (void *)ctx.image; From cd998ecd2f031cfdb88436ea12f7c6d0b09c7a80 Mon Sep 17 00:00:00 2001 From: Pavel Fedin Date: Fri, 13 Nov 2015 09:46:59 +0300 Subject: [PATCH 162/335] net: smsc911x: Reset PHY during initialization On certain hardware after software reboot the chip may get stuck and fail to reinitialize during reset. This can be fixed by ensuring that PHY is reset too. Old PHY resetting method required operational MDIO interface, therefore the chip should have been already set up. In order to be able to function during probe, it is changed to use PMT_CTRL register. The problem could be observed on SMDK5410 board. Signed-off-by: Pavel Fedin Signed-off-by: David S. Miller --- drivers/net/ethernet/smsc/smsc911x.c | 17 ++++++----------- 1 file changed, 6 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/smsc/smsc911x.c b/drivers/net/ethernet/smsc/smsc911x.c index c860c9007e49..219a99b7a631 100644 --- a/drivers/net/ethernet/smsc/smsc911x.c +++ b/drivers/net/ethernet/smsc/smsc911x.c @@ -809,22 +809,17 @@ static int smsc911x_phy_check_loopbackpkt(struct smsc911x_data *pdata) static int smsc911x_phy_reset(struct smsc911x_data *pdata) { - struct phy_device *phy_dev = pdata->phy_dev; unsigned int temp; unsigned int i = 100000; - BUG_ON(!phy_dev); - BUG_ON(!phy_dev->bus); - - SMSC_TRACE(pdata, hw, "Performing PHY BCR Reset"); - smsc911x_mii_write(phy_dev->bus, phy_dev->addr, MII_BMCR, BMCR_RESET); + temp = smsc911x_reg_read(pdata, PMT_CTRL); + smsc911x_reg_write(pdata, PMT_CTRL, temp | PMT_CTRL_PHY_RST_); do { msleep(1); - temp = smsc911x_mii_read(phy_dev->bus, phy_dev->addr, - MII_BMCR); - } while ((i--) && (temp & BMCR_RESET)); + temp = smsc911x_reg_read(pdata, PMT_CTRL); + } while ((i--) && (temp & PMT_CTRL_PHY_RST_)); - if (temp & BMCR_RESET) { + if (unlikely(temp & PMT_CTRL_PHY_RST_)) { SMSC_WARN(pdata, hw, "PHY reset failed to complete"); return -EIO; } @@ -2296,7 +2291,7 @@ static int smsc911x_init(struct net_device *dev) } /* Reset the LAN911x */ - if (smsc911x_soft_reset(pdata)) + if (smsc911x_phy_reset(pdata) || smsc911x_soft_reset(pdata)) return -ENODEV; dev->flags |= IFF_MULTICAST; From 2452cb0c6503f05ab371b26c3216c5fba54d4476 Mon Sep 17 00:00:00 2001 From: Masaru Nagai Date: Fri, 13 Nov 2015 19:24:49 +0900 Subject: [PATCH 163/335] ravb: Fix int mask value overwritten issue When RX/TX interrupt for Network Control queue and Best Effort queue is issued at the same time, the interrupt mask of Network Control queue will be reset when the mask of Best Effort queue is set. This patch fixes this problem. Signed-off-by: Masaru Nagai Signed-off-by: Yoshihiro Kaneko Acked-by: Geert Uytterhoeven Acked-by: Simon Horman Signed-off-by: David S. Miller --- drivers/net/ethernet/renesas/ravb_main.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index aa7b2083cb53..7180e26f5c67 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -733,8 +733,10 @@ static irqreturn_t ravb_interrupt(int irq, void *dev_id) ((tis & tic) & BIT(q))) { if (napi_schedule_prep(&priv->napi[q])) { /* Mask RX and TX interrupts */ - ravb_write(ndev, ric0 & ~BIT(q), RIC0); - ravb_write(ndev, tic & ~BIT(q), TIC); + ric0 &= ~BIT(q); + tic &= ~BIT(q); + ravb_write(ndev, ric0, RIC0); + ravb_write(ndev, tic, TIC); __napi_schedule(&priv->napi[q]); } else { netdev_warn(ndev, From 4114ec905de2f02b91703c484bba9b62dc3bbb87 Mon Sep 17 00:00:00 2001 From: Ivan Vecera Date: Fri, 13 Nov 2015 11:36:57 +0100 Subject: [PATCH 164/335] be2net: remove unused local rsstable array Remove rsstable array and its initialization from be_set_rss_hash_opts(). The array became unused after "e255787 be2net: Support for configurable RSS hash key". The initial RSS table is now filled and stored for later usage during Rx queue creation. Signed-off-by: Ivan Vecera Acked-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be_ethtool.c | 16 ++-------------- 1 file changed, 2 insertions(+), 14 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c index f4cb8e425853..7246eac5148a 100644 --- a/drivers/net/ethernet/emulex/benet/be_ethtool.c +++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c @@ -1062,9 +1062,7 @@ static int be_get_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd, static int be_set_rss_hash_opts(struct be_adapter *adapter, struct ethtool_rxnfc *cmd) { - struct be_rx_obj *rxo; - int status = 0, i, j; - u8 rsstable[128]; + int status; u32 rss_flags = adapter->rss_info.rss_flags; if (cmd->data != L3_RSS_FLAGS && @@ -1113,17 +1111,7 @@ static int be_set_rss_hash_opts(struct be_adapter *adapter, } if (rss_flags == adapter->rss_info.rss_flags) - return status; - - if (be_multi_rxq(adapter)) { - for (j = 0; j < 128; j += adapter->num_rss_qs) { - for_all_rss_queues(adapter, rxo, i) { - if ((j + i) >= 128) - break; - rsstable[j + i] = rxo->rss_id; - } - } - } + return 0; status = be_cmd_rss_config(adapter, adapter->rss_info.rsstable, rss_flags, 128, adapter->rss_info.rss_hkey); From d5d309815bc377ec8235df380fda6d1befbc57cc Mon Sep 17 00:00:00 2001 From: Ivan Vecera Date: Fri, 13 Nov 2015 11:36:58 +0100 Subject: [PATCH 165/335] be2net: replace hardcoded values with existing define Signed-off-by: Ivan Vecera Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be_ethtool.c | 3 ++- drivers/net/ethernet/emulex/benet/be_main.c | 2 +- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c index 7246eac5148a..734f655c99c1 100644 --- a/drivers/net/ethernet/emulex/benet/be_ethtool.c +++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c @@ -1114,7 +1114,8 @@ static int be_set_rss_hash_opts(struct be_adapter *adapter, return 0; status = be_cmd_rss_config(adapter, adapter->rss_info.rsstable, - rss_flags, 128, adapter->rss_info.rss_hkey); + rss_flags, RSS_INDIR_TABLE_LEN, + adapter->rss_info.rss_hkey); if (!status) adapter->rss_info.rss_flags = rss_flags; diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c index eb48a977f8da..b6ad02909d6b 100644 --- a/drivers/net/ethernet/emulex/benet/be_main.c +++ b/drivers/net/ethernet/emulex/benet/be_main.c @@ -3518,7 +3518,7 @@ static int be_rx_qs_create(struct be_adapter *adapter) netdev_rss_key_fill(rss_key, RSS_HASH_KEY_LEN); rc = be_cmd_rss_config(adapter, rss->rsstable, rss->rss_flags, - 128, rss_key); + RSS_INDIR_TABLE_LEN, rss_key); if (rc) { rss->rss_flags = RSS_ENABLE_NONE; return rc; From bbe14f54297467ddb23b7d1db564a2468c6ae151 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Fri, 13 Nov 2015 13:06:12 +0200 Subject: [PATCH 166/335] switchdev: bridge: Check return code is not EOPNOTSUPP When NET_SWITCHDEV=n, switchdev_port_attr_set simply returns EOPNOTSUPP. In this case we should not emit errors and warnings to the kernel log. Reported-by: Sander Eikelenboom Tested-by: Christian Borntraeger Fixes: 0bc05d585d38 ("switchdev: allow caller to explicitly request attr_set as deferred") Fixes: 6ac311ae8bfb ("Adding switchdev ageing notification on port bridged") Signed-off-by: Ido Schimmel Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller --- net/bridge/br_stp.c | 2 +- net/bridge/br_stp_if.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/bridge/br_stp.c b/net/bridge/br_stp.c index f7e8dee64fc8..5f3f64553179 100644 --- a/net/bridge/br_stp.c +++ b/net/bridge/br_stp.c @@ -48,7 +48,7 @@ void br_set_state(struct net_bridge_port *p, unsigned int state) p->state = state; err = switchdev_port_attr_set(p->dev, &attr); - if (err) + if (err && err != -EOPNOTSUPP) br_warn(p->br, "error setting offload STP state on port %u(%s)\n", (unsigned int) p->port_no, p->dev->name); } diff --git a/net/bridge/br_stp_if.c b/net/bridge/br_stp_if.c index fa53d7a89f48..5396ff08af32 100644 --- a/net/bridge/br_stp_if.c +++ b/net/bridge/br_stp_if.c @@ -50,7 +50,7 @@ void br_init_port(struct net_bridge_port *p) p->config_pending = 0; err = switchdev_port_attr_set(p->dev, &attr); - if (err) + if (err && err != -EOPNOTSUPP) netdev_err(p->dev, "failed to set HW ageing time\n"); } From 5f8dc33e8ee7e59bee3bc6dc2088807a384b285a Mon Sep 17 00:00:00 2001 From: Nikolay Aleksandrov Date: Fri, 13 Nov 2015 14:54:01 +0100 Subject: [PATCH 167/335] net: fix feature changes on devices without ndo_set_features When __netdev_update_features() was updated to ensure some features are disabled on new lower devices, an error was introduced for devices which don't have the ndo_set_features() method set. Before we'll just set the new features, but now we return an error and don't set them. Fix this by returning the old behaviour and setting err to 0 when ndo_set_features is not present. Fixes: e7868a85e1b2 ("net/core: ensure features get disabled on new lower devs") CC: Jarod Wilson CC: Jiri Pirko CC: Ido Schimmel CC: Sander Eikelenboom CC: Andy Gospodarek CC: Florian Fainelli Signed-off-by: Nikolay Aleksandrov Reviewed-by: Jiri Pirko Reviewed-by: Andy Gospodarek Reviewed-by: Jarod Wilson Tested-by: Florian Fainelli Tested-by: Dave Young Tested-by: Christian Borntraeger Signed-off-by: David S. Miller --- net/core/dev.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/core/dev.c b/net/core/dev.c index ab9b8d0d115e..4a1d198dbbff 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6426,6 +6426,8 @@ int __netdev_update_features(struct net_device *dev) if (dev->netdev_ops->ndo_set_features) err = dev->netdev_ops->ndo_set_features(dev, features); + else + err = 0; if (unlikely(err < 0)) { netdev_err(dev, From 00ee5927177792a6e139d50b6b7564d35705556a Mon Sep 17 00:00:00 2001 From: Nikolay Aleksandrov Date: Fri, 13 Nov 2015 15:20:24 +0100 Subject: [PATCH 168/335] net: fix __netdev_update_features return on ndo_set_features failure MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If ndo_set_features fails __netdev_update_features() will return -1 but this is wrong because it is expected to return 0 if no features were changed (see netdev_update_features()), which will cause a netdev notifier to be called without any actual changes. Fix this by returning 0 if ndo_set_features fails. Fixes: 6cb6a27c45ce ("net: Call netdev_features_change() from netdev_update_features()") CC: Michał Mirosław Signed-off-by: Nikolay Aleksandrov Signed-off-by: David S. Miller --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index 4a1d198dbbff..1974aee005a6 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6433,7 +6433,7 @@ int __netdev_update_features(struct net_device *dev) netdev_err(dev, "set_features() failed (%d); wanted %pNF, left %pNF\n", err, &features, &dev->features); - return -1; + return 0; } sync_lower: From 0b88393cdf6b1322522849e61f7a3328f4fd3843 Mon Sep 17 00:00:00 2001 From: Daniele Palmas Date: Fri, 13 Nov 2015 18:01:21 +0100 Subject: [PATCH 169/335] net: usb: cdc_ether: add Dell DW5580 as a mobile broadband adapter Since Dell DW5580 is a 3G modem, this patch adds the device as a mobile broadband adapter Signed-off-by: Daniele Palmas Signed-off-by: David S. Miller --- drivers/net/usb/cdc_ether.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/usb/cdc_ether.c b/drivers/net/usb/cdc_ether.c index c78d3cb1b464..3da70bf9936a 100644 --- a/drivers/net/usb/cdc_ether.c +++ b/drivers/net/usb/cdc_ether.c @@ -695,6 +695,11 @@ static const struct usb_device_id products[] = { USB_VENDOR_AND_INTERFACE_INFO(0x1bc7, USB_CLASS_COMM, USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE), .driver_info = (kernel_ulong_t) &wwan_info, +}, { + /* Dell DW5580 modules */ + USB_DEVICE_AND_INTERFACE_INFO(DELL_VENDOR_ID, 0x81ba, USB_CLASS_COMM, + USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE), + .driver_info = (kernel_ulong_t)&wwan_info, }, { USB_INTERFACE_INFO(USB_CLASS_COMM, USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE), From d37d5ec861b0d937c0bdd70f0138e2750aacccdb Mon Sep 17 00:00:00 2001 From: Shrikrishna Khare Date: Fri, 13 Nov 2015 15:42:10 -0800 Subject: [PATCH 170/335] Driver: Vmxnet3: Fix use of mfTableLen for big endian architectures Signed-off-by: Shrikrishna Khare Reported-by: Masao Uebayashi Signed-off-by: Bhavesh Davda Signed-off-by: David S. Miller --- drivers/net/vmxnet3/vmxnet3_drv.c | 7 ++++--- drivers/net/vmxnet3/vmxnet3_int.h | 4 ++-- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/net/vmxnet3/vmxnet3_drv.c b/drivers/net/vmxnet3/vmxnet3_drv.c index 46f4caddccbe..899ea4288197 100644 --- a/drivers/net/vmxnet3/vmxnet3_drv.c +++ b/drivers/net/vmxnet3/vmxnet3_drv.c @@ -2157,12 +2157,13 @@ vmxnet3_set_mc(struct net_device *netdev) if (!netdev_mc_empty(netdev)) { new_table = vmxnet3_copy_mc(netdev); if (new_table) { - rxConf->mfTableLen = cpu_to_le16( - netdev_mc_count(netdev) * ETH_ALEN); + size_t sz = netdev_mc_count(netdev) * ETH_ALEN; + + rxConf->mfTableLen = cpu_to_le16(sz); new_table_pa = dma_map_single( &adapter->pdev->dev, new_table, - rxConf->mfTableLen, + sz, PCI_DMA_TODEVICE); } diff --git a/drivers/net/vmxnet3/vmxnet3_int.h b/drivers/net/vmxnet3/vmxnet3_int.h index 3f859a55c035..4c58c83dc225 100644 --- a/drivers/net/vmxnet3/vmxnet3_int.h +++ b/drivers/net/vmxnet3/vmxnet3_int.h @@ -69,10 +69,10 @@ /* * Version numbers */ -#define VMXNET3_DRIVER_VERSION_STRING "1.4.3.0-k" +#define VMXNET3_DRIVER_VERSION_STRING "1.4.4.0-k" /* a 32-bit int, each byte encode a verion number in VMXNET3_DRIVER_VERSION */ -#define VMXNET3_DRIVER_VERSION_NUM 0x01040300 +#define VMXNET3_DRIVER_VERSION_NUM 0x01040400 #if defined(CONFIG_PCI_MSI) /* RSS only makes sense if MSI-X is supported. */ From 166e23623e7482070aa124ad805f600672377019 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sat, 14 Nov 2015 11:06:53 +0100 Subject: [PATCH 171/335] net: cavium: liquidio: constify pci_error_handlers structures This pci_error_handlers structure is never modified, like all the other pci_error_handlers structures, so declare it as const. Done with the help of Coccinelle. Signed-off-by: Julia Lawall Signed-off-by: David S. Miller --- drivers/net/ethernet/cavium/liquidio/lio_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/cavium/liquidio/lio_main.c b/drivers/net/ethernet/cavium/liquidio/lio_main.c index f683d97d7614..b89504405b72 100644 --- a/drivers/net/ethernet/cavium/liquidio/lio_main.c +++ b/drivers/net/ethernet/cavium/liquidio/lio_main.c @@ -560,7 +560,7 @@ static int liquidio_resume(struct pci_dev *pdev) #endif /* For PCI-E Advanced Error Recovery (AER) Interface */ -static struct pci_error_handlers liquidio_err_handler = { +static const struct pci_error_handlers liquidio_err_handler = { .error_detected = liquidio_pcie_error_detected, .mmio_enabled = liquidio_pcie_mmio_enabled, .slot_reset = liquidio_pcie_slot_reset, From c300366b6b978fcb84f8eeb6205e5980cc0c40c3 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sat, 14 Nov 2015 11:06:57 +0100 Subject: [PATCH 172/335] sfc: constify pci_error_handlers structures This pci_error_handlers structure is never modified, like all the other pci_error_handlers structures, so declare it as const. Done with the help of Coccinelle. Signed-off-by: Julia Lawall Signed-off-by: David S. Miller --- drivers/net/ethernet/sfc/efx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index d288f1c928de..a3c42a376741 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -3422,7 +3422,7 @@ out: * with our request for slot reset the mmio_enabled callback will never be * called, and the link_reset callback is not used by AER or EEH mechanisms. */ -static struct pci_error_handlers efx_err_handlers = { +static const struct pci_error_handlers efx_err_handlers = { .error_detected = efx_io_error_detected, .slot_reset = efx_io_slot_reset, .resume = efx_io_resume, From 027ac58e3c757277dd6cb8975d1b69c27853aa76 Mon Sep 17 00:00:00 2001 From: Ben Cartwright-Cox Date: Sat, 14 Nov 2015 15:13:58 +0000 Subject: [PATCH 173/335] raw: increment correct SNMP counters for ICMP messages Sending ICMP packets with raw sockets ends up in the SNMP counters logging the type as the first byte of the IPv4 header rather than the ICMP header. This is fixed by adding the IP Header Length to the casting into a icmphdr struct. Signed-off-by: Ben Cartwright-Cox Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/raw.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 8c0d0bdc2a7c..63e5be0abd86 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -406,10 +406,12 @@ static int raw_send_hdrinc(struct sock *sk, struct flowi4 *fl4, ip_select_ident(net, skb, NULL); iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl); + skb->transport_header += iphlen; + if (iph->protocol == IPPROTO_ICMP && + length >= iphlen + sizeof(struct icmphdr)) + icmp_out_count(net, ((struct icmphdr *) + skb_transport_header(skb))->type); } - if (iph->protocol == IPPROTO_ICMP) - icmp_out_count(net, ((struct icmphdr *) - skb_transport_header(skb))->type); err = NF_HOOK(NFPROTO_IPV4, NF_INET_LOCAL_OUT, net, sk, skb, NULL, rt->dst.dev, From d60cf616ec39141b3483304fde69acf0dcec80eb Mon Sep 17 00:00:00 2001 From: Masaru Nagai Date: Sun, 15 Nov 2015 21:34:42 +0900 Subject: [PATCH 174/335] ravb: remove unhandle int cause This driver does not handle the AVB-DMAC Receive FIFO Warning interrupt now, so the interrupt should not be enabled. Signed-off-by: Masaru Nagai Signed-off-by: Yoshihiro Kaneko Signed-off-by: David S. Miller --- drivers/net/ethernet/renesas/ravb_main.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index 7180e26f5c67..ee8d1ec61fab 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -408,8 +408,6 @@ static int ravb_dmac_init(struct net_device *ndev) /* Interrupt enable: */ /* Frame receive */ ravb_write(ndev, RIC0_FRE0 | RIC0_FRE1, RIC0); - /* Receive FIFO full warning */ - ravb_write(ndev, RIC1_RFWE, RIC1); /* Receive FIFO full error, descriptor empty */ ravb_write(ndev, RIC2_QFE0 | RIC2_QFE1 | RIC2_RFFE, RIC2); /* Frame transmitted, timestamp FIFO updated */ From ab6d7846cf80affc43b9d412fed5e25dfcf4f35d Mon Sep 17 00:00:00 2001 From: Yuval Mintz Date: Sun, 15 Nov 2015 15:02:16 +0200 Subject: [PATCH 175/335] bnx2x: Fix VLANs null-pointer for 57710, 57711 Commit 05cc5a39ddb7 "bnx2x: add vlan filtering offload" introduced a regression in regard for vlans for 57710, 57711 adapters - Loading 8021q module on a machine with such an adapter would cause a null pointer dereference, as the driver mistakenly publishes it has capabilities for vlan CTAG filtering. Reported-by: Otto Sabart Signed-off-by: Yuval Mintz Signed-off-by: Ariel Elior Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index f1d62d5dbaff..c9b036789184 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -13207,7 +13207,7 @@ static int bnx2x_init_dev(struct bnx2x *bp, struct pci_dev *pdev, /* VF with OLD Hypervisor or old PF do not support filtering */ if (IS_PF(bp)) { - if (CHIP_IS_E1x(bp)) + if (chip_is_e1x) bp->accept_any_vlan = true; else dev->hw_features |= NETIF_F_HW_VLAN_CTAG_FILTER; From 24cb7055a3066634a0f3fa0cd6a4780652905d35 Mon Sep 17 00:00:00 2001 From: Dragos Tatulea Date: Mon, 16 Nov 2015 10:52:48 +0100 Subject: [PATCH 176/335] net: switchdev: fix return code of fdb_dump stub rtnl_fdb_dump always expects an index to be returned by the ndo_fdb_dump op, but when CONFIG_NET_SWITCHDEV is off, it returns an error. Fix that by returning the given unmodified idx. A similar fix was 0890cf6cb6ab ("switchdev: fix return value of switchdev_port_fdb_dump in case of error") but for the CONFIG_NET_SWITCHDEV=y case. Fixes: 45d4122ca7cd ("switchdev: add support for fdb add/del/dump via switchdev_port_obj ops.") Signed-off-by: Dragos Tatulea Acked-by: Jiri Pirko Signed-off-by: David S. Miller --- include/net/switchdev.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/net/switchdev.h b/include/net/switchdev.h index bc865e244efe..1d22ce9f352e 100644 --- a/include/net/switchdev.h +++ b/include/net/switchdev.h @@ -323,7 +323,7 @@ static inline int switchdev_port_fdb_dump(struct sk_buff *skb, struct net_device *filter_dev, int idx) { - return -EOPNOTSUPP; + return idx; } static inline void switchdev_port_fwd_mark_set(struct net_device *dev, From 8844f97238ca6c1ca92a5d6c69f53efd361a266f Mon Sep 17 00:00:00 2001 From: Hannes Frederic Sowa Date: Mon, 16 Nov 2015 16:25:56 +0100 Subject: [PATCH 177/335] af_unix: don't append consumed skbs to sk_receive_queue In case multiple writes to a unix stream socket race we could end up in a situation where we pre-allocate a new skb for use in unix_stream_sendpage but have to free it again in the locked section because another skb has been appended meanwhile, which we must use. Accidentally we didn't clear the pointer after consuming it and so we touched freed memory while appending it to the sk_receive_queue. So, clear the pointer after consuming the skb. This bug has been found with syzkaller (http://github.com/google/syzkaller) by Dmitry Vyukov. Fixes: 869e7c62486e ("net: af_unix: implement stream sendpage support") Reported-by: Dmitry Vyukov Cc: Dmitry Vyukov Cc: Eric Dumazet Signed-off-by: Hannes Frederic Sowa Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/unix/af_unix.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 12b886f07982..a8352db5c5b5 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1800,6 +1800,7 @@ alloc_skb: * this - does no harm */ consume_skb(newskb); + newskb = NULL; } if (skb_append_pagefrags(skb, page, offset, size)) { From 88ad4175b201ae24be5e9b7752cf33c1306b64e4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Mon, 16 Nov 2015 19:16:40 +0100 Subject: [PATCH 178/335] net/core: use netdev name in warning if no parent MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A recent flaw in the netdev feature setting resulted in warnings like this one from VLAN interfaces: WARNING: CPU: 1 PID: 4975 at net/core/dev.c:2419 skb_warn_bad_offload+0xbc/0xcb() : caps=(0x00000000001b5820, 0x00000000001b5829) len=2782 data_len=0 gso_size=1348 gso_type=16 ip_summed=3 The ":" is supposed to be preceded by a driver name, but in this case it is an empty string since the device has no parent. There are many types of network devices without a parent. The anonymous warnings for these devices can be hard to debug. Log the network device name instead in these cases to assist further debugging. This is mostly similar to how __netdev_printk() handles orphan devices. Signed-off-by: Bjørn Mork Signed-off-by: David S. Miller --- net/core/dev.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 1974aee005a6..5dbc86ea6b58 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2403,17 +2403,20 @@ static void skb_warn_bad_offload(const struct sk_buff *skb) { static const netdev_features_t null_features = 0; struct net_device *dev = skb->dev; - const char *driver = ""; + const char *name = ""; if (!net_ratelimit()) return; - if (dev && dev->dev.parent) - driver = dev_driver_string(dev->dev.parent); - + if (dev) { + if (dev->dev.parent) + name = dev_driver_string(dev->dev.parent); + else + name = netdev_name(dev); + } WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d gso_size=%d " "gso_type=%d ip_summed=%d\n", - driver, dev ? &dev->features : &null_features, + name, dev ? &dev->features : &null_features, skb->sk ? &skb->sk->sk_route_caps : &null_features, skb->len, skb->data_len, skb_shinfo(skb)->gso_size, skb_shinfo(skb)->gso_type, skb->ip_summed); From 7750130d93decff06120df0d8ea024ff8a038a21 Mon Sep 17 00:00:00 2001 From: Pavel Fedin Date: Mon, 16 Nov 2015 17:51:34 +0300 Subject: [PATCH 179/335] net: thunder: Check for driver data in nicvf_remove() In some cases the crash is caused by nicvf_remove() being called from outside. For example, if we try to feed the device to vfio after the probe has failed for some reason. So, move the check to better place. Signed-off-by: Pavel Fedin Signed-off-by: David S. Miller --- drivers/net/ethernet/cavium/thunder/nicvf_main.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_main.c b/drivers/net/ethernet/cavium/thunder/nicvf_main.c index 372c39e5bcbd..7f709cbdcd87 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@ -1583,8 +1583,14 @@ err_disable_device: static void nicvf_remove(struct pci_dev *pdev) { struct net_device *netdev = pci_get_drvdata(pdev); - struct nicvf *nic = netdev_priv(netdev); - struct net_device *pnetdev = nic->pnicvf->netdev; + struct nicvf *nic; + struct net_device *pnetdev; + + if (!netdev) + return; + + nic = netdev_priv(netdev); + pnetdev = nic->pnicvf->netdev; /* Check if this Qset is assigned to different VF. * If yes, clean primary and all secondary Qsets. @@ -1600,9 +1606,6 @@ static void nicvf_remove(struct pci_dev *pdev) static void nicvf_shutdown(struct pci_dev *pdev) { - if (!pci_get_drvdata(pdev)) - return; - nicvf_remove(pdev); } From 41033f029e393a64e81966cbe34d66c6cf8a2e7e Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Mon, 16 Nov 2015 13:09:10 -0500 Subject: [PATCH 180/335] snmp: Remove duplicate OUTMCAST stat increment the OUTMCAST stat is double incremented, getting bumped once in the mcast code itself, and again in the common ip output path. Remove the mcast bump, as its not needed Validated by the reporter, with good results Signed-off-by: Neil Horman Reported-by: Claus Jensen CC: Claus Jensen CC: David Miller Signed-off-by: David S. Miller --- net/ipv6/mcast.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c index 124338a39e29..5ee56d0a8699 100644 --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1651,7 +1651,6 @@ out: if (!err) { ICMP6MSGOUT_INC_STATS(net, idev, ICMPV6_MLD2_REPORT); ICMP6_INC_STATS(net, idev, ICMP6_MIB_OUTMSGS); - IP6_UPD_PO_STATS(net, idev, IPSTATS_MIB_OUTMCAST, payload_len); } else { IP6_INC_STATS(net, idev, IPSTATS_MIB_OUTDISCARDS); } @@ -2015,7 +2014,6 @@ out: if (!err) { ICMP6MSGOUT_INC_STATS(net, idev, type); ICMP6_INC_STATS(net, idev, ICMP6_MIB_OUTMSGS); - IP6_UPD_PO_STATS(net, idev, IPSTATS_MIB_OUTMCAST, full_len); } else IP6_INC_STATS(net, idev, IPSTATS_MIB_OUTDISCARDS); From 680513cc0accf14cb447663b65e02ae307ac3811 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 10 Sep 2015 15:03:50 +0200 Subject: [PATCH 181/335] drm/amdgpu: wait interruptible when semaphores are disabled v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Otherwise debugging locked up processes isn't possible. v2: rebased Signed-off-by: Christian König Reviewed-by: Alex Deucher (v1) --- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index a6697fd05217..dd005c336c97 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c @@ -302,8 +302,14 @@ int amdgpu_sync_rings(struct amdgpu_sync *sync, return -EINVAL; } - if (amdgpu_enable_scheduler || !amdgpu_enable_semaphores || - (count >= AMDGPU_NUM_SYNCS)) { + if (amdgpu_enable_scheduler || !amdgpu_enable_semaphores) { + r = fence_wait(&fence->base, true); + if (r) + return r; + continue; + } + + if (count >= AMDGPU_NUM_SYNCS) { /* not enough room, wait manually */ r = fence_wait(&fence->base, false); if (r) From bbf0b34578a71f42d19bd6c966f09221044921fe Mon Sep 17 00:00:00 2001 From: Junwei Zhang Date: Sun, 6 Sep 2015 14:00:46 +0800 Subject: [PATCH 182/335] drm/amdgpu: remove the unnecessary parameter adev for amdgpu_sa_bo_new() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Junwei Zhang Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 7 +++---- drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c | 3 +-- drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c | 2 +- 4 files changed, 6 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 3902e7a9f8cc..9e25edafa721 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -62,7 +62,7 @@ int amdgpu_ib_get(struct amdgpu_ring *ring, struct amdgpu_vm *vm, int r; if (size) { - r = amdgpu_sa_bo_new(adev, &adev->ring_tmp_bo, + r = amdgpu_sa_bo_new(&adev->ring_tmp_bo, &ib->sa_bo, size, 256); if (r) { dev_err(adev->dev, "failed to get a new IB (%d)\n", r); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 3c2ff4567798..ea756e77b023 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -189,10 +189,9 @@ int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager); int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager); -int amdgpu_sa_bo_new(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager, - struct amdgpu_sa_bo **sa_bo, - unsigned size, unsigned align); +int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, + struct amdgpu_sa_bo **sa_bo, + unsigned size, unsigned align); void amdgpu_sa_bo_free(struct amdgpu_device *adev, struct amdgpu_sa_bo **sa_bo, struct fence *fence); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c index 0212b31dc194..8b88edb0434b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c @@ -311,8 +311,7 @@ static bool amdgpu_sa_bo_next_hole(struct amdgpu_sa_manager *sa_manager, return false; } -int amdgpu_sa_bo_new(struct amdgpu_device *adev, - struct amdgpu_sa_manager *sa_manager, +int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, struct amdgpu_sa_bo **sa_bo, unsigned size, unsigned align) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c index ff3ca52ec6fe..1caaf201b708 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c @@ -40,7 +40,7 @@ int amdgpu_semaphore_create(struct amdgpu_device *adev, if (*semaphore == NULL) { return -ENOMEM; } - r = amdgpu_sa_bo_new(adev, &adev->ring_tmp_bo, + r = amdgpu_sa_bo_new(&adev->ring_tmp_bo, &(*semaphore)->sa_bo, 8, 8); if (r) { kfree(*semaphore); From 6f9b36cd2497660e254c67ed131f0b297ed8bf40 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Thu, 12 Nov 2015 10:26:57 -0800 Subject: [PATCH 183/335] PM / wakeirq: check that wake IRQ is valid before accepting it Check that IRQ number passed to dev_pm_set_wake_irq() and dev_pm_set_dedicated_wake_irq() is valid (not negative) before accepting it. Signed-off-by: Dmitry Torokhov Signed-off-by: Rafael J. Wysocki --- drivers/base/power/wakeirq.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/base/power/wakeirq.c b/drivers/base/power/wakeirq.c index eb6e67451dec..0d77cd6fd8d1 100644 --- a/drivers/base/power/wakeirq.c +++ b/drivers/base/power/wakeirq.c @@ -68,6 +68,9 @@ int dev_pm_set_wake_irq(struct device *dev, int irq) struct wake_irq *wirq; int err; + if (irq < 0) + return -EINVAL; + wirq = kzalloc(sizeof(*wirq), GFP_KERNEL); if (!wirq) return -ENOMEM; @@ -167,6 +170,9 @@ int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq) struct wake_irq *wirq; int err; + if (irq < 0) + return -EINVAL; + wirq = kzalloc(sizeof(*wirq), GFP_KERNEL); if (!wirq) return -ENOMEM; From c3f45d322cbd379c46466cc2ecab7e2d719b22ed Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Sun, 15 Nov 2015 22:36:11 +0100 Subject: [PATCH 184/335] dl2k: Add support for IP1000A-based cards Add support for IP1000A chips to dl2k driver. IP1000A chip looks like a TC9020 with integrated PHY. This allows IP1000A chips to work reliably because the ipg driver is buggy - it loses packets under load and then completely stops transmitting data. Tested with Asus NX1101 v2.0 at 10, 100 and 1000Mbps: vendor=0x13f0 device=0x1023 (rev 0x41) subsystem vendor=0x1043 device=0x8180 MAC address registers access needed to be changed from 8-bit to 16-bit because 8-bit does not work on IP1000A. 8-bit access is not even allowed in the TC9020 datasheet (although it worked). 16-bit access works on both. Tested that it does not break D-Link DGE-550T (DL-2000 chip, probably a rebranded TC9020): vendor=0x1186 device=0x4000 (rev 0x0c) subsystem vendor=0x1186 device=0x4000 Signed-off-by: Ondrej Zary Reviewed-by: Andy Shevchenko Signed-off-by: David S. Miller --- drivers/net/ethernet/dlink/Kconfig | 5 +-- drivers/net/ethernet/dlink/dl2k.c | 55 ++++++++++++++++++++++++++++-- drivers/net/ethernet/dlink/dl2k.h | 15 +++++++- 3 files changed, 69 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/dlink/Kconfig b/drivers/net/ethernet/dlink/Kconfig index f6e858d0b9d4..ebdc83247bb6 100644 --- a/drivers/net/ethernet/dlink/Kconfig +++ b/drivers/net/ethernet/dlink/Kconfig @@ -17,15 +17,16 @@ config NET_VENDOR_DLINK if NET_VENDOR_DLINK config DL2K - tristate "DL2000/TC902x-based Gigabit Ethernet support" + tristate "DL2000/TC902x/IP1000A-based Gigabit Ethernet support" depends on PCI select CRC32 ---help--- - This driver supports DL2000/TC902x-based Gigabit ethernet cards, + This driver supports DL2000/TC902x/IP1000A-based Gigabit ethernet cards, which includes D-Link DGE-550T Gigabit Ethernet Adapter. D-Link DL2000-based Gigabit Ethernet Adapter. Sundance/Tamarack TC902x Gigabit Ethernet Adapter. + ICPlus IP1000A-based cards To compile this driver as a module, choose M here: the module will be called dl2k. diff --git a/drivers/net/ethernet/dlink/dl2k.c b/drivers/net/ethernet/dlink/dl2k.c index cf0a5fcdaaaf..ccca4799c27b 100644 --- a/drivers/net/ethernet/dlink/dl2k.c +++ b/drivers/net/ethernet/dlink/dl2k.c @@ -253,6 +253,19 @@ rio_probe1 (struct pci_dev *pdev, const struct pci_device_id *ent) if (err) goto err_out_unmap_rx; + if (np->chip_id == CHIP_IP1000A && + (np->pdev->revision == 0x40 || np->pdev->revision == 0x41)) { + /* PHY magic taken from ipg driver, undocumented registers */ + mii_write(dev, np->phy_addr, 31, 0x0001); + mii_write(dev, np->phy_addr, 27, 0x01e0); + mii_write(dev, np->phy_addr, 31, 0x0002); + mii_write(dev, np->phy_addr, 27, 0xeb8e); + mii_write(dev, np->phy_addr, 31, 0x0000); + mii_write(dev, np->phy_addr, 30, 0x005e); + /* advertise 1000BASE-T half & full duplex, prefer MASTER */ + mii_write(dev, np->phy_addr, MII_CTRL1000, 0x0700); + } + /* Fiber device? */ np->phy_media = (dr16(ASICCtrl) & PhyMedia) ? 1 : 0; np->link_status = 0; @@ -361,6 +374,11 @@ parse_eeprom (struct net_device *dev) for (i = 0; i < 6; i++) dev->dev_addr[i] = psrom->mac_addr[i]; + if (np->chip_id == CHIP_IP1000A) { + np->led_mode = psrom->led_mode; + return 0; + } + if (np->pdev->vendor != PCI_VENDOR_ID_DLINK) { return 0; } @@ -406,6 +424,28 @@ parse_eeprom (struct net_device *dev) return 0; } +static void rio_set_led_mode(struct net_device *dev) +{ + struct netdev_private *np = netdev_priv(dev); + void __iomem *ioaddr = np->ioaddr; + u32 mode; + + if (np->chip_id != CHIP_IP1000A) + return; + + mode = dr32(ASICCtrl); + mode &= ~(IPG_AC_LED_MODE_BIT_1 | IPG_AC_LED_MODE | IPG_AC_LED_SPEED); + + if (np->led_mode & 0x01) + mode |= IPG_AC_LED_MODE; + if (np->led_mode & 0x02) + mode |= IPG_AC_LED_MODE_BIT_1; + if (np->led_mode & 0x08) + mode |= IPG_AC_LED_SPEED; + + dw32(ASICCtrl, mode); +} + static int rio_open (struct net_device *dev) { @@ -424,6 +464,8 @@ rio_open (struct net_device *dev) GlobalReset | DMAReset | FIFOReset | NetworkReset | HostReset); mdelay(10); + rio_set_led_mode(dev); + /* DebugCtrl bit 4, 5, 9 must set */ dw32(DebugCtrl, dr32(DebugCtrl) | 0x0230); @@ -433,9 +475,13 @@ rio_open (struct net_device *dev) alloc_list (dev); - /* Get station address */ - for (i = 0; i < 6; i++) - dw8(StationAddr0 + i, dev->dev_addr[i]); + /* Set station address */ + /* 16 or 32-bit access is required by TC9020 datasheet but 8-bit works + * too. However, it doesn't work on IP1000A so we use 16-bit access. + */ + for (i = 0; i < 3; i++) + dw16(StationAddr0 + 2 * i, + cpu_to_le16(((u16 *)dev->dev_addr)[i])); set_multicast (dev); if (np->coalesce) { @@ -780,6 +826,7 @@ tx_error (struct net_device *dev, int tx_status) break; mdelay (1); } + rio_set_led_mode(dev); rio_free_tx (dev, 1); /* Reset TFDListPtr */ dw32(TFDListPtr0, np->tx_ring_dma + @@ -799,6 +846,7 @@ tx_error (struct net_device *dev, int tx_status) break; mdelay (1); } + rio_set_led_mode(dev); /* Let TxStartThresh stay default value */ } /* Maximum Collisions */ @@ -965,6 +1013,7 @@ rio_error (struct net_device *dev, int int_status) dev->name, int_status); dw16(ASICCtrl + 2, GlobalReset | HostReset); mdelay (500); + rio_set_led_mode(dev); } } diff --git a/drivers/net/ethernet/dlink/dl2k.h b/drivers/net/ethernet/dlink/dl2k.h index 23c07b007069..8f4f61262d5c 100644 --- a/drivers/net/ethernet/dlink/dl2k.h +++ b/drivers/net/ethernet/dlink/dl2k.h @@ -211,6 +211,10 @@ enum ASICCtrl_HiWord_bits { ResetBusy = 0x0400, }; +#define IPG_AC_LED_MODE BIT(14) +#define IPG_AC_LED_SPEED BIT(27) +#define IPG_AC_LED_MODE_BIT_1 BIT(29) + /* Transmit Frame Control bits */ enum TFC_bits { DwordAlign = 0x00000000, @@ -332,7 +336,10 @@ typedef struct t_SROM { u16 asic_ctrl; /* 0x02 */ u16 sub_vendor_id; /* 0x04 */ u16 sub_system_id; /* 0x06 */ - u16 reserved1[12]; /* 0x08-0x1f */ + u16 pci_base_1; /* 0x08 (IP1000A only) */ + u16 pci_base_2; /* 0x0a (IP1000A only) */ + u16 led_mode; /* 0x0c (IP1000A only) */ + u16 reserved1[9]; /* 0x0e-0x1f */ u8 mac_addr[6]; /* 0x20-0x25 */ u8 reserved2[10]; /* 0x26-0x2f */ u8 sib[204]; /* 0x30-0xfb */ @@ -397,6 +404,7 @@ struct netdev_private { u16 advertising; /* NWay media advertisement */ u16 negotiate; /* Negotiated media */ int phy_addr; /* PHY addresses. */ + u16 led_mode; /* LED mode read from EEPROM (IP1000A only) */ }; /* The station address location in the EEPROM. */ @@ -407,10 +415,15 @@ struct netdev_private { class_mask of the class are honored during the comparison. driver_data Data private to the driver. */ +#define CHIP_IP1000A 1 static const struct pci_device_id rio_pci_tbl[] = { {0x1186, 0x4000, PCI_ANY_ID, PCI_ANY_ID, }, {0x13f0, 0x1021, PCI_ANY_ID, PCI_ANY_ID, }, + { PCI_VDEVICE(SUNDANCE, 0x1023), CHIP_IP1000A }, + { PCI_VDEVICE(SUNDANCE, 0x2021), CHIP_IP1000A }, + { PCI_VDEVICE(DLINK, 0x9021), CHIP_IP1000A }, + { PCI_VDEVICE(DLINK, 0x4020), CHIP_IP1000A }, { } }; MODULE_DEVICE_TABLE (pci, rio_pci_tbl); From f1a454a37618b819f2528ccd234f77a02b3a6016 Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Sun, 15 Nov 2015 22:36:12 +0100 Subject: [PATCH 185/335] ipg: Remove ipg driver Now that IP1000A chips are supported by dl2k driver, the buggy ipg driver can be removed. Signed-off-by: Ondrej Zary Signed-off-by: David S. Miller --- MAINTAINERS | 7 - drivers/net/ethernet/Kconfig | 1 - drivers/net/ethernet/Makefile | 1 - drivers/net/ethernet/icplus/Kconfig | 13 - drivers/net/ethernet/icplus/Makefile | 5 - drivers/net/ethernet/icplus/ipg.c | 2300 -------------------------- drivers/net/ethernet/icplus/ipg.h | 748 --------- 7 files changed, 3075 deletions(-) delete mode 100644 drivers/net/ethernet/icplus/Kconfig delete mode 100644 drivers/net/ethernet/icplus/Makefile delete mode 100644 drivers/net/ethernet/icplus/ipg.c delete mode 100644 drivers/net/ethernet/icplus/ipg.h diff --git a/MAINTAINERS b/MAINTAINERS index 35fe7ae0492e..45320675a460 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5709,13 +5709,6 @@ M: Juanjo Ciarlante S: Maintained F: net/ipv4/netfilter/ipt_MASQUERADE.c -IP1000A 10/100/1000 GIGABIT ETHERNET DRIVER -M: Francois Romieu -M: Sorbica Shieh -L: netdev@vger.kernel.org -S: Maintained -F: drivers/net/ethernet/icplus/ipg.* - IPATH DRIVER M: Mike Marciniszyn L: linux-rdma@vger.kernel.org diff --git a/drivers/net/ethernet/Kconfig b/drivers/net/ethernet/Kconfig index 05aa7597dab9..955d06b9cdba 100644 --- a/drivers/net/ethernet/Kconfig +++ b/drivers/net/ethernet/Kconfig @@ -78,7 +78,6 @@ source "drivers/net/ethernet/ibm/Kconfig" source "drivers/net/ethernet/intel/Kconfig" source "drivers/net/ethernet/i825xx/Kconfig" source "drivers/net/ethernet/xscale/Kconfig" -source "drivers/net/ethernet/icplus/Kconfig" config JME tristate "JMicron(R) PCI-Express Gigabit Ethernet support" diff --git a/drivers/net/ethernet/Makefile b/drivers/net/ethernet/Makefile index ddfc808110a1..4a2ee98738f0 100644 --- a/drivers/net/ethernet/Makefile +++ b/drivers/net/ethernet/Makefile @@ -41,7 +41,6 @@ obj-$(CONFIG_NET_VENDOR_IBM) += ibm/ obj-$(CONFIG_NET_VENDOR_INTEL) += intel/ obj-$(CONFIG_NET_VENDOR_I825XX) += i825xx/ obj-$(CONFIG_NET_VENDOR_XSCALE) += xscale/ -obj-$(CONFIG_IP1000) += icplus/ obj-$(CONFIG_JME) += jme.o obj-$(CONFIG_KORINA) += korina.o obj-$(CONFIG_LANTIQ_ETOP) += lantiq_etop.o diff --git a/drivers/net/ethernet/icplus/Kconfig b/drivers/net/ethernet/icplus/Kconfig deleted file mode 100644 index 14a66e9d2e26..000000000000 --- a/drivers/net/ethernet/icplus/Kconfig +++ /dev/null @@ -1,13 +0,0 @@ -# -# IC Plus device configuration -# - -config IP1000 - tristate "IP1000 Gigabit Ethernet support" - depends on PCI - select MII - ---help--- - This driver supports IP1000 gigabit Ethernet cards. - - To compile this driver as a module, choose M here: the module - will be called ipg. This is recommended. diff --git a/drivers/net/ethernet/icplus/Makefile b/drivers/net/ethernet/icplus/Makefile deleted file mode 100644 index 5bc87c1f36aa..000000000000 --- a/drivers/net/ethernet/icplus/Makefile +++ /dev/null @@ -1,5 +0,0 @@ -# -# Makefile for the IC Plus device drivers -# - -obj-$(CONFIG_IP1000) += ipg.o diff --git a/drivers/net/ethernet/icplus/ipg.c b/drivers/net/ethernet/icplus/ipg.c deleted file mode 100644 index c3b6af83f070..000000000000 --- a/drivers/net/ethernet/icplus/ipg.c +++ /dev/null @@ -1,2300 +0,0 @@ -/* - * ipg.c: Device Driver for the IP1000 Gigabit Ethernet Adapter - * - * Copyright (C) 2003, 2007 IC Plus Corp - * - * Original Author: - * - * Craig Rich - * Sundance Technology, Inc. - * www.sundanceti.com - * craig_rich@sundanceti.com - * - * Current Maintainer: - * - * Sorbica Shieh. - * http://www.icplus.com.tw - * sorbica@icplus.com.tw - * - * Jesse Huang - * http://www.icplus.com.tw - * jesse@icplus.com.tw - */ - -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include -#include -#include -#include -#include -#include - -#include - -#define IPG_RX_RING_BYTES (sizeof(struct ipg_rx) * IPG_RFDLIST_LENGTH) -#define IPG_TX_RING_BYTES (sizeof(struct ipg_tx) * IPG_TFDLIST_LENGTH) -#define IPG_RESET_MASK \ - (IPG_AC_GLOBAL_RESET | IPG_AC_RX_RESET | IPG_AC_TX_RESET | \ - IPG_AC_DMA | IPG_AC_FIFO | IPG_AC_NETWORK | IPG_AC_HOST | \ - IPG_AC_AUTO_INIT) - -#define ipg_w32(val32, reg) iowrite32((val32), ioaddr + (reg)) -#define ipg_w16(val16, reg) iowrite16((val16), ioaddr + (reg)) -#define ipg_w8(val8, reg) iowrite8((val8), ioaddr + (reg)) - -#define ipg_r32(reg) ioread32(ioaddr + (reg)) -#define ipg_r16(reg) ioread16(ioaddr + (reg)) -#define ipg_r8(reg) ioread8(ioaddr + (reg)) - -enum { - netdev_io_size = 128 -}; - -#include "ipg.h" -#define DRV_NAME "ipg" - -MODULE_AUTHOR("IC Plus Corp. 2003"); -MODULE_DESCRIPTION("IC Plus IP1000 Gigabit Ethernet Adapter Linux Driver"); -MODULE_LICENSE("GPL"); - -/* - * Defaults - */ -#define IPG_MAX_RXFRAME_SIZE 0x0600 -#define IPG_RXFRAG_SIZE 0x0600 -#define IPG_RXSUPPORT_SIZE 0x0600 -#define IPG_IS_JUMBO false - -/* - * Variable record -- index by leading revision/length - * Revision/Length(=N*4), Address1, Data1, Address2, Data2,...,AddressN,DataN - */ -static const unsigned short DefaultPhyParam[] = { - /* 11/12/03 IP1000A v1-3 rev=0x40 */ - /*-------------------------------------------------------------------------- - (0x4000|(15*4)), 31, 0x0001, 27, 0x01e0, 31, 0x0002, 22, 0x85bd, 24, 0xfff2, - 27, 0x0c10, 28, 0x0c10, 29, 0x2c10, 31, 0x0003, 23, 0x92f6, - 31, 0x0000, 23, 0x003d, 30, 0x00de, 20, 0x20e7, 9, 0x0700, - --------------------------------------------------------------------------*/ - /* 12/17/03 IP1000A v1-4 rev=0x40 */ - (0x4000 | (07 * 4)), 31, 0x0001, 27, 0x01e0, 31, 0x0002, 27, 0xeb8e, 31, - 0x0000, - 30, 0x005e, 9, 0x0700, - /* 01/09/04 IP1000A v1-5 rev=0x41 */ - (0x4100 | (07 * 4)), 31, 0x0001, 27, 0x01e0, 31, 0x0002, 27, 0xeb8e, 31, - 0x0000, - 30, 0x005e, 9, 0x0700, - 0x0000 -}; - -static const char * const ipg_brand_name[] = { - "IC PLUS IP1000 1000/100/10 based NIC", - "Sundance Technology ST2021 based NIC", - "Tamarack Microelectronics TC9020/9021 based NIC", - "D-Link NIC IP1000A" -}; - -static const struct pci_device_id ipg_pci_tbl[] = { - { PCI_VDEVICE(SUNDANCE, 0x1023), 0 }, - { PCI_VDEVICE(SUNDANCE, 0x2021), 1 }, - { PCI_VDEVICE(DLINK, 0x9021), 2 }, - { PCI_VDEVICE(DLINK, 0x4020), 3 }, - { 0, } -}; - -MODULE_DEVICE_TABLE(pci, ipg_pci_tbl); - -static inline void __iomem *ipg_ioaddr(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - return sp->ioaddr; -} - -#ifdef IPG_DEBUG -static void ipg_dump_rfdlist(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - u32 offset; - - IPG_DEBUG_MSG("_dump_rfdlist\n"); - - netdev_info(dev, "rx_current = %02x\n", sp->rx_current); - netdev_info(dev, "rx_dirty = %02x\n", sp->rx_dirty); - netdev_info(dev, "RFDList start address = %016lx\n", - (unsigned long)sp->rxd_map); - netdev_info(dev, "RFDListPtr register = %08x%08x\n", - ipg_r32(IPG_RFDLISTPTR1), ipg_r32(IPG_RFDLISTPTR0)); - - for (i = 0; i < IPG_RFDLIST_LENGTH; i++) { - offset = (u32) &sp->rxd[i].next_desc - (u32) sp->rxd; - netdev_info(dev, "%02x %04x RFDNextPtr = %016lx\n", - i, offset, (unsigned long)sp->rxd[i].next_desc); - offset = (u32) &sp->rxd[i].rfs - (u32) sp->rxd; - netdev_info(dev, "%02x %04x RFS = %016lx\n", - i, offset, (unsigned long)sp->rxd[i].rfs); - offset = (u32) &sp->rxd[i].frag_info - (u32) sp->rxd; - netdev_info(dev, "%02x %04x frag_info = %016lx\n", - i, offset, (unsigned long)sp->rxd[i].frag_info); - } -} - -static void ipg_dump_tfdlist(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - u32 offset; - - IPG_DEBUG_MSG("_dump_tfdlist\n"); - - netdev_info(dev, "tx_current = %02x\n", sp->tx_current); - netdev_info(dev, "tx_dirty = %02x\n", sp->tx_dirty); - netdev_info(dev, "TFDList start address = %016lx\n", - (unsigned long) sp->txd_map); - netdev_info(dev, "TFDListPtr register = %08x%08x\n", - ipg_r32(IPG_TFDLISTPTR1), ipg_r32(IPG_TFDLISTPTR0)); - - for (i = 0; i < IPG_TFDLIST_LENGTH; i++) { - offset = (u32) &sp->txd[i].next_desc - (u32) sp->txd; - netdev_info(dev, "%02x %04x TFDNextPtr = %016lx\n", - i, offset, (unsigned long)sp->txd[i].next_desc); - - offset = (u32) &sp->txd[i].tfc - (u32) sp->txd; - netdev_info(dev, "%02x %04x TFC = %016lx\n", - i, offset, (unsigned long) sp->txd[i].tfc); - offset = (u32) &sp->txd[i].frag_info - (u32) sp->txd; - netdev_info(dev, "%02x %04x frag_info = %016lx\n", - i, offset, (unsigned long) sp->txd[i].frag_info); - } -} -#endif - -static void ipg_write_phy_ctl(void __iomem *ioaddr, u8 data) -{ - ipg_w8(IPG_PC_RSVD_MASK & data, PHY_CTRL); - ndelay(IPG_PC_PHYCTRLWAIT_NS); -} - -static void ipg_drive_phy_ctl_low_high(void __iomem *ioaddr, u8 data) -{ - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_LO | data); - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_HI | data); -} - -static void send_three_state(void __iomem *ioaddr, u8 phyctrlpolarity) -{ - phyctrlpolarity |= (IPG_PC_MGMTDATA & 0) | IPG_PC_MGMTDIR; - - ipg_drive_phy_ctl_low_high(ioaddr, phyctrlpolarity); -} - -static void send_end(void __iomem *ioaddr, u8 phyctrlpolarity) -{ - ipg_w8((IPG_PC_MGMTCLK_LO | (IPG_PC_MGMTDATA & 0) | IPG_PC_MGMTDIR | - phyctrlpolarity) & IPG_PC_RSVD_MASK, PHY_CTRL); -} - -static u16 read_phy_bit(void __iomem *ioaddr, u8 phyctrlpolarity) -{ - u16 bit_data; - - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_LO | phyctrlpolarity); - - bit_data = ((ipg_r8(PHY_CTRL) & IPG_PC_MGMTDATA) >> 1) & 1; - - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_HI | phyctrlpolarity); - - return bit_data; -} - -/* - * Read a register from the Physical Layer device located - * on the IPG NIC, using the IPG PHYCTRL register. - */ -static int mdio_read(struct net_device *dev, int phy_id, int phy_reg) -{ - void __iomem *ioaddr = ipg_ioaddr(dev); - /* - * The GMII mangement frame structure for a read is as follows: - * - * |Preamble|st|op|phyad|regad|ta| data |idle| - * |< 32 1s>|01|10|AAAAA|RRRRR|z0|DDDDDDDDDDDDDDDD|z | - * - * <32 1s> = 32 consecutive logic 1 values - * A = bit of Physical Layer device address (MSB first) - * R = bit of register address (MSB first) - * z = High impedance state - * D = bit of read data (MSB first) - * - * Transmission order is 'Preamble' field first, bits transmitted - * left to right (first to last). - */ - struct { - u32 field; - unsigned int len; - } p[] = { - { GMII_PREAMBLE, 32 }, /* Preamble */ - { GMII_ST, 2 }, /* ST */ - { GMII_READ, 2 }, /* OP */ - { phy_id, 5 }, /* PHYAD */ - { phy_reg, 5 }, /* REGAD */ - { 0x0000, 2 }, /* TA */ - { 0x0000, 16 }, /* DATA */ - { 0x0000, 1 } /* IDLE */ - }; - unsigned int i, j; - u8 polarity, data; - - polarity = ipg_r8(PHY_CTRL); - polarity &= (IPG_PC_DUPLEX_POLARITY | IPG_PC_LINK_POLARITY); - - /* Create the Preamble, ST, OP, PHYAD, and REGAD field. */ - for (j = 0; j < 5; j++) { - for (i = 0; i < p[j].len; i++) { - /* For each variable length field, the MSB must be - * transmitted first. Rotate through the field bits, - * starting with the MSB, and move each bit into the - * the 1st (2^1) bit position (this is the bit position - * corresponding to the MgmtData bit of the PhyCtrl - * register for the IPG). - * - * Example: ST = 01; - * - * First write a '0' to bit 1 of the PhyCtrl - * register, then write a '1' to bit 1 of the - * PhyCtrl register. - * - * To do this, right shift the MSB of ST by the value: - * [field length - 1 - #ST bits already written] - * then left shift this result by 1. - */ - data = (p[j].field >> (p[j].len - 1 - i)) << 1; - data &= IPG_PC_MGMTDATA; - data |= polarity | IPG_PC_MGMTDIR; - - ipg_drive_phy_ctl_low_high(ioaddr, data); - } - } - - send_three_state(ioaddr, polarity); - - read_phy_bit(ioaddr, polarity); - - /* - * For a read cycle, the bits for the next two fields (TA and - * DATA) are driven by the PHY (the IPG reads these bits). - */ - for (i = 0; i < p[6].len; i++) { - p[6].field |= - (read_phy_bit(ioaddr, polarity) << (p[6].len - 1 - i)); - } - - send_three_state(ioaddr, polarity); - send_three_state(ioaddr, polarity); - send_three_state(ioaddr, polarity); - send_end(ioaddr, polarity); - - /* Return the value of the DATA field. */ - return p[6].field; -} - -/* - * Write to a register from the Physical Layer device located - * on the IPG NIC, using the IPG PHYCTRL register. - */ -static void mdio_write(struct net_device *dev, int phy_id, int phy_reg, int val) -{ - void __iomem *ioaddr = ipg_ioaddr(dev); - /* - * The GMII mangement frame structure for a read is as follows: - * - * |Preamble|st|op|phyad|regad|ta| data |idle| - * |< 32 1s>|01|10|AAAAA|RRRRR|z0|DDDDDDDDDDDDDDDD|z | - * - * <32 1s> = 32 consecutive logic 1 values - * A = bit of Physical Layer device address (MSB first) - * R = bit of register address (MSB first) - * z = High impedance state - * D = bit of write data (MSB first) - * - * Transmission order is 'Preamble' field first, bits transmitted - * left to right (first to last). - */ - struct { - u32 field; - unsigned int len; - } p[] = { - { GMII_PREAMBLE, 32 }, /* Preamble */ - { GMII_ST, 2 }, /* ST */ - { GMII_WRITE, 2 }, /* OP */ - { phy_id, 5 }, /* PHYAD */ - { phy_reg, 5 }, /* REGAD */ - { 0x0002, 2 }, /* TA */ - { val & 0xffff, 16 }, /* DATA */ - { 0x0000, 1 } /* IDLE */ - }; - unsigned int i, j; - u8 polarity, data; - - polarity = ipg_r8(PHY_CTRL); - polarity &= (IPG_PC_DUPLEX_POLARITY | IPG_PC_LINK_POLARITY); - - /* Create the Preamble, ST, OP, PHYAD, and REGAD field. */ - for (j = 0; j < 7; j++) { - for (i = 0; i < p[j].len; i++) { - /* For each variable length field, the MSB must be - * transmitted first. Rotate through the field bits, - * starting with the MSB, and move each bit into the - * the 1st (2^1) bit position (this is the bit position - * corresponding to the MgmtData bit of the PhyCtrl - * register for the IPG). - * - * Example: ST = 01; - * - * First write a '0' to bit 1 of the PhyCtrl - * register, then write a '1' to bit 1 of the - * PhyCtrl register. - * - * To do this, right shift the MSB of ST by the value: - * [field length - 1 - #ST bits already written] - * then left shift this result by 1. - */ - data = (p[j].field >> (p[j].len - 1 - i)) << 1; - data &= IPG_PC_MGMTDATA; - data |= polarity | IPG_PC_MGMTDIR; - - ipg_drive_phy_ctl_low_high(ioaddr, data); - } - } - - /* The last cycle is a tri-state, so read from the PHY. */ - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_LO | polarity); - ipg_r8(PHY_CTRL); - ipg_write_phy_ctl(ioaddr, IPG_PC_MGMTCLK_HI | polarity); -} - -static void ipg_set_led_mode(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - u32 mode; - - mode = ipg_r32(ASIC_CTRL); - mode &= ~(IPG_AC_LED_MODE_BIT_1 | IPG_AC_LED_MODE | IPG_AC_LED_SPEED); - - if ((sp->led_mode & 0x03) > 1) - mode |= IPG_AC_LED_MODE_BIT_1; /* Write Asic Control Bit 29 */ - - if ((sp->led_mode & 0x01) == 1) - mode |= IPG_AC_LED_MODE; /* Write Asic Control Bit 14 */ - - if ((sp->led_mode & 0x08) == 8) - mode |= IPG_AC_LED_SPEED; /* Write Asic Control Bit 27 */ - - ipg_w32(mode, ASIC_CTRL); -} - -static void ipg_set_phy_set(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - int physet; - - physet = ipg_r8(PHY_SET); - physet &= ~(IPG_PS_MEM_LENB9B | IPG_PS_MEM_LEN9 | IPG_PS_NON_COMPDET); - physet |= ((sp->led_mode & 0x70) >> 4); - ipg_w8(physet, PHY_SET); -} - -static int ipg_reset(struct net_device *dev, u32 resetflags) -{ - /* Assert functional resets via the IPG AsicCtrl - * register as specified by the 'resetflags' input - * parameter. - */ - void __iomem *ioaddr = ipg_ioaddr(dev); - unsigned int timeout_count = 0; - - IPG_DEBUG_MSG("_reset\n"); - - ipg_w32(ipg_r32(ASIC_CTRL) | resetflags, ASIC_CTRL); - - /* Delay added to account for problem with 10Mbps reset. */ - mdelay(IPG_AC_RESETWAIT); - - while (IPG_AC_RESET_BUSY & ipg_r32(ASIC_CTRL)) { - mdelay(IPG_AC_RESETWAIT); - if (++timeout_count > IPG_AC_RESET_TIMEOUT) - return -ETIME; - } - /* Set LED Mode in Asic Control */ - ipg_set_led_mode(dev); - - /* Set PHYSet Register Value */ - ipg_set_phy_set(dev); - return 0; -} - -/* Find the GMII PHY address. */ -static int ipg_find_phyaddr(struct net_device *dev) -{ - unsigned int phyaddr, i; - - for (i = 0; i < 32; i++) { - u32 status; - - /* Search for the correct PHY address among 32 possible. */ - phyaddr = (IPG_NIC_PHY_ADDRESS + i) % 32; - - /* 10/22/03 Grace change verify from GMII_PHY_STATUS to - GMII_PHY_ID1 - */ - - status = mdio_read(dev, phyaddr, MII_BMSR); - - if ((status != 0xFFFF) && (status != 0)) - return phyaddr; - } - - return 0x1f; -} - -/* - * Configure IPG based on result of IEEE 802.3 PHY - * auto-negotiation. - */ -static int ipg_config_autoneg(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int txflowcontrol; - unsigned int rxflowcontrol; - unsigned int fullduplex; - u32 mac_ctrl_val; - u32 asicctrl; - u8 phyctrl; - const char *speed; - const char *duplex; - const char *tx_desc; - const char *rx_desc; - - IPG_DEBUG_MSG("_config_autoneg\n"); - - asicctrl = ipg_r32(ASIC_CTRL); - phyctrl = ipg_r8(PHY_CTRL); - mac_ctrl_val = ipg_r32(MAC_CTRL); - - /* Set flags for use in resolving auto-negotiation, assuming - * non-1000Mbps, half duplex, no flow control. - */ - fullduplex = 0; - txflowcontrol = 0; - rxflowcontrol = 0; - - /* To accommodate a problem in 10Mbps operation, - * set a global flag if PHY running in 10Mbps mode. - */ - sp->tenmbpsmode = 0; - - /* Determine actual speed of operation. */ - switch (phyctrl & IPG_PC_LINK_SPEED) { - case IPG_PC_LINK_SPEED_10MBPS: - speed = "10Mbps"; - sp->tenmbpsmode = 1; - break; - case IPG_PC_LINK_SPEED_100MBPS: - speed = "100Mbps"; - break; - case IPG_PC_LINK_SPEED_1000MBPS: - speed = "1000Mbps"; - break; - default: - speed = "undefined!"; - return 0; - } - - netdev_info(dev, "Link speed = %s\n", speed); - if (sp->tenmbpsmode == 1) - netdev_info(dev, "10Mbps operational mode enabled\n"); - - if (phyctrl & IPG_PC_DUPLEX_STATUS) { - fullduplex = 1; - txflowcontrol = 1; - rxflowcontrol = 1; - } - - /* Configure full duplex, and flow control. */ - if (fullduplex == 1) { - - /* Configure IPG for full duplex operation. */ - - duplex = "full"; - - mac_ctrl_val |= IPG_MC_DUPLEX_SELECT_FD; - - if (txflowcontrol == 1) { - tx_desc = ""; - mac_ctrl_val |= IPG_MC_TX_FLOW_CONTROL_ENABLE; - } else { - tx_desc = "no "; - mac_ctrl_val &= ~IPG_MC_TX_FLOW_CONTROL_ENABLE; - } - - if (rxflowcontrol == 1) { - rx_desc = ""; - mac_ctrl_val |= IPG_MC_RX_FLOW_CONTROL_ENABLE; - } else { - rx_desc = "no "; - mac_ctrl_val &= ~IPG_MC_RX_FLOW_CONTROL_ENABLE; - } - } else { - duplex = "half"; - tx_desc = "no "; - rx_desc = "no "; - mac_ctrl_val &= (~IPG_MC_DUPLEX_SELECT_FD & - ~IPG_MC_TX_FLOW_CONTROL_ENABLE & - ~IPG_MC_RX_FLOW_CONTROL_ENABLE); - } - - netdev_info(dev, "setting %s duplex, %sTX, %sRX flow control\n", - duplex, tx_desc, rx_desc); - ipg_w32(mac_ctrl_val, MAC_CTRL); - - return 0; -} - -/* Determine and configure multicast operation and set - * receive mode for IPG. - */ -static void ipg_nic_set_multicast_list(struct net_device *dev) -{ - void __iomem *ioaddr = ipg_ioaddr(dev); - struct netdev_hw_addr *ha; - unsigned int hashindex; - u32 hashtable[2]; - u8 receivemode; - - IPG_DEBUG_MSG("_nic_set_multicast_list\n"); - - receivemode = IPG_RM_RECEIVEUNICAST | IPG_RM_RECEIVEBROADCAST; - - if (dev->flags & IFF_PROMISC) { - /* NIC to be configured in promiscuous mode. */ - receivemode = IPG_RM_RECEIVEALLFRAMES; - } else if ((dev->flags & IFF_ALLMULTI) || - ((dev->flags & IFF_MULTICAST) && - (netdev_mc_count(dev) > IPG_MULTICAST_HASHTABLE_SIZE))) { - /* NIC to be configured to receive all multicast - * frames. */ - receivemode |= IPG_RM_RECEIVEMULTICAST; - } else if ((dev->flags & IFF_MULTICAST) && !netdev_mc_empty(dev)) { - /* NIC to be configured to receive selected - * multicast addresses. */ - receivemode |= IPG_RM_RECEIVEMULTICASTHASH; - } - - /* Calculate the bits to set for the 64 bit, IPG HASHTABLE. - * The IPG applies a cyclic-redundancy-check (the same CRC - * used to calculate the frame data FCS) to the destination - * address all incoming multicast frames whose destination - * address has the multicast bit set. The least significant - * 6 bits of the CRC result are used as an addressing index - * into the hash table. If the value of the bit addressed by - * this index is a 1, the frame is passed to the host system. - */ - - /* Clear hashtable. */ - hashtable[0] = 0x00000000; - hashtable[1] = 0x00000000; - - /* Cycle through all multicast addresses to filter. */ - netdev_for_each_mc_addr(ha, dev) { - /* Calculate CRC result for each multicast address. */ - hashindex = crc32_le(0xffffffff, ha->addr, - ETH_ALEN); - - /* Use only the least significant 6 bits. */ - hashindex = hashindex & 0x3F; - - /* Within "hashtable", set bit number "hashindex" - * to a logic 1. - */ - set_bit(hashindex, (void *)hashtable); - } - - /* Write the value of the hashtable, to the 4, 16 bit - * HASHTABLE IPG registers. - */ - ipg_w32(hashtable[0], HASHTABLE_0); - ipg_w32(hashtable[1], HASHTABLE_1); - - ipg_w8(IPG_RM_RSVD_MASK & receivemode, RECEIVE_MODE); - - IPG_DEBUG_MSG("ReceiveMode = %x\n", ipg_r8(RECEIVE_MODE)); -} - -static int ipg_io_config(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = ipg_ioaddr(dev); - u32 origmacctrl; - u32 restoremacctrl; - - IPG_DEBUG_MSG("_io_config\n"); - - origmacctrl = ipg_r32(MAC_CTRL); - - restoremacctrl = origmacctrl | IPG_MC_STATISTICS_ENABLE; - - /* Based on compilation option, determine if FCS is to be - * stripped on receive frames by IPG. - */ - if (!IPG_STRIP_FCS_ON_RX) - restoremacctrl |= IPG_MC_RCV_FCS; - - /* Determine if transmitter and/or receiver are - * enabled so we may restore MACCTRL correctly. - */ - if (origmacctrl & IPG_MC_TX_ENABLED) - restoremacctrl |= IPG_MC_TX_ENABLE; - - if (origmacctrl & IPG_MC_RX_ENABLED) - restoremacctrl |= IPG_MC_RX_ENABLE; - - /* Transmitter and receiver must be disabled before setting - * IFSSelect. - */ - ipg_w32((origmacctrl & (IPG_MC_RX_DISABLE | IPG_MC_TX_DISABLE)) & - IPG_MC_RSVD_MASK, MAC_CTRL); - - /* Now that transmitter and receiver are disabled, write - * to IFSSelect. - */ - ipg_w32((origmacctrl & IPG_MC_IFS_96BIT) & IPG_MC_RSVD_MASK, MAC_CTRL); - - /* Set RECEIVEMODE register. */ - ipg_nic_set_multicast_list(dev); - - ipg_w16(sp->max_rxframe_size, MAX_FRAME_SIZE); - - ipg_w8(IPG_RXDMAPOLLPERIOD_VALUE, RX_DMA_POLL_PERIOD); - ipg_w8(IPG_RXDMAURGENTTHRESH_VALUE, RX_DMA_URGENT_THRESH); - ipg_w8(IPG_RXDMABURSTTHRESH_VALUE, RX_DMA_BURST_THRESH); - ipg_w8(IPG_TXDMAPOLLPERIOD_VALUE, TX_DMA_POLL_PERIOD); - ipg_w8(IPG_TXDMAURGENTTHRESH_VALUE, TX_DMA_URGENT_THRESH); - ipg_w8(IPG_TXDMABURSTTHRESH_VALUE, TX_DMA_BURST_THRESH); - ipg_w16((IPG_IE_HOST_ERROR | IPG_IE_TX_DMA_COMPLETE | - IPG_IE_TX_COMPLETE | IPG_IE_INT_REQUESTED | - IPG_IE_UPDATE_STATS | IPG_IE_LINK_EVENT | - IPG_IE_RX_DMA_COMPLETE | IPG_IE_RX_DMA_PRIORITY), INT_ENABLE); - ipg_w16(IPG_FLOWONTHRESH_VALUE, FLOW_ON_THRESH); - ipg_w16(IPG_FLOWOFFTHRESH_VALUE, FLOW_OFF_THRESH); - - /* IPG multi-frag frame bug workaround. - * Per silicon revision B3 eratta. - */ - ipg_w16(ipg_r16(DEBUG_CTRL) | 0x0200, DEBUG_CTRL); - - /* IPG TX poll now bug workaround. - * Per silicon revision B3 eratta. - */ - ipg_w16(ipg_r16(DEBUG_CTRL) | 0x0010, DEBUG_CTRL); - - /* IPG RX poll now bug workaround. - * Per silicon revision B3 eratta. - */ - ipg_w16(ipg_r16(DEBUG_CTRL) | 0x0020, DEBUG_CTRL); - - /* Now restore MACCTRL to original setting. */ - ipg_w32(IPG_MC_RSVD_MASK & restoremacctrl, MAC_CTRL); - - /* Disable unused RMON statistics. */ - ipg_w32(IPG_RZ_ALL, RMON_STATISTICS_MASK); - - /* Disable unused MIB statistics. */ - ipg_w32(IPG_SM_MACCONTROLFRAMESXMTD | IPG_SM_MACCONTROLFRAMESRCVD | - IPG_SM_BCSTOCTETXMTOK_BCSTFRAMESXMTDOK | IPG_SM_TXJUMBOFRAMES | - IPG_SM_MCSTOCTETXMTOK_MCSTFRAMESXMTDOK | IPG_SM_RXJUMBOFRAMES | - IPG_SM_BCSTOCTETRCVDOK_BCSTFRAMESRCVDOK | - IPG_SM_UDPCHECKSUMERRORS | IPG_SM_TCPCHECKSUMERRORS | - IPG_SM_IPCHECKSUMERRORS, STATISTICS_MASK); - - return 0; -} - -/* - * Create a receive buffer within system memory and update - * NIC private structure appropriately. - */ -static int ipg_get_rxbuff(struct net_device *dev, int entry) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - struct ipg_rx *rxfd = sp->rxd + entry; - struct sk_buff *skb; - u64 rxfragsize; - - IPG_DEBUG_MSG("_get_rxbuff\n"); - - skb = netdev_alloc_skb_ip_align(dev, sp->rxsupport_size); - if (!skb) { - sp->rx_buff[entry] = NULL; - return -ENOMEM; - } - - /* Save the address of the sk_buff structure. */ - sp->rx_buff[entry] = skb; - - rxfd->frag_info = cpu_to_le64(pci_map_single(sp->pdev, skb->data, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE)); - - /* Set the RFD fragment length. */ - rxfragsize = sp->rxfrag_size; - rxfd->frag_info |= cpu_to_le64((rxfragsize << 48) & IPG_RFI_FRAGLEN); - - return 0; -} - -static int init_rfdlist(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - - IPG_DEBUG_MSG("_init_rfdlist\n"); - - for (i = 0; i < IPG_RFDLIST_LENGTH; i++) { - struct ipg_rx *rxfd = sp->rxd + i; - - if (sp->rx_buff[i]) { - pci_unmap_single(sp->pdev, - le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - dev_kfree_skb_irq(sp->rx_buff[i]); - sp->rx_buff[i] = NULL; - } - - /* Clear out the RFS field. */ - rxfd->rfs = 0x0000000000000000; - - if (ipg_get_rxbuff(dev, i) < 0) { - /* - * A receive buffer was not ready, break the - * RFD list here. - */ - IPG_DEBUG_MSG("Cannot allocate Rx buffer\n"); - - /* Just in case we cannot allocate a single RFD. - * Should not occur. - */ - if (i == 0) { - netdev_err(dev, "No memory available for RFD list\n"); - return -ENOMEM; - } - } - - rxfd->next_desc = cpu_to_le64(sp->rxd_map + - sizeof(struct ipg_rx)*(i + 1)); - } - sp->rxd[i - 1].next_desc = cpu_to_le64(sp->rxd_map); - - sp->rx_current = 0; - sp->rx_dirty = 0; - - /* Write the location of the RFDList to the IPG. */ - ipg_w32((u32) sp->rxd_map, RFD_LIST_PTR_0); - ipg_w32(0x00000000, RFD_LIST_PTR_1); - - return 0; -} - -static void init_tfdlist(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - - IPG_DEBUG_MSG("_init_tfdlist\n"); - - for (i = 0; i < IPG_TFDLIST_LENGTH; i++) { - struct ipg_tx *txfd = sp->txd + i; - - txfd->tfc = cpu_to_le64(IPG_TFC_TFDDONE); - - if (sp->tx_buff[i]) { - dev_kfree_skb_irq(sp->tx_buff[i]); - sp->tx_buff[i] = NULL; - } - - txfd->next_desc = cpu_to_le64(sp->txd_map + - sizeof(struct ipg_tx)*(i + 1)); - } - sp->txd[i - 1].next_desc = cpu_to_le64(sp->txd_map); - - sp->tx_current = 0; - sp->tx_dirty = 0; - - /* Write the location of the TFDList to the IPG. */ - IPG_DDEBUG_MSG("Starting TFDListPtr = %08x\n", - (u32) sp->txd_map); - ipg_w32((u32) sp->txd_map, TFD_LIST_PTR_0); - ipg_w32(0x00000000, TFD_LIST_PTR_1); - - sp->reset_current_tfd = 1; -} - -/* - * Free all transmit buffers which have already been transferred - * via DMA to the IPG. - */ -static void ipg_nic_txfree(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - unsigned int released, pending, dirty; - - IPG_DEBUG_MSG("_nic_txfree\n"); - - pending = sp->tx_current - sp->tx_dirty; - dirty = sp->tx_dirty % IPG_TFDLIST_LENGTH; - - for (released = 0; released < pending; released++) { - struct sk_buff *skb = sp->tx_buff[dirty]; - struct ipg_tx *txfd = sp->txd + dirty; - - IPG_DEBUG_MSG("TFC = %016lx\n", (unsigned long) txfd->tfc); - - /* Look at each TFD's TFC field beginning - * at the last freed TFD up to the current TFD. - * If the TFDDone bit is set, free the associated - * buffer. - */ - if (!(txfd->tfc & cpu_to_le64(IPG_TFC_TFDDONE))) - break; - - /* Free the transmit buffer. */ - if (skb) { - pci_unmap_single(sp->pdev, - le64_to_cpu(txfd->frag_info) & ~IPG_TFI_FRAGLEN, - skb->len, PCI_DMA_TODEVICE); - - dev_kfree_skb_irq(skb); - - sp->tx_buff[dirty] = NULL; - } - dirty = (dirty + 1) % IPG_TFDLIST_LENGTH; - } - - sp->tx_dirty += released; - - if (netif_queue_stopped(dev) && - (sp->tx_current != (sp->tx_dirty + IPG_TFDLIST_LENGTH))) { - netif_wake_queue(dev); - } -} - -static void ipg_tx_timeout(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - - ipg_reset(dev, IPG_AC_TX_RESET | IPG_AC_DMA | IPG_AC_NETWORK | - IPG_AC_FIFO); - - spin_lock_irq(&sp->lock); - - /* Re-configure after DMA reset. */ - if (ipg_io_config(dev) < 0) - netdev_info(dev, "Error during re-configuration\n"); - - init_tfdlist(dev); - - spin_unlock_irq(&sp->lock); - - ipg_w32((ipg_r32(MAC_CTRL) | IPG_MC_TX_ENABLE) & IPG_MC_RSVD_MASK, - MAC_CTRL); -} - -/* - * For TxComplete interrupts, free all transmit - * buffers which have already been transferred via DMA - * to the IPG. - */ -static void ipg_nic_txcleanup(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - - IPG_DEBUG_MSG("_nic_txcleanup\n"); - - for (i = 0; i < IPG_TFDLIST_LENGTH; i++) { - /* Reading the TXSTATUS register clears the - * TX_COMPLETE interrupt. - */ - u32 txstatusdword = ipg_r32(TX_STATUS); - - IPG_DEBUG_MSG("TxStatus = %08x\n", txstatusdword); - - /* Check for Transmit errors. Error bits only valid if - * TX_COMPLETE bit in the TXSTATUS register is a 1. - */ - if (!(txstatusdword & IPG_TS_TX_COMPLETE)) - break; - - /* If in 10Mbps mode, indicate transmit is ready. */ - if (sp->tenmbpsmode) { - netif_wake_queue(dev); - } - - /* Transmit error, increment stat counters. */ - if (txstatusdword & IPG_TS_TX_ERROR) { - IPG_DEBUG_MSG("Transmit error\n"); - sp->stats.tx_errors++; - } - - /* Late collision, re-enable transmitter. */ - if (txstatusdword & IPG_TS_LATE_COLLISION) { - IPG_DEBUG_MSG("Late collision on transmit\n"); - ipg_w32((ipg_r32(MAC_CTRL) | IPG_MC_TX_ENABLE) & - IPG_MC_RSVD_MASK, MAC_CTRL); - } - - /* Maximum collisions, re-enable transmitter. */ - if (txstatusdword & IPG_TS_TX_MAX_COLL) { - IPG_DEBUG_MSG("Maximum collisions on transmit\n"); - ipg_w32((ipg_r32(MAC_CTRL) | IPG_MC_TX_ENABLE) & - IPG_MC_RSVD_MASK, MAC_CTRL); - } - - /* Transmit underrun, reset and re-enable - * transmitter. - */ - if (txstatusdword & IPG_TS_TX_UNDERRUN) { - IPG_DEBUG_MSG("Transmitter underrun\n"); - sp->stats.tx_fifo_errors++; - ipg_reset(dev, IPG_AC_TX_RESET | IPG_AC_DMA | - IPG_AC_NETWORK | IPG_AC_FIFO); - - /* Re-configure after DMA reset. */ - if (ipg_io_config(dev) < 0) { - netdev_info(dev, "Error during re-configuration\n"); - } - init_tfdlist(dev); - - ipg_w32((ipg_r32(MAC_CTRL) | IPG_MC_TX_ENABLE) & - IPG_MC_RSVD_MASK, MAC_CTRL); - } - } - - ipg_nic_txfree(dev); -} - -/* Provides statistical information about the IPG NIC. */ -static struct net_device_stats *ipg_nic_get_stats(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - u16 temp1; - u16 temp2; - - IPG_DEBUG_MSG("_nic_get_stats\n"); - - /* Check to see if the NIC has been initialized via nic_open, - * before trying to read statistic registers. - */ - if (!netif_running(dev)) - return &sp->stats; - - sp->stats.rx_packets += ipg_r32(IPG_FRAMESRCVDOK); - sp->stats.tx_packets += ipg_r32(IPG_FRAMESXMTDOK); - sp->stats.rx_bytes += ipg_r32(IPG_OCTETRCVOK); - sp->stats.tx_bytes += ipg_r32(IPG_OCTETXMTOK); - temp1 = ipg_r16(IPG_FRAMESLOSTRXERRORS); - sp->stats.rx_errors += temp1; - sp->stats.rx_missed_errors += temp1; - temp1 = ipg_r32(IPG_SINGLECOLFRAMES) + ipg_r32(IPG_MULTICOLFRAMES) + - ipg_r32(IPG_LATECOLLISIONS); - temp2 = ipg_r16(IPG_CARRIERSENSEERRORS); - sp->stats.collisions += temp1; - sp->stats.tx_dropped += ipg_r16(IPG_FRAMESABORTXSCOLLS); - sp->stats.tx_errors += ipg_r16(IPG_FRAMESWEXDEFERRAL) + - ipg_r32(IPG_FRAMESWDEFERREDXMT) + temp1 + temp2; - sp->stats.multicast += ipg_r32(IPG_MCSTOCTETRCVDOK); - - /* detailed tx_errors */ - sp->stats.tx_carrier_errors += temp2; - - /* detailed rx_errors */ - sp->stats.rx_length_errors += ipg_r16(IPG_INRANGELENGTHERRORS) + - ipg_r16(IPG_FRAMETOOLONGERRORS); - sp->stats.rx_crc_errors += ipg_r16(IPG_FRAMECHECKSEQERRORS); - - /* Unutilized IPG statistic registers. */ - ipg_r32(IPG_MCSTFRAMESRCVDOK); - - return &sp->stats; -} - -/* Restore used receive buffers. */ -static int ipg_nic_rxrestore(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - const unsigned int curr = sp->rx_current; - unsigned int dirty = sp->rx_dirty; - - IPG_DEBUG_MSG("_nic_rxrestore\n"); - - for (dirty = sp->rx_dirty; curr - dirty > 0; dirty++) { - unsigned int entry = dirty % IPG_RFDLIST_LENGTH; - - /* rx_copybreak may poke hole here and there. */ - if (sp->rx_buff[entry]) - continue; - - /* Generate a new receive buffer to replace the - * current buffer (which will be released by the - * Linux system). - */ - if (ipg_get_rxbuff(dev, entry) < 0) { - IPG_DEBUG_MSG("Cannot allocate new Rx buffer\n"); - - break; - } - - /* Reset the RFS field. */ - sp->rxd[entry].rfs = 0x0000000000000000; - } - sp->rx_dirty = dirty; - - return 0; -} - -/* use jumboindex and jumbosize to control jumbo frame status - * initial status is jumboindex=-1 and jumbosize=0 - * 1. jumboindex = -1 and jumbosize=0 : previous jumbo frame has been done. - * 2. jumboindex != -1 and jumbosize != 0 : jumbo frame is not over size and receiving - * 3. jumboindex = -1 and jumbosize != 0 : jumbo frame is over size, already dump - * previous receiving and need to continue dumping the current one - */ -enum { - NORMAL_PACKET, - ERROR_PACKET -}; - -enum { - FRAME_NO_START_NO_END = 0, - FRAME_WITH_START = 1, - FRAME_WITH_END = 10, - FRAME_WITH_START_WITH_END = 11 -}; - -static void ipg_nic_rx_free_skb(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - unsigned int entry = sp->rx_current % IPG_RFDLIST_LENGTH; - - if (sp->rx_buff[entry]) { - struct ipg_rx *rxfd = sp->rxd + entry; - - pci_unmap_single(sp->pdev, - le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - dev_kfree_skb_irq(sp->rx_buff[entry]); - sp->rx_buff[entry] = NULL; - } -} - -static int ipg_nic_rx_check_frame_type(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - struct ipg_rx *rxfd = sp->rxd + (sp->rx_current % IPG_RFDLIST_LENGTH); - int type = FRAME_NO_START_NO_END; - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_FRAMESTART) - type += FRAME_WITH_START; - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_FRAMEEND) - type += FRAME_WITH_END; - return type; -} - -static int ipg_nic_rx_check_error(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - unsigned int entry = sp->rx_current % IPG_RFDLIST_LENGTH; - struct ipg_rx *rxfd = sp->rxd + entry; - - if (IPG_DROP_ON_RX_ETH_ERRORS && (le64_to_cpu(rxfd->rfs) & - (IPG_RFS_RXFIFOOVERRUN | IPG_RFS_RXRUNTFRAME | - IPG_RFS_RXALIGNMENTERROR | IPG_RFS_RXFCSERROR | - IPG_RFS_RXOVERSIZEDFRAME | IPG_RFS_RXLENGTHERROR))) { - IPG_DEBUG_MSG("Rx error, RFS = %016lx\n", - (unsigned long) rxfd->rfs); - - /* Increment general receive error statistic. */ - sp->stats.rx_errors++; - - /* Increment detailed receive error statistics. */ - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFIFOOVERRUN) { - IPG_DEBUG_MSG("RX FIFO overrun occurred\n"); - - sp->stats.rx_fifo_errors++; - } - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXRUNTFRAME) { - IPG_DEBUG_MSG("RX runt occurred\n"); - sp->stats.rx_length_errors++; - } - - /* Do nothing for IPG_RFS_RXOVERSIZEDFRAME, - * error count handled by a IPG statistic register. - */ - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXALIGNMENTERROR) { - IPG_DEBUG_MSG("RX alignment error occurred\n"); - sp->stats.rx_frame_errors++; - } - - /* Do nothing for IPG_RFS_RXFCSERROR, error count - * handled by a IPG statistic register. - */ - - /* Free the memory associated with the RX - * buffer since it is erroneous and we will - * not pass it to higher layer processes. - */ - if (sp->rx_buff[entry]) { - pci_unmap_single(sp->pdev, - le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - - dev_kfree_skb_irq(sp->rx_buff[entry]); - sp->rx_buff[entry] = NULL; - } - return ERROR_PACKET; - } - return NORMAL_PACKET; -} - -static void ipg_nic_rx_with_start_and_end(struct net_device *dev, - struct ipg_nic_private *sp, - struct ipg_rx *rxfd, unsigned entry) -{ - struct ipg_jumbo *jumbo = &sp->jumbo; - struct sk_buff *skb; - int framelen; - - if (jumbo->found_start) { - dev_kfree_skb_irq(jumbo->skb); - jumbo->found_start = 0; - jumbo->current_size = 0; - jumbo->skb = NULL; - } - - /* 1: found error, 0 no error */ - if (ipg_nic_rx_check_error(dev) != NORMAL_PACKET) - return; - - skb = sp->rx_buff[entry]; - if (!skb) - return; - - /* accept this frame and send to upper layer */ - framelen = le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFRAMELEN; - if (framelen > sp->rxfrag_size) - framelen = sp->rxfrag_size; - - skb_put(skb, framelen); - skb->protocol = eth_type_trans(skb, dev); - skb_checksum_none_assert(skb); - netif_rx(skb); - sp->rx_buff[entry] = NULL; -} - -static void ipg_nic_rx_with_start(struct net_device *dev, - struct ipg_nic_private *sp, - struct ipg_rx *rxfd, unsigned entry) -{ - struct ipg_jumbo *jumbo = &sp->jumbo; - struct pci_dev *pdev = sp->pdev; - struct sk_buff *skb; - - /* 1: found error, 0 no error */ - if (ipg_nic_rx_check_error(dev) != NORMAL_PACKET) - return; - - /* accept this frame and send to upper layer */ - skb = sp->rx_buff[entry]; - if (!skb) - return; - - if (jumbo->found_start) - dev_kfree_skb_irq(jumbo->skb); - - pci_unmap_single(pdev, le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - - skb_put(skb, sp->rxfrag_size); - - jumbo->found_start = 1; - jumbo->current_size = sp->rxfrag_size; - jumbo->skb = skb; - - sp->rx_buff[entry] = NULL; -} - -static void ipg_nic_rx_with_end(struct net_device *dev, - struct ipg_nic_private *sp, - struct ipg_rx *rxfd, unsigned entry) -{ - struct ipg_jumbo *jumbo = &sp->jumbo; - - /* 1: found error, 0 no error */ - if (ipg_nic_rx_check_error(dev) == NORMAL_PACKET) { - struct sk_buff *skb = sp->rx_buff[entry]; - - if (!skb) - return; - - if (jumbo->found_start) { - int framelen, endframelen; - - framelen = le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFRAMELEN; - - endframelen = framelen - jumbo->current_size; - if (framelen > sp->rxsupport_size) - dev_kfree_skb_irq(jumbo->skb); - else { - memcpy(skb_put(jumbo->skb, endframelen), - skb->data, endframelen); - - jumbo->skb->protocol = - eth_type_trans(jumbo->skb, dev); - - skb_checksum_none_assert(jumbo->skb); - netif_rx(jumbo->skb); - } - } - - jumbo->found_start = 0; - jumbo->current_size = 0; - jumbo->skb = NULL; - - ipg_nic_rx_free_skb(dev); - } else { - dev_kfree_skb_irq(jumbo->skb); - jumbo->found_start = 0; - jumbo->current_size = 0; - jumbo->skb = NULL; - } -} - -static void ipg_nic_rx_no_start_no_end(struct net_device *dev, - struct ipg_nic_private *sp, - struct ipg_rx *rxfd, unsigned entry) -{ - struct ipg_jumbo *jumbo = &sp->jumbo; - - /* 1: found error, 0 no error */ - if (ipg_nic_rx_check_error(dev) == NORMAL_PACKET) { - struct sk_buff *skb = sp->rx_buff[entry]; - - if (skb) { - if (jumbo->found_start) { - jumbo->current_size += sp->rxfrag_size; - if (jumbo->current_size <= sp->rxsupport_size) { - memcpy(skb_put(jumbo->skb, - sp->rxfrag_size), - skb->data, sp->rxfrag_size); - } - } - ipg_nic_rx_free_skb(dev); - } - } else { - dev_kfree_skb_irq(jumbo->skb); - jumbo->found_start = 0; - jumbo->current_size = 0; - jumbo->skb = NULL; - } -} - -static int ipg_nic_rx_jumbo(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - unsigned int curr = sp->rx_current; - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - - IPG_DEBUG_MSG("_nic_rx\n"); - - for (i = 0; i < IPG_MAXRFDPROCESS_COUNT; i++, curr++) { - unsigned int entry = curr % IPG_RFDLIST_LENGTH; - struct ipg_rx *rxfd = sp->rxd + entry; - - if (!(rxfd->rfs & cpu_to_le64(IPG_RFS_RFDDONE))) - break; - - switch (ipg_nic_rx_check_frame_type(dev)) { - case FRAME_WITH_START_WITH_END: - ipg_nic_rx_with_start_and_end(dev, sp, rxfd, entry); - break; - case FRAME_WITH_START: - ipg_nic_rx_with_start(dev, sp, rxfd, entry); - break; - case FRAME_WITH_END: - ipg_nic_rx_with_end(dev, sp, rxfd, entry); - break; - case FRAME_NO_START_NO_END: - ipg_nic_rx_no_start_no_end(dev, sp, rxfd, entry); - break; - } - } - - sp->rx_current = curr; - - if (i == IPG_MAXRFDPROCESS_COUNT) { - /* There are more RFDs to process, however the - * allocated amount of RFD processing time has - * expired. Assert Interrupt Requested to make - * sure we come back to process the remaining RFDs. - */ - ipg_w32(ipg_r32(ASIC_CTRL) | IPG_AC_INT_REQUEST, ASIC_CTRL); - } - - ipg_nic_rxrestore(dev); - - return 0; -} - -static int ipg_nic_rx(struct net_device *dev) -{ - /* Transfer received Ethernet frames to higher network layers. */ - struct ipg_nic_private *sp = netdev_priv(dev); - unsigned int curr = sp->rx_current; - void __iomem *ioaddr = sp->ioaddr; - struct ipg_rx *rxfd; - unsigned int i; - - IPG_DEBUG_MSG("_nic_rx\n"); - -#define __RFS_MASK \ - cpu_to_le64(IPG_RFS_RFDDONE | IPG_RFS_FRAMESTART | IPG_RFS_FRAMEEND) - - for (i = 0; i < IPG_MAXRFDPROCESS_COUNT; i++, curr++) { - unsigned int entry = curr % IPG_RFDLIST_LENGTH; - struct sk_buff *skb = sp->rx_buff[entry]; - unsigned int framelen; - - rxfd = sp->rxd + entry; - - if (((rxfd->rfs & __RFS_MASK) != __RFS_MASK) || !skb) - break; - - /* Get received frame length. */ - framelen = le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFRAMELEN; - - /* Check for jumbo frame arrival with too small - * RXFRAG_SIZE. - */ - if (framelen > sp->rxfrag_size) { - IPG_DEBUG_MSG - ("RFS FrameLen > allocated fragment size\n"); - - framelen = sp->rxfrag_size; - } - - if ((IPG_DROP_ON_RX_ETH_ERRORS && (le64_to_cpu(rxfd->rfs) & - (IPG_RFS_RXFIFOOVERRUN | IPG_RFS_RXRUNTFRAME | - IPG_RFS_RXALIGNMENTERROR | IPG_RFS_RXFCSERROR | - IPG_RFS_RXOVERSIZEDFRAME | IPG_RFS_RXLENGTHERROR)))) { - - IPG_DEBUG_MSG("Rx error, RFS = %016lx\n", - (unsigned long int) rxfd->rfs); - - /* Increment general receive error statistic. */ - sp->stats.rx_errors++; - - /* Increment detailed receive error statistics. */ - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFIFOOVERRUN) { - IPG_DEBUG_MSG("RX FIFO overrun occurred\n"); - sp->stats.rx_fifo_errors++; - } - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXRUNTFRAME) { - IPG_DEBUG_MSG("RX runt occurred\n"); - sp->stats.rx_length_errors++; - } - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXOVERSIZEDFRAME) ; - /* Do nothing, error count handled by a IPG - * statistic register. - */ - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXALIGNMENTERROR) { - IPG_DEBUG_MSG("RX alignment error occurred\n"); - sp->stats.rx_frame_errors++; - } - - if (le64_to_cpu(rxfd->rfs) & IPG_RFS_RXFCSERROR) ; - /* Do nothing, error count handled by a IPG - * statistic register. - */ - - /* Free the memory associated with the RX - * buffer since it is erroneous and we will - * not pass it to higher layer processes. - */ - if (skb) { - __le64 info = rxfd->frag_info; - - pci_unmap_single(sp->pdev, - le64_to_cpu(info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - - dev_kfree_skb_irq(skb); - } - } else { - - /* Adjust the new buffer length to accommodate the size - * of the received frame. - */ - skb_put(skb, framelen); - - /* Set the buffer's protocol field to Ethernet. */ - skb->protocol = eth_type_trans(skb, dev); - - /* The IPG encountered an error with (or - * there were no) IP/TCP/UDP checksums. - * This may or may not indicate an invalid - * IP/TCP/UDP frame was received. Let the - * upper layer decide. - */ - skb_checksum_none_assert(skb); - - /* Hand off frame for higher layer processing. - * The function netif_rx() releases the sk_buff - * when processing completes. - */ - netif_rx(skb); - } - - /* Assure RX buffer is not reused by IPG. */ - sp->rx_buff[entry] = NULL; - } - - /* - * If there are more RFDs to process and the allocated amount of RFD - * processing time has expired, assert Interrupt Requested to make - * sure we come back to process the remaining RFDs. - */ - if (i == IPG_MAXRFDPROCESS_COUNT) - ipg_w32(ipg_r32(ASIC_CTRL) | IPG_AC_INT_REQUEST, ASIC_CTRL); - -#ifdef IPG_DEBUG - /* Check if the RFD list contained no receive frame data. */ - if (!i) - sp->EmptyRFDListCount++; -#endif - while ((le64_to_cpu(rxfd->rfs) & IPG_RFS_RFDDONE) && - !((le64_to_cpu(rxfd->rfs) & IPG_RFS_FRAMESTART) && - (le64_to_cpu(rxfd->rfs) & IPG_RFS_FRAMEEND))) { - unsigned int entry = curr++ % IPG_RFDLIST_LENGTH; - - rxfd = sp->rxd + entry; - - IPG_DEBUG_MSG("Frame requires multiple RFDs\n"); - - /* An unexpected event, additional code needed to handle - * properly. So for the time being, just disregard the - * frame. - */ - - /* Free the memory associated with the RX - * buffer since it is erroneous and we will - * not pass it to higher layer processes. - */ - if (sp->rx_buff[entry]) { - pci_unmap_single(sp->pdev, - le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - dev_kfree_skb_irq(sp->rx_buff[entry]); - } - - /* Assure RX buffer is not reused by IPG. */ - sp->rx_buff[entry] = NULL; - } - - sp->rx_current = curr; - - /* Check to see if there are a minimum number of used - * RFDs before restoring any (should improve performance.) - */ - if ((curr - sp->rx_dirty) >= IPG_MINUSEDRFDSTOFREE) - ipg_nic_rxrestore(dev); - - return 0; -} - -static void ipg_reset_after_host_error(struct work_struct *work) -{ - struct ipg_nic_private *sp = - container_of(work, struct ipg_nic_private, task.work); - struct net_device *dev = sp->dev; - - /* - * Acknowledge HostError interrupt by resetting - * IPG DMA and HOST. - */ - ipg_reset(dev, IPG_AC_GLOBAL_RESET | IPG_AC_HOST | IPG_AC_DMA); - - init_rfdlist(dev); - init_tfdlist(dev); - - if (ipg_io_config(dev) < 0) { - netdev_info(dev, "Cannot recover from PCI error\n"); - schedule_delayed_work(&sp->task, HZ); - } -} - -static irqreturn_t ipg_interrupt_handler(int irq, void *dev_inst) -{ - struct net_device *dev = dev_inst; - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int handled = 0; - u16 status; - - IPG_DEBUG_MSG("_interrupt_handler\n"); - - if (sp->is_jumbo) - ipg_nic_rxrestore(dev); - - spin_lock(&sp->lock); - - /* Get interrupt source information, and acknowledge - * some (i.e. TxDMAComplete, RxDMAComplete, RxEarly, - * IntRequested, MacControlFrame, LinkEvent) interrupts - * if issued. Also, all IPG interrupts are disabled by - * reading IntStatusAck. - */ - status = ipg_r16(INT_STATUS_ACK); - - IPG_DEBUG_MSG("IntStatusAck = %04x\n", status); - - /* Shared IRQ of remove event. */ - if (!(status & IPG_IS_RSVD_MASK)) - goto out_enable; - - handled = 1; - - if (unlikely(!netif_running(dev))) - goto out_unlock; - - /* If RFDListEnd interrupt, restore all used RFDs. */ - if (status & IPG_IS_RFD_LIST_END) { - IPG_DEBUG_MSG("RFDListEnd Interrupt\n"); - - /* The RFD list end indicates an RFD was encountered - * with a 0 NextPtr, or with an RFDDone bit set to 1 - * (indicating the RFD is not read for use by the - * IPG.) Try to restore all RFDs. - */ - ipg_nic_rxrestore(dev); - -#ifdef IPG_DEBUG - /* Increment the RFDlistendCount counter. */ - sp->RFDlistendCount++; -#endif - } - - /* If RFDListEnd, RxDMAPriority, RxDMAComplete, or - * IntRequested interrupt, process received frames. */ - if ((status & IPG_IS_RX_DMA_PRIORITY) || - (status & IPG_IS_RFD_LIST_END) || - (status & IPG_IS_RX_DMA_COMPLETE) || - (status & IPG_IS_INT_REQUESTED)) { -#ifdef IPG_DEBUG - /* Increment the RFD list checked counter if interrupted - * only to check the RFD list. */ - if (status & (~(IPG_IS_RX_DMA_PRIORITY | IPG_IS_RFD_LIST_END | - IPG_IS_RX_DMA_COMPLETE | IPG_IS_INT_REQUESTED) & - (IPG_IS_HOST_ERROR | IPG_IS_TX_DMA_COMPLETE | - IPG_IS_LINK_EVENT | IPG_IS_TX_COMPLETE | - IPG_IS_UPDATE_STATS))) - sp->RFDListCheckedCount++; -#endif - - if (sp->is_jumbo) - ipg_nic_rx_jumbo(dev); - else - ipg_nic_rx(dev); - } - - /* If TxDMAComplete interrupt, free used TFDs. */ - if (status & IPG_IS_TX_DMA_COMPLETE) - ipg_nic_txfree(dev); - - /* TxComplete interrupts indicate one of numerous actions. - * Determine what action to take based on TXSTATUS register. - */ - if (status & IPG_IS_TX_COMPLETE) - ipg_nic_txcleanup(dev); - - /* If UpdateStats interrupt, update Linux Ethernet statistics */ - if (status & IPG_IS_UPDATE_STATS) - ipg_nic_get_stats(dev); - - /* If HostError interrupt, reset IPG. */ - if (status & IPG_IS_HOST_ERROR) { - IPG_DDEBUG_MSG("HostError Interrupt\n"); - - schedule_delayed_work(&sp->task, 0); - } - - /* If LinkEvent interrupt, resolve autonegotiation. */ - if (status & IPG_IS_LINK_EVENT) { - if (ipg_config_autoneg(dev) < 0) - netdev_info(dev, "Auto-negotiation error\n"); - } - - /* If MACCtrlFrame interrupt, do nothing. */ - if (status & IPG_IS_MAC_CTRL_FRAME) - IPG_DEBUG_MSG("MACCtrlFrame interrupt\n"); - - /* If RxComplete interrupt, do nothing. */ - if (status & IPG_IS_RX_COMPLETE) - IPG_DEBUG_MSG("RxComplete interrupt\n"); - - /* If RxEarly interrupt, do nothing. */ - if (status & IPG_IS_RX_EARLY) - IPG_DEBUG_MSG("RxEarly interrupt\n"); - -out_enable: - /* Re-enable IPG interrupts. */ - ipg_w16(IPG_IE_TX_DMA_COMPLETE | IPG_IE_RX_DMA_COMPLETE | - IPG_IE_HOST_ERROR | IPG_IE_INT_REQUESTED | IPG_IE_TX_COMPLETE | - IPG_IE_LINK_EVENT | IPG_IE_UPDATE_STATS, INT_ENABLE); -out_unlock: - spin_unlock(&sp->lock); - - return IRQ_RETVAL(handled); -} - -static void ipg_rx_clear(struct ipg_nic_private *sp) -{ - unsigned int i; - - for (i = 0; i < IPG_RFDLIST_LENGTH; i++) { - if (sp->rx_buff[i]) { - struct ipg_rx *rxfd = sp->rxd + i; - - dev_kfree_skb_irq(sp->rx_buff[i]); - sp->rx_buff[i] = NULL; - pci_unmap_single(sp->pdev, - le64_to_cpu(rxfd->frag_info) & ~IPG_RFI_FRAGLEN, - sp->rx_buf_sz, PCI_DMA_FROMDEVICE); - } - } -} - -static void ipg_tx_clear(struct ipg_nic_private *sp) -{ - unsigned int i; - - for (i = 0; i < IPG_TFDLIST_LENGTH; i++) { - if (sp->tx_buff[i]) { - struct ipg_tx *txfd = sp->txd + i; - - pci_unmap_single(sp->pdev, - le64_to_cpu(txfd->frag_info) & ~IPG_TFI_FRAGLEN, - sp->tx_buff[i]->len, PCI_DMA_TODEVICE); - - dev_kfree_skb_irq(sp->tx_buff[i]); - - sp->tx_buff[i] = NULL; - } - } -} - -static int ipg_nic_open(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - struct pci_dev *pdev = sp->pdev; - int rc; - - IPG_DEBUG_MSG("_nic_open\n"); - - sp->rx_buf_sz = sp->rxsupport_size; - - /* Check for interrupt line conflicts, and request interrupt - * line for IPG. - * - * IMPORTANT: Disable IPG interrupts prior to registering - * IRQ. - */ - ipg_w16(0x0000, INT_ENABLE); - - /* Register the interrupt line to be used by the IPG within - * the Linux system. - */ - rc = request_irq(pdev->irq, ipg_interrupt_handler, IRQF_SHARED, - dev->name, dev); - if (rc < 0) { - netdev_info(dev, "Error when requesting interrupt\n"); - goto out; - } - - dev->irq = pdev->irq; - - rc = -ENOMEM; - - sp->rxd = dma_alloc_coherent(&pdev->dev, IPG_RX_RING_BYTES, - &sp->rxd_map, GFP_KERNEL); - if (!sp->rxd) - goto err_free_irq_0; - - sp->txd = dma_alloc_coherent(&pdev->dev, IPG_TX_RING_BYTES, - &sp->txd_map, GFP_KERNEL); - if (!sp->txd) - goto err_free_rx_1; - - rc = init_rfdlist(dev); - if (rc < 0) { - netdev_info(dev, "Error during configuration\n"); - goto err_free_tx_2; - } - - init_tfdlist(dev); - - rc = ipg_io_config(dev); - if (rc < 0) { - netdev_info(dev, "Error during configuration\n"); - goto err_release_tfdlist_3; - } - - /* Resolve autonegotiation. */ - if (ipg_config_autoneg(dev) < 0) - netdev_info(dev, "Auto-negotiation error\n"); - - /* initialize JUMBO Frame control variable */ - sp->jumbo.found_start = 0; - sp->jumbo.current_size = 0; - sp->jumbo.skb = NULL; - - /* Enable transmit and receive operation of the IPG. */ - ipg_w32((ipg_r32(MAC_CTRL) | IPG_MC_RX_ENABLE | IPG_MC_TX_ENABLE) & - IPG_MC_RSVD_MASK, MAC_CTRL); - - netif_start_queue(dev); -out: - return rc; - -err_release_tfdlist_3: - ipg_tx_clear(sp); - ipg_rx_clear(sp); -err_free_tx_2: - dma_free_coherent(&pdev->dev, IPG_TX_RING_BYTES, sp->txd, sp->txd_map); -err_free_rx_1: - dma_free_coherent(&pdev->dev, IPG_RX_RING_BYTES, sp->rxd, sp->rxd_map); -err_free_irq_0: - free_irq(pdev->irq, dev); - goto out; -} - -static int ipg_nic_stop(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - struct pci_dev *pdev = sp->pdev; - - IPG_DEBUG_MSG("_nic_stop\n"); - - netif_stop_queue(dev); - - IPG_DUMPTFDLIST(dev); - - do { - (void) ipg_r16(INT_STATUS_ACK); - - ipg_reset(dev, IPG_AC_GLOBAL_RESET | IPG_AC_HOST | IPG_AC_DMA); - - synchronize_irq(pdev->irq); - } while (ipg_r16(INT_ENABLE) & IPG_IE_RSVD_MASK); - - ipg_rx_clear(sp); - - ipg_tx_clear(sp); - - pci_free_consistent(pdev, IPG_RX_RING_BYTES, sp->rxd, sp->rxd_map); - pci_free_consistent(pdev, IPG_TX_RING_BYTES, sp->txd, sp->txd_map); - - free_irq(pdev->irq, dev); - - return 0; -} - -static netdev_tx_t ipg_nic_hard_start_xmit(struct sk_buff *skb, - struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int entry = sp->tx_current % IPG_TFDLIST_LENGTH; - unsigned long flags; - struct ipg_tx *txfd; - - IPG_DDEBUG_MSG("_nic_hard_start_xmit\n"); - - /* If in 10Mbps mode, stop the transmit queue so - * no more transmit frames are accepted. - */ - if (sp->tenmbpsmode) - netif_stop_queue(dev); - - if (sp->reset_current_tfd) { - sp->reset_current_tfd = 0; - entry = 0; - } - - txfd = sp->txd + entry; - - sp->tx_buff[entry] = skb; - - /* Clear all TFC fields, except TFDDONE. */ - txfd->tfc = cpu_to_le64(IPG_TFC_TFDDONE); - - /* Specify the TFC field within the TFD. */ - txfd->tfc |= cpu_to_le64(IPG_TFC_WORDALIGNDISABLED | - (IPG_TFC_FRAMEID & sp->tx_current) | - (IPG_TFC_FRAGCOUNT & (1 << 24))); - /* - * 16--17 (WordAlign) <- 3 (disable), - * 0--15 (FrameId) <- sp->tx_current, - * 24--27 (FragCount) <- 1 - */ - - /* Request TxComplete interrupts at an interval defined - * by the constant IPG_FRAMESBETWEENTXCOMPLETES. - * Request TxComplete interrupt for every frame - * if in 10Mbps mode to accommodate problem with 10Mbps - * processing. - */ - if (sp->tenmbpsmode) - txfd->tfc |= cpu_to_le64(IPG_TFC_TXINDICATE); - txfd->tfc |= cpu_to_le64(IPG_TFC_TXDMAINDICATE); - /* Based on compilation option, determine if FCS is to be - * appended to transmit frame by IPG. - */ - if (!(IPG_APPEND_FCS_ON_TX)) - txfd->tfc |= cpu_to_le64(IPG_TFC_FCSAPPENDDISABLE); - - /* Based on compilation option, determine if IP, TCP and/or - * UDP checksums are to be added to transmit frame by IPG. - */ - if (IPG_ADD_IPCHECKSUM_ON_TX) - txfd->tfc |= cpu_to_le64(IPG_TFC_IPCHECKSUMENABLE); - - if (IPG_ADD_TCPCHECKSUM_ON_TX) - txfd->tfc |= cpu_to_le64(IPG_TFC_TCPCHECKSUMENABLE); - - if (IPG_ADD_UDPCHECKSUM_ON_TX) - txfd->tfc |= cpu_to_le64(IPG_TFC_UDPCHECKSUMENABLE); - - /* Based on compilation option, determine if VLAN tag info is to be - * inserted into transmit frame by IPG. - */ - if (IPG_INSERT_MANUAL_VLAN_TAG) { - txfd->tfc |= cpu_to_le64(IPG_TFC_VLANTAGINSERT | - ((u64) IPG_MANUAL_VLAN_VID << 32) | - ((u64) IPG_MANUAL_VLAN_CFI << 44) | - ((u64) IPG_MANUAL_VLAN_USERPRIORITY << 45)); - } - - /* The fragment start location within system memory is defined - * by the sk_buff structure's data field. The physical address - * of this location within the system's virtual memory space - * is determined using the IPG_HOST2BUS_MAP function. - */ - txfd->frag_info = cpu_to_le64(pci_map_single(sp->pdev, skb->data, - skb->len, PCI_DMA_TODEVICE)); - - /* The length of the fragment within system memory is defined by - * the sk_buff structure's len field. - */ - txfd->frag_info |= cpu_to_le64(IPG_TFI_FRAGLEN & - ((u64) (skb->len & 0xffff) << 48)); - - /* Clear the TFDDone bit last to indicate the TFD is ready - * for transfer to the IPG. - */ - txfd->tfc &= cpu_to_le64(~IPG_TFC_TFDDONE); - - spin_lock_irqsave(&sp->lock, flags); - - sp->tx_current++; - - mmiowb(); - - ipg_w32(IPG_DC_TX_DMA_POLL_NOW, DMA_CTRL); - - if (sp->tx_current == (sp->tx_dirty + IPG_TFDLIST_LENGTH)) - netif_stop_queue(dev); - - spin_unlock_irqrestore(&sp->lock, flags); - - return NETDEV_TX_OK; -} - -static void ipg_set_phy_default_param(unsigned char rev, - struct net_device *dev, int phy_address) -{ - unsigned short length; - unsigned char revision; - const unsigned short *phy_param; - unsigned short address, value; - - phy_param = &DefaultPhyParam[0]; - length = *phy_param & 0x00FF; - revision = (unsigned char)((*phy_param) >> 8); - phy_param++; - while (length != 0) { - if (rev == revision) { - while (length > 1) { - address = *phy_param; - value = *(phy_param + 1); - phy_param += 2; - mdio_write(dev, phy_address, address, value); - length -= 4; - } - break; - } else { - phy_param += length / 2; - length = *phy_param & 0x00FF; - revision = (unsigned char)((*phy_param) >> 8); - phy_param++; - } - } -} - -static int read_eeprom(struct net_device *dev, int eep_addr) -{ - void __iomem *ioaddr = ipg_ioaddr(dev); - unsigned int i; - int ret = 0; - u16 value; - - value = IPG_EC_EEPROM_READOPCODE | (eep_addr & 0xff); - ipg_w16(value, EEPROM_CTRL); - - for (i = 0; i < 1000; i++) { - u16 data; - - mdelay(10); - data = ipg_r16(EEPROM_CTRL); - if (!(data & IPG_EC_EEPROM_BUSY)) { - ret = ipg_r16(EEPROM_DATA); - break; - } - } - return ret; -} - -static void ipg_init_mii(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - struct mii_if_info *mii_if = &sp->mii_if; - int phyaddr; - - mii_if->dev = dev; - mii_if->mdio_read = mdio_read; - mii_if->mdio_write = mdio_write; - mii_if->phy_id_mask = 0x1f; - mii_if->reg_num_mask = 0x1f; - - mii_if->phy_id = phyaddr = ipg_find_phyaddr(dev); - - if (phyaddr != 0x1f) { - u16 mii_phyctrl, mii_1000cr; - - mii_1000cr = mdio_read(dev, phyaddr, MII_CTRL1000); - mii_1000cr |= ADVERTISE_1000FULL | ADVERTISE_1000HALF | - GMII_PHY_1000BASETCONTROL_PreferMaster; - mdio_write(dev, phyaddr, MII_CTRL1000, mii_1000cr); - - mii_phyctrl = mdio_read(dev, phyaddr, MII_BMCR); - - /* Set default phyparam */ - ipg_set_phy_default_param(sp->pdev->revision, dev, phyaddr); - - /* Reset PHY */ - mii_phyctrl |= BMCR_RESET | BMCR_ANRESTART; - mdio_write(dev, phyaddr, MII_BMCR, mii_phyctrl); - - } -} - -static int ipg_hw_init(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - void __iomem *ioaddr = sp->ioaddr; - unsigned int i; - int rc; - - /* Read/Write and Reset EEPROM Value */ - /* Read LED Mode Configuration from EEPROM */ - sp->led_mode = read_eeprom(dev, 6); - - /* Reset all functions within the IPG. Do not assert - * RST_OUT as not compatible with some PHYs. - */ - rc = ipg_reset(dev, IPG_RESET_MASK); - if (rc < 0) - goto out; - - ipg_init_mii(dev); - - /* Read MAC Address from EEPROM */ - for (i = 0; i < 3; i++) - sp->station_addr[i] = read_eeprom(dev, 16 + i); - - for (i = 0; i < 3; i++) - ipg_w16(sp->station_addr[i], STATION_ADDRESS_0 + 2*i); - - /* Set station address in ethernet_device structure. */ - dev->dev_addr[0] = ipg_r16(STATION_ADDRESS_0) & 0x00ff; - dev->dev_addr[1] = (ipg_r16(STATION_ADDRESS_0) & 0xff00) >> 8; - dev->dev_addr[2] = ipg_r16(STATION_ADDRESS_1) & 0x00ff; - dev->dev_addr[3] = (ipg_r16(STATION_ADDRESS_1) & 0xff00) >> 8; - dev->dev_addr[4] = ipg_r16(STATION_ADDRESS_2) & 0x00ff; - dev->dev_addr[5] = (ipg_r16(STATION_ADDRESS_2) & 0xff00) >> 8; -out: - return rc; -} - -static int ipg_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - int rc; - - mutex_lock(&sp->mii_mutex); - rc = generic_mii_ioctl(&sp->mii_if, if_mii(ifr), cmd, NULL); - mutex_unlock(&sp->mii_mutex); - - return rc; -} - -static int ipg_nic_change_mtu(struct net_device *dev, int new_mtu) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - int err; - - /* Function to accommodate changes to Maximum Transfer Unit - * (or MTU) of IPG NIC. Cannot use default function since - * the default will not allow for MTU > 1500 bytes. - */ - - IPG_DEBUG_MSG("_nic_change_mtu\n"); - - /* - * Check that the new MTU value is between 68 (14 byte header, 46 byte - * payload, 4 byte FCS) and 10 KB, which is the largest supported MTU. - */ - if (new_mtu < 68 || new_mtu > 10240) - return -EINVAL; - - err = ipg_nic_stop(dev); - if (err) - return err; - - dev->mtu = new_mtu; - - sp->max_rxframe_size = new_mtu; - - sp->rxfrag_size = new_mtu; - if (sp->rxfrag_size > 4088) - sp->rxfrag_size = 4088; - - sp->rxsupport_size = sp->max_rxframe_size; - - if (new_mtu > 0x0600) - sp->is_jumbo = true; - else - sp->is_jumbo = false; - - return ipg_nic_open(dev); -} - -static int ipg_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - int rc; - - mutex_lock(&sp->mii_mutex); - rc = mii_ethtool_gset(&sp->mii_if, cmd); - mutex_unlock(&sp->mii_mutex); - - return rc; -} - -static int ipg_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - int rc; - - mutex_lock(&sp->mii_mutex); - rc = mii_ethtool_sset(&sp->mii_if, cmd); - mutex_unlock(&sp->mii_mutex); - - return rc; -} - -static int ipg_nway_reset(struct net_device *dev) -{ - struct ipg_nic_private *sp = netdev_priv(dev); - int rc; - - mutex_lock(&sp->mii_mutex); - rc = mii_nway_restart(&sp->mii_if); - mutex_unlock(&sp->mii_mutex); - - return rc; -} - -static const struct ethtool_ops ipg_ethtool_ops = { - .get_settings = ipg_get_settings, - .set_settings = ipg_set_settings, - .nway_reset = ipg_nway_reset, -}; - -static void ipg_remove(struct pci_dev *pdev) -{ - struct net_device *dev = pci_get_drvdata(pdev); - struct ipg_nic_private *sp = netdev_priv(dev); - - IPG_DEBUG_MSG("_remove\n"); - - /* Un-register Ethernet device. */ - unregister_netdev(dev); - - pci_iounmap(pdev, sp->ioaddr); - - pci_release_regions(pdev); - - free_netdev(dev); - pci_disable_device(pdev); -} - -static const struct net_device_ops ipg_netdev_ops = { - .ndo_open = ipg_nic_open, - .ndo_stop = ipg_nic_stop, - .ndo_start_xmit = ipg_nic_hard_start_xmit, - .ndo_get_stats = ipg_nic_get_stats, - .ndo_set_rx_mode = ipg_nic_set_multicast_list, - .ndo_do_ioctl = ipg_ioctl, - .ndo_tx_timeout = ipg_tx_timeout, - .ndo_change_mtu = ipg_nic_change_mtu, - .ndo_set_mac_address = eth_mac_addr, - .ndo_validate_addr = eth_validate_addr, -}; - -static int ipg_probe(struct pci_dev *pdev, const struct pci_device_id *id) -{ - unsigned int i = id->driver_data; - struct ipg_nic_private *sp; - struct net_device *dev; - void __iomem *ioaddr; - int rc; - - rc = pci_enable_device(pdev); - if (rc < 0) - goto out; - - pr_info("%s: %s\n", pci_name(pdev), ipg_brand_name[i]); - - pci_set_master(pdev); - - rc = pci_set_dma_mask(pdev, DMA_BIT_MASK(40)); - if (rc < 0) { - rc = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); - if (rc < 0) { - pr_err("%s: DMA config failed\n", pci_name(pdev)); - goto err_disable_0; - } - } - - /* - * Initialize net device. - */ - dev = alloc_etherdev(sizeof(struct ipg_nic_private)); - if (!dev) { - rc = -ENOMEM; - goto err_disable_0; - } - - sp = netdev_priv(dev); - spin_lock_init(&sp->lock); - mutex_init(&sp->mii_mutex); - - sp->is_jumbo = IPG_IS_JUMBO; - sp->rxfrag_size = IPG_RXFRAG_SIZE; - sp->rxsupport_size = IPG_RXSUPPORT_SIZE; - sp->max_rxframe_size = IPG_MAX_RXFRAME_SIZE; - - /* Declare IPG NIC functions for Ethernet device methods. - */ - dev->netdev_ops = &ipg_netdev_ops; - SET_NETDEV_DEV(dev, &pdev->dev); - dev->ethtool_ops = &ipg_ethtool_ops; - - rc = pci_request_regions(pdev, DRV_NAME); - if (rc) - goto err_free_dev_1; - - ioaddr = pci_iomap(pdev, 1, pci_resource_len(pdev, 1)); - if (!ioaddr) { - pr_err("%s: cannot map MMIO\n", pci_name(pdev)); - rc = -EIO; - goto err_release_regions_2; - } - - /* Save the pointer to the PCI device information. */ - sp->ioaddr = ioaddr; - sp->pdev = pdev; - sp->dev = dev; - - INIT_DELAYED_WORK(&sp->task, ipg_reset_after_host_error); - - pci_set_drvdata(pdev, dev); - - rc = ipg_hw_init(dev); - if (rc < 0) - goto err_unmap_3; - - rc = register_netdev(dev); - if (rc < 0) - goto err_unmap_3; - - netdev_info(dev, "Ethernet device registered\n"); -out: - return rc; - -err_unmap_3: - pci_iounmap(pdev, ioaddr); -err_release_regions_2: - pci_release_regions(pdev); -err_free_dev_1: - free_netdev(dev); -err_disable_0: - pci_disable_device(pdev); - goto out; -} - -static struct pci_driver ipg_pci_driver = { - .name = IPG_DRIVER_NAME, - .id_table = ipg_pci_tbl, - .probe = ipg_probe, - .remove = ipg_remove, -}; - -module_pci_driver(ipg_pci_driver); diff --git a/drivers/net/ethernet/icplus/ipg.h b/drivers/net/ethernet/icplus/ipg.h deleted file mode 100644 index de606281f97b..000000000000 --- a/drivers/net/ethernet/icplus/ipg.h +++ /dev/null @@ -1,748 +0,0 @@ -/* - * Include file for Gigabit Ethernet device driver for Network - * Interface Cards (NICs) utilizing the Tamarack Microelectronics - * Inc. IPG Gigabit or Triple Speed Ethernet Media Access - * Controller. - */ -#ifndef __LINUX_IPG_H -#define __LINUX_IPG_H - -#include - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -/* - * Constants - */ - -/* GMII based PHY IDs */ -#define NS 0x2000 -#define MARVELL 0x0141 -#define ICPLUS_PHY 0x243 - -/* NIC Physical Layer Device MII register fields. */ -#define MII_PHY_SELECTOR_IEEE8023 0x0001 -#define MII_PHY_TECHABILITYFIELD 0x1FE0 - -/* GMII_PHY_1000 need to set to prefer master */ -#define GMII_PHY_1000BASETCONTROL_PreferMaster 0x0400 - -/* NIC Physical Layer Device GMII constants. */ -#define GMII_PREAMBLE 0xFFFFFFFF -#define GMII_ST 0x1 -#define GMII_READ 0x2 -#define GMII_WRITE 0x1 -#define GMII_TA_READ_MASK 0x1 -#define GMII_TA_WRITE 0x2 - -/* I/O register offsets. */ -enum ipg_regs { - DMA_CTRL = 0x00, - RX_DMA_STATUS = 0x08, /* Unused + reserved */ - TFD_LIST_PTR_0 = 0x10, - TFD_LIST_PTR_1 = 0x14, - TX_DMA_BURST_THRESH = 0x18, - TX_DMA_URGENT_THRESH = 0x19, - TX_DMA_POLL_PERIOD = 0x1a, - RFD_LIST_PTR_0 = 0x1c, - RFD_LIST_PTR_1 = 0x20, - RX_DMA_BURST_THRESH = 0x24, - RX_DMA_URGENT_THRESH = 0x25, - RX_DMA_POLL_PERIOD = 0x26, - DEBUG_CTRL = 0x2c, - ASIC_CTRL = 0x30, - FIFO_CTRL = 0x38, /* Unused */ - FLOW_OFF_THRESH = 0x3c, - FLOW_ON_THRESH = 0x3e, - EEPROM_DATA = 0x48, - EEPROM_CTRL = 0x4a, - EXPROM_ADDR = 0x4c, /* Unused */ - EXPROM_DATA = 0x50, /* Unused */ - WAKE_EVENT = 0x51, /* Unused */ - COUNTDOWN = 0x54, /* Unused */ - INT_STATUS_ACK = 0x5a, - INT_ENABLE = 0x5c, - INT_STATUS = 0x5e, /* Unused */ - TX_STATUS = 0x60, - MAC_CTRL = 0x6c, - VLAN_TAG = 0x70, /* Unused */ - PHY_SET = 0x75, - PHY_CTRL = 0x76, - STATION_ADDRESS_0 = 0x78, - STATION_ADDRESS_1 = 0x7a, - STATION_ADDRESS_2 = 0x7c, - MAX_FRAME_SIZE = 0x86, - RECEIVE_MODE = 0x88, - HASHTABLE_0 = 0x8c, - HASHTABLE_1 = 0x90, - RMON_STATISTICS_MASK = 0x98, - STATISTICS_MASK = 0x9c, - RX_JUMBO_FRAMES = 0xbc, /* Unused */ - TCP_CHECKSUM_ERRORS = 0xc0, /* Unused */ - IP_CHECKSUM_ERRORS = 0xc2, /* Unused */ - UDP_CHECKSUM_ERRORS = 0xc4, /* Unused */ - TX_JUMBO_FRAMES = 0xf4 /* Unused */ -}; - -/* Ethernet MIB statistic register offsets. */ -#define IPG_OCTETRCVOK 0xA8 -#define IPG_MCSTOCTETRCVDOK 0xAC -#define IPG_BCSTOCTETRCVOK 0xB0 -#define IPG_FRAMESRCVDOK 0xB4 -#define IPG_MCSTFRAMESRCVDOK 0xB8 -#define IPG_BCSTFRAMESRCVDOK 0xBE -#define IPG_MACCONTROLFRAMESRCVD 0xC6 -#define IPG_FRAMETOOLONGERRORS 0xC8 -#define IPG_INRANGELENGTHERRORS 0xCA -#define IPG_FRAMECHECKSEQERRORS 0xCC -#define IPG_FRAMESLOSTRXERRORS 0xCE -#define IPG_OCTETXMTOK 0xD0 -#define IPG_MCSTOCTETXMTOK 0xD4 -#define IPG_BCSTOCTETXMTOK 0xD8 -#define IPG_FRAMESXMTDOK 0xDC -#define IPG_MCSTFRAMESXMTDOK 0xE0 -#define IPG_FRAMESWDEFERREDXMT 0xE4 -#define IPG_LATECOLLISIONS 0xE8 -#define IPG_MULTICOLFRAMES 0xEC -#define IPG_SINGLECOLFRAMES 0xF0 -#define IPG_BCSTFRAMESXMTDOK 0xF6 -#define IPG_CARRIERSENSEERRORS 0xF8 -#define IPG_MACCONTROLFRAMESXMTDOK 0xFA -#define IPG_FRAMESABORTXSCOLLS 0xFC -#define IPG_FRAMESWEXDEFERRAL 0xFE - -/* RMON statistic register offsets. */ -#define IPG_ETHERSTATSCOLLISIONS 0x100 -#define IPG_ETHERSTATSOCTETSTRANSMIT 0x104 -#define IPG_ETHERSTATSPKTSTRANSMIT 0x108 -#define IPG_ETHERSTATSPKTS64OCTESTSTRANSMIT 0x10C -#define IPG_ETHERSTATSPKTS65TO127OCTESTSTRANSMIT 0x110 -#define IPG_ETHERSTATSPKTS128TO255OCTESTSTRANSMIT 0x114 -#define IPG_ETHERSTATSPKTS256TO511OCTESTSTRANSMIT 0x118 -#define IPG_ETHERSTATSPKTS512TO1023OCTESTSTRANSMIT 0x11C -#define IPG_ETHERSTATSPKTS1024TO1518OCTESTSTRANSMIT 0x120 -#define IPG_ETHERSTATSCRCALIGNERRORS 0x124 -#define IPG_ETHERSTATSUNDERSIZEPKTS 0x128 -#define IPG_ETHERSTATSFRAGMENTS 0x12C -#define IPG_ETHERSTATSJABBERS 0x130 -#define IPG_ETHERSTATSOCTETS 0x134 -#define IPG_ETHERSTATSPKTS 0x138 -#define IPG_ETHERSTATSPKTS64OCTESTS 0x13C -#define IPG_ETHERSTATSPKTS65TO127OCTESTS 0x140 -#define IPG_ETHERSTATSPKTS128TO255OCTESTS 0x144 -#define IPG_ETHERSTATSPKTS256TO511OCTESTS 0x148 -#define IPG_ETHERSTATSPKTS512TO1023OCTESTS 0x14C -#define IPG_ETHERSTATSPKTS1024TO1518OCTESTS 0x150 - -/* RMON statistic register equivalents. */ -#define IPG_ETHERSTATSMULTICASTPKTSTRANSMIT 0xE0 -#define IPG_ETHERSTATSBROADCASTPKTSTRANSMIT 0xF6 -#define IPG_ETHERSTATSMULTICASTPKTS 0xB8 -#define IPG_ETHERSTATSBROADCASTPKTS 0xBE -#define IPG_ETHERSTATSOVERSIZEPKTS 0xC8 -#define IPG_ETHERSTATSDROPEVENTS 0xCE - -/* Serial EEPROM offsets */ -#define IPG_EEPROM_CONFIGPARAM 0x00 -#define IPG_EEPROM_ASICCTRL 0x01 -#define IPG_EEPROM_SUBSYSTEMVENDORID 0x02 -#define IPG_EEPROM_SUBSYSTEMID 0x03 -#define IPG_EEPROM_STATIONADDRESS0 0x10 -#define IPG_EEPROM_STATIONADDRESS1 0x11 -#define IPG_EEPROM_STATIONADDRESS2 0x12 - -/* Register & data structure bit masks */ - -/* PCI register masks. */ - -/* IOBaseAddress */ -#define IPG_PIB_RSVD_MASK 0xFFFFFE01 -#define IPG_PIB_IOBASEADDRESS 0xFFFFFF00 -#define IPG_PIB_IOBASEADDRIND 0x00000001 - -/* MemBaseAddress */ -#define IPG_PMB_RSVD_MASK 0xFFFFFE07 -#define IPG_PMB_MEMBASEADDRIND 0x00000001 -#define IPG_PMB_MEMMAPTYPE 0x00000006 -#define IPG_PMB_MEMMAPTYPE0 0x00000002 -#define IPG_PMB_MEMMAPTYPE1 0x00000004 -#define IPG_PMB_MEMBASEADDRESS 0xFFFFFE00 - -/* ConfigStatus */ -#define IPG_CS_RSVD_MASK 0xFFB0 -#define IPG_CS_CAPABILITIES 0x0010 -#define IPG_CS_66MHZCAPABLE 0x0020 -#define IPG_CS_FASTBACK2BACK 0x0080 -#define IPG_CS_DATAPARITYREPORTED 0x0100 -#define IPG_CS_DEVSELTIMING 0x0600 -#define IPG_CS_SIGNALEDTARGETABORT 0x0800 -#define IPG_CS_RECEIVEDTARGETABORT 0x1000 -#define IPG_CS_RECEIVEDMASTERABORT 0x2000 -#define IPG_CS_SIGNALEDSYSTEMERROR 0x4000 -#define IPG_CS_DETECTEDPARITYERROR 0x8000 - -/* TFD data structure masks. */ - -/* TFDList, TFC */ -#define IPG_TFC_RSVD_MASK 0x0000FFFF9FFFFFFFULL -#define IPG_TFC_FRAMEID 0x000000000000FFFFULL -#define IPG_TFC_WORDALIGN 0x0000000000030000ULL -#define IPG_TFC_WORDALIGNTODWORD 0x0000000000000000ULL -#define IPG_TFC_WORDALIGNTOWORD 0x0000000000020000ULL -#define IPG_TFC_WORDALIGNDISABLED 0x0000000000030000ULL -#define IPG_TFC_TCPCHECKSUMENABLE 0x0000000000040000ULL -#define IPG_TFC_UDPCHECKSUMENABLE 0x0000000000080000ULL -#define IPG_TFC_IPCHECKSUMENABLE 0x0000000000100000ULL -#define IPG_TFC_FCSAPPENDDISABLE 0x0000000000200000ULL -#define IPG_TFC_TXINDICATE 0x0000000000400000ULL -#define IPG_TFC_TXDMAINDICATE 0x0000000000800000ULL -#define IPG_TFC_FRAGCOUNT 0x000000000F000000ULL -#define IPG_TFC_VLANTAGINSERT 0x0000000010000000ULL -#define IPG_TFC_TFDDONE 0x0000000080000000ULL -#define IPG_TFC_VID 0x00000FFF00000000ULL -#define IPG_TFC_CFI 0x0000100000000000ULL -#define IPG_TFC_USERPRIORITY 0x0000E00000000000ULL - -/* TFDList, FragInfo */ -#define IPG_TFI_RSVD_MASK 0xFFFF00FFFFFFFFFFULL -#define IPG_TFI_FRAGADDR 0x000000FFFFFFFFFFULL -#define IPG_TFI_FRAGLEN 0xFFFF000000000000ULL - -/* RFD data structure masks. */ - -/* RFDList, RFS */ -#define IPG_RFS_RSVD_MASK 0x0000FFFFFFFFFFFFULL -#define IPG_RFS_RXFRAMELEN 0x000000000000FFFFULL -#define IPG_RFS_RXFIFOOVERRUN 0x0000000000010000ULL -#define IPG_RFS_RXRUNTFRAME 0x0000000000020000ULL -#define IPG_RFS_RXALIGNMENTERROR 0x0000000000040000ULL -#define IPG_RFS_RXFCSERROR 0x0000000000080000ULL -#define IPG_RFS_RXOVERSIZEDFRAME 0x0000000000100000ULL -#define IPG_RFS_RXLENGTHERROR 0x0000000000200000ULL -#define IPG_RFS_VLANDETECTED 0x0000000000400000ULL -#define IPG_RFS_TCPDETECTED 0x0000000000800000ULL -#define IPG_RFS_TCPERROR 0x0000000001000000ULL -#define IPG_RFS_UDPDETECTED 0x0000000002000000ULL -#define IPG_RFS_UDPERROR 0x0000000004000000ULL -#define IPG_RFS_IPDETECTED 0x0000000008000000ULL -#define IPG_RFS_IPERROR 0x0000000010000000ULL -#define IPG_RFS_FRAMESTART 0x0000000020000000ULL -#define IPG_RFS_FRAMEEND 0x0000000040000000ULL -#define IPG_RFS_RFDDONE 0x0000000080000000ULL -#define IPG_RFS_TCI 0x0000FFFF00000000ULL - -/* RFDList, FragInfo */ -#define IPG_RFI_RSVD_MASK 0xFFFF00FFFFFFFFFFULL -#define IPG_RFI_FRAGADDR 0x000000FFFFFFFFFFULL -#define IPG_RFI_FRAGLEN 0xFFFF000000000000ULL - -/* I/O Register masks. */ - -/* RMON Statistics Mask */ -#define IPG_RZ_ALL 0x0FFFFFFF - -/* Statistics Mask */ -#define IPG_SM_ALL 0x0FFFFFFF -#define IPG_SM_OCTETRCVOK_FRAMESRCVDOK 0x00000001 -#define IPG_SM_MCSTOCTETRCVDOK_MCSTFRAMESRCVDOK 0x00000002 -#define IPG_SM_BCSTOCTETRCVDOK_BCSTFRAMESRCVDOK 0x00000004 -#define IPG_SM_RXJUMBOFRAMES 0x00000008 -#define IPG_SM_TCPCHECKSUMERRORS 0x00000010 -#define IPG_SM_IPCHECKSUMERRORS 0x00000020 -#define IPG_SM_UDPCHECKSUMERRORS 0x00000040 -#define IPG_SM_MACCONTROLFRAMESRCVD 0x00000080 -#define IPG_SM_FRAMESTOOLONGERRORS 0x00000100 -#define IPG_SM_INRANGELENGTHERRORS 0x00000200 -#define IPG_SM_FRAMECHECKSEQERRORS 0x00000400 -#define IPG_SM_FRAMESLOSTRXERRORS 0x00000800 -#define IPG_SM_OCTETXMTOK_FRAMESXMTOK 0x00001000 -#define IPG_SM_MCSTOCTETXMTOK_MCSTFRAMESXMTDOK 0x00002000 -#define IPG_SM_BCSTOCTETXMTOK_BCSTFRAMESXMTDOK 0x00004000 -#define IPG_SM_FRAMESWDEFERREDXMT 0x00008000 -#define IPG_SM_LATECOLLISIONS 0x00010000 -#define IPG_SM_MULTICOLFRAMES 0x00020000 -#define IPG_SM_SINGLECOLFRAMES 0x00040000 -#define IPG_SM_TXJUMBOFRAMES 0x00080000 -#define IPG_SM_CARRIERSENSEERRORS 0x00100000 -#define IPG_SM_MACCONTROLFRAMESXMTD 0x00200000 -#define IPG_SM_FRAMESABORTXSCOLLS 0x00400000 -#define IPG_SM_FRAMESWEXDEFERAL 0x00800000 - -/* Countdown */ -#define IPG_CD_RSVD_MASK 0x0700FFFF -#define IPG_CD_COUNT 0x0000FFFF -#define IPG_CD_COUNTDOWNSPEED 0x01000000 -#define IPG_CD_COUNTDOWNMODE 0x02000000 -#define IPG_CD_COUNTINTENABLED 0x04000000 - -/* TxDMABurstThresh */ -#define IPG_TB_RSVD_MASK 0xFF - -/* TxDMAUrgentThresh */ -#define IPG_TU_RSVD_MASK 0xFF - -/* TxDMAPollPeriod */ -#define IPG_TP_RSVD_MASK 0xFF - -/* RxDMAUrgentThresh */ -#define IPG_RU_RSVD_MASK 0xFF - -/* RxDMAPollPeriod */ -#define IPG_RP_RSVD_MASK 0xFF - -/* ReceiveMode */ -#define IPG_RM_RSVD_MASK 0x3F -#define IPG_RM_RECEIVEUNICAST 0x01 -#define IPG_RM_RECEIVEMULTICAST 0x02 -#define IPG_RM_RECEIVEBROADCAST 0x04 -#define IPG_RM_RECEIVEALLFRAMES 0x08 -#define IPG_RM_RECEIVEMULTICASTHASH 0x10 -#define IPG_RM_RECEIVEIPMULTICAST 0x20 - -/* PhySet */ -#define IPG_PS_MEM_LENB9B 0x01 -#define IPG_PS_MEM_LEN9 0x02 -#define IPG_PS_NON_COMPDET 0x04 - -/* PhyCtrl */ -#define IPG_PC_RSVD_MASK 0xFF -#define IPG_PC_MGMTCLK_LO 0x00 -#define IPG_PC_MGMTCLK_HI 0x01 -#define IPG_PC_MGMTCLK 0x01 -#define IPG_PC_MGMTDATA 0x02 -#define IPG_PC_MGMTDIR 0x04 -#define IPG_PC_DUPLEX_POLARITY 0x08 -#define IPG_PC_DUPLEX_STATUS 0x10 -#define IPG_PC_LINK_POLARITY 0x20 -#define IPG_PC_LINK_SPEED 0xC0 -#define IPG_PC_LINK_SPEED_10MBPS 0x40 -#define IPG_PC_LINK_SPEED_100MBPS 0x80 -#define IPG_PC_LINK_SPEED_1000MBPS 0xC0 - -/* DMACtrl */ -#define IPG_DC_RSVD_MASK 0xC07D9818 -#define IPG_DC_RX_DMA_COMPLETE 0x00000008 -#define IPG_DC_RX_DMA_POLL_NOW 0x00000010 -#define IPG_DC_TX_DMA_COMPLETE 0x00000800 -#define IPG_DC_TX_DMA_POLL_NOW 0x00001000 -#define IPG_DC_TX_DMA_IN_PROG 0x00008000 -#define IPG_DC_RX_EARLY_DISABLE 0x00010000 -#define IPG_DC_MWI_DISABLE 0x00040000 -#define IPG_DC_TX_WRITE_BACK_DISABLE 0x00080000 -#define IPG_DC_TX_BURST_LIMIT 0x00700000 -#define IPG_DC_TARGET_ABORT 0x40000000 -#define IPG_DC_MASTER_ABORT 0x80000000 - -/* ASICCtrl */ -#define IPG_AC_RSVD_MASK 0x07FFEFF2 -#define IPG_AC_EXP_ROM_SIZE 0x00000002 -#define IPG_AC_PHY_SPEED10 0x00000010 -#define IPG_AC_PHY_SPEED100 0x00000020 -#define IPG_AC_PHY_SPEED1000 0x00000040 -#define IPG_AC_PHY_MEDIA 0x00000080 -#define IPG_AC_FORCED_CFG 0x00000700 -#define IPG_AC_D3RESETDISABLE 0x00000800 -#define IPG_AC_SPEED_UP_MODE 0x00002000 -#define IPG_AC_LED_MODE 0x00004000 -#define IPG_AC_RST_OUT_POLARITY 0x00008000 -#define IPG_AC_GLOBAL_RESET 0x00010000 -#define IPG_AC_RX_RESET 0x00020000 -#define IPG_AC_TX_RESET 0x00040000 -#define IPG_AC_DMA 0x00080000 -#define IPG_AC_FIFO 0x00100000 -#define IPG_AC_NETWORK 0x00200000 -#define IPG_AC_HOST 0x00400000 -#define IPG_AC_AUTO_INIT 0x00800000 -#define IPG_AC_RST_OUT 0x01000000 -#define IPG_AC_INT_REQUEST 0x02000000 -#define IPG_AC_RESET_BUSY 0x04000000 -#define IPG_AC_LED_SPEED 0x08000000 -#define IPG_AC_LED_MODE_BIT_1 0x20000000 - -/* EepromCtrl */ -#define IPG_EC_RSVD_MASK 0x83FF -#define IPG_EC_EEPROM_ADDR 0x00FF -#define IPG_EC_EEPROM_OPCODE 0x0300 -#define IPG_EC_EEPROM_SUBCOMMAD 0x0000 -#define IPG_EC_EEPROM_WRITEOPCODE 0x0100 -#define IPG_EC_EEPROM_READOPCODE 0x0200 -#define IPG_EC_EEPROM_ERASEOPCODE 0x0300 -#define IPG_EC_EEPROM_BUSY 0x8000 - -/* FIFOCtrl */ -#define IPG_FC_RSVD_MASK 0xC001 -#define IPG_FC_RAM_TEST_MODE 0x0001 -#define IPG_FC_TRANSMITTING 0x4000 -#define IPG_FC_RECEIVING 0x8000 - -/* TxStatus */ -#define IPG_TS_RSVD_MASK 0xFFFF00DD -#define IPG_TS_TX_ERROR 0x00000001 -#define IPG_TS_LATE_COLLISION 0x00000004 -#define IPG_TS_TX_MAX_COLL 0x00000008 -#define IPG_TS_TX_UNDERRUN 0x00000010 -#define IPG_TS_TX_IND_REQD 0x00000040 -#define IPG_TS_TX_COMPLETE 0x00000080 -#define IPG_TS_TX_FRAMEID 0xFFFF0000 - -/* WakeEvent */ -#define IPG_WE_WAKE_PKT_ENABLE 0x01 -#define IPG_WE_MAGIC_PKT_ENABLE 0x02 -#define IPG_WE_LINK_EVT_ENABLE 0x04 -#define IPG_WE_WAKE_POLARITY 0x08 -#define IPG_WE_WAKE_PKT_EVT 0x10 -#define IPG_WE_MAGIC_PKT_EVT 0x20 -#define IPG_WE_LINK_EVT 0x40 -#define IPG_WE_WOL_ENABLE 0x80 - -/* IntEnable */ -#define IPG_IE_RSVD_MASK 0x1FFE -#define IPG_IE_HOST_ERROR 0x0002 -#define IPG_IE_TX_COMPLETE 0x0004 -#define IPG_IE_MAC_CTRL_FRAME 0x0008 -#define IPG_IE_RX_COMPLETE 0x0010 -#define IPG_IE_RX_EARLY 0x0020 -#define IPG_IE_INT_REQUESTED 0x0040 -#define IPG_IE_UPDATE_STATS 0x0080 -#define IPG_IE_LINK_EVENT 0x0100 -#define IPG_IE_TX_DMA_COMPLETE 0x0200 -#define IPG_IE_RX_DMA_COMPLETE 0x0400 -#define IPG_IE_RFD_LIST_END 0x0800 -#define IPG_IE_RX_DMA_PRIORITY 0x1000 - -/* IntStatus */ -#define IPG_IS_RSVD_MASK 0x1FFF -#define IPG_IS_INTERRUPT_STATUS 0x0001 -#define IPG_IS_HOST_ERROR 0x0002 -#define IPG_IS_TX_COMPLETE 0x0004 -#define IPG_IS_MAC_CTRL_FRAME 0x0008 -#define IPG_IS_RX_COMPLETE 0x0010 -#define IPG_IS_RX_EARLY 0x0020 -#define IPG_IS_INT_REQUESTED 0x0040 -#define IPG_IS_UPDATE_STATS 0x0080 -#define IPG_IS_LINK_EVENT 0x0100 -#define IPG_IS_TX_DMA_COMPLETE 0x0200 -#define IPG_IS_RX_DMA_COMPLETE 0x0400 -#define IPG_IS_RFD_LIST_END 0x0800 -#define IPG_IS_RX_DMA_PRIORITY 0x1000 - -/* MACCtrl */ -#define IPG_MC_RSVD_MASK 0x7FE33FA3 -#define IPG_MC_IFS_SELECT 0x00000003 -#define IPG_MC_IFS_4352BIT 0x00000003 -#define IPG_MC_IFS_1792BIT 0x00000002 -#define IPG_MC_IFS_1024BIT 0x00000001 -#define IPG_MC_IFS_96BIT 0x00000000 -#define IPG_MC_DUPLEX_SELECT 0x00000020 -#define IPG_MC_DUPLEX_SELECT_FD 0x00000020 -#define IPG_MC_DUPLEX_SELECT_HD 0x00000000 -#define IPG_MC_TX_FLOW_CONTROL_ENABLE 0x00000080 -#define IPG_MC_RX_FLOW_CONTROL_ENABLE 0x00000100 -#define IPG_MC_RCV_FCS 0x00000200 -#define IPG_MC_FIFO_LOOPBACK 0x00000400 -#define IPG_MC_MAC_LOOPBACK 0x00000800 -#define IPG_MC_AUTO_VLAN_TAGGING 0x00001000 -#define IPG_MC_AUTO_VLAN_UNTAGGING 0x00002000 -#define IPG_MC_COLLISION_DETECT 0x00010000 -#define IPG_MC_CARRIER_SENSE 0x00020000 -#define IPG_MC_STATISTICS_ENABLE 0x00200000 -#define IPG_MC_STATISTICS_DISABLE 0x00400000 -#define IPG_MC_STATISTICS_ENABLED 0x00800000 -#define IPG_MC_TX_ENABLE 0x01000000 -#define IPG_MC_TX_DISABLE 0x02000000 -#define IPG_MC_TX_ENABLED 0x04000000 -#define IPG_MC_RX_ENABLE 0x08000000 -#define IPG_MC_RX_DISABLE 0x10000000 -#define IPG_MC_RX_ENABLED 0x20000000 -#define IPG_MC_PAUSED 0x40000000 - -/* - * Tune - */ - -/* Assign IPG_APPEND_FCS_ON_TX > 0 for auto FCS append on TX. */ -#define IPG_APPEND_FCS_ON_TX 1 - -/* Assign IPG_APPEND_FCS_ON_TX > 0 for auto FCS strip on RX. */ -#define IPG_STRIP_FCS_ON_RX 1 - -/* Assign IPG_DROP_ON_RX_ETH_ERRORS > 0 to drop RX frames with - * Ethernet errors. - */ -#define IPG_DROP_ON_RX_ETH_ERRORS 1 - -/* Assign IPG_INSERT_MANUAL_VLAN_TAG > 0 to insert VLAN tags manually - * (via TFC). - */ -#define IPG_INSERT_MANUAL_VLAN_TAG 0 - -/* Assign IPG_ADD_IPCHECKSUM_ON_TX > 0 for auto IP checksum on TX. */ -#define IPG_ADD_IPCHECKSUM_ON_TX 0 - -/* Assign IPG_ADD_TCPCHECKSUM_ON_TX > 0 for auto TCP checksum on TX. - * DO NOT USE FOR SILICON REVISIONS B3 AND EARLIER. - */ -#define IPG_ADD_TCPCHECKSUM_ON_TX 0 - -/* Assign IPG_ADD_UDPCHECKSUM_ON_TX > 0 for auto UDP checksum on TX. - * DO NOT USE FOR SILICON REVISIONS B3 AND EARLIER. - */ -#define IPG_ADD_UDPCHECKSUM_ON_TX 0 - -/* If inserting VLAN tags manually, assign the IPG_MANUAL_VLAN_xx - * constants as desired. - */ -#define IPG_MANUAL_VLAN_VID 0xABC -#define IPG_MANUAL_VLAN_CFI 0x1 -#define IPG_MANUAL_VLAN_USERPRIORITY 0x5 - -#define IPG_IO_REG_RANGE 0xFF -#define IPG_MEM_REG_RANGE 0x154 -#define IPG_DRIVER_NAME "Sundance Technology IPG Triple-Speed Ethernet" -#define IPG_NIC_PHY_ADDRESS 0x01 -#define IPG_DMALIST_ALIGN_PAD 0x07 -#define IPG_MULTICAST_HASHTABLE_SIZE 0x40 - -/* Number of milliseconds to wait after issuing a software reset. - * 0x05 <= IPG_AC_RESETWAIT to account for proper 10Mbps operation. - */ -#define IPG_AC_RESETWAIT 0x05 - -/* Number of IPG_AC_RESETWAIT timeperiods before declaring timeout. */ -#define IPG_AC_RESET_TIMEOUT 0x0A - -/* Minimum number of nanoseconds used to toggle MDC clock during - * MII/GMII register access. - */ -#define IPG_PC_PHYCTRLWAIT_NS 200 - -#define IPG_TFDLIST_LENGTH 0x100 - -/* Number of frames between TxDMAComplete interrupt. - * 0 < IPG_FRAMESBETWEENTXDMACOMPLETES <= IPG_TFDLIST_LENGTH - */ -#define IPG_FRAMESBETWEENTXDMACOMPLETES 0x1 - -#define IPG_RFDLIST_LENGTH 0x100 - -/* Maximum number of RFDs to process per interrupt. - * 1 < IPG_MAXRFDPROCESS_COUNT < IPG_RFDLIST_LENGTH - */ -#define IPG_MAXRFDPROCESS_COUNT 0x80 - -/* Minimum margin between last freed RFD, and current RFD. - * 1 < IPG_MINUSEDRFDSTOFREE < IPG_RFDLIST_LENGTH - */ -#define IPG_MINUSEDRFDSTOFREE 0x80 - -/* specify the jumbo frame maximum size - * per unit is 0x600 (the rx_buffer size that one RFD can carry) - */ -#define MAX_JUMBOSIZE 0x8 /* max is 12K */ - -/* Key register values loaded at driver start up. */ - -/* TXDMAPollPeriod is specified in 320ns increments. - * - * Value Time - * --------------------- - * 0x00-0x01 320ns - * 0x03 ~1us - * 0x1F ~10us - * 0xFF ~82us - */ -#define IPG_TXDMAPOLLPERIOD_VALUE 0x26 - -/* TxDMAUrgentThresh specifies the minimum amount of - * data in the transmit FIFO before asserting an - * urgent transmit DMA request. - * - * Value Min TxFIFO occupied space before urgent TX request - * --------------------------------------------------------------- - * 0x00-0x04 128 bytes (1024 bits) - * 0x27 1248 bytes (~10000 bits) - * 0x30 1536 bytes (12288 bits) - * 0xFF 8192 bytes (65535 bits) - */ -#define IPG_TXDMAURGENTTHRESH_VALUE 0x04 - -/* TxDMABurstThresh specifies the minimum amount of - * free space in the transmit FIFO before asserting an - * transmit DMA request. - * - * Value Min TxFIFO free space before TX request - * ---------------------------------------------------- - * 0x00-0x08 256 bytes - * 0x30 1536 bytes - * 0xFF 8192 bytes - */ -#define IPG_TXDMABURSTTHRESH_VALUE 0x30 - -/* RXDMAPollPeriod is specified in 320ns increments. - * - * Value Time - * --------------------- - * 0x00-0x01 320ns - * 0x03 ~1us - * 0x1F ~10us - * 0xFF ~82us - */ -#define IPG_RXDMAPOLLPERIOD_VALUE 0x01 - -/* RxDMAUrgentThresh specifies the minimum amount of - * free space within the receive FIFO before asserting - * a urgent receive DMA request. - * - * Value Min RxFIFO free space before urgent RX request - * --------------------------------------------------------------- - * 0x00-0x04 128 bytes (1024 bits) - * 0x27 1248 bytes (~10000 bits) - * 0x30 1536 bytes (12288 bits) - * 0xFF 8192 bytes (65535 bits) - */ -#define IPG_RXDMAURGENTTHRESH_VALUE 0x30 - -/* RxDMABurstThresh specifies the minimum amount of - * occupied space within the receive FIFO before asserting - * a receive DMA request. - * - * Value Min TxFIFO free space before TX request - * ---------------------------------------------------- - * 0x00-0x08 256 bytes - * 0x30 1536 bytes - * 0xFF 8192 bytes - */ -#define IPG_RXDMABURSTTHRESH_VALUE 0x30 - -/* FlowOnThresh specifies the maximum amount of occupied - * space in the receive FIFO before a PAUSE frame with - * maximum pause time transmitted. - * - * Value Max RxFIFO occupied space before PAUSE - * --------------------------------------------------- - * 0x0000 0 bytes - * 0x0740 29,696 bytes - * 0x07FF 32,752 bytes - */ -#define IPG_FLOWONTHRESH_VALUE 0x0740 - -/* FlowOffThresh specifies the minimum amount of occupied - * space in the receive FIFO before a PAUSE frame with - * zero pause time is transmitted. - * - * Value Max RxFIFO occupied space before PAUSE - * --------------------------------------------------- - * 0x0000 0 bytes - * 0x00BF 3056 bytes - * 0x07FF 32,752 bytes - */ -#define IPG_FLOWOFFTHRESH_VALUE 0x00BF - -/* - * Miscellaneous macros. - */ - -/* Macros for printing debug statements. */ -#ifdef IPG_DEBUG -# define IPG_DEBUG_MSG(fmt, args...) \ -do { \ - if (0) \ - printk(KERN_DEBUG "IPG: " fmt, ##args); \ -} while (0) -# define IPG_DDEBUG_MSG(fmt, args...) \ - printk(KERN_DEBUG "IPG: " fmt, ##args) -# define IPG_DUMPRFDLIST(args) ipg_dump_rfdlist(args) -# define IPG_DUMPTFDLIST(args) ipg_dump_tfdlist(args) -#else -# define IPG_DEBUG_MSG(fmt, args...) \ -do { \ - if (0) \ - printk(KERN_DEBUG "IPG: " fmt, ##args); \ -} while (0) -# define IPG_DDEBUG_MSG(fmt, args...) \ -do { \ - if (0) \ - printk(KERN_DEBUG "IPG: " fmt, ##args); \ -} while (0) -# define IPG_DUMPRFDLIST(args) -# define IPG_DUMPTFDLIST(args) -#endif - -/* - * End miscellaneous macros. - */ - -/* Transmit Frame Descriptor. The IPG supports 15 fragments, - * however Linux requires only a single fragment. Note, each - * TFD field is 64 bits wide. - */ -struct ipg_tx { - __le64 next_desc; - __le64 tfc; - __le64 frag_info; -}; - -/* Receive Frame Descriptor. Note, each RFD field is 64 bits wide. - */ -struct ipg_rx { - __le64 next_desc; - __le64 rfs; - __le64 frag_info; -}; - -struct ipg_jumbo { - int found_start; - int current_size; - struct sk_buff *skb; -}; - -/* Structure of IPG NIC specific data. */ -struct ipg_nic_private { - void __iomem *ioaddr; - struct ipg_tx *txd; - struct ipg_rx *rxd; - dma_addr_t txd_map; - dma_addr_t rxd_map; - struct sk_buff *tx_buff[IPG_TFDLIST_LENGTH]; - struct sk_buff *rx_buff[IPG_RFDLIST_LENGTH]; - unsigned int tx_current; - unsigned int tx_dirty; - unsigned int rx_current; - unsigned int rx_dirty; - bool is_jumbo; - struct ipg_jumbo jumbo; - unsigned long rxfrag_size; - unsigned long rxsupport_size; - unsigned long max_rxframe_size; - unsigned int rx_buf_sz; - struct pci_dev *pdev; - struct net_device *dev; - struct net_device_stats stats; - spinlock_t lock; - int tenmbpsmode; - - u16 led_mode; - u16 station_addr[3]; /* Station Address in EEPROM Reg 0x10..0x12 */ - - struct mutex mii_mutex; - struct mii_if_info mii_if; - int reset_current_tfd; -#ifdef IPG_DEBUG - int RFDlistendCount; - int RFDListCheckedCount; - int EmptyRFDListCount; -#endif - struct delayed_work task; -}; - -#endif /* __LINUX_IPG_H */ From add68d6aa9e2492e51707ca603ada5b26c626757 Mon Sep 17 00:00:00 2001 From: Chris Bainbridge Date: Thu, 12 Nov 2015 18:05:37 +0000 Subject: [PATCH 186/335] ACPI / SMBus: Fix boot stalls / high CPU caused by reentrant code In the SBS initialisation, a reentrant call to wait_event_timeout() causes an intermittent boot stall of several minutes usually following the "Switching to clocksource tsc" message. Another symptom of this bug is high CPU usage from programs (Firefox, upowerd) querying the battery state. This is caused by: 1. drivers/acpi/sbshc.c wait_transaction_complete() calls wait_event_timeout(): if (wait_event_timeout(hc->wait, smb_check_done(hc), msecs_to_jiffies(timeout))) 2. ___wait_event sets task state to uninterruptible 3. ___wait_event calls the "condition" smb_check_done() 4. smb_check_done (sbshc.c) calls through to ec_read() in drivers/acpi/ec.c 5. ec_guard() is reached which calls wait_event_timeout() if (wait_event_timeout(ec->wait, ec_transaction_completed(ec), guard)) ie. wait_event_timeout() is being called again inside evaluation of the previous wait_event_timeout() condition 5. The EC IRQ handler calls wake_up() and wakes up the sleeping task in ec_guard() 6. The task is now in state running even though the wait "condition" is still being evaluated 7. The "condition" check returns false so ___wait_event calls schedule_timeout() 8. Since the task state is running, the scheduler immediately schedules it again 9. This loop usually repeats for around 250 seconds even though the original wait_event_timeout was only 1000ms. The timeout is incorrect because each call to schedule_timeout() usually returns immediately, taking less than 1ms, so the jiffies timeout counter is not decremented. The task is now stuck in a running state, and so is highly likely to be immediately rescheduled, which takes less than a jiffy. The loop will never exit if all schedule_timeout() calls take less than a jiffy. Fix this by replacing SMBus reads in the wait_event_timeout condition with checks of a boolean value that is updated by the EC query handler. Link: https://bugzilla.kernel.org/show_bug.cgi?id=107191 Link: https://lkml.org/lkml/2015/11/6/776 Signed-off-by: Chris Bainbridge Signed-off-by: Rafael J. Wysocki --- drivers/acpi/sbshc.c | 26 +++++++------------------- 1 file changed, 7 insertions(+), 19 deletions(-) diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c index bf034f8b7c1a..e2900518cf7e 100644 --- a/drivers/acpi/sbshc.c +++ b/drivers/acpi/sbshc.c @@ -30,6 +30,7 @@ struct acpi_smb_hc { u8 query_bit; smbus_alarm_callback callback; void *context; + bool done; }; static int acpi_smbus_hc_add(struct acpi_device *device); @@ -100,27 +101,11 @@ static inline int smb_hc_write(struct acpi_smb_hc *hc, u8 address, u8 data) return ec_write(hc->offset + address, data); } -static inline int smb_check_done(struct acpi_smb_hc *hc) -{ - union acpi_smb_status status = {.raw = 0}; - smb_hc_read(hc, ACPI_SMB_STATUS, &status.raw); - return status.fields.done && (status.fields.status == SMBUS_OK); -} - static int wait_transaction_complete(struct acpi_smb_hc *hc, int timeout) { - if (wait_event_timeout(hc->wait, smb_check_done(hc), - msecs_to_jiffies(timeout))) + if (wait_event_timeout(hc->wait, hc->done, msecs_to_jiffies(timeout))) return 0; - /* - * After the timeout happens, OS will try to check the status of SMbus. - * If the status is what OS expected, it will be regarded as the bogus - * timeout. - */ - if (smb_check_done(hc)) - return 0; - else - return -ETIME; + return -ETIME; } static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol, @@ -135,6 +120,7 @@ static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol, } mutex_lock(&hc->lock); + hc->done = false; if (macbook) udelay(5); if (smb_hc_read(hc, ACPI_SMB_PROTOCOL, &temp)) @@ -235,8 +221,10 @@ static int smbus_alarm(void *context) if (smb_hc_read(hc, ACPI_SMB_STATUS, &status.raw)) return 0; /* Check if it is only a completion notify */ - if (status.fields.done) + if (status.fields.done && status.fields.status == SMBUS_OK) { + hc->done = true; wake_up(&hc->wait); + } if (!status.fields.alarm) return 0; mutex_lock(&hc->lock); From a76032e0abef027df83f7f053580aa297de54818 Mon Sep 17 00:00:00 2001 From: Chris Bainbridge Date: Thu, 12 Nov 2015 19:24:47 +0000 Subject: [PATCH 187/335] Revert "ACPI / SBS: Add 5 us delay to fix SBS hangs on MacBook" Revert commit 3349fb64b292 (ACPI / SBS: Add 5 us delay to fix SBS hangs on MacBook), since the delay introduced by it is not necessary any more after commit add68d6aa9e2 (ACPI / SMBus: Fix boot stalls / high CPU caused by reentrant code). Signed-off-by: Chris Bainbridge [ rjw: Changelog ] Signed-off-by: Rafael J. Wysocki --- drivers/acpi/sbshc.c | 22 ---------------------- 1 file changed, 22 deletions(-) diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c index e2900518cf7e..2fa8304171e0 100644 --- a/drivers/acpi/sbshc.c +++ b/drivers/acpi/sbshc.c @@ -14,7 +14,6 @@ #include #include #include -#include #include "sbshc.h" #define PREFIX "ACPI: " @@ -89,8 +88,6 @@ enum acpi_smb_offset { ACPI_SMB_ALARM_DATA = 0x26, /* 2 bytes alarm data */ }; -static bool macbook; - static inline int smb_hc_read(struct acpi_smb_hc *hc, u8 address, u8 *data) { return ec_read(hc->offset + address, data); @@ -121,8 +118,6 @@ static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol, mutex_lock(&hc->lock); hc->done = false; - if (macbook) - udelay(5); if (smb_hc_read(hc, ACPI_SMB_PROTOCOL, &temp)) goto end; if (temp) { @@ -250,29 +245,12 @@ extern int acpi_ec_add_query_handler(struct acpi_ec *ec, u8 query_bit, acpi_handle handle, acpi_ec_query_func func, void *data); -static int macbook_dmi_match(const struct dmi_system_id *d) -{ - pr_debug("Detected MacBook, enabling workaround\n"); - macbook = true; - return 0; -} - -static struct dmi_system_id acpi_smbus_dmi_table[] = { - { macbook_dmi_match, "Apple MacBook", { - DMI_MATCH(DMI_BOARD_VENDOR, "Apple"), - DMI_MATCH(DMI_PRODUCT_NAME, "MacBook") }, - }, - { }, -}; - static int acpi_smbus_hc_add(struct acpi_device *device) { int status; unsigned long long val; struct acpi_smb_hc *hc; - dmi_check_system(acpi_smbus_dmi_table); - if (!device) return -EINVAL; From 4981c2b7abfe92a7ad3c9888b8a1ada552d49406 Mon Sep 17 00:00:00 2001 From: Markus Elfring Date: Sun, 15 Nov 2015 22:42:27 +0100 Subject: [PATCH 188/335] ACPI-EC: Drop unnecessary check made before calling acpi_ec_delete_query() The acpi_ec_delete_query() function tests whether its argument is NULL and then returns immediately. Thus the test around the call is not needed. This issue was detected by using the Coccinelle software. Signed-off-by: Markus Elfring [ rjw: Subject ] Signed-off-by: Rafael J. Wysocki --- drivers/acpi/ec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c index f61a7c834540..b420fb46669d 100644 --- a/drivers/acpi/ec.c +++ b/drivers/acpi/ec.c @@ -1103,7 +1103,7 @@ static int acpi_ec_query(struct acpi_ec *ec, u8 *data) } err_exit: - if (result && q) + if (result) acpi_ec_delete_query(q); if (data) *data = value; From a3088abc68f79ac30e6e8ccb07a15c8093e39684 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Tue, 17 Nov 2015 00:24:14 +0100 Subject: [PATCH 189/335] HID: wacom: fixup quirks setup for WACOM_DEVICETYPE_PAD Given that INTUOSHT < BAMBOO_PT features->type >= INTUOSHT || features->type <= BAMBOO_PT condition is always true, and therefore device_type is under certain circumstances wrongly set with WACOM_DEVICETYPE_PAD bit set. Fix the condition so that it actually represents the range as intended. Reported-by: David Binderman Signed-off-by: Jiri Kosina --- drivers/hid/wacom_wac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c index 0ff22dbe334c..01a4f05c1642 100644 --- a/drivers/hid/wacom_wac.c +++ b/drivers/hid/wacom_wac.c @@ -2481,7 +2481,7 @@ void wacom_setup_device_quirks(struct wacom *wacom) if (features->pktlen == WACOM_PKGLEN_BBTOUCH3) { if (features->touch_max) features->device_type |= WACOM_DEVICETYPE_TOUCH; - if (features->type >= INTUOSHT || features->type <= BAMBOO_PT) + if (features->type >= INTUOSHT && features->type <= BAMBOO_PT) features->device_type |= WACOM_DEVICETYPE_PAD; features->x_max = 4096; From 1285734c7a776b6195f8aaac6fa6acce26b5b97a Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 16 Nov 2015 15:34:15 -0800 Subject: [PATCH 190/335] MAINTAINERS: brcmnand: Add Broadcom internal mailing-list The Broadcom NAND driver is used by many different groups at Broadcom now, so use the same mailing-list we use for other areas of the kernel. Signed-off-by: Florian Fainelli Signed-off-by: Brian Norris --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index e9caa4b28828..da691aad7031 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2450,6 +2450,7 @@ F: drivers/firmware/broadcom/* BROADCOM STB NAND FLASH DRIVER M: Brian Norris L: linux-mtd@lists.infradead.org +L: bcm-kernel-feedback-list@broadcom.com S: Maintained F: drivers/mtd/nand/brcmnand/ From cf89752645e47d86ba8a4157f4b121fcb33434c5 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 12 Nov 2015 11:46:23 +0000 Subject: [PATCH 191/335] FS-Cache: Add missing initialization of ret in cachefiles_write_page() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit fs/cachefiles/rdwr.c: In function ‘cachefiles_write_page’: fs/cachefiles/rdwr.c:882: warning: ‘ret’ may be used uninitialized in this function If the jump to label "error" is taken, "ret" will indeed be uninitialized, and random stack data may be printed by the debug code. Fixes: 102f4d900c9c8f5e ("FS-Cache: Handle a write to the page immediately beyond the EOF marker") Signed-off-by: Geert Uytterhoeven Signed-off-by: David Howells Signed-off-by: Al Viro --- fs/cachefiles/rdwr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c index 7a6b02f72787..c0f3da3926a0 100644 --- a/fs/cachefiles/rdwr.c +++ b/fs/cachefiles/rdwr.c @@ -879,7 +879,7 @@ int cachefiles_write_page(struct fscache_storage *op, struct page *page) loff_t pos, eof; size_t len; void *data; - int ret; + int ret = -ENOBUFS; ASSERT(op != NULL); ASSERT(page != NULL); From e62d6e244f759556b569801c38dc4ca27a58bd18 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Thu, 12 Nov 2015 11:46:33 +0000 Subject: [PATCH 192/335] MAINTAINERS: linux-cachefs@redhat.com is moderated for non-subscribers Signed-off-by: Geert Uytterhoeven Signed-off-by: David Howells Signed-off-by: Al Viro --- MAINTAINERS | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index e9caa4b28828..78d9cbf96683 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2546,7 +2546,7 @@ F: arch/c6x/ CACHEFILES: FS-CACHE BACKEND FOR CACHING ON MOUNTED FILESYSTEMS M: David Howells -L: linux-cachefs@redhat.com +L: linux-cachefs@redhat.com (moderated for non-subscribers) S: Supported F: Documentation/filesystems/caching/cachefiles.txt F: fs/cachefiles/ @@ -4559,7 +4559,7 @@ F: include/linux/frontswap.h FS-CACHE: LOCAL CACHING FOR NETWORK FILESYSTEMS M: David Howells -L: linux-cachefs@redhat.com +L: linux-cachefs@redhat.com (moderated for non-subscribers) S: Supported F: Documentation/filesystems/caching/ F: fs/fscache/ From ee82c9ed41e896bd47e121d87e4628de0f2656a3 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Sun, 15 Nov 2015 16:06:32 -0800 Subject: [PATCH 193/335] dax: disable pmd mappings While dax pmd mappings are functional in the nominal path they trigger kernel crashes in the following paths: BUG: unable to handle kernel paging request at ffffea0004098000 IP: [] follow_trans_huge_pmd+0x117/0x3b0 [..] Call Trace: [] follow_page_mask+0x2d3/0x380 [] __get_user_pages+0xe8/0x6f0 [] get_user_pages_unlocked+0x165/0x1e0 [] get_user_pages_fast+0xa1/0x1b0 kernel BUG at arch/x86/mm/gup.c:131! [..] Call Trace: [] gup_pud_range+0x1bc/0x220 [] get_user_pages_fast+0x124/0x1b0 BUG: unable to handle kernel paging request at ffffea0004088000 IP: [] copy_huge_pmd+0x159/0x350 [..] Call Trace: [] copy_page_range+0x34c/0x9f0 [] copy_process+0x1b7f/0x1e10 [] _do_fork+0x91/0x590 All of these paths are interpreting a dax pmd mapping as a transparent huge page and making the assumption that the pfn is covered by the memmap, i.e. that the pfn has an associated struct page. PTE mappings do not suffer the same fate since they have the _PAGE_SPECIAL flag to cause the gup path to fault. We can do something similar for the PMD path, or otherwise defer pmd support for cases where a struct page is available. For now, 4.4-rc and -stable need to disable dax pmd support by default. For development the "depends on BROKEN" line can be removed from CONFIG_FS_DAX_PMD. Cc: Cc: Jan Kara Cc: Dave Chinner Cc: Matthew Wilcox Cc: Kirill A. Shutemov Reported-by: Ross Zwisler Signed-off-by: Dan Williams --- fs/Kconfig | 6 ++++++ fs/dax.c | 4 ++++ 2 files changed, 10 insertions(+) diff --git a/fs/Kconfig b/fs/Kconfig index da3f32f1a4e4..6ce72d8d1ee1 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -46,6 +46,12 @@ config FS_DAX or if unsure, say N. Saying Y will increase the size of the kernel by about 5kB. +config FS_DAX_PMD + bool + default FS_DAX + depends on FS_DAX + depends on BROKEN + endif # BLOCK # Posix ACL utility routines diff --git a/fs/dax.c b/fs/dax.c index d1e5cb7311a1..43671b68220e 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -541,6 +541,10 @@ int __dax_pmd_fault(struct vm_area_struct *vma, unsigned long address, unsigned long pfn; int result = 0; + /* dax pmd mappings are broken wrt gup and fork */ + if (!IS_ENABLED(CONFIG_FS_DAX_PMD)) + return VM_FAULT_FALLBACK; + /* Fall back to PTEs if we're going to COW */ if (write && !(vma->vm_flags & VM_SHARED)) return VM_FAULT_FALLBACK; From 24e79d0ddc416184594aa8193023c15d85ab15db Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Wed, 11 Nov 2015 11:29:07 +0100 Subject: [PATCH 194/335] drm/core: Set legacy_cursor_update in drm_atomic_helper_disable_plane. legacy_cursor_update was being set in restore_fbdev_mode_atomic which was probably unintended. Fix this by only setting it in the function that needs it. This oversight was introduced in commit bbb1e52402b2a288b09ae37e8182599931c7e9df Author: Rob Clark Date: Tue Aug 25 15:35:58 2015 -0400 drm/fb-helper: atomic restore_fbdev_mode()... Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Vetter [Jani: checkpatch fix] Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447237751-9663-2-git-send-email-maarten.lankhorst@ubuntu.com --- drivers/gpu/drm/drm_atomic_helper.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 0c6f62168776..b5efdfde641b 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1553,6 +1553,9 @@ retry: goto fail; } + if (plane_state->crtc && (plane == plane->crtc->cursor)) + plane_state->state->legacy_cursor_update = true; + ret = __drm_atomic_helper_disable_plane(plane, plane_state); if (ret != 0) goto fail; @@ -1605,9 +1608,6 @@ int __drm_atomic_helper_disable_plane(struct drm_plane *plane, plane_state->src_h = 0; plane_state->src_w = 0; - if (plane->crtc && (plane == plane->crtc->cursor)) - plane_state->state->legacy_cursor_update = true; - return 0; } From 4572372847680ee04f184df916d5cf007c94ff7e Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Wed, 11 Nov 2015 11:29:08 +0100 Subject: [PATCH 195/335] drm/core: Fix old_fb handling in drm_mode_atomic_ioctl. plane_mask should be cleared inside the retry loop, because it gets reset on every retry. Without this fix the plane->fb refcounting might get out of sync on retries, resulting in either leaked memory or use-after-free. Signed-off-by: Maarten Lankhorst Cc: stable@vger.kernel.org #v4.3 Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447237751-9663-3-git-send-email-maarten.lankhorst@ubuntu.com --- drivers/gpu/drm/drm_atomic.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index 7bb3845d9974..0ac31b1ecb67 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -1446,7 +1446,7 @@ int drm_mode_atomic_ioctl(struct drm_device *dev, struct drm_plane *plane; struct drm_crtc *crtc; struct drm_crtc_state *crtc_state; - unsigned plane_mask = 0; + unsigned plane_mask; int ret = 0; unsigned int i, j; @@ -1486,6 +1486,7 @@ int drm_mode_atomic_ioctl(struct drm_device *dev, state->allow_modeset = !!(arg->flags & DRM_MODE_ATOMIC_ALLOW_MODESET); retry: + plane_mask = 0; copied_objs = 0; copied_props = 0; From 0f45c26fc302c02b0576db37d4849baa53a2bb41 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Wed, 11 Nov 2015 11:29:09 +0100 Subject: [PATCH 196/335] drm/atomic: add a drm_atomic_clean_old_fb helper. This is useful for all the boilerplate code about cleaning old_fb. Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447237751-9663-4-git-send-email-maarten.lankhorst@ubuntu.com --- drivers/gpu/drm/drm_atomic.c | 58 +++++++++++++++++++++++++----------- include/drm/drm_atomic.h | 3 ++ 2 files changed, 43 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index 0ac31b1ecb67..aeee083c7f95 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -1432,6 +1432,45 @@ static int atomic_set_prop(struct drm_atomic_state *state, return ret; } +/** + * drm_atomic_update_old_fb -- Unset old_fb pointers and set plane->fb pointers. + * + * @dev: drm device to check. + * @plane_mask: plane mask for planes that were updated. + * @ret: return value, can be -EDEADLK for a retry. + * + * Before doing an update plane->old_fb is set to plane->fb, + * but before dropping the locks old_fb needs to be set to NULL + * and plane->fb updated. This is a common operation for each + * atomic update, so this call is split off as a helper. + */ +void drm_atomic_clean_old_fb(struct drm_device *dev, + unsigned plane_mask, + int ret) +{ + struct drm_plane *plane; + + /* if succeeded, fixup legacy plane crtc/fb ptrs before dropping + * locks (ie. while it is still safe to deref plane->state). We + * need to do this here because the driver entry points cannot + * distinguish between legacy and atomic ioctls. + */ + drm_for_each_plane_mask(plane, dev, plane_mask) { + if (ret == 0) { + struct drm_framebuffer *new_fb = plane->state->fb; + if (new_fb) + drm_framebuffer_reference(new_fb); + plane->fb = new_fb; + plane->crtc = plane->state->crtc; + + if (plane->old_fb) + drm_framebuffer_unreference(plane->old_fb); + } + plane->old_fb = NULL; + } +} +EXPORT_SYMBOL(drm_atomic_clean_old_fb); + int drm_mode_atomic_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { @@ -1577,24 +1616,7 @@ retry: } out: - /* if succeeded, fixup legacy plane crtc/fb ptrs before dropping - * locks (ie. while it is still safe to deref plane->state). We - * need to do this here because the driver entry points cannot - * distinguish between legacy and atomic ioctls. - */ - drm_for_each_plane_mask(plane, dev, plane_mask) { - if (ret == 0) { - struct drm_framebuffer *new_fb = plane->state->fb; - if (new_fb) - drm_framebuffer_reference(new_fb); - plane->fb = new_fb; - plane->crtc = plane->state->crtc; - - if (plane->old_fb) - drm_framebuffer_unreference(plane->old_fb); - } - plane->old_fb = NULL; - } + drm_atomic_clean_old_fb(dev, plane_mask, ret); if (ret && arg->flags & DRM_MODE_PAGE_FLIP_EVENT) { /* diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h index e67aeac2aee0..4b74c97d297a 100644 --- a/include/drm/drm_atomic.h +++ b/include/drm/drm_atomic.h @@ -136,6 +136,9 @@ drm_atomic_connectors_for_crtc(struct drm_atomic_state *state, void drm_atomic_legacy_backoff(struct drm_atomic_state *state); +void +drm_atomic_clean_old_fb(struct drm_device *dev, unsigned plane_mask, int ret); + int __must_check drm_atomic_check_only(struct drm_atomic_state *state); int __must_check drm_atomic_commit(struct drm_atomic_state *state); int __must_check drm_atomic_async_commit(struct drm_atomic_state *state); From f72c6b33edda439acff81a0da612f2bcd1f46f35 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Wed, 11 Nov 2015 11:29:10 +0100 Subject: [PATCH 197/335] drm/core: Fix old_fb handling in restore_fbdev_mode_atomic. Don't touch plane->old_fb/fb without having the right locks held. Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447237751-9663-5-git-send-email-maarten.lankhorst@ubuntu.com --- drivers/gpu/drm/drm_fb_helper.c | 21 ++++++--------------- 1 file changed, 6 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c index e673c13c7391..abd50863506e 100644 --- a/drivers/gpu/drm/drm_fb_helper.c +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -342,6 +342,7 @@ static int restore_fbdev_mode_atomic(struct drm_fb_helper *fb_helper) struct drm_plane *plane; struct drm_atomic_state *state; int i, ret; + unsigned plane_mask; state = drm_atomic_state_alloc(dev); if (!state) @@ -349,11 +350,10 @@ static int restore_fbdev_mode_atomic(struct drm_fb_helper *fb_helper) state->acquire_ctx = dev->mode_config.acquire_ctx; retry: + plane_mask = 0; drm_for_each_plane(plane, dev) { struct drm_plane_state *plane_state; - plane->old_fb = plane->fb; - plane_state = drm_atomic_get_plane_state(state, plane); if (IS_ERR(plane_state)) { ret = PTR_ERR(plane_state); @@ -362,6 +362,9 @@ retry: plane_state->rotation = BIT(DRM_ROTATE_0); + plane->old_fb = plane->fb; + plane_mask |= 1 << drm_plane_index(plane); + /* disable non-primary: */ if (plane->type == DRM_PLANE_TYPE_PRIMARY) continue; @@ -382,19 +385,7 @@ retry: ret = drm_atomic_commit(state); fail: - drm_for_each_plane(plane, dev) { - if (ret == 0) { - struct drm_framebuffer *new_fb = plane->state->fb; - if (new_fb) - drm_framebuffer_reference(new_fb); - plane->fb = new_fb; - plane->crtc = plane->state->crtc; - - if (plane->old_fb) - drm_framebuffer_unreference(plane->old_fb); - } - plane->old_fb = NULL; - } + drm_atomic_clean_old_fb(dev, plane_mask, ret); if (ret == -EDEADLK) goto backoff; From 07d3bad6c1210bd21e85d084807ef4ee4ac43a78 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Wed, 11 Nov 2015 11:29:11 +0100 Subject: [PATCH 198/335] drm/core: Fix old_fb handling in pan_display_atomic. Don't touch plane->old_fb/fb without having the right locks held. Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447237751-9663-6-git-send-email-maarten.lankhorst@ubuntu.com --- drivers/gpu/drm/drm_fb_helper.c | 30 ++++++++---------------------- 1 file changed, 8 insertions(+), 22 deletions(-) diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c index abd50863506e..69cbab5e5c81 100644 --- a/drivers/gpu/drm/drm_fb_helper.c +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -1227,7 +1227,9 @@ static int pan_display_atomic(struct fb_var_screeninfo *var, struct drm_fb_helper *fb_helper = info->par; struct drm_device *dev = fb_helper->dev; struct drm_atomic_state *state; + struct drm_plane *plane; int i, ret; + unsigned plane_mask; state = drm_atomic_state_alloc(dev); if (!state) @@ -1235,19 +1237,22 @@ static int pan_display_atomic(struct fb_var_screeninfo *var, state->acquire_ctx = dev->mode_config.acquire_ctx; retry: + plane_mask = 0; for(i = 0; i < fb_helper->crtc_count; i++) { struct drm_mode_set *mode_set; mode_set = &fb_helper->crtc_info[i].mode_set; - mode_set->crtc->primary->old_fb = mode_set->crtc->primary->fb; - mode_set->x = var->xoffset; mode_set->y = var->yoffset; ret = __drm_atomic_helper_set_config(mode_set, state); if (ret != 0) goto fail; + + plane = mode_set->crtc->primary; + plane_mask |= drm_plane_index(plane); + plane->old_fb = plane->fb; } ret = drm_atomic_commit(state); @@ -1259,26 +1264,7 @@ retry: fail: - for(i = 0; i < fb_helper->crtc_count; i++) { - struct drm_mode_set *mode_set; - struct drm_plane *plane; - - mode_set = &fb_helper->crtc_info[i].mode_set; - plane = mode_set->crtc->primary; - - if (ret == 0) { - struct drm_framebuffer *new_fb = plane->state->fb; - - if (new_fb) - drm_framebuffer_reference(new_fb); - plane->fb = new_fb; - plane->crtc = plane->state->crtc; - - if (plane->old_fb) - drm_framebuffer_unreference(plane->old_fb); - } - plane->old_fb = NULL; - } + drm_atomic_clean_old_fb(dev, plane_mask, ret); if (ret == -EDEADLK) goto backoff; From 4fee9f364b9b99f76732f2a6fd6df679a237fa74 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Mon, 16 Nov 2015 11:18:14 +0100 Subject: [PATCH 199/335] arm64: mm: use correct mapping granularity under DEBUG_RODATA When booting a 64k pages kernel that is built with CONFIG_DEBUG_RODATA and resides at an offset that is not a multiple of 512 MB, the rounding that occurs in __map_memblock() and fixup_executable() results in incorrect regions being mapped. The following snippet from /sys/kernel/debug/kernel_page_tables shows how, when the kernel is loaded 2 MB above the base of DRAM at 0x40000000, the first 2 MB of memory (which may be inaccessible from non-secure EL1 or just reserved by the firmware) is inadvertently mapped into the end of the module region. ---[ Modules start ]--- 0xfffffdffffe00000-0xfffffe0000000000 2M RW NX ... UXN MEM/NORMAL ---[ Modules end ]--- ---[ Kernel Mapping ]--- 0xfffffe0000000000-0xfffffe0000090000 576K RW NX ... UXN MEM/NORMAL 0xfffffe0000090000-0xfffffe0000200000 1472K ro x ... UXN MEM/NORMAL 0xfffffe0000200000-0xfffffe0000800000 6M ro x ... UXN MEM/NORMAL 0xfffffe0000800000-0xfffffe0000810000 64K ro x ... UXN MEM/NORMAL 0xfffffe0000810000-0xfffffe0000a00000 1984K RW NX ... UXN MEM/NORMAL 0xfffffe0000a00000-0xfffffe00ffe00000 4084M RW NX ... UXN MEM/NORMAL The same issue is likely to occur on 16k pages kernels whose load address is not a multiple of 32 MB (i.e., SECTION_SIZE). So round to SWAPPER_BLOCK_SIZE instead of SECTION_SIZE. Fixes: da141706aea5 ("arm64: add better page protections to arm64") Signed-off-by: Ard Biesheuvel Acked-by: Mark Rutland Acked-by: Laura Abbott Cc: # 4.0+ Signed-off-by: Catalin Marinas --- arch/arm64/mm/mmu.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c index e3f563c81c48..32ddd893da9a 100644 --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -362,8 +362,8 @@ static void __init __map_memblock(phys_addr_t start, phys_addr_t end) * for now. This will get more fine grained later once all memory * is mapped */ - unsigned long kernel_x_start = round_down(__pa(_stext), SECTION_SIZE); - unsigned long kernel_x_end = round_up(__pa(__init_end), SECTION_SIZE); + unsigned long kernel_x_start = round_down(__pa(_stext), SWAPPER_BLOCK_SIZE); + unsigned long kernel_x_end = round_up(__pa(__init_end), SWAPPER_BLOCK_SIZE); if (end < kernel_x_start) { create_mapping(start, __phys_to_virt(start), @@ -451,18 +451,18 @@ static void __init fixup_executable(void) { #ifdef CONFIG_DEBUG_RODATA /* now that we are actually fully mapped, make the start/end more fine grained */ - if (!IS_ALIGNED((unsigned long)_stext, SECTION_SIZE)) { + if (!IS_ALIGNED((unsigned long)_stext, SWAPPER_BLOCK_SIZE)) { unsigned long aligned_start = round_down(__pa(_stext), - SECTION_SIZE); + SWAPPER_BLOCK_SIZE); create_mapping(aligned_start, __phys_to_virt(aligned_start), __pa(_stext) - aligned_start, PAGE_KERNEL); } - if (!IS_ALIGNED((unsigned long)__init_end, SECTION_SIZE)) { + if (!IS_ALIGNED((unsigned long)__init_end, SWAPPER_BLOCK_SIZE)) { unsigned long aligned_end = round_up(__pa(__init_end), - SECTION_SIZE); + SWAPPER_BLOCK_SIZE); create_mapping(__pa(__init_end), (unsigned long)__init_end, aligned_end - __pa(__init_end), PAGE_KERNEL); From 1dccb598df549d892b6450c261da54cdd7af44b4 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 17:25:48 +0100 Subject: [PATCH 200/335] arm64: simplify dma_get_ops Including linux/acpi.h from asm/dma-mapping.h causes tons of compile-time warnings, e.g. drivers/isdn/mISDN/dsp_ecdis.h:43:0: warning: "FALSE" redefined drivers/isdn/mISDN/dsp_ecdis.h:44:0: warning: "TRUE" redefined drivers/net/fddi/skfp/h/targetos.h:62:0: warning: "TRUE" redefined drivers/net/fddi/skfp/h/targetos.h:63:0: warning: "FALSE" redefined However, it looks like the dependency should not even there as I do not see why __generic_dma_ops() cares about whether we have an ACPI based system or not. The current behavior is to fall back to the global dma_ops when a device has not set its own dma_ops, but only for DT based systems. This seems dangerous, as a random device might have different requirements regarding IOMMU or coherency, so we should really never have that fallback and just forbid DMA when we have not initialized DMA for a device. This removes the global dma_ops variable and the special-casing for ACPI, and just returns the dma ops that got set for the device, or the dummy_dma_ops if none were present. The original code has apparently been copied from arm32 where we rely on it for ISA devices things like the floppy controller, but we should have no such devices on ARM64. Signed-off-by: Arnd Bergmann [catalin.marinas@arm.com: removed acpi_disabled check in arch_setup_dma_ops()] Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/dma-mapping.h | 13 +++---------- arch/arm64/mm/dma-mapping.c | 16 ++++------------ 2 files changed, 7 insertions(+), 22 deletions(-) diff --git a/arch/arm64/include/asm/dma-mapping.h b/arch/arm64/include/asm/dma-mapping.h index 54d0ead41afc..61e08f360e31 100644 --- a/arch/arm64/include/asm/dma-mapping.h +++ b/arch/arm64/include/asm/dma-mapping.h @@ -18,7 +18,6 @@ #ifdef __KERNEL__ -#include #include #include @@ -26,22 +25,16 @@ #include #define DMA_ERROR_CODE (~(dma_addr_t)0) -extern struct dma_map_ops *dma_ops; extern struct dma_map_ops dummy_dma_ops; static inline struct dma_map_ops *__generic_dma_ops(struct device *dev) { - if (unlikely(!dev)) - return dma_ops; - else if (dev->archdata.dma_ops) + if (dev && dev->archdata.dma_ops) return dev->archdata.dma_ops; - else if (acpi_disabled) - return dma_ops; /* - * When ACPI is enabled, if arch_set_dma_ops is not called, - * we will disable device DMA capability by setting it - * to dummy_dma_ops. + * We expect no ISA devices, and all other DMA masters are expected to + * have someone call arch_setup_dma_ops at device creation time. */ return &dummy_dma_ops; } diff --git a/arch/arm64/mm/dma-mapping.c b/arch/arm64/mm/dma-mapping.c index 97fd714035f9..7963aa4b5d28 100644 --- a/arch/arm64/mm/dma-mapping.c +++ b/arch/arm64/mm/dma-mapping.c @@ -18,6 +18,7 @@ */ #include +#include #include #include #include @@ -28,9 +29,6 @@ #include -struct dma_map_ops *dma_ops; -EXPORT_SYMBOL(dma_ops); - static pgprot_t __get_dma_pgprot(struct dma_attrs *attrs, pgprot_t prot, bool coherent) { @@ -515,13 +513,7 @@ EXPORT_SYMBOL(dummy_dma_ops); static int __init arm64_dma_init(void) { - int ret; - - dma_ops = &swiotlb_dma_ops; - - ret = atomic_pool_init(); - - return ret; + return atomic_pool_init(); } arch_initcall(arm64_dma_init); @@ -991,8 +983,8 @@ static void __iommu_setup_dma_ops(struct device *dev, u64 dma_base, u64 size, void arch_setup_dma_ops(struct device *dev, u64 dma_base, u64 size, struct iommu_ops *iommu, bool coherent) { - if (!acpi_disabled && !dev->archdata.dma_ops) - dev->archdata.dma_ops = dma_ops; + if (!dev->archdata.dma_ops) + dev->archdata.dma_ops = &swiotlb_dma_ops; dev->archdata.dma_coherent = coherent; __iommu_setup_dma_ops(dev, dma_base, size, iommu); From adc235aff6fcf5ae4d8dc00faf30a07632b9725b Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 17:32:15 +0100 Subject: [PATCH 201/335] arm64: do not include ptrace.h from compat.h including ptrace.h brings a definition of BITS_PER_PAGE into device drivers and cause a build warning in allmodconfig builds: drivers/block/drbd/drbd_bitmap.c:482:0: warning: "BITS_PER_PAGE" redefined #define BITS_PER_PAGE (1UL << (PAGE_SHIFT + 3)) This uses a slightly different way to express current_pt_regs() that avoids the use of the header and gets away with the already included asm/ptrace.h. Signed-off-by: Arnd Bergmann Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/compat.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm64/include/asm/compat.h b/arch/arm64/include/asm/compat.h index 7fbed6919b54..eb8432bb82b8 100644 --- a/arch/arm64/include/asm/compat.h +++ b/arch/arm64/include/asm/compat.h @@ -23,7 +23,6 @@ */ #include #include -#include #define COMPAT_USER_HZ 100 #ifdef __AARCH64EB__ @@ -234,7 +233,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -#define compat_user_stack_pointer() (user_stack_pointer(current_pt_regs())) +#define compat_user_stack_pointer() (user_stack_pointer(task_pt_regs(current))) static inline void __user *arch_compat_alloc_user_space(long len) { From 0eece2b22849c90b730815c893425a36b9d10fd5 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Mon, 16 Nov 2015 19:13:26 +0000 Subject: [PATCH 202/335] irqchip/gic: Make sure all interrupts are deactivated at boot When booting a GIC/GICv3 based system, we have no idea what state the firmware (or previous kernel in the case of kexec) has left the GIC, and some interrupts may still be active. In order to garantee that we have a clean state, make sure the active bits are cleared at init time. Signed-off-by: Marc Zyngier Cc: Cc: Jason Cooper Cc: Russell King Link: http://lkml.kernel.org/r/1447701208-18150-3-git-send-email-marc.zyngier@arm.com Signed-off-by: Thomas Gleixner --- drivers/irqchip/irq-gic-common.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/irqchip/irq-gic-common.c b/drivers/irqchip/irq-gic-common.c index 44a077f3a4a2..f174ce0ca361 100644 --- a/drivers/irqchip/irq-gic-common.c +++ b/drivers/irqchip/irq-gic-common.c @@ -84,12 +84,15 @@ void __init gic_dist_config(void __iomem *base, int gic_irqs, writel_relaxed(GICD_INT_DEF_PRI_X4, base + GIC_DIST_PRI + i); /* - * Disable all interrupts. Leave the PPI and SGIs alone - * as they are enabled by redistributor registers. + * Deactivate and disable all SPIs. Leave the PPI and SGIs + * alone as they are in the redistributor registers on GICv3. */ - for (i = 32; i < gic_irqs; i += 32) + for (i = 32; i < gic_irqs; i += 32) { writel_relaxed(GICD_INT_EN_CLR_X32, - base + GIC_DIST_ENABLE_CLEAR + i / 8); + base + GIC_DIST_ACTIVE_CLEAR + i / 8); + writel_relaxed(GICD_INT_EN_CLR_X32, + base + GIC_DIST_ENABLE_CLEAR + i / 8); + } if (sync_access) sync_access(); @@ -102,7 +105,9 @@ void gic_cpu_config(void __iomem *base, void (*sync_access)(void)) /* * Deal with the banked PPI and SGI interrupts - disable all * PPI interrupts, ensure all SGI interrupts are enabled. + * Make sure everything is deactivated. */ + writel_relaxed(GICD_INT_EN_CLR_X32, base + GIC_DIST_ACTIVE_CLEAR); writel_relaxed(GICD_INT_EN_CLR_PPI, base + GIC_DIST_ENABLE_CLEAR); writel_relaxed(GICD_INT_EN_SET_SGI, base + GIC_DIST_ENABLE_SET); From 92eda4ad4371225d6ccf9cded74315547e9a3153 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Mon, 16 Nov 2015 19:13:27 +0000 Subject: [PATCH 203/335] irqchip/gic: Clear enable bits before restoring them When restoring the GIC state (after a suspend/resume cycle, for example), the driver directly writes the 'enabled' state it has saved by accessing GICD_ISENABLERn, which performs an OR operation between the value present in the register and the value we write. If whatever code that has run before we reentered the kernel has enabled an interrupt that was previously disabled, we won't restore that disabled state. Making sure we first clear the register (by writting to GICD_ICENABLERn) before restoring the enabled state. Signed-off-by: Marc Zyngier Cc: Cc: Jason Cooper Cc: Russell King Link: http://lkml.kernel.org/r/1447701208-18150-4-git-send-email-marc.zyngier@arm.com Signed-off-by: Thomas Gleixner --- drivers/irqchip/irq-gic.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c index 515c823c1c95..bc846e7fec44 100644 --- a/drivers/irqchip/irq-gic.c +++ b/drivers/irqchip/irq-gic.c @@ -604,9 +604,12 @@ static void gic_dist_restore(unsigned int gic_nr) writel_relaxed(gic_data[gic_nr].saved_spi_target[i], dist_base + GIC_DIST_TARGET + i * 4); - for (i = 0; i < DIV_ROUND_UP(gic_irqs, 32); i++) + for (i = 0; i < DIV_ROUND_UP(gic_irqs, 32); i++) { + writel_relaxed(GICD_INT_EN_CLR_X32, + dist_base + GIC_DIST_ENABLE_CLEAR + i * 4); writel_relaxed(gic_data[gic_nr].saved_spi_enable[i], dist_base + GIC_DIST_ENABLE_SET + i * 4); + } writel_relaxed(GICD_ENABLE, dist_base + GIC_DIST_CTRL); } @@ -654,8 +657,11 @@ static void gic_cpu_restore(unsigned int gic_nr) return; ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_enable); - for (i = 0; i < DIV_ROUND_UP(32, 32); i++) + for (i = 0; i < DIV_ROUND_UP(32, 32); i++) { + writel_relaxed(GICD_INT_EN_CLR_X32, + dist_base + GIC_DIST_ENABLE_CLEAR + i * 4); writel_relaxed(ptr[i], dist_base + GIC_DIST_ENABLE_SET + i * 4); + } ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_conf); for (i = 0; i < DIV_ROUND_UP(32, 16); i++) From 1c7d4dd46ee048afe19e1294634df6fa66862519 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Mon, 16 Nov 2015 19:13:28 +0000 Subject: [PATCH 204/335] irqchip/gic: Add save/restore of the active state When using EOImode==1, we may mark interrupts as being forwarded to a virtual machine. In that case, the interrupt is left active while being passed to the VM. If we suspend the system before the VM has deactivated the interrupt, the active state will be lost (which may be very annoying, as this may result in spurious interrupts and a confused guest). To avoid this, save and restore the active state together with the rest of the GIC registers. Signed-off-by: Marc Zyngier Cc: Cc: Jason Cooper Cc: Russell King Link: http://lkml.kernel.org/r/1447701208-18150-5-git-send-email-marc.zyngier@arm.com Signed-off-by: Thomas Gleixner --- drivers/irqchip/irq-gic.c | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c index bc846e7fec44..abf2ffaed392 100644 --- a/drivers/irqchip/irq-gic.c +++ b/drivers/irqchip/irq-gic.c @@ -73,9 +73,11 @@ struct gic_chip_data { union gic_base cpu_base; #ifdef CONFIG_CPU_PM u32 saved_spi_enable[DIV_ROUND_UP(1020, 32)]; + u32 saved_spi_active[DIV_ROUND_UP(1020, 32)]; u32 saved_spi_conf[DIV_ROUND_UP(1020, 16)]; u32 saved_spi_target[DIV_ROUND_UP(1020, 4)]; u32 __percpu *saved_ppi_enable; + u32 __percpu *saved_ppi_active; u32 __percpu *saved_ppi_conf; #endif struct irq_domain *domain; @@ -566,6 +568,10 @@ static void gic_dist_save(unsigned int gic_nr) for (i = 0; i < DIV_ROUND_UP(gic_irqs, 32); i++) gic_data[gic_nr].saved_spi_enable[i] = readl_relaxed(dist_base + GIC_DIST_ENABLE_SET + i * 4); + + for (i = 0; i < DIV_ROUND_UP(gic_irqs, 32); i++) + gic_data[gic_nr].saved_spi_active[i] = + readl_relaxed(dist_base + GIC_DIST_ACTIVE_SET + i * 4); } /* @@ -611,6 +617,13 @@ static void gic_dist_restore(unsigned int gic_nr) dist_base + GIC_DIST_ENABLE_SET + i * 4); } + for (i = 0; i < DIV_ROUND_UP(gic_irqs, 32); i++) { + writel_relaxed(GICD_INT_EN_CLR_X32, + dist_base + GIC_DIST_ACTIVE_CLEAR + i * 4); + writel_relaxed(gic_data[gic_nr].saved_spi_active[i], + dist_base + GIC_DIST_ACTIVE_SET + i * 4); + } + writel_relaxed(GICD_ENABLE, dist_base + GIC_DIST_CTRL); } @@ -634,6 +647,10 @@ static void gic_cpu_save(unsigned int gic_nr) for (i = 0; i < DIV_ROUND_UP(32, 32); i++) ptr[i] = readl_relaxed(dist_base + GIC_DIST_ENABLE_SET + i * 4); + ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_active); + for (i = 0; i < DIV_ROUND_UP(32, 32); i++) + ptr[i] = readl_relaxed(dist_base + GIC_DIST_ACTIVE_SET + i * 4); + ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_conf); for (i = 0; i < DIV_ROUND_UP(32, 16); i++) ptr[i] = readl_relaxed(dist_base + GIC_DIST_CONFIG + i * 4); @@ -663,6 +680,13 @@ static void gic_cpu_restore(unsigned int gic_nr) writel_relaxed(ptr[i], dist_base + GIC_DIST_ENABLE_SET + i * 4); } + ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_active); + for (i = 0; i < DIV_ROUND_UP(32, 32); i++) { + writel_relaxed(GICD_INT_EN_CLR_X32, + dist_base + GIC_DIST_ACTIVE_CLEAR + i * 4); + writel_relaxed(ptr[i], dist_base + GIC_DIST_ACTIVE_SET + i * 4); + } + ptr = raw_cpu_ptr(gic_data[gic_nr].saved_ppi_conf); for (i = 0; i < DIV_ROUND_UP(32, 16); i++) writel_relaxed(ptr[i], dist_base + GIC_DIST_CONFIG + i * 4); @@ -716,6 +740,10 @@ static void __init gic_pm_init(struct gic_chip_data *gic) sizeof(u32)); BUG_ON(!gic->saved_ppi_enable); + gic->saved_ppi_active = __alloc_percpu(DIV_ROUND_UP(32, 32) * 4, + sizeof(u32)); + BUG_ON(!gic->saved_ppi_active); + gic->saved_ppi_conf = __alloc_percpu(DIV_ROUND_UP(32, 16) * 4, sizeof(u32)); BUG_ON(!gic->saved_ppi_conf); From ba8af3e592f7175b5f9a92d2cfcc00b82097d1be Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Mon, 16 Nov 2015 12:49:14 +0100 Subject: [PATCH 205/335] drm/i915: Clear intel_crtc->atomic before updating it. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If an atomic update fails intel_crtc->atomic may have have some values left from the last atomic check. One example is atomic->wait_for_vblank, which results in spurious errors in kms_flip. [ 1551.892708] ------------[ cut here ]------------ [ 1551.892721] WARNING: CPU: 3 PID: 4179 at ../drivers/gpu/drm/drm_irq.c:1199 drm_wait_one_vblank+0x197/0x1a0 [drm]() [ 1551.892722] vblank not available on crtc 2, ret=-22 [ 1551.892751] Modules linked in: snd_hda_intel i915 drm_kms_helper drm intel_gtt i2c_algo_bit cfbfillrect syscopyarea cfbimgblt sysfillrect sysimgblt fb_sys_fops cfbcopyarea agpgart cfg80211 binfmt_misc snd_hda_codec_hdmi intel_rapl iosf_mbi x86_pkg_temp_thermal coretemp kvm_intel snd_hda_codec_realtek kvm snd_hda_codec_generic iTCO_wdt aesni_intel aes_x86_64 glue_helper lrw snd_hda_codec gf128mul ablk_helper cryptd snd_hwdep psmouse snd_hda_core pcspkr snd_pcm snd_timer snd lpc_ich i2c_i801 mfd_core soundcore wmi evdev [last unloaded: drm] [ 1551.892753] CPU: 3 PID: 4179 Comm: kms_pipe_crc_ba Tainted: G U W 4.3.0-reg+ #6 [ 1551.892754] Hardware name: /DZ77BH-55K, BIOS BHZ7710H.86A.0100.2013.0517.0942 05/17/2013 [ 1551.892758] ffffffffa03128d8 ffff8800cec73890 ffffffff812c0f3c ffff8800cec738d8 [ 1551.892760] ffff8800cec738c8 ffffffff8104ff36 ffff880116ae2290 0000000000000002 [ 1551.892762] ffff8800d39fcda0 ffff8800d038b4d0 ffff8800d42b5550 ffff8800cec73928 [ 1551.892763] Call Trace: [ 1551.892768] [] dump_stack+0x4e/0x82 [ 1551.892771] [] warn_slowpath_common+0x86/0xc0 [ 1551.892773] [] warn_slowpath_fmt+0x4c/0x50 [ 1551.892781] [] ? drm_vblank_get+0x78/0xd0 [drm] [ 1551.892787] [] drm_wait_one_vblank+0x197/0x1a0 [drm] [ 1551.892813] [] intel_post_plane_update+0xef/0x120 [i915] [ 1551.892832] [] intel_atomic_commit+0x4c2/0x1600 [i915] [ 1551.892862] [] ? drm_atomic_check_only+0x147/0x5e0 [drm] [ 1551.892872] [] ? drm_atomic_add_affected_connectors+0x27/0xf0 [drm] [ 1551.892881] [] drm_atomic_commit+0x37/0x60 [drm] [ 1551.892887] [] restore_fbdev_mode+0x28a/0x2c0 [drm_kms_helper] [ 1551.892895] [] drm_fb_helper_restore_fbdev_mode_unlocked+0x33/0x80 [drm_kms_helper] [ 1551.892900] [] drm_fb_helper_set_par+0x2d/0x50 [drm_kms_helper] [ 1551.892920] [] intel_fbdev_set_par+0x1a/0x60 [i915] [ 1551.892923] [] fb_set_var+0x1a7/0x3f0 [ 1551.892927] [] ? trace_hardirqs_on_caller+0x12f/0x1c0 [ 1551.892931] [] fbcon_blank+0x212/0x2f0 [ 1551.892935] [] do_unblank_screen+0xba/0x1d0 [ 1551.892937] [] vt_ioctl+0x13d5/0x1450 [ 1551.892940] [] ? preempt_count_sub+0x41/0x50 [ 1551.892943] [] tty_ioctl+0x423/0xe30 [ 1551.892947] [] do_vfs_ioctl+0x301/0x560 [ 1551.892949] [] ? putname+0x53/0x60 [ 1551.892952] [] ? __fget_light+0x66/0x90 [ 1551.892955] [] SyS_ioctl+0x79/0x90 [ 1551.892958] [] entry_SYSCALL_64_fastpath+0x12/0x6f [ 1551.892961] ---[ end trace 3e764d4b6628c91c ]--- Testcase: kms_flip Reported-and-tested-by: Ville Syrjälä Cc: stable@vger.kernel.org #v4.3 Signed-off-by: Maarten Lankhorst Reviewed-by: Daniel Stone Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/5649C2BA.6080300@mblankhorst.nl --- drivers/gpu/drm/i915/intel_display.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index bcf227356f78..442c1d220276 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -13034,6 +13034,9 @@ static int intel_atomic_check(struct drm_device *dev, struct intel_crtc_state *pipe_config = to_intel_crtc_state(crtc_state); + memset(&to_intel_crtc(crtc)->atomic, 0, + sizeof(struct intel_crtc_atomic_commit)); + /* Catch I915_MODE_FLAG_INHERITED */ if (crtc_state->mode.private_flags != crtc->state->mode.private_flags) crtc_state->mode_changed = true; From 83926117273cd6d9ffa1b5ff0cc566ad38648b70 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Mon, 16 Nov 2015 17:02:34 +0200 Subject: [PATCH 206/335] drm: Fix primary plane size for stereo doubled modes for legacy setcrtc MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Properly double the hdisplay/vdisplay timings that we use as the primary plane size with stereo doubled modes. Otherwise the modeset gets rejected on machines where the primary plane must be fullscreen, and on the rest only the first eye would get a visible plane. Cc: Daniel Vetter Cc: stable@vger.kernel.org #v3.19+ Fixes: 042652ed9599 ("drm/atomic-helper: implementatations for legacy interfaces") Signed-off-by: Ville Syrjälä Link: http://patchwork.freedesktop.org/patch/msgid/1447686157-29607-1-git-send-email-ville.syrjala@linux.intel.com Testcase: igt/kms_3d Signed-off-by: Daniel Vetter Signed-off-by: Jani Nikula --- drivers/gpu/drm/drm_atomic_helper.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index b5efdfde641b..cd398a245d2e 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1741,6 +1741,7 @@ int __drm_atomic_helper_set_config(struct drm_mode_set *set, struct drm_crtc_state *crtc_state; struct drm_plane_state *primary_state; struct drm_crtc *crtc = set->crtc; + int hdisplay, vdisplay; int ret; crtc_state = drm_atomic_get_crtc_state(state, crtc); @@ -1783,19 +1784,21 @@ int __drm_atomic_helper_set_config(struct drm_mode_set *set, if (ret != 0) return ret; + drm_crtc_get_hv_timing(set->mode, &hdisplay, &vdisplay); + drm_atomic_set_fb_for_plane(primary_state, set->fb); primary_state->crtc_x = 0; primary_state->crtc_y = 0; - primary_state->crtc_h = set->mode->vdisplay; - primary_state->crtc_w = set->mode->hdisplay; + primary_state->crtc_h = vdisplay; + primary_state->crtc_w = hdisplay; primary_state->src_x = set->x << 16; primary_state->src_y = set->y << 16; if (primary_state->rotation & (BIT(DRM_ROTATE_90) | BIT(DRM_ROTATE_270))) { - primary_state->src_h = set->mode->hdisplay << 16; - primary_state->src_w = set->mode->vdisplay << 16; + primary_state->src_h = hdisplay << 16; + primary_state->src_w = vdisplay << 16; } else { - primary_state->src_h = set->mode->vdisplay << 16; - primary_state->src_w = set->mode->hdisplay << 16; + primary_state->src_h = vdisplay << 16; + primary_state->src_w = hdisplay << 16; } commit: From 76dc3769d7c3cdcfa7c4c7768a7cb87cd91af12f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 11 Nov 2015 19:11:28 +0200 Subject: [PATCH 207/335] drm/i915: Don't clobber the addfb2 ioctl params MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We try to convert the old way of of specifying fb tiling (obj->tiling) into the new fb modifiers. We store the result in the passed in mode_cmd structure. But that structure comes directly from the addfb2 ioctl, and gets copied back out to userspace, which means we're clobbering the modifiers that the user provided (all 0 since the DRM_MODE_FB_MODIFIERS flag wasn't even set by the user). Hence if the user reuses the struct for another addfb2, the ioctl will be rejected since it's now asking for some modifiers w/o the flag set. Fix the problem by making a copy of the user provided structure. We can play any games we want with the copy. IGT-Version: 1.12-git (x86_64) (Linux: 4.4.0-rc1-stereo+ x86_64) ... Subtest basic-X-tiled: SUCCESS (0.001s) Test assertion failure function pitch_tests, file kms_addfb_basic.c:167: Failed assertion: drmIoctl(fd, DRM_IOCTL_MODE_ADDFB2, &f) == 0 Last errno: 22, Invalid argument Stack trace: #0 [__igt_fail_assert+0x101] #1 [pitch_tests+0x619] #2 [__real_main426+0x2f] #3 [main+0x23] #4 [__libc_start_main+0xf0] #5 [_start+0x29] #6 [+0x29] Subtest framebuffer-vs-set-tiling failed. **** DEBUG **** Test assertion failure function pitch_tests, file kms_addfb_basic.c:167: Failed assertion: drmIoctl(fd, DRM_IOCTL_MODE_ADDFB2, &f) == 0 Last errno: 22, Invalid argument **** END **** Subtest framebuffer-vs-set-tiling: FAIL (0.003s) ... IGT-Version: 1.12-git (x86_64) (Linux: 4.4.0-rc1-stereo+ x86_64) Subtest framebuffer-vs-set-tiling: SUCCESS (0.000s) Cc: stable@vger.kernel.org # v4.1+ Cc: Daniel Vetter Cc: Tvrtko Ursulin Fixes: 2a80eada326f ("drm/i915: Add fb format modifier support") Testcase: igt/kms_addfb_basic/clobbered-modifier Signed-off-by: Ville Syrjälä Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447261890-3960-1-git-send-email-ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/intel_display.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 442c1d220276..8f80bbeb1313 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14369,16 +14369,17 @@ static int intel_framebuffer_init(struct drm_device *dev, static struct drm_framebuffer * intel_user_framebuffer_create(struct drm_device *dev, struct drm_file *filp, - struct drm_mode_fb_cmd2 *mode_cmd) + struct drm_mode_fb_cmd2 *user_mode_cmd) { struct drm_i915_gem_object *obj; + struct drm_mode_fb_cmd2 mode_cmd = *user_mode_cmd; obj = to_intel_bo(drm_gem_object_lookup(dev, filp, - mode_cmd->handles[0])); + mode_cmd.handles[0])); if (&obj->base == NULL) return ERR_PTR(-ENOENT); - return intel_framebuffer_create(dev, mode_cmd, obj); + return intel_framebuffer_create(dev, &mode_cmd, obj); } #ifndef CONFIG_DRM_FBDEV_EMULATION From de818bd4522c40ea02a81b387d2fa86f989c9623 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Tue, 17 Nov 2015 11:50:51 +0000 Subject: [PATCH 208/335] arm64: kernel: pause/unpause function graph tracer in cpu_suspend() The function graph tracer adds instrumentation that is required to trace both entry and exit of a function. In particular the function graph tracer updates the "return address" of a function in order to insert a trace callback on function exit. Kernel power management functions like cpu_suspend() are called upon power down entry with functions called "finishers" that are in turn called to trigger the power down sequence but they may not return to the kernel through the normal return path. When the core resumes from low-power it returns to the cpu_suspend() function through the cpu_resume path, which leaves the trace stack frame set-up by the function tracer in an incosistent state upon return to the kernel when tracing is enabled. This patch fixes the issue by pausing/resuming the function graph tracer on the thread executing cpu_suspend() (ie the function call that subsequently triggers the "suspend finishers"), so that the function graph tracer state is kept consistent across functions that enter power down states and never return by effectively disabling graph tracer while they are executing. Fixes: 819e50e25d0c ("arm64: Add ftrace support") Signed-off-by: Lorenzo Pieralisi Reported-by: Catalin Marinas Reported-by: AKASHI Takahiro Suggested-by: Steven Rostedt Acked-by: Steven Rostedt Cc: Will Deacon Cc: # 3.16+ Signed-off-by: Catalin Marinas --- arch/arm64/kernel/suspend.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/kernel/suspend.c b/arch/arm64/kernel/suspend.c index fce95e17cf7f..1095aa483a1c 100644 --- a/arch/arm64/kernel/suspend.c +++ b/arch/arm64/kernel/suspend.c @@ -1,3 +1,4 @@ +#include #include #include #include @@ -70,6 +71,13 @@ int cpu_suspend(unsigned long arg, int (*fn)(unsigned long)) */ local_dbg_save(flags); + /* + * Function graph tracer state gets incosistent when the kernel + * calls functions that never return (aka suspend finishers) hence + * disable graph tracing during their execution. + */ + pause_graph_tracing(); + /* * mm context saved on the stack, it will be restored when * the cpu comes out of reset through the identity mapped @@ -111,6 +119,8 @@ int cpu_suspend(unsigned long arg, int (*fn)(unsigned long)) hw_breakpoint_restore(NULL); } + unpause_graph_tracing(); + /* * Restore pstate flags. OS lock and mdscr have been already * restored, so from this point onwards, debugging is fully From f16593034a30bcbef8f63c37d0f2b9e1a0902c2d Mon Sep 17 00:00:00 2001 From: Douglas Anderson Date: Tue, 20 Oct 2015 16:33:53 -0700 Subject: [PATCH 209/335] usb: dwc2: host: Fix ahbcfg for rk3066 The comment for ahbcfg for rk3066 parameters (also used for rk3288) claimed that ahbcfg was INCR16, but it wasn't. Since the bits weren't shifted properly, the 0x7 ended up being masked and we ended up programming 0x3 for the HBstLen. Let's set it to INCR16 properly. As per Wu Liang Feng at Rockchip this may increase transmission efficiency. I did blackbox tests with writing 0s to a USB-based SD reader (forcefully capping CPU Freq to try to measure efficiency): cd /sys/devices/system/cpu/cpu0/cpufreq echo userspace > scaling_governor echo 126000 > scaling_setspeed for i in $(seq 10); do dd if=/dev/zero of=/dev/sdb bs=1M count=750 done With the above tests I found that speeds went from ~15MB/s to ~18MB/s. Note that most other tests I did (including reading from the same USB reader) didn't show any difference in performance. Tested-by: Heiko Stuebner Acked-by: John Youn Reviewed-by: Liangfeng Wu Signed-off-by: Douglas Anderson Signed-off-by: Felipe Balbi --- drivers/usb/dwc2/platform.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/usb/dwc2/platform.c b/drivers/usb/dwc2/platform.c index 5859b0fa19ee..e61d773cf65e 100644 --- a/drivers/usb/dwc2/platform.c +++ b/drivers/usb/dwc2/platform.c @@ -108,7 +108,8 @@ static const struct dwc2_core_params params_rk3066 = { .host_ls_low_power_phy_clk = -1, .ts_dline = -1, .reload_ctl = -1, - .ahbcfg = 0x7, /* INCR16 */ + .ahbcfg = GAHBCFG_HBSTLEN_INCR16 << + GAHBCFG_HBSTLEN_SHIFT, .uframe_sched = -1, .external_id_pin_ctl = -1, .hibernation = -1, From 1fb7f12d5bec1b0a699c85d724bd5184a01fa2dd Mon Sep 17 00:00:00 2001 From: Douglas Anderson Date: Thu, 22 Oct 2015 13:05:03 -0700 Subject: [PATCH 210/335] usb: dwc2: host: Fix remote wakeup when not in DWC2_L2 In commit 734643dfbdde ("usb: dwc2: host: add flag to reflect bus state") we changed dwc2_port_suspend() not to set the lx_state anymore (instead it sets the new bus_suspended variable). This introduced a bug where we would fail to detect device insertions if: 1. Plug empty hub into dwc2 2. Plug USB flash drive into the empty hub. 3. Wait a few seconds 4. Unplug USB flash drive 5. Less than 2 seconds after step 4, plug the USB flash drive in again. The dwc2_hcd_rem_wakeup() function should have been changed to look at the new bus_suspended variable. Let's fix it. Since commit b46146d59fda ("usb: dwc2: host: resume root hub on remote wakeup") talks about needing the root hub resumed if the bus was suspended, we'll include it in our test. It appears that the "port_l1_change" should only be set to 1 if we were in DWC2_L1 (the driver currently never sets this), so we'll update the former "else" case based on this test. Fixes: 734643dfbdde ("usb: dwc2: host: add flag to reflect bus state") Acked-by: John Youn Tested-by: Gregory Herrero Reviewed-by: Heiko Stuebner Signed-off-by: Douglas Anderson Signed-off-by: Felipe Balbi --- drivers/usb/dwc2/hcd.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/usb/dwc2/hcd.c b/drivers/usb/dwc2/hcd.c index e79baf73c234..571c21727ff9 100644 --- a/drivers/usb/dwc2/hcd.c +++ b/drivers/usb/dwc2/hcd.c @@ -324,12 +324,13 @@ void dwc2_hcd_disconnect(struct dwc2_hsotg *hsotg) */ static void dwc2_hcd_rem_wakeup(struct dwc2_hsotg *hsotg) { - if (hsotg->lx_state == DWC2_L2) { + if (hsotg->bus_suspended) { hsotg->flags.b.port_suspend_change = 1; usb_hcd_resume_root_hub(hsotg->priv); - } else { - hsotg->flags.b.port_l1_change = 1; } + + if (hsotg->lx_state == DWC2_L1) + hsotg->flags.b.port_l1_change = 1; } /** @@ -1428,8 +1429,8 @@ static void dwc2_wakeup_detected(unsigned long data) dev_dbg(hsotg->dev, "Clear Resume: HPRT0=%0x\n", dwc2_readl(hsotg->regs + HPRT0)); - hsotg->bus_suspended = 0; dwc2_hcd_rem_wakeup(hsotg); + hsotg->bus_suspended = 0; /* Change to L0 state */ hsotg->lx_state = DWC2_L0; From 5e216d54b4862611c663e1e3e8ca1db1930ef948 Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Tue, 3 Nov 2015 15:57:38 +0800 Subject: [PATCH 211/335] usb: gadget: f_loopback: fix the warning during the enumeration The current code tries to allocate memory with GFP_KERNEL at interrupt context, it would show below warning during the enumeration when I test it with chipidea hardware, change GFP flag as GFP_ATOMIC can fix this issue. [ 40.438237] zero gadget: high-speed config #2: loopback [ 40.444924] ------------[ cut here ]------------ [ 40.449609] WARNING: CPU: 0 PID: 0 at kernel/locking/lockdep.c:2755 lockdep_trace_alloc+0x108/0x128() [ 40.461715] DEBUG_LOCKS_WARN_ON(irqs_disabled_flags(flags)) [ 40.467130] Modules linked in: [ 40.470216] usb_f_ss_lb g_zero libcomposite evbug [ 40.473822] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.3.0-rc5-00168-gb730aaf #604 [ 40.481496] Hardware name: Freescale i.MX6 SoloX (Device Tree) [ 40.487345] Backtrace: [ 40.489857] [<80014e94>] (dump_backtrace) from [<80015088>] (show_stack+0x18/0x1c) [ 40.497445] r6:80b67a80 r5:00000000 r4:00000000 r3:00000000 [ 40.503234] [<80015070>] (show_stack) from [<802e27b4>] (dump_stack+0x8c/0xa4) [ 40.510503] [<802e2728>] (dump_stack) from [<8002cfe8>] (warn_slowpath_common+0x80/0xbc) [ 40.518612] r6:8007510c r5:00000009 r4:80b49c88 r3:00000001 [ 40.524396] [<8002cf68>] (warn_slowpath_common) from [<8002d05c>] (warn_slowpath_fmt+0x38/0x40) [ 40.533109] r8:bcfdef80 r7:bdb705cc r6:000080d0 r5:be001e80 r4:809cc278 [ 40.539965] [<8002d028>] (warn_slowpath_fmt) from [<8007510c>] (lockdep_trace_alloc+0x108/0x128) [ 40.548766] r3:809d0128 r2:809cc278 [ 40.552401] r4:600b0193 [ 40.554990] [<80075004>] (lockdep_trace_alloc) from [<801093d4>] (kmem_cache_alloc+0x28/0x15c) [ 40.563618] r4:000080d0 r3:80b4aa8c [ 40.567270] [<801093ac>] (kmem_cache_alloc) from [<804d95e4>] (ep_alloc_request+0x58/0x68) [ 40.575550] r10:7f01f104 r9:00000001 r8:bcfdef80 r7:bdb705cc r6:bc178700 r5:00000000 [ 40.583512] r4:bcfdef80 r3:813c0a38 [ 40.587183] [<804d958c>] (ep_alloc_request) from [<7f01f7ec>] (loopback_set_alt+0x114/0x21c [usb_f_ss_lb]) [ 40.596929] [<7f01f6d8>] (loopback_set_alt [usb_f_ss_lb]) from [<7f006910>] (composite_setup+0xbd0/0x17e8 [libcomposite]) [ 40.607902] r10:bd3a2c0c r9:00000000 r8:bcfdef80 r7:bc178700 r6:bdb702d0 r5:bcfdefdc [ 40.615866] r4:7f0199b4 r3:00000002 [ 40.619542] [<7f005d40>] (composite_setup [libcomposite]) from [<804dae88>] (udc_irq+0x784/0xd1c) [ 40.628431] r10:80bb5619 r9:c0876140 r8:00012001 r7:bdb71010 r6:bdb70568 r5:00010001 [ 40.636392] r4:bdb70014 [ 40.638985] [<804da704>] (udc_irq) from [<804d64f8>] (ci_irq+0x5c/0x118) [ 40.645702] r10:80bb5619 r9:be11e000 r8:00000117 r7:00000000 r6:bdb71010 r5:be11e060 [ 40.653666] r4:bdb70010 [ 40.656261] [<804d649c>] (ci_irq) from [<8007f638>] (handle_irq_event_percpu+0x7c/0x13c) [ 40.664367] r6:00000000 r5:be11e060 r4:bdb05cc0 r3:804d649c [ 40.670149] [<8007f5bc>] (handle_irq_event_percpu) from [<8007f740>] (handle_irq_event+0x48/0x6c) [ 40.679036] r10:00000000 r9:be008000 r8:00000001 r7:00000000 r6:bdb05cc0 r5:be11e060 [ 40.686998] r4:be11e000 [ 40.689581] [<8007f6f8>] (handle_irq_event) from [<80082850>] (handle_fasteoi_irq+0xd4/0x1b0) [ 40.698120] r6:80b56a30 r5:be11e060 r4:be11e000 r3:00000000 [ 40.703898] [<8008277c>] (handle_fasteoi_irq) from [<8007ec04>] (generic_handle_irq+0x28/0x3c) [ 40.712524] r7:00000000 r6:80b4aaf4 r5:00000117 r4:80b445fc [ 40.718304] [<8007ebdc>] (generic_handle_irq) from [<8007ef20>] (__handle_domain_irq+0x6c/0xe8) [ 40.727033] [<8007eeb4>] (__handle_domain_irq) from [<800095d4>] (gic_handle_irq+0x48/0x94) [ 40.735402] r9:c080f100 r8:80b4ac6c r7:c080e100 r6:80b67d40 r5:80b49f00 r4:c080e10c [ 40.743290] [<8000958c>] (gic_handle_irq) from [<80015d38>] (__irq_svc+0x58/0x78) [ 40.750791] Exception stack(0x80b49f00 to 0x80b49f48) [ 40.755873] 9f00: 00000001 00000001 00000000 80024320 80b48000 80b4a9d0 80b4a984 80b433e4 [ 40.764078] 9f20: 00000001 807f4680 00000000 80b49f5c 80b49f20 80b49f50 80071ca4 800113fc [ 40.772272] 9f40: 200b0013 ffffffff [ 40.775776] r9:807f4680 r8:00000001 r7:80b49f34 r6:ffffffff r5:200b0013 r4:800113fc [ 40.783677] [<800113d4>] (arch_cpu_idle) from [<8006c5bc>] (default_idle_call+0x28/0x38) [ 40.791798] [<8006c594>] (default_idle_call) from [<8006c6dc>] (cpu_startup_entry+0x110/0x1b0) [ 40.800445] [<8006c5cc>] (cpu_startup_entry) from [<807e95dc>] (rest_init+0x12c/0x168) [ 40.808376] r7:80b4a8c0 r3:807f4b7c [ 40.812030] [<807e94b0>] (rest_init) from [<80ad7cc0>] (start_kernel+0x360/0x3d4) [ 40.819528] r5:80bcb000 r4:80bcb050 [ 40.823171] [<80ad7960>] (start_kernel) from [<8000807c>] (0x8000807c) It fixes commit 91c42b0da8e3 ("usb: gadget: loopback: Fix looping back logic implementation"). Cc: # v3.18+ Signed-off-by: Peter Chen Reviewed-by: Krzysztof Opasiak Signed-off-by: Felipe Balbi --- drivers/usb/gadget/function/f_loopback.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/gadget/function/f_loopback.c b/drivers/usb/gadget/function/f_loopback.c index 23933bdf2d9d..ddc3aad886b7 100644 --- a/drivers/usb/gadget/function/f_loopback.c +++ b/drivers/usb/gadget/function/f_loopback.c @@ -329,7 +329,7 @@ static int alloc_requests(struct usb_composite_dev *cdev, for (i = 0; i < loop->qlen && result == 0; i++) { result = -ENOMEM; - in_req = usb_ep_alloc_request(loop->in_ep, GFP_KERNEL); + in_req = usb_ep_alloc_request(loop->in_ep, GFP_ATOMIC); if (!in_req) goto fail; From 68fe05e2a45181ce02ab2698930b37f20487bd24 Mon Sep 17 00:00:00 2001 From: Bin Liu Date: Fri, 6 Nov 2015 12:08:56 -0600 Subject: [PATCH 212/335] usb: musb: fix tx fifo flush handling Here are a few changes in musb_h_tx_flush_fifo(). - It has been observed that sometimes (if not always) musb is unable to flush tx fifo during urb dequeue when disconnect a device. But it seems to be harmless, since the tx fifo flush is done again in musb_ep_program() when re-use the hw_ep. But the WARN() floods the console in the case when multiple tx urbs are queued, so change it to dev_WARN_ONCE(). - applications could queue up many tx urbs, then the 1ms delay could causes minutes of delay in device disconnect. So remove it to get better user experience. The 1ms delay does not help the flushing anyway. - cleanup the debug code - related to lastcsr. ---- Note: The tx fifo flush issue has been observed during device disconnect on AM335x. To reproduce the issue, ensure tx urb(s) are queued when unplug the usb device which is connected to AM335x usb host port. I found using a usb-ethernet device and running iperf (client on AM335x) has very high chance to trigger the problem. Better to turn on dev_dbg() in musb_cleanup_urb() with CPPI enabled to see the issue when aborting the tx channel. Signed-off-by: Bin Liu Signed-off-by: Felipe Balbi --- drivers/usb/musb/musb_host.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c index 26c65e66cc0f..795a45b1b25b 100644 --- a/drivers/usb/musb/musb_host.c +++ b/drivers/usb/musb/musb_host.c @@ -112,22 +112,32 @@ static void musb_h_tx_flush_fifo(struct musb_hw_ep *ep) struct musb *musb = ep->musb; void __iomem *epio = ep->regs; u16 csr; - u16 lastcsr = 0; int retries = 1000; csr = musb_readw(epio, MUSB_TXCSR); while (csr & MUSB_TXCSR_FIFONOTEMPTY) { - if (csr != lastcsr) - dev_dbg(musb->controller, "Host TX FIFONOTEMPTY csr: %02x\n", csr); - lastcsr = csr; csr |= MUSB_TXCSR_FLUSHFIFO | MUSB_TXCSR_TXPKTRDY; musb_writew(epio, MUSB_TXCSR, csr); csr = musb_readw(epio, MUSB_TXCSR); - if (WARN(retries-- < 1, + + /* + * FIXME: sometimes the tx fifo flush failed, it has been + * observed during device disconnect on AM335x. + * + * To reproduce the issue, ensure tx urb(s) are queued when + * unplug the usb device which is connected to AM335x usb + * host port. + * + * I found using a usb-ethernet device and running iperf + * (client on AM335x) has very high chance to trigger it. + * + * Better to turn on dev_dbg() in musb_cleanup_urb() with + * CPPI enabled to see the issue when aborting the tx channel. + */ + if (dev_WARN_ONCE(musb->controller, retries-- < 1, "Could not flush host TX%d fifo: csr: %04x\n", ep->epnum, csr)) return; - mdelay(1); } } From b9e51b2b1fda19143f48d182ed7a2943f21e1ae4 Mon Sep 17 00:00:00 2001 From: Ben McCauley Date: Mon, 16 Nov 2015 10:47:24 -0600 Subject: [PATCH 213/335] usb: dwc3: gadget: let us set lower max_speed In some SoCs, dwc3 is implemented as a USB2.0 only core, meaning that it can't ever achieve SuperSpeed. Currect driver always sets gadget.max_speed to USB_SPEED_SUPER unconditionally. This can causes issues to some Host stacks where the host will issue a GetBOS() request and we will reply with a BOS containing Superspeed Capability Descriptor. At least Windows seems to be upset by this fact and prints a warning that we should connect $this device to another port. [ balbi@ti.com : rewrote entire commit, including source code comment to make a lot clearer what the problem is ] Cc: Signed-off-by: Ben McCauley Signed-off-by: Felipe Balbi --- drivers/usb/dwc3/gadget.c | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c index 55ba447fdf8b..e24a01cc98df 100644 --- a/drivers/usb/dwc3/gadget.c +++ b/drivers/usb/dwc3/gadget.c @@ -2744,11 +2744,33 @@ int dwc3_gadget_init(struct dwc3 *dwc) } dwc->gadget.ops = &dwc3_gadget_ops; - dwc->gadget.max_speed = USB_SPEED_SUPER; dwc->gadget.speed = USB_SPEED_UNKNOWN; dwc->gadget.sg_supported = true; dwc->gadget.name = "dwc3-gadget"; + /* + * FIXME We might be setting max_speed to revision < DWC3_REVISION_220A) + dwc3_trace(trace_dwc3_gadget, + "Changing max_speed on rev %08x\n", + dwc->revision); + + dwc->gadget.max_speed = dwc->maximum_speed; + /* * Per databook, DWC3 needs buffer size to be aligned to MaxPacketSize * on ep out. From 89d99aea94a5bf8eae530cd77df40904d382946c Mon Sep 17 00:00:00 2001 From: LABBE Corentin Date: Thu, 12 Nov 2015 08:45:52 +0100 Subject: [PATCH 214/335] usb: phy: phy-mxs-usb: fix a possible NULL dereference of_match_device could return NULL, and so cause a NULL pointer dereference later. Signed-off-by: LABBE Corentin Signed-off-by: Felipe Balbi --- drivers/usb/phy/phy-mxs-usb.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/usb/phy/phy-mxs-usb.c b/drivers/usb/phy/phy-mxs-usb.c index 4d863ebc117c..b7536af777ab 100644 --- a/drivers/usb/phy/phy-mxs-usb.c +++ b/drivers/usb/phy/phy-mxs-usb.c @@ -452,10 +452,13 @@ static int mxs_phy_probe(struct platform_device *pdev) struct clk *clk; struct mxs_phy *mxs_phy; int ret; - const struct of_device_id *of_id = - of_match_device(mxs_phy_dt_ids, &pdev->dev); + const struct of_device_id *of_id; struct device_node *np = pdev->dev.of_node; + of_id = of_match_device(mxs_phy_dt_ids, &pdev->dev); + if (!of_id) + return -ENODEV; + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); base = devm_ioremap_resource(&pdev->dev, res); if (IS_ERR(base)) From 51676c8d6daac6dd11f768cb26db8d5b572f65b4 Mon Sep 17 00:00:00 2001 From: Bin Liu Date: Fri, 13 Nov 2015 15:45:24 -0600 Subject: [PATCH 215/335] usb: musb: enable usb_dma parameter Change the permission of usb_dma parameter so it can be used for runtime debug without reboot. Signed-off-by: Bin Liu Signed-off-by: Felipe Balbi --- drivers/usb/musb/musb_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c index 3de9087ecad4..18cfc0a361cb 100644 --- a/drivers/usb/musb/musb_core.c +++ b/drivers/usb/musb/musb_core.c @@ -1668,7 +1668,7 @@ EXPORT_SYMBOL_GPL(musb_interrupt); static bool use_dma = 1; /* "modprobe ... use_dma=0" etc */ -module_param(use_dma, bool, 0); +module_param(use_dma, bool, 0644); MODULE_PARM_DESC(use_dma, "enable/disable use of DMA"); void musb_dma_completion(struct musb *musb, u8 epnum, u8 transmit) From d134c48d889ddceadf4c990e6f3df16b816ed5d4 Mon Sep 17 00:00:00 2001 From: Douglas Gilbert Date: Mon, 16 Nov 2015 19:22:08 +0100 Subject: [PATCH 216/335] usb: gadget: atmel_usba_udc: Expose correct device speed Following changes that appeared in lk 4.0.0, the gadget udc driver for some ARM based Atmel SoCs (e.g. at91sam9x5 and sama5d3 families) incorrectly deduced full-speed USB link speed even when the hardware had negotiated a high-speed link. The fix is to make sure that the UDPHS Interrupt Enable Register value does not mask the SPEED bit in the Interrupt Status Register. For a mass storage gadget this problem lead to failures when the host had a USB 3 port with the xhci_hcd driver. If the host was a USB 2 port using the ehci_hcd driver then the mass storage gadget worked (but probably at a lower speed than it should have). Signed-off-by: Douglas Gilbert Reviewed-by: Boris Brezillon Cc: #4.0+ Fixes: 9870d895ad87 ("usb: atmel_usba_udc: Mask status with enabled irqs") Signed-off-by: Felipe Balbi --- drivers/usb/gadget/udc/atmel_usba_udc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.c b/drivers/usb/gadget/udc/atmel_usba_udc.c index f0f2b066ac08..f92f5aff0dd5 100644 --- a/drivers/usb/gadget/udc/atmel_usba_udc.c +++ b/drivers/usb/gadget/udc/atmel_usba_udc.c @@ -1633,7 +1633,7 @@ static irqreturn_t usba_udc_irq(int irq, void *devid) spin_lock(&udc->lock); int_enb = usba_int_enb_get(udc); - status = usba_readl(udc, INT_STA) & int_enb; + status = usba_readl(udc, INT_STA) & (int_enb | USBA_HIGH_SPEED); DBG(DBG_INT, "irq, status=%#08x\n", status); if (status & USBA_DET_SUSPEND) { From 6c606fa32c4272894248c1abbc15257fc04250be Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Timo=20Ter=C3=A4s?= Date: Mon, 16 Nov 2015 14:36:32 +0200 Subject: [PATCH 217/335] via-velocity: unconditionally drop frames with bad l2 length MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit By default the driver allowed incorrect frames to be received. What is worse the code does not handle very short frames correctly. The FCS length is unconditionally subtracted, and the underflow can cause skb_put to be called with large number after implicit cast to unsigned. And indeed, an skb_over_panic() was observed with via-velocity. This removes the module parameter as it does not work in it's current state, and should be implemented via NETIF_F_RXALL if needed. Suggested-by: Francois Romieu Signed-off-by: Timo Teräs Signed-off-by: David S. Miller --- drivers/net/ethernet/via/via-velocity.c | 24 +++--------------------- 1 file changed, 3 insertions(+), 21 deletions(-) diff --git a/drivers/net/ethernet/via/via-velocity.c b/drivers/net/ethernet/via/via-velocity.c index ae68afd50a15..f38696ceee74 100644 --- a/drivers/net/ethernet/via/via-velocity.c +++ b/drivers/net/ethernet/via/via-velocity.c @@ -345,13 +345,6 @@ VELOCITY_PARAM(flow_control, "Enable flow control ability"); */ VELOCITY_PARAM(speed_duplex, "Setting the speed and duplex mode"); -#define VAL_PKT_LEN_DEF 0 -/* ValPktLen[] is used for setting the checksum offload ability of NIC. - 0: Receive frame with invalid layer 2 length (Default) - 1: Drop frame with invalid layer 2 length -*/ -VELOCITY_PARAM(ValPktLen, "Receiving or Drop invalid 802.3 frame"); - #define WOL_OPT_DEF 0 #define WOL_OPT_MIN 0 #define WOL_OPT_MAX 7 @@ -494,7 +487,6 @@ static void velocity_get_options(struct velocity_opt *opts, int index, velocity_set_int_opt(&opts->flow_cntl, flow_control[index], FLOW_CNTL_MIN, FLOW_CNTL_MAX, FLOW_CNTL_DEF, "flow_control", devname); velocity_set_bool_opt(&opts->flags, IP_byte_align[index], IP_ALIG_DEF, VELOCITY_FLAGS_IP_ALIGN, "IP_byte_align", devname); - velocity_set_bool_opt(&opts->flags, ValPktLen[index], VAL_PKT_LEN_DEF, VELOCITY_FLAGS_VAL_PKT_LEN, "ValPktLen", devname); velocity_set_int_opt((int *) &opts->spd_dpx, speed_duplex[index], MED_LNK_MIN, MED_LNK_MAX, MED_LNK_DEF, "Media link mode", devname); velocity_set_int_opt(&opts->wol_opts, wol_opts[index], WOL_OPT_MIN, WOL_OPT_MAX, WOL_OPT_DEF, "Wake On Lan options", devname); opts->numrx = (opts->numrx & ~3); @@ -2055,8 +2047,9 @@ static int velocity_receive_frame(struct velocity_info *vptr, int idx) int pkt_len = le16_to_cpu(rd->rdesc0.len) & 0x3fff; struct sk_buff *skb; - if (rd->rdesc0.RSR & (RSR_STP | RSR_EDP)) { - VELOCITY_PRT(MSG_LEVEL_VERBOSE, KERN_ERR " %s : the received frame spans multiple RDs.\n", vptr->netdev->name); + if (unlikely(rd->rdesc0.RSR & (RSR_STP | RSR_EDP | RSR_RL))) { + if (rd->rdesc0.RSR & (RSR_STP | RSR_EDP)) + VELOCITY_PRT(MSG_LEVEL_VERBOSE, KERN_ERR " %s : the received frame spans multiple RDs.\n", vptr->netdev->name); stats->rx_length_errors++; return -EINVAL; } @@ -2069,17 +2062,6 @@ static int velocity_receive_frame(struct velocity_info *vptr, int idx) dma_sync_single_for_cpu(vptr->dev, rd_info->skb_dma, vptr->rx.buf_sz, DMA_FROM_DEVICE); - /* - * Drop frame not meeting IEEE 802.3 - */ - - if (vptr->flags & VELOCITY_FLAGS_VAL_PKT_LEN) { - if (rd->rdesc0.RSR & RSR_RL) { - stats->rx_length_errors++; - return -EINVAL; - } - } - velocity_rx_csum(rd, skb); if (velocity_rx_copy(&skb, pkt_len, vptr) < 0) { From a6e18ff111701b4ff6947605bfbe9594ec42a6e8 Mon Sep 17 00:00:00 2001 From: Vlad Yasevich Date: Mon, 16 Nov 2015 15:43:44 -0500 Subject: [PATCH 218/335] vlan: Fix untag operations of stacked vlans with REORDER_HEADER off When we have multiple stacked vlan devices all of which have turned off REORDER_HEADER flag, the untag operation does not locate the ethernet addresses correctly for nested vlans. The reason is that in case of REORDER_HEADER flag being off, the outer vlan headers are put back and the mac_len is adjusted to account for the presense of the header. Then, the subsequent untag operation, for the next level vlan, always use VLAN_ETH_HLEN to locate the begining of the ethernet header and that ends up being a multiple of 4 bytes short of the actuall beginning of the mac header (the multiple depending on the how many vlan encapsulations ethere are). As a reslult, if there are multiple levles of vlan devices with REODER_HEADER being off, the recevied packets end up being dropped. To solve this, we use skb->mac_len as the offset. The value is always set on receive path and starts out as a ETH_HLEN. The value is also updated when the vlan header manupations occur so we know it will be correct. Signed-off-by: Vladislav Yasevich Signed-off-by: David S. Miller --- net/core/skbuff.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index aa41e6dd6429..152b9c70e252 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -4268,7 +4268,8 @@ static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb) return NULL; } - memmove(skb->data - ETH_HLEN, skb->data - VLAN_ETH_HLEN, 2 * ETH_ALEN); + memmove(skb->data - ETH_HLEN, skb->data - skb->mac_len, + 2 * ETH_ALEN); skb->mac_header += VLAN_HLEN; return skb; } From 28f9ee22bcdd84726dbf6267d0b58f254166b900 Mon Sep 17 00:00:00 2001 From: Vlad Yasevich Date: Mon, 16 Nov 2015 15:43:45 -0500 Subject: [PATCH 219/335] vlan: Do not put vlan headers back on bridge and macvlan ports When a vlan is configured with REORDER_HEADER set to 0, the vlan header is put back into the packet and makes it appear that the vlan header is still there even after it's been processed. This posses a problem for bridge and macvlan ports. The packets passed to those device may be forwarded and at the time of the forward, vlan headers end up being unexpectedly present. With the patch, we make sure that we do not put the vlan header back (when REORDER_HEADER is 0) if a bridge or macvlan has been configured on top of the vlan device. Signed-off-by: Vladislav Yasevich Signed-off-by: David S. Miller --- include/linux/netdevice.h | 5 +++++ net/8021q/vlan_core.c | 4 +++- 2 files changed, 8 insertions(+), 1 deletion(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index cc221b967687..67bfac1abfc1 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -3857,6 +3857,11 @@ static inline bool netif_is_bridge_master(const struct net_device *dev) return dev->priv_flags & IFF_EBRIDGE; } +static inline bool netif_is_bridge_port(const struct net_device *dev) +{ + return dev->priv_flags & IFF_BRIDGE_PORT; +} + static inline bool netif_is_ovs_master(const struct net_device *dev) { return dev->priv_flags & IFF_OPENVSWITCH; diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c index 496b27588493..e2ed69850489 100644 --- a/net/8021q/vlan_core.c +++ b/net/8021q/vlan_core.c @@ -30,7 +30,9 @@ bool vlan_do_receive(struct sk_buff **skbp) skb->pkt_type = PACKET_HOST; } - if (!(vlan_dev_priv(vlan_dev)->flags & VLAN_FLAG_REORDER_HDR)) { + if (!(vlan_dev_priv(vlan_dev)->flags & VLAN_FLAG_REORDER_HDR) && + !netif_is_macvlan_port(vlan_dev) && + !netif_is_bridge_port(vlan_dev)) { unsigned int offset = skb->data - skb_mac_header(skb); /* From cf554ada0be7077906aa9a17faf151ff66e3cb8e Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Mon, 16 Nov 2015 22:34:26 +0100 Subject: [PATCH 220/335] ipvlan: fix leak in ipvlan_rcv_frame Pass a **skb to ipvlan_rcv_frame so that if skb_share_check returns a new skb, we actually use it during further processing. It's safe to ignore the new skb in the ipvlan_xmit_* functions, because they call ipvlan_rcv_frame with local == true, so that dev_forward_skb is called and always takes ownership of the skb. Fixes: 2ad7bf363841 ("ipvlan: Initial check-in of the IPVLAN driver.") Signed-off-by: Sabrina Dubroca Signed-off-by: David S. Miller --- drivers/net/ipvlan/ipvlan_core.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/net/ipvlan/ipvlan_core.c b/drivers/net/ipvlan/ipvlan_core.c index d50887e3df6d..a9148e46dad0 100644 --- a/drivers/net/ipvlan/ipvlan_core.c +++ b/drivers/net/ipvlan/ipvlan_core.c @@ -254,7 +254,7 @@ acct: } } -static int ipvlan_rcv_frame(struct ipvl_addr *addr, struct sk_buff *skb, +static int ipvlan_rcv_frame(struct ipvl_addr *addr, struct sk_buff **pskb, bool local) { struct ipvl_dev *ipvlan = addr->master; @@ -262,6 +262,7 @@ static int ipvlan_rcv_frame(struct ipvl_addr *addr, struct sk_buff *skb, unsigned int len; rx_handler_result_t ret = RX_HANDLER_CONSUMED; bool success = false; + struct sk_buff *skb = *pskb; len = skb->len + ETH_HLEN; if (unlikely(!(dev->flags & IFF_UP))) { @@ -273,6 +274,7 @@ static int ipvlan_rcv_frame(struct ipvl_addr *addr, struct sk_buff *skb, if (!skb) goto out; + *pskb = skb; skb->dev = dev; skb->pkt_type = PACKET_HOST; @@ -486,7 +488,7 @@ static int ipvlan_xmit_mode_l3(struct sk_buff *skb, struct net_device *dev) addr = ipvlan_addr_lookup(ipvlan->port, lyr3h, addr_type, true); if (addr) - return ipvlan_rcv_frame(addr, skb, true); + return ipvlan_rcv_frame(addr, &skb, true); out: skb->dev = ipvlan->phy_dev; @@ -506,7 +508,7 @@ static int ipvlan_xmit_mode_l2(struct sk_buff *skb, struct net_device *dev) if (lyr3h) { addr = ipvlan_addr_lookup(ipvlan->port, lyr3h, addr_type, true); if (addr) - return ipvlan_rcv_frame(addr, skb, true); + return ipvlan_rcv_frame(addr, &skb, true); } skb = skb_share_check(skb, GFP_ATOMIC); if (!skb) @@ -589,7 +591,7 @@ static rx_handler_result_t ipvlan_handle_mode_l3(struct sk_buff **pskb, addr = ipvlan_addr_lookup(port, lyr3h, addr_type, true); if (addr) - ret = ipvlan_rcv_frame(addr, skb, false); + ret = ipvlan_rcv_frame(addr, pskb, false); out: return ret; @@ -626,7 +628,7 @@ static rx_handler_result_t ipvlan_handle_mode_l2(struct sk_buff **pskb, addr = ipvlan_addr_lookup(port, lyr3h, addr_type, true); if (addr) - ret = ipvlan_rcv_frame(addr, skb, false); + ret = ipvlan_rcv_frame(addr, pskb, false); } return ret; From a534dc529853c69e94994aa47c1d80a03ce2c11d Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Mon, 16 Nov 2015 22:44:53 +0100 Subject: [PATCH 221/335] ipvlan: fix use after free of skb ipvlan_handle_frame is a rx_handler, and when it returns a value other than RX_HANDLER_CONSUMED (here, NET_RX_DROP aka RX_HANDLER_ANOTHER), __netif_receive_skb_core expects that the skb still exists and will process it further, but we just freed it. Fixes: 2ad7bf363841 ("ipvlan: Initial check-in of the IPVLAN driver.") Signed-off-by: Sabrina Dubroca Signed-off-by: David S. Miller --- drivers/net/ipvlan/ipvlan_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ipvlan/ipvlan_core.c b/drivers/net/ipvlan/ipvlan_core.c index a9148e46dad0..8c48bb2a94ea 100644 --- a/drivers/net/ipvlan/ipvlan_core.c +++ b/drivers/net/ipvlan/ipvlan_core.c @@ -653,5 +653,5 @@ rx_handler_result_t ipvlan_handle_frame(struct sk_buff **pskb) WARN_ONCE(true, "ipvlan_handle_frame() called for mode = [%hx]\n", port->mode); kfree_skb(skb); - return NET_RX_DROP; + return RX_HANDLER_CONSUMED; } From e639b8d8a7a728f0b05ef2df6cb6b45dc3d4e556 Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Mon, 16 Nov 2015 22:54:20 +0100 Subject: [PATCH 222/335] macvlan: fix leak in macvlan_handle_frame Reset pskb in macvlan_handle_frame in case skb_share_check returned a clone. Fixes: 8a4eb5734e8d ("net: introduce rx_handler results and logic around that") Signed-off-by: Sabrina Dubroca Signed-off-by: David S. Miller --- drivers/net/macvlan.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index 86f6c6292c27..06c8bfeaccd6 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -415,6 +415,7 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb) skb = ip_check_defrag(dev_net(skb->dev), skb, IP_DEFRAG_MACVLAN); if (!skb) return RX_HANDLER_CONSUMED; + *pskb = skb; eth = eth_hdr(skb); macvlan_forward_source(skb, port, eth->h_source); src = macvlan_hash_lookup(port, eth->h_source); @@ -456,6 +457,7 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb) goto out; } + *pskb = skb; skb->dev = dev; skb->pkt_type = PACKET_HOST; From ec0738db8d7a9bd6653d9f9d81b201a519144d63 Mon Sep 17 00:00:00 2001 From: Yang Shi Date: Mon, 16 Nov 2015 14:35:35 -0800 Subject: [PATCH 223/335] arm64: bpf: make BPF prologue and epilogue align with ARM64 AAPCS Save and restore FP/LR in BPF prog prologue and epilogue, save SP to FP in prologue in order to get the correct stack backtrace. However, ARM64 JIT used FP (x29) as eBPF fp register, FP is subjected to change during function call so it may cause the BPF prog stack base address change too. Use x25 to replace FP as BPF stack base register (fp). Since x25 is callee saved register, so it will keep intact during function call. It is initialized in BPF prog prologue when BPF prog is started to run everytime. Save and restore x25/x26 in BPF prologue and epilogue to keep them intact for the outside of BPF. Actually, x26 is unnecessary, but SP requires 16 bytes alignment. So, the BPF stack layout looks like: high original A64_SP => 0:+-----+ BPF prologue |FP/LR| current A64_FP => -16:+-----+ | ... | callee saved registers +-----+ | | x25/x26 BPF fp register => -80:+-----+ | | | ... | BPF prog stack | | | | current A64_SP => +-----+ | | | ... | Function call stack | | +-----+ low CC: Zi Shen Lim CC: Xi Wang Signed-off-by: Yang Shi Acked-by: Zi Shen Lim Signed-off-by: David S. Miller --- arch/arm64/net/bpf_jit_comp.c | 44 +++++++++++++++++++++++++++++++---- 1 file changed, 39 insertions(+), 5 deletions(-) diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index 64a8bc12f4a2..a6905b8d12d3 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -50,7 +50,7 @@ static const int bpf2a64[] = { [BPF_REG_8] = A64_R(21), [BPF_REG_9] = A64_R(22), /* read-only frame pointer to access stack */ - [BPF_REG_FP] = A64_FP, + [BPF_REG_FP] = A64_R(25), /* temporary register for internal BPF JIT */ [TMP_REG_1] = A64_R(23), [TMP_REG_2] = A64_R(24), @@ -155,16 +155,47 @@ static void build_prologue(struct jit_ctx *ctx) stack_size += 4; /* extra for skb_copy_bits buffer */ stack_size = STACK_ALIGN(stack_size); + /* + * BPF prog stack layout + * + * high + * original A64_SP => 0:+-----+ BPF prologue + * |FP/LR| + * current A64_FP => -16:+-----+ + * | ... | callee saved registers + * +-----+ + * | | x25/x26 + * BPF fp register => -80:+-----+ + * | | + * | ... | BPF prog stack + * | | + * | | + * current A64_SP => +-----+ + * | | + * | ... | Function call stack + * | | + * +-----+ + * low + * + */ + + /* Save FP and LR registers to stay align with ARM64 AAPCS */ + emit(A64_PUSH(A64_FP, A64_LR, A64_SP), ctx); + emit(A64_MOV(1, A64_FP, A64_SP), ctx); + /* Save callee-saved register */ emit(A64_PUSH(r6, r7, A64_SP), ctx); emit(A64_PUSH(r8, r9, A64_SP), ctx); if (ctx->tmp_used) emit(A64_PUSH(tmp1, tmp2, A64_SP), ctx); - /* Set up frame pointer */ + /* Save fp (x25) and x26. SP requires 16 bytes alignment */ + emit(A64_PUSH(fp, A64_R(26), A64_SP), ctx); + + /* Set up BPF prog stack base register (x25) */ emit(A64_MOV(1, fp, A64_SP), ctx); - /* Set up BPF stack */ + /* Set up function call stack */ emit(A64_SUB_I(1, A64_SP, A64_SP, stack_size), ctx); /* Clear registers A and X */ @@ -190,14 +221,17 @@ static void build_epilogue(struct jit_ctx *ctx) /* We're done with BPF stack */ emit(A64_ADD_I(1, A64_SP, A64_SP, stack_size), ctx); + /* Restore fs (x25) and x26 */ + emit(A64_POP(fp, A64_R(26), A64_SP), ctx); + /* Restore callee-saved register */ if (ctx->tmp_used) emit(A64_POP(tmp1, tmp2, A64_SP), ctx); emit(A64_POP(r8, r9, A64_SP), ctx); emit(A64_POP(r6, r7, A64_SP), ctx); - /* Restore frame pointer */ - emit(A64_MOV(1, fp, A64_SP), ctx); + /* Restore FP/LR registers */ + emit(A64_POP(A64_FP, A64_LR, A64_SP), ctx); /* Set return value */ emit(A64_MOV(1, A64_R(0), r0), ctx); From 819ec8e1f349f73bdf65bf33a364538e59007a9a Mon Sep 17 00:00:00 2001 From: Andrew Lunn Date: Mon, 16 Nov 2015 23:34:41 +0100 Subject: [PATCH 224/335] phy: marvell: Add support for 88E1540 PHY The 88E1540 can be found embedded in the Marvell 88E6352 switch. It is compatible with the 88E1510, so add support for it, using the 88E1510 specific functions. Signed-off-by: Andrew Lunn Signed-off-by: David S. Miller --- drivers/net/phy/marvell.c | 16 ++++++++++++++++ include/linux/marvell_phy.h | 1 + 2 files changed, 17 insertions(+) diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c index 5de8d5827536..0240552b50f3 100644 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@ -1153,6 +1153,21 @@ static struct phy_driver marvell_drivers[] = { .suspend = &genphy_suspend, .driver = { .owner = THIS_MODULE }, }, + { + .phy_id = MARVELL_PHY_ID_88E1540, + .phy_id_mask = MARVELL_PHY_ID_MASK, + .name = "Marvell 88E1540", + .features = PHY_GBIT_FEATURES, + .flags = PHY_HAS_INTERRUPT, + .config_aneg = &m88e1510_config_aneg, + .read_status = &marvell_read_status, + .ack_interrupt = &marvell_ack_interrupt, + .config_intr = &marvell_config_intr, + .did_interrupt = &m88e1121_did_interrupt, + .resume = &genphy_resume, + .suspend = &genphy_suspend, + .driver = { .owner = THIS_MODULE }, + }, { .phy_id = MARVELL_PHY_ID_88E3016, .phy_id_mask = MARVELL_PHY_ID_MASK, @@ -1186,6 +1201,7 @@ static struct mdio_device_id __maybe_unused marvell_tbl[] = { { MARVELL_PHY_ID_88E1318S, MARVELL_PHY_ID_MASK }, { MARVELL_PHY_ID_88E1116R, MARVELL_PHY_ID_MASK }, { MARVELL_PHY_ID_88E1510, MARVELL_PHY_ID_MASK }, + { MARVELL_PHY_ID_88E1540, MARVELL_PHY_ID_MASK }, { MARVELL_PHY_ID_88E3016, MARVELL_PHY_ID_MASK }, { } }; diff --git a/include/linux/marvell_phy.h b/include/linux/marvell_phy.h index e6982ac3200d..a57f0dfb6db7 100644 --- a/include/linux/marvell_phy.h +++ b/include/linux/marvell_phy.h @@ -16,6 +16,7 @@ #define MARVELL_PHY_ID_88E1318S 0x01410e90 #define MARVELL_PHY_ID_88E1116R 0x01410e40 #define MARVELL_PHY_ID_88E1510 0x01410dd0 +#define MARVELL_PHY_ID_88E1540 0x01410eb0 #define MARVELL_PHY_ID_88E3016 0x01410e60 /* struct phy_device dev_flags definitions */ From 321beec5047af83db90c88114b7e664b156f49fe Mon Sep 17 00:00:00 2001 From: Andrew Lunn Date: Mon, 16 Nov 2015 23:36:46 +0100 Subject: [PATCH 225/335] net: phy: Use interrupts when available in NOLINK state The NOLINK state will poll the phy once a second to see if the link has come up. If the phy has an interrupt line, this polling can be skipped, since the phy should interrupt when the link returns. Signed-off-by: Andrew Lunn Signed-off-by: David S. Miller --- drivers/net/phy/phy.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c index adb48abafc87..48ce6ef400fe 100644 --- a/drivers/net/phy/phy.c +++ b/drivers/net/phy/phy.c @@ -863,6 +863,9 @@ void phy_state_machine(struct work_struct *work) needs_aneg = true; break; case PHY_NOLINK: + if (phy_interrupt_is_valid(phydev)) + break; + err = phy_read_status(phydev); if (err) break; From 4194b4914a824c21bcad9305d6180825f4dc1f5e Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Tue, 17 Nov 2015 10:38:36 +0100 Subject: [PATCH 226/335] packet: Don't check frames_per_block against negative values rb->frames_per_block is an unsigned int, thus can never be negative. Also fix spacing in the calculation of frames_per_block. Signed-off-by: Tobias Klauser Signed-off-by: David S. Miller --- net/packet/af_packet.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 242bce1cf0f3..533981d49290 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -4121,8 +4121,8 @@ static int packet_set_ring(struct sock *sk, union tpacket_req_u *req_u, if (unlikely(req->tp_frame_size & (TPACKET_ALIGNMENT - 1))) goto out; - rb->frames_per_block = req->tp_block_size/req->tp_frame_size; - if (unlikely(rb->frames_per_block <= 0)) + rb->frames_per_block = req->tp_block_size / req->tp_frame_size; + if (unlikely(rb->frames_per_block == 0)) goto out; if (unlikely((rb->frames_per_block * req->tp_block_nr) != req->tp_frame_nr)) From 90836b67e2171f48be696bdd3c4006d1e6f031fc Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Tue, 17 Nov 2015 10:40:21 +0100 Subject: [PATCH 227/335] packet: Use PAGE_ALIGNED macro Use PAGE_ALIGNED(...) instead of open-coding it. Signed-off-by: Tobias Klauser Signed-off-by: David S. Miller --- net/packet/af_packet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 533981d49290..1cf928fb573e 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -4109,7 +4109,7 @@ static int packet_set_ring(struct sock *sk, union tpacket_req_u *req_u, err = -EINVAL; if (unlikely((int)req->tp_block_size <= 0)) goto out; - if (unlikely(req->tp_block_size & (PAGE_SIZE - 1))) + if (unlikely(!PAGE_ALIGNED(req->tp_block_size))) goto out; if (po->tp_version >= TPACKET_V3 && (int)(req->tp_block_size - From 19125c1a4fd873421698876f34a06dbe18d25f17 Mon Sep 17 00:00:00 2001 From: Martin Zhang Date: Tue, 17 Nov 2015 20:49:30 +0800 Subject: [PATCH 228/335] net: use skb_clone to avoid alloc_pages failure. 1. new skb only need dst and ip address(v4 or v6). 2. skb_copy may need high order pages, which is very rare on long running server. Signed-off-by: Junwei Zhang Signed-off-by: Martin Zhang Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/neighbour.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 1aa8437ed6c4..e6af42da28d9 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -857,7 +857,7 @@ static void neigh_probe(struct neighbour *neigh) struct sk_buff *skb = skb_peek_tail(&neigh->arp_queue); /* keep skb alive even if arp_queue overflows */ if (skb) - skb = skb_copy(skb, GFP_ATOMIC); + skb = skb_clone(skb, GFP_ATOMIC); write_unlock(&neigh->lock); neigh->ops->solicit(neigh, skb); atomic_inc(&neigh->probes); From b22b941b2c253a20e1d000c671594c4f3f0a3858 Mon Sep 17 00:00:00 2001 From: Hannes Frederic Sowa Date: Tue, 17 Nov 2015 14:16:52 +0100 Subject: [PATCH 229/335] rtnetlink: fix frame size warning in rtnl_fill_ifinfo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the following warning: CC net/core/rtnetlink.o net/core/rtnetlink.c: In function ‘rtnl_fill_ifinfo’: net/core/rtnetlink.c:1308:1: warning: the frame size of 2864 bytes is larger than 2048 bytes [-Wframe-larger-than=] } ^ by splitting up the huge rtnl_fill_ifinfo into some smaller ones, so we don't have the huge frame allocations at the same time. Cc: Eric Dumazet Signed-off-by: Hannes Frederic Sowa Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/rtnetlink.c | 274 ++++++++++++++++++++++++------------------- 1 file changed, 152 insertions(+), 122 deletions(-) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 504bd17b7456..34ba7a08876d 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -1045,15 +1045,156 @@ static int rtnl_phys_switch_id_fill(struct sk_buff *skb, struct net_device *dev) return 0; } +static noinline_for_stack int rtnl_fill_stats(struct sk_buff *skb, + struct net_device *dev) +{ + const struct rtnl_link_stats64 *stats; + struct rtnl_link_stats64 temp; + struct nlattr *attr; + + stats = dev_get_stats(dev, &temp); + + attr = nla_reserve(skb, IFLA_STATS, + sizeof(struct rtnl_link_stats)); + if (!attr) + return -EMSGSIZE; + + copy_rtnl_link_stats(nla_data(attr), stats); + + attr = nla_reserve(skb, IFLA_STATS64, + sizeof(struct rtnl_link_stats64)); + if (!attr) + return -EMSGSIZE; + + copy_rtnl_link_stats64(nla_data(attr), stats); + + return 0; +} + +static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb, + struct net_device *dev, + int vfs_num, + struct nlattr *vfinfo) +{ + struct ifla_vf_rss_query_en vf_rss_query_en; + struct ifla_vf_link_state vf_linkstate; + struct ifla_vf_spoofchk vf_spoofchk; + struct ifla_vf_tx_rate vf_tx_rate; + struct ifla_vf_stats vf_stats; + struct ifla_vf_trust vf_trust; + struct ifla_vf_vlan vf_vlan; + struct ifla_vf_rate vf_rate; + struct nlattr *vf, *vfstats; + struct ifla_vf_mac vf_mac; + struct ifla_vf_info ivi; + + /* Not all SR-IOV capable drivers support the + * spoofcheck and "RSS query enable" query. Preset to + * -1 so the user space tool can detect that the driver + * didn't report anything. + */ + ivi.spoofchk = -1; + ivi.rss_query_en = -1; + ivi.trusted = -1; + memset(ivi.mac, 0, sizeof(ivi.mac)); + /* The default value for VF link state is "auto" + * IFLA_VF_LINK_STATE_AUTO which equals zero + */ + ivi.linkstate = 0; + if (dev->netdev_ops->ndo_get_vf_config(dev, vfs_num, &ivi)) + return 0; + + vf_mac.vf = + vf_vlan.vf = + vf_rate.vf = + vf_tx_rate.vf = + vf_spoofchk.vf = + vf_linkstate.vf = + vf_rss_query_en.vf = + vf_trust.vf = ivi.vf; + + memcpy(vf_mac.mac, ivi.mac, sizeof(ivi.mac)); + vf_vlan.vlan = ivi.vlan; + vf_vlan.qos = ivi.qos; + vf_tx_rate.rate = ivi.max_tx_rate; + vf_rate.min_tx_rate = ivi.min_tx_rate; + vf_rate.max_tx_rate = ivi.max_tx_rate; + vf_spoofchk.setting = ivi.spoofchk; + vf_linkstate.link_state = ivi.linkstate; + vf_rss_query_en.setting = ivi.rss_query_en; + vf_trust.setting = ivi.trusted; + vf = nla_nest_start(skb, IFLA_VF_INFO); + if (!vf) { + nla_nest_cancel(skb, vfinfo); + return -EMSGSIZE; + } + if (nla_put(skb, IFLA_VF_MAC, sizeof(vf_mac), &vf_mac) || + nla_put(skb, IFLA_VF_VLAN, sizeof(vf_vlan), &vf_vlan) || + nla_put(skb, IFLA_VF_RATE, sizeof(vf_rate), + &vf_rate) || + nla_put(skb, IFLA_VF_TX_RATE, sizeof(vf_tx_rate), + &vf_tx_rate) || + nla_put(skb, IFLA_VF_SPOOFCHK, sizeof(vf_spoofchk), + &vf_spoofchk) || + nla_put(skb, IFLA_VF_LINK_STATE, sizeof(vf_linkstate), + &vf_linkstate) || + nla_put(skb, IFLA_VF_RSS_QUERY_EN, + sizeof(vf_rss_query_en), + &vf_rss_query_en) || + nla_put(skb, IFLA_VF_TRUST, + sizeof(vf_trust), &vf_trust)) + return -EMSGSIZE; + memset(&vf_stats, 0, sizeof(vf_stats)); + if (dev->netdev_ops->ndo_get_vf_stats) + dev->netdev_ops->ndo_get_vf_stats(dev, vfs_num, + &vf_stats); + vfstats = nla_nest_start(skb, IFLA_VF_STATS); + if (!vfstats) { + nla_nest_cancel(skb, vf); + nla_nest_cancel(skb, vfinfo); + return -EMSGSIZE; + } + if (nla_put_u64(skb, IFLA_VF_STATS_RX_PACKETS, + vf_stats.rx_packets) || + nla_put_u64(skb, IFLA_VF_STATS_TX_PACKETS, + vf_stats.tx_packets) || + nla_put_u64(skb, IFLA_VF_STATS_RX_BYTES, + vf_stats.rx_bytes) || + nla_put_u64(skb, IFLA_VF_STATS_TX_BYTES, + vf_stats.tx_bytes) || + nla_put_u64(skb, IFLA_VF_STATS_BROADCAST, + vf_stats.broadcast) || + nla_put_u64(skb, IFLA_VF_STATS_MULTICAST, + vf_stats.multicast)) + return -EMSGSIZE; + nla_nest_end(skb, vfstats); + nla_nest_end(skb, vf); + return 0; +} + +static int rtnl_fill_link_ifmap(struct sk_buff *skb, struct net_device *dev) +{ + struct rtnl_link_ifmap map = { + .mem_start = dev->mem_start, + .mem_end = dev->mem_end, + .base_addr = dev->base_addr, + .irq = dev->irq, + .dma = dev->dma, + .port = dev->if_port, + }; + if (nla_put(skb, IFLA_MAP, sizeof(map), &map)) + return -EMSGSIZE; + + return 0; +} + static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev, int type, u32 pid, u32 seq, u32 change, unsigned int flags, u32 ext_filter_mask) { struct ifinfomsg *ifm; struct nlmsghdr *nlh; - struct rtnl_link_stats64 temp; - const struct rtnl_link_stats64 *stats; - struct nlattr *attr, *af_spec; + struct nlattr *af_spec; struct rtnl_af_ops *af_ops; struct net_device *upper_dev = netdev_master_upper_dev_get(dev); @@ -1096,18 +1237,8 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev, nla_put_u8(skb, IFLA_PROTO_DOWN, dev->proto_down)) goto nla_put_failure; - if (1) { - struct rtnl_link_ifmap map = { - .mem_start = dev->mem_start, - .mem_end = dev->mem_end, - .base_addr = dev->base_addr, - .irq = dev->irq, - .dma = dev->dma, - .port = dev->if_port, - }; - if (nla_put(skb, IFLA_MAP, sizeof(map), &map)) - goto nla_put_failure; - } + if (rtnl_fill_link_ifmap(skb, dev)) + goto nla_put_failure; if (dev->addr_len) { if (nla_put(skb, IFLA_ADDRESS, dev->addr_len, dev->dev_addr) || @@ -1124,128 +1255,27 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev, if (rtnl_phys_switch_id_fill(skb, dev)) goto nla_put_failure; - attr = nla_reserve(skb, IFLA_STATS, - sizeof(struct rtnl_link_stats)); - if (attr == NULL) + if (rtnl_fill_stats(skb, dev)) goto nla_put_failure; - stats = dev_get_stats(dev, &temp); - copy_rtnl_link_stats(nla_data(attr), stats); - - attr = nla_reserve(skb, IFLA_STATS64, - sizeof(struct rtnl_link_stats64)); - if (attr == NULL) - goto nla_put_failure; - copy_rtnl_link_stats64(nla_data(attr), stats); - if (dev->dev.parent && (ext_filter_mask & RTEXT_FILTER_VF) && nla_put_u32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent))) goto nla_put_failure; - if (dev->netdev_ops->ndo_get_vf_config && dev->dev.parent - && (ext_filter_mask & RTEXT_FILTER_VF)) { + if (dev->netdev_ops->ndo_get_vf_config && dev->dev.parent && + ext_filter_mask & RTEXT_FILTER_VF) { int i; - - struct nlattr *vfinfo, *vf, *vfstats; + struct nlattr *vfinfo; int num_vfs = dev_num_vf(dev->dev.parent); vfinfo = nla_nest_start(skb, IFLA_VFINFO_LIST); if (!vfinfo) goto nla_put_failure; for (i = 0; i < num_vfs; i++) { - struct ifla_vf_info ivi; - struct ifla_vf_mac vf_mac; - struct ifla_vf_vlan vf_vlan; - struct ifla_vf_rate vf_rate; - struct ifla_vf_tx_rate vf_tx_rate; - struct ifla_vf_spoofchk vf_spoofchk; - struct ifla_vf_link_state vf_linkstate; - struct ifla_vf_rss_query_en vf_rss_query_en; - struct ifla_vf_stats vf_stats; - struct ifla_vf_trust vf_trust; - - /* - * Not all SR-IOV capable drivers support the - * spoofcheck and "RSS query enable" query. Preset to - * -1 so the user space tool can detect that the driver - * didn't report anything. - */ - ivi.spoofchk = -1; - ivi.rss_query_en = -1; - ivi.trusted = -1; - memset(ivi.mac, 0, sizeof(ivi.mac)); - /* The default value for VF link state is "auto" - * IFLA_VF_LINK_STATE_AUTO which equals zero - */ - ivi.linkstate = 0; - if (dev->netdev_ops->ndo_get_vf_config(dev, i, &ivi)) - break; - vf_mac.vf = - vf_vlan.vf = - vf_rate.vf = - vf_tx_rate.vf = - vf_spoofchk.vf = - vf_linkstate.vf = - vf_rss_query_en.vf = - vf_trust.vf = ivi.vf; - - memcpy(vf_mac.mac, ivi.mac, sizeof(ivi.mac)); - vf_vlan.vlan = ivi.vlan; - vf_vlan.qos = ivi.qos; - vf_tx_rate.rate = ivi.max_tx_rate; - vf_rate.min_tx_rate = ivi.min_tx_rate; - vf_rate.max_tx_rate = ivi.max_tx_rate; - vf_spoofchk.setting = ivi.spoofchk; - vf_linkstate.link_state = ivi.linkstate; - vf_rss_query_en.setting = ivi.rss_query_en; - vf_trust.setting = ivi.trusted; - vf = nla_nest_start(skb, IFLA_VF_INFO); - if (!vf) { - nla_nest_cancel(skb, vfinfo); + if (rtnl_fill_vfinfo(skb, dev, i, vfinfo)) goto nla_put_failure; - } - if (nla_put(skb, IFLA_VF_MAC, sizeof(vf_mac), &vf_mac) || - nla_put(skb, IFLA_VF_VLAN, sizeof(vf_vlan), &vf_vlan) || - nla_put(skb, IFLA_VF_RATE, sizeof(vf_rate), - &vf_rate) || - nla_put(skb, IFLA_VF_TX_RATE, sizeof(vf_tx_rate), - &vf_tx_rate) || - nla_put(skb, IFLA_VF_SPOOFCHK, sizeof(vf_spoofchk), - &vf_spoofchk) || - nla_put(skb, IFLA_VF_LINK_STATE, sizeof(vf_linkstate), - &vf_linkstate) || - nla_put(skb, IFLA_VF_RSS_QUERY_EN, - sizeof(vf_rss_query_en), - &vf_rss_query_en) || - nla_put(skb, IFLA_VF_TRUST, - sizeof(vf_trust), &vf_trust)) - goto nla_put_failure; - memset(&vf_stats, 0, sizeof(vf_stats)); - if (dev->netdev_ops->ndo_get_vf_stats) - dev->netdev_ops->ndo_get_vf_stats(dev, i, - &vf_stats); - vfstats = nla_nest_start(skb, IFLA_VF_STATS); - if (!vfstats) { - nla_nest_cancel(skb, vf); - nla_nest_cancel(skb, vfinfo); - goto nla_put_failure; - } - if (nla_put_u64(skb, IFLA_VF_STATS_RX_PACKETS, - vf_stats.rx_packets) || - nla_put_u64(skb, IFLA_VF_STATS_TX_PACKETS, - vf_stats.tx_packets) || - nla_put_u64(skb, IFLA_VF_STATS_RX_BYTES, - vf_stats.rx_bytes) || - nla_put_u64(skb, IFLA_VF_STATS_TX_BYTES, - vf_stats.tx_bytes) || - nla_put_u64(skb, IFLA_VF_STATS_BROADCAST, - vf_stats.broadcast) || - nla_put_u64(skb, IFLA_VF_STATS_MULTICAST, - vf_stats.multicast)) - goto nla_put_failure; - nla_nest_end(skb, vfstats); - nla_nest_end(skb, vf); } + nla_nest_end(skb, vfinfo); } From a3a116e04cc6a94d595ead4e956ab1bc1d2f4746 Mon Sep 17 00:00:00 2001 From: Hannes Frederic Sowa Date: Tue, 17 Nov 2015 15:10:59 +0100 Subject: [PATCH 230/335] af_unix: take receive queue lock while appending new skb While possibly in future we don't necessarily need to use sk_buff_head.lock this is a rather larger change, as it affects the af_unix fd garbage collector, diag and socket cleanups. This is too much for a stable patch. For the time being grab sk_buff_head.lock without disabling bh and irqs, so don't use locked skb_queue_tail. Fixes: 869e7c62486e ("net: af_unix: implement stream sendpage support") Cc: Eric Dumazet Signed-off-by: Hannes Frederic Sowa Reported-by: Eric Dumazet Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/unix/af_unix.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index a8352db5c5b5..955ec152cb71 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1813,8 +1813,11 @@ alloc_skb: skb->truesize += size; atomic_add(size, &sk->sk_wmem_alloc); - if (newskb) + if (newskb) { + spin_lock(&other->sk_receive_queue.lock); __skb_queue_tail(&other->sk_receive_queue, newskb); + spin_unlock(&other->sk_receive_queue.lock); + } unix_state_unlock(other); mutex_unlock(&unix_sk(other)->readlock); From 17b85d29e82cc3c874a497a8bc5764d6a2b043e2 Mon Sep 17 00:00:00 2001 From: Nikolay Aleksandrov Date: Tue, 17 Nov 2015 15:49:06 +0100 Subject: [PATCH 231/335] net/core: revert "net: fix __netdev_update_features return.." and add comment MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit 00ee59271777 ("net: fix __netdev_update_features return on ndo_set_features failure") and adds a comment explaining why it's okay to return a value other than 0 upon error. Some drivers might actually change flags and return an error so it's better to fire a spurious notification rather than miss these. CC: Michał Mirosław Signed-off-by: Nikolay Aleksandrov Signed-off-by: David S. Miller --- net/core/dev.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index 5dbc86ea6b58..ae00b894e675 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6436,7 +6436,10 @@ int __netdev_update_features(struct net_device *dev) netdev_err(dev, "set_features() failed (%d); wanted %pNF, left %pNF\n", err, &features, &dev->features); - return 0; + /* return non-0 since some features might have changed and + * it's better to fire a spurious notification than miss it + */ + return -1; } sync_lower: From e7523a497d48a9921983a80670f7a02dc4639d41 Mon Sep 17 00:00:00 2001 From: Or Gerlitz Date: Tue, 17 Nov 2015 18:25:07 +0200 Subject: [PATCH 232/335] MAINTAINERS: Update Mellanox's Eth NIC driver entries Eugenia (Jenny) Emantayev is replacing Amir Vadai as the mlx4 Ethernet driver maintainer. Saeed Mahameed is assigned to maintain mlx5 Eth functionality. Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- MAINTAINERS | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 45320675a460..902cdab54de0 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -6914,13 +6914,21 @@ F: drivers/scsi/megaraid.* F: drivers/scsi/megaraid/ MELLANOX ETHERNET DRIVER (mlx4_en) -M: Amir Vadai +M: Eugenia Emantayev L: netdev@vger.kernel.org S: Supported W: http://www.mellanox.com Q: http://patchwork.ozlabs.org/project/netdev/list/ F: drivers/net/ethernet/mellanox/mlx4/en_* +MELLANOX ETHERNET DRIVER (mlx5e) +M: Saeed Mahameed +L: netdev@vger.kernel.org +S: Supported +W: http://www.mellanox.com +Q: http://patchwork.ozlabs.org/project/netdev/list/ +F: drivers/net/ethernet/mellanox/mlx5/core/en_* + MELLANOX ETHERNET SWITCH DRIVERS M: Jiri Pirko M: Ido Schimmel From 91276ae2ea7d588cb70ed54fb0b00b732aa7ea04 Mon Sep 17 00:00:00 2001 From: kbuild test robot Date: Thu, 22 Oct 2015 11:12:26 +0800 Subject: [PATCH 233/335] drm/vc4: vc4_plane_duplicate_state() can be static Signed-off-by: Fengguang Wu Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_plane.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c index cdd8b10c0147..2db509210d62 100644 --- a/drivers/gpu/drm/vc4/vc4_plane.c +++ b/drivers/gpu/drm/vc4/vc4_plane.c @@ -70,7 +70,7 @@ static bool plane_enabled(struct drm_plane_state *state) return state->fb && state->crtc; } -struct drm_plane_state *vc4_plane_duplicate_state(struct drm_plane *plane) +static struct drm_plane_state *vc4_plane_duplicate_state(struct drm_plane *plane) { struct vc4_plane_state *vc4_state; @@ -97,8 +97,8 @@ struct drm_plane_state *vc4_plane_duplicate_state(struct drm_plane *plane) return &vc4_state->base; } -void vc4_plane_destroy_state(struct drm_plane *plane, - struct drm_plane_state *state) +static void vc4_plane_destroy_state(struct drm_plane *plane, + struct drm_plane_state *state) { struct vc4_plane_state *vc4_state = to_vc4_plane_state(state); @@ -108,7 +108,7 @@ void vc4_plane_destroy_state(struct drm_plane *plane, } /* Called during init to allocate the plane's atomic state. */ -void vc4_plane_reset(struct drm_plane *plane) +static void vc4_plane_reset(struct drm_plane *plane) { struct vc4_plane_state *vc4_state; From e2ab3218f46bce3380189ba4ac2a8cbdb2ae6272 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 23 Oct 2015 07:39:45 +0200 Subject: [PATCH 234/335] drm/vc4: fix platform_no_drv_owner.cocci warnings drivers/gpu/drm/vc4/vc4_drv.c:248:3-8: No need to set .owner here. The core will do it. Remove .owner field if calls are used which set it automatically Generated by: scripts/coccinelle/api/platform_no_drv_owner.cocci Signed-off-by: Fengguang Wu Signed-off-by: Julia Lawall Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_drv.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c index 6e730605edcc..d5db9e0f3b73 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.c +++ b/drivers/gpu/drm/vc4/vc4_drv.c @@ -259,7 +259,6 @@ static struct platform_driver vc4_platform_driver = { .remove = vc4_platform_drm_remove, .driver = { .name = "vc4-drm", - .owner = THIS_MODULE, .of_match_table = vc4_of_match, }, }; From 2fa8e90433cdc2d9f09883f45fd2942dc042d6a6 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 23 Oct 2015 07:38:00 +0200 Subject: [PATCH 235/335] drm/vc4: fix itnull.cocci warnings Connector cannot be null because it is a list entry, ie accessed at an offset from the positions of the list structure pointers themselves. Generated by: scripts/coccinelle/iterators/itnull.cocci Signed-off-by: Fengguang Wu Signed-off-by: Julia Lawall Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_crtc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index 7a9f4768591e..f794c5b6996b 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -168,7 +168,7 @@ static int vc4_get_clock_select(struct drm_crtc *crtc) struct drm_connector *connector; drm_for_each_connector(connector, crtc->dev) { - if (connector && connector->state->crtc == crtc) { + if (connector->state->crtc == crtc) { struct drm_encoder *encoder = connector->encoder; struct vc4_encoder *vc4_encoder = to_vc4_encoder(encoder); From 79513237da7ecc0981e7d962a784625b32cabfdc Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 4 Nov 2015 16:21:40 +0300 Subject: [PATCH 236/335] drm/vc4: checking for NULL instead of IS_ERR vc4_plane_init() returns an ERR_PTR on error, it doesn't return NULL. This was obviously intended because the next lines call PTR_ERR(primary_plane) already. Fixes: c8b75bca92cb ('Eric Anholt ') Signed-off-by: Dan Carpenter Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_crtc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index f794c5b6996b..f77fa9ebab29 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -591,14 +591,14 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data) * that will take too much. */ primary_plane = vc4_plane_init(drm, DRM_PLANE_TYPE_PRIMARY); - if (!primary_plane) { + if (IS_ERR(primary_plane)) { dev_err(dev, "failed to construct primary plane\n"); ret = PTR_ERR(primary_plane); goto err; } cursor_plane = vc4_plane_init(drm, DRM_PLANE_TYPE_CURSOR); - if (!cursor_plane) { + if (IS_ERR(cursor_plane)) { dev_err(dev, "failed to construct cursor plane\n"); ret = PTR_ERR(cursor_plane); goto err_primary; From 36f4f69a092d766581f2442ddffea051e1ef464e Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Fri, 23 Oct 2015 10:24:11 +0100 Subject: [PATCH 237/335] drm/vc4: Fix some failure to track __iomem decorations on pointers. Caught by the kbuild test robot. v2: Fix the +i I dropped in the first version. Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_crtc.c | 3 ++- drivers/gpu/drm/vc4/vc4_hvs.c | 8 ++++---- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index f77fa9ebab29..265064c62d49 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -401,7 +401,8 @@ static void vc4_crtc_atomic_flush(struct drm_crtc *crtc, dlist_next++; HVS_WRITE(SCALER_DISPLISTX(vc4_crtc->channel), - (u32 *)vc4_crtc->dlist - (u32 *)vc4->hvs->dlist); + (u32 __iomem *)vc4_crtc->dlist - + (u32 __iomem *)vc4->hvs->dlist); /* Make the next display list start after ours. */ vc4_crtc->dlist_size -= (dlist_next - vc4_crtc->dlist); diff --git a/drivers/gpu/drm/vc4/vc4_hvs.c b/drivers/gpu/drm/vc4/vc4_hvs.c index ab1673f672a4..8098c5b21ba4 100644 --- a/drivers/gpu/drm/vc4/vc4_hvs.c +++ b/drivers/gpu/drm/vc4/vc4_hvs.c @@ -75,10 +75,10 @@ void vc4_hvs_dump_state(struct drm_device *dev) for (i = 0; i < 64; i += 4) { DRM_INFO("0x%08x (%s): 0x%08x 0x%08x 0x%08x 0x%08x\n", i * 4, i < HVS_BOOTLOADER_DLIST_END ? "B" : "D", - ((uint32_t *)vc4->hvs->dlist)[i + 0], - ((uint32_t *)vc4->hvs->dlist)[i + 1], - ((uint32_t *)vc4->hvs->dlist)[i + 2], - ((uint32_t *)vc4->hvs->dlist)[i + 3]); + readl((u32 __iomem *)vc4->hvs->dlist + i + 0), + readl((u32 __iomem *)vc4->hvs->dlist + i + 1), + readl((u32 __iomem *)vc4->hvs->dlist + i + 2), + readl((u32 __iomem *)vc4->hvs->dlist + i + 3)); } } From bf893acc1675516fabf238a5ee3b52859c5687dc Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Fri, 23 Oct 2015 10:36:27 +0100 Subject: [PATCH 238/335] drm/vc4: Make sure that planes aren't scaled. We would scan out the memory around them if an upscale was attempted, and would just scan out incorrectly for downscaling. Signed-off-by: Eric Anholt --- drivers/gpu/drm/vc4/vc4_plane.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c index 2db509210d62..887f3caad0be 100644 --- a/drivers/gpu/drm/vc4/vc4_plane.c +++ b/drivers/gpu/drm/vc4/vc4_plane.c @@ -157,6 +157,16 @@ static int vc4_plane_mode_set(struct drm_plane *plane, int crtc_w = state->crtc_w; int crtc_h = state->crtc_h; + if (state->crtc_w << 16 != state->src_w || + state->crtc_h << 16 != state->src_h) { + /* We don't support scaling yet, which involves + * allocating the LBM memory for scaling temporary + * storage, and putting filter kernels in the HVS + * context. + */ + return -EINVAL; + } + if (crtc_x < 0) { offset += drm_format_plane_cpp(fb->pixel_format, 0) * -crtc_x; crtc_w += crtc_x; From 1ef0623371e0a39a476fb05e575089cf48178f5c Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 17 Nov 2015 13:29:49 -0200 Subject: [PATCH 239/335] kernel-doc: Make it compatible with Perl versions below 5.12 again Changeset 4d73270192ec('scripts/kernel-doc: Replacing highlights hash by an array') broke compatibility of the kernel-doc script with older versions of perl by using "keys ARRAY" syntax with is available only on Perl 5.12 or newer, according with: http://perldoc.perl.org/functions/keys.html Restore backward compatibility by replacing "foreach my $k (keys ARRAY)" by a C-like variant: "for (my $k = 0; $k < !ARRAY; $k++)" Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Jonathan Corbet --- scripts/kernel-doc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/kernel-doc b/scripts/kernel-doc index 125b906cd1d4..638a38e1b419 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -2711,7 +2711,7 @@ $kernelversion = get_kernel_version(); # generate a sequence of code that will splice in highlighting information # using the s// operator. -foreach my $k (keys @highlights) { +for (my $k = 0; $k < @highlights; $k++) { my $pattern = $highlights[$k][0]; my $result = $highlights[$k][1]; # print STDERR "scanning pattern:$pattern, highlight:($result)\n"; From ae3e57ae26cdcc85728bb566f999bcb9a7cc6954 Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Wed, 16 Sep 2015 09:40:51 +0800 Subject: [PATCH 240/335] usb: chipidea: imx: refine clock operations to adapt for all platforms Some i.mx platforms need three clocks to let controller work, but others only need one, refine clock operation to adapt for all platforms, it fixes a regression found at i.mx27. Signed-off-by: Peter Chen Tested-by: Fabio Estevam Cc: #v4.1+ --- drivers/usb/chipidea/ci_hdrc_imx.c | 133 ++++++++++++++++++++++++----- 1 file changed, 114 insertions(+), 19 deletions(-) diff --git a/drivers/usb/chipidea/ci_hdrc_imx.c b/drivers/usb/chipidea/ci_hdrc_imx.c index 6ccbf60cdd5c..7c9b63064ea0 100644 --- a/drivers/usb/chipidea/ci_hdrc_imx.c +++ b/drivers/usb/chipidea/ci_hdrc_imx.c @@ -84,6 +84,12 @@ struct ci_hdrc_imx_data { struct imx_usbmisc_data *usbmisc_data; bool supports_runtime_pm; bool in_lpm; + /* SoC before i.mx6 (except imx23/imx28) needs three clks */ + bool need_three_clks; + struct clk *clk_ipg; + struct clk *clk_ahb; + struct clk *clk_per; + /* --------------------------------- */ }; /* Common functions shared by usbmisc drivers */ @@ -135,6 +141,102 @@ static struct imx_usbmisc_data *usbmisc_get_init_data(struct device *dev) } /* End of common functions shared by usbmisc drivers*/ +static int imx_get_clks(struct device *dev) +{ + struct ci_hdrc_imx_data *data = dev_get_drvdata(dev); + int ret = 0; + + data->clk_ipg = devm_clk_get(dev, "ipg"); + if (IS_ERR(data->clk_ipg)) { + /* If the platform only needs one clocks */ + data->clk = devm_clk_get(dev, NULL); + if (IS_ERR(data->clk)) { + ret = PTR_ERR(data->clk); + dev_err(dev, + "Failed to get clks, err=%ld,%ld\n", + PTR_ERR(data->clk), PTR_ERR(data->clk_ipg)); + return ret; + } + return ret; + } + + data->clk_ahb = devm_clk_get(dev, "ahb"); + if (IS_ERR(data->clk_ahb)) { + ret = PTR_ERR(data->clk_ahb); + dev_err(dev, + "Failed to get ahb clock, err=%d\n", ret); + return ret; + } + + data->clk_per = devm_clk_get(dev, "per"); + if (IS_ERR(data->clk_per)) { + ret = PTR_ERR(data->clk_per); + dev_err(dev, + "Failed to get per clock, err=%d\n", ret); + return ret; + } + + data->need_three_clks = true; + return ret; +} + +static int imx_prepare_enable_clks(struct device *dev) +{ + struct ci_hdrc_imx_data *data = dev_get_drvdata(dev); + int ret = 0; + + if (data->need_three_clks) { + ret = clk_prepare_enable(data->clk_ipg); + if (ret) { + dev_err(dev, + "Failed to prepare/enable ipg clk, err=%d\n", + ret); + return ret; + } + + ret = clk_prepare_enable(data->clk_ahb); + if (ret) { + dev_err(dev, + "Failed to prepare/enable ahb clk, err=%d\n", + ret); + clk_disable_unprepare(data->clk_ipg); + return ret; + } + + ret = clk_prepare_enable(data->clk_per); + if (ret) { + dev_err(dev, + "Failed to prepare/enable per clk, err=%d\n", + ret); + clk_disable_unprepare(data->clk_ahb); + clk_disable_unprepare(data->clk_ipg); + return ret; + } + } else { + ret = clk_prepare_enable(data->clk); + if (ret) { + dev_err(dev, + "Failed to prepare/enable clk, err=%d\n", + ret); + return ret; + } + } + + return ret; +} + +static void imx_disable_unprepare_clks(struct device *dev) +{ + struct ci_hdrc_imx_data *data = dev_get_drvdata(dev); + + if (data->need_three_clks) { + clk_disable_unprepare(data->clk_per); + clk_disable_unprepare(data->clk_ahb); + clk_disable_unprepare(data->clk_ipg); + } else { + clk_disable_unprepare(data->clk); + } +} static int ci_hdrc_imx_probe(struct platform_device *pdev) { @@ -153,23 +255,18 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) if (!data) return -ENOMEM; + platform_set_drvdata(pdev, data); data->usbmisc_data = usbmisc_get_init_data(&pdev->dev); if (IS_ERR(data->usbmisc_data)) return PTR_ERR(data->usbmisc_data); - data->clk = devm_clk_get(&pdev->dev, NULL); - if (IS_ERR(data->clk)) { - dev_err(&pdev->dev, - "Failed to get clock, err=%ld\n", PTR_ERR(data->clk)); - return PTR_ERR(data->clk); - } - - ret = clk_prepare_enable(data->clk); - if (ret) { - dev_err(&pdev->dev, - "Failed to prepare or enable clock, err=%d\n", ret); + ret = imx_get_clks(&pdev->dev); + if (ret) + return ret; + + ret = imx_prepare_enable_clks(&pdev->dev); + if (ret) return ret; - } data->phy = devm_usb_get_phy_by_phandle(&pdev->dev, "fsl,usbphy", 0); if (IS_ERR(data->phy)) { @@ -212,8 +309,6 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) goto disable_device; } - platform_set_drvdata(pdev, data); - if (data->supports_runtime_pm) { pm_runtime_set_active(&pdev->dev); pm_runtime_enable(&pdev->dev); @@ -226,7 +321,7 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) disable_device: ci_hdrc_remove_device(data->ci_pdev); err_clk: - clk_disable_unprepare(data->clk); + imx_disable_unprepare_clks(&pdev->dev); return ret; } @@ -240,7 +335,7 @@ static int ci_hdrc_imx_remove(struct platform_device *pdev) pm_runtime_put_noidle(&pdev->dev); } ci_hdrc_remove_device(data->ci_pdev); - clk_disable_unprepare(data->clk); + imx_disable_unprepare_clks(&pdev->dev); return 0; } @@ -252,7 +347,7 @@ static int imx_controller_suspend(struct device *dev) dev_dbg(dev, "at %s\n", __func__); - clk_disable_unprepare(data->clk); + imx_disable_unprepare_clks(dev); data->in_lpm = true; return 0; @@ -270,7 +365,7 @@ static int imx_controller_resume(struct device *dev) return 0; } - ret = clk_prepare_enable(data->clk); + ret = imx_prepare_enable_clks(dev); if (ret) return ret; @@ -285,7 +380,7 @@ static int imx_controller_resume(struct device *dev) return 0; clk_disable: - clk_disable_unprepare(data->clk); + imx_disable_unprepare_clks(dev); return ret; } From facf47ee6b4d07d43c3bfd6f0762f1b28f64703a Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Wed, 16 Sep 2015 09:35:06 +0800 Subject: [PATCH 241/335] ARM: dts: imx27.dtsi: change the clock information for usb For imx27, it needs three clocks to let the controller work, the old code is wrong, and usbmisc has not included clock handling code any more. Without this patch, it will cause below data abort when accessing usbmisc registers. usbcore: registered new interface driver usb-storage Unhandled fault: external abort on non-linefetch (0x008) at 0xf4424600 pgd = c0004000 [f4424600] *pgd=10000452(bad) Internal error: : 8 [#1] PREEMPT ARM Modules linked in: CPU: 0 PID: 1 Comm: swapper Not tainted 4.1.0-next-20150701-dirty #3089 Hardware name: Freescale i.MX27 (Device Tree Support) task: c7832b60 ti: c783e000 task.ti: c783e000 PC is at usbmisc_imx27_init+0x4c/0xbc LR is at usbmisc_imx27_init+0x40/0xbc pc : [] lr : [] psr: 60000093 sp : c783fe08 ip : 00000000 fp : 00000000 r10: c0576434 r9 : 0000009c r8 : c7a773a0 r7 : 01000000 r6 : 60000013 r5 : c7a776f0 r4 : c7a773f0 r3 : f4424600 r2 : 00000000 r1 : 00000001 r0 : 00000001 Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment kernel Control: 0005317f Table: a0004000 DAC: 00000017 Process swapper (pid: 1, stack limit = 0xc783e190) Stack: (0xc783fe08 to 0xc7840000) Signed-off-by: Peter Chen Reported-by: Fabio Estevam Tested-by: Fabio Estevam Cc: #v4.1+ Acked-by: Shawn Guo --- arch/arm/boot/dts/imx27.dtsi | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/arch/arm/boot/dts/imx27.dtsi b/arch/arm/boot/dts/imx27.dtsi index feb9d34b239c..f818ea483aeb 100644 --- a/arch/arm/boot/dts/imx27.dtsi +++ b/arch/arm/boot/dts/imx27.dtsi @@ -486,7 +486,10 @@ compatible = "fsl,imx27-usb"; reg = <0x10024000 0x200>; interrupts = <56>; - clocks = <&clks IMX27_CLK_USB_IPG_GATE>; + clocks = <&clks IMX27_CLK_USB_IPG_GATE>, + <&clks IMX27_CLK_USB_AHB_GATE>, + <&clks IMX27_CLK_USB_DIV>; + clock-names = "ipg", "ahb", "per"; fsl,usbmisc = <&usbmisc 0>; status = "disabled"; }; @@ -495,7 +498,10 @@ compatible = "fsl,imx27-usb"; reg = <0x10024200 0x200>; interrupts = <54>; - clocks = <&clks IMX27_CLK_USB_IPG_GATE>; + clocks = <&clks IMX27_CLK_USB_IPG_GATE>, + <&clks IMX27_CLK_USB_AHB_GATE>, + <&clks IMX27_CLK_USB_DIV>; + clock-names = "ipg", "ahb", "per"; fsl,usbmisc = <&usbmisc 1>; dr_mode = "host"; status = "disabled"; @@ -505,7 +511,10 @@ compatible = "fsl,imx27-usb"; reg = <0x10024400 0x200>; interrupts = <55>; - clocks = <&clks IMX27_CLK_USB_IPG_GATE>; + clocks = <&clks IMX27_CLK_USB_IPG_GATE>, + <&clks IMX27_CLK_USB_AHB_GATE>, + <&clks IMX27_CLK_USB_DIV>; + clock-names = "ipg", "ahb", "per"; fsl,usbmisc = <&usbmisc 2>; dr_mode = "host"; status = "disabled"; @@ -515,7 +524,6 @@ #index-cells = <1>; compatible = "fsl,imx27-usbmisc"; reg = <0x10024600 0x200>; - clocks = <&clks IMX27_CLK_USB_AHB_GATE>; }; sahara2: sahara@10025000 { From 251b3c8b57481bcecd3f753108e36e7389ce12ac Mon Sep 17 00:00:00 2001 From: Li Jun Date: Tue, 13 Oct 2015 18:23:31 +0800 Subject: [PATCH 242/335] usb: chipidea: debug: disable usb irq while role switch Since the ci->role will be set after the host role start is complete, there will be nobody cared irq during start host if usb irq enabled. This error can be reproduced on i.mx6 sololite EVK board by: 1. disable otg id irq(IDIE) and disable all real otg properties of usbotg1 in dts. 2. boot up the board with ID cable and usb device connected. 3. echo gadget > /sys/kernel/debug/ci_hdrc.0/role 4. echo host > /sys/kernel/debug/ci_hdrc.0/role 5. irq 212: nobody cared. Cc: # v3.10+ Signed-off-by: Li Jun Signed-off-by: Peter Chen --- drivers/usb/chipidea/debug.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/usb/chipidea/debug.c b/drivers/usb/chipidea/debug.c index 080b7be3daf0..58c8485a0715 100644 --- a/drivers/usb/chipidea/debug.c +++ b/drivers/usb/chipidea/debug.c @@ -322,8 +322,10 @@ static ssize_t ci_role_write(struct file *file, const char __user *ubuf, return -EINVAL; pm_runtime_get_sync(ci->dev); + disable_irq(ci->irq); ci_role_stop(ci); ret = ci_role_start(ci, role); + enable_irq(ci->irq); pm_runtime_put_sync(ci->dev); return ret ? ret : count; From 85da852df66e5e0d3aba761b0fece7c958ff0685 Mon Sep 17 00:00:00 2001 From: Li Jun Date: Fri, 12 Dec 2014 09:11:42 +0800 Subject: [PATCH 243/335] usb: chipidea: otg: gadget module load and unload support This patch is to support load and unload gadget driver in full OTG mode. Signed-off-by: Li Jun Signed-off-by: Peter Chen Tested-by: Jiada Wang Cc: #v4.0+ --- drivers/usb/chipidea/udc.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/drivers/usb/chipidea/udc.c b/drivers/usb/chipidea/udc.c index 8223fe73ea85..391a1225b0ba 100644 --- a/drivers/usb/chipidea/udc.c +++ b/drivers/usb/chipidea/udc.c @@ -1751,6 +1751,22 @@ static int ci_udc_start(struct usb_gadget *gadget, return retval; } +static void ci_udc_stop_for_otg_fsm(struct ci_hdrc *ci) +{ + if (!ci_otg_is_fsm_mode(ci)) + return; + + mutex_lock(&ci->fsm.lock); + if (ci->fsm.otg->state == OTG_STATE_A_PERIPHERAL) { + ci->fsm.a_bidl_adis_tmout = 1; + ci_hdrc_otg_fsm_start(ci); + } else if (ci->fsm.otg->state == OTG_STATE_B_PERIPHERAL) { + ci->fsm.protocol = PROTO_UNDEF; + ci->fsm.otg->state = OTG_STATE_UNDEFINED; + } + mutex_unlock(&ci->fsm.lock); +} + /** * ci_udc_stop: unregister a gadget driver */ @@ -1775,6 +1791,7 @@ static int ci_udc_stop(struct usb_gadget *gadget) ci->driver = NULL; spin_unlock_irqrestore(&ci->lock, flags); + ci_udc_stop_for_otg_fsm(ci); return 0; } From 090bc267ea1013bbb33778b343b4acd78b9c5d9b Mon Sep 17 00:00:00 2001 From: LABBE Corentin Date: Thu, 12 Nov 2015 08:43:33 +0100 Subject: [PATCH 244/335] usb: chipidea: usbmisc_imx: fix a possible NULL dereference of_match_device could return NULL, and so cause a NULL pointer dereference later. Renaming tmp_dev to of_id (like all others do) in the process. Reported-by: coverity (CID 1324135) Signed-off-by: LABBE Corentin Signed-off-by: Peter Chen --- drivers/usb/chipidea/usbmisc_imx.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/usb/chipidea/usbmisc_imx.c b/drivers/usb/chipidea/usbmisc_imx.c index fcea4eb36eee..ab8b027e8cc8 100644 --- a/drivers/usb/chipidea/usbmisc_imx.c +++ b/drivers/usb/chipidea/usbmisc_imx.c @@ -500,7 +500,11 @@ static int usbmisc_imx_probe(struct platform_device *pdev) { struct resource *res; struct imx_usbmisc *data; - struct of_device_id *tmp_dev; + const struct of_device_id *of_id; + + of_id = of_match_device(usbmisc_imx_dt_ids, &pdev->dev); + if (!of_id) + return -ENODEV; data = devm_kzalloc(&pdev->dev, sizeof(*data), GFP_KERNEL); if (!data) @@ -513,9 +517,7 @@ static int usbmisc_imx_probe(struct platform_device *pdev) if (IS_ERR(data->base)) return PTR_ERR(data->base); - tmp_dev = (struct of_device_id *) - of_match_device(usbmisc_imx_dt_ids, &pdev->dev); - data->ops = (const struct usbmisc_ops *)tmp_dev->data; + data->ops = (const struct usbmisc_ops *)of_id->data; platform_set_drvdata(pdev, data); return 0; From 6f51bc340d2a1c71a2409f80f3e60fe2c44e35ae Mon Sep 17 00:00:00 2001 From: LABBE Corentin Date: Thu, 12 Nov 2015 08:43:34 +0100 Subject: [PATCH 245/335] usb: chipidea: imx: fix a possible NULL dereference of_match_device could return NULL, and so cause a NULL pointer dereference later. Reported-by: coverity (CID 1324138) Signed-off-by: LABBE Corentin Signed-off-by: Peter Chen --- drivers/usb/chipidea/ci_hdrc_imx.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/usb/chipidea/ci_hdrc_imx.c b/drivers/usb/chipidea/ci_hdrc_imx.c index 7c9b63064ea0..5a048b7b92e8 100644 --- a/drivers/usb/chipidea/ci_hdrc_imx.c +++ b/drivers/usb/chipidea/ci_hdrc_imx.c @@ -247,9 +247,14 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) .flags = CI_HDRC_SET_NON_ZERO_TTHA, }; int ret; - const struct of_device_id *of_id = - of_match_device(ci_hdrc_imx_dt_ids, &pdev->dev); - const struct ci_hdrc_imx_platform_flag *imx_platform_flag = of_id->data; + const struct of_device_id *of_id; + const struct ci_hdrc_imx_platform_flag *imx_platform_flag; + + of_id = of_match_device(ci_hdrc_imx_dt_ids, &pdev->dev); + if (!of_id) + return -ENODEV; + + imx_platform_flag = of_id->data; data = devm_kzalloc(&pdev->dev, sizeof(*data), GFP_KERNEL); if (!data) From 0f94592efd36213c961145fe1ab0c3bc323ec053 Mon Sep 17 00:00:00 2001 From: Mika Kuoppala Date: Tue, 17 Nov 2015 18:14:26 +0200 Subject: [PATCH 246/335] drm/i915: Fix gpu frequency change tracing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With gen < 9 we have had always 50Mhz units as our hw ratio. With gen >= 9 the hw ratio changed to 16.667Mhz (50/3). The result was that our gpu frequency tracing started to output values 3 times larger than expected due to hardcoded scaling value. Fix this by using Use intel_gpu_freq() when generating Mhz value from ratio for 'intel_gpu_freq_change' trace event. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92591 Cc: stable@vger.kernel.org # v4.3+ Reported-by: Eero Tamminen Signed-off-by: Mika Kuoppala Reviewed-by: Ville Syrjälä Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447776866-29384-1-git-send-email-mika.kuoppala@intel.com --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 7b47da805685..071a76b9ac52 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4449,7 +4449,7 @@ static void gen6_set_rps(struct drm_device *dev, u8 val) POSTING_READ(GEN6_RPNSWREQ); dev_priv->rps.cur_freq = val; - trace_intel_gpu_freq_change(val * 50); + trace_intel_gpu_freq_change(intel_gpu_freq(dev_priv, val)); } static void valleyview_set_rps(struct drm_device *dev, u8 val) From 65da0a8e34a857f2ba9ccb91dc8f8f964cf938b7 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Tue, 17 Nov 2015 09:53:31 +0100 Subject: [PATCH 247/335] arm64: use non-global mappings for UEFI runtime regions As pointed out by Russell King in response to the proposed ARM version of this code, the sequence to switch between the UEFI runtime mapping and current's actual userland mapping (and vice versa) is potentially unsafe, since it leaves a time window between the switch to the new page tables and the TLB flush where speculative accesses may hit on stale global TLB entries. So instead, use non-global mappings, and perform the switch via the ordinary ASID-aware context switch routines. Signed-off-by: Ard Biesheuvel Acked-by: Will Deacon Reviewed-by: Mark Rutland Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/mmu_context.h | 2 +- arch/arm64/kernel/efi.c | 14 +++++--------- 2 files changed, 6 insertions(+), 10 deletions(-) diff --git a/arch/arm64/include/asm/mmu_context.h b/arch/arm64/include/asm/mmu_context.h index c0e87898ba96..24165784b803 100644 --- a/arch/arm64/include/asm/mmu_context.h +++ b/arch/arm64/include/asm/mmu_context.h @@ -101,7 +101,7 @@ static inline void cpu_set_default_tcr_t0sz(void) #define destroy_context(mm) do { } while(0) void check_and_switch_context(struct mm_struct *mm, unsigned int cpu); -#define init_new_context(tsk,mm) ({ atomic64_set(&mm->context.id, 0); 0; }) +#define init_new_context(tsk,mm) ({ atomic64_set(&(mm)->context.id, 0); 0; }) /* * This is called when "tsk" is about to enter lazy TLB mode. diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c index de46b50f4cdf..fc5508e0df57 100644 --- a/arch/arm64/kernel/efi.c +++ b/arch/arm64/kernel/efi.c @@ -224,6 +224,8 @@ static bool __init efi_virtmap_init(void) { efi_memory_desc_t *md; + init_new_context(NULL, &efi_mm); + for_each_efi_memory_desc(&memmap, md) { u64 paddr, npages, size; pgprot_t prot; @@ -254,7 +256,8 @@ static bool __init efi_virtmap_init(void) else prot = PAGE_KERNEL; - create_pgd_mapping(&efi_mm, paddr, md->virt_addr, size, prot); + create_pgd_mapping(&efi_mm, paddr, md->virt_addr, size, + __pgprot(pgprot_val(prot) | PTE_NG)); } return true; } @@ -329,14 +332,7 @@ core_initcall(arm64_dmi_init); static void efi_set_pgd(struct mm_struct *mm) { - if (mm == &init_mm) - cpu_set_reserved_ttbr0(); - else - cpu_switch_mm(mm->pgd, mm); - - local_flush_tlb_all(); - if (icache_is_aivivt()) - __local_flush_icache_all(); + switch_mm(NULL, mm, NULL); } void efi_virtmap_load(void) From 08c6781cfaa196d4b257ec043c17e8746d9284e3 Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Mon, 16 Nov 2015 13:12:48 +0100 Subject: [PATCH 248/335] arm64: crypto: reduce priority of core AES cipher The asynchronous, merged implementations of AES in CBC, CTR and XTS modes are preferred when available (i.e., when instantiating ablkciphers explicitly). However, the synchronous core AES cipher combined with the generic CBC mode implementation will produce a 'cbc(aes)' blkcipher that is callable asynchronously as well. To prevent this implementation from being used when the accelerated asynchronous implemenation is also available, lower its priority to 250 (i.e., below the asynchronous module's priority of 300). Signed-off-by: Ard Biesheuvel Signed-off-by: Catalin Marinas --- arch/arm64/crypto/aes-ce-cipher.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/crypto/aes-ce-cipher.c b/arch/arm64/crypto/aes-ce-cipher.c index ce47792a983d..f7bd9bf0bbb3 100644 --- a/arch/arm64/crypto/aes-ce-cipher.c +++ b/arch/arm64/crypto/aes-ce-cipher.c @@ -237,7 +237,7 @@ EXPORT_SYMBOL(ce_aes_setkey); static struct crypto_alg aes_alg = { .cra_name = "aes", .cra_driver_name = "aes-ce", - .cra_priority = 300, + .cra_priority = 250, .cra_flags = CRYPTO_ALG_TYPE_CIPHER, .cra_blocksize = AES_BLOCK_SIZE, .cra_ctxsize = sizeof(struct crypto_aes_ctx), From 0b2aa5b80bbf4d0fb5daa1fb83ff637daa12d552 Mon Sep 17 00:00:00 2001 From: Laura Abbott Date: Thu, 12 Nov 2015 12:21:10 -0800 Subject: [PATCH 249/335] arm64: Fix R/O permissions in mark_rodata_ro The permissions in mark_rodata_ro trigger a build error with STRICT_MM_TYPECHECKS. Fix this by introducing PAGE_KERNEL_ROX for the same reasons as PAGE_KERNEL_RO. From Ard: "PAGE_KERNEL_EXEC has PTE_WRITE set as well, making the range writeable under the ARMv8.1 DBM feature, that manages the dirty bit in hardware (writing to a page with the PTE_RDONLY and PTE_WRITE bits both set will clear the PTE_RDONLY bit in that case)" Signed-off-by: Laura Abbott Acked-by: Ard Biesheuvel Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/pgtable.h | 1 + arch/arm64/mm/mmu.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h index 9819a9426b69..7e074f93f383 100644 --- a/arch/arm64/include/asm/pgtable.h +++ b/arch/arm64/include/asm/pgtable.h @@ -81,6 +81,7 @@ extern void __pgd_error(const char *file, int line, unsigned long val); #define PAGE_KERNEL __pgprot(_PAGE_DEFAULT | PTE_PXN | PTE_UXN | PTE_DIRTY | PTE_WRITE) #define PAGE_KERNEL_RO __pgprot(_PAGE_DEFAULT | PTE_PXN | PTE_UXN | PTE_DIRTY | PTE_RDONLY) +#define PAGE_KERNEL_ROX __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_RDONLY) #define PAGE_KERNEL_EXEC __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE) #define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT) diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c index 32ddd893da9a..abb66f84d4ac 100644 --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -475,7 +475,7 @@ void mark_rodata_ro(void) { create_mapping_late(__pa(_stext), (unsigned long)_stext, (unsigned long)_etext - (unsigned long)_stext, - PAGE_KERNEL_EXEC | PTE_RDONLY); + PAGE_KERNEL_ROX); } #endif From 00490c22b1b59b168d89de798f0790bfd9541515 Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst Date: Mon, 16 Nov 2015 14:42:12 +0100 Subject: [PATCH 250/335] drm/i915: Consider SPLL as another shared pll, v2. When diagnosing a unrelated bug for someone on irc, it would seem the hardware can be brought up by the BIOS with the embedded displayport using the SPLL for spread spectrum. Right now this is not handled well in i915, and it calculates the crtc needs to be reprogrammed on the first modeset without SSC, but the SPLL itself was kept active. Fix this by exposing SPLL as a shared pll that will not be returned by intel_get_shared_dpll; you have to know it exists to use it. Changes since v1: - Create a separate dpll_hw_state.spll for spll, and use separate pll functions for spll. Tested-by: Emil Renner Berthing Tested-by: Gabriel Feceoru Reviewed-by: Daniel Vetter Signed-off-by: Maarten Lankhorst Link: http://patchwork.freedesktop.org/patch/msgid/1447681332-6318-1-git-send-email-maarten.lankhorst@linux.intel.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_drv.h | 3 ++ drivers/gpu/drm/i915/intel_crt.c | 31 ++---------- drivers/gpu/drm/i915/intel_ddi.c | 75 ++++++++++++++++++++++++---- drivers/gpu/drm/i915/intel_display.c | 15 ++++-- 4 files changed, 83 insertions(+), 41 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 8afda459a26e..84b5db257f1b 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -351,6 +351,8 @@ enum intel_dpll_id { /* hsw/bdw */ DPLL_ID_WRPLL1 = 0, DPLL_ID_WRPLL2 = 1, + DPLL_ID_SPLL = 2, + /* skl */ DPLL_ID_SKL_DPLL1 = 0, DPLL_ID_SKL_DPLL2 = 1, @@ -367,6 +369,7 @@ struct intel_dpll_hw_state { /* hsw, bdw */ uint32_t wrpll; + uint32_t spll; /* skl */ /* diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index b84aaa0bb48a..6a2c76e367a5 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -138,18 +138,6 @@ static void hsw_crt_get_config(struct intel_encoder *encoder, pipe_config->base.adjusted_mode.flags |= intel_crt_get_flags(encoder); } -static void hsw_crt_pre_enable(struct intel_encoder *encoder) -{ - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; - - WARN(I915_READ(SPLL_CTL) & SPLL_PLL_ENABLE, "SPLL already enabled\n"); - I915_WRITE(SPLL_CTL, - SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz | SPLL_PLL_SSC); - POSTING_READ(SPLL_CTL); - udelay(20); -} - /* Note: The caller is required to filter out dpms modes not supported by the * platform. */ static void intel_crt_set_dpms(struct intel_encoder *encoder, int mode) @@ -216,19 +204,6 @@ static void pch_post_disable_crt(struct intel_encoder *encoder) intel_disable_crt(encoder); } -static void hsw_crt_post_disable(struct intel_encoder *encoder) -{ - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; - uint32_t val; - - DRM_DEBUG_KMS("Disabling SPLL\n"); - val = I915_READ(SPLL_CTL); - WARN_ON(!(val & SPLL_PLL_ENABLE)); - I915_WRITE(SPLL_CTL, val & ~SPLL_PLL_ENABLE); - POSTING_READ(SPLL_CTL); -} - static void intel_enable_crt(struct intel_encoder *encoder) { struct intel_crt *crt = intel_encoder_to_crt(encoder); @@ -280,6 +255,10 @@ static bool intel_crt_compute_config(struct intel_encoder *encoder, if (HAS_DDI(dev)) { pipe_config->ddi_pll_sel = PORT_CLK_SEL_SPLL; pipe_config->port_clock = 135000 * 2; + + pipe_config->dpll_hw_state.wrpll = 0; + pipe_config->dpll_hw_state.spll = + SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz | SPLL_PLL_SSC; } return true; @@ -860,8 +839,6 @@ void intel_crt_init(struct drm_device *dev) if (HAS_DDI(dev)) { crt->base.get_config = hsw_crt_get_config; crt->base.get_hw_state = intel_ddi_get_hw_state; - crt->base.pre_enable = hsw_crt_pre_enable; - crt->base.post_disable = hsw_crt_post_disable; } else { crt->base.get_config = intel_crt_get_config; crt->base.get_hw_state = intel_crt_get_hw_state; diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index b25e99a432fb..a6752a61d99f 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -1286,6 +1286,18 @@ hsw_ddi_pll_select(struct intel_crtc *intel_crtc, } crtc_state->ddi_pll_sel = PORT_CLK_SEL_WRPLL(pll->id); + } else if (crtc_state->ddi_pll_sel == PORT_CLK_SEL_SPLL) { + struct drm_atomic_state *state = crtc_state->base.state; + struct intel_shared_dpll_config *spll = + &intel_atomic_get_shared_dpll_state(state)[DPLL_ID_SPLL]; + + if (spll->crtc_mask && + WARN_ON(spll->hw_state.spll != crtc_state->dpll_hw_state.spll)) + return false; + + crtc_state->shared_dpll = DPLL_ID_SPLL; + spll->hw_state.spll = crtc_state->dpll_hw_state.spll; + spll->crtc_mask |= 1 << intel_crtc->pipe; } return true; @@ -2437,7 +2449,7 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder) } } -static void hsw_ddi_pll_enable(struct drm_i915_private *dev_priv, +static void hsw_ddi_wrpll_enable(struct drm_i915_private *dev_priv, struct intel_shared_dpll *pll) { I915_WRITE(WRPLL_CTL(pll->id), pll->config.hw_state.wrpll); @@ -2445,8 +2457,16 @@ static void hsw_ddi_pll_enable(struct drm_i915_private *dev_priv, udelay(20); } -static void hsw_ddi_pll_disable(struct drm_i915_private *dev_priv, +static void hsw_ddi_spll_enable(struct drm_i915_private *dev_priv, struct intel_shared_dpll *pll) +{ + I915_WRITE(SPLL_CTL, pll->config.hw_state.spll); + POSTING_READ(SPLL_CTL); + udelay(20); +} + +static void hsw_ddi_wrpll_disable(struct drm_i915_private *dev_priv, + struct intel_shared_dpll *pll) { uint32_t val; @@ -2455,9 +2475,19 @@ static void hsw_ddi_pll_disable(struct drm_i915_private *dev_priv, POSTING_READ(WRPLL_CTL(pll->id)); } -static bool hsw_ddi_pll_get_hw_state(struct drm_i915_private *dev_priv, - struct intel_shared_dpll *pll, - struct intel_dpll_hw_state *hw_state) +static void hsw_ddi_spll_disable(struct drm_i915_private *dev_priv, + struct intel_shared_dpll *pll) +{ + uint32_t val; + + val = I915_READ(SPLL_CTL); + I915_WRITE(SPLL_CTL, val & ~SPLL_PLL_ENABLE); + POSTING_READ(SPLL_CTL); +} + +static bool hsw_ddi_wrpll_get_hw_state(struct drm_i915_private *dev_priv, + struct intel_shared_dpll *pll, + struct intel_dpll_hw_state *hw_state) { uint32_t val; @@ -2470,25 +2500,50 @@ static bool hsw_ddi_pll_get_hw_state(struct drm_i915_private *dev_priv, return val & WRPLL_PLL_ENABLE; } +static bool hsw_ddi_spll_get_hw_state(struct drm_i915_private *dev_priv, + struct intel_shared_dpll *pll, + struct intel_dpll_hw_state *hw_state) +{ + uint32_t val; + + if (!intel_display_power_is_enabled(dev_priv, POWER_DOMAIN_PLLS)) + return false; + + val = I915_READ(SPLL_CTL); + hw_state->spll = val; + + return val & SPLL_PLL_ENABLE; +} + + static const char * const hsw_ddi_pll_names[] = { "WRPLL 1", "WRPLL 2", + "SPLL" }; static void hsw_shared_dplls_init(struct drm_i915_private *dev_priv) { int i; - dev_priv->num_shared_dpll = 2; + dev_priv->num_shared_dpll = 3; - for (i = 0; i < dev_priv->num_shared_dpll; i++) { + for (i = 0; i < 2; i++) { dev_priv->shared_dplls[i].id = i; dev_priv->shared_dplls[i].name = hsw_ddi_pll_names[i]; - dev_priv->shared_dplls[i].disable = hsw_ddi_pll_disable; - dev_priv->shared_dplls[i].enable = hsw_ddi_pll_enable; + dev_priv->shared_dplls[i].disable = hsw_ddi_wrpll_disable; + dev_priv->shared_dplls[i].enable = hsw_ddi_wrpll_enable; dev_priv->shared_dplls[i].get_hw_state = - hsw_ddi_pll_get_hw_state; + hsw_ddi_wrpll_get_hw_state; } + + /* SPLL is special, but needs to be initialized anyway.. */ + dev_priv->shared_dplls[i].id = i; + dev_priv->shared_dplls[i].name = hsw_ddi_pll_names[i]; + dev_priv->shared_dplls[i].disable = hsw_ddi_spll_disable; + dev_priv->shared_dplls[i].enable = hsw_ddi_spll_enable; + dev_priv->shared_dplls[i].get_hw_state = hsw_ddi_spll_get_hw_state; + } static const char * const skl_ddi_pll_names[] = { diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 8f80bbeb1313..002e2a41ce1b 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -4239,6 +4239,7 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc, struct intel_shared_dpll *pll; struct intel_shared_dpll_config *shared_dpll; enum intel_dpll_id i; + int max = dev_priv->num_shared_dpll; shared_dpll = intel_atomic_get_shared_dpll_state(crtc_state->base.state); @@ -4273,9 +4274,11 @@ struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc, WARN_ON(shared_dpll[i].crtc_mask); goto found; - } + } else if (INTEL_INFO(dev_priv)->gen < 9 && HAS_DDI(dev_priv)) + /* Do not consider SPLL */ + max = 2; - for (i = 0; i < dev_priv->num_shared_dpll; i++) { + for (i = 0; i < max; i++) { pll = &dev_priv->shared_dplls[i]; /* Only want to check enabled timings first */ @@ -9725,6 +9728,8 @@ static void haswell_get_ddi_pll(struct drm_i915_private *dev_priv, case PORT_CLK_SEL_WRPLL2: pipe_config->shared_dpll = DPLL_ID_WRPLL2; break; + case PORT_CLK_SEL_SPLL: + pipe_config->shared_dpll = DPLL_ID_SPLL; } } @@ -12005,9 +12010,10 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc, pipe_config->dpll_hw_state.cfgcr1, pipe_config->dpll_hw_state.cfgcr2); } else if (HAS_DDI(dev)) { - DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: wrpll: 0x%x\n", + DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: wrpll: 0x%x spll: 0x%x\n", pipe_config->ddi_pll_sel, - pipe_config->dpll_hw_state.wrpll); + pipe_config->dpll_hw_state.wrpll, + pipe_config->dpll_hw_state.spll); } else { DRM_DEBUG_KMS("dpll_hw_state: dpll: 0x%x, dpll_md: 0x%x, " "fp0: 0x%x, fp1: 0x%x\n", @@ -12530,6 +12536,7 @@ intel_pipe_config_compare(struct drm_device *dev, PIPE_CONF_CHECK_X(dpll_hw_state.fp0); PIPE_CONF_CHECK_X(dpll_hw_state.fp1); PIPE_CONF_CHECK_X(dpll_hw_state.wrpll); + PIPE_CONF_CHECK_X(dpll_hw_state.spll); PIPE_CONF_CHECK_X(dpll_hw_state.ctrl1); PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr1); PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr2); From 2269a395790dd1ed3ae6e7708d76731e92fd46ff Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Tue, 17 Nov 2015 11:02:47 +0100 Subject: [PATCH 251/335] drm/amdgpu: fix typo in firmware name MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Christian König Reviewed-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index b93b649d9e36..98ee39001c45 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -40,7 +40,7 @@ static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); -MODULE_FIRMWARE("radeon/boniare_mc.bin"); +MODULE_FIRMWARE("radeon/bonaire_mc.bin"); MODULE_FIRMWARE("radeon/hawaii_mc.bin"); /** From 984810fc45389c545719fbb4219e8a12b27032a4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Sat, 14 Nov 2015 21:05:35 +0100 Subject: [PATCH 252/335] drm/amdgpu: cleanup scheduler command submission MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Unify the two code path again, cause they do pretty much the same thing. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Reviewed-by: Junwei Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 63 ++++++++++---------------- 2 files changed, 25 insertions(+), 39 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 0f187027c753..0ce5c1e45ec8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1210,6 +1210,7 @@ struct amdgpu_cs_parser { /* relocations */ struct amdgpu_bo_list_entry *vm_bos; struct list_head validated; + struct fence *fence; struct amdgpu_ib *ibs; uint32_t num_ibs; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 6096effd6a56..3afcf0237c25 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -439,8 +439,18 @@ static int cmp_size_smaller_first(void *priv, struct list_head *a, return (int)la->robj->tbo.num_pages - (int)lb->robj->tbo.num_pages; } -static void amdgpu_cs_parser_fini_early(struct amdgpu_cs_parser *parser, int error, bool backoff) +/** + * cs_parser_fini() - clean parser states + * @parser: parser structure holding parsing context. + * @error: error number + * + * If error is set than unvalidate buffer, otherwise just free memory + * used by parsing context. + **/ +static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bool backoff) { + unsigned i; + if (!error) { /* Sort the buffer list from the smallest to largest buffer, * which affects the order of buffers in the LRU list. @@ -455,17 +465,13 @@ static void amdgpu_cs_parser_fini_early(struct amdgpu_cs_parser *parser, int err list_sort(NULL, &parser->validated, cmp_size_smaller_first); ttm_eu_fence_buffer_objects(&parser->ticket, - &parser->validated, - &parser->ibs[parser->num_ibs-1].fence->base); + &parser->validated, + parser->fence); } else if (backoff) { ttm_eu_backoff_reservation(&parser->ticket, &parser->validated); } -} - -static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser) -{ - unsigned i; + fence_put(parser->fence); if (parser->ctx) amdgpu_ctx_put(parser->ctx); @@ -484,20 +490,6 @@ static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser) drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base); } -/** - * cs_parser_fini() - clean parser states - * @parser: parser structure holding parsing context. - * @error: error number - * - * If error is set than unvalidate buffer, otherwise just free memory - * used by parsing context. - **/ -static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bool backoff) -{ - amdgpu_cs_parser_fini_early(parser, error, backoff); - amdgpu_cs_parser_fini_late(parser); -} - static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p, struct amdgpu_vm *vm) { @@ -582,15 +574,9 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev, } r = amdgpu_bo_vm_update_pte(parser, vm); - if (r) { - goto out; - } - amdgpu_cs_sync_rings(parser); - if (!amdgpu_enable_scheduler) - r = amdgpu_ib_schedule(adev, parser->num_ibs, parser->ibs, - parser->filp); + if (!r) + amdgpu_cs_sync_rings(parser); -out: return r; } @@ -881,7 +867,7 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) goto out; } job->base.s_fence = fence; - fence_get(&fence->base); + parser.fence = fence_get(&fence->base); cs->out.handle = amdgpu_ctx_add_fence(parser.ctx, ring, &fence->base); @@ -890,17 +876,16 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) trace_amdgpu_cs_ioctl(job); amd_sched_entity_push_job(&job->base); - list_sort(NULL, &parser.validated, cmp_size_smaller_first); - ttm_eu_fence_buffer_objects(&parser.ticket, &parser.validated, - &fence->base); - fence_put(&fence->base); + } else { + struct amdgpu_fence *fence; - amdgpu_cs_parser_fini_late(&parser); - mutex_unlock(&vm->mutex); - return 0; + r = amdgpu_ib_schedule(adev, parser.num_ibs, parser.ibs, + parser.filp); + fence = parser.ibs[parser.num_ibs - 1].fence; + parser.fence = fence_get(&fence->base); + cs->out.handle = parser.ibs[parser.num_ibs - 1].sequence; } - cs->out.handle = parser.ibs[parser.num_ibs - 1].sequence; out: amdgpu_cs_parser_fini(&parser, r, reserved_buffers); mutex_unlock(&vm->mutex); From eeed25ab83bbd68b2c9dba3cc2ce5913537594f6 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Sun, 15 Nov 2015 15:50:07 +0100 Subject: [PATCH 253/335] drm/amdgpu: remove unused VM manager field MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Christian König Reviewed-by: Chunming Zhou Reviewed-by: Junwei Zhang --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 0ce5c1e45ec8..819d1473de30 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -967,8 +967,6 @@ struct amdgpu_vm_manager { u64 vram_base_offset; /* is vm enabled? */ bool enabled; - /* for hw to save the PD addr on suspend/resume */ - uint32_t saved_table_addr[AMDGPU_NUM_VM]; /* vm pte handling */ const struct amdgpu_vm_pte_funcs *vm_pte_funcs; struct amdgpu_ring *vm_pte_funcs_ring; From 8b4fb00b5db271bb1e117a078d5d7a161d8d8af1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Sun, 15 Nov 2015 16:04:16 +0100 Subject: [PATCH 254/335] drm/amdgpu: cleanup VM coding style MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the indentation and move the VM functions to the structures. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Reviewed-by: Junwei Zhang Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 98 ++++++++++++++--------------- 1 file changed, 48 insertions(+), 50 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 819d1473de30..d1e16726e225 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -916,8 +916,8 @@ struct amdgpu_ring { #define AMDGPU_VM_FAULT_STOP_ALWAYS 2 struct amdgpu_vm_pt { - struct amdgpu_bo *bo; - uint64_t addr; + struct amdgpu_bo *bo; + uint64_t addr; }; struct amdgpu_vm_id { @@ -959,19 +959,60 @@ struct amdgpu_vm { }; struct amdgpu_vm_manager { - struct fence *active[AMDGPU_NUM_VM]; - uint32_t max_pfn; + struct fence *active[AMDGPU_NUM_VM]; + uint32_t max_pfn; /* number of VMIDs */ - unsigned nvm; + unsigned nvm; /* vram base address for page table entry */ - u64 vram_base_offset; + u64 vram_base_offset; /* is vm enabled? */ - bool enabled; + bool enabled; /* vm pte handling */ const struct amdgpu_vm_pte_funcs *vm_pte_funcs; struct amdgpu_ring *vm_pte_funcs_ring; }; +int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm); +void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); +struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct list_head *head); +int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, + struct amdgpu_sync *sync); +void amdgpu_vm_flush(struct amdgpu_ring *ring, + struct amdgpu_vm *vm, + struct fence *updates); +void amdgpu_vm_fence(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct fence *fence); +uint64_t amdgpu_vm_map_gart(struct amdgpu_device *adev, uint64_t addr); +int amdgpu_vm_update_page_directory(struct amdgpu_device *adev, + struct amdgpu_vm *vm); +int amdgpu_vm_clear_freed(struct amdgpu_device *adev, + struct amdgpu_vm *vm); +int amdgpu_vm_clear_invalids(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct amdgpu_sync *sync); +int amdgpu_vm_bo_update(struct amdgpu_device *adev, + struct amdgpu_bo_va *bo_va, + struct ttm_mem_reg *mem); +void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, + struct amdgpu_bo *bo); +struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, + struct amdgpu_bo *bo); +struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct amdgpu_bo *bo); +int amdgpu_vm_bo_map(struct amdgpu_device *adev, + struct amdgpu_bo_va *bo_va, + uint64_t addr, uint64_t offset, + uint64_t size, uint32_t flags); +int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, + struct amdgpu_bo_va *bo_va, + uint64_t addr); +void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, + struct amdgpu_bo_va *bo_va); +int amdgpu_vm_free_job(struct amdgpu_job *job); + /* * context related structures */ @@ -2311,49 +2352,6 @@ int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe, long amdgpu_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); -/* - * vm - */ -int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm); -void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); -struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct list_head *head); -int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, - struct amdgpu_sync *sync); -void amdgpu_vm_flush(struct amdgpu_ring *ring, - struct amdgpu_vm *vm, - struct fence *updates); -void amdgpu_vm_fence(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct fence *fence); -uint64_t amdgpu_vm_map_gart(struct amdgpu_device *adev, uint64_t addr); -int amdgpu_vm_update_page_directory(struct amdgpu_device *adev, - struct amdgpu_vm *vm); -int amdgpu_vm_clear_freed(struct amdgpu_device *adev, - struct amdgpu_vm *vm); -int amdgpu_vm_clear_invalids(struct amdgpu_device *adev, - struct amdgpu_vm *vm, struct amdgpu_sync *sync); -int amdgpu_vm_bo_update(struct amdgpu_device *adev, - struct amdgpu_bo_va *bo_va, - struct ttm_mem_reg *mem); -void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, - struct amdgpu_bo *bo); -struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, - struct amdgpu_bo *bo); -struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_bo *bo); -int amdgpu_vm_bo_map(struct amdgpu_device *adev, - struct amdgpu_bo_va *bo_va, - uint64_t addr, uint64_t offset, - uint64_t size, uint32_t flags); -int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, - struct amdgpu_bo_va *bo_va, - uint64_t addr); -void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, - struct amdgpu_bo_va *bo_va); -int amdgpu_vm_free_job(struct amdgpu_job *job); /* * functions used by amdgpu_encoder.c */ From ea89f8c9e8ba8a7b75446eef36917da50d2186d9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Sun, 15 Nov 2015 20:52:06 +0100 Subject: [PATCH 255/335] drm/amdgpu: move VM manager clean into the VM code again MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It's not a good idea to duplicate that code. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 15 +++++++++++++++ drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c | 8 ++------ drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 8 ++------ 4 files changed, 20 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index d1e16726e225..beb74854a8a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -972,6 +972,7 @@ struct amdgpu_vm_manager { struct amdgpu_ring *vm_pte_funcs_ring; }; +void amdgpu_vm_manager_fini(struct amdgpu_device *adev); int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index a9fcc995d27e..6bb209bc0d36 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1318,3 +1318,18 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) mutex_destroy(&vm->mutex); } + +/** + * amdgpu_vm_manager_fini - cleanup VM manager + * + * @adev: amdgpu_device pointer + * + * Cleanup the VM manager and free resources. + */ +void amdgpu_vm_manager_fini(struct amdgpu_device *adev) +{ + unsigned i; + + for (i = 0; i < AMDGPU_NUM_VM; ++i) + fence_put(adev->vm_manager.active[i]); +} diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 98ee39001c45..7427d8cd4c43 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -961,12 +961,10 @@ static int gmc_v7_0_sw_init(void *handle) static int gmc_v7_0_sw_fini(void *handle) { - int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->vm_manager.enabled) { - for (i = 0; i < AMDGPU_NUM_VM; ++i) - fence_put(adev->vm_manager.active[i]); + amdgpu_vm_manager_fini(adev); gmc_v7_0_vm_fini(adev); adev->vm_manager.enabled = false; } @@ -1011,12 +1009,10 @@ static int gmc_v7_0_hw_fini(void *handle) static int gmc_v7_0_suspend(void *handle) { - int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->vm_manager.enabled) { - for (i = 0; i < AMDGPU_NUM_VM; ++i) - fence_put(adev->vm_manager.active[i]); + amdgpu_vm_manager_fini(adev); gmc_v7_0_vm_fini(adev); adev->vm_manager.enabled = false; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index c9209b45d901..cb0e50ebb528 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -980,12 +980,10 @@ static int gmc_v8_0_sw_init(void *handle) static int gmc_v8_0_sw_fini(void *handle) { - int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->vm_manager.enabled) { - for (i = 0; i < AMDGPU_NUM_VM; ++i) - fence_put(adev->vm_manager.active[i]); + amdgpu_vm_manager_fini(adev); gmc_v8_0_vm_fini(adev); adev->vm_manager.enabled = false; } @@ -1032,12 +1030,10 @@ static int gmc_v8_0_hw_fini(void *handle) static int gmc_v8_0_suspend(void *handle) { - int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; if (adev->vm_manager.enabled) { - for (i = 0; i < AMDGPU_NUM_VM; ++i) - fence_put(adev->vm_manager.active[i]); + amdgpu_vm_manager_fini(adev); gmc_v8_0_vm_fini(adev); adev->vm_manager.enabled = false; } From 1c16c0a7b26c6c905dc79c4194135ca2f360f0f5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Sat, 14 Nov 2015 21:31:40 +0100 Subject: [PATCH 256/335] drm/amdgpu: keep the owner for VMIDs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We don't need the last VM use any more, keep the owner directly. Signed-off-by: Christian König Reviewed-by: Chunming Zhou --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 8 ++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 33 +++++++++++++++----------- 2 files changed, 24 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index beb74854a8a3..a5692624070a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -925,8 +925,6 @@ struct amdgpu_vm_id { uint64_t pd_gpu_addr; /* last flushed PD/PT update */ struct fence *flushed_updates; - /* last use of vmid */ - struct fence *last_id_use; }; struct amdgpu_vm { @@ -959,7 +957,11 @@ struct amdgpu_vm { }; struct amdgpu_vm_manager { - struct fence *active[AMDGPU_NUM_VM]; + struct { + struct fence *active; + atomic_long_t owner; + } ids[AMDGPU_NUM_VM]; + uint32_t max_pfn; /* number of VMIDs */ unsigned nvm; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 6bb209bc0d36..0bdbb2480f9b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -143,10 +143,15 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, unsigned i; /* check if the id is still valid */ - if (vm_id->id && vm_id->last_id_use && - vm_id->last_id_use == adev->vm_manager.active[vm_id->id]) { - trace_amdgpu_vm_grab_id(vm_id->id, ring->idx); - return 0; + if (vm_id->id) { + unsigned id = vm_id->id; + long owner; + + owner = atomic_long_read(&adev->vm_manager.ids[id].owner); + if (owner == (long)vm) { + trace_amdgpu_vm_grab_id(vm_id->id, ring->idx); + return 0; + } } /* we definately need to flush */ @@ -154,7 +159,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, /* skip over VMID 0, since it is the system VM */ for (i = 1; i < adev->vm_manager.nvm; ++i) { - struct fence *fence = adev->vm_manager.active[i]; + struct fence *fence = adev->vm_manager.ids[i].active; struct amdgpu_ring *fring; if (fence == NULL) { @@ -176,7 +181,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (choices[i]) { struct fence *fence; - fence = adev->vm_manager.active[choices[i]]; + fence = adev->vm_manager.ids[choices[i]].active; vm_id->id = choices[i]; trace_amdgpu_vm_grab_id(choices[i], ring->idx); @@ -246,11 +251,9 @@ void amdgpu_vm_fence(struct amdgpu_device *adev, struct amdgpu_ring *ring = amdgpu_ring_from_fence(fence); unsigned vm_id = vm->ids[ring->idx].id; - fence_put(adev->vm_manager.active[vm_id]); - adev->vm_manager.active[vm_id] = fence_get(fence); - - fence_put(vm->ids[ring->idx].last_id_use); - vm->ids[ring->idx].last_id_use = fence_get(fence); + fence_put(adev->vm_manager.ids[vm_id].active); + adev->vm_manager.ids[vm_id].active = fence_get(fence); + atomic_long_set(&adev->vm_manager.ids[vm_id].owner, (long)vm); } /** @@ -1238,7 +1241,6 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { vm->ids[i].id = 0; vm->ids[i].flushed_updates = NULL; - vm->ids[i].last_id_use = NULL; } mutex_init(&vm->mutex); vm->va = RB_ROOT; @@ -1312,8 +1314,11 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) fence_put(vm->page_directory_fence); for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { + unsigned id = vm->ids[i].id; + + atomic_long_cmpxchg(&adev->vm_manager.ids[id].owner, + (long)vm, 0); fence_put(vm->ids[i].flushed_updates); - fence_put(vm->ids[i].last_id_use); } mutex_destroy(&vm->mutex); @@ -1331,5 +1336,5 @@ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) unsigned i; for (i = 0; i < AMDGPU_NUM_VM; ++i) - fence_put(adev->vm_manager.active[i]); + fence_put(adev->vm_manager.ids[i].active); } From c25867dfabf045a8148fd179fa759bb17f670e42 Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Fri, 13 Nov 2015 13:32:01 +0800 Subject: [PATCH 257/335] drm/amdgpu: add lock for interval tree in vm MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: I62b892a22af37b32e6b4aefca80a25cf45426ed2 Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 15 +++++++++++++-- 2 files changed, 15 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index a5692624070a..306f75700bf8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -954,6 +954,8 @@ struct amdgpu_vm { /* for id and flush management per ring */ struct amdgpu_vm_id ids[AMDGPU_MAX_RINGS]; + /* for interval tree */ + spinlock_t it_lock; }; struct amdgpu_vm_manager { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 0bdbb2480f9b..0513f3fed2c2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1028,7 +1028,9 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, saddr /= AMDGPU_GPU_PAGE_SIZE; eaddr /= AMDGPU_GPU_PAGE_SIZE; + spin_lock(&vm->it_lock); it = interval_tree_iter_first(&vm->va, saddr, eaddr - 1); + spin_unlock(&vm->it_lock); if (it) { struct amdgpu_bo_va_mapping *tmp; tmp = container_of(it, struct amdgpu_bo_va_mapping, it); @@ -1055,7 +1057,9 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, mapping->flags = flags; list_add(&mapping->list, &bo_va->invalids); + spin_lock(&vm->it_lock); interval_tree_insert(&mapping->it, &vm->va); + spin_unlock(&vm->it_lock); trace_amdgpu_vm_bo_map(bo_va, mapping); /* Make sure the page tables are allocated */ @@ -1101,7 +1105,9 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, error_free: list_del(&mapping->list); + spin_lock(&vm->it_lock); interval_tree_remove(&mapping->it, &vm->va); + spin_unlock(&vm->it_lock); trace_amdgpu_vm_bo_unmap(bo_va, mapping); kfree(mapping); @@ -1151,7 +1157,9 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, } list_del(&mapping->list); + spin_lock(&vm->it_lock); interval_tree_remove(&mapping->it, &vm->va); + spin_unlock(&vm->it_lock); trace_amdgpu_vm_bo_unmap(bo_va, mapping); if (valid) @@ -1187,13 +1195,17 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, list_for_each_entry_safe(mapping, next, &bo_va->valids, list) { list_del(&mapping->list); + spin_lock(&vm->it_lock); interval_tree_remove(&mapping->it, &vm->va); + spin_unlock(&vm->it_lock); trace_amdgpu_vm_bo_unmap(bo_va, mapping); list_add(&mapping->list, &vm->freed); } list_for_each_entry_safe(mapping, next, &bo_va->invalids, list) { list_del(&mapping->list); + spin_lock(&vm->it_lock); interval_tree_remove(&mapping->it, &vm->va); + spin_unlock(&vm->it_lock); kfree(mapping); } @@ -1248,7 +1260,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) INIT_LIST_HEAD(&vm->invalidated); INIT_LIST_HEAD(&vm->cleared); INIT_LIST_HEAD(&vm->freed); - + spin_lock_init(&vm->it_lock); pd_size = amdgpu_vm_directory_size(adev); pd_entries = amdgpu_vm_num_pdes(adev); @@ -1312,7 +1324,6 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) amdgpu_bo_unref(&vm->page_directory); fence_put(vm->page_directory_fence); - for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { unsigned id = vm->ids[i].id; From ef9f0a83d68cecca241ee5d61b797d598722772e Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Fri, 13 Nov 2015 13:43:22 +0800 Subject: [PATCH 258/335] drm/amdgpu: move bo_reserve out of amdgpu_vm_clear_bo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: Ifbb0c06680494bfa04d0be5e5941d31ae2e5ef28 Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 33 ++++++++++++++------------ 1 file changed, 18 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 0513f3fed2c2..e6dc19bc2dd8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -332,6 +332,8 @@ int amdgpu_vm_free_job(struct amdgpu_job *job) * * @adev: amdgpu_device pointer * @bo: bo to clear + * + * need to reserve bo first before calling it. */ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, struct amdgpu_bo *bo) @@ -343,24 +345,20 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, uint64_t addr; int r; - r = amdgpu_bo_reserve(bo, false); - if (r) - return r; - r = reservation_object_reserve_shared(bo->tbo.resv); if (r) return r; r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); if (r) - goto error_unreserve; + goto error; addr = amdgpu_bo_gpu_offset(bo); entries = amdgpu_bo_size(bo) / 8; ib = kzalloc(sizeof(struct amdgpu_ib), GFP_KERNEL); if (!ib) - goto error_unreserve; + goto error; r = amdgpu_ib_get(ring, NULL, entries * 2 + 64, ib); if (r) @@ -378,16 +376,14 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, if (!r) amdgpu_bo_fence(bo, fence, true); fence_put(fence); - if (amdgpu_enable_scheduler) { - amdgpu_bo_unreserve(bo); + if (amdgpu_enable_scheduler) return 0; - } + error_free: amdgpu_ib_free(adev, ib); kfree(ib); -error_unreserve: - amdgpu_bo_unreserve(bo); +error: return r; } @@ -1087,11 +1083,12 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, AMDGPU_GEM_DOMAIN_VRAM, AMDGPU_GEM_CREATE_NO_CPU_ACCESS, NULL, resv, &pt); - ww_mutex_unlock(&resv->lock); - if (r) + if (r) { + ww_mutex_unlock(&resv->lock); goto error_free; - + } r = amdgpu_vm_clear_bo(adev, pt); + ww_mutex_unlock(&resv->lock); if (r) { amdgpu_bo_unref(&pt); goto error_free; @@ -1280,8 +1277,14 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) NULL, NULL, &vm->page_directory); if (r) return r; - + r = amdgpu_bo_reserve(vm->page_directory, false); + if (r) { + amdgpu_bo_unref(&vm->page_directory); + vm->page_directory = NULL; + return r; + } r = amdgpu_vm_clear_bo(adev, vm->page_directory); + amdgpu_bo_unreserve(vm->page_directory); if (r) { amdgpu_bo_unref(&vm->page_directory); vm->page_directory = NULL; From 49b02b180a541d6fb31031aaffe75496e9238942 Mon Sep 17 00:00:00 2001 From: Chunming Zhou Date: Fri, 13 Nov 2015 14:18:38 +0800 Subject: [PATCH 259/335] drm/amdgpu: reserve/unreserve objects out of map/unmap operations MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change-Id: Id6514f2fb6e002437fdbe99353d5d35f4ac736c7 Signed-off-by: Chunming Zhou Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 21 +++++++++++++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 29 ++++++------------------- 2 files changed, 25 insertions(+), 25 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 16dca46314bd..00c5b580f56c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -515,6 +515,9 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, struct amdgpu_fpriv *fpriv = filp->driver_priv; struct amdgpu_bo *rbo; struct amdgpu_bo_va *bo_va; + struct ttm_validate_buffer tv, tv_pd; + struct ww_acquire_ctx ticket; + struct list_head list, duplicates; uint32_t invalid_flags, va_flags = 0; int r = 0; @@ -552,7 +555,18 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, return -ENOENT; mutex_lock(&fpriv->vm.mutex); rbo = gem_to_amdgpu_bo(gobj); - r = amdgpu_bo_reserve(rbo, false); + INIT_LIST_HEAD(&list); + INIT_LIST_HEAD(&duplicates); + tv.bo = &rbo->tbo; + tv.shared = true; + list_add(&tv.head, &list); + + if (args->operation == AMDGPU_VA_OP_MAP) { + tv_pd.bo = &fpriv->vm.page_directory->tbo; + tv_pd.shared = true; + list_add(&tv_pd.head, &list); + } + r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates); if (r) { mutex_unlock(&fpriv->vm.mutex); drm_gem_object_unreference_unlocked(gobj); @@ -561,7 +575,8 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, bo_va = amdgpu_vm_bo_find(&fpriv->vm, rbo); if (!bo_va) { - amdgpu_bo_unreserve(rbo); + ttm_eu_backoff_reservation(&ticket, &list); + drm_gem_object_unreference_unlocked(gobj); mutex_unlock(&fpriv->vm.mutex); return -ENOENT; } @@ -584,7 +599,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, default: break; } - + ttm_eu_backoff_reservation(&ticket, &list); if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE)) amdgpu_gem_va_update_vm(adev, bo_va, args->operation); mutex_unlock(&fpriv->vm.mutex); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index e6dc19bc2dd8..159ce54bbd8d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -985,7 +985,7 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev, * Add a mapping of the BO at the specefied addr into the VM. * Returns 0 for success, error for failure. * - * Object has to be reserved and gets unreserved by this function! + * Object has to be reserved and unreserved outside! */ int amdgpu_vm_bo_map(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, @@ -1001,23 +1001,18 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, /* validate the parameters */ if (saddr & AMDGPU_GPU_PAGE_MASK || offset & AMDGPU_GPU_PAGE_MASK || - size == 0 || size & AMDGPU_GPU_PAGE_MASK) { - amdgpu_bo_unreserve(bo_va->bo); + size == 0 || size & AMDGPU_GPU_PAGE_MASK) return -EINVAL; - } /* make sure object fit at this offset */ eaddr = saddr + size; - if ((saddr >= eaddr) || (offset + size > amdgpu_bo_size(bo_va->bo))) { - amdgpu_bo_unreserve(bo_va->bo); + if ((saddr >= eaddr) || (offset + size > amdgpu_bo_size(bo_va->bo))) return -EINVAL; - } last_pfn = eaddr / AMDGPU_GPU_PAGE_SIZE; if (last_pfn > adev->vm_manager.max_pfn) { dev_err(adev->dev, "va above limit (0x%08X > 0x%08X)\n", last_pfn, adev->vm_manager.max_pfn); - amdgpu_bo_unreserve(bo_va->bo); return -EINVAL; } @@ -1034,14 +1029,12 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, dev_err(adev->dev, "bo %p va 0x%010Lx-0x%010Lx conflict with " "0x%010lx-0x%010lx\n", bo_va->bo, saddr, eaddr, tmp->it.start, tmp->it.last + 1); - amdgpu_bo_unreserve(bo_va->bo); r = -EINVAL; goto error; } mapping = kmalloc(sizeof(*mapping), GFP_KERNEL); if (!mapping) { - amdgpu_bo_unreserve(bo_va->bo); r = -ENOMEM; goto error; } @@ -1067,8 +1060,6 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, if (eaddr > vm->max_pde_used) vm->max_pde_used = eaddr; - amdgpu_bo_unreserve(bo_va->bo); - /* walk over the address space and allocate the page tables */ for (pt_idx = saddr; pt_idx <= eaddr; ++pt_idx) { struct reservation_object *resv = vm->page_directory->tbo.resv; @@ -1077,18 +1068,15 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, if (vm->page_tables[pt_idx].bo) continue; - ww_mutex_lock(&resv->lock, NULL); r = amdgpu_bo_create(adev, AMDGPU_VM_PTE_COUNT * 8, AMDGPU_GPU_PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM, AMDGPU_GEM_CREATE_NO_CPU_ACCESS, NULL, resv, &pt); - if (r) { - ww_mutex_unlock(&resv->lock); + if (r) goto error_free; - } + r = amdgpu_vm_clear_bo(adev, pt); - ww_mutex_unlock(&resv->lock); if (r) { amdgpu_bo_unref(&pt); goto error_free; @@ -1122,7 +1110,7 @@ error: * Remove a mapping of the BO at the specefied addr from the VM. * Returns 0 for success, error for failure. * - * Object has to be reserved and gets unreserved by this function! + * Object has to be reserved and unreserved outside! */ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, @@ -1147,10 +1135,8 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, break; } - if (&mapping->list == &bo_va->invalids) { - amdgpu_bo_unreserve(bo_va->bo); + if (&mapping->list == &bo_va->invalids) return -ENOENT; - } } list_del(&mapping->list); @@ -1163,7 +1149,6 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, list_add(&mapping->list, &vm->freed); else kfree(mapping); - amdgpu_bo_unreserve(bo_va->bo); return 0; } From a5964396190d0c40dd549c23848c282fffa5d1f2 Mon Sep 17 00:00:00 2001 From: Rajmohan Mani Date: Wed, 18 Nov 2015 10:48:20 +0200 Subject: [PATCH 260/335] xhci: Workaround to get Intel xHCI reset working more reliably Existing Intel xHCI controllers require a delay of 1 mS, after setting the CMD_RESET bit in command register, before accessing any HC registers. This allows the HC to complete the reset operation and be ready for HC register access. Without this delay, the subsequent HC register access, may result in a system hang, very rarely. Verified CherryView / Braswell platforms go through over 5000 warm reboot cycles (which was not possible without this patch), without any xHCI reset hang. Signed-off-by: Rajmohan Mani Tested-by: Joe Lawrence Cc: stable Signed-off-by: Mathias Nyman Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/xhci.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index 6e7dc6f93978..dfa44d3e8eee 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -175,6 +175,16 @@ int xhci_reset(struct xhci_hcd *xhci) command |= CMD_RESET; writel(command, &xhci->op_regs->command); + /* Existing Intel xHCI controllers require a delay of 1 mS, + * after setting the CMD_RESET bit, and before accessing any + * HC registers. This allows the HC to complete the + * reset operation and be ready for HC register access. + * Without this delay, the subsequent HC register access, + * may result in a system hang very rarely. + */ + if (xhci->quirks & XHCI_INTEL_HOST) + udelay(1000); + ret = xhci_handshake(&xhci->op_regs->command, CMD_RESET, 0, 10 * 1000 * 1000); if (ret) From 42df7215facf27be8d53e657dd4a12d4ebad0a44 Mon Sep 17 00:00:00 2001 From: Lu Baolu Date: Wed, 18 Nov 2015 10:48:21 +0200 Subject: [PATCH 261/335] usb: xhci: fix checking ep busy for CFC Function ep_ring_is_processing() checks the dequeue pointer in endpoint context to know whether an endpoint is busy with processing TRBs. This is not correct since dequeue pointer field in an endpoint context is only valid when the endpoint is in Halted or Stopped states. This buggy code causes audio noise when playing sound with USB headset connected to host controllers which support CFC (one of xhci 1.1 features). This patch should exist in stable kernel since v4.3. Reported-and-tested-by: YD Tseng Signed-off-by: Lu Baolu Cc: stable # v4.3 Signed-off-by: Mathias Nyman Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/xhci-ring.c | 32 ++++++-------------------------- 1 file changed, 6 insertions(+), 26 deletions(-) diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index fa836251ca21..6c5e8133cf87 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -3896,28 +3896,6 @@ cleanup: return ret; } -static int ep_ring_is_processing(struct xhci_hcd *xhci, - int slot_id, unsigned int ep_index) -{ - struct xhci_virt_device *xdev; - struct xhci_ring *ep_ring; - struct xhci_ep_ctx *ep_ctx; - struct xhci_virt_ep *xep; - dma_addr_t hw_deq; - - xdev = xhci->devs[slot_id]; - xep = &xhci->devs[slot_id]->eps[ep_index]; - ep_ring = xep->ring; - ep_ctx = xhci_get_ep_ctx(xhci, xdev->out_ctx, ep_index); - - if ((le32_to_cpu(ep_ctx->ep_info) & EP_STATE_MASK) != EP_STATE_RUNNING) - return 0; - - hw_deq = le64_to_cpu(ep_ctx->deq) & ~EP_CTX_CYCLE_MASK; - return (hw_deq != - xhci_trb_virt_to_dma(ep_ring->enq_seg, ep_ring->enqueue)); -} - /* * Check transfer ring to guarantee there is enough room for the urb. * Update ISO URB start_frame and interval. @@ -3983,10 +3961,12 @@ int xhci_queue_isoc_tx_prepare(struct xhci_hcd *xhci, gfp_t mem_flags, } /* Calculate the start frame and put it in urb->start_frame. */ - if (HCC_CFC(xhci->hcc_params) && - ep_ring_is_processing(xhci, slot_id, ep_index)) { - urb->start_frame = xep->next_frame_id; - goto skip_start_over; + if (HCC_CFC(xhci->hcc_params) && !list_empty(&ep_ring->td_list)) { + if ((le32_to_cpu(ep_ctx->ep_info) & EP_STATE_MASK) == + EP_STATE_RUNNING) { + urb->start_frame = xep->next_frame_id; + goto skip_start_over; + } } start_frame = readl(&xhci->run_regs->microframe_index); From dad67d5f3d0efe01d38c6cebcb6698280e51927b Mon Sep 17 00:00:00 2001 From: Mathias Nyman Date: Wed, 18 Nov 2015 10:48:22 +0200 Subject: [PATCH 262/335] xhci: Fix a race in usb2 LPM resume, blocking U3 for usb2 devices Clear device initiated resume variables once device is fully up and running in U0 state. Resume needs to be signaled for 20ms for usb2 devices before they can be moved to U0 state. An interrupt is triggered if a device initiates resume. As we handle the event in interrupt context we can not sleep for 20ms, so we instead set a resume flag, a timestamp, and start the roothub polling. The roothub code will later move the port to U0 when it finds a port in resume state with the resume flag set, and timestamp passed by 20ms. A host initiated resume is however not done in interrupt context, and host initiated resume code will directly signal resume, wait 20ms and then move the port to U0. These two codepaths can race, if we are in the middle of a host initated resume, while sleeping for 20ms, we may handle a port event and find the port in resume state. The port event handling code will assume the resume was device initiated and set the resume flag and timestamp. Root hub code will however not catch the port in resume state again as the host initated resume code has already moved the port to U0. The resume flag and timestamp will remain set for this port preventing port from suspending again (LPM setting port to U3) Fix this for now by always clearing the device initated resume parameters once port is in U0 Cc: stable Signed-off-by: Mathias Nyman Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/xhci-hub.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/drivers/usb/host/xhci-hub.c b/drivers/usb/host/xhci-hub.c index 5d2d7e954bd4..0230965fb78c 100644 --- a/drivers/usb/host/xhci-hub.c +++ b/drivers/usb/host/xhci-hub.c @@ -782,12 +782,15 @@ static u32 xhci_get_port_status(struct usb_hcd *hcd, status |= USB_PORT_STAT_SUSPEND; } } - if ((raw_port_status & PORT_PLS_MASK) == XDEV_U0 - && (raw_port_status & PORT_POWER) - && (bus_state->suspended_ports & (1 << wIndex))) { - bus_state->suspended_ports &= ~(1 << wIndex); - if (hcd->speed < HCD_USB3) - bus_state->port_c_suspend |= 1 << wIndex; + if ((raw_port_status & PORT_PLS_MASK) == XDEV_U0 && + (raw_port_status & PORT_POWER)) { + if (bus_state->suspended_ports & (1 << wIndex)) { + bus_state->suspended_ports &= ~(1 << wIndex); + if (hcd->speed < HCD_USB3) + bus_state->port_c_suspend |= 1 << wIndex; + } + bus_state->resume_done[wIndex] = 0; + clear_bit(wIndex, &bus_state->resuming_ports); } if (raw_port_status & PORT_CONNECT) { status |= USB_PORT_STAT_CONNECTION; From c139aa60c1007429335131167a0ca181e38c5668 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Wed, 18 Nov 2015 10:13:08 +0000 Subject: [PATCH 263/335] arm64: barriers: fix smp_load_acquire to work with const arguments A newly introduced function in include/net/sock.h passes a const argument to smp_load_acquire: static inline int sk_state_load(const struct sock *sk) { return smp_load_acquire(&sk->sk_state); } This cause an allmodconfig build failure, since our underlying load-acquire implementation does not handle const types correctly: include/net/sock.h: In function 'sk_state_load': ./arch/arm64/include/asm/barrier.h:71:3: error: read-only variable '___p1' used as 'asm' output asm volatile ("ldarb %w0, %1" \ This patch fixes the problem by reusing the trick in READ_ONCE that loads via a non-const member of an anonymous union. This has the advantage of allowing us to use smp_load_acquire on packed structures (e.g. arch_spinlock_t) as well as primitive types. Cc: Arnd Bergmann Cc: David Daney Cc: Eric Dumazet Reported-by: Arnd Bergmann Reported-by: David Daney Signed-off-by: Will Deacon Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/barrier.h | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/arch/arm64/include/asm/barrier.h b/arch/arm64/include/asm/barrier.h index 624f9679f4b0..9622eb48f894 100644 --- a/arch/arm64/include/asm/barrier.h +++ b/arch/arm64/include/asm/barrier.h @@ -64,27 +64,31 @@ do { \ #define smp_load_acquire(p) \ ({ \ - typeof(*p) ___p1; \ + union { typeof(*p) __val; char __c[1]; } __u; \ compiletime_assert_atomic_type(*p); \ switch (sizeof(*p)) { \ case 1: \ asm volatile ("ldarb %w0, %1" \ - : "=r" (___p1) : "Q" (*p) : "memory"); \ + : "=r" (*(__u8 *)__u.__c) \ + : "Q" (*p) : "memory"); \ break; \ case 2: \ asm volatile ("ldarh %w0, %1" \ - : "=r" (___p1) : "Q" (*p) : "memory"); \ + : "=r" (*(__u16 *)__u.__c) \ + : "Q" (*p) : "memory"); \ break; \ case 4: \ asm volatile ("ldar %w0, %1" \ - : "=r" (___p1) : "Q" (*p) : "memory"); \ + : "=r" (*(__u32 *)__u.__c) \ + : "Q" (*p) : "memory"); \ break; \ case 8: \ asm volatile ("ldar %0, %1" \ - : "=r" (___p1) : "Q" (*p) : "memory"); \ + : "=r" (*(__u64 *)__u.__c) \ + : "Q" (*p) : "memory"); \ break; \ } \ - ___p1; \ + __u.__val; \ }) #define read_barrier_depends() do { } while(0) From 1ddaa021b000220b5f2ad023e4f15ed44990974b Mon Sep 17 00:00:00 2001 From: Kamal Dasu Date: Tue, 17 Nov 2015 13:50:08 -0500 Subject: [PATCH 264/335] MAINTAINERS: brcmnand: Add co-maintainer for Broadcom SoCs Adding myself as co-maintainer of nand controller driver for the Broadcom SoCs. Signed-off-by: Kamal Dasu Acked-by: Florian Fainelli Signed-off-by: Brian Norris --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index da691aad7031..d20d6f3d6509 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2449,6 +2449,7 @@ F: drivers/firmware/broadcom/* BROADCOM STB NAND FLASH DRIVER M: Brian Norris +M: Kamal Dasu L: linux-mtd@lists.infradead.org L: bcm-kernel-feedback-list@broadcom.com S: Maintained From b57f9f34e27bf81c97b10d6725d71824e448c37e Mon Sep 17 00:00:00 2001 From: Glen Lee Date: Thu, 5 Nov 2015 18:51:23 +0900 Subject: [PATCH 265/335] Revert "Staging: wilc1000: coreconfigurator: Drop unneeded wrapper functions" The source and destination pointers are misplaced. This will be like, ether_addr_copy(data, bssid + ADDR2); -> ether_addr_copy(bssid, data + ADDR2); and also to use ether_addr_copy, it has to be proved that src/dst address are properly aligned(2). I revert this as author agree to drop this patch. This reverts commit d4622f68db8095dd54179e3134e97812727f6b89. Signed-off-by: Glen Lee Signed-off-by: Greg Kroah-Hartman --- drivers/staging/wilc1000/coreconfigurator.c | 48 +++++++++++---------- 1 file changed, 25 insertions(+), 23 deletions(-) diff --git a/drivers/staging/wilc1000/coreconfigurator.c b/drivers/staging/wilc1000/coreconfigurator.c index e10c6ffa698a..9568bdb6319b 100644 --- a/drivers/staging/wilc1000/coreconfigurator.c +++ b/drivers/staging/wilc1000/coreconfigurator.c @@ -13,12 +13,8 @@ #include "wilc_wlan.h" #include #include -#include #define TAG_PARAM_OFFSET (MAC_HDR_LEN + TIME_STAMP_LEN + \ BEACON_INTERVAL_LEN + CAP_INFO_LEN) -#define ADDR1 4 -#define ADDR2 10 -#define ADDR3 16 /* Basic Frame Type Codes (2-bit) */ enum basic_frame_type { @@ -175,32 +171,38 @@ static inline u8 get_from_ds(u8 *header) return ((header[1] & 0x02) >> 1); } +/* This function extracts the MAC Address in 'address1' field of the MAC */ +/* header and updates the MAC Address in the allocated 'addr' variable. */ +static inline void get_address1(u8 *pu8msa, u8 *addr) +{ + memcpy(addr, pu8msa + 4, 6); +} + +/* This function extracts the MAC Address in 'address2' field of the MAC */ +/* header and updates the MAC Address in the allocated 'addr' variable. */ +static inline void get_address2(u8 *pu8msa, u8 *addr) +{ + memcpy(addr, pu8msa + 10, 6); +} + +/* This function extracts the MAC Address in 'address3' field of the MAC */ +/* header and updates the MAC Address in the allocated 'addr' variable. */ +static inline void get_address3(u8 *pu8msa, u8 *addr) +{ + memcpy(addr, pu8msa + 16, 6); +} + /* This function extracts the BSSID from the incoming WLAN packet based on */ -/* the 'from ds' bit, and updates the MAC Address in the allocated 'data' */ +/* the 'from ds' bit, and updates the MAC Address in the allocated 'addr' */ /* variable. */ static inline void get_BSSID(u8 *data, u8 *bssid) { if (get_from_ds(data) == 1) - /* - * Extract the MAC Address in 'address2' field of the MAC - * header and update the MAC Address in the allocated 'data' - * variable. - */ - ether_addr_copy(data, bssid + ADDR2); + get_address2(data, bssid); else if (get_to_ds(data) == 1) - /* - * Extract the MAC Address in 'address1' field of the MAC - * header and update the MAC Address in the allocated 'data' - * variable. - */ - ether_addr_copy(data, bssid + ADDR1); + get_address1(data, bssid); else - /* - * Extract the MAC Address in 'address3' field of the MAC - * header and update the MAC Address in the allocated 'data' - * variable. - */ - ether_addr_copy(data, bssid + ADDR3); + get_address3(data, bssid); } /* This function extracts the SSID from a beacon/probe response frame */ From 799281a3c481a738801bf17e3079a6f91df56cd3 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 18 Nov 2015 23:29:56 +0100 Subject: [PATCH 266/335] Revert "cpufreq: intel_pstate: Avoid calculation for max/min" Revert commit 4ef451487019 (cpufreq: intel_pstate: Avoid calculation for max/min) as it depends on commit 37afb0003242 (cpufreq: intel_pstate: Use ACPI perf configuration) that causes problems to happen and needs to be reverted. Conflicts: drivers/cpufreq/intel_pstate.c Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 48 ++++------------------------------ 1 file changed, 5 insertions(+), 43 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 2e31d097def6..d444000ce148 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -163,8 +163,6 @@ struct perf_limits { int max_sysfs_pct; int min_policy_pct; int min_sysfs_pct; - int max_perf_ctl; - int min_perf_ctl; }; static struct perf_limits performance_limits = { @@ -191,8 +189,6 @@ static struct perf_limits powersave_limits = { .max_sysfs_pct = 100, .min_policy_pct = 0, .min_sysfs_pct = 0, - .max_perf_ctl = 0, - .min_perf_ctl = 0, }; #ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE @@ -938,23 +934,12 @@ static void intel_pstate_get_min_max(struct cpudata *cpu, int *min, int *max) * policy, or by cpu specific default values determined through * experimentation. */ - if (limits->max_perf_ctl && limits->max_sysfs_pct >= - limits->max_policy_pct) { - *max = limits->max_perf_ctl; - } else { - max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf), - limits->max_perf)); - *max = clamp_t(int, max_perf_adj, cpu->pstate.min_pstate, - cpu->pstate.turbo_pstate); - } + max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf), limits->max_perf)); + *max = clamp_t(int, max_perf_adj, + cpu->pstate.min_pstate, cpu->pstate.turbo_pstate); - if (limits->min_perf_ctl) { - *min = limits->min_perf_ctl; - } else { - min_perf = fp_toint(mul_fp(int_tofp(max_perf), - limits->min_perf)); - *min = clamp_t(int, min_perf, cpu->pstate.min_pstate, max_perf); - } + min_perf = fp_toint(mul_fp(int_tofp(max_perf), limits->min_perf)); + *min = clamp_t(int, min_perf, cpu->pstate.min_pstate, max_perf); } static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate, bool force) @@ -1229,12 +1214,6 @@ static unsigned int intel_pstate_get(unsigned int cpu_num) static int intel_pstate_set_policy(struct cpufreq_policy *policy) { -#if IS_ENABLED(CONFIG_ACPI) - struct cpudata *cpu; - int i; -#endif - pr_debug("intel_pstate: %s max %u policy->max %u\n", __func__, - policy->cpuinfo.max_freq, policy->max); if (!policy->cpuinfo.max_freq) return -ENODEV; @@ -1270,23 +1249,6 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) limits->max_perf = div_fp(int_tofp(limits->max_perf_pct), int_tofp(100)); -#if IS_ENABLED(CONFIG_ACPI) - cpu = all_cpu_data[policy->cpu]; - for (i = 0; i < cpu->acpi_perf_data.state_count; i++) { - int control; - - control = convert_to_native_pstate_format(cpu, i); - if (control * cpu->pstate.scaling == policy->max) - limits->max_perf_ctl = control; - if (control * cpu->pstate.scaling == policy->min) - limits->min_perf_ctl = control; - } - - pr_debug("intel_pstate: max %u policy_max %u perf_ctl [0x%x-0x%x]\n", - policy->cpuinfo.max_freq, policy->max, limits->min_perf_ctl, - limits->max_perf_ctl); -#endif - if (hwp_active) intel_pstate_hwp_set(); From 6ee11e413c495390dacabd96ad462eea9de9dfbd Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 19 Nov 2015 00:20:42 +0100 Subject: [PATCH 267/335] Revert "cpufreq: intel_pstate: Use ACPI perf configuration" Revert commit 37afb0003242 (cpufreq: intel_pstate: Use ACPI perf configuration) that is reported to cause a regression to happen on a system where invalid data are returned by the ACPI _PSS object. Since that commit makes assumptions regarding the _PSS output correctness that may turn out to be overly optimistic in general, there is a concern that it may introduce regression on more systems, so it's better to revert it now and we'll revisit the underlying issue in the next cycle with a more robust solution. Conflicts: drivers/cpufreq/intel_pstate.c Fixes: 37afb0003242 (cpufreq: intel_pstate: Use ACPI perf configuration) Reported-by: Borislav Petkov Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Kconfig.x86 | 1 - drivers/cpufreq/intel_pstate.c | 171 +-------------------------------- 2 files changed, 1 insertion(+), 171 deletions(-) diff --git a/drivers/cpufreq/Kconfig.x86 b/drivers/cpufreq/Kconfig.x86 index adbd1de1cea5..c59bdcb83217 100644 --- a/drivers/cpufreq/Kconfig.x86 +++ b/drivers/cpufreq/Kconfig.x86 @@ -5,7 +5,6 @@ config X86_INTEL_PSTATE bool "Intel P state control" depends on X86 - select ACPI_PROCESSOR if ACPI help This driver provides a P state for Intel core processors. The driver implements an internal governor and will become diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index d444000ce148..ef05944d8b16 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -34,10 +34,6 @@ #include #include -#if IS_ENABLED(CONFIG_ACPI) -#include -#endif - #define BYT_RATIOS 0x66a #define BYT_VIDS 0x66b #define BYT_TURBO_RATIOS 0x66c @@ -117,9 +113,6 @@ struct cpudata { u64 prev_mperf; u64 prev_tsc; struct sample sample; -#if IS_ENABLED(CONFIG_ACPI) - struct acpi_processor_performance acpi_perf_data; -#endif }; static struct cpudata **all_cpu_data; @@ -150,7 +143,6 @@ struct cpu_defaults { static struct pstate_adjust_policy pid_params; static struct pstate_funcs pstate_funcs; static int hwp_active; -static int no_acpi_perf; struct perf_limits { int no_turbo; @@ -197,153 +189,6 @@ static struct perf_limits *limits = &performance_limits; static struct perf_limits *limits = &powersave_limits; #endif -#if IS_ENABLED(CONFIG_ACPI) -/* - * The max target pstate ratio is a 8 bit value in both PLATFORM_INFO MSR and - * in TURBO_RATIO_LIMIT MSR, which pstate driver stores in max_pstate and - * max_turbo_pstate fields. The PERF_CTL MSR contains 16 bit value for P state - * ratio, out of it only high 8 bits are used. For example 0x1700 is setting - * target ratio 0x17. The _PSS control value stores in a format which can be - * directly written to PERF_CTL MSR. But in intel_pstate driver this shift - * occurs during write to PERF_CTL (E.g. for cores core_set_pstate()). - * This function converts the _PSS control value to intel pstate driver format - * for comparison and assignment. - */ -static int convert_to_native_pstate_format(struct cpudata *cpu, int index) -{ - return cpu->acpi_perf_data.states[index].control >> 8; -} - -static int intel_pstate_init_perf_limits(struct cpufreq_policy *policy) -{ - struct cpudata *cpu; - int ret; - bool turbo_absent = false; - int max_pstate_index; - int min_pss_ctl, max_pss_ctl, turbo_pss_ctl; - int i; - - cpu = all_cpu_data[policy->cpu]; - - pr_debug("intel_pstate: default limits 0x%x 0x%x 0x%x\n", - cpu->pstate.min_pstate, cpu->pstate.max_pstate, - cpu->pstate.turbo_pstate); - - if (!cpu->acpi_perf_data.shared_cpu_map && - zalloc_cpumask_var_node(&cpu->acpi_perf_data.shared_cpu_map, - GFP_KERNEL, cpu_to_node(policy->cpu))) { - return -ENOMEM; - } - - ret = acpi_processor_register_performance(&cpu->acpi_perf_data, - policy->cpu); - if (ret) - return ret; - - /* - * Check if the control value in _PSS is for PERF_CTL MSR, which should - * guarantee that the states returned by it map to the states in our - * list directly. - */ - if (cpu->acpi_perf_data.control_register.space_id != - ACPI_ADR_SPACE_FIXED_HARDWARE) - return -EIO; - - pr_debug("intel_pstate: CPU%u - ACPI _PSS perf data\n", policy->cpu); - for (i = 0; i < cpu->acpi_perf_data.state_count; i++) - pr_debug(" %cP%d: %u MHz, %u mW, 0x%x\n", - (i == cpu->acpi_perf_data.state ? '*' : ' '), i, - (u32) cpu->acpi_perf_data.states[i].core_frequency, - (u32) cpu->acpi_perf_data.states[i].power, - (u32) cpu->acpi_perf_data.states[i].control); - - /* - * If there is only one entry _PSS, simply ignore _PSS and continue as - * usual without taking _PSS into account - */ - if (cpu->acpi_perf_data.state_count < 2) - return 0; - - turbo_pss_ctl = convert_to_native_pstate_format(cpu, 0); - min_pss_ctl = convert_to_native_pstate_format(cpu, - cpu->acpi_perf_data.state_count - 1); - /* Check if there is a turbo freq in _PSS */ - if (turbo_pss_ctl <= cpu->pstate.max_pstate && - turbo_pss_ctl > cpu->pstate.min_pstate) { - pr_debug("intel_pstate: no turbo range exists in _PSS\n"); - limits->no_turbo = limits->turbo_disabled = 1; - cpu->pstate.turbo_pstate = cpu->pstate.max_pstate; - turbo_absent = true; - } - - /* Check if the max non turbo p state < Intel P state max */ - max_pstate_index = turbo_absent ? 0 : 1; - max_pss_ctl = convert_to_native_pstate_format(cpu, max_pstate_index); - if (max_pss_ctl < cpu->pstate.max_pstate && - max_pss_ctl > cpu->pstate.min_pstate) - cpu->pstate.max_pstate = max_pss_ctl; - - /* check If min perf > Intel P State min */ - if (min_pss_ctl > cpu->pstate.min_pstate && - min_pss_ctl < cpu->pstate.max_pstate) { - cpu->pstate.min_pstate = min_pss_ctl; - policy->cpuinfo.min_freq = min_pss_ctl * cpu->pstate.scaling; - } - - if (turbo_absent) - policy->cpuinfo.max_freq = cpu->pstate.max_pstate * - cpu->pstate.scaling; - else { - policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate * - cpu->pstate.scaling; - /* - * The _PSS table doesn't contain whole turbo frequency range. - * This just contains +1 MHZ above the max non turbo frequency, - * with control value corresponding to max turbo ratio. But - * when cpufreq set policy is called, it will call with this - * max frequency, which will cause a reduced performance as - * this driver uses real max turbo frequency as the max - * frequeny. So correct this frequency in _PSS table to - * correct max turbo frequency based on the turbo ratio. - * Also need to convert to MHz as _PSS freq is in MHz. - */ - cpu->acpi_perf_data.states[0].core_frequency = - turbo_pss_ctl * 100; - } - - pr_debug("intel_pstate: Updated limits using _PSS 0x%x 0x%x 0x%x\n", - cpu->pstate.min_pstate, cpu->pstate.max_pstate, - cpu->pstate.turbo_pstate); - pr_debug("intel_pstate: policy max_freq=%d Khz min_freq = %d KHz\n", - policy->cpuinfo.max_freq, policy->cpuinfo.min_freq); - - return 0; -} - -static int intel_pstate_exit_perf_limits(struct cpufreq_policy *policy) -{ - struct cpudata *cpu; - - if (!no_acpi_perf) - return 0; - - cpu = all_cpu_data[policy->cpu]; - acpi_processor_unregister_performance(policy->cpu); - return 0; -} - -#else -static int intel_pstate_init_perf_limits(struct cpufreq_policy *policy) -{ - return 0; -} - -static int intel_pstate_exit_perf_limits(struct cpufreq_policy *policy) -{ - return 0; -} -#endif - static inline void pid_reset(struct _pid *pid, int setpoint, int busy, int deadband, int integral) { pid->setpoint = setpoint; @@ -1303,30 +1148,18 @@ static int intel_pstate_cpu_init(struct cpufreq_policy *policy) policy->cpuinfo.min_freq = cpu->pstate.min_pstate * cpu->pstate.scaling; policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate * cpu->pstate.scaling; - if (!no_acpi_perf) - intel_pstate_init_perf_limits(policy); - /* - * If there is no acpi perf data or error, we ignore and use Intel P - * state calculated limits, So this is not fatal error. - */ policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL; cpumask_set_cpu(policy->cpu, policy->cpus); return 0; } -static int intel_pstate_cpu_exit(struct cpufreq_policy *policy) -{ - return intel_pstate_exit_perf_limits(policy); -} - static struct cpufreq_driver intel_pstate_driver = { .flags = CPUFREQ_CONST_LOOPS, .verify = intel_pstate_verify_policy, .setpolicy = intel_pstate_set_policy, .get = intel_pstate_get, .init = intel_pstate_cpu_init, - .exit = intel_pstate_cpu_exit, .stop_cpu = intel_pstate_stop_cpu, .name = "intel_pstate", }; @@ -1368,6 +1201,7 @@ static void copy_cpu_funcs(struct pstate_funcs *funcs) } #if IS_ENABLED(CONFIG_ACPI) +#include static bool intel_pstate_no_acpi_pss(void) { @@ -1563,9 +1397,6 @@ static int __init intel_pstate_setup(char *str) force_load = 1; if (!strcmp(str, "hwp_only")) hwp_only = 1; - if (!strcmp(str, "no_acpi")) - no_acpi_perf = 1; - return 0; } early_param("intel_pstate", intel_pstate_setup); From 938d21a2a6370241c86d515ca574aaaa9e8812f2 Mon Sep 17 00:00:00 2001 From: Philippe Longepe Date: Mon, 9 Nov 2015 17:40:46 -0800 Subject: [PATCH 268/335] cpufreq: intel_pstate: Replace BYT with ATOM Rename symbol and function names starting with "BYT" or "byt" to start with "ATOM" or "atom", respectively, so as to make it clear that they may apply to Atom in general and not just to Baytrail (the goal is to support several Atoms architectures eventually). This should not lead to any functional changes. Signed-off-by: Philippe Longepe Signed-off-by: Stephane Gasparini Acked-by: Srinivas Pandruvada [ rjw : Changelog ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 58 +++++++++++++++++----------------- 1 file changed, 29 insertions(+), 29 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index ef05944d8b16..5f124e98f439 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -34,10 +34,10 @@ #include #include -#define BYT_RATIOS 0x66a -#define BYT_VIDS 0x66b -#define BYT_TURBO_RATIOS 0x66c -#define BYT_TURBO_VIDS 0x66d +#define ATOM_RATIOS 0x66a +#define ATOM_VIDS 0x66b +#define ATOM_TURBO_RATIOS 0x66c +#define ATOM_TURBO_VIDS 0x66d #define FRAC_BITS 8 #define int_tofp(X) ((int64_t)(X) << FRAC_BITS) @@ -528,31 +528,31 @@ static void intel_pstate_hwp_enable(struct cpudata *cpudata) wrmsrl_on_cpu(cpudata->cpu, MSR_PM_ENABLE, 0x1); } -static int byt_get_min_pstate(void) +static int atom_get_min_pstate(void) { u64 value; - rdmsrl(BYT_RATIOS, value); + rdmsrl(ATOM_RATIOS, value); return (value >> 8) & 0x7F; } -static int byt_get_max_pstate(void) +static int atom_get_max_pstate(void) { u64 value; - rdmsrl(BYT_RATIOS, value); + rdmsrl(ATOM_RATIOS, value); return (value >> 16) & 0x7F; } -static int byt_get_turbo_pstate(void) +static int atom_get_turbo_pstate(void) { u64 value; - rdmsrl(BYT_TURBO_RATIOS, value); + rdmsrl(ATOM_TURBO_RATIOS, value); return value & 0x7F; } -static void byt_set_pstate(struct cpudata *cpudata, int pstate) +static void atom_set_pstate(struct cpudata *cpudata, int pstate) { u64 val; int32_t vid_fp; @@ -577,10 +577,10 @@ static void byt_set_pstate(struct cpudata *cpudata, int pstate) wrmsrl_on_cpu(cpudata->cpu, MSR_IA32_PERF_CTL, val); } -#define BYT_BCLK_FREQS 5 -static int byt_freq_table[BYT_BCLK_FREQS] = { 833, 1000, 1333, 1167, 800}; +#define ATOM_BCLK_FREQS 5 +static int atom_freq_table[ATOM_BCLK_FREQS] = { 833, 1000, 1333, 1167, 800}; -static int byt_get_scaling(void) +static int atom_get_scaling(void) { u64 value; int i; @@ -588,16 +588,16 @@ static int byt_get_scaling(void) rdmsrl(MSR_FSB_FREQ, value); i = value & 0x3; - BUG_ON(i > BYT_BCLK_FREQS); + WARN_ON(i > ATOM_BCLK_FREQS); - return byt_freq_table[i] * 100; + return atom_freq_table[i] * 100; } -static void byt_get_vid(struct cpudata *cpudata) +static void atom_get_vid(struct cpudata *cpudata) { u64 value; - rdmsrl(BYT_VIDS, value); + rdmsrl(ATOM_VIDS, value); cpudata->vid.min = int_tofp((value >> 8) & 0x7f); cpudata->vid.max = int_tofp((value >> 16) & 0x7f); cpudata->vid.ratio = div_fp( @@ -605,7 +605,7 @@ static void byt_get_vid(struct cpudata *cpudata) int_tofp(cpudata->pstate.max_pstate - cpudata->pstate.min_pstate)); - rdmsrl(BYT_TURBO_VIDS, value); + rdmsrl(ATOM_TURBO_VIDS, value); cpudata->vid.turbo = value & 0x7f; } @@ -726,7 +726,7 @@ static struct cpu_defaults core_params = { }, }; -static struct cpu_defaults byt_params = { +static struct cpu_defaults atom_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, @@ -736,13 +736,13 @@ static struct cpu_defaults byt_params = { .i_gain_pct = 4, }, .funcs = { - .get_max = byt_get_max_pstate, - .get_max_physical = byt_get_max_pstate, - .get_min = byt_get_min_pstate, - .get_turbo = byt_get_turbo_pstate, - .set = byt_set_pstate, - .get_scaling = byt_get_scaling, - .get_vid = byt_get_vid, + .get_max = atom_get_max_pstate, + .get_max_physical = atom_get_max_pstate, + .get_min = atom_get_min_pstate, + .get_turbo = atom_get_turbo_pstate, + .set = atom_set_pstate, + .get_scaling = atom_get_scaling, + .get_vid = atom_get_vid, }, }; @@ -983,7 +983,7 @@ static void intel_pstate_timer_func(unsigned long __data) static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(0x2a, core_params), ICPU(0x2d, core_params), - ICPU(0x37, byt_params), + ICPU(0x37, atom_params), ICPU(0x3a, core_params), ICPU(0x3c, core_params), ICPU(0x3d, core_params), @@ -992,7 +992,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(0x45, core_params), ICPU(0x46, core_params), ICPU(0x47, core_params), - ICPU(0x4c, byt_params), + ICPU(0x4c, atom_params), ICPU(0x4e, core_params), ICPU(0x4f, core_params), ICPU(0x5e, core_params), From 1421df63c3cf956c69c26ac9660c6e223eeed980 Mon Sep 17 00:00:00 2001 From: Philippe Longepe Date: Mon, 9 Nov 2015 17:40:47 -0800 Subject: [PATCH 269/335] cpufreq: intel_pstate: Add separate support for Airmont cores There are two flavors of Atom cores to be supported by intel_pstate, Silvermont and Airmont, so make the driver distinguish between them by adding separate frequency tables. Separate the CPU defaults params for each of them and match the CPU IDs against them as appropriate. Signed-off-by: Philippe Longepe Signed-off-by: Stephane Gasparini Acked-by: Srinivas Pandruvada [ rjw: Subject and changelog ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 57 +++++++++++++++++++++++++++------- 1 file changed, 46 insertions(+), 11 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 5f124e98f439..001a532e342e 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -577,20 +577,35 @@ static void atom_set_pstate(struct cpudata *cpudata, int pstate) wrmsrl_on_cpu(cpudata->cpu, MSR_IA32_PERF_CTL, val); } -#define ATOM_BCLK_FREQS 5 -static int atom_freq_table[ATOM_BCLK_FREQS] = { 833, 1000, 1333, 1167, 800}; - -static int atom_get_scaling(void) +static int silvermont_get_scaling(void) { u64 value; int i; + /* Defined in Table 35-6 from SDM (Sept 2015) */ + static int silvermont_freq_table[] = { + 83300, 100000, 133300, 116700, 80000}; rdmsrl(MSR_FSB_FREQ, value); - i = value & 0x3; + i = value & 0x7; + WARN_ON(i > 4); - WARN_ON(i > ATOM_BCLK_FREQS); + return silvermont_freq_table[i]; +} - return atom_freq_table[i] * 100; +static int airmont_get_scaling(void) +{ + u64 value; + int i; + /* Defined in Table 35-10 from SDM (Sept 2015) */ + static int airmont_freq_table[] = { + 83300, 100000, 133300, 116700, 80000, + 93300, 90000, 88900, 87500}; + + rdmsrl(MSR_FSB_FREQ, value); + i = value & 0xF; + WARN_ON(i > 8); + + return airmont_freq_table[i]; } static void atom_get_vid(struct cpudata *cpudata) @@ -726,7 +741,7 @@ static struct cpu_defaults core_params = { }, }; -static struct cpu_defaults atom_params = { +static struct cpu_defaults silvermont_params = { .pid_policy = { .sample_rate_ms = 10, .deadband = 0, @@ -741,7 +756,27 @@ static struct cpu_defaults atom_params = { .get_min = atom_get_min_pstate, .get_turbo = atom_get_turbo_pstate, .set = atom_set_pstate, - .get_scaling = atom_get_scaling, + .get_scaling = silvermont_get_scaling, + .get_vid = atom_get_vid, + }, +}; + +static struct cpu_defaults airmont_params = { + .pid_policy = { + .sample_rate_ms = 10, + .deadband = 0, + .setpoint = 60, + .p_gain_pct = 14, + .d_gain_pct = 0, + .i_gain_pct = 4, + }, + .funcs = { + .get_max = atom_get_max_pstate, + .get_max_physical = atom_get_max_pstate, + .get_min = atom_get_min_pstate, + .get_turbo = atom_get_turbo_pstate, + .set = atom_set_pstate, + .get_scaling = airmont_get_scaling, .get_vid = atom_get_vid, }, }; @@ -983,7 +1018,7 @@ static void intel_pstate_timer_func(unsigned long __data) static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(0x2a, core_params), ICPU(0x2d, core_params), - ICPU(0x37, atom_params), + ICPU(0x37, silvermont_params), ICPU(0x3a, core_params), ICPU(0x3c, core_params), ICPU(0x3d, core_params), @@ -992,7 +1027,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(0x45, core_params), ICPU(0x46, core_params), ICPU(0x47, core_params), - ICPU(0x4c, atom_params), + ICPU(0x4c, airmont_params), ICPU(0x4e, core_params), ICPU(0x4f, core_params), ICPU(0x5e, core_params), From 2d4ee3036774e394d416cded9d5cf7661ffb4e4f Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 22:27:55 +0100 Subject: [PATCH 270/335] cpufreq: mediatek: fix build error The recently added mt8173 cpufreq driver relies on the cpu topology that is always present on ARM64 but optional on ARM32: drivers/cpufreq/mt8173-cpufreq.c: In function 'mtk_cpufreq_init': drivers/cpufreq/mt8173-cpufreq.c:441:30: error: 'cpu_topology' undeclared (first use in this function) cpumask_copy(policy->cpus, &cpu_topology[policy->cpu].core_sibling); This refines the Kconfig dependencies so that we can still build on ARM32, but only if COMPILE_TEST is selected and the CPU topology code is present. Signed-off-by: Arnd Bergmann Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Kconfig.arm | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index cd0391e46c6d..c97d9ea97b48 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -84,6 +84,7 @@ config ARM_KIRKWOOD_CPUFREQ config ARM_MT8173_CPUFREQ bool "Mediatek MT8173 CPUFreq support" depends on ARCH_MEDIATEK && REGULATOR + depends on ARM64 || (ARM_CPU_TOPOLOGY && COMPILE_TEST) depends on !CPU_THERMAL || THERMAL=y select PM_OPP help From 3bf7f56e70c9f0edb4f91cbb73a393c3b1d24801 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 19 Nov 2015 00:42:23 +0100 Subject: [PATCH 271/335] Revert "Documentation: kernel_parameters for Intel P state driver" Revert commit 053f56def57b (Documentation: kernel_parameters for Intel P state driver) as the code documented by it has been reverted already. Signed-off-by: Rafael J. Wysocki --- Documentation/kernel-parameters.txt | 3 --- 1 file changed, 3 deletions(-) diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 9b75e2a760de..22a4b687ea5b 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1546,9 +1546,6 @@ bytes respectively. Such letter suffixes can also be entirely omitted. hwp_only Only load intel_pstate on systems which support hardware P state control (HWP) if available. - no_acpi - Don't use ACPI processor performance control objects - _PSS and _PPC specified limits. intremap= [X86-64, Intel-IOMMU] on enable Interrupt Remapping (default) From 0df9d41ab5d43dc5b20abc8b22a6b6d098b03994 Mon Sep 17 00:00:00 2001 From: Yigal Korman Date: Mon, 16 Nov 2015 14:09:15 +0200 Subject: [PATCH 272/335] mm, dax: fix DAX deadlocks (COW fault) DAX handling of COW faults has wrong locking sequence: dax_fault does i_mmap_lock_read do_cow_fault does i_mmap_unlock_write Ross's commit[1] missed a fix[2] that Kirill added to Matthew's commit[3]. Original COW locking logic was introduced by Matthew here[4]. This should be applied to v4.3 as well. [1] 0f90cc6609c7 mm, dax: fix DAX deadlocks [2] 52a2b53ffde6 mm, dax: use i_mmap_unlock_write() in do_cow_fault() [3] 843172978bb9 dax: fix race between simultaneous faults [4] 2e4cdab0584f mm: allow page fault handlers to perform the COW Cc: Cc: Boaz Harrosh Cc: Alexander Viro Cc: Dave Chinner Cc: Jan Kara Cc: "Kirill A. Shutemov" Cc: Matthew Wilcox Acked-by: Ross Zwisler Signed-off-by: Yigal Korman Signed-off-by: Dan Williams --- mm/memory.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/mm/memory.c b/mm/memory.c index deb679c31f2a..c387430f06c3 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3015,9 +3015,9 @@ static int do_cow_fault(struct mm_struct *mm, struct vm_area_struct *vma, } else { /* * The fault handler has no page to lock, so it holds - * i_mmap_lock for write to protect against truncate. + * i_mmap_lock for read to protect against truncate. */ - i_mmap_unlock_write(vma->vm_file->f_mapping); + i_mmap_unlock_read(vma->vm_file->f_mapping); } goto uncharge_out; } @@ -3031,9 +3031,9 @@ static int do_cow_fault(struct mm_struct *mm, struct vm_area_struct *vma, } else { /* * The fault handler has no page to lock, so it holds - * i_mmap_lock for write to protect against truncate. + * i_mmap_lock for read to protect against truncate. */ - i_mmap_unlock_write(vma->vm_file->f_mapping); + i_mmap_unlock_read(vma->vm_file->f_mapping); } return ret; uncharge_out: From f6619ef7508261be2ba3ded313ccc46ce670d0d3 Mon Sep 17 00:00:00 2001 From: "Wang, Rui Y" Date: Wed, 18 Nov 2015 23:00:53 +0800 Subject: [PATCH 273/335] drm/mgag200: fix kernel hang in cursor code. The machine hang completely with the following message on the console: [ 487.777538] BUG: unable to handle kernel NULL pointer dereference at 0000000000000060 [ 487.777554] IP: [] _raw_spin_lock+0xe/0x30 [ 487.777557] PGD 42e9f7067 PUD 42f2fa067 PMD 0 [ 487.777560] Oops: 0002 [#1] SMP ... [ 487.777618] CPU: 21 PID: 3190 Comm: Xorg Tainted: G E 4.4.0-rc1-3-default+ #6 [ 487.777620] Hardware name: Intel Corporation BRICKLAND/BRICKLAND, BIOS BRHSXSD1.86B.0059.R00.1501081238 01/08/2015 [ 487.777621] task: ffff880853ae4680 ti: ffff8808696d4000 task.ti: ffff8808696d4000 [ 487.777625] RIP: 0010:[] [] _raw_spin_lock+0xe/0x30 [ 487.777627] RSP: 0018:ffff8808696d79c0 EFLAGS: 00010246 [ 487.777628] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000 [ 487.777629] RDX: 0000000000000001 RSI: 0000000000000000 RDI: 0000000000000060 [ 487.777630] RBP: ffff8808696d79e0 R08: 0000000000000000 R09: ffff88086924a780 [ 487.777631] R10: 000000000001bb40 R11: 0000000000003246 R12: 0000000000000000 [ 487.777632] R13: ffff880463a27360 R14: ffff88046ca50218 R15: 0000000000000080 [ 487.777634] FS: 00007f3f81c5a8c0(0000) GS:ffff88086f060000(0000) knlGS:0000000000000000 [ 487.777635] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 487.777636] CR2: 0000000000000060 CR3: 000000042e678000 CR4: 00000000001406e0 [ 487.777638] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 487.777639] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 487.777639] Stack: [ 487.777642] ffffffffa00eb5fa ffff8808696d7b60 ffff88086b87d800 0000000000000000 [ 487.777644] ffff8808696d7ac8 ffffffffa01694b6 ffff8808696d7ae8 ffffffff8109c8d5 [ 487.777647] ffff880469158740 ffff880463a27000 ffff88086b87d800 ffff88086b87d800 [ 487.777647] Call Trace: [ 487.777674] [] ? drm_gem_object_lookup+0x1a/0xa0 [drm] [ 487.777681] [] mga_crtc_cursor_set+0xc6/0xb60 [mgag200] [ 487.777691] [] ? find_busiest_group+0x35/0x4a0 [ 487.777696] [] ? __might_sleep+0x44/0x80 [ 487.777699] [] ? __ww_mutex_lock+0x22/0x9c [ 487.777722] [] ? drm_modeset_lock+0x34/0xf0 [drm] [ 487.777733] [] restore_fbdev_mode+0xee/0x2a0 [drm_kms_helper] [ 487.777742] [] drm_fb_helper_restore_fbdev_mode_unlocked+0x2e/0x70 [drm_kms_helper] [ 487.777748] [] drm_fb_helper_set_par+0x27/0x50 [drm_kms_helper] [ 487.777752] [] fb_set_var+0x18c/0x3f0 [ 487.777777] [] ? __ext4_handle_dirty_metadata+0x8a/0x210 [ext4] [ 487.777783] [] fbcon_blank+0x1b7/0x2b0 [ 487.777790] [] do_unblank_screen+0xb3/0x1c0 [ 487.777795] [] vt_ioctl+0x118a/0x1210 [ 487.777801] [] tty_ioctl+0x3f0/0xc90 [ 487.777808] [] ? kzfree+0x28/0x30 [ 487.777813] [] ? mntput+0x1f/0x30 [ 487.777817] [] do_vfs_ioctl+0x30d/0x570 [ 487.777822] [] ? task_work_run+0x8a/0xa0 [ 487.777825] [] SyS_ioctl+0x74/0x80 [ 487.777829] [] entry_SYSCALL_64_fastpath+0x12/0x71 [ 487.777851] Code: 65 ff 0d ce 02 a8 7e 5d c3 ba 01 00 00 00 f0 0f b1 17 85 c0 75 e8 b0 01 5d c3 0f 1f 00 65 ff 05 b1 02 a8 7e 31 c0 ba 01 00 00 00 0f b1 17 85 c0 75 01 c3 55 89 c6 48 89 e5 e8 4e f5 b1 ff 5d [ 487.777854] RIP [] _raw_spin_lock+0xe/0x30 [ 487.777855] RSP [ 487.777856] CR2: 0000000000000060 [ 487.777860] ---[ end trace 672a2cd555e0ebd3 ]--- The cursor code may be entered with file_priv == NULL && handle == NULL. The problem was introduced by: "bf89209 drm/mga200g: Hold a proper reference for cursor_set" which calls drm_gem_object_lookup(dev, file_priv...). Previously this wasn't a problem because we checked the handle. Move the check early in the function can fix the problem. Signed-off-by: Rui Wang Reviewed-by: Daniel Vetter Signed-off-by: Dave Airlie --- drivers/gpu/drm/mgag200/mgag200_cursor.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/mgag200/mgag200_cursor.c b/drivers/gpu/drm/mgag200/mgag200_cursor.c index 4f2068fe5d88..a7bf6a90eae5 100644 --- a/drivers/gpu/drm/mgag200/mgag200_cursor.c +++ b/drivers/gpu/drm/mgag200/mgag200_cursor.c @@ -70,6 +70,11 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc, BUG_ON(pixels_2 != pixels_current && pixels_2 != pixels_prev); BUG_ON(pixels_current == pixels_prev); + if (!handle || !file_priv) { + mga_hide_cursor(mdev); + return 0; + } + obj = drm_gem_object_lookup(dev, file_priv, handle); if (!obj) return -ENOENT; @@ -88,12 +93,6 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc, goto out_unreserve1; } - if (!handle) { - mga_hide_cursor(mdev); - ret = 0; - goto out1; - } - /* Move cursor buffers into VRAM if they aren't already */ if (!pixels_1->pin_count) { ret = mgag200_bo_pin(pixels_1, TTM_PL_FLAG_VRAM, From 7383123647566a813692bb37a1389c767ed89158 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 19 Nov 2015 10:26:30 +0200 Subject: [PATCH 274/335] Revert "drm/i915: skip modeset if compatible for everyone." This reverts commit 6764e9f8724f1231b4deac53b9a82286ac0830e7 Author: Maarten Lankhorst Date: Thu Aug 27 15:44:06 2015 +0200 drm/i915: skip modeset if compatible for everyone. Bring back the i915.fastboot module parameter, disabled by default, due to backlight regression on Chromebook Pixel 2015. Apparently the firmware of the Chromebook in question enables the panel but disables backlight to avoid a brief garbage scanout upon loading the kernel/module. With fastboot, we leave the backlight untouched, in this case disabled. The user would have to do a modeset (i.e. not just crank up the brightness) to enable the backlight. There is no clean fix readily available, so get back to the drawing board by reverting. [N.B. The reference below is for when the thread was included on public lists, and some of the context had already been dropped by then.] Reported-and-tested-by: Olof Johansson Acked-by: Maarten Lankhorst Acked-by: Daniel Vetter References: http://marc.info/?i=CAKMK7uES7xk05ki92oeX6gmvZWAh9f2vL7yz=6T+fGK9J3X7cQ@mail.gmail.com Fixes: 6764e9f8724f ("drm/i915: skip modeset if compatible for everyone.") Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447921590-3785-1-git-send-email-jani.nikula@intel.com --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/i915_params.c | 5 +++++ drivers/gpu/drm/i915/intel_display.c | 3 ++- 3 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 84b5db257f1b..95bb27de774f 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2651,6 +2651,7 @@ struct i915_params { int enable_cmd_parser; /* leave bools at the end to not create holes */ bool enable_hangcheck; + bool fastboot; bool prefault_disable; bool load_detect_test; bool reset; diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index 96bb23865eac..4be13a5eb932 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -40,6 +40,7 @@ struct i915_params i915 __read_mostly = { .preliminary_hw_support = IS_ENABLED(CONFIG_DRM_I915_PRELIMINARY_HW_SUPPORT), .disable_power_well = -1, .enable_ips = 1, + .fastboot = 0, .prefault_disable = 0, .load_detect_test = 0, .reset = true, @@ -133,6 +134,10 @@ MODULE_PARM_DESC(disable_power_well, module_param_named_unsafe(enable_ips, i915.enable_ips, int, 0600); MODULE_PARM_DESC(enable_ips, "Enable IPS (default: true)"); +module_param_named(fastboot, i915.fastboot, bool, 0600); +MODULE_PARM_DESC(fastboot, + "Try to skip unnecessary mode sets at boot time (default: false)"); + module_param_named_unsafe(prefault_disable, i915.prefault_disable, bool, 0600); MODULE_PARM_DESC(prefault_disable, "Disable page prefaulting for pread/pwrite/reloc (default:false). " diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 002e2a41ce1b..71860f8680f9 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -13068,7 +13068,8 @@ static int intel_atomic_check(struct drm_device *dev, if (ret) return ret; - if (intel_pipe_config_compare(state->dev, + if (i915.fastboot && + intel_pipe_config_compare(state->dev, to_intel_crtc_state(crtc->state), pipe_config, true)) { crtc_state->mode_changed = false; From 112677d683d31ebd6a8e8b02e0620ae512354b2d Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Tue, 17 Nov 2015 09:43:24 +0900 Subject: [PATCH 275/335] x86/ftrace: Add comment on static function tracing There was a confusion between update_ftrace_function() and static function tracing trampoline regarding 3rd parameter (ftrace_ops). Add a comment for clarification. Suggested-by: Steven Rostedt Signed-off-by: Namhyung Kim Cc: H. Peter Anvin Cc: Steven Rostedt Link: http://lkml.kernel.org/r/1447721004-2551-1-git-send-email-namhyung@kernel.org Signed-off-by: Thomas Gleixner --- arch/x86/kernel/mcount_64.S | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/x86/kernel/mcount_64.S b/arch/x86/kernel/mcount_64.S index 94ea120fa21f..87e1762e2bca 100644 --- a/arch/x86/kernel/mcount_64.S +++ b/arch/x86/kernel/mcount_64.S @@ -278,6 +278,12 @@ trace: /* save_mcount_regs fills in first two parameters */ save_mcount_regs + /* + * When DYNAMIC_FTRACE is not defined, ARCH_SUPPORTS_FTRACE_OPS is not + * set (see include/asm/ftrace.h and include/linux/ftrace.h). Only the + * ip and parent ip are used and the list function is called when + * function tracing is enabled. + */ call *ftrace_trace_function restore_mcount_regs From 581b7f158fe0383b492acd1ce3fb4e99d4e57808 Mon Sep 17 00:00:00 2001 From: Andrew Cooper Date: Wed, 3 Jun 2015 10:31:14 +0100 Subject: [PATCH 276/335] x86/cpu: Fix SMAP check in PVOPS environments There appears to be no formal statement of what pv_irq_ops.save_fl() is supposed to return precisely. Native returns the full flags, while lguest and Xen only return the Interrupt Flag, and both have comments by the implementations stating that only the Interrupt Flag is looked at. This may have been true when initially implemented, but no longer is. To make matters worse, the Xen PVOP leaves the upper bits undefined, making the BUG_ON() undefined behaviour. Experimentally, this now trips for 32bit PV guests on Broadwell hardware. The BUG_ON() is consistent for an individual build, but not consistent for all builds. It has also been a sitting timebomb since SMAP support was introduced. Use native_save_fl() instead, which will obtain an accurate view of the AC flag. Signed-off-by: Andrew Cooper Reviewed-by: David Vrabel Tested-by: Rusty Russell Cc: Rusty Russell Cc: Konrad Rzeszutek Wilk Cc: Boris Ostrovsky Cc: Cc: Xen-devel CC: stable@vger.kernel.org Link: http://lkml.kernel.org/r/1433323874-6927-1-git-send-email-andrew.cooper3@citrix.com Signed-off-by: Thomas Gleixner --- arch/x86/kernel/cpu/common.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 4ddd780aeac9..c2b7522cbf35 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -273,10 +273,9 @@ __setup("nosmap", setup_disable_smap); static __always_inline void setup_smap(struct cpuinfo_x86 *c) { - unsigned long eflags; + unsigned long eflags = native_save_fl(); /* This should have been cleared long ago */ - raw_local_save_flags(eflags); BUG_ON(eflags & X86_EFLAGS_AC); if (cpu_has(c, X86_FEATURE_SMAP)) { From 5481c8fb1da2a573861095fbea078f30c664d0bd Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Wed, 18 Nov 2015 18:46:48 +0100 Subject: [PATCH 277/335] drm/atomic-helper: Check encoder/crtc constraints MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This was totally lost when I originally created the atomic helpers. We probably should also check possible_clones in the helpers, but since the legacy ones didn't do that this is for a separate patch. Reported-by: Ville Syrjälä Cc: Ville Syrjälä Cc: Daniel Stone Reviewed-by: Daniel Stone Reviewed-by: Ville Syrjälä Tested-by: Ville Syrjälä Signed-off-by: Daniel Vetter Signed-off-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1447868808-10266-1-git-send-email-daniel.vetter@ffwll.ch --- drivers/gpu/drm/drm_atomic_helper.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index cd398a245d2e..e5aec45bf985 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -210,6 +210,14 @@ update_connector_routing(struct drm_atomic_state *state, int conn_idx) return -EINVAL; } + if (!drm_encoder_crtc_ok(new_encoder, connector_state->crtc)) { + DRM_DEBUG_ATOMIC("[ENCODER:%d:%s] incompatible with [CRTC:%d]\n", + new_encoder->base.id, + new_encoder->name, + connector_state->crtc->base.id); + return -EINVAL; + } + if (new_encoder == connector_state->best_encoder) { DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] keeps [ENCODER:%d:%s], now on [CRTC:%d]\n", connector->base.id, From 92e788b749862ebe9920360513a718e5dd4da7a9 Mon Sep 17 00:00:00 2001 From: Yang Shi Date: Wed, 18 Nov 2015 10:48:55 -0800 Subject: [PATCH 278/335] arm64: restore bogomips information in /proc/cpuinfo As previously reported, some userspace applications depend on bogomips showed by /proc/cpuinfo. Although there is much less legacy impact on aarch64 than arm, it does break libvirt. This patch reverts commit 326b16db9f69 ("arm64: delay: don't bother reporting bogomips in /proc/cpuinfo"), but with some tweak due to context change and without the pr_info(). Fixes: 326b16db9f69 ("arm64: delay: don't bother reporting bogomips in /proc/cpuinfo") Signed-off-by: Yang Shi Acked-by: Will Deacon Cc: # 3.12+ Signed-off-by: Catalin Marinas --- arch/arm64/kernel/cpuinfo.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/arm64/kernel/cpuinfo.c b/arch/arm64/kernel/cpuinfo.c index 706679d0a0b4..212ae6361d8b 100644 --- a/arch/arm64/kernel/cpuinfo.c +++ b/arch/arm64/kernel/cpuinfo.c @@ -30,6 +30,7 @@ #include #include #include +#include /* * In case the boot CPU is hotpluggable, we record its initial state and @@ -112,6 +113,10 @@ static int c_show(struct seq_file *m, void *v) */ seq_printf(m, "processor\t: %d\n", i); + seq_printf(m, "BogoMIPS\t: %lu.%02lu\n", + loops_per_jiffy / (500000UL/HZ), + loops_per_jiffy / (5000UL/HZ) % 100); + /* * Dump out the common processor features in a single line. * Userspace should read the hwcaps with getauxval(AT_HWCAP) From 2e6edc95382cc36423aff18a237173ad62d5ab52 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Thu, 19 Nov 2015 13:29:28 -0800 Subject: [PATCH 279/335] block: protect rw_page against device teardown Fix use after free crashes like the following: general protection fault: 0000 [#1] SMP Call Trace: [] ? pmem_do_bvec.isra.12+0xa6/0xf0 [nd_pmem] [] pmem_rw_page+0x42/0x80 [nd_pmem] [] bdev_read_page+0x50/0x60 [] do_mpage_readpage+0x510/0x770 [] ? I_BDEV+0x20/0x20 [] ? lru_cache_add+0x1c/0x50 [] mpage_readpages+0x107/0x170 [] ? I_BDEV+0x20/0x20 [] ? I_BDEV+0x20/0x20 [] blkdev_readpages+0x1d/0x20 [] __do_page_cache_readahead+0x28f/0x310 [] ? __do_page_cache_readahead+0x169/0x310 [] ? pagecache_get_page+0x2d/0x1d0 [] filemap_fault+0x396/0x530 [] __do_fault+0x4e/0xf0 [] handle_mm_fault+0x11bd/0x1b50 Cc: Cc: Jens Axboe Cc: Alexander Viro Reported-by: kbuild test robot Acked-by: Matthew Wilcox [willy: symmetry fixups] Signed-off-by: Dan Williams --- block/blk.h | 2 -- fs/block_dev.c | 18 ++++++++++++++++-- include/linux/blkdev.h | 2 ++ 3 files changed, 18 insertions(+), 4 deletions(-) diff --git a/block/blk.h b/block/blk.h index da722eb786df..c43926d3d74d 100644 --- a/block/blk.h +++ b/block/blk.h @@ -72,8 +72,6 @@ void blk_dequeue_request(struct request *rq); void __blk_queue_free_tags(struct request_queue *q); bool __blk_end_bidi_request(struct request *rq, int error, unsigned int nr_bytes, unsigned int bidi_bytes); -int blk_queue_enter(struct request_queue *q, gfp_t gfp); -void blk_queue_exit(struct request_queue *q); void blk_freeze_queue(struct request_queue *q); static inline void blk_queue_enter_live(struct request_queue *q) diff --git a/fs/block_dev.c b/fs/block_dev.c index bb0dfb1c7af1..c25639e907bd 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -390,9 +390,17 @@ int bdev_read_page(struct block_device *bdev, sector_t sector, struct page *page) { const struct block_device_operations *ops = bdev->bd_disk->fops; + int result = -EOPNOTSUPP; + if (!ops->rw_page || bdev_get_integrity(bdev)) - return -EOPNOTSUPP; - return ops->rw_page(bdev, sector + get_start_sect(bdev), page, READ); + return result; + + result = blk_queue_enter(bdev->bd_queue, GFP_KERNEL); + if (result) + return result; + result = ops->rw_page(bdev, sector + get_start_sect(bdev), page, READ); + blk_queue_exit(bdev->bd_queue); + return result; } EXPORT_SYMBOL_GPL(bdev_read_page); @@ -421,14 +429,20 @@ int bdev_write_page(struct block_device *bdev, sector_t sector, int result; int rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE; const struct block_device_operations *ops = bdev->bd_disk->fops; + if (!ops->rw_page || bdev_get_integrity(bdev)) return -EOPNOTSUPP; + result = blk_queue_enter(bdev->bd_queue, GFP_KERNEL); + if (result) + return result; + set_page_writeback(page); result = ops->rw_page(bdev, sector + get_start_sect(bdev), page, rw); if (result) end_page_writeback(page); else unlock_page(page); + blk_queue_exit(bdev->bd_queue); return result; } EXPORT_SYMBOL_GPL(bdev_write_page); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 3fe27f8d91f0..c0d2b7927c1f 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -794,6 +794,8 @@ extern int scsi_cmd_ioctl(struct request_queue *, struct gendisk *, fmode_t, extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t, struct scsi_ioctl_command __user *); +extern int blk_queue_enter(struct request_queue *q, gfp_t gfp); +extern void blk_queue_exit(struct request_queue *q); extern void blk_start_queue(struct request_queue *q); extern void blk_stop_queue(struct request_queue *q); extern void blk_sync_queue(struct request_queue *q); From 768acd64d68b232e0d2b9623d9846457355f0c27 Mon Sep 17 00:00:00 2001 From: Suravee Suthikulpanit Date: Wed, 18 Nov 2015 16:49:52 -0800 Subject: [PATCH 280/335] PCI: Fix OF logic in pci_dma_configure() This patch fixes a bug introduced by previous commit, which incorrectly checkes the of_node of the end-point device. Instead, it should check the of_node of the host bridge. Fixes: 50230713b639 ("PCI: OF: Move of_pci_dma_configure() to pci_dma_configure()") Reported-by: Robin Murphy Signed-off-by: Suravee Suthikulpanit Acked-by: Arnd Bergmann Signed-off-by: Rafael J. Wysocki --- drivers/pci/probe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index 064078e11017..dc8fbe5c09d3 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -1646,8 +1646,8 @@ static void pci_dma_configure(struct pci_dev *dev) { struct device *bridge = pci_get_host_bridge_device(dev); - if (IS_ENABLED(CONFIG_OF) && dev->dev.of_node) { - if (bridge->parent) + if (IS_ENABLED(CONFIG_OF) && + bridge->parent && bridge->parent->of_node) { of_dma_configure(&dev->dev, bridge->parent->of_node); } else if (has_acpi_companion(bridge)) { struct acpi_device *adev = to_acpi_device_node(bridge->fwnode); From 638148e20c7f8f6e95017fdc13bce8549a6925e0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Wed, 18 Nov 2015 21:12:33 +0100 Subject: [PATCH 281/335] USB: option: add XS Stick W100-2 from 4G Systems MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Thomas reports " 4gsystems sells two total different LTE-surfsticks under the same name. .. The newer version of XS Stick W100 is from "omega" .. Under windows the driver switches to the same ID, and uses MI03\6 for network and MI01\6 for modem. .. echo "1c9e 9b01" > /sys/bus/usb/drivers/qmi_wwan/new_id echo "1c9e 9b01" > /sys/bus/usb-serial/drivers/option1/new_id T: Bus=01 Lev=01 Prnt=01 Port=03 Cnt=01 Dev#= 4 Spd=480 MxCh= 0 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1 P: Vendor=1c9e ProdID=9b01 Rev=02.32 S: Manufacturer=USB Modem S: Product=USB Modem S: SerialNumber= C: #Ifs= 5 Cfg#= 1 Atr=80 MxPwr=500mA I: If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=ff Driver=option I: If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=option I: If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=option I: If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan I: If#= 4 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage Now all important things are there: wwp0s29f7u2i3 (net), ttyUSB2 (at), cdc-wdm0 (qmi), ttyUSB1 (at) There is also ttyUSB0, but it is not usable, at least not for at. The device works well with qmi and ModemManager-NetworkManager. " Reported-by: Thomas Schäfer Cc: Signed-off-by: Bjørn Mork Signed-off-by: Greg Kroah-Hartman --- drivers/usb/serial/option.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 2ab2a33e00c4..f2280606b73c 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -355,6 +355,7 @@ static void option_instat_callback(struct urb *urb); /* This is the 4G XS Stick W14 a.k.a. Mobilcom Debitel Surf-Stick * * It seems to contain a Qualcomm QSC6240/6290 chipset */ #define FOUR_G_SYSTEMS_PRODUCT_W14 0x9603 +#define FOUR_G_SYSTEMS_PRODUCT_W100 0x9b01 /* iBall 3.5G connect wireless modem */ #define IBALL_3_5G_CONNECT 0x9605 @@ -520,6 +521,11 @@ static const struct option_blacklist_info four_g_w14_blacklist = { .sendsetup = BIT(0) | BIT(1), }; +static const struct option_blacklist_info four_g_w100_blacklist = { + .sendsetup = BIT(1) | BIT(2), + .reserved = BIT(3), +}; + static const struct option_blacklist_info alcatel_x200_blacklist = { .sendsetup = BIT(0) | BIT(1), .reserved = BIT(4), @@ -1643,6 +1649,9 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE(LONGCHEER_VENDOR_ID, FOUR_G_SYSTEMS_PRODUCT_W14), .driver_info = (kernel_ulong_t)&four_g_w14_blacklist }, + { USB_DEVICE(LONGCHEER_VENDOR_ID, FOUR_G_SYSTEMS_PRODUCT_W100), + .driver_info = (kernel_ulong_t)&four_g_w100_blacklist + }, { USB_DEVICE_INTERFACE_CLASS(LONGCHEER_VENDOR_ID, SPEEDUP_PRODUCT_SU9800, 0xff) }, { USB_DEVICE(LONGCHEER_VENDOR_ID, ZOOM_PRODUCT_4597) }, { USB_DEVICE(LONGCHEER_VENDOR_ID, IBALL_3_5G_CONNECT) }, From c4f161308ac280ce5b27681ddcc2bbac6de39c20 Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Thu, 19 Nov 2015 11:47:47 +0800 Subject: [PATCH 282/335] usb: kconfig: fix warning of select USB_OTG When choose randconfig for kernel build, it reports below warning: "warning: (USB_OTG_FSM && FSL_USB2_OTG && USB_MV_OTG) selects USB_OTG which has unmet direct dependencies (USB_SUPPORT && USB && PM)" In fact, USB_OTG is visible symbol and depends on PM, so the driver needs to depend on it to reduce dependency problem. Signed-off-by: Peter Chen Reported-by: Arnd Bergmann Cc: Felipe Balbi Acked-by: Arnd Bergmann Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/Kconfig | 3 +-- drivers/usb/phy/Kconfig | 4 +--- 2 files changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/usb/core/Kconfig b/drivers/usb/core/Kconfig index a99c89e78126..dd280108758f 100644 --- a/drivers/usb/core/Kconfig +++ b/drivers/usb/core/Kconfig @@ -77,8 +77,7 @@ config USB_OTG_BLACKLIST_HUB config USB_OTG_FSM tristate "USB 2.0 OTG FSM implementation" - depends on USB - select USB_OTG + depends on USB && USB_OTG select USB_PHY help Implements OTG Finite State Machine as specified in On-The-Go diff --git a/drivers/usb/phy/Kconfig b/drivers/usb/phy/Kconfig index 173132416170..22e8ecb6bfbd 100644 --- a/drivers/usb/phy/Kconfig +++ b/drivers/usb/phy/Kconfig @@ -21,7 +21,6 @@ config AB8500_USB config FSL_USB2_OTG bool "Freescale USB OTG Transceiver Driver" depends on USB_EHCI_FSL && USB_FSL_USB2 && USB_OTG_FSM && PM - select USB_OTG select USB_PHY help Enable this to support Freescale USB OTG transceiver. @@ -168,8 +167,7 @@ config USB_QCOM_8X16_PHY config USB_MV_OTG tristate "Marvell USB OTG support" - depends on USB_EHCI_MV && USB_MV_UDC && PM - select USB_OTG + depends on USB_EHCI_MV && USB_MV_UDC && PM && USB_OTG select USB_PHY help Say Y here if you want to build Marvell USB OTG transciever From 44243ef42d77ae367e6e9c68afca3b4578d2d461 Mon Sep 17 00:00:00 2001 From: Simon Arlott Date: Sun, 15 Nov 2015 17:12:08 +0000 Subject: [PATCH 283/335] USB: MAINTAINERS: cxacru Make cxacru an orphan. I still have a few of these devices for testing but haven't had an ADSL1 connection for several years. Signed-off-by: Simon Arlott Cc: Duncan Sands Signed-off-by: Greg Kroah-Hartman --- MAINTAINERS | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index e9caa4b28828..d1ef313905bf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2929,10 +2929,9 @@ S: Maintained F: drivers/platform/x86/compal-laptop.c CONEXANT ACCESSRUNNER USB DRIVER -M: Simon Arlott L: accessrunner-general@lists.sourceforge.net W: http://accessrunner.sourceforge.net/ -S: Maintained +S: Orphan F: drivers/usb/atm/cxacru.c CONFIGFS From 19cd80a214821f4b558560ebd76bfb2c38b4f3d8 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Mon, 2 Nov 2015 10:27:00 +0100 Subject: [PATCH 284/335] usblp: do not set TASK_INTERRUPTIBLE before lock It is not permitted to set task state before lock. usblp_wwait sets the state to TASK_INTERRUPTIBLE and calls mutex_lock_interruptible. Upon return from that function, the state will be TASK_RUNNING again. This is clearly a bug and a warning is generated with LOCKDEP too: WARNING: CPU: 1 PID: 5109 at kernel/sched/core.c:7404 __might_sleep+0x7d/0x90() do not call blocking ops when !TASK_RUNNING; state=1 set at [] usblp_wwait+0xa0/0x310 [usblp] Modules linked in: ... CPU: 1 PID: 5109 Comm: captmon Tainted: G W 4.2.5-0.gef2823b-default #1 Hardware name: LENOVO 23252SG/23252SG, BIOS G2ET33WW (1.13 ) 07/24/2012 ffffffff81a4edce ffff880236ec7ba8 ffffffff81716651 0000000000000000 ffff880236ec7bf8 ffff880236ec7be8 ffffffff8106e146 0000000000000282 ffffffff81a50119 000000000000028b 0000000000000000 ffff8802dab7c508 Call Trace: ... [] warn_slowpath_fmt+0x46/0x50 [] __might_sleep+0x7d/0x90 [] mutex_lock_interruptible_nested+0x2f/0x4b0 [] usblp_wwait+0xcc/0x310 [usblp] [] usblp_write+0x72/0x350 [usblp] [] __vfs_write+0x28/0xf0 ... Commit 7f477358e2384c54b190cc3b6ce28277050a041b (usblp: Implement the ENOSPC convention) moved the set prior locking. So move it back after the lock. Signed-off-by: Jiri Slaby Fixes: 7f477358e2 ("usblp: Implement the ENOSPC convention") Acked-By: Pete Zaitcev Cc: stable Signed-off-by: Greg Kroah-Hartman --- drivers/usb/class/usblp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/class/usblp.c b/drivers/usb/class/usblp.c index 433bbc34a8a4..071964c7847f 100644 --- a/drivers/usb/class/usblp.c +++ b/drivers/usb/class/usblp.c @@ -884,11 +884,11 @@ static int usblp_wwait(struct usblp *usblp, int nonblock) add_wait_queue(&usblp->wwait, &waita); for (;;) { - set_current_state(TASK_INTERRUPTIBLE); if (mutex_lock_interruptible(&usblp->mut)) { rc = -EINTR; break; } + set_current_state(TASK_INTERRUPTIBLE); rc = usblp_wtest(usblp, nonblock); mutex_unlock(&usblp->mut); if (rc <= 0) From 2b3f34451054768a3da27ea1e86da676bf4621bd Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 10:09:18 +0100 Subject: [PATCH 285/335] parisc: Fix wrong comment regarding first pmd entry flags The first pmd entry is marked with PxD_FLAG_ATTACHED instead of _PAGE_GATEWAY. Signed-off-by: Helge Deller --- arch/parisc/include/asm/pgalloc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/parisc/include/asm/pgalloc.h b/arch/parisc/include/asm/pgalloc.h index 3edbb9fc91b4..f2fd327dce2e 100644 --- a/arch/parisc/include/asm/pgalloc.h +++ b/arch/parisc/include/asm/pgalloc.h @@ -35,7 +35,7 @@ static inline pgd_t *pgd_alloc(struct mm_struct *mm) PxD_FLAG_VALID | PxD_FLAG_ATTACHED) + (__u32)(__pa((unsigned long)pgd) >> PxD_VALUE_SHIFT)); - /* The first pmd entry also is marked with _PAGE_GATEWAY as + /* The first pmd entry also is marked with PxD_FLAG_ATTACHED as * a signal that this pmd may not be freed */ __pgd_val_set(*pgd, PxD_FLAG_ATTACHED); #endif From 1e208ae7326caa33ef7d4496d7ff1c404657a4a3 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 10:12:42 +0100 Subject: [PATCH 286/335] parisc: Drop definition of start_thread_som for HP-UX SOM binaries The definition of start_thread_som was planned to be used to execute HP-UX SOM binaries. Since HP-UX compatibility was dropped with kernel 4.0 there is no need to carry it further. Signed-off-by: Helge Deller --- arch/parisc/include/asm/processor.h | 27 --------------------------- 1 file changed, 27 deletions(-) diff --git a/arch/parisc/include/asm/processor.h b/arch/parisc/include/asm/processor.h index 54adb60c0a42..7e759ecb1343 100644 --- a/arch/parisc/include/asm/processor.h +++ b/arch/parisc/include/asm/processor.h @@ -192,33 +192,6 @@ void show_trace(struct task_struct *task, unsigned long *stack); */ typedef unsigned int elf_caddr_t; -#define start_thread_som(regs, new_pc, new_sp) do { \ - unsigned long *sp = (unsigned long *)new_sp; \ - __u32 spaceid = (__u32)current->mm->context; \ - unsigned long pc = (unsigned long)new_pc; \ - /* offset pc for priv. level */ \ - pc |= 3; \ - \ - regs->iasq[0] = spaceid; \ - regs->iasq[1] = spaceid; \ - regs->iaoq[0] = pc; \ - regs->iaoq[1] = pc + 4; \ - regs->sr[2] = LINUX_GATEWAY_SPACE; \ - regs->sr[3] = 0xffff; \ - regs->sr[4] = spaceid; \ - regs->sr[5] = spaceid; \ - regs->sr[6] = spaceid; \ - regs->sr[7] = spaceid; \ - regs->gr[ 0] = USER_PSW; \ - regs->gr[30] = ((new_sp)+63)&~63; \ - regs->gr[31] = pc; \ - \ - get_user(regs->gr[26],&sp[0]); \ - get_user(regs->gr[25],&sp[-1]); \ - get_user(regs->gr[24],&sp[-2]); \ - get_user(regs->gr[23],&sp[-3]); \ -} while(0) - /* The ELF abi wants things done a "wee bit" differently than * som does. Supporting this behavior here avoids * having our own version of create_elf_tables. From 5011a7e808c9fec643d752c5a495a48f27268a48 Mon Sep 17 00:00:00 2001 From: Alban Bedel Date: Tue, 17 Nov 2015 09:40:07 +0100 Subject: [PATCH 287/335] MIPS: ath79: Fix the DDR control initialization on ar71xx and ar934x The DDR control initialization needs to know the SoC type, however ath79_detect_sys_type() was called after ath79_ddr_ctrl_init(). Reverse the order to fix the DDR control initialization on ar71xx and ar934x. Signed-off-by: Alban Bedel Cc: Felix Fietkau Cc: Qais Yousef Cc: Andrew Bresticker CC: stable@vger.kernel.org # v4.2+ Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/11500/ Signed-off-by: Ralf Baechle --- arch/mips/ath79/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/ath79/setup.c b/arch/mips/ath79/setup.c index 1ba21204ebe0..9a0013703579 100644 --- a/arch/mips/ath79/setup.c +++ b/arch/mips/ath79/setup.c @@ -216,9 +216,9 @@ void __init plat_mem_setup(void) AR71XX_RESET_SIZE); ath79_pll_base = ioremap_nocache(AR71XX_PLL_BASE, AR71XX_PLL_SIZE); + ath79_detect_sys_type(); ath79_ddr_ctrl_init(); - ath79_detect_sys_type(); if (mips_machtype != ATH79_MACH_GENERIC_OF) detect_memory_region(0, ATH79_MEM_SIZE_MIN, ATH79_MEM_SIZE_MAX); From accbfb52d09ac00147f02c4335d1cc47665e6653 Mon Sep 17 00:00:00 2001 From: Alban Bedel Date: Tue, 17 Nov 2015 20:34:51 +0100 Subject: [PATCH 288/335] MIPS: ath79: Fix the size of the MISC INTC registers in ar9132.dtsi There is 2 registers that is 8 bytes long, not 4. Signed-off-by: Alban Bedel Cc: Thomas Gleixner Cc: Jason Cooper Cc: Marc Zyngier Cc: Alexander Couzens Cc: Joel Porquet Cc: Andrew Bresticker Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/11508/ Signed-off-by: Ralf Baechle --- arch/mips/boot/dts/qca/ar9132.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/boot/dts/qca/ar9132.dtsi b/arch/mips/boot/dts/qca/ar9132.dtsi index fb7734eadbf0..13d0439496a9 100644 --- a/arch/mips/boot/dts/qca/ar9132.dtsi +++ b/arch/mips/boot/dts/qca/ar9132.dtsi @@ -107,7 +107,7 @@ miscintc: interrupt-controller@18060010 { compatible = "qca,ar9132-misc-intc", "qca,ar7100-misc-intc"; - reg = <0x18060010 0x4>; + reg = <0x18060010 0x8>; interrupt-parent = <&cpuintc>; interrupts = <6>; From 55f1d5988c52d481dc489e29ee5e8905b18ff859 Mon Sep 17 00:00:00 2001 From: Alban Bedel Date: Tue, 17 Nov 2015 21:52:00 +0100 Subject: [PATCH 289/335] MIPS: ath79: Add a machine entry for booting OF machines As I'm using a board with a broken old bootloader I hardcoded the mips_machtype and did't notice that the machine entry was still missing. [ralf@linux-mips.org: Fixed spelling message noticed by Sergei Shtylyov .] Signed-off-by: Alban Bedel Cc: Qais Yousef Cc: Felix Fietkau Cc: Andrew Bresticker Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/11503/ Signed-off-by: Ralf Baechle --- arch/mips/ath79/setup.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/mips/ath79/setup.c b/arch/mips/ath79/setup.c index 9a0013703579..8755d618e116 100644 --- a/arch/mips/ath79/setup.c +++ b/arch/mips/ath79/setup.c @@ -281,3 +281,8 @@ MIPS_MACHINE(ATH79_MACH_GENERIC, "Generic", "Generic AR71XX/AR724X/AR913X based board", ath79_generic_init); + +MIPS_MACHINE(ATH79_MACH_GENERIC_OF, + "DTB", + "Generic AR71XX/AR724X/AR913X based board (DT)", + NULL); From 8bb6fd585d419d906fb3c228db47fbc4e4272a00 Mon Sep 17 00:00:00 2001 From: Hou Zhiqiang Date: Tue, 17 Nov 2015 17:53:18 +0800 Subject: [PATCH 290/335] i2c: imx: fix a compiling error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit drivers/i2c/busses/i2c-imx.c:978:2: error: implicit declaration of function ‘pinctrl_select_state’ [-Werror=implicit-function-declaration] pinctrl_select_state(i2c_imx->pinctrl, i2c_imx->pinctrl_pins_gpio); ^ Signed-off-by: Hou Zhiqiang Signed-off-by: Wolfram Sang --- drivers/i2c/busses/i2c-imx.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c index 1e4d99da4164..9bb0b056b25f 100644 --- a/drivers/i2c/busses/i2c-imx.c +++ b/drivers/i2c/busses/i2c-imx.c @@ -50,6 +50,7 @@ #include #include #include +#include #include #include #include From 9656eeebf3f1dd05376c4c923797369746d9a618 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Mon, 16 Nov 2015 14:42:03 +0100 Subject: [PATCH 291/335] i2c: Revert "i2c: xiic: Do not reset controller before every transfer" Commit d701667bb331 ("i2c: xiic: Do not reset controller before every transfer") removed the reinitialization of the controller before the start of each transfer. Apparently this change is not safe to make and the commit results in random I2C bus failures. An easy way to trigger the issue is to run i2cdetect. Without the patch applied: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- UU UU -- UU 3c -- -- UU 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 70: -- -- -- -- -- -- -- -- With the patch applied every other or so invocation: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 10: 10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f 20: 20 21 22 23 24 25 26 27 28 29 2a 2b 2c 2d 2e 2f 30: -- -- -- -- -- -- -- -- UU UU -- UU 3c -- -- UU 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 70: -- -- -- -- -- -- -- -- So revert the commit for now. Fixes: d701667bb331 ("i2c: xiic: Do not reset controller before every transfer") Signed-off-by: Lars-Peter Clausen Acked-by: Shubhrajyoti Datta Signed-off-by: Wolfram Sang --- drivers/i2c/busses/i2c-xiic.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/i2c/busses/i2c-xiic.c b/drivers/i2c/busses/i2c-xiic.c index e23a7b068c60..705cf69ccf42 100644 --- a/drivers/i2c/busses/i2c-xiic.c +++ b/drivers/i2c/busses/i2c-xiic.c @@ -662,6 +662,9 @@ static void __xiic_start_xfer(struct xiic_i2c *i2c) static void xiic_start_xfer(struct xiic_i2c *i2c) { + spin_lock(&i2c->lock); + xiic_reinit(i2c); + spin_unlock(&i2c->lock); __xiic_start_xfer(i2c); } From d0fe5258e6be609dab7e53335d845818684d1dcf Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Mon, 16 Nov 2015 14:42:04 +0100 Subject: [PATCH 292/335] i2c: xiic: Prevent concurrent running of the IRQ handler and __xiic_start_xfer() Prior to commit e6c9a037bc8a ("i2c: xiic: Remove the disabling of interrupts") IRQs where disabled when the initial __xiic_start_xfer() was called. After the commit the interrupt is enabled while the function is running, this means it is possible for the interrupt to be triggered while the function is still running. When this happens the internal data structures get corrupted and undefined behavior can occur like the following crash: Internal error: Oops: 17 [#1] PREEMPT SMP ARM Modules linked in: CPU: 0 PID: 2040 Comm: i2cdetect Not tainted 4.0.0-02856-g047a308 #10956 Hardware name: Xilinx Zynq Platform task: ee0c9500 ti: e99a2000 task.ti: e99a2000 PC is at __xiic_start_xfer+0x6c4/0x7c8 LR is at __xiic_start_xfer+0x690/0x7c8 pc : [] lr : [] psr: 800f0013 sp : e99a3da8 ip : 00000000 fp : 00000000 r10: 00000001 r9 : 600f0013 r8 : f0180000 r7 : f0180000 r6 : c064e444 r5 : 00000017 r4 : ee031010 r3 : 00000000 r2 : 00000000 r1 : 600f0013 r0 : 0000000f Flags: Nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 18c5387d Table: 29a5404a DAC: 00000015 Process i2cdetect (pid: 2040, stack limit = 0xe99a2210) Stack: (0xe99a3da8 to 0xe99a4000) 3da0: ee031010 00000000 00000001 ee031020 ee031224 c02bc5ec 3dc0: ee34c604 00000000 ee0c9500 e99a3dcc e99a3dd0 e99a3dd0 e99a3dd8 c069f0e8 3de0: 00000000 ee031020 c064e100 ffff90bb e99a3e48 c02b6590 ee031020 00000001 3e00: e99a3e48 ee031020 00000000 e99a3e63 00000001 c02b6ec4 00000000 00000000 3e20: 00000000 c02b7320 e99a3ef0 00000000 00000000 e99e3df0 00000000 00000000 3e40: 00000103 2814575f 0000003e c00a0000 e99a3e85 0001003e ee0c0000 e99a3e63 3e60: eefd3578 c064e61c ee0c9500 c0041e04 0000056c e9a56db8 00006e5a b6f5c000 3e80: ee0c9548 eefd0040 00000001 eefd3540 ee0c9500 eefd39a0 c064b540 ee0c9500 3ea0: 00000000 ee92b000 00000000 bef4862c ee34c600 e99ecdc0 00000720 00000003 3ec0: e99a2000 00000000 00000000 c02b8b30 00000000 00000000 00000000 e99a3f24 3ee0: b6e80000 00000000 00000000 c04257e8 00000000 e99a3f24 c02b8f08 00000703 3f00: 00000003 c02116bc ee935300 00000000 bef4862c ee34c600 e99ecdc0 c02b91f0 3f20: e99ecdc0 00000720 bef4862c eeb725f8 e99ecdc0 c00c9e2c 00000003 00000003 3f40: ee248dc0 00000000 ee248dc8 00000002 eeb7c1a8 00000000 00000000 c00bb360 3f60: 00000000 00000000 00000003 ee248dc0 bef4862c e99ecdc0 e99ecdc0 00000720 3f80: 00000003 e99a2000 00000000 c00c9f68 00000000 00000000 b6f22000 00000036 3fa0: c000dfa4 c000de20 00000000 00000000 00000003 00000720 bef4862c bef4862c 3fc0: 00000000 00000000 b6f22000 00000036 00000000 00000000 b6f60000 00000000 3fe0: 00013040 bef48614 00008cab b6ecdbe6 400f0030 00000003 2f7fd821 2f7fdc21 [] (__xiic_start_xfer) from [] (xiic_xfer+0x94/0x168) [] (xiic_xfer) from [] (__i2c_transfer+0x4c/0x7c) [] (__i2c_transfer) from [] (i2c_transfer+0x9c/0xc4) [] (i2c_transfer) from [] (i2c_smbus_xfer+0x3a0/0x4ec) [] (i2c_smbus_xfer) from [] (i2cdev_ioctl_smbus+0xb0/0x214) [] (i2cdev_ioctl_smbus) from [] (i2cdev_ioctl+0xa0/0x1d4) [] (i2cdev_ioctl) from [] (do_vfs_ioctl+0x4b0/0x5b8) [] (do_vfs_ioctl) from [] (SyS_ioctl+0x34/0x5c) [] (SyS_ioctl) from [] (ret_fast_syscall+0x0/0x34) Code: e283300c e5843210 eafffe64 e5943210 (e1d320b4) The issue can easily be reproduced by performing I2C access under high system load or IO load. To fix the issue protect the invocation to __xiic_start_xfer() form xiic_start_xfer() with the same lock that is used to protect the interrupt handler. Fixes: e6c9a037bc8a ("i2c: xiic: Remove the disabling of interrupts") Signed-off-by: Lars-Peter Clausen Reviewed-by: Shubhrajyoti Datta Signed-off-by: Wolfram Sang --- drivers/i2c/busses/i2c-xiic.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/i2c/busses/i2c-xiic.c b/drivers/i2c/busses/i2c-xiic.c index 705cf69ccf42..0b20449e48cf 100644 --- a/drivers/i2c/busses/i2c-xiic.c +++ b/drivers/i2c/busses/i2c-xiic.c @@ -664,9 +664,8 @@ static void xiic_start_xfer(struct xiic_i2c *i2c) { spin_lock(&i2c->lock); xiic_reinit(i2c); - spin_unlock(&i2c->lock); - __xiic_start_xfer(i2c); + spin_unlock(&i2c->lock); } static int xiic_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num) From c18fba23061f16dde128e10d4869ba4e88e0e81a Mon Sep 17 00:00:00 2001 From: Grygorii Strashko Date: Thu, 12 Nov 2015 15:42:26 +0200 Subject: [PATCH 293/335] i2c: fix wakeup irq parsing This patch fixes obvious copy-past error in wake up irq parsing code which leads to the fact that dev_pm_set_wake_irq() will be called with wrong IRQ number when "wakeup" IRQ is not defined in DT. Fixes: 3fffd1283927 ("i2c: allow specifying separate wakeup interrupt in device tree") Signed-off-by: Grygorii Strashko Acked-by: Dmitry Torokhov Signed-off-by: Wolfram Sang Cc: # v4.3 --- drivers/i2c/i2c-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c index 040af5cc8143..ba8eb087f224 100644 --- a/drivers/i2c/i2c-core.c +++ b/drivers/i2c/i2c-core.c @@ -715,7 +715,7 @@ static int i2c_device_probe(struct device *dev) if (wakeirq > 0 && wakeirq != client->irq) status = dev_pm_set_dedicated_wake_irq(dev, wakeirq); else if (client->irq > 0) - status = dev_pm_set_wake_irq(dev, wakeirq); + status = dev_pm_set_wake_irq(dev, client->irq); else status = 0; From cdc5a3110e7c3ae793f367285789a6bc39c962dc Mon Sep 17 00:00:00 2001 From: Alexandra Yates Date: Thu, 5 Nov 2015 11:40:25 -0800 Subject: [PATCH 294/335] i2c: i801: add Intel Lewisburg device IDs Adding Intel codename Lewisburg platform device IDs for SMBus. Signed-off-by: Alexandra Yates Reviewed-by: Jean Delvare Signed-off-by: Wolfram Sang --- Documentation/i2c/busses/i2c-i801 | 1 + drivers/i2c/busses/Kconfig | 1 + drivers/i2c/busses/i2c-i801.c | 6 ++++++ 3 files changed, 8 insertions(+) diff --git a/Documentation/i2c/busses/i2c-i801 b/Documentation/i2c/busses/i2c-i801 index 6a4b1af724f8..1bba38dd2637 100644 --- a/Documentation/i2c/busses/i2c-i801 +++ b/Documentation/i2c/busses/i2c-i801 @@ -32,6 +32,7 @@ Supported adapters: * Intel Sunrise Point-LP (PCH) * Intel DNV (SOC) * Intel Broxton (SOC) + * Intel Lewisburg (PCH) Datasheets: Publicly available at the Intel website On Intel Patsburg and later chipsets, both the normal host SMBus controller diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig index e24c2b680b47..7b0aa82ea38b 100644 --- a/drivers/i2c/busses/Kconfig +++ b/drivers/i2c/busses/Kconfig @@ -126,6 +126,7 @@ config I2C_I801 Sunrise Point-LP (PCH) DNV (SOC) Broxton (SOC) + Lewisburg (PCH) This driver can also be built as a module. If so, the module will be called i2c-i801. diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c index c306751ceadb..f62d69799a9c 100644 --- a/drivers/i2c/busses/i2c-i801.c +++ b/drivers/i2c/busses/i2c-i801.c @@ -62,6 +62,8 @@ * Sunrise Point-LP (PCH) 0x9d23 32 hard yes yes yes * DNV (SOC) 0x19df 32 hard yes yes yes * Broxton (SOC) 0x5ad4 32 hard yes yes yes + * Lewisburg (PCH) 0xa1a3 32 hard yes yes yes + * Lewisburg Supersku (PCH) 0xa223 32 hard yes yes yes * * Features supported by this driver: * Software PEC no @@ -206,6 +208,8 @@ #define PCI_DEVICE_ID_INTEL_SUNRISEPOINT_LP_SMBUS 0x9d23 #define PCI_DEVICE_ID_INTEL_DNV_SMBUS 0x19df #define PCI_DEVICE_ID_INTEL_BROXTON_SMBUS 0x5ad4 +#define PCI_DEVICE_ID_INTEL_LEWISBURG_SMBUS 0xa1a3 +#define PCI_DEVICE_ID_INTEL_LEWISBURG_SSKU_SMBUS 0xa223 struct i801_mux_config { char *gpio_chip; @@ -869,6 +873,8 @@ static const struct pci_device_id i801_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SUNRISEPOINT_LP_SMBUS) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_DNV_SMBUS) }, { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_BROXTON_SMBUS) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_LEWISBURG_SMBUS) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_LEWISBURG_SSKU_SMBUS) }, { 0, } }; From 94a58c360a45c066ab5472cfd2bf2a4ba63aa532 Mon Sep 17 00:00:00 2001 From: Rasmus Villemoes Date: Fri, 20 Nov 2015 15:56:48 -0800 Subject: [PATCH 295/335] slab.h: sprinkle __assume_aligned attributes The various allocators return aligned memory. Telling the compiler that allows it to generate better code in many cases, for example when the return value is immediately passed to memset(). Some code does become larger, but at least we win twice as much as we lose: $ scripts/bloat-o-meter /tmp/vmlinux vmlinux add/remove: 0/0 grow/shrink: 13/52 up/down: 995/-2140 (-1145) An example of the different (and smaller) code can be seen in mm_alloc(). Before: : 48 8d 78 08 lea 0x8(%rax),%rdi : 48 89 c1 mov %rax,%rcx : 48 89 c2 mov %rax,%rdx : 48 c7 00 00 00 00 00 movq $0x0,(%rax) : 48 c7 80 48 03 00 00 movq $0x0,0x348(%rax) : 00 00 00 00 : 31 c0 xor %eax,%eax : 48 83 e7 f8 and $0xfffffffffffffff8,%rdi : 48 29 f9 sub %rdi,%rcx : 81 c1 50 03 00 00 add $0x350,%ecx : c1 e9 03 shr $0x3,%ecx : f3 48 ab rep stos %rax,%es:(%rdi) After: : 48 89 c2 mov %rax,%rdx : b9 6a 00 00 00 mov $0x6a,%ecx : 31 c0 xor %eax,%eax : 48 89 d7 mov %rdx,%rdi : f3 48 ab rep stos %rax,%es:(%rdi) So gcc's strategy is to do two possibly (but not really, of course) unaligned stores to the first and last word, then do an aligned rep stos covering the middle part with a little overlap. Maybe arches which do not allow unaligned stores gain even more. I don't know if gcc can actually make use of alignments greater than 8 for anything, so one could probably drop the __assume_xyz_alignment macros and just use __assume_aligned(8). The increases in code size are mostly caused by gcc deciding to opencode strlen() using the check-four-bytes-at-a-time trick when it knows the buffer is sufficiently aligned (one function grew by 200 bytes). Now it turns out that many of these strlen() calls showing up were in fact redundant, and they're gone from -next. Applying the two patches to next-20151001 bloat-o-meter instead says add/remove: 0/0 grow/shrink: 6/52 up/down: 244/-2140 (-1896) Signed-off-by: Rasmus Villemoes Acked-by: Christoph Lameter Cc: David Rientjes Cc: Pekka Enberg Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/slab.h | 43 ++++++++++++++++++++++++++----------------- 1 file changed, 26 insertions(+), 17 deletions(-) diff --git a/include/linux/slab.h b/include/linux/slab.h index 7c82e3b307a3..96940772bb92 100644 --- a/include/linux/slab.h +++ b/include/linux/slab.h @@ -157,6 +157,24 @@ size_t ksize(const void *); #define ARCH_KMALLOC_MINALIGN __alignof__(unsigned long long) #endif +/* + * Setting ARCH_SLAB_MINALIGN in arch headers allows a different alignment. + * Intended for arches that get misalignment faults even for 64 bit integer + * aligned buffers. + */ +#ifndef ARCH_SLAB_MINALIGN +#define ARCH_SLAB_MINALIGN __alignof__(unsigned long long) +#endif + +/* + * kmalloc and friends return ARCH_KMALLOC_MINALIGN aligned + * pointers. kmem_cache_alloc and friends return ARCH_SLAB_MINALIGN + * aligned pointers. + */ +#define __assume_kmalloc_alignment __assume_aligned(ARCH_KMALLOC_MINALIGN) +#define __assume_slab_alignment __assume_aligned(ARCH_SLAB_MINALIGN) +#define __assume_page_alignment __assume_aligned(PAGE_SIZE) + /* * Kmalloc array related definitions */ @@ -286,8 +304,8 @@ static __always_inline int kmalloc_index(size_t size) } #endif /* !CONFIG_SLOB */ -void *__kmalloc(size_t size, gfp_t flags); -void *kmem_cache_alloc(struct kmem_cache *, gfp_t flags); +void *__kmalloc(size_t size, gfp_t flags) __assume_kmalloc_alignment; +void *kmem_cache_alloc(struct kmem_cache *, gfp_t flags) __assume_slab_alignment; void kmem_cache_free(struct kmem_cache *, void *); /* @@ -301,8 +319,8 @@ void kmem_cache_free_bulk(struct kmem_cache *, size_t, void **); bool kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **); #ifdef CONFIG_NUMA -void *__kmalloc_node(size_t size, gfp_t flags, int node); -void *kmem_cache_alloc_node(struct kmem_cache *, gfp_t flags, int node); +void *__kmalloc_node(size_t size, gfp_t flags, int node) __assume_kmalloc_alignment; +void *kmem_cache_alloc_node(struct kmem_cache *, gfp_t flags, int node) __assume_slab_alignment; #else static __always_inline void *__kmalloc_node(size_t size, gfp_t flags, int node) { @@ -316,12 +334,12 @@ static __always_inline void *kmem_cache_alloc_node(struct kmem_cache *s, gfp_t f #endif #ifdef CONFIG_TRACING -extern void *kmem_cache_alloc_trace(struct kmem_cache *, gfp_t, size_t); +extern void *kmem_cache_alloc_trace(struct kmem_cache *, gfp_t, size_t) __assume_slab_alignment; #ifdef CONFIG_NUMA extern void *kmem_cache_alloc_node_trace(struct kmem_cache *s, gfp_t gfpflags, - int node, size_t size); + int node, size_t size) __assume_slab_alignment; #else static __always_inline void * kmem_cache_alloc_node_trace(struct kmem_cache *s, @@ -354,10 +372,10 @@ kmem_cache_alloc_node_trace(struct kmem_cache *s, } #endif /* CONFIG_TRACING */ -extern void *kmalloc_order(size_t size, gfp_t flags, unsigned int order); +extern void *kmalloc_order(size_t size, gfp_t flags, unsigned int order) __assume_page_alignment; #ifdef CONFIG_TRACING -extern void *kmalloc_order_trace(size_t size, gfp_t flags, unsigned int order); +extern void *kmalloc_order_trace(size_t size, gfp_t flags, unsigned int order) __assume_page_alignment; #else static __always_inline void * kmalloc_order_trace(size_t size, gfp_t flags, unsigned int order) @@ -482,15 +500,6 @@ static __always_inline void *kmalloc_node(size_t size, gfp_t flags, int node) return __kmalloc_node(size, flags, node); } -/* - * Setting ARCH_SLAB_MINALIGN in arch headers allows a different alignment. - * Intended for arches that get misalignment faults even for 64 bit integer - * aligned buffers. - */ -#ifndef ARCH_SLAB_MINALIGN -#define ARCH_SLAB_MINALIGN __alignof__(unsigned long long) -#endif - struct memcg_cache_array { struct rcu_head rcu; struct kmem_cache *entries[0]; From dd7d664a2bcf93634a57d208e8e480ab7448c107 Mon Sep 17 00:00:00 2001 From: Moritz Fischer Date: Fri, 20 Nov 2015 15:56:51 -0800 Subject: [PATCH 296/335] MAINTAINERS: add Moritz as reviewer for FPGA Manager Framework Nominate myself as Reviewer. Signed-off-by: Moritz Fischer Acked-by: Alan Tull Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index b16bffabe70a..0f5fa0f978c2 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4409,6 +4409,7 @@ K: fmc_d.*register FPGA MANAGER FRAMEWORK M: Alan Tull +R: Moritz Fischer S: Maintained F: drivers/fpga/ F: include/linux/fpga/fpga-mgr.h From 5cf6a51e6062afe7cc507f32f1e5f7e6497ae844 Mon Sep 17 00:00:00 2001 From: Daniel Baluta Date: Fri, 20 Nov 2015 15:56:53 -0800 Subject: [PATCH 297/335] configfs: allow dynamic group creation This patchset introduces IIO software triggers, offers a way of configuring them via configfs and adds the IIO hrtimer based interrupt source to be used with software triggers. The architecture is now split in 3 parts, to remove all IIO trigger specific parts from IIO configfs core: (1) IIO configfs - creates the root of the IIO configfs subsys. (2) IIO software triggers - software trigger implementation, dynamically creating /config/iio/triggers group. (3) IIO hrtimer trigger - is the first interrupt source for software triggers (with syfs to follow). Each trigger type can implement its own set of attributes. Lockdep seems to be happy with the locking in configfs patch. This patch (of 5): We don't want to hardcode default groups at subsystem creation time. We export: * configfs_register_group * configfs_unregister_group to allow drivers to programatically create/destroy groups later, after module init time. This is needed for IIO configfs support. (akpm: the other 4 patches to be merged via the IIO tree) Signed-off-by: Daniel Baluta Suggested-by: Lars-Peter Clausen Reviewed-by: Christoph Hellwig Acked-by: Joel Becker Cc: Hartmut Knaack Cc: Octavian Purdila Cc: Paul Bolle Cc: Adriana Reus Cc: Cristina Opriceana Cc: Peter Meerwald Cc: Alexander Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/configfs/dir.c | 110 +++++++++++++++++++++++++++++++++++++++ include/linux/configfs.h | 10 ++++ 2 files changed, 120 insertions(+) diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index c81ce7f200a6..a7a1b218f308 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1636,6 +1636,116 @@ const struct file_operations configfs_dir_operations = { .iterate = configfs_readdir, }; +/** + * configfs_register_group - creates a parent-child relation between two groups + * @parent_group: parent group + * @group: child group + * + * link groups, creates dentry for the child and attaches it to the + * parent dentry. + * + * Return: 0 on success, negative errno code on error + */ +int configfs_register_group(struct config_group *parent_group, + struct config_group *group) +{ + struct configfs_subsystem *subsys = parent_group->cg_subsys; + struct dentry *parent; + int ret; + + mutex_lock(&subsys->su_mutex); + link_group(parent_group, group); + mutex_unlock(&subsys->su_mutex); + + parent = parent_group->cg_item.ci_dentry; + + mutex_lock_nested(&d_inode(parent)->i_mutex, I_MUTEX_PARENT); + ret = create_default_group(parent_group, group); + if (!ret) { + spin_lock(&configfs_dirent_lock); + configfs_dir_set_ready(group->cg_item.ci_dentry->d_fsdata); + spin_unlock(&configfs_dirent_lock); + } + mutex_unlock(&d_inode(parent)->i_mutex); + return ret; +} +EXPORT_SYMBOL(configfs_register_group); + +/** + * configfs_unregister_group() - unregisters a child group from its parent + * @group: parent group to be unregistered + * + * Undoes configfs_register_group() + */ +void configfs_unregister_group(struct config_group *group) +{ + struct configfs_subsystem *subsys = group->cg_subsys; + struct dentry *dentry = group->cg_item.ci_dentry; + struct dentry *parent = group->cg_item.ci_parent->ci_dentry; + + mutex_lock_nested(&d_inode(parent)->i_mutex, I_MUTEX_PARENT); + spin_lock(&configfs_dirent_lock); + configfs_detach_prep(dentry, NULL); + spin_unlock(&configfs_dirent_lock); + + configfs_detach_group(&group->cg_item); + d_inode(dentry)->i_flags |= S_DEAD; + dont_mount(dentry); + d_delete(dentry); + mutex_unlock(&d_inode(parent)->i_mutex); + + dput(dentry); + + mutex_lock(&subsys->su_mutex); + unlink_group(group); + mutex_unlock(&subsys->su_mutex); +} +EXPORT_SYMBOL(configfs_unregister_group); + +/** + * configfs_register_default_group() - allocates and registers a child group + * @parent_group: parent group + * @name: child group name + * @item_type: child item type description + * + * boilerplate to allocate and register a child group with its parent. We need + * kzalloc'ed memory because child's default_group is initially empty. + * + * Return: allocated config group or ERR_PTR() on error + */ +struct config_group * +configfs_register_default_group(struct config_group *parent_group, + const char *name, + struct config_item_type *item_type) +{ + int ret; + struct config_group *group; + + group = kzalloc(sizeof(*group), GFP_KERNEL); + if (!group) + return ERR_PTR(-ENOMEM); + config_group_init_type_name(group, name, item_type); + + ret = configfs_register_group(parent_group, group); + if (ret) { + kfree(group); + return ERR_PTR(ret); + } + return group; +} +EXPORT_SYMBOL(configfs_register_default_group); + +/** + * configfs_unregister_default_group() - unregisters and frees a child group + * @group: the group to act on + */ +void configfs_unregister_default_group(struct config_group *group) +{ + configfs_unregister_group(group); + kfree(group); +} +EXPORT_SYMBOL(configfs_unregister_default_group); + int configfs_register_subsystem(struct configfs_subsystem *subsys) { int err; diff --git a/include/linux/configfs.h b/include/linux/configfs.h index a8a335b7fce0..758a029011b1 100644 --- a/include/linux/configfs.h +++ b/include/linux/configfs.h @@ -197,6 +197,16 @@ static inline struct configfs_subsystem *to_configfs_subsystem(struct config_gro int configfs_register_subsystem(struct configfs_subsystem *subsys); void configfs_unregister_subsystem(struct configfs_subsystem *subsys); +int configfs_register_group(struct config_group *parent_group, + struct config_group *group); +void configfs_unregister_group(struct config_group *group); + +struct config_group * +configfs_register_default_group(struct config_group *parent_group, + const char *name, + struct config_item_type *item_type); +void configfs_unregister_default_group(struct config_group *group); + /* These functions can sleep and can alloc with GFP_KERNEL */ /* WARNING: These cannot be called underneath configfs callbacks!! */ int configfs_depend_item(struct configfs_subsystem *subsys, struct config_item *target); From 1491e30ed1a741009d1d38f9285f7a29e6c05c78 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 20 Nov 2015 15:56:56 -0800 Subject: [PATCH 298/335] ncpfs: don't allow negative timeouts This code causes a static checker warning because it's a user controlled variable where we cap the upper bound but not the lower bound. Let's return an -EINVAL for negative timeouts. [akpm@linux-foundation.org: remove unneeded `else'] Signed-off-by: Dan Carpenter Reviewed-by: Jan Kara Cc: Petr Vandrovec Cc: David Howells Cc: Jan Kara Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ncpfs/ioctl.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 79b113048eac..0a3f9b594602 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -525,6 +525,8 @@ static long __ncp_ioctl(struct inode *inode, unsigned int cmd, unsigned long arg switch (rqdata.cmd) { case NCP_LOCK_EX: case NCP_LOCK_SH: + if (rqdata.timeout < 0) + return -EINVAL; if (rqdata.timeout == 0) rqdata.timeout = NCP_LOCK_DEFAULT_TIMEOUT; else if (rqdata.timeout > NCP_LOCK_MAX_TIMEOUT) From 429d48622cb08b7b4908cb63b43a135be9b3db09 Mon Sep 17 00:00:00 2001 From: Naoya Horiguchi Date: Fri, 20 Nov 2015 15:56:59 -0800 Subject: [PATCH 299/335] tools/vm/page-types.c: support KPF_IDLE PageIdle is exported in include/uapi/linux/kernel-page-flags.h, so let's make page-types.c tool handle it. Signed-off-by: Naoya Horiguchi Reviewed-by: Vladimir Davydov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- tools/vm/page-types.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/vm/page-types.c b/tools/vm/page-types.c index bcf5ec760eb9..5a6016224bb9 100644 --- a/tools/vm/page-types.c +++ b/tools/vm/page-types.c @@ -128,6 +128,7 @@ static const char * const page_flag_names[] = { [KPF_THP] = "t:thp", [KPF_BALLOON] = "o:balloon", [KPF_ZERO_PAGE] = "z:zero_page", + [KPF_IDLE] = "i:idle_page", [KPF_RESERVED] = "r:reserved", [KPF_MLOCKED] = "m:mlocked", From 7511c3ede752e6dd67df20779b4e11effe102637 Mon Sep 17 00:00:00 2001 From: Jerome Marchand Date: Fri, 20 Nov 2015 15:57:02 -0800 Subject: [PATCH 300/335] mm: vmalloc: don't remove inexistent guard hole in remove_vm_area() Commit 71394fe50146 ("mm: vmalloc: add flag preventing guard hole allocation") missed a spot. Currently remove_vm_area() decreases vm->size to "remove" the guard hole page, even when it isn't present. All but one users just free the vm_struct rigth away and never access vm->size anyway. Don't touch the size in remove_vm_area() and have __vunmap() use the proper get_vm_area_size() helper. Signed-off-by: Jerome Marchand Acked-by: Andrey Ryabinin Acked-by: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/vmalloc.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index d04563480c94..8e3c9c5a3042 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -1443,7 +1443,6 @@ struct vm_struct *remove_vm_area(const void *addr) vmap_debug_free_range(va->va_start, va->va_end); kasan_free_shadow(vm); free_unmap_vmap_area(va); - vm->size -= PAGE_SIZE; return vm; } @@ -1468,8 +1467,8 @@ static void __vunmap(const void *addr, int deallocate_pages) return; } - debug_check_no_locks_freed(addr, area->size); - debug_check_no_obj_freed(addr, area->size); + debug_check_no_locks_freed(addr, get_vm_area_size(area)); + debug_check_no_obj_freed(addr, get_vm_area_size(area)); if (deallocate_pages) { int i; From 1a763615688b891246c5b0a932d7a95fea4c1a68 Mon Sep 17 00:00:00 2001 From: "Jason J. Herne" Date: Fri, 20 Nov 2015 15:57:04 -0800 Subject: [PATCH 301/335] mm: loosen MADV_NOHUGEPAGE to enable Qemu postcopy on s390 MADV_NOHUGEPAGE processing is too restrictive. kvm already disables hugepage but hugepage_madvise() takes the error path when we ask to turn on the MADV_NOHUGEPAGE bit and the bit is already on. This causes Qemu's new postcopy migration feature to fail on s390 because its first action is to madvise the guest address space as NOHUGEPAGE. This patch modifies the code so that the operation succeeds without error now. For consistency reasons do the same for MADV_HUGEPAGE. Signed-off-by: Jason J. Herne Reviewed-by: Andrea Arcangeli Acked-by: Christian Borntraeger Acked-by: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/huge_memory.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/huge_memory.c b/mm/huge_memory.c index c29ddebc8705..62fe06bb7d04 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -2009,7 +2009,7 @@ int hugepage_madvise(struct vm_area_struct *vma, /* * Be somewhat over-protective like KSM for now! */ - if (*vm_flags & (VM_HUGEPAGE | VM_NO_THP)) + if (*vm_flags & VM_NO_THP) return -EINVAL; *vm_flags &= ~VM_NOHUGEPAGE; *vm_flags |= VM_HUGEPAGE; @@ -2025,7 +2025,7 @@ int hugepage_madvise(struct vm_area_struct *vma, /* * Be somewhat over-protective like KSM for now! */ - if (*vm_flags & (VM_NOHUGEPAGE | VM_NO_THP)) + if (*vm_flags & VM_NO_THP) return -EINVAL; *vm_flags &= ~VM_HUGEPAGE; *vm_flags |= VM_NOHUGEPAGE; From 1a47de6e4a8dc2aaf3c3fb544b60730b74abe0f1 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 20 Nov 2015 15:57:07 -0800 Subject: [PATCH 302/335] various: fix pci_set_dma_mask return value checking pci_set_dma_mask returns a negative errno value, not a bool like pci_dma_supported. This of course was just a giant test for attention :) Signed-off-by: Christoph Hellwig Reported-by: Jongman Heo Tested-by: Jongman Heo [pcnet32] Acked-by: Mauro Carvalho Chehab Cc: Hans Verkuil Cc: Antti Palosaari Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/media/pci/cx23885/cx23885-core.c | 4 ++-- drivers/media/pci/cx25821/cx25821-core.c | 3 ++- drivers/media/pci/cx88/cx88-alsa.c | 4 ++-- drivers/media/pci/cx88/cx88-mpeg.c | 3 ++- drivers/media/pci/cx88/cx88-video.c | 4 ++-- drivers/media/pci/netup_unidvb/netup_unidvb_core.c | 2 +- drivers/media/pci/saa7134/saa7134-core.c | 4 ++-- drivers/media/pci/saa7164/saa7164-core.c | 4 ++-- drivers/media/pci/tw68/tw68-core.c | 4 ++-- drivers/net/ethernet/amd/pcnet32.c | 5 +++-- 10 files changed, 20 insertions(+), 17 deletions(-) diff --git a/drivers/media/pci/cx23885/cx23885-core.c b/drivers/media/pci/cx23885/cx23885-core.c index 35759a91d47d..e8f847226a19 100644 --- a/drivers/media/pci/cx23885/cx23885-core.c +++ b/drivers/media/pci/cx23885/cx23885-core.c @@ -1992,9 +1992,9 @@ static int cx23885_initdev(struct pci_dev *pci_dev, (unsigned long long)pci_resource_start(pci_dev, 0)); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev, 0xffffffff)) { + err = pci_set_dma_mask(pci_dev, 0xffffffff); + if (err) { printk("%s/0: Oops: no 32bit PCI DMA ???\n", dev->name); - err = -EIO; goto fail_context; } diff --git a/drivers/media/pci/cx25821/cx25821-core.c b/drivers/media/pci/cx25821/cx25821-core.c index dbc695f32760..0042803a9de7 100644 --- a/drivers/media/pci/cx25821/cx25821-core.c +++ b/drivers/media/pci/cx25821/cx25821-core.c @@ -1319,7 +1319,8 @@ static int cx25821_initdev(struct pci_dev *pci_dev, dev->pci_lat, (unsigned long long)dev->base_io_addr); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev, 0xffffffff)) { + err = pci_set_dma_mask(pci_dev, 0xffffffff); + if (err) { pr_err("%s/0: Oops: no 32bit PCI DMA ???\n", dev->name); err = -EIO; goto fail_irq; diff --git a/drivers/media/pci/cx88/cx88-alsa.c b/drivers/media/pci/cx88/cx88-alsa.c index 0ed1b6530374..1b5268f9bb24 100644 --- a/drivers/media/pci/cx88/cx88-alsa.c +++ b/drivers/media/pci/cx88/cx88-alsa.c @@ -890,9 +890,9 @@ static int snd_cx88_create(struct snd_card *card, struct pci_dev *pci, return err; } - if (!pci_set_dma_mask(pci,DMA_BIT_MASK(32))) { + err = pci_set_dma_mask(pci,DMA_BIT_MASK(32)); + if (err) { dprintk(0, "%s/1: Oops: no 32bit PCI DMA ???\n",core->name); - err = -EIO; cx88_core_put(core, pci); return err; } diff --git a/drivers/media/pci/cx88/cx88-mpeg.c b/drivers/media/pci/cx88/cx88-mpeg.c index 9db7767d1fe0..f34c229f9b37 100644 --- a/drivers/media/pci/cx88/cx88-mpeg.c +++ b/drivers/media/pci/cx88/cx88-mpeg.c @@ -393,7 +393,8 @@ static int cx8802_init_common(struct cx8802_dev *dev) if (pci_enable_device(dev->pci)) return -EIO; pci_set_master(dev->pci); - if (!pci_set_dma_mask(dev->pci,DMA_BIT_MASK(32))) { + err = pci_set_dma_mask(dev->pci,DMA_BIT_MASK(32)); + if (err) { printk("%s/2: Oops: no 32bit PCI DMA ???\n",dev->core->name); return -EIO; } diff --git a/drivers/media/pci/cx88/cx88-video.c b/drivers/media/pci/cx88/cx88-video.c index 0de1ad5a977d..aef9acf351f6 100644 --- a/drivers/media/pci/cx88/cx88-video.c +++ b/drivers/media/pci/cx88/cx88-video.c @@ -1314,9 +1314,9 @@ static int cx8800_initdev(struct pci_dev *pci_dev, dev->pci_lat,(unsigned long long)pci_resource_start(pci_dev,0)); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev,DMA_BIT_MASK(32))) { + err = pci_set_dma_mask(pci_dev,DMA_BIT_MASK(32)); + if (err) { printk("%s/0: Oops: no 32bit PCI DMA ???\n",core->name); - err = -EIO; goto fail_core; } dev->alloc_ctx = vb2_dma_sg_init_ctx(&pci_dev->dev); diff --git a/drivers/media/pci/netup_unidvb/netup_unidvb_core.c b/drivers/media/pci/netup_unidvb/netup_unidvb_core.c index 60b2d462f98d..3fdbd81b5580 100644 --- a/drivers/media/pci/netup_unidvb/netup_unidvb_core.c +++ b/drivers/media/pci/netup_unidvb/netup_unidvb_core.c @@ -810,7 +810,7 @@ static int netup_unidvb_initdev(struct pci_dev *pci_dev, "%s(): board vendor 0x%x, revision 0x%x\n", __func__, board_vendor, board_revision); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev, 0xffffffff)) { + if (pci_set_dma_mask(pci_dev, 0xffffffff) < 0) { dev_err(&pci_dev->dev, "%s(): 32bit PCI DMA is not supported\n", __func__); goto pci_detect_err; diff --git a/drivers/media/pci/saa7134/saa7134-core.c b/drivers/media/pci/saa7134/saa7134-core.c index e79d63eb774e..f720cea80e28 100644 --- a/drivers/media/pci/saa7134/saa7134-core.c +++ b/drivers/media/pci/saa7134/saa7134-core.c @@ -951,9 +951,9 @@ static int saa7134_initdev(struct pci_dev *pci_dev, pci_name(pci_dev), dev->pci_rev, pci_dev->irq, dev->pci_lat,(unsigned long long)pci_resource_start(pci_dev,0)); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev, DMA_BIT_MASK(32))) { + err = pci_set_dma_mask(pci_dev, DMA_BIT_MASK(32)); + if (err) { pr_warn("%s: Oops: no 32bit PCI DMA ???\n", dev->name); - err = -EIO; goto fail1; } diff --git a/drivers/media/pci/saa7164/saa7164-core.c b/drivers/media/pci/saa7164/saa7164-core.c index 8f36b48ef733..8bbd092fbe1d 100644 --- a/drivers/media/pci/saa7164/saa7164-core.c +++ b/drivers/media/pci/saa7164/saa7164-core.c @@ -1264,9 +1264,9 @@ static int saa7164_initdev(struct pci_dev *pci_dev, pci_set_master(pci_dev); /* TODO */ - if (!pci_set_dma_mask(pci_dev, 0xffffffff)) { + err = pci_set_dma_mask(pci_dev, 0xffffffff); + if (err) { printk("%s/0: Oops: no 32bit PCI DMA ???\n", dev->name); - err = -EIO; goto fail_irq; } diff --git a/drivers/media/pci/tw68/tw68-core.c b/drivers/media/pci/tw68/tw68-core.c index 8c5655d351d3..4e77618fbb2b 100644 --- a/drivers/media/pci/tw68/tw68-core.c +++ b/drivers/media/pci/tw68/tw68-core.c @@ -257,9 +257,9 @@ static int tw68_initdev(struct pci_dev *pci_dev, dev->name, pci_name(pci_dev), dev->pci_rev, pci_dev->irq, dev->pci_lat, (u64)pci_resource_start(pci_dev, 0)); pci_set_master(pci_dev); - if (!pci_set_dma_mask(pci_dev, DMA_BIT_MASK(32))) { + err = pci_set_dma_mask(pci_dev, DMA_BIT_MASK(32)); + if (err) { pr_info("%s: Oops: no 32bit PCI DMA ???\n", dev->name); - err = -EIO; goto fail1; } diff --git a/drivers/net/ethernet/amd/pcnet32.c b/drivers/net/ethernet/amd/pcnet32.c index e2afabf3a465..7ccebae9cb48 100644 --- a/drivers/net/ethernet/amd/pcnet32.c +++ b/drivers/net/ethernet/amd/pcnet32.c @@ -1500,10 +1500,11 @@ pcnet32_probe_pci(struct pci_dev *pdev, const struct pci_device_id *ent) return -ENODEV; } - if (!pci_set_dma_mask(pdev, PCNET32_DMA_MASK)) { + err = pci_set_dma_mask(pdev, PCNET32_DMA_MASK); + if (err) { if (pcnet32_debug & NETIF_MSG_PROBE) pr_err("architecture does not support 32bit PCI busmaster DMA\n"); - return -ENODEV; + return err; } if (!request_region(ioaddr, PCNET32_TOTAL_SIZE, "pcnet32_probe_pci")) { if (pcnet32_debug & NETIF_MSG_PROBE) From 50e55bf626ad3ebbca45c0c0d03eb1710a139638 Mon Sep 17 00:00:00 2001 From: Yang Shi Date: Fri, 20 Nov 2015 15:57:10 -0800 Subject: [PATCH 303/335] mm/page-writeback.c: initialize m_dirty to avoid compile warning When building kernel with gcc 5.2, the below warning is raised: mm/page-writeback.c: In function 'balance_dirty_pages.isra.10': mm/page-writeback.c:1545:17: warning: 'm_dirty' may be used uninitialized in this function [-Wmaybe-uninitialized] unsigned long m_dirty, m_thresh, m_bg_thresh; The m_dirty{thresh, bg_thresh} are initialized in the block of "if (mdtc)", so if mdts is null, they won't be initialized before being used. Initialize m_dirty to zero, also initialize m_thresh and m_bg_thresh to keep consistency. They are used later by if condition: !mdtc || m_dirty <= dirty_freerun_ceiling(m_thresh, m_bg_thresh) If mdtc is null, dirty_freerun_ceiling will not be called at all, so the initialization will not change any behavior other than just ceasing the compile warning. (akpm: the patch actually reduces .text size by ~20 bytes on gcc-4.x.y) [akpm@linux-foundation.org: add comment] Signed-off-by: Yang Shi Cc: Tejun Heo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page-writeback.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 2c90357c34ea..3e4d65445fa7 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -1542,7 +1542,9 @@ static void balance_dirty_pages(struct address_space *mapping, for (;;) { unsigned long now = jiffies; unsigned long dirty, thresh, bg_thresh; - unsigned long m_dirty, m_thresh, m_bg_thresh; + unsigned long m_dirty = 0; /* stop bogus uninit warnings */ + unsigned long m_thresh = 0; + unsigned long m_bg_thresh = 0; /* * Unstable writes are a feature of certain networked From 1817889e3b2cc1db8abb595712095129ff9156c1 Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Fri, 20 Nov 2015 15:57:13 -0800 Subject: [PATCH 304/335] mm/hugetlbfs: fix bugs in fallocate hole punch of areas with holes Hugh Dickins pointed out problems with the new hugetlbfs fallocate hole punch code. These problems are in the routine remove_inode_hugepages and mostly occur in the case where there are holes in the range of pages to be removed. These holes could be the result of a previous hole punch or simply sparse allocation. The current code could access pages outside the specified range. remove_inode_hugepages handles both hole punch and truncate operations. Page index handling was fixed/cleaned up so that the loop index always matches the page being processed. The code now only makes a single pass through the range of pages as it was determined page faults could not race with truncate. A cond_resched() was added after removing up to PAGEVEC_SIZE pages. Some totally unnecessary code in hugetlbfs_fallocate() that remained from early development was also removed. Tested with fallocate tests submitted here: http://librelist.com/browser//libhugetlbfs/2015/6/25/patch-tests-add-tests-for-fallocate-system-call/ And, some ftruncate tests under development Fixes: b5cec28d36f5 ("hugetlbfs: truncate_hugepages() takes a range of pages") Signed-off-by: Mike Kravetz Acked-by: Hugh Dickins Cc: Dave Hansen Cc: Naoya Horiguchi Cc: Davidlohr Bueso Cc: "Hillf Danton" Cc: [4.3] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hugetlbfs/inode.c | 65 ++++++++++++++++++++++---------------------- 1 file changed, 32 insertions(+), 33 deletions(-) diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 316adb968b65..de4bdfac0cec 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -332,12 +332,17 @@ static void remove_huge_page(struct page *page) * truncation is indicated by end of range being LLONG_MAX * In this case, we first scan the range and release found pages. * After releasing pages, hugetlb_unreserve_pages cleans up region/reserv - * maps and global counts. + * maps and global counts. Page faults can not race with truncation + * in this routine. hugetlb_no_page() prevents page faults in the + * truncated range. It checks i_size before allocation, and again after + * with the page table lock for the page held. The same lock must be + * acquired to unmap a page. * hole punch is indicated if end is not LLONG_MAX * In the hole punch case we scan the range and release found pages. * Only when releasing a page is the associated region/reserv map * deleted. The region/reserv map for ranges without associated - * pages are not modified. + * pages are not modified. Page faults can race with hole punch. + * This is indicated if we find a mapped page. * Note: If the passed end of range value is beyond the end of file, but * not LLONG_MAX this routine still performs a hole punch operation. */ @@ -361,46 +366,37 @@ static void remove_inode_hugepages(struct inode *inode, loff_t lstart, next = start; while (next < end) { /* - * Make sure to never grab more pages that we - * might possibly need. + * Don't grab more pages than the number left in the range. */ if (end - next < lookup_nr) lookup_nr = end - next; /* - * This pagevec_lookup() may return pages past 'end', - * so we must check for page->index > end. + * When no more pages are found, we are done. */ - if (!pagevec_lookup(&pvec, mapping, next, lookup_nr)) { - if (next == start) - break; - next = start; - continue; - } + if (!pagevec_lookup(&pvec, mapping, next, lookup_nr)) + break; for (i = 0; i < pagevec_count(&pvec); ++i) { struct page *page = pvec.pages[i]; u32 hash; + /* + * The page (index) could be beyond end. This is + * only possible in the punch hole case as end is + * max page offset in the truncate case. + */ + next = page->index; + if (next >= end) + break; + hash = hugetlb_fault_mutex_hash(h, current->mm, &pseudo_vma, mapping, next, 0); mutex_lock(&hugetlb_fault_mutex_table[hash]); lock_page(page); - if (page->index >= end) { - unlock_page(page); - mutex_unlock(&hugetlb_fault_mutex_table[hash]); - next = end; /* we are done */ - break; - } - - /* - * If page is mapped, it was faulted in after being - * unmapped. Do nothing in this race case. In the - * normal case page is not mapped. - */ - if (!page_mapped(page)) { + if (likely(!page_mapped(page))) { bool rsv_on_error = !PagePrivate(page); /* * We must free the huge page and remove @@ -421,17 +417,23 @@ static void remove_inode_hugepages(struct inode *inode, loff_t lstart, hugetlb_fix_reserve_counts( inode, rsv_on_error); } + } else { + /* + * If page is mapped, it was faulted in after + * being unmapped. It indicates a race between + * hole punch and page fault. Do nothing in + * this case. Getting here in a truncate + * operation is a bug. + */ + BUG_ON(truncate_op); } - if (page->index > next) - next = page->index; - - ++next; unlock_page(page); - mutex_unlock(&hugetlb_fault_mutex_table[hash]); } + ++next; huge_pagevec_release(&pvec); + cond_resched(); } if (truncate_op) @@ -647,9 +649,6 @@ static long hugetlbfs_fallocate(struct file *file, int mode, loff_t offset, if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size) i_size_write(inode, offset + len); inode->i_ctime = CURRENT_TIME; - spin_lock(&inode->i_lock); - inode->i_private = NULL; - spin_unlock(&inode->i_lock); out: mutex_unlock(&inode->i_mutex); return error; From 928a477102c4fc6739883415b66987207e3502f4 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Fri, 20 Nov 2015 15:57:15 -0800 Subject: [PATCH 305/335] fat: fix fake_offset handling on error path For the root directory, . and .. are faked (using dir_emit_dots()) and ctx->pos is reset from 2 to 0. A corrupted root directory could cause fat_get_entry() to fail, but ->iterate() (fat_readdir()) reports progress to the VFS (with ctx->pos rewound to 0), so any following calls to ->iterate() continue to return the same entries again and again. The result is that userspace will never see the end of the directory, causing e.g. 'ls' to hang in a getdents() loop. [hirofumi@mail.parknet.co.jp: cleanup and make sure to correct fake_offset] Reported-by: Vegard Nossum Tested-by: Vegard Nossum Signed-off-by: Richard Weinberger Signed-off-by: OGAWA Hirofumi Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fat/dir.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 4afc4d9d2e41..8b2127ffb226 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -610,9 +610,9 @@ parse_record: int status = fat_parse_long(inode, &cpos, &bh, &de, &unicode, &nr_slots); if (status < 0) { - ctx->pos = cpos; + bh = NULL; ret = status; - goto out; + goto end_of_dir; } else if (status == PARSE_INVALID) goto record_end; else if (status == PARSE_NOT_LONGNAME) @@ -654,8 +654,9 @@ parse_record: fill_len = short_len; start_filldir: - if (!fake_offset) - ctx->pos = cpos - (nr_slots + 1) * sizeof(struct msdos_dir_entry); + ctx->pos = cpos - (nr_slots + 1) * sizeof(struct msdos_dir_entry); + if (fake_offset && ctx->pos < 2) + ctx->pos = 2; if (!memcmp(de->name, MSDOS_DOT, MSDOS_NAME)) { if (!dir_emit_dot(file, ctx)) @@ -681,14 +682,19 @@ record_end: fake_offset = 0; ctx->pos = cpos; goto get_new; + end_of_dir: - ctx->pos = cpos; + if (fake_offset && cpos < 2) + ctx->pos = 2; + else + ctx->pos = cpos; fill_failed: brelse(bh); if (unicode) __putname(unicode); out: mutex_unlock(&sbi->s_lock); + return ret; } From 459372545c9c0d6f491e280dccc8a54a61b60e56 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Fri, 20 Nov 2015 15:57:18 -0800 Subject: [PATCH 306/335] kasan: fix kmemleak false-positive in kasan_module_alloc() Kmemleak reports the following leak: unreferenced object 0xfffffbfff41ea000 (size 20480): comm "modprobe", pid 65199, jiffies 4298875551 (age 542.568s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [] kmemleak_alloc+0x4e/0xc0 [] __vmalloc_node_range+0x4b8/0x740 [] kasan_module_alloc+0x72/0xc0 [] module_alloc+0x78/0xb0 [] module_alloc_update_bounds+0x14/0x70 [] layout_and_allocate+0x16f4/0x3c90 [] load_module+0x2ff/0x6690 [] SyS_finit_module+0x136/0x170 [] system_call_fastpath+0x16/0x1b [] 0xffffffffffffffff kasan_module_alloc() allocates shadow memory for module and frees it on module unloading. It doesn't store the pointer to allocated shadow memory because it could be calculated from the shadowed address, i.e. kasan_mem_to_shadow(addr). Since kmemleak cannot find pointer to allocated shadow, it thinks that memory leaked. Use kmemleak_ignore() to tell kmemleak that this is not a leak and shadow memory doesn't contain any pointers. Signed-off-by: Andrey Ryabinin Acked-by: Catalin Marinas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/kasan/kasan.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index d41b21bce6a0..bc0a8d8b8f42 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include @@ -444,6 +445,7 @@ int kasan_module_alloc(void *addr, size_t size) if (ret) { find_vm_area(addr)->flags |= VM_KASAN; + kmemleak_ignore(ret); return 0; } From 9d8a765211335cfdad464b90fb19f546af5706ae Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Fri, 20 Nov 2015 15:57:21 -0800 Subject: [PATCH 307/335] kernel/signal.c: unexport sigsuspend() sigsuspend() is nowhere used except in signal.c itself, so we can mark it static do not pollute the global namespace. But this patch is more than a boring cleanup patch, it fixes a real issue on UserModeLinux. UML has a special console driver to display ttys using xterm, or other terminal emulators, on the host side. Vegard reported that sometimes UML is unable to spawn a xterm and he's facing the following warning: WARNING: CPU: 0 PID: 908 at include/linux/thread_info.h:128 sigsuspend+0xab/0xc0() It turned out that this warning makes absolutely no sense as the UML xterm code calls sigsuspend() on the host side, at least it tries. But as the kernel itself offers a sigsuspend() symbol the linker choose this one instead of the glibc wrapper. Interestingly this code used to work since ever but always blocked signals on the wrong side. Some recent kernel change made the WARN_ON() trigger and uncovered the bug. It is a wonderful example of how much works by chance on computers. :-) Fixes: 68f3f16d9ad0f1 ("new helper: sigsuspend()") Signed-off-by: Richard Weinberger Reported-by: Vegard Nossum Tested-by: Vegard Nossum Acked-by: Oleg Nesterov Cc: [3.5+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/signal.h | 1 - kernel/signal.c | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/include/linux/signal.h b/include/linux/signal.h index ab1e0392b5ac..92557bbce7e7 100644 --- a/include/linux/signal.h +++ b/include/linux/signal.h @@ -239,7 +239,6 @@ extern int sigprocmask(int, sigset_t *, sigset_t *); extern void set_current_blocked(sigset_t *); extern void __set_current_blocked(const sigset_t *); extern int show_unhandled_signals; -extern int sigsuspend(sigset_t *); struct sigaction { #ifndef __ARCH_HAS_IRIX_SIGACTION diff --git a/kernel/signal.c b/kernel/signal.c index c0b01fe24bbd..f3f1f7a972fd 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3503,7 +3503,7 @@ SYSCALL_DEFINE0(pause) #endif -int sigsuspend(sigset_t *set) +static int sigsuspend(sigset_t *set) { current->saved_sigmask = current->blocked; set_current_blocked(set); From 7625b3a0007decf2b135cb47ca67abc78a7b1bc1 Mon Sep 17 00:00:00 2001 From: Vitaly Kuznetsov Date: Fri, 20 Nov 2015 15:57:24 -0800 Subject: [PATCH 308/335] kernel/panic.c: turn off locks debug before releasing console lock Commit 08d78658f393 ("panic: release stale console lock to always get the logbuf printed out") introduced an unwanted bad unlock balance report when panic() is called directly and not from OOPS (e.g. from out_of_memory()). The difference is that in case of OOPS we disable locks debug in oops_enter() and on direct panic call nobody does that. Fixes: 08d78658f393 ("panic: release stale console lock to always get the logbuf printed out") Reported-by: kernel test robot Signed-off-by: Vitaly Kuznetsov Cc: HATAYAMA Daisuke Cc: Masami Hiramatsu Cc: Jiri Kosina Cc: Baoquan He Cc: Prarit Bhargava Cc: Xie XiuQi Cc: Seth Jennings Cc: "K. Y. Srinivasan" Cc: Jan Kara Cc: Petr Mladek Cc: Yasuaki Ishimatsu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/panic.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/panic.c b/kernel/panic.c index 4579dbb7ed87..4b150bc0c6c1 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -152,8 +152,11 @@ void panic(const char *fmt, ...) * We may have ended up stopping the CPU holding the lock (in * smp_send_stop()) while still having some valuable data in the console * buffer. Try to acquire the lock then release it regardless of the - * result. The release will also print the buffers out. + * result. The release will also print the buffers out. Locks debug + * should be disabled to avoid reporting bad unlock balance when + * panic() is not being callled from OOPS. */ + debug_locks_off(); console_trylock(); console_unlock(); From 875fa6fbb8d8d5bbf8cc2fa7b340a716f50ee0c7 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 20 Nov 2015 15:57:27 -0800 Subject: [PATCH 309/335] PM/OPP: add entry in MAINTAINERS Add entry for operating performance points into MAINTAINERS file. This will also allow get_maintainers to list OPP stakeholders properly. Signed-off-by: Viresh Kumar Acked-by: Stephen Boyd Cc: Rafael Wysocki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 0f5fa0f978c2..bab1b03a19fd 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7903,6 +7903,18 @@ S: Maintained F: net/openvswitch/ F: include/uapi/linux/openvswitch.h +OPERATING PERFORMANCE POINTS (OPP) +M: Viresh Kumar +M: Nishanth Menon +M: Stephen Boyd +L: linux-pm@vger.kernel.org +S: Maintained +T: git git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm.git +F: drivers/base/power/opp/ +F: include/linux/pm_opp.h +F: Documentation/power/opp.txt +F: Documentation/devicetree/bindings/opp/ + OPL4 DRIVER M: Clemens Ladisch L: alsa-devel@alsa-project.org (moderated for non-subscribers) From 8f1eb48758aacf6c1ffce18179295adbf3bd7640 Mon Sep 17 00:00:00 2001 From: Junxiao Bi Date: Fri, 20 Nov 2015 15:57:30 -0800 Subject: [PATCH 310/335] ocfs2: fix umask ignored issue New created file's mode is not masked with umask, and this makes umask not work for ocfs2 volume. Fixes: 702e5bc ("ocfs2: use generic posix ACL infrastructure") Signed-off-by: Junxiao Bi Cc: Gang He Cc: Mark Fasheh Cc: Joel Becker Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/namei.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 3b48ac25d8a7..a03f6f433075 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -372,6 +372,8 @@ static int ocfs2_mknod(struct inode *dir, mlog_errno(status); goto leave; } + /* update inode->i_mode after mask with "umask". */ + inode->i_mode = mode; handle = ocfs2_start_trans(osb, ocfs2_mknod_credits(osb->sb, S_ISDIR(mode), From 21fa8442799945beaca074cb5bcf7cfe24969d59 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 20 Nov 2015 15:57:32 -0800 Subject: [PATCH 311/335] mm: fix up sparse warning in gfpflags_allow_blocking sparse says: include/linux/gfp.h:274:26: warning: incorrect type in return expression (different base types) include/linux/gfp.h:274:26: expected bool include/linux/gfp.h:274:26: got restricted gfp_t ...add a forced cast to silence the warning. Signed-off-by: Jeff Layton Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/gfp.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/gfp.h b/include/linux/gfp.h index 6523109e136d..8942af0813e3 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -271,7 +271,7 @@ static inline int gfpflags_to_migratetype(const gfp_t gfp_flags) static inline bool gfpflags_allow_blocking(const gfp_t gfp_flags) { - return gfp_flags & __GFP_DIRECT_RECLAIM; + return (bool __force)(gfp_flags & __GFP_DIRECT_RECLAIM); } #ifdef CONFIG_HIGHMEM From a380a3c75529a5c42b78c0d64a46404f8cb0c0d1 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Fri, 20 Nov 2015 15:57:35 -0800 Subject: [PATCH 312/335] slub: create new ___slab_alloc function that can be called with irqs disabled Bulk alloc needs a function like that because it enables interrupts before calling __slab_alloc which promptly disables them again using the expensive local_irq_save(). Signed-off-by: Christoph Lameter Cc: Jesper Dangaard Brouer Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Cc: Alexander Duyck Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 44 +++++++++++++++++++++++++++++--------------- 1 file changed, 29 insertions(+), 15 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index 7cb4bf9ae320..2a952751bb50 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2295,23 +2295,15 @@ static inline void *get_freelist(struct kmem_cache *s, struct page *page) * And if we were unable to get a new slab from the partial slab lists then * we need to allocate a new slab. This is the slowest path since it involves * a call to the page allocator and the setup of a new slab. + * + * Version of __slab_alloc to use when we know that interrupts are + * already disabled (which is the case for bulk allocation). */ -static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, +static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, unsigned long addr, struct kmem_cache_cpu *c) { void *freelist; struct page *page; - unsigned long flags; - - local_irq_save(flags); -#ifdef CONFIG_PREEMPT - /* - * We may have been preempted and rescheduled on a different - * cpu before disabling interrupts. Need to reload cpu area - * pointer. - */ - c = this_cpu_ptr(s->cpu_slab); -#endif page = c->page; if (!page) @@ -2369,7 +2361,6 @@ load_freelist: VM_BUG_ON(!c->page->frozen); c->freelist = get_freepointer(s, freelist); c->tid = next_tid(c->tid); - local_irq_restore(flags); return freelist; new_slab: @@ -2386,7 +2377,6 @@ new_slab: if (unlikely(!freelist)) { slab_out_of_memory(s, gfpflags, node); - local_irq_restore(flags); return NULL; } @@ -2402,10 +2392,34 @@ new_slab: deactivate_slab(s, page, get_freepointer(s, freelist)); c->page = NULL; c->freelist = NULL; - local_irq_restore(flags); return freelist; } +/* + * Another one that disabled interrupt and compensates for possible + * cpu changes by refetching the per cpu area pointer. + */ +static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + unsigned long addr, struct kmem_cache_cpu *c) +{ + void *p; + unsigned long flags; + + local_irq_save(flags); +#ifdef CONFIG_PREEMPT + /* + * We may have been preempted and rescheduled on a different + * cpu before disabling interrupts. Need to reload cpu area + * pointer. + */ + c = this_cpu_ptr(s->cpu_slab); +#endif + + p = ___slab_alloc(s, gfpflags, node, addr, c); + local_irq_restore(flags); + return p; +} + /* * Inlined fastpath so that allocation functions (kmalloc, kmem_cache_alloc) * have the fastpath folded into their functions. So no function call From 87098373e244840e00bd1c93884c1d917411597e Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Fri, 20 Nov 2015 15:57:38 -0800 Subject: [PATCH 313/335] slub: avoid irqoff/on in bulk allocation Use the new function that can do allocation while interrupts are disabled. Avoids irq on/off sequences. Signed-off-by: Christoph Lameter Cc: Jesper Dangaard Brouer Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Cc: Alexander Duyck Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index 2a952751bb50..23f9d8d26422 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2818,30 +2818,23 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void *object = c->freelist; if (unlikely(!object)) { - local_irq_enable(); /* * Invoking slow path likely have side-effect * of re-populating per CPU c->freelist */ - p[i] = __slab_alloc(s, flags, NUMA_NO_NODE, + p[i] = ___slab_alloc(s, flags, NUMA_NO_NODE, _RET_IP_, c); - if (unlikely(!p[i])) { - __kmem_cache_free_bulk(s, i, p); - return false; - } - local_irq_disable(); + if (unlikely(!p[i])) + goto error; + c = this_cpu_ptr(s->cpu_slab); continue; /* goto for-loop */ } /* kmem_cache debug support */ s = slab_pre_alloc_hook(s, flags); - if (unlikely(!s)) { - __kmem_cache_free_bulk(s, i, p); - c->tid = next_tid(c->tid); - local_irq_enable(); - return false; - } + if (unlikely(!s)) + goto error; c->freelist = get_freepointer(s, object); p[i] = object; @@ -2861,6 +2854,11 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, } return true; + +error: + __kmem_cache_free_bulk(s, i, p); + local_irq_enable(); + return false; } EXPORT_SYMBOL(kmem_cache_alloc_bulk); From b4a64718797b84b64a6ddf3d4183c29c2e79ef1d Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:41 -0800 Subject: [PATCH 314/335] slub: mark the dangling ifdef #else of CONFIG_SLUB_DEBUG The #ifdef of CONFIG_SLUB_DEBUG is located very far from the associated #else. For readability mark it with a comment. Signed-off-by: Jesper Dangaard Brouer Acked-by: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Cc: Alexander Duyck Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/slub.c b/mm/slub.c index 23f9d8d26422..a0c1365f6426 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -1204,7 +1204,7 @@ unsigned long kmem_cache_flags(unsigned long object_size, return flags; } -#else +#else /* !CONFIG_SLUB_DEBUG */ static inline void setup_object_debug(struct kmem_cache *s, struct page *page, void *object) {} From ee0c1a65cf95230d5eb3d9de94fd2ead9a428c67 Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Wed, 11 Nov 2015 08:03:54 -0500 Subject: [PATCH 315/335] tty: Fix tty_send_xchar() lock order inversion The correct lock order is atomic_write_lock => termios_rwsem, as established by tty_write() => n_tty_write(). Fixes: c274f6ef1c666 ("tty: Hold termios_rwsem for tcflow(TCIxxx)") Reported-and-Tested-by: Dmitry Vyukov Cc: # v3.18+ Signed-off-by: Peter Hurley Signed-off-by: Greg Kroah-Hartman --- drivers/tty/tty_io.c | 4 ++++ drivers/tty/tty_ioctl.c | 4 ---- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/tty/tty_io.c b/drivers/tty/tty_io.c index 0c41dbcb90b8..bcc8e1e8bb72 100644 --- a/drivers/tty/tty_io.c +++ b/drivers/tty/tty_io.c @@ -1282,18 +1282,22 @@ int tty_send_xchar(struct tty_struct *tty, char ch) int was_stopped = tty->stopped; if (tty->ops->send_xchar) { + down_read(&tty->termios_rwsem); tty->ops->send_xchar(tty, ch); + up_read(&tty->termios_rwsem); return 0; } if (tty_write_lock(tty, 0) < 0) return -ERESTARTSYS; + down_read(&tty->termios_rwsem); if (was_stopped) start_tty(tty); tty->ops->write(tty, &ch, 1); if (was_stopped) stop_tty(tty); + up_read(&tty->termios_rwsem); tty_write_unlock(tty); return 0; } diff --git a/drivers/tty/tty_ioctl.c b/drivers/tty/tty_ioctl.c index 9c5aebfe7053..1445dd39aa62 100644 --- a/drivers/tty/tty_ioctl.c +++ b/drivers/tty/tty_ioctl.c @@ -1147,16 +1147,12 @@ int n_tty_ioctl_helper(struct tty_struct *tty, struct file *file, spin_unlock_irq(&tty->flow_lock); break; case TCIOFF: - down_read(&tty->termios_rwsem); if (STOP_CHAR(tty) != __DISABLED_CHAR) retval = tty_send_xchar(tty, STOP_CHAR(tty)); - up_read(&tty->termios_rwsem); break; case TCION: - down_read(&tty->termios_rwsem); if (START_CHAR(tty) != __DISABLED_CHAR) retval = tty_send_xchar(tty, START_CHAR(tty)); - up_read(&tty->termios_rwsem); break; default: return -EINVAL; From 17a69219713b5063698038151fea9167b5b0c7b4 Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Sun, 8 Nov 2015 07:53:06 -0500 Subject: [PATCH 316/335] tty: Fix direct use of tty buffer work Recent abstraction of tty buffer work introduced api to manage tty input kworker; use it. Fixes: e176058f0de5 ("tty: Abstract tty buffer work") Signed-off-by: Peter Hurley Signed-off-by: Greg Kroah-Hartman --- drivers/tty/tty_ldisc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/tty/tty_ldisc.c b/drivers/tty/tty_ldisc.c index 5af8f1874c1a..629e3c865072 100644 --- a/drivers/tty/tty_ldisc.c +++ b/drivers/tty/tty_ldisc.c @@ -592,7 +592,7 @@ int tty_set_ldisc(struct tty_struct *tty, int ldisc) /* Restart the work queue in case no characters kick it off. Safe if already running */ - schedule_work(&tty->port->buf.work); + tty_buffer_restart_work(tty->port); tty_unlock(tty); return retval; From 83f7fa6c2976a9db9edf3abf7749176ca36fb4cc Mon Sep 17 00:00:00 2001 From: Simon Arlott Date: Sun, 15 Nov 2015 16:14:32 +0000 Subject: [PATCH 317/335] bcm63xx_uart: Use the device name when registering an interrupt Use the device name when registering an interrupt so that multiple ports don't all have the same interrupt name. Signed-off-by: Simon Arlott Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/bcm63xx_uart.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/tty/serial/bcm63xx_uart.c b/drivers/tty/serial/bcm63xx_uart.c index 681e0f3d5e0e..a1c0a89d9c7f 100644 --- a/drivers/tty/serial/bcm63xx_uart.c +++ b/drivers/tty/serial/bcm63xx_uart.c @@ -474,7 +474,7 @@ static int bcm_uart_startup(struct uart_port *port) /* register irq and enable rx interrupts */ ret = request_irq(port->irq, bcm_uart_interrupt, 0, - bcm_uart_type(port), port); + dev_name(port->dev), port); if (ret) return ret; bcm_uart_writel(port, UART_RX_INT_MASK, UART_IR_REG); From bec8bf61dc911ebe04a7997db2c44ed51d8ade9d Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Tue, 20 Oct 2015 09:55:21 -0400 Subject: [PATCH 318/335] serial: fsl_lpuart: Fix earlycon support Earlycon support for Freescale lpuart should only be enabled when console support is enabled. Fixes: 1d59b382f1c4 ("serial: fsl_lpuart: add earlycon support") Acked-by: Stefan Agner Signed-off-by: Peter Hurley Acked-by: Arnd Bergmann Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig index 1aec4404062d..f38beb28e7ae 100644 --- a/drivers/tty/serial/Kconfig +++ b/drivers/tty/serial/Kconfig @@ -1539,7 +1539,6 @@ config SERIAL_FSL_LPUART tristate "Freescale lpuart serial port support" depends on HAS_DMA select SERIAL_CORE - select SERIAL_EARLYCON help Support for the on-chip lpuart on some Freescale SOCs. @@ -1547,6 +1546,7 @@ config SERIAL_FSL_LPUART_CONSOLE bool "Console on Freescale lpuart serial port" depends on SERIAL_FSL_LPUART=y select SERIAL_CORE_CONSOLE + select SERIAL_EARLYCON help If you have enabled the lpuart serial port on the Freescale SoCs, you can make it the console by answering Y to this option. From f356d7a7ddb5ea545e81c84eecfdf1b5ab4647fc Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Sun, 1 Nov 2015 18:32:56 -0800 Subject: [PATCH 319/335] serial: etraxfs-uart: Fix crash MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since commit 7d8c70d8048c ("serial: mctrl-gpio: rename init function"), crisv32 either do not build or crash as follows. Unable to handle kernel NULL pointer dereference Linux 4.3.0-rc7-next-20151101 #1 Sun Nov 1 11:41:28 PST 2015 ... Call Trace: [] show_stack+0x0/0x9e [] printk+0x0/0x2c [] show_registers+0x14a/0x1c2 [] printk+0x0/0x2c [] die_if_kernel+0x7c/0x9e [] do_page_fault+0x32e/0x3e6 [] of_get_property+0x0/0x2c [] of_irq_parse_raw+0x12a/0x376 [] of_get_property+0x0/0x2c [] get_page_from_freelist+0x73e/0x856 [] of_get_property+0x0/0x2c [] d_mmu_refill+0x10a/0x112 [] devm_kmalloc+0x40/0x56 [] add_dr+0xc/0x1c [] devm_add_action+0x2/0x4e [] mctrl_gpio_init_noauto+0x1c/0x76 [] mctrl_gpio_init+0x22/0x110 The function call in the etraxfs-uart driver was not renamed, possibly due to interference with commit 7b9c5162c182 ("serial: etraxfs-uart: use mctrl_gpio helpers for handling modem signals"). Fixes: 7d8c70d8048c ("serial: mctrl-gpio: rename init function") Signed-off-by: Guenter Roeck Acked-by: Uwe Kleine-König Acked-by: Niklas Cassel Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/etraxfs-uart.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/tty/serial/etraxfs-uart.c b/drivers/tty/serial/etraxfs-uart.c index 6813e316e9ff..2f80bc7e44fb 100644 --- a/drivers/tty/serial/etraxfs-uart.c +++ b/drivers/tty/serial/etraxfs-uart.c @@ -894,7 +894,7 @@ static int etraxfs_uart_probe(struct platform_device *pdev) up->regi_ser = of_iomap(np, 0); up->port.dev = &pdev->dev; - up->gpios = mctrl_gpio_init(&pdev->dev, 0); + up->gpios = mctrl_gpio_init_noauto(&pdev->dev, 0); if (IS_ERR(up->gpios)) return PTR_ERR(up->gpios); From 6b2a3d628aa752f0ab825fc6d4d07b09e274d1c1 Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Sun, 8 Nov 2015 08:52:31 -0500 Subject: [PATCH 320/335] tty: audit: Fix audit source MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The data to audit/record is in the 'from' buffer (ie., the input read buffer). Fixes: 72586c6061ab ("n_tty: Fix auditing support for cannonical mode") Cc: stable # 4.1+ Cc: Miloslav Trmač Signed-off-by: Peter Hurley Acked-by: Laura Abbott Signed-off-by: Greg Kroah-Hartman --- drivers/tty/n_tty.c | 2 +- drivers/tty/tty_audit.c | 2 +- include/linux/tty.h | 6 +++--- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/tty/n_tty.c b/drivers/tty/n_tty.c index 13844261cd5f..ed776149261e 100644 --- a/drivers/tty/n_tty.c +++ b/drivers/tty/n_tty.c @@ -169,7 +169,7 @@ static inline int tty_copy_to_user(struct tty_struct *tty, { struct n_tty_data *ldata = tty->disc_data; - tty_audit_add_data(tty, to, n, ldata->icanon); + tty_audit_add_data(tty, from, n, ldata->icanon); return copy_to_user(to, from, n); } diff --git a/drivers/tty/tty_audit.c b/drivers/tty/tty_audit.c index 90ca082935f6..3d245cd3d8e6 100644 --- a/drivers/tty/tty_audit.c +++ b/drivers/tty/tty_audit.c @@ -265,7 +265,7 @@ static struct tty_audit_buf *tty_audit_buf_get(struct tty_struct *tty, * * Audit @data of @size from @tty, if necessary. */ -void tty_audit_add_data(struct tty_struct *tty, unsigned char *data, +void tty_audit_add_data(struct tty_struct *tty, const void *data, size_t size, unsigned icanon) { struct tty_audit_buf *buf; diff --git a/include/linux/tty.h b/include/linux/tty.h index 5b04b0a5375b..5e31f1b99037 100644 --- a/include/linux/tty.h +++ b/include/linux/tty.h @@ -607,7 +607,7 @@ extern void n_tty_inherit_ops(struct tty_ldisc_ops *ops); /* tty_audit.c */ #ifdef CONFIG_AUDIT -extern void tty_audit_add_data(struct tty_struct *tty, unsigned char *data, +extern void tty_audit_add_data(struct tty_struct *tty, const void *data, size_t size, unsigned icanon); extern void tty_audit_exit(void); extern void tty_audit_fork(struct signal_struct *sig); @@ -615,8 +615,8 @@ extern void tty_audit_tiocsti(struct tty_struct *tty, char ch); extern void tty_audit_push(struct tty_struct *tty); extern int tty_audit_push_current(void); #else -static inline void tty_audit_add_data(struct tty_struct *tty, - unsigned char *data, size_t size, unsigned icanon) +static inline void tty_audit_add_data(struct tty_struct *tty, const void *data, + size_t size, unsigned icanon) { } static inline void tty_audit_tiocsti(struct tty_struct *tty, char ch) From fa70045e9df47d2e7a698e5f7a0d21423a58cdaa Mon Sep 17 00:00:00 2001 From: Heikki Krogerus Date: Thu, 12 Nov 2015 15:21:23 +0200 Subject: [PATCH 321/335] serial: 8250_mid: Add missing dependency 8250_mid uses rational_best_approximation() function, so the driver needs to select CONFIG_RATIONAL option. This fixes build error when CONFIG_RATIONAL is not enabled: drivers/built-in.o: In function `mid8250_set_termios': 8250_mid.c:(.text+0x10169a): undefined reference to `rational_best_approximation' Reported-by: Randy Dunlap Signed-off-by: Heikki Krogerus Acked-by: Andy Shevchenko Signed-off-by: Arnd Bergmann Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/8250/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/tty/serial/8250/Kconfig b/drivers/tty/serial/8250/Kconfig index e6f5e12a2d83..6412f1455beb 100644 --- a/drivers/tty/serial/8250/Kconfig +++ b/drivers/tty/serial/8250/Kconfig @@ -373,6 +373,7 @@ config SERIAL_8250_MID depends on SERIAL_8250 && PCI select HSU_DMA if SERIAL_8250_DMA select HSU_DMA_PCI if X86_INTEL_MID + select RATIONAL help Selecting this option will enable handling of the extra features present on the UART found on Intel Medfield SOC and various other From bd63acf9e9a6b48e9c6c15f0f413f85a481cb5ef Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 16 Nov 2015 16:48:09 +0100 Subject: [PATCH 322/335] serial: export fsl8250_handle_irq fsl8250_handle_irq is now used by the of_serial driver, and that fails if it is a loadable module: ERROR: "fsl8250_handle_irq" [drivers/tty/serial/of_serial.ko] undefined! This exports the symbol to avoid randconfig errors. Signed-off-by: Arnd Bergmann Fixes: d43b54d269d2 ("serial: Enable Freescale 16550 workaround on arm") Cc: Scott Wood Signed-off-by: Jeff Mahoney Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/8250/8250_fsl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/tty/serial/8250/8250_fsl.c b/drivers/tty/serial/8250/8250_fsl.c index c0533a57ec53..910bfee5a88b 100644 --- a/drivers/tty/serial/8250/8250_fsl.c +++ b/drivers/tty/serial/8250/8250_fsl.c @@ -60,3 +60,4 @@ int fsl8250_handle_irq(struct uart_port *port) spin_unlock_irqrestore(&up->port.lock, flags); return 1; } +EXPORT_SYMBOL_GPL(fsl8250_handle_irq); From dcbf0d299c00ed4f82ea8d6e359ad88a5182f9b8 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sun, 22 Nov 2015 12:14:14 +0100 Subject: [PATCH 323/335] parisc: Drop unused MADV_xxxK_PAGES flags from asm/mman.h Drop the MADV_xxK_PAGES flags, which were never used and were from a proposed API which was never integrated into the generic Linux kernel code. Cc: stable@vger.kernel.org Signed-off-by: Helge Deller --- arch/parisc/include/uapi/asm/mman.h | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/arch/parisc/include/uapi/asm/mman.h b/arch/parisc/include/uapi/asm/mman.h index 294d251ca7b2..2ae13ce592e8 100644 --- a/arch/parisc/include/uapi/asm/mman.h +++ b/arch/parisc/include/uapi/asm/mman.h @@ -46,16 +46,6 @@ #define MADV_DONTFORK 10 /* don't inherit across fork */ #define MADV_DOFORK 11 /* do inherit across fork */ -/* The range 12-64 is reserved for page size specification. */ -#define MADV_4K_PAGES 12 /* Use 4K pages */ -#define MADV_16K_PAGES 14 /* Use 16K pages */ -#define MADV_64K_PAGES 16 /* Use 64K pages */ -#define MADV_256K_PAGES 18 /* Use 256K pages */ -#define MADV_1M_PAGES 20 /* Use 1 Megabyte pages */ -#define MADV_4M_PAGES 22 /* Use 4 Megabyte pages */ -#define MADV_16M_PAGES 24 /* Use 16 Megabyte pages */ -#define MADV_64M_PAGES 26 /* Use 64 Megabyte pages */ - #define MADV_MERGEABLE 65 /* KSM may merge identical pages */ #define MADV_UNMERGEABLE 66 /* KSM may not merge identical pages */ From 1f25ad26d65b3740f44d6e03edcd34a5f7b58850 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 15:46:52 +0100 Subject: [PATCH 324/335] parisc: Add defines for Huge page support Huge pages on parisc will have the same size as one pmd table, which is on a 64bit kernel 2MB on a kernel with 4K kernel page sizes, and on a 32bit kernel 4MB when used with 4K kernel pages. Since parisc does not physically supports 2MB huge page sizes, emulate it with two consecutive 1MB page sizes instead. Keeping the same huge page size as one pmd will allow us to add transparent huge page support later on. Bit 21 in the pte flags was unused and will now be used to mark a page as huge page (_PAGE_HPAGE_BIT). Signed-off-by: Helge Deller --- arch/parisc/include/asm/page.h | 13 ++++++++++++- arch/parisc/include/asm/pgtable.h | 20 +++++++++++++++++--- 2 files changed, 29 insertions(+), 4 deletions(-) diff --git a/arch/parisc/include/asm/page.h b/arch/parisc/include/asm/page.h index 60d5d174dfe4..80e742a1c162 100644 --- a/arch/parisc/include/asm/page.h +++ b/arch/parisc/include/asm/page.h @@ -145,11 +145,22 @@ extern int npmem_ranges; #endif /* CONFIG_DISCONTIGMEM */ #ifdef CONFIG_HUGETLB_PAGE -#define HPAGE_SHIFT 22 /* 4MB (is this fixed?) */ +#define HPAGE_SHIFT PMD_SHIFT /* fixed for transparent huge pages */ #define HPAGE_SIZE ((1UL) << HPAGE_SHIFT) #define HPAGE_MASK (~(HPAGE_SIZE - 1)) #define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT) + +#if defined(CONFIG_64BIT) && defined(CONFIG_PARISC_PAGE_SIZE_4KB) +# define REAL_HPAGE_SHIFT 20 /* 20 = 1MB */ +# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_1M +#elif !defined(CONFIG_64BIT) && defined(CONFIG_PARISC_PAGE_SIZE_4KB) +# define REAL_HPAGE_SHIFT 22 /* 22 = 4MB */ +# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_4M +#else +# define REAL_HPAGE_SHIFT 24 /* 24 = 16MB */ +# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_16M #endif +#endif /* CONFIG_HUGETLB_PAGE */ #define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT) diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index f93c4a4e6580..638317ae276e 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -167,7 +167,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr) #define _PAGE_NO_CACHE_BIT 24 /* (0x080) Uncached Page (U bit) */ #define _PAGE_ACCESSED_BIT 23 /* (0x100) Software: Page Accessed */ #define _PAGE_PRESENT_BIT 22 /* (0x200) Software: translation valid */ -/* bit 21 was formerly the FLUSH bit but is now unused */ +#define _PAGE_HPAGE_BIT 21 /* (0x400) Software: Huge Page */ #define _PAGE_USER_BIT 20 /* (0x800) Software: User accessible page */ /* N.B. The bits are defined in terms of a 32 bit word above, so the */ @@ -194,6 +194,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr) #define _PAGE_NO_CACHE (1 << xlate_pabit(_PAGE_NO_CACHE_BIT)) #define _PAGE_ACCESSED (1 << xlate_pabit(_PAGE_ACCESSED_BIT)) #define _PAGE_PRESENT (1 << xlate_pabit(_PAGE_PRESENT_BIT)) +#define _PAGE_HUGE (1 << xlate_pabit(_PAGE_HPAGE_BIT)) #define _PAGE_USER (1 << xlate_pabit(_PAGE_USER_BIT)) #define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_READ | _PAGE_WRITE | _PAGE_DIRTY | _PAGE_ACCESSED) @@ -217,7 +218,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr) #define PxD_FLAG_VALID (1 << xlate_pabit(_PxD_VALID_BIT)) #define PxD_FLAG_MASK (0xf) #define PxD_FLAG_SHIFT (4) -#define PxD_VALUE_SHIFT (8) /* (PAGE_SHIFT-PxD_FLAG_SHIFT) */ +#define PxD_VALUE_SHIFT (PFN_PTE_SHIFT-PxD_FLAG_SHIFT) #ifndef __ASSEMBLY__ @@ -362,6 +363,18 @@ static inline pte_t pte_mkyoung(pte_t pte) { pte_val(pte) |= _PAGE_ACCESSED; ret static inline pte_t pte_mkwrite(pte_t pte) { pte_val(pte) |= _PAGE_WRITE; return pte; } static inline pte_t pte_mkspecial(pte_t pte) { return pte; } +/* + * Huge pte definitions. + */ +#ifdef CONFIG_HUGETLB_PAGE +#define pte_huge(pte) (pte_val(pte) & _PAGE_HUGE) +#define pte_mkhuge(pte) (__pte(pte_val(pte) | _PAGE_HUGE)) +#else +#define pte_huge(pte) (0) +#define pte_mkhuge(pte) (pte) +#endif + + /* * Conversion functions: convert a page and protection to a page entry, * and a page entry and page directory to the page they refer to. @@ -410,8 +423,9 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot) /* Find an entry in the second-level page table.. */ #if CONFIG_PGTABLE_LEVELS == 3 +#define pmd_index(addr) (((addr) >> PMD_SHIFT) & (PTRS_PER_PMD - 1)) #define pmd_offset(dir,address) \ -((pmd_t *) pgd_page_vaddr(*(dir)) + (((address)>>PMD_SHIFT) & (PTRS_PER_PMD-1))) +((pmd_t *) pgd_page_vaddr(*(dir)) + pmd_index(address)) #else #define pmd_offset(dir,addr) ((pmd_t *) dir) #endif From 4182d0cdf853fb044b969318289ae9f451f69c86 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 10:50:01 +0100 Subject: [PATCH 325/335] parisc: Initialize the fault vector earlier in the boot process. A fault vector on parisc needs to be 2K aligned. Furthermore the checksum of the fault vector needs to sum up to 0 which is being calculated and written at runtime. Up to now we aligned both PA20 and PA11 fault vectors on the same 4K page in order to easily write the checksum after having mapped the kernel read-only (by mapping this page only as read-write). But when we want to map the kernel text and data on huge pages this makes things harder. So, simplify it by aligning both fault vectors on 2K boundries and write the checksum before we map the page read-only. Signed-off-by: Helge Deller --- arch/parisc/kernel/entry.S | 11 +++-------- arch/parisc/kernel/setup.c | 3 +++ arch/parisc/kernel/traps.c | 35 +++++++++++++++-------------------- 3 files changed, 21 insertions(+), 28 deletions(-) diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index c5ef4081b01d..b2fdc44da0d5 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -646,17 +646,12 @@ /* - * Align fault_vector_20 on 4K boundary so that both - * fault_vector_11 and fault_vector_20 are on the - * same page. This is only necessary as long as we - * write protect the kernel text, which we may stop - * doing once we use large page translations to cover - * the static part of the kernel address space. + * Fault_vectors are architecturally required to be aligned on a 2K + * boundary */ .text - - .align 4096 + .align 2048 ENTRY(fault_vector_20) /* First vector is invalid (0) */ diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index 72a3c658ad7b..f097762d3922 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -377,6 +377,7 @@ arch_initcall(parisc_init); void start_parisc(void) { extern void start_kernel(void); + extern void early_trap_init(void); int ret, cpunum; struct pdc_coproc_cfg coproc_cfg; @@ -397,6 +398,8 @@ void start_parisc(void) panic("must have an fpu to boot linux"); } + early_trap_init(); /* initialize checksum of fault_vector */ + start_kernel(); // not reached } diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index b99b39f1da02..553b09855cfd 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -807,7 +807,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) } -int __init check_ivt(void *iva) +void __init initialize_ivt(const void *iva) { extern u32 os_hpmc_size; extern const u32 os_hpmc[]; @@ -818,8 +818,8 @@ int __init check_ivt(void *iva) u32 *hpmcp; u32 length; - if (strcmp((char *)iva, "cows can fly")) - return -1; + if (strcmp((const char *)iva, "cows can fly")) + panic("IVT invalid"); ivap = (u32 *)iva; @@ -839,28 +839,23 @@ int __init check_ivt(void *iva) check += ivap[i]; ivap[5] = -check; - - return 0; } + +/* early_trap_init() is called before we set up kernel mappings and + * write-protect the kernel */ +void __init early_trap_init(void) +{ + extern const void fault_vector_20; + #ifndef CONFIG_64BIT -extern const void fault_vector_11; + extern const void fault_vector_11; + initialize_ivt(&fault_vector_11); #endif -extern const void fault_vector_20; + + initialize_ivt(&fault_vector_20); +} void __init trap_init(void) { - void *iva; - - if (boot_cpu_data.cpu_type >= pcxu) - iva = (void *) &fault_vector_20; - else -#ifdef CONFIG_64BIT - panic("Can't boot 64-bit OS on PA1.1 processor!"); -#else - iva = (void *) &fault_vector_11; -#endif - - if (check_ivt(iva)) - panic("IVT invalid"); } From 332b42e4eb6e955f3be0bbbf1f272aa943954d98 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 11:17:27 +0100 Subject: [PATCH 326/335] parisc: Increase initial kernel mapping to 32MB on 64bit kernel For the 64bit kernel the initially 16 MB kernel memory might become too small if you build a kernel with many modules built-in and with kernel text and data areas mapped on huge pages. This patch increases the initial mapping to 32MB for 64bit kernels and keeps 16MB for 32bit kernels. Signed-off-by: Helge Deller --- arch/parisc/include/asm/pgtable.h | 6 +++++- arch/parisc/kernel/head.S | 4 ++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index 638317ae276e..d8534f95915a 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -83,7 +83,11 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr) printk("%s:%d: bad pgd %08lx.\n", __FILE__, __LINE__, (unsigned long)pgd_val(e)) /* This is the size of the initially mapped kernel memory */ -#define KERNEL_INITIAL_ORDER 24 /* 0 to 1<<24 = 16MB */ +#ifdef CONFIG_64BIT +#define KERNEL_INITIAL_ORDER 25 /* 1<<25 = 32MB */ +#else +#define KERNEL_INITIAL_ORDER 24 /* 1<<24 = 16MB */ +#endif #define KERNEL_INITIAL_SIZE (1 << KERNEL_INITIAL_ORDER) #if CONFIG_PGTABLE_LEVELS == 3 diff --git a/arch/parisc/kernel/head.S b/arch/parisc/kernel/head.S index e7d64527aff9..75aa0db9f69e 100644 --- a/arch/parisc/kernel/head.S +++ b/arch/parisc/kernel/head.S @@ -69,7 +69,7 @@ $bss_loop: stw,ma %arg2,4(%r1) stw,ma %arg3,4(%r1) - /* Initialize startup VM. Just map first 8/16 MB of memory */ + /* Initialize startup VM. Just map first 16/32 MB of memory */ load32 PA(swapper_pg_dir),%r4 mtctl %r4,%cr24 /* Initialize kernel root pointer */ mtctl %r4,%cr25 /* Initialize user root pointer */ @@ -107,7 +107,7 @@ $bss_loop: /* Now initialize the PTEs themselves. We use RWX for * everything ... it will get remapped correctly later */ ldo 0+_PAGE_KERNEL_RWX(%r0),%r3 /* Hardwired 0 phys addr start */ - ldi (1<<(KERNEL_INITIAL_ORDER-PAGE_SHIFT)),%r11 /* PFN count */ + load32 (1<<(KERNEL_INITIAL_ORDER-PAGE_SHIFT)),%r11 /* PFN count */ load32 PA(pg0),%r1 $pgt_fill_loop: From 337685e556c6f080bf4775950e3b9493852715f8 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Fri, 20 Nov 2015 11:22:32 +0100 Subject: [PATCH 327/335] parisc: Use long branch to do_syscall_trace_exit Use the 22bit instead of the 17bit branch instruction on a 64bit kernel to reach the do_syscall_trace_exit function from the gateway page. A huge page enabled kernel may need the additional branch distance bits. Signed-off-by: Helge Deller --- arch/parisc/kernel/syscall.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/parisc/kernel/syscall.S b/arch/parisc/kernel/syscall.S index 0b8d26d3ba43..3fbd7252a4b2 100644 --- a/arch/parisc/kernel/syscall.S +++ b/arch/parisc/kernel/syscall.S @@ -369,7 +369,7 @@ tracesys_exit: ldo -16(%r30),%r29 /* Reference param save area */ #endif ldo TASK_REGS(%r1),%r26 - bl do_syscall_trace_exit,%r2 + BL do_syscall_trace_exit,%r2 STREG %r28,TASK_PT_GR28(%r1) /* save return value now */ ldo -THREAD_SZ_ALGN-FRAME_SIZE(%r30),%r1 /* get task ptr */ LDREG TI_TASK(%r1), %r1 @@ -390,7 +390,7 @@ tracesys_sigexit: #ifdef CONFIG_64BIT ldo -16(%r30),%r29 /* Reference param save area */ #endif - bl do_syscall_trace_exit,%r2 + BL do_syscall_trace_exit,%r2 ldo TASK_REGS(%r1),%r26 ldil L%syscall_exit_rfi,%r1 From 736d2169338a50c8814efc186b5423aee43b0c68 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sun, 22 Nov 2015 00:07:06 +0100 Subject: [PATCH 328/335] parisc: Add Huge Page and HUGETLBFS support This patch adds huge page support to allow userspace to allocate huge pages and to use hugetlbfs filesystem on 32- and 64-bit Linux kernels. A later patch will add kernel support to map kernel text and data on huge pages. The only requirement is, that the kernel needs to be compiled for a PA8X00 CPU (PA2.0 architecture). Older PA1.X CPUs do not support variable page sizes. 64bit Kernels are compiled for PA2.0 by default. Technically on parisc multiple physical huge pages may be needed to emulate standard 2MB huge pages. Signed-off-by: Helge Deller --- arch/parisc/Kconfig | 3 + arch/parisc/include/asm/hugetlb.h | 85 ++++++++++++++++ arch/parisc/kernel/entry.S | 45 ++++++--- arch/parisc/kernel/setup.c | 11 +- arch/parisc/mm/Makefile | 1 + arch/parisc/mm/hugetlbpage.c | 161 ++++++++++++++++++++++++++++++ 6 files changed, 291 insertions(+), 15 deletions(-) create mode 100644 arch/parisc/include/asm/hugetlb.h create mode 100644 arch/parisc/mm/hugetlbpage.c diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig index c36546959e86..729f89163bc3 100644 --- a/arch/parisc/Kconfig +++ b/arch/parisc/Kconfig @@ -108,6 +108,9 @@ config PGTABLE_LEVELS default 3 if 64BIT && PARISC_PAGE_SIZE_4KB default 2 +config SYS_SUPPORTS_HUGETLBFS + def_bool y if PA20 + source "init/Kconfig" source "kernel/Kconfig.freezer" diff --git a/arch/parisc/include/asm/hugetlb.h b/arch/parisc/include/asm/hugetlb.h new file mode 100644 index 000000000000..7d56a9ccb752 --- /dev/null +++ b/arch/parisc/include/asm/hugetlb.h @@ -0,0 +1,85 @@ +#ifndef _ASM_PARISC64_HUGETLB_H +#define _ASM_PARISC64_HUGETLB_H + +#include +#include + + +void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, pte_t pte); + +pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, + pte_t *ptep); + +static inline int is_hugepage_only_range(struct mm_struct *mm, + unsigned long addr, + unsigned long len) { + return 0; +} + +/* + * If the arch doesn't supply something else, assume that hugepage + * size aligned regions are ok without further preparation. + */ +static inline int prepare_hugepage_range(struct file *file, + unsigned long addr, unsigned long len) +{ + if (len & ~HPAGE_MASK) + return -EINVAL; + if (addr & ~HPAGE_MASK) + return -EINVAL; + return 0; +} + +static inline void hugetlb_free_pgd_range(struct mmu_gather *tlb, + unsigned long addr, unsigned long end, + unsigned long floor, + unsigned long ceiling) +{ + free_pgd_range(tlb, addr, end, floor, ceiling); +} + +static inline void huge_ptep_clear_flush(struct vm_area_struct *vma, + unsigned long addr, pte_t *ptep) +{ +} + +static inline int huge_pte_none(pte_t pte) +{ + return pte_none(pte); +} + +static inline pte_t huge_pte_wrprotect(pte_t pte) +{ + return pte_wrprotect(pte); +} + +static inline void huge_ptep_set_wrprotect(struct mm_struct *mm, + unsigned long addr, pte_t *ptep) +{ + pte_t old_pte = *ptep; + set_huge_pte_at(mm, addr, ptep, pte_wrprotect(old_pte)); +} + +static inline int huge_ptep_set_access_flags(struct vm_area_struct *vma, + unsigned long addr, pte_t *ptep, + pte_t pte, int dirty) +{ + int changed = !pte_same(*ptep, pte); + if (changed) { + set_huge_pte_at(vma->vm_mm, addr, ptep, pte); + flush_tlb_page(vma, addr); + } + return changed; +} + +static inline pte_t huge_ptep_get(pte_t *ptep) +{ + return *ptep; +} + +static inline void arch_clear_hugepage_flags(struct page *page) +{ +} + +#endif /* _ASM_PARISC64_HUGETLB_H */ diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index b2fdc44da0d5..623496c11756 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -502,21 +502,38 @@ STREG \pte,0(\ptp) .endm + /* We have (depending on the page size): + * - 38 to 52-bit Physical Page Number + * - 12 to 26-bit page offset + */ /* bitshift difference between a PFN (based on kernel's PAGE_SIZE) * to a CPU TLB 4k PFN (4k => 12 bits to shift) */ - #define PAGE_ADD_SHIFT (PAGE_SHIFT-12) + #define PAGE_ADD_SHIFT (PAGE_SHIFT-12) + #define PAGE_ADD_HUGE_SHIFT (REAL_HPAGE_SHIFT-12) /* Drop prot bits and convert to page addr for iitlbt and idtlbt */ - .macro convert_for_tlb_insert20 pte + .macro convert_for_tlb_insert20 pte,tmp +#ifdef CONFIG_HUGETLB_PAGE + copy \pte,\tmp + extrd,u \tmp,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\ + 64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte + + depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\ + (63-58)+PAGE_ADD_SHIFT,\pte + extrd,u,*= \tmp,_PAGE_HPAGE_BIT+32,1,%r0 + depdi _HUGE_PAGE_SIZE_ENCODING_DEFAULT,63,\ + (63-58)+PAGE_ADD_HUGE_SHIFT,\pte +#else /* Huge pages disabled */ extrd,u \pte,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\ 64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\ (63-58)+PAGE_ADD_SHIFT,\pte +#endif .endm /* Convert the pte and prot to tlb insertion values. How * this happens is quite subtle, read below */ - .macro make_insert_tlb spc,pte,prot + .macro make_insert_tlb spc,pte,prot,tmp space_to_prot \spc \prot /* create prot id from space */ /* The following is the real subtlety. This is depositing * T <-> _PAGE_REFTRAP @@ -553,7 +570,7 @@ depdi 1,12,1,\prot /* Drop prot bits and convert to page addr for iitlbt and idtlbt */ - convert_for_tlb_insert20 \pte + convert_for_tlb_insert20 \pte \tmp .endm /* Identical macro to make_insert_tlb above, except it @@ -1142,7 +1159,7 @@ dtlb_miss_20w: tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20w update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 idtlbt pte,prot @@ -1168,7 +1185,7 @@ nadtlb_miss_20w: tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20w update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 idtlbt pte,prot @@ -1262,7 +1279,7 @@ dtlb_miss_20: tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20 update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 f_extend pte,t1 @@ -1290,7 +1307,7 @@ nadtlb_miss_20: tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20 update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 f_extend pte,t1 @@ -1399,7 +1416,7 @@ itlb_miss_20w: tlb_lock spc,ptp,pte,t0,t1,itlb_fault update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 iitlbt pte,prot @@ -1423,7 +1440,7 @@ naitlb_miss_20w: tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20w update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 iitlbt pte,prot @@ -1509,7 +1526,7 @@ itlb_miss_20: tlb_lock spc,ptp,pte,t0,t1,itlb_fault update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 f_extend pte,t1 @@ -1529,7 +1546,7 @@ naitlb_miss_20: tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20 update_accessed ptp,pte,t0,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 f_extend pte,t1 @@ -1561,7 +1578,7 @@ dbit_trap_20w: tlb_lock spc,ptp,pte,t0,t1,dbit_fault update_dirty ptp,pte,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 idtlbt pte,prot @@ -1605,7 +1622,7 @@ dbit_trap_20: tlb_lock spc,ptp,pte,t0,t1,dbit_fault update_dirty ptp,pte,t1 - make_insert_tlb spc,pte,prot + make_insert_tlb spc,pte,prot,t1 f_extend pte,t1 diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index f097762d3922..f7ea626e29c9 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -130,7 +130,16 @@ void __init setup_arch(char **cmdline_p) printk(KERN_INFO "The 32-bit Kernel has started...\n"); #endif - printk(KERN_INFO "Default page size is %dKB.\n", (int)(PAGE_SIZE / 1024)); + printk(KERN_INFO "Kernel default page size is %d KB. Huge pages ", + (int)(PAGE_SIZE / 1024)); +#ifdef CONFIG_HUGETLB_PAGE + printk(KERN_CONT "enabled with %d MB physical and %d MB virtual size", + 1 << (REAL_HPAGE_SHIFT - 20), 1 << (HPAGE_SHIFT - 20)); +#else + printk(KERN_CONT "disabled"); +#endif + printk(KERN_CONT ".\n"); + pdc_console_init(); diff --git a/arch/parisc/mm/Makefile b/arch/parisc/mm/Makefile index 758ceefb373a..134393de69d2 100644 --- a/arch/parisc/mm/Makefile +++ b/arch/parisc/mm/Makefile @@ -3,3 +3,4 @@ # obj-y := init.o fault.o ioremap.o +obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o diff --git a/arch/parisc/mm/hugetlbpage.c b/arch/parisc/mm/hugetlbpage.c new file mode 100644 index 000000000000..f6fdc77a72bd --- /dev/null +++ b/arch/parisc/mm/hugetlbpage.c @@ -0,0 +1,161 @@ +/* + * PARISC64 Huge TLB page support. + * + * This parisc implementation is heavily based on the SPARC and x86 code. + * + * Copyright (C) 2015 Helge Deller + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + + +unsigned long +hugetlb_get_unmapped_area(struct file *file, unsigned long addr, + unsigned long len, unsigned long pgoff, unsigned long flags) +{ + struct hstate *h = hstate_file(file); + + if (len & ~huge_page_mask(h)) + return -EINVAL; + if (len > TASK_SIZE) + return -ENOMEM; + + if (flags & MAP_FIXED) + if (prepare_hugepage_range(file, addr, len)) + return -EINVAL; + + if (addr) + addr = ALIGN(addr, huge_page_size(h)); + + /* we need to make sure the colouring is OK */ + return arch_get_unmapped_area(file, addr, len, pgoff, flags); +} + + +pte_t *huge_pte_alloc(struct mm_struct *mm, + unsigned long addr, unsigned long sz) +{ + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + pte_t *pte = NULL; + + /* We must align the address, because our caller will run + * set_huge_pte_at() on whatever we return, which writes out + * all of the sub-ptes for the hugepage range. So we have + * to give it the first such sub-pte. + */ + addr &= HPAGE_MASK; + + pgd = pgd_offset(mm, addr); + pud = pud_alloc(mm, pgd, addr); + if (pud) { + pmd = pmd_alloc(mm, pud, addr); + if (pmd) + pte = pte_alloc_map(mm, NULL, pmd, addr); + } + return pte; +} + +pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr) +{ + pgd_t *pgd; + pud_t *pud; + pmd_t *pmd; + pte_t *pte = NULL; + + addr &= HPAGE_MASK; + + pgd = pgd_offset(mm, addr); + if (!pgd_none(*pgd)) { + pud = pud_offset(pgd, addr); + if (!pud_none(*pud)) { + pmd = pmd_offset(pud, addr); + if (!pmd_none(*pmd)) + pte = pte_offset_map(pmd, addr); + } + } + return pte; +} + +/* Purge data and instruction TLB entries. Must be called holding + * the pa_tlb_lock. The TLB purge instructions are slow on SMP + * machines since the purge must be broadcast to all CPUs. + */ +static inline void purge_tlb_entries_huge(struct mm_struct *mm, unsigned long addr) +{ + int i; + + /* We may use multiple physical huge pages (e.g. 2x1 MB) to emulate + * Linux standard huge pages (e.g. 2 MB) */ + BUILD_BUG_ON(REAL_HPAGE_SHIFT > HPAGE_SHIFT); + + addr &= HPAGE_MASK; + addr |= _HUGE_PAGE_SIZE_ENCODING_DEFAULT; + + for (i = 0; i < (1 << (HPAGE_SHIFT-REAL_HPAGE_SHIFT)); i++) { + mtsp(mm->context, 1); + pdtlb(addr); + if (unlikely(split_tlb)) + pitlb(addr); + addr += (1UL << REAL_HPAGE_SHIFT); + } +} + +void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, pte_t entry) +{ + unsigned long addr_start; + int i; + + addr &= HPAGE_MASK; + addr_start = addr; + + for (i = 0; i < (1 << HUGETLB_PAGE_ORDER); i++) { + /* Directly write pte entry. We could call set_pte_at(mm, addr, ptep, entry) + * instead, but then we get double locking on pa_tlb_lock. */ + *ptep = entry; + ptep++; + + /* Drop the PAGE_SIZE/non-huge tlb entry */ + purge_tlb_entries(mm, addr); + + addr += PAGE_SIZE; + pte_val(entry) += PAGE_SIZE; + } + + purge_tlb_entries_huge(mm, addr_start); +} + + +pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, + pte_t *ptep) +{ + pte_t entry; + + entry = *ptep; + set_huge_pte_at(mm, addr, ptep, __pte(0)); + + return entry; +} + +int pmd_huge(pmd_t pmd) +{ + return 0; +} + +int pud_huge(pud_t pud) +{ + return 0; +} From 41b85a1163386f8328ad570f383973cb3975d2fa Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sun, 22 Nov 2015 00:07:44 +0100 Subject: [PATCH 329/335] parisc: Map kernel text and data on huge pages Adjust the linker script and map_pages() to map kernel text and data on physical 1MB huge/large pages. Signed-off-by: Helge Deller --- arch/parisc/kernel/asm-offsets.c | 8 +++++++ arch/parisc/kernel/vmlinux.lds.S | 9 ++++--- arch/parisc/mm/init.c | 40 ++++++++++++++------------------ 3 files changed, 31 insertions(+), 26 deletions(-) diff --git a/arch/parisc/kernel/asm-offsets.c b/arch/parisc/kernel/asm-offsets.c index 59001cea13f9..d2f62570a7b1 100644 --- a/arch/parisc/kernel/asm-offsets.c +++ b/arch/parisc/kernel/asm-offsets.c @@ -289,6 +289,14 @@ int main(void) DEFINE(ASM_PTE_ENTRY_SIZE, PTE_ENTRY_SIZE); DEFINE(ASM_PFN_PTE_SHIFT, PFN_PTE_SHIFT); DEFINE(ASM_PT_INITIAL, PT_INITIAL); + BLANK(); + /* HUGEPAGE_SIZE is only used in vmlinux.lds.S to align kernel text + * and kernel data on physical huge pages */ +#ifdef CONFIG_HUGETLB_PAGE + DEFINE(HUGEPAGE_SIZE, 1UL << REAL_HPAGE_SHIFT); +#else + DEFINE(HUGEPAGE_SIZE, PAGE_SIZE); +#endif BLANK(); DEFINE(EXCDATA_IP, offsetof(struct exception_data, fault_ip)); DEFINE(EXCDATA_SPACE, offsetof(struct exception_data, fault_space)); diff --git a/arch/parisc/kernel/vmlinux.lds.S b/arch/parisc/kernel/vmlinux.lds.S index 0dacc5ca555a..308f29081d46 100644 --- a/arch/parisc/kernel/vmlinux.lds.S +++ b/arch/parisc/kernel/vmlinux.lds.S @@ -60,7 +60,7 @@ SECTIONS EXIT_DATA } PERCPU_SECTION(8) - . = ALIGN(PAGE_SIZE); + . = ALIGN(HUGEPAGE_SIZE); __init_end = .; /* freed after init ends here */ @@ -116,7 +116,7 @@ SECTIONS * that we can properly leave these * as writable */ - . = ALIGN(PAGE_SIZE); + . = ALIGN(HUGEPAGE_SIZE); data_start = .; EXCEPTION_TABLE(8) @@ -135,8 +135,11 @@ SECTIONS _edata = .; /* BSS */ - BSS_SECTION(PAGE_SIZE, PAGE_SIZE, 8) + BSS_SECTION(PAGE_SIZE, PAGE_SIZE, PAGE_SIZE) + /* bootmap is allocated in setup_bootmem() directly behind bss. */ + + . = ALIGN(HUGEPAGE_SIZE); _end = . ; STABS_DEBUG diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index c229427fa546..ac90df1119bd 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -407,15 +407,11 @@ static void __init map_pages(unsigned long start_vaddr, unsigned long vaddr; unsigned long ro_start; unsigned long ro_end; - unsigned long fv_addr; - unsigned long gw_addr; - extern const unsigned long fault_vector_20; - extern void * const linux_gateway_page; + unsigned long kernel_end; ro_start = __pa((unsigned long)_text); ro_end = __pa((unsigned long)&data_start); - fv_addr = __pa((unsigned long)&fault_vector_20) & PAGE_MASK; - gw_addr = __pa((unsigned long)&linux_gateway_page) & PAGE_MASK; + kernel_end = __pa((unsigned long)&_end); end_paddr = start_paddr + size; @@ -473,24 +469,25 @@ static void __init map_pages(unsigned long start_vaddr, for (tmp2 = start_pte; tmp2 < PTRS_PER_PTE; tmp2++, pg_table++) { pte_t pte; - /* - * Map the fault vector writable so we can - * write the HPMC checksum. - */ if (force) pte = __mk_pte(address, pgprot); - else if (parisc_text_address(vaddr) && - address != fv_addr) + else if (parisc_text_address(vaddr)) { pte = __mk_pte(address, PAGE_KERNEL_EXEC); + if (address >= ro_start && address < kernel_end) + pte = pte_mkhuge(pte); + } else #if defined(CONFIG_PARISC_PAGE_SIZE_4KB) - if (address >= ro_start && address < ro_end - && address != fv_addr - && address != gw_addr) - pte = __mk_pte(address, PAGE_KERNEL_RO); - else + if (address >= ro_start && address < ro_end) { + pte = __mk_pte(address, PAGE_KERNEL_EXEC); + pte = pte_mkhuge(pte); + } else #endif + { pte = __mk_pte(address, pgprot); + if (address >= ro_start && address < kernel_end) + pte = pte_mkhuge(pte); + } if (address >= end_paddr) { if (force) @@ -534,15 +531,12 @@ void free_initmem(void) /* force the kernel to see the new TLB entries */ __flush_tlb_range(0, init_begin, init_end); - /* Attempt to catch anyone trying to execute code here - * by filling the page with BRK insns. - */ - memset((void *)init_begin, 0x00, init_end - init_begin); + /* finally dump all the instructions which were cached, since the * pages are no-longer executable */ flush_icache_range(init_begin, init_end); - free_initmem_default(-1); + free_initmem_default(POISON_FREE_INITMEM); /* set up a new led state on systems shipped LED State panel */ pdc_chassis_send_status(PDC_CHASSIS_DIRECT_BCOMPLETE); @@ -712,8 +706,8 @@ static void __init pagetable_init(void) unsigned long size; start_paddr = pmem_ranges[range].start_pfn << PAGE_SHIFT; - end_paddr = start_paddr + (pmem_ranges[range].pages << PAGE_SHIFT); size = pmem_ranges[range].pages << PAGE_SHIFT; + end_paddr = start_paddr + size; map_pages((unsigned long)__va(start_paddr), start_paddr, size, PAGE_KERNEL, 0); From 81084651d73737988355f167065fab8a73574db1 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:46 -0800 Subject: [PATCH 330/335] slub: support for bulk free with SLUB freelists Make it possible to free a freelist with several objects by adjusting API of slab_free() and __slab_free() to have head, tail and an objects counter (cnt). Tail being NULL indicate single object free of head object. This allow compiler inline constant propagation in slab_free() and slab_free_freelist_hook() to avoid adding any overhead in case of single object free. This allows a freelist with several objects (all within the same slab-page) to be free'ed using a single locked cmpxchg_double in __slab_free() and with an unlocked cmpxchg_double in slab_free(). Object debugging on the free path is also extended to handle these freelists. When CONFIG_SLUB_DEBUG is enabled it will also detect if objects don't belong to the same slab-page. These changes are needed for the next patch to bulk free the detached freelists it introduces and constructs. Micro benchmarking showed no performance reduction due to this change, when debugging is turned off (compiled with CONFIG_SLUB_DEBUG). Signed-off-by: Jesper Dangaard Brouer Signed-off-by: Alexander Duyck Acked-by: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 85 +++++++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 67 insertions(+), 18 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index a0c1365f6426..d52f0d0ab712 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -1065,11 +1065,15 @@ bad: return 0; } +/* Supports checking bulk free of a constructed freelist */ static noinline struct kmem_cache_node *free_debug_processing( - struct kmem_cache *s, struct page *page, void *object, + struct kmem_cache *s, struct page *page, + void *head, void *tail, int bulk_cnt, unsigned long addr, unsigned long *flags) { struct kmem_cache_node *n = get_node(s, page_to_nid(page)); + void *object = head; + int cnt = 0; spin_lock_irqsave(&n->list_lock, *flags); slab_lock(page); @@ -1077,6 +1081,9 @@ static noinline struct kmem_cache_node *free_debug_processing( if (!check_slab(s, page)) goto fail; +next_object: + cnt++; + if (!check_valid_pointer(s, page, object)) { slab_err(s, page, "Invalid object pointer 0x%p", object); goto fail; @@ -1107,8 +1114,19 @@ static noinline struct kmem_cache_node *free_debug_processing( if (s->flags & SLAB_STORE_USER) set_track(s, object, TRACK_FREE, addr); trace(s, page, object, 0); + /* Freepointer not overwritten by init_object(), SLAB_POISON moved it */ init_object(s, object, SLUB_RED_INACTIVE); + + /* Reached end of constructed freelist yet? */ + if (object != tail) { + object = get_freepointer(s, object); + goto next_object; + } out: + if (cnt != bulk_cnt) + slab_err(s, page, "Bulk freelist count(%d) invalid(%d)\n", + bulk_cnt, cnt); + slab_unlock(page); /* * Keep node_lock to preserve integrity @@ -1212,7 +1230,8 @@ static inline int alloc_debug_processing(struct kmem_cache *s, struct page *page, void *object, unsigned long addr) { return 0; } static inline struct kmem_cache_node *free_debug_processing( - struct kmem_cache *s, struct page *page, void *object, + struct kmem_cache *s, struct page *page, + void *head, void *tail, int bulk_cnt, unsigned long addr, unsigned long *flags) { return NULL; } static inline int slab_pad_check(struct kmem_cache *s, struct page *page) @@ -1308,6 +1327,29 @@ static inline void slab_free_hook(struct kmem_cache *s, void *x) kasan_slab_free(s, x); } +static inline void slab_free_freelist_hook(struct kmem_cache *s, + void *head, void *tail) +{ +/* + * Compiler cannot detect this function can be removed if slab_free_hook() + * evaluates to nothing. Thus, catch all relevant config debug options here. + */ +#if defined(CONFIG_KMEMCHECK) || \ + defined(CONFIG_LOCKDEP) || \ + defined(CONFIG_DEBUG_KMEMLEAK) || \ + defined(CONFIG_DEBUG_OBJECTS_FREE) || \ + defined(CONFIG_KASAN) + + void *object = head; + void *tail_obj = tail ? : head; + + do { + slab_free_hook(s, object); + } while ((object != tail_obj) && + (object = get_freepointer(s, object))); +#endif +} + static void setup_object(struct kmem_cache *s, struct page *page, void *object) { @@ -2583,10 +2625,11 @@ EXPORT_SYMBOL(kmem_cache_alloc_node_trace); * handling required then we can return immediately. */ static void __slab_free(struct kmem_cache *s, struct page *page, - void *x, unsigned long addr) + void *head, void *tail, int cnt, + unsigned long addr) + { void *prior; - void **object = (void *)x; int was_frozen; struct page new; unsigned long counters; @@ -2596,7 +2639,8 @@ static void __slab_free(struct kmem_cache *s, struct page *page, stat(s, FREE_SLOWPATH); if (kmem_cache_debug(s) && - !(n = free_debug_processing(s, page, x, addr, &flags))) + !(n = free_debug_processing(s, page, head, tail, cnt, + addr, &flags))) return; do { @@ -2606,10 +2650,10 @@ static void __slab_free(struct kmem_cache *s, struct page *page, } prior = page->freelist; counters = page->counters; - set_freepointer(s, object, prior); + set_freepointer(s, tail, prior); new.counters = counters; was_frozen = new.frozen; - new.inuse--; + new.inuse -= cnt; if ((!new.inuse || !prior) && !was_frozen) { if (kmem_cache_has_cpu_partial(s) && !prior) { @@ -2640,7 +2684,7 @@ static void __slab_free(struct kmem_cache *s, struct page *page, } while (!cmpxchg_double_slab(s, page, prior, counters, - object, new.counters, + head, new.counters, "__slab_free")); if (likely(!n)) { @@ -2705,15 +2749,20 @@ slab_empty: * * If fastpath is not possible then fall back to __slab_free where we deal * with all sorts of special processing. + * + * Bulk free of a freelist with several objects (all pointing to the + * same page) possible by specifying head and tail ptr, plus objects + * count (cnt). Bulk free indicated by tail pointer being set. */ -static __always_inline void slab_free(struct kmem_cache *s, - struct page *page, void *x, unsigned long addr) +static __always_inline void slab_free(struct kmem_cache *s, struct page *page, + void *head, void *tail, int cnt, + unsigned long addr) { - void **object = (void *)x; + void *tail_obj = tail ? : head; struct kmem_cache_cpu *c; unsigned long tid; - slab_free_hook(s, x); + slab_free_freelist_hook(s, head, tail); redo: /* @@ -2732,19 +2781,19 @@ redo: barrier(); if (likely(page == c->page)) { - set_freepointer(s, object, c->freelist); + set_freepointer(s, tail_obj, c->freelist); if (unlikely(!this_cpu_cmpxchg_double( s->cpu_slab->freelist, s->cpu_slab->tid, c->freelist, tid, - object, next_tid(tid)))) { + head, next_tid(tid)))) { note_cmpxchg_failure("slab_free", s, tid); goto redo; } stat(s, FREE_FASTPATH); } else - __slab_free(s, page, x, addr); + __slab_free(s, page, head, tail_obj, cnt, addr); } @@ -2753,7 +2802,7 @@ void kmem_cache_free(struct kmem_cache *s, void *x) s = cache_from_obj(s, x); if (!s) return; - slab_free(s, virt_to_head_page(x), x, _RET_IP_); + slab_free(s, virt_to_head_page(x), x, NULL, 1, _RET_IP_); trace_kmem_cache_free(_RET_IP_, x); } EXPORT_SYMBOL(kmem_cache_free); @@ -2788,7 +2837,7 @@ void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p) c->tid = next_tid(c->tid); local_irq_enable(); /* Slowpath: overhead locked cmpxchg_double_slab */ - __slab_free(s, page, object, _RET_IP_); + __slab_free(s, page, object, object, 1, _RET_IP_); local_irq_disable(); c = this_cpu_ptr(s->cpu_slab); } @@ -3523,7 +3572,7 @@ void kfree(const void *x) __free_kmem_pages(page, compound_order(page)); return; } - slab_free(page->slab_cache, page, object, _RET_IP_); + slab_free(page->slab_cache, page, object, NULL, 1, _RET_IP_); } EXPORT_SYMBOL(kfree); From d0ecd894e3d5f768a84403b34019c4a7daa05882 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:49 -0800 Subject: [PATCH 331/335] slub: optimize bulk slowpath free by detached freelist This change focus on improving the speed of object freeing in the "slowpath" of kmem_cache_free_bulk. The calls slab_free (fastpath) and __slab_free (slowpath) have been extended with support for bulk free, which amortize the overhead of the (locked) cmpxchg_double. To use the new bulking feature, we build what I call a detached freelist. The detached freelist takes advantage of three properties: 1) the free function call owns the object that is about to be freed, thus writing into this memory is synchronization-free. 2) many freelist's can co-exist side-by-side in the same slab-page each with a separate head pointer. 3) it is the visibility of the head pointer that needs synchronization. Given these properties, the brilliant part is that the detached freelist can be constructed without any need for synchronization. The freelist is constructed directly in the page objects, without any synchronization needed. The detached freelist is allocated on the stack of the function call kmem_cache_free_bulk. Thus, the freelist head pointer is not visible to other CPUs. All objects in a SLUB freelist must belong to the same slab-page. Thus, constructing the detached freelist is about matching objects that belong to the same slab-page. The bulk free array is scanned is a progressive manor with a limited look-ahead facility. Kmem debug support is handled in call of slab_free(). Notice kmem_cache_free_bulk no longer need to disable IRQs. This only slowed down single free bulk with approx 3 cycles. Performance data: Benchmarked[1] obj size 256 bytes on CPU i7-4790K @ 4.00GHz SLUB fastpath single object quick reuse: 47 cycles(tsc) 11.931 ns To get stable and comparable numbers, the kernel have been booted with "slab_merge" (this also improve performance for larger bulk sizes). Performance data, compared against fallback bulking: bulk - fallback bulk - improvement with this patch 1 - 62 cycles(tsc) 15.662 ns - 49 cycles(tsc) 12.407 ns- improved 21.0% 2 - 55 cycles(tsc) 13.935 ns - 30 cycles(tsc) 7.506 ns - improved 45.5% 3 - 53 cycles(tsc) 13.341 ns - 23 cycles(tsc) 5.865 ns - improved 56.6% 4 - 52 cycles(tsc) 13.081 ns - 20 cycles(tsc) 5.048 ns - improved 61.5% 8 - 50 cycles(tsc) 12.627 ns - 18 cycles(tsc) 4.659 ns - improved 64.0% 16 - 49 cycles(tsc) 12.412 ns - 17 cycles(tsc) 4.495 ns - improved 65.3% 30 - 49 cycles(tsc) 12.484 ns - 18 cycles(tsc) 4.533 ns - improved 63.3% 32 - 50 cycles(tsc) 12.627 ns - 18 cycles(tsc) 4.707 ns - improved 64.0% 34 - 96 cycles(tsc) 24.243 ns - 23 cycles(tsc) 5.976 ns - improved 76.0% 48 - 83 cycles(tsc) 20.818 ns - 21 cycles(tsc) 5.329 ns - improved 74.7% 64 - 74 cycles(tsc) 18.700 ns - 20 cycles(tsc) 5.127 ns - improved 73.0% 128 - 90 cycles(tsc) 22.734 ns - 27 cycles(tsc) 6.833 ns - improved 70.0% 158 - 99 cycles(tsc) 24.776 ns - 30 cycles(tsc) 7.583 ns - improved 69.7% 250 - 104 cycles(tsc) 26.089 ns - 37 cycles(tsc) 9.280 ns - improved 64.4% Performance data, compared current in-kernel bulking: bulk - curr in-kernel - improvement with this patch 1 - 46 cycles(tsc) - 49 cycles(tsc) - improved (cycles:-3) -6.5% 2 - 27 cycles(tsc) - 30 cycles(tsc) - improved (cycles:-3) -11.1% 3 - 21 cycles(tsc) - 23 cycles(tsc) - improved (cycles:-2) -9.5% 4 - 18 cycles(tsc) - 20 cycles(tsc) - improved (cycles:-2) -11.1% 8 - 17 cycles(tsc) - 18 cycles(tsc) - improved (cycles:-1) -5.9% 16 - 18 cycles(tsc) - 17 cycles(tsc) - improved (cycles: 1) 5.6% 30 - 18 cycles(tsc) - 18 cycles(tsc) - improved (cycles: 0) 0.0% 32 - 18 cycles(tsc) - 18 cycles(tsc) - improved (cycles: 0) 0.0% 34 - 78 cycles(tsc) - 23 cycles(tsc) - improved (cycles:55) 70.5% 48 - 60 cycles(tsc) - 21 cycles(tsc) - improved (cycles:39) 65.0% 64 - 49 cycles(tsc) - 20 cycles(tsc) - improved (cycles:29) 59.2% 128 - 69 cycles(tsc) - 27 cycles(tsc) - improved (cycles:42) 60.9% 158 - 79 cycles(tsc) - 30 cycles(tsc) - improved (cycles:49) 62.0% 250 - 86 cycles(tsc) - 37 cycles(tsc) - improved (cycles:49) 57.0% Performance with normal SLUB merging is significantly slower for larger bulking. This is believed to (primarily) be an effect of not having to share the per-CPU data-structures, as tuning per-CPU size can achieve similar performance. bulk - slab_nomerge - normal SLUB merge 1 - 49 cycles(tsc) - 49 cycles(tsc) - merge slower with cycles:0 2 - 30 cycles(tsc) - 30 cycles(tsc) - merge slower with cycles:0 3 - 23 cycles(tsc) - 23 cycles(tsc) - merge slower with cycles:0 4 - 20 cycles(tsc) - 20 cycles(tsc) - merge slower with cycles:0 8 - 18 cycles(tsc) - 18 cycles(tsc) - merge slower with cycles:0 16 - 17 cycles(tsc) - 17 cycles(tsc) - merge slower with cycles:0 30 - 18 cycles(tsc) - 23 cycles(tsc) - merge slower with cycles:5 32 - 18 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:4 34 - 23 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:-1 48 - 21 cycles(tsc) - 22 cycles(tsc) - merge slower with cycles:1 64 - 20 cycles(tsc) - 48 cycles(tsc) - merge slower with cycles:28 128 - 27 cycles(tsc) - 57 cycles(tsc) - merge slower with cycles:30 158 - 30 cycles(tsc) - 59 cycles(tsc) - merge slower with cycles:29 250 - 37 cycles(tsc) - 56 cycles(tsc) - merge slower with cycles:19 Joint work with Alexander Duyck. [1] https://github.com/netoptimizer/prototype-kernel/blob/master/kernel/mm/slab_bulk_test01.c [akpm@linux-foundation.org: BUG_ON -> WARN_ON;return] Signed-off-by: Jesper Dangaard Brouer Signed-off-by: Alexander Duyck Acked-by: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 113 ++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 81 insertions(+), 32 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index d52f0d0ab712..c17c5202864d 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2807,44 +2807,93 @@ void kmem_cache_free(struct kmem_cache *s, void *x) } EXPORT_SYMBOL(kmem_cache_free); +struct detached_freelist { + struct page *page; + void *tail; + void *freelist; + int cnt; +}; + +/* + * This function progressively scans the array with free objects (with + * a limited look ahead) and extract objects belonging to the same + * page. It builds a detached freelist directly within the given + * page/objects. This can happen without any need for + * synchronization, because the objects are owned by running process. + * The freelist is build up as a single linked list in the objects. + * The idea is, that this detached freelist can then be bulk + * transferred to the real freelist(s), but only requiring a single + * synchronization primitive. Look ahead in the array is limited due + * to performance reasons. + */ +static int build_detached_freelist(struct kmem_cache *s, size_t size, + void **p, struct detached_freelist *df) +{ + size_t first_skipped_index = 0; + int lookahead = 3; + void *object; + + /* Always re-init detached_freelist */ + df->page = NULL; + + do { + object = p[--size]; + } while (!object && size); + + if (!object) + return 0; + + /* Start new detached freelist */ + set_freepointer(s, object, NULL); + df->page = virt_to_head_page(object); + df->tail = object; + df->freelist = object; + p[size] = NULL; /* mark object processed */ + df->cnt = 1; + + while (size) { + object = p[--size]; + if (!object) + continue; /* Skip processed objects */ + + /* df->page is always set at this point */ + if (df->page == virt_to_head_page(object)) { + /* Opportunity build freelist */ + set_freepointer(s, object, df->freelist); + df->freelist = object; + df->cnt++; + p[size] = NULL; /* mark object processed */ + + continue; + } + + /* Limit look ahead search */ + if (!--lookahead) + break; + + if (!first_skipped_index) + first_skipped_index = size + 1; + } + + return first_skipped_index; +} + + /* Note that interrupts must be enabled when calling this function. */ void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p) { - struct kmem_cache_cpu *c; - struct page *page; - int i; + if (WARN_ON(!size)) + return; - local_irq_disable(); - c = this_cpu_ptr(s->cpu_slab); + do { + struct detached_freelist df; - for (i = 0; i < size; i++) { - void *object = p[i]; + size = build_detached_freelist(s, size, p, &df); + if (unlikely(!df.page)) + continue; - BUG_ON(!object); - /* kmem cache debug support */ - s = cache_from_obj(s, object); - if (unlikely(!s)) - goto exit; - slab_free_hook(s, object); - - page = virt_to_head_page(object); - - if (c->page == page) { - /* Fastpath: local CPU free */ - set_freepointer(s, object, c->freelist); - c->freelist = object; - } else { - c->tid = next_tid(c->tid); - local_irq_enable(); - /* Slowpath: overhead locked cmpxchg_double_slab */ - __slab_free(s, page, object, object, 1, _RET_IP_); - local_irq_disable(); - c = this_cpu_ptr(s->cpu_slab); - } - } -exit: - c->tid = next_tid(c->tid); - local_irq_enable(); + slab_free(s, df.page, df.freelist, df.tail, df.cnt, _RET_IP_); + } while (likely(size)); } EXPORT_SYMBOL(kmem_cache_free_bulk); From 03ec0ed57ffc77720b811dbb6d44733b58360d9f Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:52 -0800 Subject: [PATCH 332/335] slub: fix kmem cgroup bug in kmem_cache_alloc_bulk The call slab_pre_alloc_hook() interacts with kmemgc and is not allowed to be called several times inside the bulk alloc for loop, due to the call to memcg_kmem_get_cache(). This would result in hitting the VM_BUG_ON in __memcg_kmem_get_cache. As suggested by Vladimir Davydov, change slab_post_alloc_hook() to be able to handle an array of objects. A subtle detail is, loop iterator "i" in slab_post_alloc_hook() must have same type (size_t) as size argument. This helps the compiler to easier realize that it can remove the loop, when all debug statements inside loop evaluates to nothing. Note, this is only an issue because the kernel is compiled with GCC option: -fno-strict-overflow In slab_alloc_node() the compiler inlines and optimizes the invocation of slab_post_alloc_hook(s, flags, 1, &object) by removing the loop and access object directly. Signed-off-by: Jesper Dangaard Brouer Reported-by: Vladimir Davydov Suggested-by: Vladimir Davydov Reviewed-by: Vladimir Davydov Cc: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index c17c5202864d..ce1797623391 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -1292,14 +1292,21 @@ static inline struct kmem_cache *slab_pre_alloc_hook(struct kmem_cache *s, return memcg_kmem_get_cache(s, flags); } -static inline void slab_post_alloc_hook(struct kmem_cache *s, - gfp_t flags, void *object) +static inline void slab_post_alloc_hook(struct kmem_cache *s, gfp_t flags, + size_t size, void **p) { + size_t i; + flags &= gfp_allowed_mask; - kmemcheck_slab_alloc(s, flags, object, slab_ksize(s)); - kmemleak_alloc_recursive(object, s->object_size, 1, s->flags, flags); + for (i = 0; i < size; i++) { + void *object = p[i]; + + kmemcheck_slab_alloc(s, flags, object, slab_ksize(s)); + kmemleak_alloc_recursive(object, s->object_size, 1, + s->flags, flags); + kasan_slab_alloc(s, object); + } memcg_kmem_put_cache(s); - kasan_slab_alloc(s, object); } static inline void slab_free_hook(struct kmem_cache *s, void *x) @@ -2475,7 +2482,7 @@ static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, static __always_inline void *slab_alloc_node(struct kmem_cache *s, gfp_t gfpflags, int node, unsigned long addr) { - void **object; + void *object; struct kmem_cache_cpu *c; struct page *page; unsigned long tid; @@ -2554,7 +2561,7 @@ redo: if (unlikely(gfpflags & __GFP_ZERO) && object) memset(object, 0, s->object_size); - slab_post_alloc_hook(s, gfpflags, object); + slab_post_alloc_hook(s, gfpflags, 1, &object); return object; } @@ -2904,6 +2911,10 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, struct kmem_cache_cpu *c; int i; + /* memcg and kmem_cache debug support */ + s = slab_pre_alloc_hook(s, flags); + if (unlikely(!s)) + return false; /* * Drain objects in the per cpu slab, while disabling local * IRQs, which protects against PREEMPT and interrupts @@ -2928,17 +2939,8 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, c = this_cpu_ptr(s->cpu_slab); continue; /* goto for-loop */ } - - /* kmem_cache debug support */ - s = slab_pre_alloc_hook(s, flags); - if (unlikely(!s)) - goto error; - c->freelist = get_freepointer(s, object); p[i] = object; - - /* kmem_cache debug support */ - slab_post_alloc_hook(s, flags, object); } c->tid = next_tid(c->tid); local_irq_enable(); @@ -2951,11 +2953,13 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, memset(p[j], 0, s->object_size); } + /* memcg and kmem_cache debug support */ + slab_post_alloc_hook(s, flags, size, p); return true; - error: - __kmem_cache_free_bulk(s, i, p); local_irq_enable(); + slab_post_alloc_hook(s, flags, i, p); + __kmem_cache_free_bulk(s, i, p); return false; } EXPORT_SYMBOL(kmem_cache_alloc_bulk); From 033745189b1bae3fc931beeaf48604ee7c259309 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:55 -0800 Subject: [PATCH 333/335] slub: add missing kmem cgroup support to kmem_cache_free_bulk Initial implementation missed support for kmem cgroup support in kmem_cache_free_bulk() call, add this. If CONFIG_MEMCG_KMEM is not enabled, the compiler should be smart enough to not add any asm code. Incoming bulk free objects can belong to different kmem cgroups, and object free call can happen at a later point outside memcg context. Thus, we need to keep the orig kmem_cache, to correctly verify if a memcg object match against its "root_cache" (s->memcg_params.root_cache). Signed-off-by: Jesper Dangaard Brouer Reviewed-by: Vladimir Davydov Cc: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/mm/slub.c b/mm/slub.c index ce1797623391..34847044dfe5 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2887,13 +2887,17 @@ static int build_detached_freelist(struct kmem_cache *s, size_t size, /* Note that interrupts must be enabled when calling this function. */ -void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p) +void kmem_cache_free_bulk(struct kmem_cache *orig_s, size_t size, void **p) { if (WARN_ON(!size)) return; do { struct detached_freelist df; + struct kmem_cache *s; + + /* Support for memcg */ + s = cache_from_obj(orig_s, p[size - 1]); size = build_detached_freelist(s, size, p, &df); if (unlikely(!df.page)) From 865762a8119e74b5f0e236d2d8eaaf8be9292a06 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 20 Nov 2015 15:57:58 -0800 Subject: [PATCH 334/335] slab/slub: adjust kmem_cache_alloc_bulk API Adjust kmem_cache_alloc_bulk API before we have any real users. Adjust API to return type 'int' instead of previously type 'bool'. This is done to allow future extension of the bulk alloc API. A future extension could be to allow SLUB to stop at a page boundary, when specified by a flag, and then return the number of objects. The advantage of this approach, would make it easier to make bulk alloc run without local IRQs disabled. With an approach of cmpxchg "stealing" the entire c->freelist or page->freelist. To avoid overshooting we would stop processing at a slab-page boundary. Else we always end up returning some objects at the cost of another cmpxchg. To keep compatible with future users of this API linking against an older kernel when using the new flag, we need to return the number of allocated objects with this API change. Signed-off-by: Jesper Dangaard Brouer Cc: Vladimir Davydov Acked-by: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/slab.h | 2 +- mm/slab.c | 2 +- mm/slab.h | 2 +- mm/slab_common.c | 6 +++--- mm/slob.c | 2 +- mm/slub.c | 8 ++++---- 6 files changed, 11 insertions(+), 11 deletions(-) diff --git a/include/linux/slab.h b/include/linux/slab.h index 96940772bb92..2037a861e367 100644 --- a/include/linux/slab.h +++ b/include/linux/slab.h @@ -316,7 +316,7 @@ void kmem_cache_free(struct kmem_cache *, void *); * Note that interrupts must be enabled when calling these functions. */ void kmem_cache_free_bulk(struct kmem_cache *, size_t, void **); -bool kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **); +int kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **); #ifdef CONFIG_NUMA void *__kmalloc_node(size_t size, gfp_t flags, int node) __assume_kmalloc_alignment; diff --git a/mm/slab.c b/mm/slab.c index e0819fa96559..4765c97ce690 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -3419,7 +3419,7 @@ void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p) } EXPORT_SYMBOL(kmem_cache_free_bulk); -bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, +int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void **p) { return __kmem_cache_alloc_bulk(s, flags, size, p); diff --git a/mm/slab.h b/mm/slab.h index 27492eb678f7..7b6087197997 100644 --- a/mm/slab.h +++ b/mm/slab.h @@ -170,7 +170,7 @@ ssize_t slabinfo_write(struct file *file, const char __user *buffer, * may be allocated or freed using these operations. */ void __kmem_cache_free_bulk(struct kmem_cache *, size_t, void **); -bool __kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **); +int __kmem_cache_alloc_bulk(struct kmem_cache *, gfp_t, size_t, void **); #ifdef CONFIG_MEMCG_KMEM /* diff --git a/mm/slab_common.c b/mm/slab_common.c index d88e97c10a2e..3c6a86b4ec25 100644 --- a/mm/slab_common.c +++ b/mm/slab_common.c @@ -112,7 +112,7 @@ void __kmem_cache_free_bulk(struct kmem_cache *s, size_t nr, void **p) kmem_cache_free(s, p[i]); } -bool __kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t nr, +int __kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t nr, void **p) { size_t i; @@ -121,10 +121,10 @@ bool __kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t nr, void *x = p[i] = kmem_cache_alloc(s, flags); if (!x) { __kmem_cache_free_bulk(s, i, p); - return false; + return 0; } } - return true; + return i; } #ifdef CONFIG_MEMCG_KMEM diff --git a/mm/slob.c b/mm/slob.c index 0d7e5df74d1f..17e8f8cc7c53 100644 --- a/mm/slob.c +++ b/mm/slob.c @@ -617,7 +617,7 @@ void kmem_cache_free_bulk(struct kmem_cache *s, size_t size, void **p) } EXPORT_SYMBOL(kmem_cache_free_bulk); -bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, +int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, void **p) { return __kmem_cache_alloc_bulk(s, flags, size, p); diff --git a/mm/slub.c b/mm/slub.c index 34847044dfe5..46997517406e 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2909,8 +2909,8 @@ void kmem_cache_free_bulk(struct kmem_cache *orig_s, size_t size, void **p) EXPORT_SYMBOL(kmem_cache_free_bulk); /* Note that interrupts must be enabled when calling this function. */ -bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, - void **p) +int kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, + void **p) { struct kmem_cache_cpu *c; int i; @@ -2959,12 +2959,12 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *s, gfp_t flags, size_t size, /* memcg and kmem_cache debug support */ slab_post_alloc_hook(s, flags, size, p); - return true; + return i; error: local_irq_enable(); slab_post_alloc_hook(s, flags, i, p); __kmem_cache_free_bulk(s, i, p); - return false; + return 0; } EXPORT_SYMBOL(kmem_cache_alloc_bulk); From 1ec218373b8ebda821aec00bb156a9c94fad9cd4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 22 Nov 2015 16:45:59 -0800 Subject: [PATCH 335/335] Linux 4.4-rc2 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 3a0234f50f36..2ffdf9d6f339 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 4 PATCHLEVEL = 4 SUBLEVEL = 0 -EXTRAVERSION = -rc1 +EXTRAVERSION = -rc2 NAME = Blurry Fish Butt # *DOCUMENTATION*