cfq_queue->ref is used with queue_lock hold, so ref doesn't need to be an atomic
and atomic operation is slower.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The "error" field in block_bio_complete is not assigned, leaving the memory area
uninitialized (keeping garbage data). Pass an additional tracepoint argument to
this event to initialize this field.
Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
CC: Steven Rostedt <rostedt@goodmis.org>
CC: Frederic Weisbecker <fweisbec@gmail.com>
CC: Ingo Molnar <mingo@elte.hu>
CC: Thomas Gleixner <tglx@linutronix.de>
CC: Li Zefan <lizf@cn.fujitsu.com>
CC: Alan.Brunelle@hp.com
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
We can't use krefs since it's apparently restricted to very basic
reference counting.
This reverts commit e4a683c8.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
/proc/diskstats would display a strange output as follows.
$ cat /proc/diskstats |grep sda
8 0 sda 90524 7579 102154 20464 0 0 0 0 0 14096 20089
8 1 sda1 19085 1352 21841 4209 0 0 0 0 4294967064 15689 4293424691
~~~~~~~~~~
8 2 sda2 71252 3624 74891 15950 0 0 0 0 232 23995 1562390
8 3 sda3 54 487 2188 92 0 0 0 0 0 88 92
8 4 sda4 4 0 8 0 0 0 0 0 0 0 0
8 5 sda5 81 2027 2130 138 0 0 0 0 0 87 137
Its reason is the wrong way of accounting hd_struct->in_flight. When a bio is
merged into a request belongs to different partition by ELEVATOR_FRONT_MERGE.
The detailed root cause is as follows.
Assuming that there are two partition, sda1 and sda2.
1. A request for sda2 is in request_queue. Hence sda1's hd_struct->in_flight
is 0 and sda2's one is 1.
| hd_struct->in_flight
---------------------------
sda1 | 0
sda2 | 1
---------------------------
2. A bio belongs to sda1 is issued and is merged into the request mentioned on
step1 by ELEVATOR_BACK_MERGE. The first sector of the request is changed
from sda2 region to sda1 region. However the two partition's
hd_struct->in_flight are not changed.
| hd_struct->in_flight
---------------------------
sda1 | 0
sda2 | 1
---------------------------
3. The request is finished and blk_account_io_done() is called. In this case,
sda2's hd_struct->in_flight, not a sda1's one, is decremented.
| hd_struct->in_flight
---------------------------
sda1 | -1
sda2 | 1
---------------------------
The patch fixes the problem by caching the partition lookup
inside the request structure, hence making sure that the increment
and decrement will always happen on the same partition struct. This
also speeds up IO with accounting enabled, since it cuts down on
the number of lookups we have to do.
Also add a refcount to struct hd_struct to keep the partition in
memory as long as users exist. We use kref_test_and_get() to ensure
we don't add a reference to a partition which is going away.
Signed-off-by: Jerome Marchand <jmarchan@redhat.com>
Signed-off-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Add kref_test_and_get() function, which atomically add a reference only if
refcount is not zero. This prevent to add a reference to an object that is
already being removed.
Signed-off-by: Jerome Marchand <jmarchan@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Work items processed by kintegrityd_wq won't block much, may burn a
lot of CPU cycles and affect IO latency. Use alloc_workqueue() to
mark it highpri and CPU intensive with max concurrency of 1.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
kblockd is used for unplugging and may affect IO latency and
throughput and the max number of concurrent work items are bound by
the number of block devices. Make it HIGHPRI workqueue w/ default max
concurrency.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
This patch fixes a spelling error in a source code comment and removes
superfluous braces in the function exit_io_context().
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Cc: Jens Axboe <jaxboe@fusionio.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Commit a8adbe3 forgot to remove the return variable, kill it.
drivers/block/loop.c: In function 'lo_splice_actor':
drivers/block/loop.c:398: warning: unused variable 'ret'
[...]
fs/nfsd/vfs.c: In function 'nfsd_splice_actor':
fs/nfsd/vfs.c:848: warning: unused variable 'ret'
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The major/minor device numbers are always defined and used as `unsigned'.
Signed-off-by: Yang Zhang <kthreadd@gmail.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
When cfq_choose_cfqg() is called in select_queue(), there must be at least one
backlogged CFQ queue waiting for dispatching, hence there must be at least one
backlogged CFQ group on service tree. So we never call choose_service_tree()
with cfqg == NULL.
Signed-off-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
This patch pulls calls to buf->ops->confirm() from all actors passed
(also indirectly) to splice_from_pipe_feed().
Is avoiding the call to buf->ops->confirm() while splice()ing to
/dev/null is an intentional optimization? No other user does that
and this will remove this special case.
Against current linux.git 6313e3c217.
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
If priority is changed, continuing to check workload_expires and service tree
count of the previous workload does not make sense. We should always choose
the workload with lowest key of new priority in such case.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
It's able to check whether a CFQ group on a service tree by
checking "cfqg->rb_node". There's no need to maintain an
extra flag here.
Signed-off-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
When a cfq group is running, it won't be dequeued from service tree, so
there's no need to store the active one in st->active. Just gid rid of it.
Signed-off-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The addition of CONFIG_SECURITY_DMESG_RESTRICT resulted in a build
failure when CONFIG_PRINTK=n. This is because the capabilities code
which used the new option was built even though the variable in question
didn't exist.
The patch here fixes this by moving the capabilities checks out of the
LSM and into the caller. All (known) LSMs should have been calling the
capabilities hook already so it actually makes the code organization
better to eliminate the hook altogether.
Signed-off-by: Eric Paris <eparis@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6:
arm: omap1: devices: need to return with a value
OMAP1: camera.h: add missing include
omap: dma: Add read-back to DMA interrupt handler to avoid spuriousinterrupts
OMAP2: Devkit8000: Fix mmc regulator failure
* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
hwmon: (w83795) Check for BEEP pin availability
hwmon: (w83795) Clear intrusion alarm immediately
hwmon: (w83795) Read the intrusion state properly
hwmon: (w83795) Print the actual temperature channels as sources
hwmon: (w83795) List all usable temperature sources
hwmon: (w83795) Expose fan control method
hwmon: (w83795) Fix fan control mode attributes
hwmon: (lm95241) Check validity of input values
hwmon: Change mail address of Hans J. Koch
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jbarnes/pci-2.6:
PCI: sysfs: fix printk warnings
PCI: fix pci_bus_alloc_resource() hang, prefer positive decode
PCI: read current power state at enable time
PCI: fix size checks for mmap() on /proc/bus/pci files
x86/PCI: coalesce overlapping host bridge windows
PCI hotplug: ibmphp: Add check to prevent reading beyond mapped area
Make sure I2C adapters being registered have the required struct
fields set. If they don't, problems will happen later.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
It's about time to make it clear that i2c_adapter.id is deprecated.
Hopefully this will remind the last user to move over to a different
strategy.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Jarod Wilson <jarod@redhat.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
Drivers don't need to include <linux/i2c-id.h>, especially not when
they don't use anything that header file provides.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Michael Hunold <michael@mihu.de>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Delete unused I2C adapter IDs. Special cases are:
* I2C_HW_B_RIVA was still set in driver rivafb, however no other
driver is ever looking for this value, so we can safely remove it.
* I2C_HW_B_HDPVR is used in staging driver lirc_zilog, however no
adapter ID is ever set to this value, so the code in question never
runs. As the code additionally expects that I2C_HW_B_HDPVR may not
be defined, we can delete it now and let the lirc_zilog driver
maintainer rewrite this piece of code.
Big thanks for Hans Verkuil for doing all the hard work :)
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Jarod Wilson <jarod@redhat.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
A few new i2c-drivers came into the kernel which clear the clientdata-pointer
on exit. This is obsolete meanwhile, so fix it and hope the word will spread.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Alan Cox <alan@linux.intel.com>
Acked-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Move the logging bits from kernel.h into printk.h so that
there is a bit more logical separation of the generic from
the printk logging specific parts.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The fix in commit 6b4e81db25 ("i8k: Tell gcc that *regs gets
clobbered") to work around the gcc miscompiling i8k.c to add "+m
(*regs)" caused register pressure problems and a build failure.
Changing the 'asm' statement to 'asm volatile' instead should prevent
that and works around the gcc bug as well, so we can remove the "+m".
[ Background on the gcc bug: a memory clobber fails to mark the function
the asm resides in as non-pure (aka "__attribute__((const))"), so if
the function does nothing else that triggers the non-pure logic, gcc
will think that that function has no side effects at all. As a result,
callers will be mis-compiled.
Adding the "+m" made gcc see that it's not a pure function, and so
does "asm volatile". The problem was never really the need to mark
"*regs" as changed, since the memory clobber did that part - the
problem was just a bug in the gcc "pure" function analysis - Linus ]
Signed-off-by: Jim Bos <jim876@xs4all.nl>
Acked-by: Jakub Jelinek <jakub@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On the W83795ADG, there's a single pin for BEEP and OVT#, so you
can't have both. Check the configuration and don't create beep
attributes when BEEP pin is not available.
The W83795G has a dedicated BEEP pin so the functionality is always
available there.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
When asked to clear the intrusion alarm, do so immediately. We have to
invalidate the cache to make sure the new status will be read. But we
also have to read from the status register once to clear the pending
alarm, as writing to CLR_CHS surprising won't clear it automatically.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
We can't read the intrusion state from the real-time alarm registers
as we do for all other alarm flags, because real-time alarm bits don't
stick (by definition) and the intrusion state has to stick until
explicitly cleared (otherwise it has little value.)
So we have to use the interrupt status register instead, which is read
from the same address but with a configuration bit flipped in another
register.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Don't expose raw register values to user-space. Decode and encode
temperature channels selected as temperature sources as needed.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Temperature sources are not correlated directly with temperature
channels. A look-up table is required to find out which temperature
sources can be used depending on which temperature channels (both
analog and digital) are enabled.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
Expose fan control method (DC vs. PWM) using the standard sysfs
attributes. I've made it read-only as the board should be wired for
a given mode, the BIOS should have set up the chip for this mode, and
you shouldn't have to change it. But it would be easy enough to make
it changeable if someone comes up with a use case.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
There were two bugs:
* Speed cruise mode was improperly reported for all fans but fan1.
* Fan control method (PWM vs. DC) was mixed with the control mode.
It will be added back as a separate attribute, as per the standard
sysfs interface.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
This clears the following build-time warnings I was seeing:
drivers/hwmon/lm95241.c: In function "set_interval":
drivers/hwmon/lm95241.c:132:15: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_max2":
drivers/hwmon/lm95241.c:278:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_max1":
drivers/hwmon/lm95241.c:277:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_min2":
drivers/hwmon/lm95241.c:249:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_min1":
drivers/hwmon/lm95241.c:248:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_type2":
drivers/hwmon/lm95241.c:220:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
drivers/hwmon/lm95241.c: In function "set_type1":
drivers/hwmon/lm95241.c:219:1: warning: ignoring return value of "strict_strtol", declared with attribute warn_unused_result
This also fixes a small race in set_interval() as a side effect: by
working with a temporary local variable we prevent data->interval from
being accessed at a time it contains the interval value in the wrong
unit.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Davide Rizzo <elpa.rizzo@gmail.com>
My old mail address doesn't exist anymore. This changes all occurrences
to my new address.
Signed-off-by: Hans J. Koch <hjk@hansjkoch.de>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
o Allow hierarchical cgroup creation for blkio controller
o Currently we disallow it as both the io controller policies (throttling
as well as proportion bandwidth) do not support hierarhical accounting
and control. But the flip side is that blkio controller can not be used with
libvirt as libvirt creates a cgroup hierarchy deeper than 1 level.
<top-level-cgroup-dir>/<controller>/libvirt/qemu/<virtual-machine-groups>
o So this patch will allow creation of cgroup hierarhcy but at the backend
everything will be treated as flat. So if somebody created a an hierarchy
like as follows.
root
/ \
test1 test2
|
test3
CFQ and throttling will practically treat all groups at same level.
pivot
/ | \ \
root test1 test2 test3
o Once we have actual support for hierarchical accounting and control
then we can introduce another cgroup tunable file "blkio.use_hierarchy"
which will be 0 by default but if user wants to enforce hierarhical
control then it can be set to 1. This way there should not be any
ABI problems down the line.
o The only not so pretty part is introduction of extra file "use_hierarchy"
down the line. Kame-san had mentioned that hierarhical accounting is
expensive in memory controller hence they keep it off by default. I
suspect same will be the case for IO controller also as for each IO
completion we shall have to account IO through hierarchy up to the root.
if yes, then it probably is not a very bad idea to introduce this extra
file so that it will be used only when somebody needs it and some people
might enable hierarchy only in part of the hierarchy.
o This is how basically memory controller also uses "use_hierarhcy" and
they also allowed creation of hierarchies when actual backend support
was not available.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Reviewed-by: Gui Jianfeng <guijianfeng@cn.fujitsu.com>
Reviewed-by: Ciju Rajan K <ciju@linux.vnet.ibm.com>
Tested-by: Ciju Rajan K <ciju@linux.vnet.ibm.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Cast pci_resource_start() and pci_resource_len() to u64 for printk.
drivers/pci/pci-sysfs.c:753: warning: format '%16Lx' expects type 'long long unsigned int', but argument 9 has type 'resource_size_t'
drivers/pci/pci-sysfs.c:753: warning: format '%16Lx' expects type 'long long unsigned int', but argument 10 has type 'resource_size_t'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
* 'fbdev-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6:
fsl-diu-fb: drop dead ioctl define
MAINTAINERS: Add an fbdev git tree entry.
OMAP: DSS: Fix documentation regarding 'vram' kernel parameter
OMAP: VRAM: Fix boot-time memory allocation
OMAP: VRAM: improve VRAM error prints
sisfb: limit POST memory test according to PCI resource length
fbdev: sh_mobile_lcdc: use correct number of modes, when using the default
fbdev: sh_mobile_lcdc: use the standard CEA-861 720p timing
fbdev: sh_mobile_hdmi: properly clean up modedb on monitor unplug
* 'sh-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
sh: intc: Fix up build failure introduced by radix tree changes.
MAINTAINERS: update the sh git tree entry.
sh: clkfwk: fix up compiler warnings.
sh: intc: Fix up initializers for gcc 4.5.
rtc: rtc-sh - fix a memory leak
* 'rmobile-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
ARM: mach-shmobile: ap4evb: add fsib 44100Hz rate
MAINTAINERS: update the ARM SH-Mobile git tree entry.
ARM: mach-shmobile: ap4evb: Mark NOR boot loader partitions read-only.
ARM: mach-shmobile: intc-sh7372: fix interrupt number
This area of the code has always been a bit delicate due to the
subtleties of lock ordering. The problem is that for "normal"
alloc/dealloc, we always grab the inode locks first and the rgrp lock
later.
In order to ensure no races in looking up the unlinked, but still
allocated inodes, we need to hold the rgrp lock when we do the lookup,
which means that we can't take the inode glock.
The solution is to borrow the technique already used by NFS to solve
what is essentially the same problem (given an inode number, look up
the inode carefully, checking that it really is in the expected
state).
We cannot do that directly from the allocation code (lock ordering
again) so we give the job to the pre-existing delete workqueue and
carry on with the allocation as normal.
If we find there is no space, we do a journal flush (required anyway
if space from a deallocation is to be released) which should block
against the pending deallocations, so we should always get the space
back.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Shaohua Li
Jeff Moyer
Jens Axboe
Jerome Marchand
Tejun Heo
Bart Van Assche
Yang Zhang
Gui Jianfeng
Michał Mirosław
Mike Snitzer
Linus Torvalds
Eric Paris
Jean Delvare
Wolfram Sang
Jim Bos
Hans J. Koch
Vivek Goyal
Randy Dunlap
Steven Whitehouse
Paul Mundt