linux/drivers
Alexey Kardashevskiy a68bd1267b powerpc/powernv/ioda: Allocate indirect TCE levels on demand
At the moment we allocate the entire TCE table, twice (hardware part and
userspace translation cache). This normally works as we normally have
contigous memory and the guest will map entire RAM for 64bit DMA.

However if we have sparse RAM (one example is a memory device), then
we will allocate TCEs which will never be used as the guest only maps
actual memory for DMA. If it is a single level TCE table, there is nothing
we can really do but if it a multilevel table, we can skip allocating
TCEs we know we won't need.

This adds ability to allocate only first level, saving memory.

This changes iommu_table::free() to avoid allocating of an extra level;
iommu_table::set() will do this when needed.

This adds @alloc parameter to iommu_table::exchange() to tell the callback
if it can allocate an extra level; the flag is set to "false" for
the realmode KVM handlers of H_PUT_TCE hcalls and the callback returns
H_TOO_HARD.

This still requires the entire table to be counted in mm::locked_vm.

To be conservative, this only does on-demand allocation when
the usespace cache table is requested which is the case of VFIO.

The example math for a system replicating a powernv setup with NVLink2
in a guest:
16GB RAM mapped at 0x0
128GB GPU RAM window (16GB of actual RAM) mapped at 0x244000000000

the table to cover that all with 64K pages takes:
(((0x244000000000 + 0x2000000000) >> 16)*8)>>20 = 4556MB

If we allocate only necessary TCE levels, we will only need:
(((0x400000000 + 0x400000000) >> 16)*8)>>20 = 4MB (plus some for indirect
levels).

Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
2018-07-16 22:53:11 +10:00
..
accessibility
acpi pwm: Changes for v4.18-rc1 2018-06-14 16:25:43 +09:00
amba
android treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
ata treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
atm treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
auxdisplay treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
base Additional power management updates for 4.18-rc1 2018-06-13 07:24:18 -07:00
bcma
block The main piece is a set of libceph changes that revamps how OSD 2018-06-15 07:24:58 +09:00
bluetooth
bus - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
cdrom treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
char docs: Fix some broken references 2018-06-15 18:10:01 -03:00
clk docs: Fix some broken references 2018-06-15 18:10:01 -03:00
clocksource treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
connector Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2018-06-06 18:39:49 -07:00
cpufreq Additional power management updates for 4.18-rc1 2018-06-13 07:24:18 -07:00
cpuidle powerpc updates for 4.18 2018-06-07 10:23:33 -07:00
crypto treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
dax libnvdimm for 4.18 2018-06-08 17:21:52 -07:00
dca
devfreq treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
dio
dma fix a series of Documentation/ broken file name references 2018-06-15 18:10:01 -03:00
dma-buf
edac treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
eisa
extcon treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
firewire treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
firmware Merge branch 'akpm' (patches from Andrew) 2018-06-15 08:51:42 +09:00
fmc treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
fpga
fsi
gpio treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
gpu Solve a series of broken links for files under Documentation: 2018-06-17 05:25:18 +09:00
hid docs: fix broken references with multiple hints 2018-06-15 18:10:01 -03:00
hsi
hv treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
hwmon treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
hwspinlock hwspinlock updates for v4.18 2018-06-11 12:09:19 -07:00
hwtracing treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
i2c Merge branch 'i2c/for-4.18' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux 2018-06-14 16:21:46 +09:00
ide treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
idle
iio treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
infiniband treewide: Use array_size() in kvzalloc_node() 2018-06-12 16:19:22 -07:00
input docs: Fix some broken references 2018-06-15 18:10:01 -03:00
iommu - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
ipack treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
irqchip treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
isdn treewide: Use array_size() in vmalloc() 2018-06-12 16:19:22 -07:00
leds treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
lightnvm docs: Fix some broken references 2018-06-15 18:10:01 -03:00
macintosh powerpc updates for 4.18 2018-06-07 10:23:33 -07:00
mailbox treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
mcb
md docs: Fix some broken references 2018-06-15 18:10:01 -03:00
media Solve a series of broken links for files under Documentation: 2018-06-17 05:25:18 +09:00
memory - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
memstick treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
message treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
mfd Merge branch 'i2c/for-4.18' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux 2018-06-14 16:21:46 +09:00
misc powerpc/64s: Remove POWER9 DD1 support 2018-07-16 11:37:21 +10:00
mmc treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
mtd - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
mux
net Solve a series of broken links for files under Documentation: 2018-06-17 05:25:18 +09:00
nfc treewide: devm_kmalloc() -> devm_kmalloc_array() 2018-06-12 16:19:22 -07:00
ntb - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
nubus
nvdimm Merge branch 'for-4.18/mcsafe' into libnvdimm-for-next 2018-06-08 15:16:44 -07:00
nvme Merge branch 'nvme-4.18' of git://git.infradead.org/nvme into for-linus 2018-06-15 08:11:05 -06:00
nvmem treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
of - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
opp treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
oprofile treewide: Use array_size() in vmalloc() 2018-06-12 16:19:22 -07:00
parisc
parport docs: Fix some broken references 2018-06-15 18:10:01 -03:00
pci - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
pcmcia treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
perf
phy Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2018-06-06 18:39:49 -07:00
pinctrl treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
platform fix a series of Documentation/ broken file name references 2018-06-15 18:10:01 -03:00
pnp media updates for v4.18-rc1 2018-06-07 12:34:37 -07:00
power treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
powercap treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
pps
ps3
ptp
pwm pwm: Changes for v4.18-rc1 2018-06-14 16:25:43 +09:00
rapidio treewide: Use array_size() in vmalloc() 2018-06-12 16:19:22 -07:00
ras
regulator treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
remoteproc treewide: use PHYS_ADDR_MAX to avoid type casting ULLONG_MAX 2018-06-15 07:55:25 +09:00
reset - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
rpmsg
rtc - New Device Support 2018-06-11 07:20:17 -07:00
s390 treewide: Use array_size() in vzalloc() 2018-06-12 16:19:22 -07:00
sbus fix a series of Documentation/ broken file name references 2018-06-15 18:10:01 -03:00
scsi SCSI fixes on 20180613 2018-06-14 16:35:32 +09:00
sfi
sh treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
siox
slimbus treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
sn
soc treewide: use PHYS_ADDR_MAX to avoid type casting ULLONG_MAX 2018-06-15 07:55:25 +09:00
soundwire docs: Fix more broken references 2018-06-15 18:11:26 -03:00
spi treewide: devm_kzalloc() -> devm_kcalloc() 2018-06-12 16:19:22 -07:00
spmi
ssb
staging media: v4l: fix broken video4linux docs locations 2018-06-15 18:10:01 -03:00
target treewide: Use array_size() in vzalloc() 2018-06-12 16:19:22 -07:00
tc
tee
thermal - Error path bug fix for overflow tests (Dan) 2018-06-12 18:28:00 -07:00
thunderbolt
tty vfs/y2038: inode timestamps conversion to timespec64 2018-06-15 07:31:07 +09:00
uio treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
usb vfs/y2038: inode timestamps conversion to timespec64 2018-06-15 07:31:07 +09:00
uwb treewide: kmalloc() -> kmalloc_array() 2018-06-12 16:19:22 -07:00
vfio powerpc/powernv/ioda: Allocate indirect TCE levels on demand 2018-07-16 22:53:11 +10:00
vhost virtio, vhost: features, fixes 2018-06-16 06:35:02 +09:00
video Solve a series of broken links for files under Documentation: 2018-06-17 05:25:18 +09:00
virt treewide: kzalloc() -> kcalloc() 2018-06-12 16:19:22 -07:00
virtio virtio, vhost: features, fixes 2018-06-16 06:35:02 +09:00
visorbus
vlynq
vme
w1
watchdog MIPS changes for 4.18 2018-06-12 12:56:02 -07:00
xen treewide: kvmalloc() -> kvmalloc_array() 2018-06-12 16:19:22 -07:00
zorro - Introduce arithmetic overflow test helper functions (Rasmus) 2018-06-06 17:27:14 -07:00
Kconfig
Makefile