Ying Huang
966a967116
smp: Avoid using two cache lines for struct call_single_data ...
struct call_single_data is used in IPIs to transfer information between
CPUs. Its size is bigger than sizeof(unsigned long) and less than
cache line size. Currently it is not allocated with any explicit alignment
requirements. This makes it possible for allocated call_single_data to
cross two cache lines, which results in double the number of the cache lines
that need to be transferred among CPUs.
This can be fixed by requiring call_single_data to be aligned with the
size of call_single_data. Currently the size of call_single_data is the
power of 2. If we add new fields to call_single_data, we may need to
add padding to make sure the size of new definition is the power of 2
as well.
Fortunately, this is enforced by GCC, which will report bad sizes.
To set alignment requirements of call_single_data to the size of
call_single_data, a struct definition and a typedef is used.
To test the effect of the patch, I used the vm-scalability multiple
thread swap test case (swap-w-seq-mt). The test will create multiple
threads and each thread will eat memory until all RAM and part of swap
is used, so that huge number of IPIs are triggered when unmapping
memory. In the test, the throughput of memory writing improves ~5%
compared with misaligned call_single_data, because of faster IPIs.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Huang, Ying <ying.huang@intel.com>
[ Add call_single_data_t and align with size of call_single_data. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lu@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/87bmnqd6lz.fsf@yhuang-mobile.sh.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-29 15:14:38 +02:00
Merge branches 'acpica-fix', 'acpi-ec-fix' and 'acpi-properties-fix'
2017-08-24 22:22:53 +02:00
binder: Use wake up hint for synchronous transactions.
2017-07-17 14:44:19 +02:00
libata: fix a couple of doc build warnings
2017-07-31 08:03:06 -07:00
atm: zatm: Fix an error handling path in 'zatm_init_one()'
2017-07-18 11:37:46 -07:00
driver core fixes for 4.13-rc5
2017-08-13 12:44:18 -07:00
smp: Avoid using two cache lines for struct call_single_data
2017-08-29 15:14:38 +02:00
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next
2017-07-05 12:31:59 -07:00
bus: uniphier-system-bus: set up registers when resuming
2017-08-04 12:57:18 +02:00
random: fix warning message on ia64 and parisc
2017-08-08 09:36:46 -07:00
clk: keystone: sci-clk: Fix sci_clk_get
2017-08-02 18:37:26 -07:00
Merge branch 'linus' into locking/core, to pick up fixes
2017-08-25 11:04:51 +02:00
cpufreq: intel_pstate: report correct CPU frequencies during trace
2017-08-11 01:25:53 +02:00
smp: Avoid using two cache lines for struct call_single_data
2017-08-29 15:14:38 +02:00
crypto: ixp4xx - Fix error handling path in 'aead_perform()'
2017-08-09 20:01:33 +08:00
- A few DM integrity fixes that improve performance. One that address
2017-07-28 12:17:17 -07:00
PM / devfreq: constify attribute_group structures.
2017-07-06 10:17:24 +09:00
dmaengine updates for 4.13-rc1
2017-07-08 12:36:50 -07:00
dma-buf/sync_file: Allow multiple sync_files to wrap a single dma-fence
2017-07-31 10:55:24 -03:00
efi: avoid fortify checks in EFI stub
2017-07-12 16:26:02 -07:00
drivers/fsi: fix fsi_slave_mode prototype
2017-07-17 16:13:54 +02:00
gpio: mvebu: Fix cause computation in irq handler
2017-08-14 15:00:43 +02:00
Merge branch 'linus' into locking/core, to pick up fixes
2017-08-25 11:04:51 +02:00
HID: ortek: add one more buggy device
2017-07-24 17:38:21 +02:00
vmbus: re-enable channel tasklet
2017-07-17 15:00:47 +02:00
hwmon: (applesmc) Avoid buffer overruns
2017-07-15 16:38:56 -07:00
i2c: allow i2c-versatile for ARM MPS platforms
2017-07-31 17:05:16 +02:00
ide: avoid warning for timings calculation
2017-07-21 04:37:22 +01:00
First set of IIO fixes for the 4.13 cycle.
2017-07-23 20:54:31 -07:00
IB/mlx5: Always return success for RoCE modify port
2017-08-24 15:33:33 -04:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
2017-08-17 13:45:44 -07:00
iommu/arm-smmu: fix null-pointer dereference in arm_smmu_add_device
2017-08-11 16:56:51 +02:00
irqchip fixes for 4.13
2017-08-14 09:34:10 +02:00
mISDN: Fix null pointer dereference at mISDN_FsmNew
2017-08-11 14:56:23 -07:00
LED updates for 4.13
2017-07-06 11:32:40 -07:00
lightnvm: pblk: advance bio according to lba index
2017-07-28 08:06:00 -06:00
Merge branch 'work.misc-set_fs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
2017-07-05 13:13:32 -07:00
mailbox: pcc: Fix crash when request PCC channel 0
2017-07-26 02:11:47 +02:00
MD: not clear ->safemode for external metadata array
2017-08-11 20:42:06 -07:00
media: platform: davinci: drop VPFE_CMD_S_CCDC_RAW_PARAMS
2017-07-26 06:14:33 -04:00
memory: atmel-ebi: Fix smc cycle xlate converter
2017-07-26 22:37:54 +02:00
ARM: SoC fixes for 4.13
2017-08-24 14:01:18 -07:00
mei: exclude device from suspend direct complete optimization
2017-08-10 14:13:18 -07:00
mmc: block: fix lockdep splat when removing mmc_block module
2017-08-09 13:19:44 +02:00
mtd: blkdevs: Fix mtd block write failure
2017-08-12 14:53:24 -07:00
mux: mux-core: unregister mux_class in mux_exit()
2017-07-17 16:38:35 +02:00
smp: Avoid using two cache lines for struct call_single_data
2017-08-29 15:14:38 +02:00
ntb: Add error path/handling to Debug FS entry creation
2017-07-06 11:30:08 -04:00
libnvdimm: fix badblock range handling of ARS range
2017-07-17 11:43:58 -07:00
nvme-pci: set cqe_seen on polled completions
2017-08-18 09:19:39 +02:00
nvmem: rockchip-efuse: amend compatible rk322x-efuse to rk3228-efuse
2017-07-17 16:15:57 +02:00
of: fix DMA mask generation
2017-08-17 10:23:45 +02:00
parisc: pci memory bar assignment fails with 64bit kernels on dino/cujo
2017-08-16 09:50:39 +02:00
PCI: Allow PCI express root ports to find themselves
2017-08-18 16:14:37 -07:00
drivers/perf: arm_pmu: Request PMU SPIs with IRQF_PER_CPU
2017-07-27 13:43:22 +01:00
phy: bcm-ns-usb3: fix MDIO_BUS dependency
2017-07-27 17:20:19 -07:00
Pin control fixes for the v4.13 cycle:
2017-08-09 14:30:34 -07:00
platform/x86: intel-vbtn: match power button on press rather than release
2017-08-05 14:37:19 -07:00
This is the bulk of GPIO changes for the v4.13 series:
2017-07-07 12:40:27 -07:00
power supply and reset changes for the v4.13 series (part 2)
2017-07-13 11:47:59 -07:00
ptp: introduce ptp auxiliary worker
2017-08-01 15:22:55 -07:00
pwm: Changes for v4.13-rc1
2017-07-13 11:49:52 -07:00
arm64 updates for 4.13:
2017-07-05 17:09:27 -07:00
rpmsg updates for v4.13
2017-07-06 15:38:31 -07:00
rtc: ds1307: fix regmap config
2017-08-21 11:08:03 +02:00
s390/qeth: fix L3 next-hop in xmit qeth hdr
2017-08-07 11:24:37 -07:00
sbus: Convert to using %pOF instead of full_name
2017-07-20 12:37:10 -07:00
SCSI fixes on 20170823
2017-08-23 11:34:40 -07:00
drivers/sh/intc/virq.c: delete an error message for a failed memory allocation in add_virq_to_pirq()
2017-07-06 16:24:30 -07:00
soc: ti: knav: Add a NULL pointer check for kdev in knav_pool_create
2017-08-21 09:19:50 +02:00
Merge branch 'for-spi' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
2017-07-08 10:41:53 -07:00
spmi: pmic-arb: Always allocate ppid_to_apid table
2017-07-17 15:00:47 +02:00
staging/iio fixes for 4.13-rc5
2017-08-13 12:30:17 -07:00
target: Fix node_acl demo-mode + uncached dynamic shutdown regression
2017-08-09 20:55:19 -07:00
Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux
2017-07-14 13:12:32 -07:00
char/misc fixes for 4.13-rc5
2017-08-13 12:41:58 -07:00
pty: Repair TIOCGPTPEER
2017-08-24 13:23:03 -07:00
USB fixes for 4.13-rc5
2017-08-13 12:27:42 -07:00
vfio/pci: Fix handling of RC integrated endpoint PCIe capability size
2017-07-27 10:39:33 -06:00
Revert "vhost: cache used event for better performance"
2017-07-29 14:15:56 -07:00
efifb: allow user to disable write combined mapping.
2017-07-31 18:45:41 +02:00
virtio-balloon: coding format cleanup
2017-07-25 16:37:35 +03:00
w1: omap-hdq: fix error return code in omap_hdq_probe()
2017-07-17 16:48:15 +02:00
Merge git://www.linux-watchdog.org/linux-watchdog
2017-07-11 09:59:37 -07:00
Kbuild fixes for v4.13
2017-08-24 14:22:27 -07:00
966a967116