Age | Commit message (Collapse) | Author |
|
Several bitmaps are declared in arch/mips/include/asm/gic.h, but the
scope of their use is limited to arch/mips/kernel/irq-gic.c. Move the
declarations from the header file to the C file.
Signed-off-by: Jeffrey Deans <jeffrey.deans@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7372/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Detect if the core supports unique exception codes for the
Read-Inhibit and Execute-Inhibit exceptions and set the
option accordingly. The RI/XI exception support is detected
by setting the 27th bit (IEC) of the PageGrain C0 register
and reading back the value of that register to verify the
bit is enabled.
Signed-off-by: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7340/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Use the regular tlb_do_page_fault_0 (no write) handler to handle
the RI and XI exceptions. Also skip the RI/XI validation check
on TLB load handler since it's redundant when the CPU has
unique RI/XI exceptions.
Singed-off-by: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7339/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
MIPSr5 added support for unique exception codes for the Read-Inhibit
and Execute-Inhibit exceptions.
Signed-off-by: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7338/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
The Hardware Page Table Walker aims to speed up TLB refill exceptions
by handling them in the hardware level instead of having a software
TLB refill handler. However, a TLB refill exception can still be
thrown in certain cases such as, synchronus exceptions, or address
translation or memory errors during the HTW operation. As a result of
which, HTW must not be considered a complete replacement for the TLB
refill software handler, but rather a fast-path for it.
For HTW to work, the PWBase register must contain the task's page
global directory address so the HTW will kick in on TLB refill
exceptions.
Due to HTW being a separate engine embedded deep in the CPU pipeline,
we need to restart the HTW everytime a PTE changes to avoid HTW
fetching a old entry from the page tables. It's also necessary to
restart the HTW on context switches to prevent it from fetching a
page from the previous process. Finally, since HTW is using the
entryhi register to write the translations to the TLB, it's necessary
to stop the HTW whenever the entryhi changes (eg for tlb probe
perations) and enable it back afterwards.
== Performance ==
The following trivial test was used to measure the performance of the
HTW. Using the same root filesystem, the following command was used
to measure the number of tlb refill handler executions with and
without (using 'nohtw' kernel parameter) HTW support. The kernel was
modified to use a scratch register as a counter for the TLB refill
exceptions.
find /usr -type f -exec ls -lh {} \;
HTW Enabled:
TLB refill exceptions: 12306
HTW Disabled:
TLB refill exceptions: 17805
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: Markos Chandras <markos.chandras@imgtec.com>
Patchwork: https://patchwork.linux-mips.org/patch/7336/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Detect if the core implements the HTW and set the option accordingly.
Also, add a new kernel parameter called 'nohtw' allowing
the user to disable the htw support and fallback to the software
refill handler.
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7335/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7326/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Moreover, report hardware page table walker support as 'htw' in the ASE
list of /proc/cpuinfo, if the core implements this feature.
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7334/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Long integers which are 4 bytes in MIPS32 can't hold new CPU
options anymore, so the type of the 'options' variable is changed
to unsigned long long which allows 32 more cpu options to be defined
for MIPS32
Also, re-arrange the 'options' struct member to avoid potential 4-byte
alignment gap in the middle of the struct.
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7324/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Add cases in perf_event_mipsxx.c for CPU_P5600. All the event numbers
listed for proAptiv also apply to P5600, so we use mipsxxcore_event_map2
and mipsxxcore_cache_map2 too, but the P5600 has 8-bit event numbers so
bit 8 (256) of the user ABI config is used for the parity bit (to
specify odd/even counter events).
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7242/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
In mipsxx_pmu_map_raw_event(), set event_id to base_id after the cpu
type conditional code to allow that code to override the base_id to use
more bits from the config and a higher bit for parity.
This will allow cores with up to 512 events between all even/odd
counters (an 8-bit event id) such as P5600 to use bit 8 for parity.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7243/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
This header defines an exported interface (the register layout used in
core dumps and the GP regset accessible with PTRACE_{GET,SET}REGSET),
therefore belongs in uapi.
Signed-off-by: Alex Smith <alex@alex-smith.me.uk>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7458/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
The struct user definition in this file is not used anywhere (the ELF
core dumper does not use that format). Therefore, remove the header and
instead enable the asm-generic user.h which is an empty header to
satisfy a few generic headers which still try to include user.h.
Signed-off-by: Alex Smith <alex@alex-smith.me.uk>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7459/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Since the core dumper now uses regsets, the old core dump functions are
now unused. Remove them.
Signed-off-by: Alex Smith <alex@alex-smith.me.uk>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7456/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
In uapi/asm/ptrace.h, a user version of pt_regs is defined wrapped in
ifndef __KERNEL__. This structure definition does not match anything
used by any kernel API, in particular it does not match the format used
by PTRACE_{GET,SET}REGS.
Therefore, replace the structure definition with one matching what is
used by PTRACE_{GET,SET}REGS. The format used by these is the same for
both 32-bit and 64-bit.
Also, change the implementation of PTRACE_{GET,SET}REGS to use this new
structure definition. The structure is renamed to user_pt_regs when
__KERNEL__ is defined to avoid conflicts with the kernel's own pt_regs.
Signed-off-by: Alex Smith <alex@alex-smith.me.uk>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7457/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
A comment in the O32/32-bit system call code is incorrect since commit
46e12c07b3b9 ("MIPS: O32 / 32-bit: Always copy 4 stack arguments.").
Remove it.
Signed-off-by: Alex Smith <alex@alex-smith.me.uk>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7455/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
On 32-bit/O32, pt_regs has a padding area at the beginning into which the
syscall arguments passed via the user stack are copied. 4 arguments
totalling 16 bytes are copied to offset 16 bytes into this area, however
the area is only 24 bytes long. This means the last 2 arguments overwrite
pt_regs->regs[{0,1}].
If a syscall function returns an error, handle_sys stores the original
syscall number in pt_regs->regs[0] for syscall restart. signal.c checks
whether regs[0] is non-zero, if it is it will check whether the syscall
return value is one of the ERESTART* codes to see if it must be
restarted.
Should a syscall be made that results in a non-zero value being copied
off the user stack into regs[0], and then returns a positive (non-error)
value that matches one of the ERESTART* error codes, this can be mistaken
for requiring a syscall restart.
While the possibility for this to occur has always existed, it is made
much more likely to occur by commit 46e12c07b3b9 ("MIPS: O32 / 32-bit:
Always copy 4 stack arguments."), since now every syscall will copy 4
arguments and overwrite regs[0], rather than just those with 7 or 8
arguments.
Since that commit, booting Debian under a 32-bit MIPS kernel almost
always results in a hang early in boot, due to a wait4 syscall returning
a PID that matches one of the ERESTART* codes, which then causes an
incorrect restart of the syscall.
The problem is fixed by increasing the size of the padding area so that
arguments copied off the stack will not overwrite pt_regs->regs[{0,1}].
Signed-off-by: Alex Smith <alex.smith@imgtec.com>
Cc: <stable@vger.kernel.org> # v3.13+
Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
Tested-by: Aurelien Jarno <aurelien@aurel32.net>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7454/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM straggler SoC fix from Olof Johansson:
"A DT bugfix for Nomadik that had an ambigouos double-inversion of a
gpio line, and one MAINTAINER URL update that might as well go in now.
We could hold off until the merge window, but then we'll just have to
mark the DT fix for stable and it just seems like in total causing
more work"
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
MAINTAINERS: Update Tegra Git URL
ARM: nomadik: fix up double inversion in DT
|
|
devel-stable
Two different fixes for the same problem making some ARM nommu configurations
not boot since 3.6-rc1. The problem is that user_addr_max returned the biggest
available RAM address which makes some copy_from_user variants fail to read
from XIP memory.
Even in the presence of one of the two fixes the other still makes sense, so
both patches are included here.
This problem was the last one preventing efm32 boot to a prompt with mainline.
|
|
Commit
Commit 1d40cfcd3442a53e98468cdb3e6d4d9a568d76cf
Author: Ralf Baechle <ralf@linux-mips.org>
Date: Fri Jul 15 15:23:23 2005 +0000
Avoid SMP cacheflushes. This is a minor optimization of startup but
will also avoid smp_call_function from doing stupid things when called
from a CPU that is not yet marked online.
missed an appropriate cache flush of TLB refill handler because that time it was
at fixed location CAC_BASE. After years the refill handler in EBASE vector
is not at that location and can be allocated in some another memory and needs
I-cache sync as other TLB exception vectors.
Besides that, the new function - local_flash_icache_range() was introduced
to avoid SMP cacheflushes.
Signed-off-by: Leonid Yegoshin <Leonid.Yegoshin@imgtec.com>
Cc: linux-mips@linux-mips.org
Cc: paul.gortmaker@windriver.com
Cc: jchandra@broadcom.com
Cc: linux-kernel@vger.kernel.org
Cc: david.daney@cavium.com
Patchwork: https://patchwork.linux-mips.org/patch/7312/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Rename usb_nop_xceiv to usb_phy_generic in platform data to match the
name change of the nop transceiver driver in commit 4525bee (usb: phy:
rename usb_nop_xceiv to usb_phy_generic).
The name change induced a kernel panic due to an unhandled kernel
unaligned access while trying to dereference musb->xceiv->io_ops in
musb_init_controller().
Signed-off-by: Apelete Seketeli <apelete@seketeli.net>
Acked-by: Lars-Peter Clausen <lars@metafoo.de>
Cc: John Crispin <blogic@openwrt.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/7263/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Commit d6d3c9afaab4 (MIPS: MT: proc: Add support for printing VPE and TC
ids) causes a link error when CONFIG_PROC_FS=n:
arch/mips/built-in.o: In function `proc_cpuinfo_notifier_init':
smp-mt.c: undefined reference to `register_proc_cpuinfo_notifier'
This is fixed by adding an ifdef around the procfs handling code
in smp-mt.c.
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Reported-by: Markos Chandras <markos.chandras@imgtec.com>
Reviewed-by: Markos Chandras <markos.chandras@imgtec.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Cc: <stable@vger.kernel.org> # >= 3.15
Patchwork: https://patchwork.linux-mips.org/patch/7244/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
The IP_PKTOPTIONS sockopt puts control messages in option_values, these
need to be handled differently in the compat case. This is already done
through the MSG_CMSG_COMPAT flag, we just need to use
compat_sys_getsockopt which sets that flag.
Signed-off-by: Sorin Dumitru <sdumitru@ixiacom.com>
Reviewed-by: James Hogan <james.hogan@imgtec.com>
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/7115/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
If a call to device_create() fails for a channel during the initialize
loop, we need to clean the devices entries already created before
leaving.
Signed-off-by: Sebastien Bourdelin <sebastien.bourdelin@savoirfairelinux.com>
Cc: linux-mips@linux-mips.org
Cc: Steven J. Hill <Steven.Hill@imgtec.com>
Cc: Deng-Cheng Zhu <dengcheng.zhu@imgtec.com>
Cc: John Crispin <blogic@openwrt.org>
Cc: Qais Yousef <Qais.Yousef@imgtec.com>
Cc: linux-kernel@vger.kernel.org
Cc: Jerome Oufella <jerome.oufella@savoirfairelinux.com>
Patchwork: https://patchwork.linux-mips.org/patch/7111/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
|
|
Due to a missing newline in the I-cache policy detection log output,
it's possible to get some ratehr unfortunate output at boot time:
CPU1: Booted secondary processor
Detected VIPT I-cache on CPU1CPU2: Booted secondary processor
Detected VIPT I-cache on CPU2CPU3: Booted secondary processor
Detected VIPT I-cache on CPU3CPU4: Booted secondary processor
Detected PIPT I-cache on CPU4CPU5: Booted secondary processor
Detected PIPT I-cache on CPU5Brought up 6 CPUs
SMP: Total of 6 processors activated.
This patch adds the missing newline to the format string, cleaning up
the output.
Fixes: 59ccc0d41b7a ("arm64: cachetype: report weakest cache policy")
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
|
|
The large segment table entry format has block of bits for the
ACC/F values for the large page. These bits are valid only if
another bit (AV bit 0x10000) of the segment table entry is set.
The ACC/F bits do not have a meaning if the AV bit is off.
This allows to put the THP splitting bit, the segment young bit
and the new segment dirty bit into the ACC/F bits as long as
the AV bit stays off. The dirty and young information is only
available if the pmd is large.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
|
|
Commit f0a3eaff71b8 (ARM64: KVM: fix big endian issue in
access_vm_reg for 32bit guest) changed the way we handle CP15
VM accesses, so that all 64bit accesses are done via vcpu_sys_reg.
This looks like a good idea as it solves indianness issues in an
elegant way, except for one small detail: the register index is
doesn't refer to the same array! We end up corrupting some random
data structure instead.
Fix this by reverting to the original code, except for the introduction
of a vcpu_cp15_64_high macro that deals with the endianness thing.
Tested on Juno with 32bit SMP guests.
Cc: Victor Kamensky <victor.kamensky@linaro.org>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
|
|
commit ec66ad66a0de87866be347b5ecc83bd46427f53b (s390/mm: enable
split page table lock for PMD level) activated the split pmd lock
for s390. Turns out that we missed one place: We also have to take
the pmd lock instead of the page table lock when we reallocate the
page tables (==> changing entries in the PMD) during sie enablement.
Cc: stable@vger.kernel.org # 3.15+
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
|
|
Fix coding style problems in arch/arm/mach-omap2/control.c.
Signed-off-by: Jeremy Vial <jvial@adeneo-embedded.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
According to the comment “restore_es3: applies to 34xx >= ES3.0" in
"arch/arm/mach-omap2/sleep34xx.S”, omap3_restore_es3 should be used
if the revision of an OMAP34xx is ES3.1.2.
Signed-off-by: Jeremy Vial <jvial@adeneo-embedded.com>
Cc: stable@vger.kernel.org
Signed-off-by: Tony Lindgren <tony@atomide.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung into next/dt
Merge "Samsung DT 2nd updates for v3.17" from Kukjin Kim:
This is based on tags/exynos-power because this DT changes
are depending PMU cleanup.
Fixes boot for exynos5260 and exynos5410,
- Since exynos cannot boot without obtaining PMU address via
DT from now on, add PMU node for exynos5260 and exynos5410
For preparing exynos5250-spring,
- move max77686 and cypress,cyapa trackpad from exynos5250-
cros-common to exynos5250-snow DT file
(Note exynos5250-spring is not included in this branch yet)
For exynos3250,
- add TMU node and remove duplicated interrupt-parent
- add missing pinctrl property for uart0 and uart1
For exynos5250-smdk5250 board
- add max77686 pmic interrupt property which is connected to
gpx3
* tag 'samsung-dt-2' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung: (28 commits)
ARM: dts: Add missing pinctrl for uart0/1 for exynos3250
ARM: dts: Remove duplicate 'interrput-parent' property for exynos3250
ARM: dts: Add TMU dt node to monitor the temperature for exynos3250
ARM: dts: Specify MAX77686 pmic interrupt for exynos5250-smdk5250
ARM: dts: cypress,cyapa trackpad is exynos5250-Snow only
ARM: dts: max77686 is exynos5250-snow only
ARM: EXYNOS: Add exynos5260 PMU compatible string to DT match table
ARM: dts: Add PMU DT node for exynos5260 SoC
ARM: EXYNOS: Add support for Exynos5410 PMU
ARM: dts: Add PMU to exynos5410
ARM: dts: Document exynos5410 PMU
ARM: EXYNOS: Move cpufreq and cpuidle device registration to init_machine
ARM: EXYNOS: Refactored code for using PMU address via DT
ARM: EXYNOS: Support cluster power off on exynos5420/5800
ARM: EXYNOS: populate suspend and powered_up callbacks for mcpm
ARM: EXYNOS: do not allow cpuidle registration for exynos5420
cpuidle: big.LITTLE: init driver for exynos5420
cpuidle: big.LITTLE: Add ARCH_EXYNOS entry in config
ARM: EXYNOS: add generic function to calculate cpu number
cpuidle: big.LITTLE: add of_device_id structure
...
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
This patch adds socfpga Ethernet filter attributes for multicast
and unicast filters per Synopsys Ethernet IP configuration chosen
by Altera for the Cyclone 5 and Arria SOC FPGAs.
Signed-off-by: Vince Bridgers <vbridgers2013@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
'arm/exynos' and 'core' into next
|
|
Removing a debug message for setting the identity map since it becomes
rather noisy after rework of the identity map code.
Signed-off-by: Matthew Rushton <mrushton@amazon.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tfiga/samsung-clk into clk-next-samsung
Samsung clock patches for 3.17
1) non-critical fixes (without need to push to stable):
d5e136a clk: samsung: Register clk provider only after registering its all clocks
305cfab clk: samsung: Make of_device_id array const
e9d5295 clk: samsung: exynos5420: Setup clocks before system suspend
f65d518 clk: samsung: trivial: Correct typo in author's name
2) Exynos CLKOUT driver:
800c979 clk: samsung: exynos4: Add missing CPU/DMC clock hierarchy
01f7ec2 clk: samsung: exynos4: Add CLKOUT clock hierarchy
1e832e5 clk: samsung: Add driver to control CLKOUT line on Exynos SoCs
d19bb39 ARM: dts: exynos: Update PMU node with CLKOUT related data
3) Clock hierarchy extensions:
17d3f1d clk: exynos4: Add PPMU IP block source clocks.
ca5b402 clk: samsung: register exynos5420 apll/kpll configuration data
4) ARM CLKDOWN functionality enablement for Exynos4 and 3250:
42773b2 clk: samsung: exynos4: Enable ARMCLK down feature
45c5b0a clk: samsung: exynos3250: Enable ARMCLK down feature
|
|
This has been run through Intel's LKP tests across a wide range
of modern sytems and workloads and it wasn't shown to make a
measurable performance difference positive or negative.
Now that we have some shiny new tracepoints, we can actually
figure out what the heck is going on.
During a kernel compile, 60% of the flush_tlb_mm_range() calls
are for a single page. It breaks down like this:
size percent percent<=
V V V
GLOBAL: 2.20% 2.20% avg cycles: 2283
1: 56.92% 59.12% avg cycles: 1276
2: 13.78% 72.90% avg cycles: 1505
3: 8.26% 81.16% avg cycles: 1880
4: 7.41% 88.58% avg cycles: 2447
5: 1.73% 90.31% avg cycles: 2358
6: 1.32% 91.63% avg cycles: 2563
7: 1.14% 92.77% avg cycles: 2862
8: 0.62% 93.39% avg cycles: 3542
9: 0.08% 93.47% avg cycles: 3289
10: 0.43% 93.90% avg cycles: 3570
11: 0.20% 94.10% avg cycles: 3767
12: 0.08% 94.18% avg cycles: 3996
13: 0.03% 94.20% avg cycles: 4077
14: 0.02% 94.23% avg cycles: 4836
15: 0.04% 94.26% avg cycles: 5699
16: 0.06% 94.32% avg cycles: 5041
17: 0.57% 94.89% avg cycles: 5473
18: 0.02% 94.91% avg cycles: 5396
19: 0.03% 94.95% avg cycles: 5296
20: 0.02% 94.96% avg cycles: 6749
21: 0.18% 95.14% avg cycles: 6225
22: 0.01% 95.15% avg cycles: 6393
23: 0.01% 95.16% avg cycles: 6861
24: 0.12% 95.28% avg cycles: 6912
25: 0.05% 95.32% avg cycles: 7190
26: 0.01% 95.33% avg cycles: 7793
27: 0.01% 95.34% avg cycles: 7833
28: 0.01% 95.35% avg cycles: 8253
29: 0.08% 95.42% avg cycles: 8024
30: 0.03% 95.45% avg cycles: 9670
31: 0.01% 95.46% avg cycles: 8949
32: 0.01% 95.46% avg cycles: 9350
33: 3.11% 98.57% avg cycles: 8534
34: 0.02% 98.60% avg cycles: 10977
35: 0.02% 98.62% avg cycles: 11400
We get in to dimishing returns pretty quickly. On pre-IvyBridge
CPUs, we used to set the limit at 8 pages, and it was set at 128
on IvyBrige. That 128 number looks pretty silly considering that
less than 0.5% of the flushes are that large.
The previous code tried to size this number based on the size of
the TLB. Good idea, but it's error-prone, needs maintenance
(which it didn't get up to now), and probably would not matter in
practice much.
Settting it to 33 means that we cover the mallopt
M_TRIM_THRESHOLD, which is the most universally common size to do
flushes.
That's the short version. Here's the long one for why I chose 33:
1. These numbers have a constant bias in the timestamps from the
tracing. Probably counts for a couple hundred cycles in each of
these tests, but it should be fairly _even_ across all of them.
The smallest delta between the tracepoints I have ever seen is
335 cycles. This is one reason the cycles/page cost goes down in
general as the flushes get larger. The true cost is nearer to
100 cycles.
2. A full flush is more expensive than a single invlpg, but not
by much (single percentages).
3. A dtlb miss is 17.1ns (~45 cycles) and a itlb miss is 13.0ns
(~34 cycles). At those rates, refilling the 512-entry dTLB takes
22,000 cycles.
4. 22,000 cycles is approximately the equivalent of doing 85
invlpg operations. But, the odds are that the TLB can
actually be filled up faster than that because TLB misses that
are close in time also tend to leverage the same caches.
6. ~98% of flushes are <=33 pages. There are a lot of flushes of
33 pages, probably because libc's M_TRIM_THRESHOLD is set to
128k (32 pages)
7. I've found no consistent data to support changing the IvyBridge
vs. SandyBridge tunable by a factor of 16
I used the performance counters on this hardware (IvyBridge i5-3320M)
to figure out the tlb miss costs:
ocperf.py stat -e dtlb_load_misses.walk_duration,dtlb_load_misses.walk_completed,dtlb_store_misses.walk_duration,dtlb_store_misses.walk_completed,itlb_misses.walk_duration,itlb_misses.walk_completed,itlb.itlb_flush
7,720,030,970 dtlb_load_misses_walk_duration [57.13%]
169,856,353 dtlb_load_misses_walk_completed [57.15%]
708,832,859 dtlb_store_misses_walk_duration [57.17%]
19,346,823 dtlb_store_misses_walk_completed [57.17%]
2,779,687,402 itlb_misses_walk_duration [57.15%]
82,241,148 itlb_misses_walk_completed [57.13%]
770,717 itlb_itlb_flush [57.11%]
Show that a dtlb miss is 17.1ns (~45 cycles) and a itlb miss is 13.0ns
(~34 cycles). At those rates, refilling the 512-entry dTLB takes
22,000 cycles. On a SandyBridge system with more cores and larger
caches, those are dtlb=13.4ns and itlb=9.5ns.
cat perf.stat.txt | perl -pe 's/,//g'
| awk '/itlb_misses_walk_duration/ { icyc+=$1 }
/itlb_misses_walk_completed/ { imiss+=$1 }
/dtlb_.*_walk_duration/ { dcyc+=$1 }
/dtlb_.*.*completed/ { dmiss+=$1 }
END {print "itlb cyc/miss: ", icyc/imiss, " dtlb cyc/miss: ", dcyc/dmiss, " ----- ", icyc,imiss, dcyc,dmiss }
On Westmere CPUs, the counters to use are: itlb_flush,itlb_misses.walk_cycles,itlb_misses.any,dtlb_misses.walk_cycles,dtlb_misses.any
The assumptions that this code went in under:
https://lkml.org/lkml/2012/6/12/119 say that a flush and a refill are
about 100ns. Being generous, that is over by a factor of 6 on the
refill side, although it is fairly close on the cost of an invlpg.
An increase of a single invlpg operation seems to lengthen the flush
range operation by about 200 cycles. Here is one example of the data
collected for flushing 10 and 11 pages (full data are below):
10: 0.43% 93.90% avg cycles: 3570 cycles/page: 357 samples: 4714
11: 0.20% 94.10% avg cycles: 3767 cycles/page: 342 samples: 2145
How to generate this table:
echo 10000 > /sys/kernel/debug/tracing/buffer_size_kb
echo x86-tsc > /sys/kernel/debug/tracing/trace_clock
echo 'reason != 0' > /sys/kernel/debug/tracing/events/tlb/tlb_flush/filter
echo 1 > /sys/kernel/debug/tracing/events/tlb/tlb_flush/enable
Pipe the trace output in to this script:
http://sr71.net/~dave/intel/201402-tlb/trace-time-diff-process.pl.txt
Note that these data were gathered with the invlpg threshold set to
150 pages. Only data points with >=50 of samples were printed:
Flush % of %<=
in flush this
pages es size
------------------------------------------------------------------------------
-1: 2.20% 2.20% avg cycles: 2283 cycles/page: xxxx samples: 23960
1: 56.92% 59.12% avg cycles: 1276 cycles/page: 1276 samples: 620895
2: 13.78% 72.90% avg cycles: 1505 cycles/page: 752 samples: 150335
3: 8.26% 81.16% avg cycles: 1880 cycles/page: 626 samples: 90131
4: 7.41% 88.58% avg cycles: 2447 cycles/page: 611 samples: 80877
5: 1.73% 90.31% avg cycles: 2358 cycles/page: 471 samples: 18885
6: 1.32% 91.63% avg cycles: 2563 cycles/page: 427 samples: 14397
7: 1.14% 92.77% avg cycles: 2862 cycles/page: 408 samples: 12441
8: 0.62% 93.39% avg cycles: 3542 cycles/page: 442 samples: 6721
9: 0.08% 93.47% avg cycles: 3289 cycles/page: 365 samples: 917
10: 0.43% 93.90% avg cycles: 3570 cycles/page: 357 samples: 4714
11: 0.20% 94.10% avg cycles: 3767 cycles/page: 342 samples: 2145
12: 0.08% 94.18% avg cycles: 3996 cycles/page: 333 samples: 864
13: 0.03% 94.20% avg cycles: 4077 cycles/page: 313 samples: 289
14: 0.02% 94.23% avg cycles: 4836 cycles/page: 345 samples: 236
15: 0.04% 94.26% avg cycles: 5699 cycles/page: 379 samples: 390
16: 0.06% 94.32% avg cycles: 5041 cycles/page: 315 samples: 643
17: 0.57% 94.89% avg cycles: 5473 cycles/page: 321 samples: 6229
18: 0.02% 94.91% avg cycles: 5396 cycles/page: 299 samples: 224
19: 0.03% 94.95% avg cycles: 5296 cycles/page: 278 samples: 367
20: 0.02% 94.96% avg cycles: 6749 cycles/page: 337 samples: 185
21: 0.18% 95.14% avg cycles: 6225 cycles/page: 296 samples: 1964
22: 0.01% 95.15% avg cycles: 6393 cycles/page: 290 samples: 83
23: 0.01% 95.16% avg cycles: 6861 cycles/page: 298 samples: 61
24: 0.12% 95.28% avg cycles: 6912 cycles/page: 288 samples: 1307
25: 0.05% 95.32% avg cycles: 7190 cycles/page: 287 samples: 533
26: 0.01% 95.33% avg cycles: 7793 cycles/page: 299 samples: 94
27: 0.01% 95.34% avg cycles: 7833 cycles/page: 290 samples: 66
28: 0.01% 95.35% avg cycles: 8253 cycles/page: 294 samples: 73
29: 0.08% 95.42% avg cycles: 8024 cycles/page: 276 samples: 846
30: 0.03% 95.45% avg cycles: 9670 cycles/page: 322 samples: 296
31: 0.01% 95.46% avg cycles: 8949 cycles/page: 288 samples: 79
32: 0.01% 95.46% avg cycles: 9350 cycles/page: 292 samples: 60
33: 3.11% 98.57% avg cycles: 8534 cycles/page: 258 samples: 33936
34: 0.02% 98.60% avg cycles: 10977 cycles/page: 322 samples: 268
35: 0.02% 98.62% avg cycles: 11400 cycles/page: 325 samples: 177
36: 0.01% 98.63% avg cycles: 11504 cycles/page: 319 samples: 161
37: 0.02% 98.65% avg cycles: 11596 cycles/page: 313 samples: 182
38: 0.02% 98.66% avg cycles: 11850 cycles/page: 311 samples: 195
39: 0.01% 98.68% avg cycles: 12158 cycles/page: 311 samples: 128
40: 0.01% 98.68% avg cycles: 11626 cycles/page: 290 samples: 78
41: 0.04% 98.73% avg cycles: 11435 cycles/page: 278 samples: 477
42: 0.01% 98.73% avg cycles: 12571 cycles/page: 299 samples: 74
43: 0.01% 98.74% avg cycles: 12562 cycles/page: 292 samples: 78
44: 0.01% 98.75% avg cycles: 12991 cycles/page: 295 samples: 108
45: 0.01% 98.76% avg cycles: 13169 cycles/page: 292 samples: 78
46: 0.02% 98.78% avg cycles: 12891 cycles/page: 280 samples: 261
47: 0.01% 98.79% avg cycles: 13099 cycles/page: 278 samples: 67
48: 0.01% 98.80% avg cycles: 13851 cycles/page: 288 samples: 77
49: 0.01% 98.80% avg cycles: 13749 cycles/page: 280 samples: 66
50: 0.01% 98.81% avg cycles: 13949 cycles/page: 278 samples: 73
52: 0.00% 98.82% avg cycles: 14243 cycles/page: 273 samples: 52
54: 0.01% 98.83% avg cycles: 15312 cycles/page: 283 samples: 87
55: 0.01% 98.84% avg cycles: 15197 cycles/page: 276 samples: 109
56: 0.02% 98.86% avg cycles: 15234 cycles/page: 272 samples: 208
57: 0.00% 98.86% avg cycles: 14888 cycles/page: 261 samples: 53
58: 0.01% 98.87% avg cycles: 15037 cycles/page: 259 samples: 59
59: 0.01% 98.87% avg cycles: 15752 cycles/page: 266 samples: 63
62: 0.00% 98.89% avg cycles: 16222 cycles/page: 261 samples: 54
64: 0.02% 98.91% avg cycles: 17179 cycles/page: 268 samples: 248
65: 0.12% 99.03% avg cycles: 18762 cycles/page: 288 samples: 1324
85: 0.00% 99.10% avg cycles: 21649 cycles/page: 254 samples: 50
127: 0.01% 99.18% avg cycles: 32397 cycles/page: 255 samples: 75
128: 0.13% 99.31% avg cycles: 31711 cycles/page: 247 samples: 1466
129: 0.18% 99.49% avg cycles: 33017 cycles/page: 255 samples: 1927
181: 0.33% 99.84% avg cycles: 2489 cycles/page: 13 samples: 3547
256: 0.05% 99.91% avg cycles: 2305 cycles/page: 9 samples: 550
512: 0.03% 99.95% avg cycles: 2133 cycles/page: 4 samples: 304
1512: 0.01% 99.99% avg cycles: 3038 cycles/page: 2 samples: 65
Here are the tlb counters during a 10-second slice of a kernel compile
for a SandyBridge system. It's better than IvyBridge, but probably
due to the larger caches since this was one of the 'X' extreme parts.
10,873,007,282 dtlb_load_misses_walk_duration
250,711,333 dtlb_load_misses_walk_completed
1,212,395,865 dtlb_store_misses_walk_duration
31,615,772 dtlb_store_misses_walk_completed
5,091,010,274 itlb_misses_walk_duration
163,193,511 itlb_misses_walk_completed
1,321,980 itlb_itlb_flush
10.008045158 seconds time elapsed
# cat perf.stat.1392743721.txt | perl -pe 's/,//g' | awk '/itlb_misses_walk_duration/ { icyc+=$1 } /itlb_misses_walk_completed/ { imiss+=$1 } /dtlb_.*_walk_duration/ { dcyc+=$1 } /dtlb_.*.*completed/ { dmiss+=$1 } END {print "itlb cyc/miss: ", icyc/imiss/3.3, " dtlb cyc/miss: ", dcyc/dmiss/3.3, " ----- ", icyc,imiss, dcyc,dmiss }'
itlb ns/miss: 9.45338 dtlb ns/miss: 12.9716
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154103.10C1115E@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
Most of the logic here is in the documentation file. Please take
a look at it.
I know we've come full-circle here back to a tunable, but this
new one is *WAY* simpler. I challenge anyone to describe in one
sentence how the old one worked. Here's the way the new one
works:
If we are flushing more pages than the ceiling, we use
the full flush, otherwise we use per-page flushes.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154101.12B52CAF@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
We don't have any good way to figure out what kinds of flushes
are being attempted. Right now, we can try to use the vm
counters, but those only tell us what we actually did with the
hardware (one-by-one vs full) and don't tell us what was actually
_requested_.
This allows us to select out "interesting" TLB flushes that we
might want to optimize (like the ranged ones) and ignore the ones
that we have very little control over (the ones at context
switch).
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154059.4C96CBA5@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
There are currently three paths through the remote flush code:
1. full invalidation
2. single page invalidation using invlpg
3. ranged invalidation using invlpg
This takes 2 and 3 and combines them in to a single path by
making the single-page one just be the start and end be start
plus a single page. This makes placement of our tracepoint easier.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154058.E0F90408@viggo.jf.intel.com
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
If we take the
if (end == TLB_FLUSH_ALL || vmflag & VM_HUGETLB) {
local_flush_tlb();
goto out;
}
path out of flush_tlb_mm_range(), we will have flushed the tlb,
but not incremented NR_TLB_LOCAL_FLUSH_ALL. This unifies the
way out of the function so that we always take a single path when
doing a full tlb flush.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154056.FF763B76@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
I think the flush_tlb_mm_range() code that tries to tune the
flush sizes based on the CPU needs to get ripped out for
several reasons:
1. It is obviously buggy. It uses mm->total_vm to judge the
task's footprint in the TLB. It should certainly be using
some measure of RSS, *NOT* ->total_vm since only resident
memory can populate the TLB.
2. Haswell, and several other CPUs are missing from the
intel_tlb_flushall_shift_set() function. Thus, it has been
demonstrated to bitrot quickly in practice.
3. It is plain wrong in my vm:
[ 0.037444] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0
[ 0.037444] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0
[ 0.037444] tlb_flushall_shift: 6
Which leads to it to never use invlpg.
4. The assumptions about TLB refill costs are wrong:
http://lkml.kernel.org/r/1337782555-8088-3-git-send-email-alex.shi@intel.com
(more on this in later patches)
5. I can not reproduce the original data: https://lkml.org/lkml/2012/5/17/59
I believe the sample times were too short. Running the
benchmark in a loop yields times that vary quite a bit.
Note that this leaves us with a static ceiling of 1 page. This
is a conservative, dumb setting, and will be revised in a later
patch.
This also removes the code which attempts to predict whether we
are flushing data or instructions. We expect instruction flushes
to be relatively rare and not worth tuning for explicitly.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154055.ABC88E89@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
The
if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
line of code is not exactly the easiest to audit, especially when
it ends up at two different indentation levels. This eliminates
one of the the copy-n-paste versions. It also gives us a unified
exit point for each path through this function. We need this in
a minute for our tracepoint.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://lkml.kernel.org/r/20140731154054.44F1CDDC@viggo.jf.intel.com
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
|
|
The enable_apic_mode() apic callback is never called, so remove it.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302352320.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
Since commit b5660ba76b41 ("x86, platforms: Remove NUMAQ") removed NUMAQ,
the setup_portio_remap() apic callback has been obsolete. Remove it.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302351480.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
Since commit b5660ba76b41 ("x86, platforms: Remove NUMAQ") removed NUMAQ,
the multi_timer_check() apic callback has been obsolete. Remove it.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302351120.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
noop_check_apicid_used() has the same implementation as
default_check_apicid_used() in the standard header file, so replace the
former with the latter.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302350450.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
The check_apicid_present() apic callback is never called, so remove it
and functions that implement it.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302350160.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
Since commit b5660ba76b41 ("x86, platforms: Remove NUMAQ") removed NUMAQ,
the mps_oem_check() apic callback has been obsolete. Remove it.
This allows generic_mps_oem_check() to be removed as well.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302349390.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
Since commit b5660ba76b41 ("x86, platforms: Remove NUMAQ") removed NUMAQ,
the smp_callin_clear_local_apic() apic callback has been obsolete.
Remove it.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302349040.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|
|
The trampoline_phys_{high,low} members of struct apic are always
initialized to DEFAULT_TRAMPOLINE_PHYS_HIGH and TRAMPOLINE_PHYS_LOW,
respectively. Hardwire the constants and remove the unneeded members.
Signed-off-by: David Rientjes <rientjes@google.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.02.1407302348330.17503@chino.kir.corp.google.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
|