mirror of
https://github.com/torvalds/linux.git
synced 2024-12-02 17:11:33 +00:00
64b609d6a6
- A set of commits which reduce the stack usage of various perf event
handling functions which allocated large data structs on stack causing
stack overflows in the worst case.
- Use the proper mechanism for detecting soft interrupts in the recursion
protection.
- Make the resursion protection simpler and more robust.
- Simplify the scheduling of event groups to make the code more robust and
prepare for fixing the issues vs. scheduling of exclusive event groups.
- Prevent event multiplexing and rotation for exclusive event groups
- Correct the perf event attribute exclusive semantics to take pinned
events, e.g. the PMU watchdog, into account
- Make the anythread filtering conditional for Intel's generic PMU
counters as it is not longer guaranteed to be supported on newer
CPUs. Check the corresponding CPUID leaf to make sure.
- Fixup a duplicate initialization in an array which was probably cause by
the usual copy & paste - forgot to edit mishap.
-----BEGIN PGP SIGNATURE-----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=FzCf
-----END PGP SIGNATURE-----
Merge tag 'perf-urgent-2020-11-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf fixes from Thomas Gleixner:
"A set of fixes for perf:
- A set of commits which reduce the stack usage of various perf
event handling functions which allocated large data structs on
stack causing stack overflows in the worst case
- Use the proper mechanism for detecting soft interrupts in the
recursion protection
- Make the resursion protection simpler and more robust
- Simplify the scheduling of event groups to make the code more
robust and prepare for fixing the issues vs. scheduling of
exclusive event groups
- Prevent event multiplexing and rotation for exclusive event groups
- Correct the perf event attribute exclusive semantics to take
pinned events, e.g. the PMU watchdog, into account
- Make the anythread filtering conditional for Intel's generic PMU
counters as it is not longer guaranteed to be supported on newer
CPUs. Check the corresponding CPUID leaf to make sure
- Fixup a duplicate initialization in an array which was probably
caused by the usual 'copy & paste - forgot to edit' mishap"
* tag 'perf-urgent-2020-11-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel/uncore: Fix Add BW copypasta
perf/x86/intel: Make anythread filter support conditional
perf: Tweak perf_event_attr::exclusive semantics
perf: Fix event multiplexing for exclusive groups
perf: Simplify group_sched_in()
perf: Simplify group_sched_out()
perf/x86: Make dummy_iregs static
perf/arch: Remove perf_sample_data::regs_user_copy
perf: Optimize get_recursion_context()
perf: Fix get_recursion_context()
perf/x86: Reduce stack usage for x86_pmu::drain_pebs()
perf: Reduce stack usage of perf_output_begin()
|
||
|---|---|---|
| .. | ||
| boot | ||
| common | ||
| configs | ||
| crypto | ||
| include | ||
| kernel | ||
| lib | ||
| mach-actions | ||
| mach-alpine | ||
| mach-artpec | ||
| mach-asm9260 | ||
| mach-aspeed | ||
| mach-at91 | ||
| mach-axxia | ||
| mach-bcm | ||
| mach-berlin | ||
| mach-clps711x | ||
| mach-cns3xxx | ||
| mach-davinci | ||
| mach-digicolor | ||
| mach-dove | ||
| mach-ebsa110 | ||
| mach-efm32 | ||
| mach-ep93xx | ||
| mach-exynos | ||
| mach-footbridge | ||
| mach-gemini | ||
| mach-highbank | ||
| mach-hisi | ||
| mach-imx | ||
| mach-integrator | ||
| mach-iop32x | ||
| mach-ixp4xx | ||
| mach-keystone | ||
| mach-lpc18xx | ||
| mach-lpc32xx | ||
| mach-mediatek | ||
| mach-meson | ||
| mach-milbeaut | ||
| mach-mmp | ||
| mach-moxart | ||
| mach-mstar | ||
| mach-mv78xx0 | ||
| mach-mvebu | ||
| mach-mxs | ||
| mach-nomadik | ||
| mach-npcm | ||
| mach-nspire | ||
| mach-omap1 | ||
| mach-omap2 | ||
| mach-orion5x | ||
| mach-oxnas | ||
| mach-picoxcell | ||
| mach-prima2 | ||
| mach-pxa | ||
| mach-qcom | ||
| mach-rda | ||
| mach-realtek | ||
| mach-realview | ||
| mach-rockchip | ||
| mach-rpc | ||
| mach-s3c | ||
| mach-s5pv210 | ||
| mach-sa1100 | ||
| mach-shmobile | ||
| mach-socfpga | ||
| mach-spear | ||
| mach-sti | ||
| mach-stm32 | ||
| mach-sunxi | ||
| mach-tango | ||
| mach-tegra | ||
| mach-u300 | ||
| mach-uniphier | ||
| mach-ux500 | ||
| mach-versatile | ||
| mach-vexpress | ||
| mach-vt8500 | ||
| mach-zx | ||
| mach-zynq | ||
| mm | ||
| net | ||
| nwfpe | ||
| oprofile | ||
| plat-omap | ||
| plat-orion | ||
| plat-pxa | ||
| plat-versatile | ||
| probes | ||
| tools | ||
| vdso | ||
| vfp | ||
| xen | ||
| Kbuild | ||
| Kconfig | ||
| Kconfig-nommu | ||
| Kconfig.assembler | ||
| Kconfig.debug | ||
| Makefile | ||