- 10 May, 2024 6 commits
-
-
https://git.linaro.org/people/daniel.lezcano/linuxThomas Gleixner authored
Pull clockevent/source updates from Daniel Lezcano: - Add the R9A09G057 compatible bindings in the DT documentation and add specific code to deal with the probe routine being called twice (Geert Uytterhoeven) - Remove unused field in the struct dmtimer in the TI driver (Christophe JAILLET) - Constify the hisi_161010101_oem_info variable in the ARM arch timer (Stephen Boyd) Link: https://lore.kernel.org/lkml/7ca1c46a-93e6-4f67-bee3-623cb56764fa@linaro.org -
Stephen Boyd authored
This isn't modified at runtime. Mark it const so it can move to read-only data. Cc: dann frazier <dann.frazier@canonical.com> Cc: Hanjun Guo <hanjun.guo@linaro.org> Cc: Marc Zyngier <maz@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Signed-off-by:
Stephen Boyd <swboyd@chromium.org> Reviewed-by:
Hanjun Guo <guohanjun@huawei.com> Signed-off-by:
Daniel Lezcano <daniel.lezcano@linaro.org> Link: https://lore.kernel.org/r/20240502233447.420888-1-swboyd@chromium.org
-
Christophe JAILLET authored
In "struct dmtimer", the 'rate' field is unused. Remove it. Found with cppcheck, unusedStructMember. Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
-
Geert Uytterhoeven authored
The Renesas OS Timer (OSTM) driver contains two probe points, of which only one should complete: 1. Early probe, using TIMER_OF_DECLARE(), to provide the sole clocksource on (arm32) RZ/A1 and RZ/A2 SoCs, 2. Normal probe, using a platform driver, to provide additional timers on (arm64 + riscv) RZ/G2L and similar SoCs. The latter is needed because using OSTM on RZ/G2L requires manipulation of its reset signal, which is not yet available at the time of early probe, causing early probe to fail with -EPROBE_DEFER. It is only enabled when building a kernel with support for the RZ/G2L family, so it does not impact RZ/A1 and RZ/A2. Hence only one probe method can complete on all affected systems. As relying on the order of initialization of subsystems inside the kernel is fragile, set the DT node's OF_POPULATED flag after a succesful early probe. This makes sure the platform driver's probe is never called after a successful early probe. Signed-off-by:Geert Uytterhoeven <geert+renesas@glider.be> Reviwed-by:
Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> Signed-off-by:
-
Lad Prabhakar authored
The RZ/V2H(P) (R9A09G057) SoC is equipped with the Generic Timer Module, also known as OSTM. Similar to the RZ/G2L SoC, the OSTM on the RZ/V2H(P) SoC requires the reset line to be deasserted before accessing any registers. Early call to ostm_init() happens through TIMER_OF_DECLARE() which always fails with -EPROBE_DEFER, as resets are not available that early in the boot process. To address this issue on the RZ/V2H(P) SoC, enable the OSTM driver to be reprobed through the platform driver probe mechanism. Signed-off-by:
-
Lad Prabhakar authored
Document the General Timer Module (a.k.a OSTM) block on Renesas RZ/V2H(P) ("R9A09G057") SoC, which is identical to the one found on the RZ/A1H and RZ/G2L SoCs. Add the "renesas,r9a09g057-ostm" compatible string for the RZ/V2H(P) SoC. Signed-off-by:Conor Dooley <conor.dooley@microchip.com> Signed-off-by:
-
- 30 Apr, 2024 1 commit
-
-
Boqun Feng authored
The definitions related to jiffies are at linux/jiffies.h, and the definitions related to ktime_t are at linux/ktime.h, since `kernel::time` provides the functionality dealing with jiffies and ktime_t, it makes sense to add links to them from Rust's time module. Signed-off-by:
Boqun Feng <boqun.feng@gmail.com> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Reviewed-by:
Alice Ryhl <aliceryhl@google.com> Acked-by:
Miguel Ojeda <ojeda@kernel.org> Link: https://lore.kernel.org/r/20240411230801.1504496-2-boqun.feng@gmail.com
-
- 29 Apr, 2024 1 commit
-
-
Borislav Petkov (AMD) authored
When doing make menuconfig and searching for the CLOCKSOURCE_WATCHDOG_MAX_SKEW_US config item, the help says: │ Symbol: CLOCKSOURCE_WATCHDOG_MAX_SKEW_US [=125] │ Type : integer │ Range : [50 1000] │ Defined at kernel/time/Kconfig:204 │ Prompt: Clocksource watchdog maximum allowable skew (in s) ^^^ │ Depends on: GENERIC_CLOCKEVENTS [=y] && CLOCKSOURCE_WATCHDOG [=y] because on some terminals, it cannot display the 'μ' char, unicode number 0x3bc. So simply write it out so that there's no trouble. Signed-off-by:
Borislav Petkov (AMD) <bp@alien8.de> Signed-off-by:
Paul E. McKenney <paulmck@kernel.org> Acked-by:
Randy Dunlap <rdunlap@infradead.org> Link: https://lore.kernel.org/r/20240428102143.26764-1-bp@kernel.org
-
- 22 Apr, 2024 2 commits
-
-
Jiapeng Chong authored
The function hrtimer_hres_active() are defined in the hrtimer.c file, but not called elsewhere, so rename __hrtimer_hres_active() to hrtimer_hres_active() and remove the old hrtimer_hres_active() function. kernel/time/hrtimer.c:653:19: warning: unused function 'hrtimer_hres_active'. Fixes: 82ccdf06 ("hrtimer: Remove unused function") Reported-by:
Abaci Robot <abaci@linux.alibaba.com> Signed-off-by:
Jiapeng Chong <jiapeng.chong@linux.alibaba.com> Signed-off-by:
Anna-Maria Behnsen <anna-maria@linutronix.de> Link: https://lore.kernel.org/r/20240418023000.130324-1-jiapeng.chong@linux.alibaba.com Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=8778
-
Anna-Maria Behnsen authored
This function was introduced with commit 60bda037 ("posix-cpu-timers: Utilize timerqueue for storage") but never used. Remove it. Signed-off-by:
-
- 11 Apr, 2024 1 commit
-
-
Alice Ryhl authored
Introduce a wrapper around `ktime_t` with a few different useful methods. Rust Binder will use these bindings to compute how many milliseconds a transaction has been active for when dumping the current state of the Binder driver. This replicates the logic in C Binder [1]. For a usage example in Rust Binder, see [2]. ktime_get() cannot be safely called in NMI context. This requirement is not checked by these abstractions, but it is intended that klint [3] or a similar tool will be used to check it in the future. Signed-off-by:
Benno Lossin <benno.lossin@proton.me> Reviewed-by:
-
- 09 Apr, 2024 3 commits
-
-
Adrian Hunter authored
U64_MAX is not in include/vdso/limits.h, although that isn't noticed on x86 because x86 includes include/linux/limits.h indirectly. However powerpc is more selective, resulting in the following build error: In file included from <command-line>: lib/vdso/gettimeofday.c: In function 'vdso_calc_ns': lib/vdso/gettimeofday.c:11:33: error: 'U64_MAX' undeclared 11 | # define VDSO_DELTA_MASK(vd) U64_MAX | ^~~~~~~ Use ULLONG_MAX instead which will work just as well and is in include/vdso/limits.h. Fixes: c8e3a8b6 ("vdso: Consolidate vdso_calc_delta()") Reported-by:Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
Adrian Hunter <adrian.hunter@intel.com> Signed-off-by:
-
Li Zhijian authored
Per filesystems/sysfs.rst, show() should only use sysfs_emit() or sysfs_emit_at() when formatting the value to be returned to user space. coccinelle complains that there are still a couple of functions that use snprintf(). Convert them to sysfs_emit(). Signed-off-by:
Li Zhijian <lizhijian@fujitsu.com> Signed-off-by:
-
Li Zhijian authored
Per filesystems/sysfs.rst, show() should only use sysfs_emit() or sysfs_emit_at() when formatting the value to be returned to user space. coccinelle complains that there are still a couple of functions that use snprintf(). Convert them to sysfs_emit(). Signed-off-by:
-
- 08 Apr, 2024 20 commits
-
-
Adrian Hunter authored
Kernel timekeeping is designed to keep the change in cycles (since the last timer interrupt) below max_cycles, which prevents multiplication overflow when converting cycles to nanoseconds. However, if timer interrupts stop, the clocksource_cyc2ns() calculation will eventually overflow. Add protection against that. Simplify by folding together clocksource_delta() and clocksource_cyc2ns() into cycles_to_nsec_safe(). Check against max_cycles, falling back to a slower higher precision calculation. Suggested-by:
-
Adrian Hunter authored
For the case !CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE, forego overflow protection in the range (mask << 1) < delta <= mask, and interpret it always as an inconsistency between CPU clock values. That allows slightly neater code, and it is on a slow path so has no effect on performance. Suggested-by:
-
Adrian Hunter authored
Kernel timekeeping is designed to keep the change in cycles (since the last timer interrupt) below max_cycles, which prevents multiplication overflow when converting cycles to nanoseconds. However, if timer interrupts stop, the calculation will eventually overflow. Add protection against that. In timekeeping_cycles_to_ns() calculation, check against max_cycles, falling back to a slower higher precision calculation. In timekeeping_forward_now(), process delta in chunks of at most max_cycles. Suggested-by:
-
Adrian Hunter authored
Open code clocksource_delta() in timekeeping_cycles_to_ns() so that overflow safety can be added efficiently. Suggested-by:
-
Adrian Hunter authored
timekeeping_delta_to_ns() is now called only from timekeeping_cycles_to_ns(), and it is not useful otherwise. Simplify the code by folding it into timekeeping_cycles_to_ns(). No functional change. Suggested-by:
-
Adrian Hunter authored
Consolidate timekeeping helpers, making use of timekeeping_cycles_to_ns() in preference to directly using timekeeping_delta_to_ns(). No functional change. Suggested-by:
-
Adrian Hunter authored
Simplify the usage of timekeeping sanity checking, in preparation for consolidating timekeeping helpers. This works towards eliminating timekeeping_delta_to_ns() in favour of timekeeping_cycles_to_ns(). No functional change. Suggested-by:
-
Adrian Hunter authored
Simplify __timekeeping_get_ns() by reusing timekeeping_cycles_to_ns(). No functional change. Suggested-by:
-
Adrian Hunter authored
Put together declaration and initialization of the local variable 'delta'. Suggested-by:
-
Adrian Hunter authored
Rename fast_tk_get_delta_ns() to __timekeeping_get_ns() to prepare for its reuse as a general timekeeping helper function. Suggested-by:
-
Adrian Hunter authored
Move timekeeping helper functions to prepare for their reuse. Suggested-by:
-
Adrian Hunter authored
Kernel timekeeping is designed to keep the change in cycles (since the last timer interrupt) below max_cycles, which prevents multiplication overflow when converting cycles to nanoseconds. However, if timer interrupts stop, the calculation will eventually overflow. Add protection against that. Select GENERIC_VDSO_OVERFLOW_PROTECT so that max_cycles is made available in the VDSO data page. Check against max_cycles, falling back to a slower higher precision calculation. Take advantage of the opportunity to move masking and negative motion check into the slow path. The result is a calculation that has similar performance as before. Newer machines showed performance benefit, whereas older Skylake-based hardware such as Intel Kaby Lake was seen <1% worse. Suggested-by:
-
Adrian Hunter authored
Kernel timekeeping is designed to keep the change in cycles (since the last timer interrupt) below max_cycles, which prevents multiplication overflow when converting cycles to nanoseconds. However, if timer interrupts stop, the calculation will eventually overflow. Add protection against that, enabled by config option CONFIG_GENERIC_VDSO_OVERFLOW_PROTECT. Check against max_cycles, falling back to a slower higher precision calculation. Suggested-by:
-
Adrian Hunter authored
Add vdso_data::max_cycles in preparation to use it to detect potential multiplication overflow. Suggested-by:
-
Adrian Hunter authored
Provide mul_u64_u32_add_u64_shr() which is a calculation that will be used by timekeeping and VDSO. Place #include <vdso/math64.h> after #include <asm/div64.h> to allow architecture-specific overrides, at least for the kernel. Signed-off-by:
-
Adrian Hunter authored
Put together declaration and initialization of local variables. Suggested-by:
-
Adrian Hunter authored
Add CONFIG_GENERIC_VDSO_OVERFLOW_PROTECT in preparation to add multiplication overflow protection to the VDSO time getter functions. Suggested-by:
-
Adrian Hunter authored
Consolidate nanoseconds calculation to simplify and reduce code duplication. Suggested-by:
-
Adrian Hunter authored
Consolidate vdso_calc_delta(), in preparation for further simplification. Suggested-by:
-
Jiapeng Chong authored
The function is defined, but not called anywhere: kernel/time/hrtimer.c:1880:20: warning: unused function '__hrtimer_peek_ahead_timers'. Remove it. Reported-by:
-
- 07 Apr, 2024 4 commits
-
-
Linus Torvalds authored
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Ingo Molnar: - Fix MCE timer reinit locking - Fix/improve CoCo guest random entropy pool init - Fix SEV-SNP late disable bugs - Fix false positive objtool build warning - Fix header dependency bug - Fix resctrl CPU offlining bug * tag 'x86-urgent-2024-04-07' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/retpoline: Add NOENDBR annotation to the SRSO dummy return thunk x86/mce: Make sure to grab mce_sysfs_mutex in set_bank() x86/CPU/AMD: Track SNP host status with cc_platform_*() x86/cc: Add cc_platform_set/_clear() helpers x86/kvm/Kconfig: Have KVM_AMD_SEV select ARCH_HAS_CC_PLATFORM x86/coco: Require seeding RNG with RDRAND on CoCo systems x86/numa/32: Include missing <asm/pgtable_areas.h> x86/resctrl: Fix uninitialized memory read when last CPU of domain goes offline
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull timer fixes from Ingo Molnar: "Fix various timer bugs: - Fix a timer migration bug that may result in missed events - Fix timer migration group hierarchy event updates - Fix a PowerPC64 build warning - Fix a handful of DocBook annotation bugs" * tag 'timers-urgent-2024-04-07' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: timers/migration: Return early on deactivation timers/migration: Fix ignored event due to missing CPU update vdso: Use CONFIG_PAGE_SHIFT in vdso/datapage.h timers: Fix text inconsistencies and spelling tick/sched: Fix struct tick_sched doc warnings tick/sched: Fix various kernel-doc warnings timers: Fix kernel-doc format and add Return values time/timekeeping: Fix kernel-doc warnings and typos time/timecounter: Fix inline documentation
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 perf fix from Ingo Molnar: "Fix a combined PEBS events bug on x86 Intel CPUs" * tag 'perf-urgent-2024-04-07' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: perf/x86/intel/ds: Don't clear ->pebs_data_cfg for the last PEBS event
-
- 06 Apr, 2024 2 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/cel/linuxLinus Torvalds authored
Pull nfsd fixes from Chuck Lever: - Address a slow memory leak with RPC-over-TCP - Prevent another NFS4ERR_DELAY loop during CREATE_SESSION * tag 'nfsd-6.9-2' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux: nfsd: hold a lighter-weight client reference over CB_RECALL_ANY SUNRPC: Fix a slow server-side memory leak with RPC-over-TCP
-
git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linuxLinus Torvalds authored
Pull i2c fix from Wolfram Sang: "A host driver build fix" * tag 'i2c-for-6.9-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux: i2c: pxa: hide unused icr_bits[] variable
-
