From da6b85598af30e9fec34d82882d7e1e39f3da769 Mon Sep 17 00:00:00 2001 From: Yongliang Gao Date: Sun, 16 Feb 2025 16:41:09 +0800 Subject: rcu/cpu_stall_cputime: fix the hardirq count for x86 architecture When counting the number of hardirqs in the x86 architecture, it is essential to add arch_irq_stat_cpu to ensure accuracy. For example, a CPU loop within the rcu_read_lock function. Before: [ 70.910184] rcu: INFO: rcu_preempt self-detected stall on CPU [ 70.910436] rcu: 3-....: (4999 ticks this GP) idle=*** [ 70.910711] rcu: hardirqs softirqs csw/system [ 70.910870] rcu: number: 0 657 0 [ 70.911024] rcu: cputime: 0 0 2498 ==> 2498(ms) [ 70.911278] rcu: (t=5001 jiffies g=3677 q=29 ncpus=8) After: [ 68.046132] rcu: INFO: rcu_preempt self-detected stall on CPU [ 68.046354] rcu: 2-....: (4999 ticks this GP) idle=*** [ 68.046628] rcu: hardirqs softirqs csw/system [ 68.046793] rcu: number: 2498 663 0 [ 68.046951] rcu: cputime: 0 0 2496 ==> 2496(ms) [ 68.047244] rcu: (t=5000 jiffies g=3825 q=4 ncpus=8) Fixes: be42f00b73a0 ("rcu: Add RCU stall diagnosis information") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202501090842.SfI6QPGS-lkp@intel.com/ Signed-off-by: Yongliang Gao Reviewed-by: Neeraj Upadhyay Link: https://lore.kernel.org/r/20250216084109.3109837-1-leonylgao@gmail.com Signed-off-by: Boqun Feng Signed-off-by: Joel Fernandes --- kernel/rcu/tree_stall.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/rcu/tree_stall.h') diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 925fcdad5dea..56b21219442b 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -435,8 +435,8 @@ static void print_cpu_stat_info(int cpu) rsr.cputime_system = kcpustat_field(kcsp, CPUTIME_SYSTEM, cpu); pr_err("\t hardirqs softirqs csw/system\n"); - pr_err("\t number: %8ld %10d %12lld\n", - kstat_cpu_irqs_sum(cpu) - rsrp->nr_hardirqs, + pr_err("\t number: %8lld %10d %12lld\n", + kstat_cpu_irqs_sum(cpu) + arch_irq_stat_cpu(cpu) - rsrp->nr_hardirqs, kstat_cpu_softirqs_sum(cpu) - rsrp->nr_softirqs, nr_context_switches_cpu(cpu) - rsrp->nr_csw); pr_err("\tcputime: %8lld %10lld %12lld ==> %d(ms)\n", -- cgit v1.2.3 From 2da20fd904f87f7bb31b79719bc3dda4093f8cdb Mon Sep 17 00:00:00 2001 From: Max Kellermann Date: Sun, 4 May 2025 20:08:31 +0200 Subject: kernel/rcu/tree_stall: add /sys/kernel/rcu_stall_count Expose a simple counter to userspace for monitoring tools. (akpm: 2536c5c7d6ae added the documentation but the code changes were lost) Link: https://lkml.kernel.org/r/20250504180831.4190860-3-max.kellermann@ionos.com Fixes: 2536c5c7d6ae ("kernel/rcu/tree_stall: add /sys/kernel/rcu_stall_count") Signed-off-by: Max Kellermann Cc: Core Minyard Cc: Doug Anderson Cc: Joel Granados Cc: Max Kellermann Cc: Song Liu Cc: Sourabh Jain Signed-off-by: Andrew Morton --- kernel/rcu/tree_stall.h | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) (limited to 'kernel/rcu/tree_stall.h') diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 56b21219442b..486c00536207 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -20,6 +20,28 @@ int sysctl_panic_on_rcu_stall __read_mostly; int sysctl_max_rcu_stall_to_panic __read_mostly; +#ifdef CONFIG_SYSFS + +static unsigned int rcu_stall_count; + +static ssize_t rcu_stall_count_show(struct kobject *kobj, struct kobj_attribute *attr, + char *page) +{ + return sysfs_emit(page, "%u\n", rcu_stall_count); +} + +static struct kobj_attribute rcu_stall_count_attr = __ATTR_RO(rcu_stall_count); + +static __init int kernel_rcu_stall_sysfs_init(void) +{ + sysfs_add_file_to_group(kernel_kobj, &rcu_stall_count_attr.attr, NULL); + return 0; +} + +late_initcall(kernel_rcu_stall_sysfs_init); + +#endif // CONFIG_SYSFS + #ifdef CONFIG_PROVE_RCU #define RCU_STALL_DELAY_DELTA (5 * HZ) #else @@ -784,6 +806,10 @@ static void check_cpu_stall(struct rcu_data *rdp) if (kvm_check_and_clear_guest_paused()) return; +#ifdef CONFIG_SYSFS + ++rcu_stall_count; +#endif + rcu_stall_notifier_call_chain(RCU_STALL_NOTIFY_NORM, (void *)j - gps); if (READ_ONCE(csd_lock_suppress_rcu_stall) && csd_lock_is_stuck()) { pr_err("INFO: %s detected stall, but suppressed full report due to a stuck CSD-lock.\n", rcu_state.name); -- cgit v1.2.3