[PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Wangyang Guo]

Read-mostly sched_clock_irqtime may share the same cacheline with
frequently updated nohz struct. Mark it as __read_mostly to avoid
false sharing issue.

Details:
We observed ~3% cycles hotspots in irqtime_account_irq when running
SPECjbb2015 in a 2-sockets system. Most of cycles spent in reading
sched_clock_irqtime, which is a read-mostly var.

perf c2c (cachelien view) shows it has false sharing with nohz struct:
     Num RmtHitm LclHitm  Offset records             Symbol
   6.25%   0.00%   0.00%   0x0       4   [k] _nohz_idle_balance.isra.0
  18.75% 100.00%   0.00%   0x8      14   [k] nohz_balance_exit_idle
   6.25%   0.00%   0.00%   0x8       8   [k] nohz_balance_enter_idle
   6.25%   0.00%   0.00%   0xc       8   [k] sched_balance_newidle
   6.25%   0.00%   0.00%  0x10      31   [k] nohz_balancer_kick
   6.25%   0.00%   0.00%  0x20      16   [k] sched_balance_newidle
  37.50%   0.00%   0.00%  0x38      50   [k] irqtime_account_irq
   6.25%   0.00%   0.00%  0x38      47   [k] account_process_tick
   6.25%   0.00%   0.00%  0x38      12   [k] account_idle_ticks

Offsets:
*  0x0 -- nohz.idle_cpu_mask (r)
*  0x8 -- nohz.nr_cpus (w)
* 0x38 -- sched_clock_irqtime (r), not in nohz, but share cacheline

The layout in /proc/kallsyms can also confirm that:
ffffffff88600d40 b nohz
ffffffff88600d68 B arch_needs_tick_broadcast
ffffffff88600d6c b __key.264
ffffffff88600d6c b __key.265
ffffffff88600d70 b dl_generation
ffffffff88600d78 b sched_clock_irqtime

With the patch applied, irqtime_account_irq hotspot disappear.

Changes since V1:
- Use __read_mostly instead of __cacheline_aligned to avoid wasting
  spaces.

History:
  v1: https://lore.kernel.org/all/20260113022958.3379650-1-wangyang.guo@intel.com/
  prev discussions: https://lore.kernel.org/all/20251211055612.4071266-1-wangyang.guo@intel.com/T/#u

Suggested-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Reported-by: Benjamin Lei <benjamin.lei@intel.com>
Reviewed-by: Tim Chen <tim.c.chen@linux.intel.com>
Reviewed-by: Tianyou Li <tianyou.li@intel.com>
Signed-off-by: Wangyang Guo <wangyang.guo@intel.com>
---
 kernel/sched/cputime.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 7097de2c8cda..fc8f8081c48f 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -25,7 +25,7 @@
  */
 DEFINE_PER_CPU(struct irqtime, cpu_irqtime);
 
-int sched_clock_irqtime;
+int sched_clock_irqtime __read_mostly;
 
 void enable_sched_clock_irqtime(void)
 {
-- 
2.47.3

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Shrikanth Hegde]

On 1/13/26 1:18 PM, Wangyang Guo wrote:
> Read-mostly sched_clock_irqtime may share the same cacheline with
> frequently updated nohz struct. Mark it as __read_mostly to avoid
> false sharing issue.
> 
> Details:
> We observed ~3% cycles hotspots in irqtime_account_irq when running
> SPECjbb2015 in a 2-sockets system. Most of cycles spent in reading
> sched_clock_irqtime, which is a read-mostly var.
> 
> perf c2c (cachelien view) shows it has false sharing with nohz struct:
>       Num RmtHitm LclHitm  Offset records             Symbol
>     6.25%   0.00%   0.00%   0x0       4   [k] _nohz_idle_balance.isra.0
>    18.75% 100.00%   0.00%   0x8      14   [k] nohz_balance_exit_idle
>     6.25%   0.00%   0.00%   0x8       8   [k] nohz_balance_enter_idle
>     6.25%   0.00%   0.00%   0xc       8   [k] sched_balance_newidle
>     6.25%   0.00%   0.00%  0x10      31   [k] nohz_balancer_kick
>     6.25%   0.00%   0.00%  0x20      16   [k] sched_balance_newidle
>    37.50%   0.00%   0.00%  0x38      50   [k] irqtime_account_irq
>     6.25%   0.00%   0.00%  0x38      47   [k] account_process_tick
>     6.25%   0.00%   0.00%  0x38      12   [k] account_idle_ticks
> 
> Offsets:
> *  0x0 -- nohz.idle_cpu_mask (r)
> *  0x8 -- nohz.nr_cpus (w)
> * 0x38 -- sched_clock_irqtime (r), not in nohz, but share cacheline
> 
> The layout in /proc/kallsyms can also confirm that:
> ffffffff88600d40 b nohz
> ffffffff88600d68 B arch_needs_tick_broadcast
> ffffffff88600d6c b __key.264
> ffffffff88600d6c b __key.265
> ffffffff88600d70 b dl_generation
> ffffffff88600d78 b sched_clock_irqtime
> 
> With the patch applied, irqtime_account_irq hotspot disappear.
> 
> Changes since V1:
> - Use __read_mostly instead of __cacheline_aligned to avoid wasting
>    spaces.
> 
> History:
>    v1: https://lore.kernel.org/all/20260113022958.3379650-1-wangyang.guo@intel.com/
>    prev discussions: https://lore.kernel.org/all/20251211055612.4071266-1-wangyang.guo@intel.com/T/#u
> 
> Suggested-by: Shrikanth Hegde <sshegde@linux.ibm.com>

I think it would appropriate to take prateek's tag here.

> Reported-by: Benjamin Lei <benjamin.lei@intel.com>
> Reviewed-by: Tim Chen <tim.c.chen@linux.intel.com>
> Reviewed-by: Tianyou Li <tianyou.li@intel.com>
> Signed-off-by: Wangyang Guo <wangyang.guo@intel.com>
> ---
>   kernel/sched/cputime.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
> index 7097de2c8cda..fc8f8081c48f 100644
> --- a/kernel/sched/cputime.c
> +++ b/kernel/sched/cputime.c
> @@ -25,7 +25,7 @@
>    */
>   DEFINE_PER_CPU(struct irqtime, cpu_irqtime);
>   
> -int sched_clock_irqtime;
> +int sched_clock_irqtime __read_mostly;
>   
>   void enable_sched_clock_irqtime(void)
>   {

Patch should be good on its own.

Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Peter Zijlstra]

On Tue, Jan 13, 2026 at 03:48:07PM +0800, Wangyang Guo wrote:
> Read-mostly sched_clock_irqtime may share the same cacheline with
> frequently updated nohz struct. Mark it as __read_mostly to avoid
> false sharing issue.
> 

Is there a reason that thing can't be a static_key or so?

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[K Prateek Nayak]

Hello Peter,

On 1/13/2026 3:37 PM, Peter Zijlstra wrote:
> On Tue, Jan 13, 2026 at 03:48:07PM +0800, Wangyang Guo wrote:
>> Read-mostly sched_clock_irqtime may share the same cacheline with
>> frequently updated nohz struct. Mark it as __read_mostly to avoid
>> false sharing issue.
>>
> 
> Is there a reason that thing can't be a static_key or so?

Yes, the exact case is stated in commit b9f2b29b9494
("sched: Don't define sched_clock_irqtime as static key")
https://lore.kernel.org/lkml/173961695743.10177.17683780278419896262.tip-bot2@tip-bot2/

-- 
Thanks and Regards,
Prateek

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Vincent Guittot]

On Tue, 13 Jan 2026 at 11:08, Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Tue, Jan 13, 2026 at 03:48:07PM +0800, Wangyang Guo wrote:
> > Read-mostly sched_clock_irqtime may share the same cacheline with
> > frequently updated nohz struct. Mark it as __read_mostly to avoid
> > false sharing issue.
> >
>
> Is there a reason that thing can't be a static_key or so?

We tried that but disable_sched_clock_irqtime can be called in atomic
context with mark_tsc_unstable()
https://lore.kernel.org/all/174161357383.14745.8770394914047302959.tip-bot2@tip-bot2/

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Peter Zijlstra]

On Tue, Jan 13, 2026 at 11:14:54AM +0100, Vincent Guittot wrote:
> On Tue, 13 Jan 2026 at 11:08, Peter Zijlstra <peterz@infradead.org> wrote:
> >
> > On Tue, Jan 13, 2026 at 03:48:07PM +0800, Wangyang Guo wrote:
> > > Read-mostly sched_clock_irqtime may share the same cacheline with
> > > frequently updated nohz struct. Mark it as __read_mostly to avoid
> > > false sharing issue.
> > >
> >
> > Is there a reason that thing can't be a static_key or so?
> 
> We tried that but disable_sched_clock_irqtime can be called in atomic
> context with mark_tsc_unstable()
> https://lore.kernel.org/all/174161357383.14745.8770394914047302959.tip-bot2@tip-bot2/

Right, but if you note, that very same path has
clear_sched_clock_stable(), and that is a static_key too.

It's just that that punts by pushing it into a workqueue. So why not
lift that workqueue and do both keys or something?

Re: [PATCH v2] sched/clock: Avoid false sharing for sched_clock_irqtime

[Vincent Guittot]

On Tue, 13 Jan 2026 at 11:36, Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Tue, Jan 13, 2026 at 11:14:54AM +0100, Vincent Guittot wrote:
> > On Tue, 13 Jan 2026 at 11:08, Peter Zijlstra <peterz@infradead.org> wrote:
> > >
> > > On Tue, Jan 13, 2026 at 03:48:07PM +0800, Wangyang Guo wrote:
> > > > Read-mostly sched_clock_irqtime may share the same cacheline with
> > > > frequently updated nohz struct. Mark it as __read_mostly to avoid
> > > > false sharing issue.
> > > >
> > >
> > > Is there a reason that thing can't be a static_key or so?
> >
> > We tried that but disable_sched_clock_irqtime can be called in atomic
> > context with mark_tsc_unstable()
> > https://lore.kernel.org/all/174161357383.14745.8770394914047302959.tip-bot2@tip-bot2/
>
> Right, but if you note, that very same path has
> clear_sched_clock_stable(), and that is a static_key too.
>
> It's just that that punts by pushing it into a workqueue. So why not
> lift that workqueue and do both keys or something?

fair enough