Re: [PATCH 04/15] powerpc/time: Prepare to stop elapsing in dynticks-idle

["Christophe Leroy (CS GROUP)" <chleroy@kernel.org>]

Hi Hegde,

Le 25/02/2026 à 08:46, Shrikanth Hegde a écrit :
> Hi Christophe,
> 
> On 2/24/26 9:11 PM, Christophe Leroy (CS GROUP) wrote:
>> Hi Hegde,
>>
>> Le 19/02/2026 à 19:30, Shrikanth Hegde a écrit :
>>>
>>>
>>> On 2/6/26 7:52 PM, Frederic Weisbecker wrote:
>>>> Currently the tick subsystem stores the idle cputime accounting in
>>>> private fields, allowing cohabitation with architecture idle vtime
>>>> accounting. The former is fetched on online CPUs, the latter on offline
>>>> CPUs.
>>>>
>>>> For consolidation purpose, architecture vtime accounting will continue
>>>> to account the cputime but will make a break when the idle tick is
>>>> stopped. The dyntick cputime accounting will then be relayed by the 
>>>> tick
>>>> subsystem so that the idle cputime is still seen advancing coherently
>>>> even when the tick isn't there to flush the idle vtime.
>>>>
>>>> Prepare for that and introduce three new APIs which will be used in
>>>> subsequent patches:
>>>>
>>>> _ vtime_dynticks_start() is deemed to be called when idle enters in
>>>>    dyntick mode. The idle cputime that elapsed so far is accumulated.
>>>>
>>>> - vtime_dynticks_stop() is deemed to be called when idle exits from
>>>>    dyntick mode. The vtime entry clocks are fast-forward to current 
>>>> time
>>>>    so that idle accounting restarts elapsing from now.
>>>>
>>>> - vtime_reset() is deemed to be called from dynticks idle IRQ entry to
>>>>    fast-forward the clock to current time so that the IRQ time is still
>>>>    accounted by vtime while nohz cputime is paused.
>>>>
>>>> Also accumulated vtime won't be flushed from dyntick-idle ticks to 
>>>> avoid
>>>> accounting twice the idle cputime, along with nohz accounting.
>>>>
>>>> Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
>>>
>>> Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>
>>>
>>>> ---
>>>>   arch/powerpc/kernel/time.c | 41 ++++++++++++++++++++++++++++++++++ 
>>>> ++++
>>>>   include/linux/vtime.h      |  6 ++++++
>>>>   2 files changed, 47 insertions(+)
>>>>
>>>> diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
>>>> index 4bbeb8644d3d..18506740f4a4 100644
>>>> --- a/arch/powerpc/kernel/time.c
>>>> +++ b/arch/powerpc/kernel/time.c
>>>> @@ -376,6 +376,47 @@ void vtime_task_switch(struct task_struct *prev)
>>>>           acct->starttime = acct0->starttime;
>>>>       }
>>>>   }
>>>> +
>>>> +#ifdef CONFIG_NO_HZ_COMMON
>>>> +/**
>>>> + * vtime_reset - Fast forward vtime entry clocks
>>>> + *
>>>> + * Called from dynticks idle IRQ entry to fast-forward the clocks 
>>>> to current time
>>>> + * so that the IRQ time is still accounted by vtime while nohz 
>>>> cputime is paused.
>>>> + */
>>>> +void vtime_reset(void)
>>>> +{
>>>> +    struct cpu_accounting_data *acct = get_accounting(current);
>>>> +
>>>> +    acct->starttime = mftb();
>>>
>>> I figured out why those huge values happen.
>>>
>>> This happens because mftb is from when the system is booted.
>>> I was doing kexec to start the new kernel and mftb wasn't getting
>>> reset.
>>>
>>> I thought about this. This is concern for pseries too, where LPAR's
>>> restart but system won't restart and mftb will continue to run 
>>> instead of
>>> reset.
>>>
>>> I think we should be using sched_clock instead of mftb here.
>>> Though we need it a few more places and some cosmetic changes around it.
>>>
>>> Note: Some values being huge exists without series for few CPUs, with 
>>> series it
>>> shows up in most of the CPUs.
>>>
>>> So I am planning send out fix below fix separately keeping your
>>> series as dependency.
>>>
>>> ---
>>>   arch/powerpc/include/asm/accounting.h |  4 ++--
>>>   arch/powerpc/include/asm/cputime.h    | 14 +++++++-------
>>>   arch/powerpc/kernel/time.c            | 22 +++++++++++-----------
>>>   3 files changed, 20 insertions(+), 20 deletions(-)
>>>
>>> diff --git a/arch/powerpc/include/asm/accounting.h b/arch/powerpc/ 
>>> include/asm/accounting.h
>>> index 6d79c31700e2..50f120646e6d 100644
>>> --- a/arch/powerpc/include/asm/accounting.h
>>> +++ b/arch/powerpc/include/asm/accounting.h
>>> @@ -21,8 +21,8 @@ struct cpu_accounting_data {
>>>       unsigned long steal_time;
>>>       unsigned long idle_time;
>>>       /* Internal counters */
>>> -    unsigned long starttime;    /* TB value snapshot */
>>> -    unsigned long starttime_user;    /* TB value on exit to usermode */
>>> +    unsigned long starttime;    /* Time value snapshot */
>>> +    unsigned long starttime_user;    /* Time value on exit to 
>>> usermode */
>>>   #ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
>>>       unsigned long startspurr;    /* SPURR value snapshot */
>>>       unsigned long utime_sspurr;    /* ->user_time when ->startspurr 
>>> set */
>>> diff --git a/arch/powerpc/include/asm/cputime.h b/arch/powerpc/ 
>>> include/ asm/cputime.h
>>> index aff858ca99c0..eb6b629b113f 100644
>>> --- a/arch/powerpc/include/asm/cputime.h
>>> +++ b/arch/powerpc/include/asm/cputime.h
>>> @@ -20,9 +20,9 @@
>>>   #include <asm/time.h>
>>>   #include <asm/param.h>
>>>   #include <asm/firmware.h>
>>> +#include <linux/sched/clock.h>
>>>
>>>   #ifdef __KERNEL__
>>> -#define cputime_to_nsecs(cputime) tb_to_ns(cputime)
>>>
>>>   /*
>>>    * PPC64 uses PACA which is task independent for storing accounting 
>>> data while
>>> @@ -44,20 +44,20 @@
>>>    */
>>>   static notrace inline void account_cpu_user_entry(void)
>>>   {
>>> -    unsigned long tb = mftb();
>>> +    unsigned long now = sched_clock();
>>
>> Now way !
>>
>> By doing that you'll kill performance for no reason. All we need when 
>> accounting time spent in kernel or in user is the difference between 
>> time at entry and time at exit, no mater what the time was at boot time.
>>
> 
> No. With this patch there will not be any performance difference.
> All it does is, instead of using mftb uses sched_clock at those places.
> 

For the record, I did some benchmark test with 
tools/testing/selftests/powerpc/benchmarks/null_syscall on powerpc 885 
microcontroller:

Without your proposed patch:

root@vgoip:~# ./null_syscall
    2729.98 ns     360.36 cycles

With your proposed patch below:

root@vgoip:~# ./null_syscall
    3370.80 ns     444.95 cycles

So as expected it is a huge regression, almost 25% more time to run the 
syscall.

Christophe


> 
> In arch/powerpc/kernel/time.c we have sched_clock().
> notrace unsigned long long sched_clock(void)
> {
>          return mulhdu(get_tb() - boot_tb, tb_to_ns_scale) << 
> tb_to_ns_shift;
> }
> 
> It does the same mftb call, and accounts only the time after boot, which is
> what /proc/stat should do as well.
> 
> "
> the amount of time, measured in units of USER_HZ
> (1/100ths of a second on most architectures
> 
> user   (1) Time spent in user mode.
> 
> idle   (4) Time spent in the idle task.  This value
>         should be USER_HZ times the second entry in
>         the /proc/uptime pseudo-file.
> "
> /proc/uptime is based on sched_clock, so i infer /proc/stat also should 
> show
> values w.r.t to boot of the OS.
> 
> 
>> Also sched_clock() returns nanoseconds which implies calculation from 
>> timebase. This is pointless CPU consumption. The current 
>> implementation calculates nanoseconds at task switch when calling 
>> vtime_flush().Your change will now do it at every kernel entry and 
>> kernel exit by calling sched_clock().
> 
> This change doesn't add any additional paths. Even without patches, mftb 
> would have
> been called in every kernel entry/exit.  See mftb usage 
> account_cpu_user_exit/enter
> 
> Now instead of mftb sched_clock is used, that's all. No additional 
> entry/exit points.
> And previously when accounting we would have done cputime_to_nsecs, now 
> that conversion
> is done automatically in sched_clock. So overall computation-wise it 
> should be same.
> 
> What i am missing to see it here?
> 
>>
>> Another point is that sched_clock() returns a long long not a long.
> 
> Thanks for pointing that out.
> 
> Ok. Let me change some of those variables into unsigned long long.
> Compiler didn't warn me, so i didn't see it.
> 
>>
>> And also sched_clock() uses get_tb() which does mftb and mftbu. Which 
>> is pointless for calculating time deltas unless your application 
>> spends hours without being re-scheduled.
>>
> 
> I didn't get this. At current also, we use mftb, that functionality 
> should be the same.
> Could you please explain how?
> 
>>
>>>       struct cpu_accounting_data *acct = raw_get_accounting(current);
>>>
>>> -    acct->utime += (tb - acct->starttime_user);
>>> -    acct->starttime = tb;
>>> +    acct->utime += (now - acct->starttime_user);
>>> +    acct->starttime = now;
>>>   }
>>>
>>>   static notrace inline void account_cpu_user_exit(void)
>>>   {
>>> -    unsigned long tb = mftb();
>>> +    unsigned long now = sched_clock();
>>>       struct cpu_accounting_data *acct = raw_get_accounting(current);
>>>
>>> -    acct->stime += (tb - acct->starttime);
>>> -    acct->starttime_user = tb;
>>> +    acct->stime += (now - acct->starttime);
>>> +    acct->starttime_user = now;
>>>   }
>>>
>>>   static notrace inline void account_stolen_time(void)
>>> diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
>>> index 18506740f4a4..fb67cdae3bcb 100644
>>> --- a/arch/powerpc/kernel/time.c
>>> +++ b/arch/powerpc/kernel/time.c
>>> @@ -215,7 +215,7 @@ static unsigned long vtime_delta(struct 
>>> cpu_accounting_data *acct,
>>>
>>>       WARN_ON_ONCE(!irqs_disabled());
>>>
>>> -    now = mftb();
>>> +    now = sched_clock();
>>>       stime = now - acct->starttime;
>>>       acct->starttime = now;
>>>
>>> @@ -299,9 +299,9 @@ static void vtime_flush_scaled(struct task_struct 
>>> *tsk,
>>>   {
>>>   #ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
>>>       if (acct->utime_scaled)
>>> -        tsk->utimescaled += cputime_to_nsecs(acct->utime_scaled);
>>> +        tsk->utimescaled += acct->utime_scaled;
>>>       if (acct->stime_scaled)
>>> -        tsk->stimescaled += cputime_to_nsecs(acct->stime_scaled);
>>> +        tsk->stimescaled += acct->stime_scaled;
>>>
>>>       acct->utime_scaled = 0;
>>>       acct->utime_sspurr = 0;
>>> @@ -321,28 +321,28 @@ void vtime_flush(struct task_struct *tsk)
>>>       struct cpu_accounting_data *acct = get_accounting(tsk);
>>>
>>>       if (acct->utime)
>>> -        account_user_time(tsk, cputime_to_nsecs(acct->utime));
>>> +        account_user_time(tsk, acct->utime);
>>>
>>>       if (acct->gtime)
>>> -        account_guest_time(tsk, cputime_to_nsecs(acct->gtime));
>>> +        account_guest_time(tsk, acct->gtime);
>>>
>>>       if (IS_ENABLED(CONFIG_PPC_SPLPAR) && acct->steal_time) {
>>> -        account_steal_time(cputime_to_nsecs(acct->steal_time));
>>> +        account_steal_time(acct->steal_time);
>>>           acct->steal_time = 0;
>>>       }
>>>
>>>       if (acct->idle_time)
>>> -        account_idle_time(cputime_to_nsecs(acct->idle_time));
>>> +        account_idle_time(acct->idle_time);
>>>
>>>       if (acct->stime)
>>> -        account_system_index_time(tsk, cputime_to_nsecs(acct->stime),
>>> +        account_system_index_time(tsk, acct->stime,
>>>                         CPUTIME_SYSTEM);
>>>
>>>       if (acct->hardirq_time)
>>> -        account_system_index_time(tsk, cputime_to_nsecs(acct- 
>>>  >hardirq_time),
>>> +        account_system_index_time(tsk, acct->hardirq_time,
>>>                         CPUTIME_IRQ);
>>>       if (acct->softirq_time)
>>> -        account_system_index_time(tsk, cputime_to_nsecs(acct- 
>>>  >softirq_time),
>>> +        account_system_index_time(tsk, acct->softirq_time,
>>>                         CPUTIME_SOFTIRQ);
>>>
>>>       vtime_flush_scaled(tsk, acct);
>>> @@ -388,7 +388,7 @@ void vtime_reset(void)
>>>   {
>>>       struct cpu_accounting_data *acct = get_accounting(current);
>>>
>>> -    acct->starttime = mftb();
>>> +    acct->starttime = sched_clock();
>>>   #ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
>>>       acct->startspurr = read_spurr(acct->starttime);
>>>   #endif
>>
> 
> PS: I measured the performance with hackbench. I don't see any degradation.
>