Re: [PATCH v15 12/16] cpuidle/poll_state: Wait for need-resched via tif_need_resched_relaxed_wait()
From: Ankur Arora
Date: Fri Sep 11 2026 - 03:21:39 EST
Haris Okanovic <harisokn@xxxxxxxxxx> writes:
> On Mon, 31 Aug 2026 13:22:47 -0700, Ankur Arora <ankur.a.arora@xxxxxxxxxx> wrote:
>> diff --git a/drivers/cpuidle/poll_state.c b/drivers/cpuidle/poll_state.c
>> index c7524e4c522a2..7443b3e971baf 100644
>> --- a/drivers/cpuidle/poll_state.c
>> +++ b/drivers/cpuidle/poll_state.c
>> @@ -6,41 +6,22 @@
>> [ ... skip 36 lines ... ]
>> - }
>> - }
>> + dev->poll_time_limit = !tif_need_resched_relaxed_wait(limit);
>> }
>> raw_local_irq_disable();
>>
>
> Tested again on AWS Graviton3/4/5 (ARM Neoverse V1/V2/V3) with haltpoll changes
> atop latest master 8ab1afb2e: still seeing average 6% improvement across various
> workloads (mysql, postgresql, cassandra, memcached) in latency sensitive configuration.
Thanks for retesting.
> The new unit tests all pass on both metal and virt variants of the same machines.
>
> I hope we can merge this soon.
Yes indeed. I'll send out one more version which addresses a few minor
bot comments (and patch-1 needs a re-review because of changes.)
But other than that this is as mergeable as it will ever be.
> Tested-by: Haris Okanovic <harisokn@xxxxxxxxxx>
Thanks
--
ankur