Messages in this thread | | | Date | Wed, 3 Aug 2022 11:18:27 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH] sched/fair: no sync wakeup from interrupt context |
| |
* Libo Chen <libo.chen@oracle.com> wrote:
> > > On 8/1/22 06:26, Ingo Molnar wrote: > > * K Prateek Nayak <kprateek.nayak@amd.com> wrote: > > > > > Hello Libo and Peter, > > > > > > tl;dr > > > > > > - We observed a major regression with tbench when testing the latest tip > > > sched/core at: > > > commit 14b3f2d9ee8d "sched/fair: Disallow sync wakeup from interrupt context" > > > Reason for the regression are the fewer affine wakeups that leaves the > > > client farther away from the data it needs to consume next primed in the > > > waker's LLC. > > > Such regressions can be expected from tasks that use sockets to communicate > > > significant amount of data especially on system with multiple LLCs. > > > > > > - Other benchmarks have a comparable behavior to the tip at previous commit > > > commit : 91caa5ae2424 "sched/core: Fix the bug that task won't enqueue > > > into core tree when update cookie" > > > > > > I'll leave more details below. > > Mel Gorman also warned about this negative side-effect in: > > > > Subject: Re: [PATCH] sched/fair: no sync wakeup from interrupt context > > Date: Fri, 15 Jul 2022 11:07:38 +0100 > > Message-ID: <20220715100738.GD3493@suse.de> > > > > https://urldefense.com/v3/__https://lore.kernel.org/all/20220715100738.GD3493@suse.de/__;!!ACWV5N9M2RV99hQ!PQsIeuK0UwII-A0xS-B3plepNniNeyw14OJowT1cYL-tnuN99MkWfg9C8P60tVFFrnxj0NEanUmEkA$ > ?? Mel was talking about a completely different thing, I brought up a > different patch that I wanted to revert and Mel thought it would hurt other > workloads which don't benefit from pulling but > as you can see, tbench somehow benefits from it, at least according to one > metric from one workload.
Yeah - but nevertheless the discussion with Mel was open-ended AFAICS, and the 'major tbench regression' report by K Prateek Nayak above still stands and needs to be investigated/understood, right?
Thanks,
Ingo
| |