On 2021-11-16 22:42:01 [+0900], Chanho Park wrote: > Hi, Hi, > > On 2021-11-16 10:37:22 [+0100], Vlastimil Babka wrote: > > > On 11/12/21 15:00, 'Sebastian Andrzej Siewior' wrote: > > > > On 2021-11-05 10:41:40 [+0900], Chanho Park wrote: > > > >> Hi, > > > > Hi, > > > > > > > >> I checked the same test on my aarch64 target. I'll do more > > > >> realistic benchmark such as compile bench. > > > >> > > > >> <5.10.73-rt54 aarch64> > > > >> root@euto-v9-sadk:~# hackbench -l 10000 > > > >> Time: 24.994 > > > >> > > > >> <5.15.0-rt17 aarch64 w/o CONFIG_SLUB_CPU_PARTIAL> > > > >> root@euto-v9-sadk:~# hackbench -l 10000 > > > >> Time: 31.372 > > > >> > > > >> <5.15.0-rt17 aarch64 w/ CONFIG_SLUB_CPU_PARTIAL> > > > >> root@euto-v9-sadk:~# hackbench -l 10000 > > > >> Time: 35.269 > > > > > > > > The last two are equal, is that on purporse? > > > > > > I see the first of last two is without SLUB_CPU_PARTIAL and the second > > > with SLUB_CPU_PARTIAL, so not equal? > > > > indeed. I didn't notice the missing o. Thanks!. > > Sorry for this confusion. > > > > > > > Any feedback on v5.14-rc3-rt1 and v5.14-rc3-rt2? v5.13-rt1 should be > > > > identical to you 5.10 numbers. > > I got below value on v5.14-rc3-rt1. > > <5.14.0-rc3-rt1 aarch64> > Time: 35.948 So this matches your v5.15 numbers with CONFIG_SLUB_CPU_PARTIAL. In v5.14-rc3-rt2 adaptive spinning was introduced. Based on these numbers it makes no difference for you. > Unfortunately, I've got below panic on v5.13-rt1. I stacked same patches for my board both v5.13-rt1 and v514-rc3-rt1. Yes, good. > Best Regards, > Chanho Park Sebastian
