Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1464416 > unrolled thread

Re: [PATCH v3 05/13] sched: Introduce SD_ASYM_CPUCAPACITY sched_domain topology flag

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH v3 05/13] sched: Introduce SD_ASYM_CPUCAPACITY  sched_domain topology flag Wanpeng Li <kernellwp@gmail.com> - 2016-08-17 10:50 +0200
    Re: [PATCH v3 05/13] sched: Introduce SD_ASYM_CPUCAPACITY  sched_domain topology flag Morten Rasmussen <morten.rasmussen@arm.com> - 2016-08-17 11:30 +0200
      Re: [PATCH v3 05/13] sched: Introduce SD_ASYM_CPUCAPACITY  sched_domain topology flag Wanpeng Li <kernellwp@gmail.com> - 2016-08-17 11:30 +0200

#1464416 — Re: [PATCH v3 05/13] sched: Introduce SD_ASYM_CPUCAPACITY sched_domain topology flag

2016-07-25 21:34 GMT+08:00 Morten Rasmussen <morten.rasmussen@arm.com>:
> Add a topology flag to the sched_domain hierarchy indicating the lowest
> domain level where the full range of cpu capacities is represented by
> the domain members for asymmetric capacity topologies (e.g. ARM
> big.LITTLE).
>
> The flag is intended to indicate that extra care should be taken when
> placing tasks on cpus and this level spans all the different types of
> cpus found in the system (no need to look further up the domain
> hierarchy). This information is currently only available through
> iterating through the capacities of all the cpus at parent levels in the
> sched_domain hierarchy.
>
> SD 2      [  0      1      2      3]  SD_ASYM_CPUCAPACITY
>
> SD 1      [  0      1] [   2      3]  !SD_ASYM_CPUCAPACITY
>
> cpu:         0      1      2      3
> capacity:  756    756   1024   1024
>
> If the topology in the example above is duplicated to create an eight
> cpu example with third sched_domain level on top (SD 3), this level
> should not have the flag set (!SD_ASYM_CPUCAPACITY) as its two group
> would both have all cpu capacities represented within them.

I didn't find the place where set SD_ASYM_CPUCAPACITY to any SDs in
this patchset, but you have testing result in cover letter, where I
miss?

Regards,
Wanpeng Li

[toc] | [next] | [standalone]

#1464434

On Wed, Aug 17, 2016 at 04:42:36PM +0800, Wanpeng Li wrote:
> 2016-07-25 21:34 GMT+08:00 Morten Rasmussen <morten.rasmussen@arm.com>:
> > Add a topology flag to the sched_domain hierarchy indicating the lowest
> > domain level where the full range of cpu capacities is represented by
> > the domain members for asymmetric capacity topologies (e.g. ARM
> > big.LITTLE).
> >
> > The flag is intended to indicate that extra care should be taken when
> > placing tasks on cpus and this level spans all the different types of
> > cpus found in the system (no need to look further up the domain
> > hierarchy). This information is currently only available through
> > iterating through the capacities of all the cpus at parent levels in the
> > sched_domain hierarchy.
> >
> > SD 2      [  0      1      2      3]  SD_ASYM_CPUCAPACITY
> >
> > SD 1      [  0      1] [   2      3]  !SD_ASYM_CPUCAPACITY
> >
> > cpu:         0      1      2      3
> > capacity:  756    756   1024   1024
> >
> > If the topology in the example above is duplicated to create an eight
> > cpu example with third sched_domain level on top (SD 3), this level
> > should not have the flag set (!SD_ASYM_CPUCAPACITY) as its two group
> > would both have all cpu capacities represented within them.
> 
> I didn't find the place where set SD_ASYM_CPUCAPACITY to any SDs in
> this patchset, but you have testing result in cover letter, where I
> miss?

The flag is supposed to be set by arch-specific code. I included a few
patches in v1 and v2 that set the flag for arch/arm. However, since they
are hopefully soon to be superseded by patches from Juri I dropped them
from the v3 posting and provided a pointer to branch containing this patch
set, Juri's patches, and few additional glue patches instead that
enabled the flag when necessary for arch/arm and arch/arm64.

Sorry for the confusion.
Morten

[toc] | [prev] | [next] | [standalone]

#1464440

2016-08-17 17:23 GMT+08:00 Morten Rasmussen <morten.rasmussen@arm.com>:
> On Wed, Aug 17, 2016 at 04:42:36PM +0800, Wanpeng Li wrote:
>> 2016-07-25 21:34 GMT+08:00 Morten Rasmussen <morten.rasmussen@arm.com>:
>> > Add a topology flag to the sched_domain hierarchy indicating the lowest
>> > domain level where the full range of cpu capacities is represented by
>> > the domain members for asymmetric capacity topologies (e.g. ARM
>> > big.LITTLE).
>> >
>> > The flag is intended to indicate that extra care should be taken when
>> > placing tasks on cpus and this level spans all the different types of
>> > cpus found in the system (no need to look further up the domain
>> > hierarchy). This information is currently only available through
>> > iterating through the capacities of all the cpus at parent levels in the
>> > sched_domain hierarchy.
>> >
>> > SD 2      [  0      1      2      3]  SD_ASYM_CPUCAPACITY
>> >
>> > SD 1      [  0      1] [   2      3]  !SD_ASYM_CPUCAPACITY
>> >
>> > cpu:         0      1      2      3
>> > capacity:  756    756   1024   1024
>> >
>> > If the topology in the example above is duplicated to create an eight
>> > cpu example with third sched_domain level on top (SD 3), this level
>> > should not have the flag set (!SD_ASYM_CPUCAPACITY) as its two group
>> > would both have all cpu capacities represented within them.
>>
>> I didn't find the place where set SD_ASYM_CPUCAPACITY to any SDs in
>> this patchset, but you have testing result in cover letter, where I
>> miss?
>
> The flag is supposed to be set by arch-specific code. I included a few
> patches in v1 and v2 that set the flag for arch/arm. However, since they
> are hopefully soon to be superseded by patches from Juri I dropped them
> from the v3 posting and provided a pointer to branch containing this patch
> set, Juri's patches, and few additional glue patches instead that
> enabled the flag when necessary for arch/arm and arch/arm64.

Ah, ok, thanks for the information. :)

Regards,
Wanpeng Li

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web