grp_spread_init_one() implementation is sub-optimal because it
traverses bitmaps from the beginning, instead of picking from the
previous iteration.

Fix it and use find_bit API where appropriate. While here, optimize
cpumasks allocation and drop unneeded cpumask_empty() call.

---
v1: https://lore.kernel.org/all/ZW5MI3rKQueLM0Bz@yury-ThinkPad/T/
v2: add few more optimizations (patches 5-6)

Yury Norov (6):
  cpumask: introduce for_each_cpu_and_from()
  lib/group_cpus: relax atomicity requirement in grp_spread_init_one()
  lib/group_cpus: optimize inner loop in grp_spread_init_one()
  lib/group_cpus: optimize outer loop in grp_spread_init_one()
  lib/cgroup_cpus.c: don't zero cpumasks in group_cpus_evenly() on
    allocation
  lib/group_cpus.c: drop unneeded cpumask_empty() call in
    __group_cpus_evenly()

 include/linux/cpumask.h | 11 ++++++++++
 include/linux/find.h    |  3 +++
 lib/group_cpus.c        | 47 +++++++++++++++++++----------------------
 3 files changed, 36 insertions(+), 25 deletions(-)

-- 
2.40.1

On Thu,  7 Dec 2023 12:38:54 -0800 Yury Norov <yury.norov@gmail.com> wrote:

> grp_spread_init_one() implementation is sub-optimal because it
> traverses bitmaps from the beginning, instead of picking from the
> previous iteration.
> 
> Fix it and use find_bit API where appropriate. While here, optimize
> cpumasks allocation and drop unneeded cpumask_empty() call.

Thanks.  This isn't my playground, but I'll grab the patches to at
least get them some testing.  Review from those who have worked
on this code would be appreciated.

On Thu, Dec 07, 2023 at 01:46:19PM -0800, Andrew Morton wrote:
> On Thu,  7 Dec 2023 12:38:54 -0800 Yury Norov <yury.norov@gmail.com> wrote:
> 
> > grp_spread_init_one() implementation is sub-optimal because it
> > traverses bitmaps from the beginning, instead of picking from the
> > previous iteration.
> > 
> > Fix it and use find_bit API where appropriate. While here, optimize
> > cpumasks allocation and drop unneeded cpumask_empty() call.
> 
> Thanks.  This isn't my playground, but I'll grab the patches to at
> least get them some testing.  Review from those who have worked
> on this code would be appreciated.

Thanks you!