breeze/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-01-12 00:42:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

cpumask: use cpumask_next_wrap() where appropriate

Browse Source 
Now that cpumask_next{_and}_wrap() is wired to generic
find_next_bit_wrap(), we can use it in cpumask_any{_and}_distribute().

This automatically makes the cpumask_*_distribute() functions to use
small_cpumask_bits instead of nr_cpumask_bits, which itself is a good
optimization.

Signed-off-by: Yury Norov <yury.norov@gmail.com>
This commit is contained in:
Yury Norov
2025-01-28 11:46:36 -05:00
parent 3268cb2e49
commit 566babe82b
1 changed files with 2 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
lib/cpumask.c
View File

@@ -171,8 +171,7 @@ unsigned int cpumask_any_and_distribute(const struct cpumask *src1p,
/* NOTE: our first selection will skip 0. */
prev = __this_cpu_read(distribute_cpu_mask_prev);
next = find_next_and_bit_wrap(cpumask_bits(src1p), cpumask_bits(src2p),
nr_cpumask_bits, prev + 1);
next = cpumask_next_and_wrap(prev, src1p, src2p);
if (next < nr_cpu_ids)
__this_cpu_write(distribute_cpu_mask_prev, next);
@@ -192,7 +191,7 @@ unsigned int cpumask_any_distribute(const struct cpumask *srcp)
/* NOTE: our first selection will skip 0. */
prev = __this_cpu_read(distribute_cpu_mask_prev);
next = find_next_bit_wrap(cpumask_bits(srcp), nr_cpumask_bits, prev + 1);
next = cpumask_next_wrap(prev, srcp);
if (next < nr_cpu_ids)
__this_cpu_write(distribute_cpu_mask_prev, next);