Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux

mglru: mm/vmscan.c: fix imprecise comments

Link: https://lkml.kernel.org/r/YzSWfFI+MOeb1ils@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

authored by

Yu Zhao and committed by
Andrew Morton
e4fea72b 14aa8b2d

+4 -5
+4 -5
mm/vmscan.c
··· 5076 5076 DEFINE_MAX_SEQ(lruvec); 5077 5077 5078 5078 if (!current_is_kswapd()) { 5079 - /* age each memcg once to ensure fairness */ 5079 + /* age each memcg at most once to ensure fairness */ 5080 5080 if (max_seq - seq > 1) 5081 5081 return true; 5082 5082 ··· 5101 5101 5102 5102 /* 5103 5103 * A minimum amount of work was done under global memory pressure. For 5104 - * kswapd, it may be overshooting. For direct reclaim, the target isn't 5105 - * met, and yet the allocation may still succeed, since kswapd may have 5106 - * caught up. In either case, it's better to stop now, and restart if 5107 - * necessary. 5104 + * kswapd, it may be overshooting. For direct reclaim, the allocation 5105 + * may succeed if all suitable zones are somewhat safe. In either case, 5106 + * it's better to stop now, and restart later if necessary. 5108 5107 */ 5109 5108 for (i = 0; i <= sc->reclaim_idx; i++) { 5110 5109 unsigned long wmark;