mm/zsmalloc: clarify class per-fullness zspage counts

We always use insert_zspage() and remove_zspage() to update zspage's
fullness location, which will account correctly.

But this special async free path use "splice" instead of remove_zspage(),
so the per-fullness zspage count for ZS_INUSE_RATIO_0 won't decrease.

Clean things up by decreasing when iterate over the zspage free list.

This doesn't actually fix anything.  ZS_INUSE_RATIO_0 is just a
"placeholder" which is never used anywhere.

Link: https://lkml.kernel.org/r/20240627075959.611783-1-chengming.zhou@linux.dev
Signed-off-by: Chengming Zhou <chengming.zhou@linux.dev>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
This commit is contained in:
Chengming Zhou 2024-06-27 15:59:58 +08:00 committed by Andrew Morton
parent 81510a0eaa
commit 538148f9ba

View File

@ -1883,6 +1883,7 @@ static void async_free_zspage(struct work_struct *work)
class = zspage_class(pool, zspage);
spin_lock(&class->lock);
class_stat_dec(class, ZS_INUSE_RATIO_0, 1);
__free_zspage(pool, class, zspage);
spin_unlock(&class->lock);
}