[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210922091153.GA80396@kvm.asia-northeast3-a.c.our-ratio-313919.internal>
Date: Wed, 22 Sep 2021 09:11:53 +0000
From: Hyeonggon Yoo <42.hyeyoo@...il.com>
To: Matthew Wilcox <willy@...radead.org>
Cc: linux-mm@...ck.org, Christoph Lameter <cl@...ux.com>,
Pekka Enberg <penberg@...nel.org>,
David Rientjes <rientjes@...gle.com>,
Joonsoo Kim <iamjoonsoo.kim@....com>,
Andrew Morton <akpm@...ux-foundation.org>,
Vlastimil Babka <vbabka@...e.cz>, linux-kernel@...r.kernel.org,
Jens Axboe <axboe@...nel.dk>,
John Garry <john.garry@...wei.com>,
linux-block@...r.kernel.org, netdev@...r.kernel.org
Subject: Re: [RFC v2 PATCH] mm, sl[au]b: Introduce lockless cache
> @@ -491,13 +492,13 @@ void kmem_cache_free_cached(struct kmem_cache *s, void *p)
> cache = get_cpu_ptr(s->cache);
> if (cache->size < KMEM_LOCKLESS_CACHE_QUEUE_SIZE) {
> cache->queue[cache->size++] = p;
> - put_cpu_ptr(s->cache);
> - return ;
> + } else {
> + kmem_cache_free_bulk(s,
> + KMEM_LOCKLESS_CACHE_BATCHCOUNT,
> + cache->queue - KMEM_LOCKLESS_CACHE_BATCHCOUNT);
> + cache->size -= KMEM_LOCKLESS_CACHE_BATCHCOUNT;
> }
> put_cpu_ptr(s->cache);
> -
> - /* Is there better way to do this? */
> - kmem_cache_free(s, p);
> }
> EXPORT_SYMBOL(kmem_cache_free_cached);
Sent you a wrong code.
Above was buggy code from some hours ago
because of cache->queue - KMEM_LOCKLESS_CACHE_BATCHCOUNT.
So that is now:
cache = get_cpu_ptr(s->cache);
if (cache->size < KMEM_LOCKLESS_CACHE_QUEUE_SIZE) {
cache->queue[cache->size++] = p;
} else {
kmem_cache_free_bulk(s,
KMEM_LOCKLESS_CACHE_BATCHCOUNT,
cache->queue + KMEM_LOCKLESS_CACHE_QUEUE_SIZE
- KMEM_LOCKLESS_CACHE_BATCHCOUNT);
cache->size -= KMEM_LOCKLESS_CACHE_BATCHCOUNT;
}
put_cpu_ptr(s->cache);
Powered by blists - more mailing lists