From: Vlastimil Babka <vbabka@suse.cz>
To: Hyeonggon Yoo <42.hyeyoo@gmail.com>, linux-mm@kvack.org
Cc: Christoph Lameter <cl@linux.com>,
Pekka Enberg <penberg@kernel.org>,
David Rientjes <rientjes@google.com>,
Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Andrew Morton <akpm@linux-foundation.org>,
Marco Elver <elver@google.com>,
Matthew WilCox <willy@infradead.org>,
Roman Gushchin <roman.gushchin@linux.dev>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 5/5] mm/slub: refactor deactivate_slab()
Date: Fri, 4 Mar 2022 20:01:20 +0100 [thread overview]
Message-ID: <6763e97b-88bf-59b0-c80e-26c3846531fc@suse.cz> (raw)
In-Reply-To: <20220304063427.372145-6-42.hyeyoo@gmail.com>
On 3/4/22 07:34, Hyeonggon Yoo wrote:
> Simplify deactivate_slab() by unlocking n->list_lock and retrying
> cmpxchg_double() when cmpxchg_double() fails, and perform
> add_{partial,full} only when it succeed.
>
> Releasing and taking n->list_lock again here is not harmful as SLUB
> avoids deactivating slabs as much as possible.
>
> [ vbabka@suse.cz: perform add_{partial,full} when cmpxchg_double()
> succeed. ]
>
> Signed-off-by: Hyeonggon Yoo <42.hyeyoo@gmail.com>
Looks good, just noticed a tiny issue.
> ---
> mm/slub.c | 81 ++++++++++++++++++++++---------------------------------
> 1 file changed, 32 insertions(+), 49 deletions(-)
>
> diff --git a/mm/slub.c b/mm/slub.c
> index f9ae983a3dc6..c1a693ec5874 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -2344,8 +2344,8 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab,
> {
> enum slab_modes { M_NONE, M_PARTIAL, M_FULL, M_FREE };
> struct kmem_cache_node *n = get_node(s, slab_nid(slab));
> - int lock = 0, free_delta = 0;
> - enum slab_modes l = M_NONE, m = M_NONE;
> + int free_delta = 0;
> + enum slab_modes mode = M_NONE;
> void *nextfree, *freelist_iter, *freelist_tail;
> int tail = DEACTIVATE_TO_HEAD;
> unsigned long flags = 0;
> @@ -2387,14 +2387,10 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab,
> * Ensure that the slab is unfrozen while the list presence
> * reflects the actual number of objects during unfreeze.
> *
> - * We setup the list membership and then perform a cmpxchg
> - * with the count. If there is a mismatch then the slab
> - * is not unfrozen but the slab is on the wrong list.
> - *
> - * Then we restart the process which may have to remove
> - * the slab from the list that we just put it on again
> - * because the number of objects in the slab may have
> - * changed.
> + * We first perform cmpxchg holding lock and insert to list
> + * when it succeed. If there is mismatch then slub is not
> + * unfrozen and number of objects in the slab may have changed.
> + * Then release lock and retry cmpxchg again.
> */
> redo:
>
> @@ -2414,57 +2410,44 @@ static void deactivate_slab(struct kmem_cache *s, struct slab *slab,
> new.frozen = 0;
>
> if (!new.inuse && n->nr_partial >= s->min_partial)
> - m = M_FREE;
> + mode = M_FREE;
> else if (new.freelist) {
> - m = M_PARTIAL;
> - if (!lock) {
> - lock = 1;
> - /*
> - * Taking the spinlock removes the possibility that
> - * acquire_slab() will see a slab that is frozen
> - */
> - spin_lock_irqsave(&n->list_lock, flags);
> - }
> - } else {
> - m = M_FULL;
> - if (kmem_cache_debug_flags(s, SLAB_STORE_USER) && !lock) {
This used to set m = M_FULL; always.
> - lock = 1;
> - /*
> - * This also ensures that the scanning of full
> - * slabs from diagnostic functions will not see
> - * any frozen slabs.
> - */
> - spin_lock_irqsave(&n->list_lock, flags);
> - }
> + mode = M_PARTIAL;
> + /*
> + * Taking the spinlock removes the possibility that
> + * acquire_slab() will see a slab that is frozen
> + */
> + spin_lock_irqsave(&n->list_lock, flags);
> + } else if (kmem_cache_debug_flags(s, SLAB_STORE_USER)) {
> + mode = M_FULL;
Now you only set it for SLAB_STORE_USER caches.
> + /*
> + * This also ensures that the scanning of full
> + * slabs from diagnostic functions will not see
> + * any frozen slabs.
> + */
> + spin_lock_irqsave(&n->list_lock, flags);
> }
>
> - if (l != m) {
> - if (l == M_PARTIAL)
> - remove_partial(n, slab);
> - else if (l == M_FULL)
> - remove_full(s, n, slab);
> -
> - if (m == M_PARTIAL)
> - add_partial(n, slab, tail);
> - else if (m == M_FULL)
> - add_full(s, n, slab);
> - }
>
> - l = m;
> if (!cmpxchg_double_slab(s, slab,
> old.freelist, old.counters,
> new.freelist, new.counters,
> - "unfreezing slab"))
> + "unfreezing slab")) {
> + if (mode == M_PARTIAL || mode == M_FULL)
> + spin_unlock_irqrestore(&n->list_lock, flags);
> goto redo;
> + }
>
> - if (lock)
> - spin_unlock_irqrestore(&n->list_lock, flags);
>
> - if (m == M_PARTIAL)
> + if (mode == M_PARTIAL) {
> + add_partial(n, slab, tail);
> + spin_unlock_irqrestore(&n->list_lock, flags);
> stat(s, tail);
> - else if (m == M_FULL)
> + } else if (mode == M_FULL) {
> + add_full(s, n, slab);
> + spin_unlock_irqrestore(&n->list_lock, flags);
> stat(s, DEACTIVATE_FULL);
As a result, full slabs without SLAB_STORE_USER will not count
DEACTIVATE_FULL anymore.
I guess the easiest way to solve it is to e.g. add a M_FULL_NOLIST mode that
only does the DEACTIVATE_FULL counting.
> - else if (m == M_FREE) {
> + } else if (mode == M_FREE) {
> stat(s, DEACTIVATE_EMPTY);
> discard_slab(s, slab);
> stat(s, FREE_SLAB);
next prev parent reply other threads:[~2022-03-04 19:01 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-03-04 6:34 [PATCH v2 0/5] slab cleanups Hyeonggon Yoo
2022-03-04 6:34 ` [PATCH v2 1/5] mm/slab: kmalloc: pass requests larger than order-1 page to page allocator Hyeonggon Yoo
2022-03-04 12:45 ` Vlastimil Babka
2022-03-05 5:10 ` Hyeonggon Yoo
2022-03-04 6:34 ` [PATCH v2 2/5] mm/sl[au]b: unify __ksize() Hyeonggon Yoo
2022-03-04 18:25 ` Vlastimil Babka
2022-03-05 4:02 ` Hyeonggon Yoo
2022-03-04 6:34 ` [PATCH v2 3/5] mm/sl[auo]b: move definition of __ksize() to mm/slab.h Hyeonggon Yoo
2022-03-04 18:29 ` Vlastimil Babka
2022-03-05 4:03 ` Hyeonggon Yoo
2022-03-04 6:34 ` [PATCH v2 4/5] mm/slub: limit number of node partial slabs only in cache creation Hyeonggon Yoo
2022-03-04 18:33 ` Vlastimil Babka
2022-03-04 6:34 ` [PATCH v2 5/5] mm/slub: refactor deactivate_slab() Hyeonggon Yoo
2022-03-04 19:01 ` Vlastimil Babka [this message]
2022-03-05 4:21 ` Hyeonggon Yoo
2022-03-04 11:50 ` [PATCH v2 0/5] slab cleanups Marco Elver
2022-03-04 12:02 ` Hyeonggon Yoo
2022-03-04 13:11 ` Marco Elver
2022-03-04 16:42 ` Vlastimil Babka
2022-03-04 16:45 ` Marco Elver
2022-03-05 4:00 ` Hyeonggon Yoo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=6763e97b-88bf-59b0-c80e-26c3846531fc@suse.cz \
--to=vbabka@suse.cz \
--cc=42.hyeyoo@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=cl@linux.com \
--cc=elver@google.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=penberg@kernel.org \
--cc=rientjes@google.com \
--cc=roman.gushchin@linux.dev \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).