From: David Hildenbrand <david@redhat.com>
To: Vlastimil Babka <vbabka@suse.cz>, linux-mm@kvack.org
Cc: linux-kernel@vger.kernel.org, Michal Hocko <mhocko@kernel.org>,
Pavel Tatashin <pasha.tatashin@soleen.com>,
Oscar Salvador <osalvador@suse.de>,
Joonsoo Kim <iamjoonsoo.kim@lge.com>
Subject: Re: [PATCH 4/9] mm, page_alloc: simplify pageset_update()
Date: Fri, 25 Sep 2020 12:23:20 +0200 [thread overview]
Message-ID: <96cb97df-fe33-5a99-6618-1560467f8ffa@redhat.com> (raw)
In-Reply-To: <20200922143712.12048-5-vbabka@suse.cz>
On 22.09.20 16:37, Vlastimil Babka wrote:
> pageset_update() attempts to update pcplist's high and batch values in a way
> that readers don't observe batch > high. It uses smp_wmb() to order the updates
> in a way to achieve this. However, without proper pairing read barriers in
> readers this guarantee doesn't hold, and there are no such barriers in
> e.g. free_unref_page_commit().
>
> Commit 88e8ac11d2ea ("mm, page_alloc: fix core hung in free_pcppages_bulk()")
> already showed this is problematic, and solved this by ultimately only trusing
> pcp->count of the current cpu with interrupts disabled.
>
> The update dance with unpaired write barriers thus makes no sense. Replace
> them with plain WRITE_ONCE to prevent store tearing, and document that the
> values can change asynchronously and should not be trusted for correctness.
>
> All current readers appear to be OK after 88e8ac11d2ea. Convert them to
> READ_ONCE to prevent unnecessary read tearing, but mainly to alert anybody
> making future changes to the code that special care is needed.
>
> Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
> ---
> mm/page_alloc.c | 40 ++++++++++++++++++----------------------
> 1 file changed, 18 insertions(+), 22 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 76c2b4578723..99b74c1c2b0a 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -1297,7 +1297,7 @@ static void free_pcppages_bulk(struct zone *zone, int count,
> {
> int migratetype = 0;
> int batch_free = 0;
> - int prefetch_nr = 0;
> + int prefetch_nr = READ_ONCE(pcp->batch);
> bool isolated_pageblocks;
> struct page *page, *tmp;
> LIST_HEAD(head);
> @@ -1348,8 +1348,10 @@ static void free_pcppages_bulk(struct zone *zone, int count,
> * avoid excessive prefetching due to large count, only
> * prefetch buddy for the first pcp->batch nr of pages.
> */
> - if (prefetch_nr++ < pcp->batch)
> + if (prefetch_nr) {
> prefetch_buddy(page);
> + prefetch_nr--;
> + }
> } while (--count && --batch_free && !list_empty(list));
> }
>
> @@ -3131,10 +3133,8 @@ static void free_unref_page_commit(struct page *page, unsigned long pfn)
> pcp = &this_cpu_ptr(zone->pageset)->pcp;
> list_add(&page->lru, &pcp->lists[migratetype]);
> pcp->count++;
> - if (pcp->count >= pcp->high) {
> - unsigned long batch = READ_ONCE(pcp->batch);
> - free_pcppages_bulk(zone, batch, pcp);
> - }
> + if (pcp->count >= READ_ONCE(pcp->high))
> + free_pcppages_bulk(zone, READ_ONCE(pcp->batch), pcp);
> }
>
> /*
> @@ -3318,7 +3318,7 @@ static struct page *__rmqueue_pcplist(struct zone *zone, int migratetype,
> do {
> if (list_empty(list)) {
> pcp->count += rmqueue_bulk(zone, 0,
> - pcp->batch, list,
> + READ_ONCE(pcp->batch), list,
> migratetype, alloc_flags);
> if (unlikely(list_empty(list)))
> return NULL;
> @@ -6174,13 +6174,16 @@ static int zone_batchsize(struct zone *zone)
> }
>
> /*
> - * pcp->high and pcp->batch values are related and dependent on one another:
> - * ->batch must never be higher then ->high.
> - * The following function updates them in a safe manner without read side
> - * locking.
> + * pcp->high and pcp->batch values are related and generally batch is lower
> + * than high. They are also related to pcp->count such that count is lower
> + * than high, and as soon as it reaches high, the pcplist is flushed.
> *
> - * Any new users of pcp->batch and pcp->high should ensure they can cope with
> - * those fields changing asynchronously (acording to the above rule).
> + * However, guaranteeing these relations at all times would require e.g. write
> + * barriers here but also careful usage of read barriers at the read side, and
> + * thus be prone to error and bad for performance. Thus the update only prevents
> + * store tearing. Any new users of pcp->batch and pcp->high should ensure they
> + * can cope with those fields changing asynchronously, and fully trust only the
> + * pcp->count field on the local CPU with interrupts disabled.
> *
> * mutex_is_locked(&pcp_batch_high_lock) required when calling this function
> * outside of boot time (or some other assurance that no concurrent updaters
> @@ -6189,15 +6192,8 @@ static int zone_batchsize(struct zone *zone)
> static void pageset_update(struct per_cpu_pages *pcp, unsigned long high,
> unsigned long batch)
> {
> - /* start with a fail safe value for batch */
> - pcp->batch = 1;
> - smp_wmb();
> -
> - /* Update high, then batch, in order */
> - pcp->high = high;
> - smp_wmb();
> -
> - pcp->batch = batch;
> + WRITE_ONCE(pcp->batch, batch);
> + WRITE_ONCE(pcp->high, high);
> }
>
> static void pageset_init(struct per_cpu_pageset *p)
>
I *think* this is okay and obviously simplifies things. But to be 100%
sure, I'll have to rely on your judgment :)
Acked-by: David Hildenbrand <david@redhat.com>
--
Thanks,
David / dhildenb
next prev parent reply other threads:[~2020-09-25 10:23 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-22 14:37 [PATCH 0/9] disable pcplists during memory offline Vlastimil Babka
2020-09-22 14:37 ` [PATCH 1/9] mm, page_alloc: clean up pageset high and batch update Vlastimil Babka
2020-09-25 10:18 ` David Hildenbrand
2020-10-05 12:03 ` Michal Hocko
2020-09-22 14:37 ` [PATCH 2/9] mm, page_alloc: calculate pageset high and batch once per zone Vlastimil Babka
2020-10-05 12:52 ` Michal Hocko
2020-10-06 22:04 ` Vlastimil Babka
2020-09-22 14:37 ` [PATCH 3/9] mm, page_alloc: remove setup_pageset() Vlastimil Babka
2020-09-25 10:19 ` David Hildenbrand
2020-10-05 12:59 ` Michal Hocko
2020-10-06 22:11 ` Vlastimil Babka
2020-09-22 14:37 ` [PATCH 4/9] mm, page_alloc: simplify pageset_update() Vlastimil Babka
2020-09-25 10:23 ` David Hildenbrand [this message]
2020-10-05 13:20 ` Michal Hocko
2020-09-22 14:37 ` [PATCH 5/9] mm, page_alloc: make per_cpu_pageset accessible only after init Vlastimil Babka
2020-09-25 10:25 ` David Hildenbrand
2020-10-05 13:24 ` Michal Hocko
2020-10-06 22:28 ` Vlastimil Babka
2020-09-22 14:37 ` [PATCH 6/9] mm, page_alloc: cache pageset high and batch in struct zone Vlastimil Babka
2020-09-25 10:34 ` David Hildenbrand
2020-10-06 22:31 ` Vlastimil Babka
2020-10-05 13:28 ` Michal Hocko
2020-10-06 22:34 ` Vlastimil Babka
2020-09-22 14:37 ` [PATCH 7/9] mm, page_alloc: move draining pcplists to page isolation users Vlastimil Babka
2020-09-25 10:39 ` David Hildenbrand
2020-10-05 13:57 ` Michal Hocko
2020-09-22 14:37 ` [PATCH 8/9] mm, page_alloc: drain all pcplists during memory offline Vlastimil Babka
2020-09-25 10:46 ` David Hildenbrand
2020-10-05 14:03 ` Michal Hocko
2020-09-22 14:37 ` [PATCH 9/9] mm, page_alloc: optionally disable pcplists during page isolation Vlastimil Babka
2020-09-25 10:53 ` David Hildenbrand
2020-09-25 10:54 ` David Hildenbrand
2020-09-25 11:10 ` Vlastimil Babka
2020-10-01 8:47 ` David Hildenbrand
2020-10-05 14:05 ` Michal Hocko
2020-10-05 14:22 ` Vlastimil Babka
2020-10-05 16:56 ` Michal Hocko
2020-10-06 8:34 ` Michal Hocko
2020-10-06 8:40 ` David Hildenbrand
2020-10-06 10:05 ` Michal Hocko
2020-09-22 17:15 ` [PATCH 0/9] disable pcplists during memory offline David Hildenbrand
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=96cb97df-fe33-5a99-6618-1560467f8ffa@redhat.com \
--to=david@redhat.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=osalvador@suse.de \
--cc=pasha.tatashin@soleen.com \
--cc=vbabka@suse.cz \
/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).