From: Mike Kravetz <mike.kravetz@oracle.com>
To: Wei Yang <richard.weiyang@linux.alibaba.com>
Cc: Michal Hocko <mhocko@suse.com>, Baoquan He <bhe@redhat.com>,
akpm@linux-foundation.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 10/10] mm/hugetlb: not necessary to abuse temporary page to workaround the nasty free_huge_page
Date: Tue, 11 Aug 2020 16:25:01 -0700 [thread overview]
Message-ID: <5125772b-2939-e71f-da4a-374cb74c9061@oracle.com> (raw)
In-Reply-To: <20200811231932.GA33666@L-31X9LVDL-1304.local>
On 8/11/20 4:19 PM, Wei Yang wrote:
> On Tue, Aug 11, 2020 at 02:43:28PM -0700, Mike Kravetz wrote:
>> Subject: [PATCH] hugetlb: optimize race error return in
>> alloc_surplus_huge_page
>>
>> The routine alloc_surplus_huge_page() could race with with a pool
>> size change. If this happens, the allocated page may not be needed.
>> To free the page, the current code will 'Abuse temporary page to
>> workaround the nasty free_huge_page codeflow'. Instead, directly
>> call the low level routine that free_huge_page uses. This works
>> out well because the page is new, we hold the only reference and
>> already hold the hugetlb_lock.
>>
>> Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
>> ---
>> mm/hugetlb.c | 13 ++++++++-----
>> 1 file changed, 8 insertions(+), 5 deletions(-)
>>
>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
>> index 590111ea6975..ac89b91fba86 100644
>> --- a/mm/hugetlb.c
>> +++ b/mm/hugetlb.c
>> @@ -1923,14 +1923,17 @@ static struct page *alloc_surplus_huge_page(struct hstate *h, gfp_t gfp_mask,
>> /*
>> * We could have raced with the pool size change.
>> * Double check that and simply deallocate the new page
>> - * if we would end up overcommiting the surpluses. Abuse
>> - * temporary page to workaround the nasty free_huge_page
>> - * codeflow
>> + * if we would end up overcommiting the surpluses.
>> */
>> if (h->surplus_huge_pages >= h->nr_overcommit_huge_pages) {
>> - SetPageHugeTemporary(page);
>> + /*
>> + * Since this page is new, we hold the only reference, and
>> + * we already hold the hugetlb_lock call the low level free
>> + * page routine. This saves at least a lock roundtrip.
>
> The change looks good to me, while I may not understand the "lock roundtrip".
> You mean we don't need to release the hugetlb_lock?
Correct.
Normally we would free the page via free_huge_page() processing. To do that
we need to drop hugetlb_lock and call put_page/free_huge_page which will
need to acquire the hugetlb_lock again.
--
Mike Kravetz
>
>> + */
>> + (void)put_page_testzero(page); /* don't call destructor */
>> + update_and_free_page(h, page);
>> spin_unlock(&hugetlb_lock);
>> - put_page(page);
>> return NULL;
>> } else {
>> h->surplus_huge_pages++;
>> --
>> 2.25.4
next prev parent reply other threads:[~2020-08-11 23:25 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-07 9:12 [PATCH 00/10] mm/hugetlb: code refine and simplification Wei Yang
2020-08-07 9:12 ` [PATCH 01/10] mm/hugetlb: not necessary to coalesce regions recursively Wei Yang
2020-08-07 12:47 ` Baoquan He
2020-08-10 20:22 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 02/10] mm/hugetlb: make sure to get NULL when list is empty Wei Yang
2020-08-07 12:49 ` Baoquan He
2020-08-07 14:28 ` Wei Yang
2020-08-10 0:57 ` Baoquan He
2020-08-10 20:28 ` Mike Kravetz
2020-08-10 23:05 ` Wei Yang
2020-08-07 9:12 ` [PATCH 03/10] mm/hugetlb: use list_splice to merge two list at once Wei Yang
2020-08-07 12:53 ` Baoquan He
2020-08-10 21:07 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 04/10] mm/hugetlb: count file_region to be added when regions_needed != NULL Wei Yang
2020-08-07 12:54 ` Baoquan He
2020-08-10 21:46 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 05/10] mm/hugetlb: remove the redundant check on non_swap_entry() Wei Yang
2020-08-07 12:55 ` Baoquan He
2020-08-07 14:28 ` Wei Yang
2020-08-07 9:12 ` [PATCH 06/10] mm/hugetlb: remove redundant huge_pte_alloc() in hugetlb_fault() Wei Yang
2020-08-07 12:59 ` Baoquan He
2020-08-10 22:00 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 07/10] mm/hugetlb: a page from buddy is not on any list Wei Yang
2020-08-07 13:06 ` Baoquan He
2020-08-10 22:25 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 08/10] mm/hugetlb: return non-isolated page in the loop instead of break and check Wei Yang
2020-08-07 13:09 ` Baoquan He
2020-08-07 14:32 ` Wei Yang
2020-08-10 22:55 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 09/10] mm/hugetlb: narrow the hugetlb_lock protection area during preparing huge page Wei Yang
2020-08-07 13:12 ` Baoquan He
2020-08-10 23:02 ` Mike Kravetz
2020-08-07 9:12 ` [PATCH 10/10] mm/hugetlb: not necessary to abuse temporary page to workaround the nasty free_huge_page Wei Yang
2020-08-10 2:17 ` Baoquan He
2020-08-11 0:19 ` Mike Kravetz
2020-08-11 1:51 ` Baoquan He
2020-08-11 6:54 ` Michal Hocko
2020-08-11 21:43 ` Mike Kravetz
2020-08-11 23:19 ` Wei Yang
2020-08-11 23:25 ` Mike Kravetz [this message]
2020-08-12 5:40 ` Baoquan He
2020-08-13 11:46 ` Michal Hocko
2020-08-17 3:04 ` Wei Yang
2020-08-11 23:55 ` Baoquan He
2020-08-07 22:25 ` [PATCH 00/10] mm/hugetlb: code refine and simplification Mike Kravetz
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=5125772b-2939-e71f-da4a-374cb74c9061@oracle.com \
--to=mike.kravetz@oracle.com \
--cc=akpm@linux-foundation.org \
--cc=bhe@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.com \
--cc=richard.weiyang@linux.alibaba.com \
/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).