[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <e4d6dc06-30cb-1f92-32a9-d9057bcbf6b0@intel.com>
Date: Fri, 18 Feb 2022 08:39:44 +0800
From: Yin Fengwei <fengwei.yin@...el.com>
To: David Hildenbrand <david@...hat.com>,
Mike Kravetz <mike.kravetz@...cle.com>, <linux-mm@...ck.org>,
<linux-kernel@...r.kernel.org>
CC: Naoya Horiguchi <naoya.horiguchi@...ux.dev>,
Axel Rasmussen <axelrasmussen@...gle.com>,
Mina Almasry <almasrymina@...gle.com>,
"Michal Hocko" <mhocko@...e.com>, Peter Xu <peterx@...hat.com>,
Andrea Arcangeli <aarcange@...hat.com>,
Shuah Khan <shuah@...nel.org>, Mike Rapoport <rppt@...nel.org>,
Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH v3 1/3] mm: enable MADV_DONTNEED for hugetlb mappings
On 2/17/2022 4:58 PM, David Hildenbrand wrote:
>> In man page of mmap, NOTE for "Huge page (Huge TLB) mappings":
>>
>> "For munmap(), addr, and length must both be a multiple of the
>> underlying huge page size."
>>
>> Should we apply same rule to MADV_DONTNEED? Thanks.
> madvise() already has different rules than mmap() for ordinary pages, so
> we'd much rather try staying consistent with madvise() rules.
Thanks for clarification.
Regards
Yin, Fengwei
Powered by blists - more mailing lists