[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <9a26ce23-9ac7-b111-b700-db0905203c73@redhat.com>
Date: Mon, 31 Jul 2023 21:22:26 +0200
From: David Hildenbrand <david@...hat.com>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Peter Xu <peterx@...hat.com>, linux-kernel@...r.kernel.org,
linux-mm@...ck.org, linux-fsdevel@...r.kernel.org,
Andrew Morton <akpm@...ux-foundation.org>,
liubo <liubo254@...wei.com>,
Matthew Wilcox <willy@...radead.org>,
Hugh Dickins <hughd@...gle.com>,
Jason Gunthorpe <jgg@...pe.ca>,
John Hubbard <jhubbard@...dia.com>,
Mel Gorman <mgorman@...e.de>
Subject: Re: [PATCH v1 0/4] smaps / mm/gup: fix gup_can_follow_protnone
fallout
On 31.07.23 21:07, Linus Torvalds wrote:
> On Mon, 31 Jul 2023 at 12:00, David Hildenbrand <david@...hat.com> wrote:
>>
>> So in the "!writable" case, we would always call
>> get_user_pages_unlocked() and never honor NUMA faults.
>
> Ok, so kvm doesn't just use the fast version. Oh well. It was an idea..
Certainly an interesting one, thanks for thinking about possible
alternatives! Unfortunately, KVM is an advanced GUP user to managed
secondary MMUs.
I'll get the FOLL_HONOR_NUMA_FAULT patches ready tomorrow and we can
discuss if that looks better.
(btw, the whole reasoning about "HW would refuse to use these pages" was
exactly the whole reason I went into the FOLL_FORCE direction ... but
it's really better to make FOLL_FORCE deal with VMA protection only)
--
Cheers,
David / dhildenb
Powered by blists - more mailing lists