From: "Paul E. McKenney" <paulmck@kernel.org>
To: John Hubbard <jhubbard@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>, Minchan Kim <minchan@kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
linux-mm <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
John Dias <joaodias@google.com>,
David Hildenbrand <david@redhat.com>
Subject: Re: [PATCH v4] mm: fix is_pinnable_page against on cma page
Date: Tue, 17 May 2022 13:21:15 -0700 [thread overview]
Message-ID: <20220517202115.GE1790663@paulmck-ThinkPad-P17-Gen-1> (raw)
In-Reply-To: <b6eef200-43d1-7913-21ed-176b05fcb4fe@nvidia.com>
On Tue, May 17, 2022 at 01:12:02PM -0700, John Hubbard wrote:
> On 5/17/22 12:28, Jason Gunthorpe wrote:
> > > If you compare this to the snippet above, you'll see that there is
> > > an extra mov statement, and that one dereferences a pointer from
> > > %rax:
> > >
> > > mov (%rax),%rbx
> >
> > That is the same move as:
> >
> > mov 0x8(%rdx,%rax,8),%rbx
> >
> > Except that the EA calculation was done in advance and stored in rax.
> >
> > lea isn't a memory reference, it is just computing the pointer value
> > that 0x8(%rdx,%rax,8) represents. ie the lea computes
> >
> > %rax = %rdx + %rax*8 + 6
> >
> > Which is then fed into the mov. Maybe it is an optimization to allow
> > one pipe to do the shr and an other to the EA - IDK, seems like a
> > random thing for the compiler to do.
Maybe an optimization suppressed due to the volatile nature of the
load? If so, perhaps it might be considered a compiler bug. Though
it is quite difficult to get optimization bugs involving volatile
to be taken seriously.
> Apologies for getting that wrong, and thanks for walking me through the
> asm.
>
> [...]
> >
> > Paul can correct me, but I understand we do not have a list of allowed
> > operations that are exempted from the READ_ONCE() requirement. ie it
> > is not just conditional branching that requires READ_ONCE().
> >
> > This is why READ_ONCE() must always be on the memory load, because the
> > point is to sanitize away the uncertainty that comes with an unlocked
> > read of unstable memory contents. READ_ONCE() samples the value in
> > memory, and removes all tearing, multiload, etc "instability" that may
> > effect down stream computations. In this way down stream compulations
> > become reliable.
> >
> > Jason
>
> So then:
Works for me!
Acked-by: Paul E. McKenney <paulmck@kernel.org>
Thanx, Paul
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 0e42038382c1..b404f87e2682 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -482,7 +482,12 @@ unsigned long __get_pfnblock_flags_mask(const struct page *page,
> word_bitidx = bitidx / BITS_PER_LONG;
> bitidx &= (BITS_PER_LONG-1);
>
> - word = bitmap[word_bitidx];
> + /*
> + * This races, without locks, with set_pageblock_migratetype(). Ensure
> + * a consistent (non-tearing) read of the memory array, so that results,
> + * even though racy, are not corrupted.
> + */
> + word = READ_ONCE(bitmap[word_bitidx]);
> return (word >> bitidx) & mask;
> }
>
>
> thanks,
> --
> John Hubbard
> NVIDIA
next prev parent reply other threads:[~2022-05-17 20:21 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-10 21:17 [PATCH v4] mm: fix is_pinnable_page against on cma page Minchan Kim
2022-05-10 22:56 ` John Hubbard
2022-05-10 23:31 ` Minchan Kim
2022-05-10 23:58 ` John Hubbard
2022-05-11 0:09 ` Minchan Kim
2022-05-11 4:32 ` John Hubbard
2022-05-11 21:46 ` Minchan Kim
2022-05-11 22:25 ` John Hubbard
2022-05-11 22:37 ` Minchan Kim
2022-05-11 22:49 ` John Hubbard
2022-05-11 23:08 ` Minchan Kim
2022-05-11 23:13 ` John Hubbard
2022-05-11 23:15 ` Minchan Kim
2022-05-11 23:28 ` Minchan Kim
2022-05-11 23:33 ` John Hubbard
2022-05-11 23:45 ` Paul E. McKenney
2022-05-11 23:57 ` John Hubbard
2022-05-12 0:12 ` Paul E. McKenney
2022-05-12 0:12 ` John Hubbard
2022-05-12 0:22 ` Paul E. McKenney
2022-05-12 0:26 ` Minchan Kim
2022-05-12 0:34 ` John Hubbard
2022-05-12 0:49 ` Paul E. McKenney
2022-05-12 1:02 ` John Hubbard
2022-05-12 1:03 ` Minchan Kim
2022-05-12 1:08 ` John Hubbard
2022-05-12 2:18 ` John Hubbard
2022-05-12 3:44 ` Minchan Kim
2022-05-12 4:47 ` John Hubbard
2022-05-17 14:00 ` Jason Gunthorpe
2022-05-17 18:12 ` John Hubbard
2022-05-17 19:28 ` Jason Gunthorpe
2022-05-17 20:12 ` John Hubbard
2022-05-17 20:21 ` Paul E. McKenney [this message]
2022-05-23 16:33 ` Minchan Kim
2022-05-24 2:55 ` John Hubbard
2022-05-24 5:16 ` Minchan Kim
2022-05-24 6:22 ` John Hubbard
2022-05-24 14:19 ` Jason Gunthorpe
2022-05-24 15:43 ` Minchan Kim
2022-05-24 15:48 ` Jason Gunthorpe
2022-05-24 16:37 ` Paul E. McKenney
2022-05-24 16:59 ` Minchan Kim
2022-05-12 3:57 ` Paul E. McKenney
2022-05-12 1:03 ` Minchan Kim
2022-05-12 0:35 ` Paul E. McKenney
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=20220517202115.GE1790663@paulmck-ThinkPad-P17-Gen-1 \
--to=paulmck@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=david@redhat.com \
--cc=jgg@ziepe.ca \
--cc=jhubbard@nvidia.com \
--cc=joaodias@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=minchan@kernel.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).