From: Alan Stern <stern@rowland.harvard.edu>
To: Will Deacon <will@kernel.org>
Cc: "Paul E. McKenney" <paulmck@kernel.org>,
parri.andrea@gmail.com, peterz@infradead.org,
boqun.feng@gmail.com, npiggin@gmail.com, dhowells@redhat.com,
j.alglave@ucl.ac.uk, luc.maranget@inria.fr, akiyks@gmail.com,
dlustig@nvidia.com, joel@joelfernandes.org,
viro@zeniv.linux.org.uk, linux-kernel@vger.kernel.org,
linux-arch@vger.kernel.org
Subject: Re: Litmus test for question from Al Viro
Date: Mon, 5 Oct 2020 10:23:51 -0400 [thread overview]
Message-ID: <20201005142351.GB376584@rowland.harvard.edu> (raw)
In-Reply-To: <20201005091247.GA23575@willie-the-truck>
On Mon, Oct 05, 2020 at 10:12:48AM +0100, Will Deacon wrote:
> On Mon, Oct 05, 2020 at 09:20:03AM +0100, Will Deacon wrote:
> > On Sun, Oct 04, 2020 at 10:38:46PM -0400, Alan Stern wrote:
> > > Considering the bug in herd7 pointed out by Akira, we should rewrite P1 as:
> > >
> > > P1(int *x, int *y)
> > > {
> > > int r2;
> > >
> > > r = READ_ONCE(*y);
> >
> > (r2?)
> >
> > > WRITE_ONCE(*x, r2);
> > > }
> > >
> > > Other than that, this is fine.
> >
> > But yes, module the typo, I agree that this rewrite is much better than the
> > proposal above. The definition of control dependencies on arm64 (per the Arm
> > ARM [1]) isn't entirely clear that it provides order if the WRITE is
> > executed on both paths of the branch, and I believe there are ongoing
> > efforts to try to tighten that up. I'd rather keep _that_ topic separate
> > from the "bug in herd" topic to avoid extra confusion.
>
> Ah, now I see that you're changing P1 here, not P0. So I'm now nervous
> about claiming that this is a bug in herd without input from Jade or Luc,
> as it does unfortunately tie into the definition of control dependencies
> and it could be a deliberate choice.
I think you misunderstood. The bug in herd7 affects the way it handles
P1, not P0. With
r2 = READ_ONCE(*y);
WRITE_ONCE(*x, r2);
herd7 generates a data dependency from the read to the write. With
WRITE_ONCE(*x, READ_ONCE(*y));
it doesn't generate any dependency, even though the code does exactly
the same thing as far as the memory model is concerned. That's the bug
I was referring to.
The failure to recognize the dependency in P0 should be considered a
combined limitation of the memory model and herd7. It's not a simple
mistake that can be fixed by a small rewrite of herd7; rather it's a
deliberate choice we made based on herd7's inherent design. We
explicitly said that control dependencies extend only to the code in the
branches of an "if" statement; anything beyond the end of the statement
is not considered to be dependent.
Alan
next prev parent reply other threads:[~2020-10-05 14:23 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-01 4:51 Litmus test for question from Al Viro Paul E. McKenney
2020-10-01 16:15 ` Alan Stern
2020-10-01 16:36 ` Al Viro
2020-10-01 18:39 ` Alan Stern
2020-10-01 19:29 ` Al Viro
2020-10-01 21:30 ` Paul E. McKenney
2020-10-03 2:01 ` Alan Stern
2020-10-03 13:22 ` Alan Stern
2020-10-03 15:16 ` Akira Yokosawa
2020-10-03 17:13 ` Bug in herd7 [Was: Re: Litmus test for question from Al Viro] Alan Stern
2020-10-03 22:50 ` Akira Yokosawa
2020-10-04 1:40 ` [PATCH] tools: memory-model: Document that the LKMM can easily miss control dependencies Alan Stern
2020-10-04 21:07 ` joel
2020-10-04 23:12 ` Paul E. McKenney
2020-10-05 15:15 ` Bug in herd7 [Was: Re: Litmus test for question from Al Viro] Luc Maranget
2020-10-05 15:53 ` Alan Stern
2020-10-05 16:52 ` Paul E. McKenney
2020-10-05 18:19 ` Alan Stern
2020-10-05 19:18 ` Paul E. McKenney
2020-10-05 19:48 ` Alan Stern
2020-10-06 16:39 ` Paul E. McKenney
2020-10-06 17:05 ` Alan Stern
2020-10-07 17:50 ` Paul E. McKenney
2020-10-07 19:40 ` Alan Stern
2020-10-07 22:38 ` Paul E. McKenney
2020-10-08 2:25 ` Alan Stern
2020-10-08 2:50 ` Paul E. McKenney
2020-10-08 14:01 ` Alan Stern
2020-10-08 18:32 ` Paul E. McKenney
2020-10-05 15:54 ` Paul E. McKenney
2020-10-04 23:31 ` Litmus test for question from Al Viro Paul E. McKenney
2020-10-05 2:38 ` Alan Stern
2020-10-05 8:20 ` Will Deacon
2020-10-05 9:12 ` Will Deacon
2020-10-05 14:01 ` Paul E. McKenney
2020-10-05 14:23 ` Alan Stern [this message]
2020-10-05 15:13 ` Will Deacon
2020-10-05 15:16 ` Alan Stern
2020-10-05 15:35 ` Peter Zijlstra
2020-10-05 15:49 ` Paul E. McKenney
2020-10-05 14:16 ` Alan Stern
2020-10-05 14:03 ` Paul E. McKenney
2020-10-05 14:24 ` Alan Stern
2020-10-05 14:44 ` joel
2020-10-05 15:55 ` Paul E. McKenney
2020-10-05 8:36 ` David Laight
2020-10-05 13:59 ` Paul E. McKenney
2020-10-03 16:08 ` joel
2020-10-03 16:11 ` joel
2020-10-04 23:13 ` Paul E. McKenney
2020-10-03 2:35 ` Jon Masters
2020-10-04 23:32 ` 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=20201005142351.GB376584@rowland.harvard.edu \
--to=stern@rowland.harvard.edu \
--cc=akiyks@gmail.com \
--cc=boqun.feng@gmail.com \
--cc=dhowells@redhat.com \
--cc=dlustig@nvidia.com \
--cc=j.alglave@ucl.ac.uk \
--cc=joel@joelfernandes.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luc.maranget@inria.fr \
--cc=npiggin@gmail.com \
--cc=parri.andrea@gmail.com \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=viro@zeniv.linux.org.uk \
--cc=will@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).