[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20190909062111.GA14572@gondor.apana.org.au>
Date: Mon, 9 Sep 2019 16:21:11 +1000
From: Herbert Xu <herbert@...dor.apana.org.au>
To: Peter Zijlstra <peterz@...radead.org>
Cc: torvalds@...ux-foundation.org, mathieu.desnoyers@...icios.com,
tglx@...utronix.de, joel@...lfernandes.org,
stern@...land.harvard.edu, rostedt@...dmis.org,
valentin.schneider@....com, linux-kernel@...r.kernel.org,
paulmck@...ux.ibm.com, boqun.feng@...il.com, will.deacon@....com,
dhowells@...hat.com
Subject: Re: [PATCH 1/1] Fix: trace sched switch start/stop racy updates
Peter Zijlstra <peterz@...radead.org> wrote:
>
> Paulmck actually has an example of that somewhere; ISTR that particular
> case actually got fixed by GCC, but I'd really _love_ for some compiler
> people (both GCC and LLVM) to state that their respective compilers will
> not do load/store tearing for machine word sized load/stores.
>
> Without this written guarantee (which supposedly was in older GCC
> manuals but has since gone missing), I'm loathe to rely on it.
IIRC in that case gcc actually broke atomic writes even with a
volatile keyword. So even WRITE_ONCE wouldn't have saved us.
Cheers,
--
Email: Herbert Xu <herbert@...dor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Powered by blists - more mailing lists