This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Transition to C11 atomics and memory model
- From: Torvald Riegel <triegel at redhat dot com>
- To: "Joseph S. Myers" <joseph at codesourcery dot com>
- Cc: GLIBC Devel <libc-alpha at sourceware dot org>
- Date: Wed, 17 Sep 2014 16:07:16 +0200
- Subject: Re: Transition to C11 atomics and memory model
- Authentication-results: sourceware.org; auth=none
- References: <1410719669 dot 4967 dot 160 dot camel at triegel dot csb> <1410904080 dot 4967 dot 400 dot camel at triegel dot csb> <Pine dot LNX dot 4 dot 64 dot 1409162159110 dot 4182 at digraph dot polyomino dot org dot uk>
On Tue, 2014-09-16 at 22:02 +0000, Joseph S. Myers wrote:
> On Tue, 16 Sep 2014, Torvald Riegel wrote:
>
> > * mips use of __sync_lock_test_and_set for both _acq and _rel is
> > surprising. Is this __sync builtin different on mips, or is the _rel
> > variant broken and needs a stronger barrier?
>
> It sounds like a bug - specific to the case of (MIPS16, GCC before 4.7),
> as that's the only case where __sync_* get used on MIPS, and that case may
> not be widely used. Please file in Bugzilla. (__sync_* for MIPS16 then
> in turn ends up using out-of-line libgcc functions that call __sync_*
> built as non-MIPS16 code, but I see nothing special there to give a
> stronger barrier.)
>
https://sourceware.org/bugzilla/show_bug.cgi?id=17404