This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH 0/4] Provide C11 atomic operations
- From: Torvald Riegel <triegel at redhat dot com>
- To: Joseph Myers <joseph at codesourcery dot com>
- Cc: GLIBC Devel <libc-alpha at sourceware dot org>
- Date: Fri, 21 Nov 2014 14:11:55 +0100
- Subject: Re: [PATCH 0/4] Provide C11 atomic operations
- Authentication-results: sourceware.org; auth=none
- References: <1414617613 dot 10085 dot 23 dot camel at triegel dot csb> <1416495778 dot 1771 dot 34 dot camel at triegel dot csb> <alpine dot DEB dot 2 dot 10 dot 1411202048200 dot 5914 at digraph dot polyomino dot org dot uk>
On Thu, 2014-11-20 at 20:50 +0000, Joseph Myers wrote:
> On Thu, 20 Nov 2014, Torvald Riegel wrote:
> > I have committed this patch after some more code comparison for x86_64
> > pthread_once using a current GCC, and a regression check on x86_64. The
> > fast path has identical code. On the slow path, there are some minor
> > differences in which instructions are used, but I didn't spot anything
> > hat looked significant.
> > I've also fixed the Changelog; Adhemerval, thanks for spotting this.
> What is the basis for __HAVE_64B_ATOMICS in
> sysdeps/microblaze/bits/atomic.h? It looks to me as if the existing
> 64-bit operations there all abort.
It's a mistake. Thanks for spotting this. I have committed the
diff --git a/ChangeLog b/ChangeLog
index 76fda28..b0041ea 100644
@@ -1,3 +1,8 @@
+2014-11-21 Torvald Riegel <email@example.com>
+ * sysdeps/microblaze/bits/atomic.h (__HAVE_64B_ATOMICS): Fix
+ by setting it to 0. 64b atomics are not supported currently.
2014-11-21 Alexandre Oliva <firstname.lastname@example.org>
diff --git a/sysdeps/microblaze/bits/atomic.h
index 395162d..315c872 100644
@@ -35,7 +35,7 @@ typedef uintptr_t uatomicptr_t;
typedef intmax_t atomic_max_t;
typedef uintmax_t uatomic_max_t;
-#define __HAVE_64B_ATOMICS 1
+#define __HAVE_64B_ATOMICS 0
#define USE_ATOMIC_COMPILER_BUILTINS 0