This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug tree-optimization/47460] Inconsistent behaviour of __sync_fetch_and_add builtin?
- From: "redi at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 25 Jan 2011 16:41:02 +0000
- Subject: [Bug tree-optimization/47460] Inconsistent behaviour of __sync_fetch_and_add builtin?
- Auto-submitted: auto-generated
- References: <bug-47460-4@http.gcc.gnu.org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47460
--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-01-25 16:40:50 UTC ---
This seems more appropriate as a question to the gcc-help list requesting
clarification, not a bug report.
To answer part of it ...
(In reply to comment #0)
> I get the unexpected (for me) inconsistent behaviour of the
> __sync_fetch_and_add builtin with the program below. My main confusion is
> around the missing __sync_val_compare_and_swap_{1,2,4} when not explicitely
> specifying the architecture in GCC 4.4.5, but availability in all other tried
> variants.
You need to look at the output of 'gcc -v' to see if each GCC was built with
--with-arch, which specifies a default value for -march.
Prior to GCC 4.5, the default arch was i386 unless GCC was built with the
--with-arch switch. Since GCC 4.5 the default arch is implied by the target
machine.