bug-gnulib
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: gnulib:test-stdalign failing on powerpc-darwin8, gcc-4.0


From: David Fang
Subject: Re: gnulib:test-stdalign failing on powerpc-darwin8, gcc-4.0
Date: Tue, 16 Dec 2014 19:15:27 -0500 (EST)

Hi Paul,

On 12/15/2014 04:47 PM, David Fang wrote:
_alignas and _Alignas are broken on your platform, even with gnulib trying to work around the breakage.

 Is there a problem in the compiler, or rather some incorrect assumption
 about the ABI in the test itself?  This is the system compiler, after all.

Most likely the compiler is broken. Some older GCCs do have a bug here. Gnulib knows about the breakage and attempts to work around it, but apparently this is not working for you, and you need to figure out why.

Does config.log report stdalign.h as working? If so, that's a bug in the Gnulib test in m4/stdalign.m4: it should detect the bug in your system. If not, you need to investigate why the Gnulib workaround does not in fact work around your compiler bug.

In my build area for gnuTLS, the config.log reports:

configure:22801: checking for working stdalign.h
configure:22839: gcc -c -g -O2 -I/sw/include conftest.c >&5
conftest.c:191:22: error: stdalign.h: No such file or directory

since the compiler pre-dates C11.
Does this mean that it falls back to the distributed copy in gl/stdalign.h? That file matches the current stdalign.h in gnulib's git-master, except for comments at the top.

I don't see any evidence of the replacement stdalign.h being tested in config.log.

Fang

--
David Fang
http://www.csl.cornell.edu/~fang/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]