bug-autoconf
[Top][All Lists]
Advanced

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

Re: bug in autoconf-2.64


From: Ralf Wildenhues
Subject: Re: bug in autoconf-2.64
Date: Thu, 24 Feb 2011 07:24:35 +0100
User-agent: Mutt/1.5.20 (2010-08-04)

* Mike Stump wrote on Thu, Feb 24, 2011 at 12:06:07AM CET:
> On Feb 23, 2011, at 1:37 PM, Eric Blake wrote:
> > Are you on a machine with SSE4.2 instructions?
> 
> Core 2 Duo.  darwin10 (aka SnowLeopard).  No glibc.

Here no SSE4.2 either.  In an up to date build tree of the commit that
exposes the issue, I can turn the build failure into success if I preset

gl_cv_func_strstr_linear=yes gl_cv_func_strstr_works_always=yes

and I can flip between success and failure by editing config.status
like this: set REPLACE_STRSTR to 1, add strstr.lo to M4_LTLIBOBJS
(and strstr.o to M4_LIBOBJS), removing lib/string.h, rerunning
config.status and make.

IOW, it looks like the replacement code in strstr.c and str-two-way.h
has a bug (or glibc strchr, which seems rather unlikely).  Besides
copyright year bumps, these two files have not been updated since in
gnulib.

Cheers,
Ralf



reply via email to

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