Re: [PATCH] Correct link mode tag inference (Was: tag inference broken?)

[Scott James Remnant]

On Wed, 2003-12-31 at 00:48, Bob Friesenhahn wrote:

> On Wed, 31 Dec 2003, Scott James Remnant wrote:
> 
> > I'd like to commit this, it seems to fix the problem.  Are there any
> > objections?
> 
> If libtool still passes its test suite, then I have no objection.
> 
Without the patch, at the moment libtool can't even built itself if CC
contains multiple arguments.  For example's sake, I'll (incorrectly)
stick -march=athlon in there as well as prefixing with ccache.


Before the patch:

$ ./configure && make && make check
        :
====================
All 108 tests passed
====================

$ ./configure CC="ccache gcc -march=athlon" && make && make check
        :
/bin/sh ../libtool --mode=link ccache gcc -march=athlon  -g -O2   -o libltdl.la 
-rpath /usr/local/lib -no-undefined -version-info 4:0:1 ltdl.lo -ldl
libtool: link: unable to infer tagged configuration
libtool: link: specify a tag with `--tag'
make[2]: *** [libltdl.la] Error 1


After the patch:

$ ./configure && make && make check
        :
====================
All 108 tests passed
====================

$ ./configure CC="ccache gcc -march=athlon" && make && make check
        :
====================
All 108 tests passed
====================


I've noticed the patch wasn't my usual -p1 with ChangeLog as I thought
it was, just a quick one I mailed to the list while discussing it. 
Here's the ChangeLog entry:

2003-12-31  Scott James Remnant  <address@hidden>
 
        * ltmain.sh: Infer tagged configuration for link mode with the
        full compiler arguments available rather than only the first.
 
Scott
-- 
Have you ever, ever felt like this?
Had strange things happen?  Are you going round the twist?

signature.asc
Description: This is a digitally signed message part