Hi Gary,
* Gary V. Vaughan wrote on Thu, Nov 17, 2005 at 04:32:43PM CET:
> Ralf Wildenhues wrote:
> >>>What I don't know yet: should lt~obsolete.m4 also be installed in
> >>>$aclocaldir? Can this do any harm perchance?
> >
> >I still need to think
Backport from HEAD to branch-1-5.
Cheers,
Ralf
* libtool.m4 (AC_LIBTOOL_PROG_LD_SHLIBS)
(AC_LIBTOOL_LANG_CXX_CONFIG) [aix4/aix5 !ia64]: Set
whole_archive_flag_spec to $convenience so the convenience libs
are really added to the link cmdline.
Missing backpor
Sorry for the delay.
* Gary V. Vaughan wrote on Fri, Sep 16, 2005 at 08:26:22PM CEST:
> Ralf Wildenhues wrote:
> > On a related note, `-pg' should be handled better as well, noted by Tor
> > Lillqvist[2]. One issue here is that it changes predep_objects,
> > pos
1) This patch from HEAD:
| 2004-09-26 Albert Chin-A-Young <[EMAIL PROTECTED]>
|
| * m4/libtool.m4 (_LT_LINKER_SHLIBS) [aix]: In archive_expsym_cmds,
| move -bnoentry linker option before $compiler_flags to honor any
| possible -e entries in LDFLAGS.
http://lists.gnu.org/
Hi Albert, Dan,
* Albert Chin wrote on Tue, Nov 22, 2005 at 07:59:31PM CET:
> On Sun, Nov 20, 2005 at 08:08:50PM +0100, Ralf Wildenhues wrote:
> > 1) This patch from HEAD:
> >
> > has been applied to the C case only (and not backported to branch-1-5).
> > Shoul
Hi Max,
* Max Bowsher wrote on Thu, Nov 24, 2005 at 01:23:41PM CET:
>
> With --disable-static, when libtool falls back to producing a static
> archive due to undefined symbols, it produces an empty archive.
Thank you for the bug report. I have applied the following, very
similar backport from H
This is necessary to complete the Solaris -xarch=v9 bug fix.
Applied to branch-1-5.
Cheers,
Ralf
2005-11-24 Albert Chin-A-Young <[EMAIL PROTECTED]>
* libtool.m4, ltmain.in: When calling $LTCC, pass default $CFLAGS
through with $LTCFLAGS.
Backport from HEAD 2004-09-05.
In --mode=clean, we should not remove files which are not present,
to facilitate omitting the -f in "rm -f". Applied to both branches.
Cheers,
Ralf
* libltdl/config/ltmain.m4sh (func_mode_uninstall): In clean
mode, don't remove non-existent libfoo.lai.
Index: libltdl/config/ltma
Hi Peter,
* Peter O'Gorman wrote on Thu, Nov 24, 2005 at 04:32:33PM CET:
>
> Libtool should accept -isysroot to help people build fat projects on
> Mac OS X.
Erm, you mean -isysroot not --sysroot?
Why is -isysroot relevant for the linker?
(Am I being dense again because I don't understand fat arc
* Ralf Wildenhues wrote on Thu, Nov 24, 2005 at 06:15:54PM CET:
> * Peter O'Gorman wrote on Thu, Nov 24, 2005 at 04:32:33PM CET:
> >
> > Libtool should accept -isysroot to help people build fat projects on
> > Mac OS X.
>
> Erm, you mean -isysroot not --sysroot?
[ from the libtool list ]
* Ralf Wildenhues wrote on Thu, Nov 24, 2005 at 09:44:02AM CET:
> With response file support in GCC [1] we need to adjust Libtool
> accordingly. Minimally to let the option through as below, but
> ideally we should probably parse its contents.
Applied to
indir:"*) ;;
+ *) dllsearchpath="$dllsearchpath:$testbindir";;
+ esac
;;
esac
done
2005-11-25 Charles Wilson <[EMAIL PROTECTED]>,
Peter Ekberg <[EMAIL PROTECTED]>,
Ralf Wildenhu
> * Albert Chin wrote on Tue, Nov 22, 2005 at 07:59:31PM CET:
> > On Sun, Nov 20, 2005 at 08:08:50PM +0100, Ralf Wildenhues wrote:
> > > 1) This patch from HEAD:
> > >
> > > has been applied to the C case only (and not backported to branch-1-5).
> > >
Applied this to HEAD. Fixing the rest of this mess (including Charles'
other observed ltdl bug) is independent and will follow in another
patch.
Cheers,
Ralf
* Ralf Wildenhues wrote on Sun, Nov 13, 2005 at 11:11:56PM CET:
>
> 2005-11-13 Eric Blake <[EMAIL PROTECTED]>,
&
Moving from libtool to libtool-patches:
* Ralf Wildenhues wrote on Thu, Nov 24, 2005 at 06:50:54PM CET:
> > * Peter O'Gorman wrote on Thu, Nov 24, 2005 at 03:29:24PM CET:
> > > cd . && /bin/sh ./config.status libtool
> > > config.status: executing libtool
This patch should allow the libltdl directory to be named differently
also in nonrecursive mode. (Hi Bob!)
It always copies Makefile.inc, which is the only reasonable thing to do
in this situation anyway. It also always updates it, which might be
debatable, but seems safe to me.
OK to apply to
Applied to branch-1-5 (already fixed in HEAD).
Cheers,
Ralf
* libtool.m4: Fix comment typo.
* tests/dryrun.test: Avoid `directory exists' warning.
Index: libtool.m4
===
RCS file: /cvsroot/libtool/libtool/Attic/libtoo
Hi Peter,
* Peter O'Gorman wrote on Sat, Nov 26, 2005 at 02:25:34PM:
Okay?
Sure. Both HEAD and branch-1-5.
Out of curiosity, what are the failures with using $CC here
(thus possibly the C++ compiler) and of omitting flags?
Cheers,
Ralf
2005-11-26 Peter O'Gorman <[EMAIL PROTECTED]>
* Ralf Wildenhues wrote on Sat, Nov 26, 2005 at 10:40:26AM CET:
> This patch should allow the libltdl directory to be named differently
> also in nonrecursive mode.
Here's one that should work even if the subdir path includes the string
`libltdl'. OK?
Cheers,
Ralf
>
* Ralf Wildenhues wrote on Fri, Nov 25, 2005 at 07:36:23PM CET:
>
> I have applied these patches to branch-1-5. All backports of HEAD
> patches, some took more than one patch in HEAD to fix at once, but
> it makes little sense to relive those errand states. While the diffs
> look
Hi Peter,
* Peter O'Gorman wrote on Mon, Nov 28, 2005 at 04:43:44PM CET:
> Hi,
> I had been telling those who asked for years to put -arch flags etc in
> CC/CXX env vars before calling configure. My recent -isysroot patch broke
> the ability to do that, apple's linker rather stupidly dies on som
* Ralf Wildenhues wrote on Mon, Nov 28, 2005 at 05:41:44PM CET:
>
> Can it ever be useful to add more than one of those?
> Like this:
> -framework foo -framework bar
>
> If yes, then you should
> prev=darwin_framework
> framework_arg=$arg
> and, in the
>
Hi Peter,
* Peter O'Gorman wrote on Thu, Dec 01, 2005 at 04:59:32PM CET:
> Ralf Wildenhues wrote:
> >Also, I wonder why you leave out CFLAGS and such, so why not check
> >against $compile_command?
> You are right, of course. The original patch, by checking for CC and th
Hi Peter,
* Peter O'Gorman wrote on Fri, Dec 02, 2005 at 02:17:59AM CET:
> Ralf Wildenhues wrote:
> >
> >OK. But I don't see where we make sure $compile_command never gets
> >duplicates.. anyway, I guess this part is fine, you'd have seen it.
>
> archiv
[ moving from the autoconf and the ports at openbsd lists ]
* Ralf Wildenhues wrote on Thu, Dec 01, 2005 at 08:43:31PM CET:
> This one is the reason for a subtle CVS Libtool testsuite failure..
To be specific: demo-noinst-link.test would FAIL, because
'libtool --mode=link' bogu
* Ralf Wildenhues wrote on Fri, Dec 02, 2005 at 04:36:55PM CET:
>
> Any reason against applying these two patches to HEAD and branch-1-5
> respectively?
I didn't think so. ;)
Applied.
Cheers,
Ralf
> HEAD:
> * libltdl/config/general.m4sh (func_show_eval):
>
[ moving from the libtool list ]
Applied to HEAD.
Cheers,
Ralf
Basic support for PIE (position-independent executables).
* libltdl/config/ltmain.m4sh (func_generate_dlsyms):
Don't pass PIE flag for compilation of the symbol file object.
(func_mode_compile): Pass
Hi Gary,
* Gary V. Vaughan wrote on Thu, Dec 01, 2005 at 06:01:16PM CET:
> Ralf Wildenhues wrote:
> >Here's one that should work even if the subdir path includes the string
> >`libltdl'. OK?
>
> Nice work! Untested, but looks fine from inspection. Please
* Ralf Wildenhues wrote on Sat, Dec 03, 2005 at 09:50:23AM CET:
>
> > >> * libtoolize.m4sh (func_fixup_Makefile_inc): New function.
> > >> (main): Call it in nonrecursive mode to mangle Makefile.inc.
> > >> * tests/nonrecursive.at: adju
Hi Gary,
Oops, forgot about this one.
* Gary V. Vaughan wrote on Thu, Oct 27, 2005 at 12:09:22PM CEST:
> Ralf Wildenhues wrote:
> >OK to apply? The actual fix for Solaris/CVS HEAD is to follow soon.
>
> Yes please!
Thanks. Done.
There'll be quite a bit of systems/compi
Hi Peter,
* Peter O'Gorman wrote on Sat, Dec 03, 2005 at 03:00:38PM CET:
> Peter O'Gorman wrote:
> | I have been persuaded that -single_module should be the default (probably
> | only for HEAD for now patch coming soon).
>
> Okay for HEAD?
No testing done, and little experience with darwin, so i
With Gary's wiki, the TODO file has grown a bit outdated.
Well, all the better that a few entries are obsolete now. :)
Note the
wrong $ECHO after `config.status --recheck'
is actually fixed in CVS Autoconf by the change on 2005-11-10,
so you need a version newer than that, and also you need to
Hi Gary,
Somewhat older:
* Gary V. Vaughan wrote on Thu, Nov 17, 2005 at 05:30:51PM CET:
> Ralf Wildenhues wrote:
> >>It's a bug when
> >> libtoolize --ltdl
> >>succeeds although it did not find the libltdl source files.
> There is code in func_copy_cb tha
Applied to HEAD. We no longer use any of this.
Cheers,
Ralf
* libtoolize.m4sh (func_copy_all_files)
(glob_exclude_pkgaux_files): Removed, since unused.
Index: libtoolize.m4sh
===
RCS file: /cvsroot/libtool/libtool/
Applied to HEAD.
Rationale for not waiting for approval: triviality. :)
Cheers,
Ralf
* README, README.alpha, libtoolize.m4sh: Typos.
* tests/am-subdir.at, tests/duplicate_members.at: Make all
banners end with a dot, for symmetry.
Index: README
===
The test for README.alpha was bogus. Nothing copied it over into
$(distdir). Now we do that manually. Note you can't put it in
EXTRA_DIST either, because then, it won't be present in the tarball,
(since we may move it over to `README', or delete it), then `distcheck'
will naturally fail, because
I believe I have the distribution issues with libltdl sorted out.
In fact, along the way I fixed a few other things.
I'll be replying to this mail with a number of patches, separated by
logical changes, and *should* work in the order posted (I hope).
- two patches for compatibility with older Auto
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> I'll be replying to this mail with a number of patches, separated by
> logical changes, and *should* work in the order posted (I hope).
> - two patches for compatibility with older Automake
*snip*
Not really
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> I'll be replying to this mail with a number of patches, separated by
> logical changes, and *should* work in the order posted (I hope).
> - two patches for compatibility with older Automake
*snip*
OK, this
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> I'll be replying to this mail with a number of patches, separated by
> logical changes, and *should* work in the order posted (I hope).
> - several (dunno how many yet) patches to fix the actual `make dis
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
> - several (dunno how many yet) patches to fix the actual `make dist'
> stuff for clients.
This patch ensures that libltdl/aclocal.m4 and libltdl/Makefile.in are
always up to date. These dependencies are not real in
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> - several (dunno how many yet) patches to fix the actual `make dist'
> stuff for clients.
This patch is more subtle than it looks.
First, we note that, in the subpackage case, we do not install
Makefile.inc.
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> - several (dunno how many yet) patches to fix the actual `make dist'
> stuff for clients.
This patch is the last of the queue, and the most intertwined. Maybe I
should make the effort to rip it apart --
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 04:25:28PM CET:
>
> This patch ensures that libltdl/aclocal.m4 and libltdl/Makefile.in are
> always up to date. These dependencies are not real in the Libtool
> source tree, but they will be in the client source tree, in the
> subp
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 04:00:44PM CET:
> -# In the past libtool.m4 has provided macros via AC_DEFUN (or AU_DEFUN)
> +# In the past libtool.m4 has provided macros via AC_DEFUN (or AC_DEFUN)
> # which have later been changed to m4_define as they aren'
6 wasn't the last one then..
* Ralf Wildenhues wrote on Sun, Dec 04, 2005 at 03:56:15PM CET:
>
> README/README.alpha updates on how to use the new testsuite, and
> how to report failures it may show,
OK to apply?
Cheers,
Ralf
* README.alpha (Reporting bugs): S
With this patch, the info documentation builds with texinfo-4.2
(tested), probably older, instead of >= 4.6.
Applied to HEAD.
Cheers,
Ralf
* doc/libtool.texi (References): Do not use `@/' for the benefit
of older `makeinfo', break lines instead.
Index: doc/libtool.texi
=
Hi Bob,
* Bob Friesenhahn wrote on Mon, Dec 05, 2005 at 02:55:35AM CET:
> Ralph provided me with some patches to get development libtool
> integrated into development GraphicsMagick using a non-recursive
> build, but I have run into a hang-up. The problem is that libtool
> tests include code l
Hi Bob,
* Bob Friesenhahn wrote on Mon, Dec 05, 2005 at 05:39:31PM CET:
> I notice that libltdl/lt__glibc.h conditionally includes .
> HAVE_ARGZ_H is not defined under Solaris in the configuration header
> file. Regardless, when this file is encountered, it seems that
> HAVE_ARGZ_H is somehow
* Bob Friesenhahn wrote on Mon, Dec 05, 2005 at 05:00:47PM CET:
>
> So the only solution using released tools is to pollute the top of the
> source tree with copies of files. How unpleasant!
Erm. Libtool 2.0 isn't released either. Not yet.
> Is there any ETA for a new Autoconf release?
I ca
* Bob Friesenhahn wrote on Mon, Dec 05, 2005 at 06:45:26PM CET:
> On Mon, 5 Dec 2005, Bob Friesenhahn wrote:
> >I see that Makefile.inc includes
> >
> >AM_CPPFLAGS += -DLT_CONFIG_H='<$(LT_CONFIG_H)>' \
> > -DLTDL -I. -I$(srcdir) -Ilibltdl \
> > -I
* Bob Friesenhahn wrote on Tue, Dec 06, 2005 at 03:48:04AM CET:
> On Mon, 5 Dec 2005, Bob Friesenhahn wrote:
> >
> >So I am at a loss as to why Automake fails to build ltdl/libltdlc.la
> >before attempting to use it.
>
> FYI, a quick fix to the build-order problem is
>
> # Convince libltdl to
* Ralf Wildenhues wrote on Tue, Dec 06, 2005 at 05:39:57AM CET:
> * Bob Friesenhahn wrote on Tue, Dec 06, 2005 at 03:48:04AM CET:
> > On Mon, 5 Dec 2005, Bob Friesenhahn wrote:
> > >
> > >So I am at a loss as to why Automake fails to build ltdl/libltdlc.la
>
Hi Peter,
* Peter O'Gorman wrote on Tue, Dec 06, 2005 at 03:41:10PM CET:
> Ralf Wildenhues wrote:
> >
> >Note indenting is inconsistenly wrt. TABs/spaces.
>
> Which is it, spaces or tabs? Sometimes lines start with 8 spaces, sometimes
> a tab. If editor is not
* Eric Blake wrote on Tue, Dec 06, 2005 at 09:13:42PM CET:
> > Using latest CVS autoconf/automake I am encountering this sort of
> > problem when bootstrapping/configuring libtool:
> >
> > config.status: executing libtool commands
> > sed: can't read libltdl/config/ltmain.sh: No such file or dire
Hi Eric,
* Eric Blake wrote on Wed, Dec 07, 2005 at 02:39:34PM CET:
> According to Ralf Wildenhues on 12/7/2005 2:30 AM:
> >>>In fact, while running the bootstrap script, there are complaints
> >>>about this "missing" file for every direc
* Marc Espie wrote on Tue, Dec 06, 2005 at 08:27:47PM CET:
> The preserve_args fragment was missing, thus the generated relink commands
> were bogus.
I have applied this to branch-1-5.
Cheers,
Ralf
2005-12-07 Marc Espie <[EMAIL PROTECTED]>
* ltmain.in: Preserve both parts of `--tag TA
Hi Peter,
* Peter O'Gorman wrote on Wed, Dec 07, 2005 at 02:25:23PM CET:
>
> I just applied this patch to HEAD. Fixes the archive_expsym_cmds too.
Thanks!
> Note that I do not have any idea if xlc support still works, I added
> it a couple of years ago based on a trial version of the compiler,
Hi Eric,
* Eric Blake wrote on Thu, Dec 08, 2005 at 02:49:19PM CET:
>
> libltdl/ltdl.c:2042: warning: passing arg 1 of `free' discards qualifiers
> from pointer target type
>
> So, should we change libltdl/libtldl/lt__alloc.h to have the FREE(mem)
> macro cast away const by calling free ((void *)
[ see http://article.gmane.org/gmane.os.openbsd.ports/15455 ]
* Marc Espie wrote on Fri, Dec 09, 2005 at 12:13:45AM CET:
>
> This includes the former --tag CXX fix, plus the part that makes
> --tag=disable-static
> --tag=disable-shared
>
> independent from other tags by delaying their effect.
l
Hi Peter,
* Peter O'Gorman wrote on Sat, Dec 10, 2005 at 02:58:59PM CET:
> Peter O'Gorman wrote:
*snip*
>
> Thanks for not shouting at me for this!
I haven't even looked at it yet.. ;-)
> Here is a patch that has a slightly better chance of doing the right
> thing.
Looks good, although for loo
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 11:06:16AM CET:
>
> I'm back in the UK for a few weeks :-D
Cool.
First of all, let me thank you for all your reviews.
I'll apply everything soon, but commenting on one issue for now only:
> Ralf Wildenhues wrote:
>
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 11:06:16AM CET:
> Ralf Wildenhues wrote:
> >
> >Not really sure about how to handle old Automake versions.
> >For now I have only done the latter. That way, the mechanism to detect
> >this and flag `expected failure&#
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 11:12:07AM CET:
> Ralf Wildenhues wrote:
> > * libltdl/m4/lt~obsolete.m4: Update documentation for aclocal
> > versions before 1.7. Change all macros to use AC_DEFUN instead
> > of AU_DEFUN.
> &
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 12:41:28PM CET:
> Ralf Wildenhues wrote:
> > * tests/testsuite.at (PREPARE_TESTS): New variable `tst_dist',
> > default to empty.
> > * tests/nonrecursive.at, tests/recursive.at,
> > tests/st
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 12:52:22PM CET:
> Ralf Wildenhues wrote:
> >
> >* Makefile.am (libltdl/Makefile.in): also depend on
> >libltdl/aclocal.m4, so that we guarantee timestamp consistency.
> >(libltdl/st
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 01:14:40PM CET:
> Ralf Wildenhues wrote:
> >First, we note that, in the subpackage case, we do not install
> >Makefile.inc. Since we have to be correct for this case, remove
> >`Makefile.inc' from EXTRA_DIST. R
Hi Gary,
* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 02:00:55PM CET:
> Ralf Wildenhues wrote:
> >6 wasn't the last one then..
>
> ;-) There is never a last one! :-p
Nope. The last couple of weeks have also re-filled the queue of
things-yet-to-fix..
> >* Ralf Wi
* Gary V. Vaughan wrote on Fri, Dec 16, 2005 at 06:14:44PM CET:
> Ralf Wildenhues wrote:
> >* Gary V. Vaughan wrote on Wed, Dec 14, 2005 at 12:41:28PM CET:
> >>
> >>As an improvement, maybe we should test all possibilities in each
> >>relevant test:
> >
Hi Kurt,
* Kurt Roeckx wrote on Tue, Dec 13, 2005 at 06:17:28PM CET:
> On Tue, Dec 13, 2005 at 10:26:22AM +0100, Ralf Wildenhues wrote:
> >
> > > old-m4-iface.at:135: $AUTOCONF --force
> > > stderr:
> > > Usage: autoconf [-h] [--help] [-m dir] [--macrodir=di
Hi Nicolas,
* Nicolas Kaiser wrote on Wed, Dec 14, 2005 at 04:07:16PM CET:
> Some documentation typos.
Thanks! Applied to branch-1-5 (all fixed in HEAD already).
Cheers,
Ralf
2005-12-16 Nicolas Kaiser <[EMAIL PROTECTED]>
* doc/libtool.texi: Fix some documentation typos.
Index: doc/
* Marc Espie wrote on Thu, Dec 15, 2005 at 07:46:41PM CET:
>
> On OpenBSD, the ld.so path is /usr/lib, and never includes /lib. Thus,
> do not put /lib there... this is more correct, and removes some useless
> tests...
I assume that Brad's comment was about the other part of the patch
only(?). S
Sorry for the delay.
* Peter O'Gorman wrote on Fri, Dec 09, 2005 at 01:41:55PM CET:
> Ralf Wildenhues wrote:
> | [ see http://article.gmane.org/gmane.os.openbsd.ports/15455 ]
> |
> | * Marc Espie wrote on Fri, Dec 09, 2005 at 12:13:45AM CET:
> |
> |>This includes the
[ moving from libtool list ]
* Ralf Wildenhues wrote on Sun, Dec 18, 2005 at 01:53:10PM CET:
>
> Issues:
> 1) There is no default value for the CXX tag (should be unsupported).
>Trivial bug.
Applied to HEAD and branch-1-5.
Cheers,
Ralf
HEAD:
* libltdl/m
Sorry for the delay.
* Ralf Wildenhues wrote on Sun, Nov 13, 2005 at 08:59:20PM CET:
> Regarding the AC_PROG_NM issue:
>
> * Ralf Wildenhues wrote on Sat, Nov 05, 2005 at 08:54:26AM CET:
> >
> > Then AFAICS there's only the /usr/ccs/bin/elf path issue left, and t
Hi Olly,
Sorry for the delay.
* Olly Betts wrote on Sat, Dec 10, 2005 at 05:11:14AM CET:
> On 2005-12-03, Ralf Wildenhues <[EMAIL PROTECTED]> wrote:
> > There'll be quite a bit of systems/compilers where this will fail.
> > I wonder whether libtool should just go
[ moving from the ports at openbsd, and libtool lists ]
> * Jacob Meuser wrote on Tue, Nov 29, 2005 at 05:19:44AM CET:
> > > Ralf Wildenhues wrote:
> > >
> > > | 1) Move all paths to uninstalled libraries (in the correct order) before
Applied to branch-1-5 as belo
Applied to branch-1-5 and HEAD, respectively.
Cheers,
Ralf
branch-1-5:
* ltmain.in :
Initialize properly.
Index: ltmain.in
===
RCS file: /cvsroot/libtool/libtool/Attic/ltmain.in,v
retrieving revision 1.334.2.112
diff
urs if noone complains.)
Cheers,
Ralf
2005-12-15 Peter O'Gorman <[EMAIL PROTECTED]>,
Ralf Wildenhues <[EMAIL PROTECTED]>
* ltmain.in (link mode): Don't die if -L does not
exist. Put in notinst_path though, so it won't end up in the
Applied to branch-1-5.
Cheers,
Ralf
GNU libtool 1.5.22 is released.
* configure.ac: Bump version.
* NEWS, THANKS: Updated.
* config.guess, config.sub, install-sh, texinfo.tex: Updated
from external sources.
* libtool.m4, ltdl.m4: Bump serial.
Applied to branch-1-5 (already fixed in HEAD).
Cheers,
Ralf
* ltmain.in (func_mktempdir): New, backported from HEAD.
(link mode): Use it. Fixes potential denial of service through
malicious other local user.
Reported by James Kosin <[EMAIL PROTECTED]>.
Index: lt
OK, time I messed up. I bootstrapped 1.5.22 with CVS versions of
Autoconf and Automake. Darn.
What now? Leave things as they are? Jump to 1.5.24?
Given the fixed bugs in autotools I'm leaning towards the former,
but I'm undecided on this.
Cheers, and sorry,
Ralf
Hi Peter,
* Peter O'Gorman wrote on Mon, Dec 19, 2005 at 12:14:22AM CET:
> Ralf Wildenhues wrote:
> | OK, time I messed up. I bootstrapped 1.5.22 with CVS versions of
> | Autoconf and Automake. Darn.
> |
> | What now? Leave things as they are? Jump to 1.5.24?
>
[ getting libtool-patches into play; this thread is archived at
http://sourceforge.net/mailarchive/forum.php?thread_id=9231830&forum_id=5119 ]
* Pierre Ossman wrote on Mon, Dec 19, 2005 at 01:40:22PM CET:
> Ralf Wildenhues wrote:
> >
> >OK, thanks for doing this. Now I
* James Kosin wrote on Mon, Dec 19, 2005 at 05:02:33PM CET:
>
> It ran through the tests and worked. Sorry, for the late response, I
> don't have access on weekends to rebuild the package. This weekend I
> graduated from colledge and Sunday is usually busy for me anyway.
Well, congratulations
Hi Eric,
* Eric Blake wrote on Mon, Dec 19, 2005 at 04:08:39PM CET:
> Missed one. Apply this in addition to the previous mail.
D'oh. I don't know how I could forget that. :-(
Thanks for the patch, applied to both branches.
Cheers,
Ralf
> 2005-12-19 Eric Blake <[EMAIL PROTECTED]>
>
>
* Eric Blake wrote on Tue, Dec 20, 2005 at 02:21:40PM CET:
> According to Ralf Wildenhues on 12/19/2005 9:31 AM:
> >
> > Thanks for the patch, applied to both branches.
>
> You got the correct ChangeLog, but only applied my patch to HACKING. You
> need to also appl
[ adding subscribers-only mingw-users back -- they may have useful
input on this ]
Hi Bob,
* Bob Friesenhahn wrote on Mon, Dec 19, 2005 at 06:46:21PM CET:
On Mon, 19 Dec 2005, Ralf Wildenhues wrote:
>
>Question to libtool folks: OK to apply this patch?
There are other reasons not t
* Keith MARSHALL wrote on Thu, Dec 22, 2005 at 06:17:57PM CET:
> Ralf Wildenhues wrote, quoting Bob Frieshahn:
> >
> > OK, so the only remaining case really is MSYS/MinGW itself. Do we
> > know of win32 installations that provide a `file' command that does
> > not
Hi Keith,
* Keith MARSHALL wrote on Fri, Dec 23, 2005 at 11:17:08AM CET:
> Ralf Wildenhues wrote:
> > http://sourceforge.net/mailarchive/message.php?msg_id=13069136 which
>
> If I interpret that correctly, you are specifically asking for `join'
> and `paste', from G
* Ralf Wildenhues wrote on Thu, Dec 29, 2005 at 02:46:14PM CET:
> * Edward Chernenko wrote on Wed, Dec 28, 2005 at 12:34:06PM CET:
> > Hi, my system is LFS/Linux, C compiler TCC 0.9.23 and
> Interestingly, tcc uncovered unportable use of parentheses around string
> literal initia
Hi Bob,
* Bob Friesenhahn wrote on Thu, Dec 29, 2005 at 05:01:18AM CET:
> I configured CVS libtool to use the prefix /usr/local/autotools-devel
> and autoconf/automake also use that prefix and come first in the PATH.
> I edited libltdl/Makefile.inc and installed libtool. The file
> /usr/local/a
Applied to HEAD. Without this the colon ended up at the very beginning
of LD_LIBRARY_PATH instead of between the uninstalled paths, as in
LD_LIBRARY_PATH=:/some/temp/path/.libs$LD_LIBRARY_PATH
(not present in branch-1-5).
Cheers,
Ralf
* libltdl/config/ltmain.m4sh (func_mode_link) :
Hi Bob,
* Bob Friesenhahn wrote on Thu, Dec 29, 2005 at 05:51:49PM CET:
> On Thu, 29 Dec 2005, Ralf Wildenhues wrote:
> >
> >I haven't looked in libtoolize or Makefile.am for looming bugs. Very
> >respectfully would I like to suggest that what you are seeing *could*
>
* Bob Friesenhahn wrote on Thu, Dec 29, 2005 at 06:35:27PM CET:
> On Thu, 29 Dec 2005, Ralf Wildenhues wrote:
> >>It seems likely that 'make' after configure was not sufficient to
> >>regenerate some files. I did a 'make clean' and re-configured a
Hi Bob,
* Bob Friesenhahn wrote on Thu, Dec 29, 2005 at 09:26:23PM CET:
> This proposed patch is necessary in order to be able to use the
> Makefile.inc installed by 'libtoolize --nonrecursive' since otherwise
> components are requested which are not installed. Ok to apply?
See inline comments
Hi Alexis,
* Alexis Wilke wrote on Sat, Dec 31, 2005 at 08:04:05AM CET:
>
> There is a really bad bug in LT_DLMUTEX_GETERROR() which checks if the
> lt_dlmutex_seterror_func [see SET] instead of the lt_dlmutex_geterror_func
> [see, here GET].
Thanks for this bug report. I have applied this patc
Hi Bob,
* Bob Friesenhahn wrote on Fri, Dec 30, 2005 at 04:33:39PM CET:
> On Fri, 30 Dec 2005, Ralf Wildenhues wrote:
>
> >>--- libltdl/Makefile.inc16 Dec 2005 16:36:20 - 1.5
> >>+++ libltdl/Makefile.inc29 Dec 2005 20:24:40 -
> >>@@ -27
Hi Gary,
Is it beneficial for arch to do the ChangeLog rotation with it
rather than from within CVS?
Cheers,
Ralf
Hi Kurt, Gary,
* Kurt Roeckx wrote on Thu, Dec 29, 2005 at 07:52:53PM CET:
>
> It seems 1.5 and head are a little bit different when it comes to
> linking a binary. It seems that for linking a binary it does the
> same as in case of a conviences library, and adds all
> depedency_libs to deplibs.
1201 - 1300 of 2537 matches
Mail list logo
