GNU bug report logs -
#8441
Patches making libtool-2.4-1 build under GNU/Hurd
Previous Next
To reply to this bug, email your comments to 8441 AT debbugs.gnu.org.
Toggle the display of automated, internal messages from the tracker.
Report forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Thu, 07 Apr 2011 17:18:01 GMT)
Full text and
rfc822 format available.
Acknowledgement sent
to
srs <at> kth.se:
New bug report received and forwarded. Copy sent to
bug-libtool <at> gnu.org.
(Thu, 07 Apr 2011 17:18:02 GMT)
Full text and
rfc822 format available.
Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):
[Message part 1 (text/plain, inline)]
merge 603971 612204
found 612204 2.4-1
thanks
Package: libtool
Version: 2.4-1
Severity: important
Tags: patch
User: debian-hurd <at> lists.debian.org
Usertags: hurd
Updated patches to make libtool-2.4 build under GNU/Hurd and the
corresponding series file.
The first defines shlibpath_overrides_runpath for gnu.
The second disables the test 102 of a low setting of max_cmd_len since
it has no meaning under GNU/Hurd.
The third updates the previous link_all_deplibs patch adding gnu to the
host-os list combining the patch in #603971. Therefore the two bugs are
merged.
[series (text/plain, attachment)]
[define_shlibpath_overrides_runpath4hurd.patch (text/x-patch, attachment)]
[disable_cmdline_wrap4hurd.patch (text/x-patch, attachment)]
[link_all_deplibs.patch (text/x-patch, attachment)]
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Thu, 07 Apr 2011 18:07:02 GMT)
Full text and
rfc822 format available.
Message #8 received at 8441 <at> debbugs.gnu.org (full text, mbox):
Hi Svante,
* Svante Signell wrote on Thu, Apr 07, 2011 at 06:24:20PM CEST:
> The first defines shlibpath_overrides_runpath for gnu.
Is this always the correct setting? binutils can be configured with
either as default.
> # shlibpath_overrides_runpath is set to 'unknown' in libtool.m4
Oh. Then you have a good point. This should go upstream.
> # and not defined under $host_os ="gnu"
> # This patch make the tests/*demo* run.
> --- libtool-2.4/libltdt/m4/libtool.m4.orig 2011-02-03 21:33:56.000000000 +0100
> +++ libtool-2.4/libltdl/m4/libtool.m4 2011-02-03 21:43:46.000000000 +0100
> @@ -2325,6 +2325,7 @@
> library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}'
> soname_spec='${libname}${release}${shared_ext}$major'
> shlibpath_var=LD_LIBRARY_PATH
> + shlibpath_overrides_runpath=no
> hardcode_into_libs=yes
> ;;
> # This test is broken for hurd-i386
> # Skip for $host_os = "gnu", since max_cmd_len=-1 to indicate that
> # max_cmd_len is not limited under GNU Hurd.
> #
> --- libtool-2.4/tests/cmdline_wrap.at.orig 2010-05-20 23:18:41.000000000 +0200
> +++ libtool-2.4/tests/cmdline_wrap.at 2011-02-06 11:08:29.000000000 +0100
> @@ -30,6 +30,9 @@
> m4_ifdef([AT_CAPTURE_FILE],
> [AT_CAPTURE_FILE([testsuite.log])])
>
> +# Setting low max_cmd_len on "hurd-386" does not make sense"
> +AT_CHECK([if test "$host_os" == "gnu" ; then exit 77; fi])
Please use = not ==, the former is Posix test, the latter bash-specific.
Does this test actually fail on hurd-i386, or is this just because you
consider it wasted time? In the former case, please send the patch
upstream but also show a cut-n-paste of the failing test log.
Thanks!
Ralf
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Thu, 07 Apr 2011 18:22:02 GMT)
Full text and
rfc822 format available.
Message #11 received at 8441 <at> debbugs.gnu.org (full text, mbox):
On Thu, 2011-04-07 at 20:06 +0200, Ralf Wildenhues wrote:
> Hi Svante,
>
> * Svante Signell wrote on Thu, Apr 07, 2011 at 06:24:20PM CEST:
> > The first defines shlibpath_overrides_runpath for gnu.
>
> Is this always the correct setting? binutils can be configured with
> either as default.
>
> > # shlibpath_overrides_runpath is set to 'unknown' in libtool.m4
>
> Oh. Then you have a good point. This should go upstream.
OK!
> > # and not defined under $host_os ="gnu"
> > # This patch make the tests/*demo* run.
> > --- libtool-2.4/libltdt/m4/libtool.m4.orig 2011-02-03 21:33:56.000000000 +0100
> > +++ libtool-2.4/libltdl/m4/libtool.m4 2011-02-03 21:43:46.000000000 +0100
> > @@ -2325,6 +2325,7 @@
> > library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}'
> > soname_spec='${libname}${release}${shared_ext}$major'
> > shlibpath_var=LD_LIBRARY_PATH
> > + shlibpath_overrides_runpath=no
> > hardcode_into_libs=yes
> > ;;
>
> > # This test is broken for hurd-i386
> > # Skip for $host_os = "gnu", since max_cmd_len=-1 to indicate that
> > # max_cmd_len is not limited under GNU Hurd.
> > #
> > --- libtool-2.4/tests/cmdline_wrap.at.orig 2010-05-20 23:18:41.000000000 +0200
> > +++ libtool-2.4/tests/cmdline_wrap.at 2011-02-06 11:08:29.000000000 +0100
> > @@ -30,6 +30,9 @@
> > m4_ifdef([AT_CAPTURE_FILE],
> > [AT_CAPTURE_FILE([testsuite.log])])
> >
> > +# Setting low max_cmd_len on "hurd-386" does not make sense"
> > +AT_CHECK([if test "$host_os" == "gnu" ; then exit 77; fi])
>
> Please use = not ==, the former is Posix test, the latter bash-specific.
OK, do you want me to send an update?
> Does this test actually fail on hurd-i386, or is this just because you
> consider it wasted time? In the former case, please send the patch
> upstream but also show a cut-n-paste of the failing test log.
Yes, this test fails without the patch. I'll build an unpatched version
and submit the failing test log. BTW: Where is upstream, is
bug-libtool <at> gnu.org sufficient?
Thanks,
Svante
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Thu, 07 Apr 2011 18:25:02 GMT)
Full text and
rfc822 format available.
Message #14 received at 8441 <at> debbugs.gnu.org (full text, mbox):
* Svante Signell wrote on Thu, Apr 07, 2011 at 08:20:55PM CEST:
> On Thu, 2011-04-07 at 20:06 +0200, Ralf Wildenhues wrote:
> > > --- libtool-2.4/tests/cmdline_wrap.at.orig 2010-05-20 23:18:41.000000000 +0200
> > > +++ libtool-2.4/tests/cmdline_wrap.at 2011-02-06 11:08:29.000000000 +0100
> > > @@ -30,6 +30,9 @@
> > > +# Setting low max_cmd_len on "hurd-386" does not make sense"
> > > +AT_CHECK([if test "$host_os" == "gnu" ; then exit 77; fi])
> >
> > Please use = not ==, the former is Posix test, the latter bash-specific.
>
> OK, do you want me to send an update?
I can't speak for Kurt, I only help maintain upstream. ;-)
> > Does this test actually fail on hurd-i386, or is this just because you
> > consider it wasted time? In the former case, please send the patch
> > upstream but also show a cut-n-paste of the failing test log.
>
> Yes, this test fails without the patch. I'll build an unpatched version
> and submit the failing test log. BTW: Where is upstream, is
> bug-libtool <at> gnu.org sufficient?
Sure, but you can just as well send a patch right to libtool-patches
(first-post moderation, but subscribing is not required).
Thanks,
Ralf
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Fri, 08 Apr 2011 07:03:02 GMT)
Full text and
rfc822 format available.
Message #17 received at 8441 <at> debbugs.gnu.org (full text, mbox):
[Message part 1 (text/plain, inline)]
On Thu, 2011-04-07 at 20:24 +0200, Ralf Wildenhues wrote:
> * Svante Signell wrote on Thu, Apr 07, 2011 at 08:20:55PM CEST:
> > On Thu, 2011-04-07 at 20:06 +0200, Ralf Wildenhues wrote:
> > > > --- libtool-2.4/tests/cmdline_wrap.at.orig 2010-05-20 23:18:41.000000000 +0200
> > > > +++ libtool-2.4/tests/cmdline_wrap.at 2011-02-06 11:08:29.000000000 +0100
> > > > @@ -30,6 +30,9 @@
>
> > > > +# Setting low max_cmd_len on "hurd-386" does not make sense"
> > > > +AT_CHECK([if test "$host_os" == "gnu" ; then exit 77; fi])
> > >
> > > Please use = not ==, the former is Posix test, the latter bash-specific.
> >
> > OK, do you want me to send an update?
>
> I can't speak for Kurt, I only help maintain upstream. ;-)
Updated!
> > > Does this test actually fail on hurd-i386, or is this just because you
> > > consider it wasted time? In the former case, please send the patch
> > > upstream but also show a cut-n-paste of the failing test log.
> >
> > Yes, this test fails without the patch. I'll build an unpatched version
> > and submit the failing test log. BTW: Where is upstream, is
> > bug-libtool <at> gnu.org sufficient?
>
> Sure, but you can just as well send a patch right to libtool-patches
> (first-post moderation, but subscribing is not required).
I tried libtool-patches, but I had to register to be able to submit
there. Comments are embedded in the patches.
The test fails without the patch, see below (partial info I found in the
tests/testsuite.log file on test 112):
(where to find the complete log if this test, in
tests/testsuite.dir/112/testsuite.lo or
tests/testsuite.dir/112/tests/testsuite.dir/testsuilt.log? Both files
are very large)
112: cmdline_wrap.at:28 Run tests with low max_cmd_len
recursive
112. cmdline_wrap.at:28: testing ...
./cmdline_wrap.at:43: $CONFIG_SHELL $abs_srcdir/testsuite -k libtool
$INNER_TESTSUITEFLAGS
stderr:
ERROR: 43 tests were run,
4 failed (3 expected failures).
10 tests were skipped.
[define_shlibpath_overrides_runpath4hurd.patch (text/x-patch, attachment)]
[disable_cmdline_wrap4hurd.patch (text/x-patch, attachment)]
[link_all_deplibs.patch (text/x-patch, attachment)]
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Fri, 08 Apr 2011 14:38:02 GMT)
Full text and
rfc822 format available.
Message #20 received at 8441 <at> debbugs.gnu.org (full text, mbox):
On Fri, 2011-04-08 at 09:01 +0200, Svante Signell wrote:
> On Thu, 2011-04-07 at 20:24 +0200, Ralf Wildenhues wrote:
> > * Svante Signell wrote on Thu, Apr 07, 2011 at 08:20:55PM CEST:
> > > On Thu, 2011-04-07 at 20:06 +0200, Ralf Wildenhues wrote:
> > > > > --- libtool-2.4/tests/cmdline_wrap.at.orig 2010-05-20 23:18:41.000000000 +0200
> > > > > +++ libtool-2.4/tests/cmdline_wrap.at 2011-02-06 11:08:29.000000000 +0100
> > > > > @@ -30,6 +30,9 @@
> >
> > > > > +# Setting low max_cmd_len on "hurd-386" does not make sense"
> > > > > +AT_CHECK([if test "$host_os" == "gnu" ; then exit 77; fi])
> > > >
> > > > Please use = not ==, the former is Posix test, the latter bash-specific.
> > >
> > > OK, do you want me to send an update?
> >
> > I can't speak for Kurt, I only help maintain upstream. ;-)
>
> Updated!
>
> > > > Does this test actually fail on hurd-i386, or is this just because you
> > > > consider it wasted time? In the former case, please send the patch
> > > > upstream but also show a cut-n-paste of the failing test log.
> > >
> > > Yes, this test fails without the patch. I'll build an unpatched version
> > > and submit the failing test log.
I was wrong about the cmdline_wrap.at build failure, supplied in
disable_cmdline_wrap4hurd.patch. Without the patch libtool built fine:
The test 112 (and 106) failures was fixed by the updated
link_all_deplibs.patch. Sorry about that, but building (and especially
testing) libtool takes a very long time for each patch change in my qemu
VM.
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Fri, 08 Apr 2011 16:17:05 GMT)
Full text and
rfc822 format available.
Message #23 received at 8441 <at> debbugs.gnu.org (full text, mbox):
On Fri, Apr 08, 2011 at 04:37:48PM +0200, Svante Signell wrote:
>
> I was wrong about the cmdline_wrap.at build failure, supplied in
> disable_cmdline_wrap4hurd.patch. Without the patch libtool built fine:
> The test 112 (and 106) failures was fixed by the updated
> link_all_deplibs.patch. Sorry about that, but building (and especially
> testing) libtool takes a very long time for each patch change in my qemu
> VM.
That test basicly reruns all, or most, of the other tests again,
so if any of those failed, that test will also fail.
Kurt
Information forwarded
to
owner <at> debbugs.gnu.org, bug-libtool <at> gnu.org:
bug#8441; Package
libtool.
(Sun, 10 Apr 2011 08:21:02 GMT)
Full text and
rfc822 format available.
Message #26 received at 8441 <at> debbugs.gnu.org (full text, mbox):
* Svante Signell wrote on Fri, Apr 08, 2011 at 09:01:56AM CEST:
> # shlibpath_overrides_runpath is set to 'unknown' in libtool.m4
> # and not defined under $host_os ="gnu"
> # This patch make the tests/*demo* run.
> --- libtool-2.4/libltdt/m4/libtool.m4.orig 2011-02-03 21:33:56.000000000 +0100
> +++ libtool-2.4/libltdl/m4/libtool.m4 2011-02-03 21:43:46.000000000 +0100
> @@ -2325,6 +2325,7 @@
> library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}'
> soname_spec='${libname}${release}${shared_ext}$major'
> shlibpath_var=LD_LIBRARY_PATH
> + shlibpath_overrides_runpath=no
> hardcode_into_libs=yes
> ;;
Thank you. This should let the low-cmdline test pass as well, so it
need not be disabled any more (except for non coffee drinking purposes
at least ;-)
I'm pushing the patch below in your name and adding you to THANKS.
The '(tiny change)' annotation is just to denote that you haven't
exchanged copyright papers with the FSF yet.
Cheers,
Ralf
2011-04-10 Svante Signell <...> (tiny change)
Set shlibpath_overrides_runpath for the Hurd.
* libltdl/m4/libtool.m4 (_LT_SYS_DYNAMIC_LINKER) [gnu]
<shlibpath_overrides_runpath>: Set to no.
* THANKS: Update.
diff --git a/libltdl/m4/libtool.m4 b/libltdl/m4/libtool.m4
index 5cc027b..2ed41b7 100644
--- a/libltdl/m4/libtool.m4
+++ b/libltdl/m4/libtool.m4
@@ -2519,6 +2519,7 @@ gnu*)
library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}'
soname_spec='${libname}${release}${shared_ext}$major'
shlibpath_var=LD_LIBRARY_PATH
+ shlibpath_overrides_runpath=no
hardcode_into_libs=yes
;;
Merged 8441 8443 8444.
Request was from
Glenn Morris <rgm <at> gnu.org>
to
control <at> debbugs.gnu.org.
(Sat, 27 Aug 2011 01:30:02 GMT)
Full text and
rfc822 format available.
This bug report was last modified 13 years and 306 days ago.
Previous Next
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.