Bug#459101: marked as done (gnat-4.1: FTBFS with dash: CANNOT FIND debian/*FFI*)
Your message dated Sun, 13 Jan 2008 15:36:51 +0100
with message-id <18314.8707.471997.427301@gargle.gargle.HOWL>
and subject line Bug#459101: gnat-4.1: FTBFS with dash: CANNOT FIND debian/*FFI*
has caused the attached Bug report to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere. Please contact me immediately.)
Debian bug tracking system administrator
(administrator, Debian Bugs database)
--- Begin Message ---
Package: gnat-4.1
version: 4.1.2-17
User: debian-qa@lists.debian.org
Usertags: qa-ftbfs-dash-20080103 qa-ftbfs-dash
Hi,
During a rebuild of all packages in sid using /bin/dash as /bin/sh, your package failed to build.
Relevant part:
rm -f debian/substvars.local.tmp
( \
echo 'gcc:Version=4.1.2-2'; \
echo 'gcc:EpochVersion=1:4.1.2-17'; \
echo 'gcc:SoftVersion=4.1.2-2'; \
echo 'gpc:Version='; \
echo 'gdc:Version='; \
echo 'gcj:Version=4.1.2-17'; \
echo 'gcj:SoftVersion=4.1.2-2'; \
echo 'gnat:Version=4.1.2-17'; \
echo 'binutils:Version=2.17cvs20070426'; \
echo 'dep:libgcc=libgcc1 (>= 1:4.1.2-17)'; \
echo 'dep:libgccbiarch=lib64gcc1 (>= 1:4.1.2-17)'; \
echo 'dep:libcdev=libc6-dev (>= 2.5)'; \
echo 'dep:libcbiarch=libc6-amd64 (>= 2.5)'; \
echo 'dep:libcbiarchdev=libc6-dev-amd64'; \
echo 'dep:libunwinddev='; \
echo 'dep:libcxxbiarch=lib64stdc++6 (>= ${gcc:Version})'; \
echo 'dep:libcxxbiarchdbg=lib64stdc++6-4.1-dbg'; \
echo 'dep:libgnat=libgnat-4.1 (= 4.1.2-17)'; \
) > debian/substvars.local.tmp
( \
echo 'gcc:multilib=gcc-4.1-multilib'; \
echo 'gxx:multilib=g++-4.1-multilib'; \
echo 'gobjc:multilib=gobjc-4.1-multilib'; \
echo 'gobjcxx:multilib=gobjc++-4.1-multilib'; \
echo 'gfortran:multilib=gfortran-4.1-multilib'; \
) >> debian/substvars.local.tmp
echo 'dep:libfortranbiarch=lib64gfortran1 (>= ${gcc:Version})' \
>> debian/substvars.local.tmp
echo 'dep:gcj=gcc-4.1 (>= 4.1.2-2)' \
>> debian/substvars.local.tmp
[ -e debian/substvars.local ] \
&& cmp -s debian/substvars.local debian/substvars.local.tmp \
&& rm -f debian/substvars.local.tmp && exit 0; \
mv debian/substvars.local.tmp debian/substvars.local; \
touch stamps/03-control-stamp
fs=`echo debian/*-BV* debian/*GCJ* debian/*CXX* debian/*LC* debian/*FFI*|sort -u`; \
for f in $fs; do \
[ -f $f ] || echo "CANNOT FIND $f"; \
[ -f $f ] || continue; \
f2=$(echo $f | sed 's/BV/4.1/;s/PV//;s/CXX/6/;s/LGCJ/7-1/;s/GCJ/7/;s/LC/1/;s/FFI/4/'); \
sed -e 's/@BV@/4.1/g' \
-e 's/@PV@//g' \
-e 's/@CXX@/6/g' \
-e 's/@LGCJ@/7-1/g' \
-e 's/@GCJ@/7/g' \
-e 's/@LC@/1/g' \
-e 's/@FFI@/4/g' \
-e 's/@browser_dirs@/firefox iceape iceweasel mozilla midbrowser xulrunner/g' \
-e 's/@GFDL@/#/g' \
$f > $f2; \
touch -r $f $f2; \
done
CANNOT FIND debian/*FFI*
make[1]: Leaving directory `/build/user/gnat-4.1-4.1.2'
/usr/bin/make -f debian/rules2 stamps/04-configure-stamp
make[1]: Entering directory `/build/user/gnat-4.1-4.1.2'
The full build log is available from:
http://people.debian.org/~lucas/logs/2008/01/03.dash/
Rumors say that Ubuntu is using dash as /bin/sh on their buildds, so a patch
might be available. Check <http://packages.qa.debian.org/gnat-4.1>, or
directly on launchpad: <https://launchpad.net/ubuntu/+source/gnat-4.1>.
A list of current common problems and possible solutions is available at
http://wiki.debian.org/qa.debian.org/FTBFS . You're welcome to contribute!
--
| Lucas Nussbaum
| lucas@lucas-nussbaum.net http://www.lucas-nussbaum.net/ |
| jabber: lucas@nussbaum.fr GPG: 1024D/023B3F4F |
--- End Message ---
--- Begin Message ---
There's nothing to fix; the relevant part reads:
[ -f $$f ] || echo "CANNOT FIND $$f"; \
Lucas, please feel free to reopen the report providing the really
"relevant" part of the build log ;)
Matthias
Ludovic Brenta writes:
> The problem is that debian/rules.conf assumes there are lib*FFI* files
> that need to produce lib*ffi4* files, but that is not true. Instead,
> the lib*ffi4* files are already there and are not generated.
>
> I don't know precisely what triggers this bug, but this is somewhere
> in the generation of debian/control from debian/control.m4.
>
> One approach to fix it is to remove debian/*FFI* from the first line
> of the versioned-files rule in debian/rules.conf. Another is to
> replace the existing lib*ffi4* files with more generic lib*FFI* files.
>
> Matthias, what are your intentions?
--- End Message ---
Reply to: