[Bug 1127700] New: bash patch bash50-001/bash50-002 seems to break env expansion
http://bugzilla.suse.com/show_bug.cgi?id=1127700 Bug ID: 1127700 Summary: bash patch bash50-001/bash50-002 seems to break env expansion Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening@forge.provo.novell.com Reporter: dmueller@suse.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- Tony Jones made me aware that using build(1) on tumbleweed is no longer possible for his userid. It is not really clear whats special about his user but we digged this far: build(1) at some point runs this function: recipe_setup_spec() { TOPDIR=`chroot $BUILD_ROOT su -c "rpm --eval '%_topdir'" - $BUILD_USER` if test -z "$TOPDIR"; then cleanup_and_exit 1 "Error: TOPDIR empty" fi test "$DO_INIT_TOPDIR" = false || rm -rf "$BUILD_ROOT$TOPDIR" for i in BUILD RPMS/`uname -m` RPMS/i386 RPMS/noarch SOURCES SPECS SRPMS BUILDROOT OTHER ; do mkdir -p $BUILD_ROOT$TOPDIR/$i done chown -R "$ABUILD_UID:$ABUILD_GID" "$BUILD_ROOT$TOPDIR" mkdir -p $BUILD_ROOT$TOPDIR/SOURCES echo ":: MYSRCDIR: $MYSRCDIR" ls -ld "$MYSRCDIR"/* cp -p "$MYSRCDIR"/* $BUILD_ROOT$TOPDIR/SOURCES/ } what we observed is that cp -p "$MYSRCDIR"/* $BUILD_ROOT/ .. expands to cp -p <empty> $BUILD_ROOT leading to an error message. we have validated that $MYSRCDIR is actually properly set (as can be seen by the echo before". with other words ls -ld "$MYSRCDIR"/* expands to *empty*, while ls -ld $MYSRCDIR/* properly expands. MYSRCDIR is without spaces or other weird characters (the only non-normal thing in there is a literal '.'). We have validated that building bash without bash50-001/bash50-002 patches fixes this problem. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1127700
Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c1
Dr. Werner Fink
We have validated that building bash without bash50-001/bash50-002 patches fixes this problem.
Compare with bug#1125818 which value does MYSRCDIR have in this function recipe_setup_spec() as I've to have a reproducer which I don have now. Simply to use the bashbug script to be able to inform upstream here -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c2
--- Comment #2 from Dr. Werner Fink
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c3
--- Comment #3 from Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c4
Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c5
--- Comment #5 from Dr. Werner Fink
your reporducer script, even if adjusted, does not reproduce.
I have a full log with env(1) output, any idea how to reduce it to the testcase?
The question is how the affected script does read the variable MYSRCDIR ... by environment or any other method like sourcing an other text file. It would be very helpful to see the environment of the script performing the shell function recipe_setup_spec() -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c6
--- Comment #6 from Dr. Werner Fink
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c7
Dirk Mueller
http://bugzilla.suse.com/show_bug.cgi?id=1127700
Dr. Werner Fink
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c8
--- Comment #8 from Swamp Workflow Management
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c9
--- Comment #9 from Swamp Workflow Management
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c10
--- Comment #10 from Swamp Workflow Management
http://bugzilla.suse.com/show_bug.cgi?id=1127700
http://bugzilla.suse.com/show_bug.cgi?id=1127700#c11
Dr. Werner Fink
participants (1)
-
bugzilla_noreply@novell.com