davfs2 umount.davfs crashes when compiled with gcc 9

[Joseph Yasi]

davfs2 uses the fscanf format string "%a[0-9]". "%a" is part of C89, but not C99, and tells fscanf to do the string allocation.

The fscanf man page states:
Note that the a modifier is not available if the program is compiled with gcc -std=c99 or gcc -D_ISOC99_SOURCE (unless _GNU_SOURCE is also specified), in which case the a is interpreted as a specifier for floating-point numbers (see above).

Support for the m modifier was added to glibc starting with version 2.7, and new programs should use that modifier instead of a.

Switching the a to an m prevents the crash.

[aros (aros-redhat-bugs)]

Description of problem: while trying to unmount a davfs filesystem davfs crashes

Version-Release number of selected component (if applicable): davfs2-1.5.4-9.fc30.x86_64

How reproducible: always

systemd-coredump[1456]: Process 1454 (umount.davfs) of user 1000 dumped core.

                                           Stack trace of thread 1454:
                                           #0 0x00007f1ec9a425e5 __strlen_avx2 (libc.so.6)
                                           #1 0x00007f1ec9abe588 ne_concat (libneon.so.27)
                                           #2 0x000055b5dfc0c608 n/a (umount.davfs)
                                           #3 0x00007f1ec9909f43 __libc_start_main (libc.so.6)
                                           #4 0x000055b5dfc0ca3e n/a (umount.davfs)

Gentoo users have already found this bug and it could be solved by using -std=c89 compilation flag

https://forums.gentoo.org/viewtopic-p-8331344.html?sid=86512aac8febf8259dee6fcee5bf1345

[aros (aros-redhat-bugs)]

It's a dupe of bug 1710968 however I am curious why it hasn't yet been solved.

[aros (aros-redhat-bugs)]

Hopefully a barrage of emails will force the fedora developer responsible for this component to finally fix this error for good.

[aros (aros-redhat-bugs)]

The original bug report was filed on 2019-05-16 16:16:50 UTC, i.e. FIVE MONTHS AGO.

What is going on?

[aros (aros-redhat-bugs)]

This is yet another comment to attract a modicum of attention.

[aros (aros-redhat-bugs)]

Hopefully six emails are going to be enough to force someone do something.

[fschwarz (fschwarz-redhat-bugs)]

Artem: I'm sorry that nobody reacted on your bug report - though I can not help directly. As you noted this is likely the same issue as bug 1710968 you should close it as such. Keeping duplicate bug reports around does not help anyone. Actually the chance of getting a reaction is higher if you keep the communication in one place.

However there is something you do to get this fixed:
1. Try davfs2 1.5.5 (F31 though) which Filipe Rosset created as a response to bug 1736871.

2. The other reporter indicated that unmounting still does not work. If you experience the same issue this is how you can help:
   - file a *support* ticket upstream: http://savannah.nongnu.org/support/?group=davfs2
   - once upstream has confirmed the issue/has committed a fix please ask upstream to release a new version and get back to use so we can push a new version to Fedora.

Unfortunately Fedora maintainers often do not have the time to debug upstream bugs (though I can't speak for Filipe). I hope you can understand that most of us would like to do more but this is only a volunteer activity...

Thank you for helping Fedora :-)

[fschwarz (fschwarz-redhat-bugs)]

btw: please check https://bugzilla.redhat.com/show_bug.cgi?id=1710968#c4 if you are willing to try a test build.

[fschwarz (fschwarz-redhat-bugs)]

I talked to the upstream developer: A bugfix release is expected to be released within two weeks. I'll try to get this into Fedora shortly afterwards but I could use some testers (you'll be notified in this bug about a new update in updates-testing).

[updates (updates-redhat-bugs)]

FEDORA-2019-d2b5b244ee has been submitted as an update to Fedora 31. https://bodhi.fedoraproject.org/updates/FEDORA-2019-d2b5b244ee

[fschwarz (fschwarz-redhat-bugs)]

Please try the linked update from the updates-testing repo and give karma if this update fixes the bug for you.

[updates (updates-redhat-bugs)]

davfs2-1.5.6-1.fc31 has been pushed to the Fedora 31 testing repository. If problems still persist, please make note of it in this bug report.
See https://fedoraproject.org/wiki/QA:Updates_Testing for
instructions on how to install test updates.
You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2019-d2b5b244ee

[Joseph Yasi (joe-yasi)]

This crash happens in eoan with davfs2 1.5.5 which was built with gcc 9.

[Joseph Yasi (joe-yasi)]

Upstream bug: http://savannah.nongnu.org/bugs/index.php?56286

[Joseph Yasi (joe-yasi)]

Upstream fix: http://cvs.savannah.nongnu.org/viewvc/davfs2/davfs2/src/umount_davfs.c?r1=1.2.4.6&r2=1.2.4.7

This is fixed in upstream davfs2 1.5.6. A version bump will fix this package.

[Ubuntu Foundations Team Bug Bot (crichton)]

The attachment "Fix crash in umount due to fscanf format" seems to be a patch. If it isn't, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are a member of the ~ubuntu-reviewers, unsubscribe the team.

[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issues please contact him.]

[updates (updates-redhat-bugs)]

davfs2-1.5.6-1.fc31 has been pushed to the Fedora 31 stable repository. If problems still persist, please make note of it in this bug report.

[Joseph Yasi (joe-yasi)]

This bug causes umount to crash for davfs filesystem. I can't unmount them. The patch is simple and upstream. It looks like no one tested this after upgrading the compiler to gcc 9. Upgrading the package to davfs2 1.5.6 fixes this crash.

Is this going to make it for focal?

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Daniel von Obernitz (itzonban)]

Will there be a release of 1.5.6 for Ubuntu Focal soon? The fix is out there for almost six months now :(

[Woodrow Shen (woodrow-shen)]

Hi,

1.5.6-0.1 is available now on groovy, and I'm working on focal SRU and waiting for sponsoring.

[David Scherfgen (d-scherfgen)]

Any news here? It has been broken for so long time now.

[Woodrow Shen (woodrow-shen)]

I'd like to remove myself from assignee as this is still blocked by sponsoring queue, unless someone who has MOTU can help to proceed the SRU.