github-starred/firejail

Fork 0

mirror of https://github.com/netblue30/firejail.git synced 2026-05-15 14:16:14 -06:00

[GH-ISSUE #1096] Firefox 51.0 segfaults with --private-dev option #751

New issue

Closed

opened 2026-05-05 06:34:19 -06:00 by gitea-mirror · 6 comments

gitea-mirror commented

2026-05-05 06:34:19 -06:00

Owner

Originally created by @acrobat1 on GitHub (Feb 10, 2017).
Original GitHub issue: https://github.com/netblue30/firejail/issues/1096

Firefox 51.0 segfaults when is launched and when it's closed if using the --private-dev option

firefox[4203]: segfault at 0 ip 00007f5281ee721a sp 00007ffc6a3e1670 error 6 in libxul.so

To be precise, it does not crash but it throws this error multiple times (using --tracelog) when it starts and when it exits - which I've never seen up to this 51.0 version (64bit).

I've tested with --noprofile and it seems consistent in Firejail 0.9.44.8.

Originally created by @acrobat1 on GitHub (Feb 10, 2017). Original GitHub issue: https://github.com/netblue30/firejail/issues/1096 Firefox 51.0 segfaults when is launched and when it's closed if using the **--private-dev** option `firefox[4203]: segfault at 0 ip 00007f5281ee721a sp 00007ffc6a3e1670 error 6 in libxul.so` To be precise, it does not crash but it throws this error multiple times (using --tracelog) when it starts and when it exits - which I've never seen up to this 51.0 version (64bit). I've tested with **--noprofile** and it seems consistent in Firejail 0.9.44.8.

gitea-mirror

2026-05-05 06:34:19 -06:00

closed this issue
added the
information_old
label

gitea-mirror commented

2026-05-05 06:34:24 -06:00

Author

Owner

@netblue30 commented on GitHub (Feb 11, 2017):

I have it running fine here (regular AMD driver from Debian stable). Maybe is because of --tracelog? Build a custom profile like this:

$ mkdir -p ~/.config/firejail
$ echo "ignore private-dev" > ~/.config/firejail/firefox.profile
$ echo "include /etc/firejail/firefox.profile" >> ~/.config/firejail/firefox.profile

I'll keep an eye on it if anybody is seeing the problem, in the worst case we remove private-dev from the profile.

@netblue30 commented on GitHub (Feb 11, 2017): I have it running fine here (regular AMD driver from Debian stable). Maybe is because of --tracelog? Build a custom profile like this: ````` $ mkdir -p ~/.config/firejail $ echo "ignore private-dev" > ~/.config/firejail/firefox.profile $ echo "include /etc/firejail/firefox.profile" >> ~/.config/firejail/firefox.profile ````` I'll keep an eye on it if anybody is seeing the problem, in the worst case we remove private-dev from the profile.

gitea-mirror commented

2026-05-05 06:34:26 -06:00

Author

Owner

@acrobat1 commented on GitHub (Feb 21, 2017):

Hi, sorry for my late reply : in the meantime I've tried with a "vanilla" Firefox 51 installation (=no modifications whatsoever and clean profile with no extensions) and this very short command line firejail --noprofile --private-dev and it segfaults consistently , then if I remove the private-dev argument it doesn't.

I too was thinking about --tracelog, but as far as I can see it really is triggered just by --private-dev , and it only started with Firefox 51 - I've never seen it before.

Would this test above be comparable to what you were suggesting ?

Do you have any ideas about to why Firefox would do this from version 51.0 onwards but not before ?

BTW, I too am on Debian stable with no weird drivers or custom kernels, actually it's a pretty standard installation : at the moment, I really can't think of any reason why I should experience this if no one else here is .

@acrobat1 commented on GitHub (Feb 21, 2017): Hi, sorry for my late reply : in the meantime I've tried with a "vanilla" Firefox 51 installation (=no modifications whatsoever and clean profile with no extensions) and this very short command line **firejail --noprofile --private-dev** and it segfaults consistently , then if I remove the **private-dev** argument it doesn't. I too was thinking about --tracelog, but as far as I can see it really is triggered just by **--private-dev** , and it only started with Firefox 51 - I've never seen it before. Would this test above be comparable to what you were suggesting ? Do you have any ideas about to why Firefox would do this from version 51.0 onwards but not before ? BTW, I too am on Debian stable with no weird drivers or custom kernels, actually it's a pretty standard installation : at the moment, I really can't think of any reason why I should experience this if no one else here is .

gitea-mirror commented

2026-05-05 06:34:29 -06:00

Author

Owner

@acrobat1 commented on GitHub (Mar 16, 2017):

It just occurred to me that in firejail.config I've set chroot-desktop no : could this be the reason of this issue?

@acrobat1 commented on GitHub (Mar 16, 2017): It just occurred to me that in firejail.config I've set **chroot-desktop no** : could this be the reason of this issue?

gitea-mirror commented

2026-05-05 06:34:31 -06:00

Author

Owner

@netblue30 commented on GitHub (Mar 21, 2017):

That chroot-desktop should make a difference only if you are using --chroot on the command line.

@netblue30 commented on GitHub (Mar 21, 2017): That chroot-desktop should make a difference only if you are using --chroot on the command line.

gitea-mirror commented

2026-05-05 06:34:33 -06:00

Author

Owner

@acrobat1 commented on GitHub (Mar 28, 2017):

Well, I'm not using --chroot on the command line .. so , until someone can figure this out, the only thing for me left to do is to run firefox without --private-dev option (or put up with that segfault message, which doesn't cause any visible issue) .

BTW, what we exactly lose by dropping --private-dev ? Is that a major drawback security-wise ?

@acrobat1 commented on GitHub (Mar 28, 2017): Well, I'm not using --chroot on the command line .. so , until someone can figure this out, the only thing for me left to do is to run firefox without **--private-dev** option (or put up with that segfault message, which doesn't cause any visible issue) . BTW, what we exactly lose by dropping **--private-dev** ? Is that a major drawback security-wise ?

gitea-mirror commented

2026-05-05 06:34:35 -06:00

Author

Owner

@netblue30 commented on GitHub (Mar 28, 2017):

Not too much, it is a secondary feature. Nothing important like seccomp and whitelisting home directory. I wouldn't worry about it.

@netblue30 commented on GitHub (Mar 28, 2017): Not too much, it is a secondary feature. Nothing important like seccomp and whitelisting home directory. I wouldn't worry about it.

gitea-mirror referenced this issue

2026-05-05 10:07:02 -06:00

[PR #751] [MERGED] Fix spelling error #3739

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/firejail#751

No description provided.

Rows
Columns