Historical bug tracker for the Pacman package manager.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues
This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
FS#19011 - pacman segfault (2.6.33 / Testing)
Attached to Project:
Pacman
Opened by Michael C (afterlife) - Friday, 09 April 2010, 16:28 GMT
Last edited by Dan McGee (toofishes) - Saturday, 10 April 2010, 15:42 GMT
Opened by Michael C (afterlife) - Friday, 09 April 2010, 16:28 GMT
Last edited by Dan McGee (toofishes) - Saturday, 10 April 2010, 15:42 GMT
|
DetailsSummary and Info:
pacman -Syu fails to update (Segmentation fault) While updating mirrors (pacman -Sy) and then installing all updates (pacman -Su) pacman will fail after downloading all updates and starting upgrade as such: --8<---------------------------------------------------------- checking package integrity... error: segmentation fault Internal pacman error: Segmentation fault. Please submit a full bug report with --debug if appropriate. ------------------------------------------------------->8----- running pacman -Syu --debug produces the following output (truncated) --8<---------------------------------------------------------- debug: loading target '/var/cache/pacman/pkg/openoffice-base-3.2.0-3-i686.pkg.tar.xz' debug: openoffice-base: syntax error in description file line 5 error: segmentation fault Internal pacman error: Segmentation fault. ------------------------------------------------------->8----- This will happen even after cleaning the cache and trying multiple times. Misc info: $uname -a Linux chiaki 2.6.33-ARCH #1 SMP PREEMPT Mon Mar 15 18:36:58 UTC 2010 i686 Intel(R) Pentium(R) 4 CPU 3.06GHz GenuineIntel GNU/Linux Repos open: testing, core, extra, comm and archlinuxfr I am aware I am using testing and as such I'm not making a fuss just wanted to contribute incase anyone else seems to have this problem. Any other info please feel free to ask for. Kind regards, Mike. Steps to Reproduce: pacman -Syu with openoffice installed and running the testing repo. |
This task depends upon
Closed by Dan McGee (toofishes)
Saturday, 10 April 2010, 15:42 GMT
Reason for closing: Deferred
Additional comments about closing: Cannot repro, if it happens in the future we'll see what we can do.
Saturday, 10 April 2010, 15:42 GMT
Reason for closing: Deferred
Additional comments about closing: Cannot repro, if it happens in the future we'll see what we can do.
Can you try to decompress the file manually with tar and bsdtar ?
$ mkdir /tmp/ooo && cd /tmp/ooo
$ tar xf /var/cache/pacman/pkg/openoffice-base-3.2.0-3-x86_64.pkg.tar.xz
$ bsdtar xf /var/cache/pacman/pkg/openoffice-base-3.2.0-3-x86_64.pkg.tar.xz
You could just test the file using tf instead of xf.
And you can just test the xz part with xz -t /var/cache/pacman/pkg/openoffice-base-3.2.0-3-x86_64.pkg.tar.xz
as a result this bug has 'gone' for me :/ I am at a loss to explain it other than 'between sync' oddity...
Best close this one, just a bermuda triangle bug :)
Regards,
Mike.
debug: loading target '/var/cache/pacman/pkg/openoffice-base-3.2.0-3-i686.pkg.tar.xz'
debug: openoffice-base: syntax error in description file line 5
error: segmentation fault
Internal pacman error: Segmentation fault.
We should probably look into why that got triggered and see what we can figure out.
debug: openoffice-base: syntax error in description file line 5
That's just because pacman does not handle pkgbase when parsing PKGINFO.
And normally the message that is printed after that one should be "sorting package filelist"
And that's all in pkg_load in be_package.c . So I thought of a memory problem.. But then libarchive should maybe return an error.
Maybe a bug in libarchive.
Well that's just wild guess, and if it cannot be reproduced, I am afraid we won't get far.
Happy to provide any other info though.