FS#61465 - [pacman] Assertion upon interrupt.

Attached to Project: Pacman
Opened by Chris Billington (chrisjbillington) - Saturday, 19 January 2019, 17:18 GMT
Last edited by Andrew Gregory (andrewgregory) - Sunday, 20 January 2019, 02:10 GMT
Task Type Bug Report
Category General
Status Closed
Assigned To No-one
Architecture All
Severity Low
Priority Normal
Reported Version 5.1.2
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Summary and Info:

I encountered an assertion in pacman v5.1.2.

I ran the command 'pacman -Fs ca'. Normally this command outputs:


$ pacman -Fs ca
extra/easy-rsa 3.0.5-1
etc/easy-rsa/x509-types/ca
extra/xkeyboard-config 2.25-1
usr/share/X11/xkb/symbols/ca
usr/share/X11/xkb/symbols/sun_vndr/ca
community/espeak 1:1.48.04-2
usr/share/espeak-data/voices/europe/ca
community/nltk-data 3.3-1
usr/share/nltk_data/corpora/swadesh/ca
community/python-langdetect 1.0.7-3
usr/lib/python3.7/site-packages/langdetect/profiles/ca
community/python2-langdetect 1.0.7-3
usr/lib/python2.7/site-packages/langdetect/profiles/ca

However, I hit ctrl-C after the "extra" results printed, but before the "community" results printed, and got an assertion:

$ pacman -Fs ca
extra/easy-rsa 3.0.5-1
etc/easy-rsa/x509-types/ca
extra/xkeyboard-config 2.25-1
usr/share/X11/xkb/symbols/ca
usr/share/X11/xkb/symbols/sun_vndr/ca
^Cpacman: malloc.c:2385: sysmalloc: Assertion `(old_top == initial_top (av) && old_size == 0) || ((unsigned long) (old_size) >= MINSIZE && prev_inuse (old_top) && ((unsigned long) old_end & (pagesize - 1)) == 0)' failed.
Aborted (core dumped)

The printing pauses briefly in between printing the "extra", and this is when I hit ctrl-C. I am able to reproduce the same result four times (out of ~100) by repeatedly running the command and hitting ctrl-C, but it is obviously timing-sensitive. It also occurred once searching for another filename ('test')

I also got the following after hitting ctrl-C a tiny bit later after pacman had printed all the results but before it exited:

$ pacman -Fs ca
extra/easy-rsa 3.0.5-1
etc/easy-rsa/x509-types/ca
extra/xkeyboard-config 2.25-1
usr/share/X11/xkb/symbols/ca
usr/share/X11/xkb/symbols/sun_vndr/ca
community/espeak 1:1.48.04-2
usr/share/espeak-data/voices/europe/ca
community/nltk-data 3.3-1
usr/share/nltk_data/corpora/swadesh/ca
community/python-langdetect 1.0.7-3
usr/lib/python3.7/site-packages/langdetect/profiles/ca
community/python2-langdetect 1.0.7-3
usr/lib/python2.7/site-packages/langdetect/profiles/ca
^Cmalloc(): memory corruption
Aborted (core dumped)



Steps to Reproduce:

* while True:
* run 'pacman -Fs ca'
* hit ctrl-C at various points in time.
* cross your fingers/press your thumbs
This task depends upon

Closed by  Andrew Gregory (andrewgregory)
Sunday, 20 January 2019, 02:10 GMT
Reason for closing:  Not a bug
Comment by Andrew Gregory (andrewgregory) - Sunday, 20 January 2019, 00:27 GMT
I cannot reproduce this, nor can I see how pacman could possibly be causing this. Can you provide a stack trace?
Comment by Chris Billington (chrisjbillington) - Sunday, 20 January 2019, 01:31 GMT
Upon getting a backtrace from the core dump, I found that the crash was in a function from a library I had LD_PRELOADed (I use a hack to make stderr red by intercepting write calls to stderr).

It's almost certainly nothing to do with pacman, sorry for the noise. I'll clear my LD_PRELOAD before reporting bugs next time!

Loading...