sreadahead assert failure: *** buffer overflow detected ***: /sbin/sreadahead terminated

Bug #452782 reported by A York
56
This bug affects 9 people
Affects Status Importance Assigned to Milestone
sreadahead (Ubuntu)
Invalid
Medium
Unassigned

Bug Description

Binary package hint: sreadahead

Description: Ubuntu karmic (development branch)
Release: 9.10
Last updated 2009-10-15

sreadahead 1.0-5

Expected desktop to appear without a crash symbol.

Crash symbol appeared, leading to this report.

ProblemType: Crash
Architecture: i386
AssertionMessage: *** buffer overflow detected ***: /sbin/sreadahead terminated
Date: Thu Oct 15 16:58:51 2009
DistroRelease: Ubuntu 9.10
ExecutablePath: /sbin/sreadahead
Package: sreadahead 1.0-5
ProcCmdline: /sbin/sreadahead -t 0
ProcEnviron: PATH=(custom, no user)
ProcVersionSignature: Ubuntu 2.6.31-14.47-generic
Signal: 6
SourcePackage: sreadahead
StacktraceTop:
 __kernel_vsyscall ()
 __pause_nocancel () from /lib/tls/i686/cmov/libpthread.so.0
 ?? ()
 __libc_start_main () from /lib/tls/i686/cmov/libc.so.6
 ?? ()
Title: sreadahead assert failure: *** buffer overflow detected ***: /sbin/sreadahead terminated
Uname: Linux 2.6.31-14-generic i686
UserGroups:

Revision history for this message
A York (dropax) wrote :
Revision history for this message
Apport retracing service (apport) wrote : Stacktrace.txt (retraced)

StacktraceTop:__kernel_vsyscall ()
__pause_nocancel () from /lib/tls/i686/cmov/libpthread.so.0
main (argc=3, argv=0xbf8c82c4) at sreadahead.c:703

Revision history for this message
Apport retracing service (apport) wrote : ThreadStacktrace.txt (retraced)
Changed in sreadahead (Ubuntu):
importance: Undecided → Medium
tags: removed: need-i386-retrace
A York (dropax)
visibility: private → public
Revision history for this message
A York (dropax) wrote :

I read the man page and the .conf file in /etc/init and wondered if stopping the profiling earlier would help.

Varied the run time allowed for profiling by changing the "-t 0" in /etc/init/sreadahead.conf and found that for times under 60 seconds it completed and created a /var/lib/sreadahead/pack file.

The size of the pack file for various run times:

-t 10 76208
-t 20 111760
-t 30 159456
-t 40 231440
-t 50 290752
-t 60 pack not created, crashed as described in this bug.

So it looks like when profiling, it doesn't terminate gracefully when the pack file is full, or there's a file on my system between 50 and 60 seconds with a path length > 176, or some other sort of corrupting characteristic, or ?

I contemplated the debug (-d) option, but reading the init documentation and discussion of forking left me wondering just where the sysout would get written. If debug output from a crashing run would help, I'll need help configuring where sysout goes, and/or where to look for it afterwards.

Revision history for this message
A York (dropax) wrote :

Today no longer repeatable. Emptied recycle bin, nothing notable in there except one 600mb .iso file, . Does gnome iterate through the recycle bin when initializing the desktop, and sreadahead was watching every file read?

Revision history for this message
dino99 (9d9) wrote :

This version is outdated and no more supported

Changed in sreadahead (Ubuntu):
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.