JS tests failing in gjs on i386

Bug #537903 reported by Micah Gersten
46
This bug affects 8 people
Affects Status Importance Assigned to Milestone
gjs
Fix Released
Medium
gjs (Ubuntu)
Fix Released
High
Chris Coulson
Lucid
Fix Released
High
Chris Coulson

Bug Description

Binary package hint: gjs

For some reason, this works right on amd64. It also works with GJS_DISABLE_JIT.
RAOF generated some debug information that I'm attaching. Apparently the finalizer isn't being called when JIT is enabled.

Related branches

Revision history for this message
Micah Gersten (micahg) wrote :
Revision history for this message
Micah Gersten (micahg) wrote :
Revision history for this message
Micah Gersten (micahg) wrote :

Assigning to myself at the moment and importance of Low as we can disable JIT if we need to.

Changed in gjs (Ubuntu):
assignee: nobody → Micah Gersten (micahg)
importance: Undecided → Low
status: New → Triaged
Revision history for this message
Chris Halse Rogers (raof) wrote :

The tests fail on armel, but for a different reason - it dies with a SIGILL in libmozjs and has a corrupted stack:

(gdb) run
Starting program: /home/chrishr/gjs-0.5/_build/.libs/lt-gjs-unit
[Thread debugging using libthread_db enabled]
/js/Mainloop:
Program received signal SIGILL, Illegal instruction.
0x402a3b70 in str_match (cx=0x23650, argc=1, vp=0xbefa9aa4) at jsstr.cpp:1458
1458 jsstr.cpp: No such file or directory.
 in jsstr.cpp
Current language: auto
The current source language is "auto; currently c++".
(gdb) bt full
#0 0x402a3b70 in str_match (cx=0x23650, argc=1, vp=0xbefa9aa4) at jsstr.cpp:1458
No locals.
#1 0xbefabf90 in ?? ()
No symbol table info available.
#2 0xbefabf90 in ?? ()
No symbol table info available.
Backtrace stopped: previous frame identical to this frame (corrupt stack?)

Changed in gjs (Ubuntu Lucid):
importance: Low → Medium
Changed in gjs (Ubuntu Lucid):
importance: Medium → High
Changed in gjs (Ubuntu Lucid):
assignee: Micah Gersten (micahg) → Chris Coulson (chrisccoulson)
status: Triaged → In Progress
Changed in gjs (Ubuntu Lucid):
status: In Progress → Fix Committed
Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package gjs - 0.5-1ubuntu2

---------------
gjs (0.5-1ubuntu2) lucid; urgency=low

  * debian/patches/01_disable_memcheck.patch:
    - Don't make leaks fatal, see b.g.o #616193. Fixes LP: #537903
  * debian/control:
    - Build-depend on quilt
  * debian/rules:
    - Include quilt.make and add patch/unpatch targets
  * Transition to xulrunner 1.9.2:
    - bump minimum xulrunner depends to 1.9.2
    - update debian/rules
 -- Chris Coulson <email address hidden> Mon, 19 Apr 2010 16:36:54 +0100

Changed in gjs (Ubuntu Lucid):
status: Fix Committed → Fix Released
Revision history for this message
Chris Coulson (chrisccoulson) wrote :

Note, I'll look at the armel issue separately. I haven't recreated that yet as I couldn't get access to an armel chroot today

summary: - JS tests failing in gjs on i386 and armel
+ JS tests failing in gjs on i386
Changed in gjs:
status: Unknown → New
Changed in gjs:
importance: Unknown → Medium
Changed in gjs:
status: New → Confirmed
Changed in gjs:
status: Confirmed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

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