Closed
Bug 734763
Opened 13 years ago
Closed 13 years ago
"Assertion failure: rt->gcMode != JSGC_MODE_GLOBAL" with certain GC prefs and forceShrinkingGC
Categories
(Core :: JavaScript Engine, defect)
Tracking
()
RESOLVED
FIXED
mozilla13
People
(Reporter: jruderman, Assigned: billm)
References
Details
(Keywords: assertion, regression, testcase)
Attachments
(2 files)
|
50 bytes,
text/html
|
Details | |
|
904 bytes,
patch
|
gwagner
:
review+
|
Details | Diff | Splinter Review |
1. Install the new version of
https://www.squarefree.com/extensions/domFuzzLite3.xpi
2. Set
user_pref("javascript.options.mem.gc_incremental", false);
user_pref("javascript.options.mem.gc_per_compartment", false);
3. Load the testcase.
Result:
Assertion failure: rt->gcMode != JSGC_MODE_GLOBAL, at jsgc.cpp:3600
Based on the stack trace, I suspect this is a regression from the following chunk of bug 730447's patch:
@ MaybeGC
+ if (comp->gcMallocAndFreeBytes >= comp->gcTriggerMallocAndFreeBytes) {
+ GCSlice(cx, comp, GC_NORMAL, gcreason::MAYBEGC);
+ return;
+ }
| Assignee | ||
Comment 1•13 years ago
|
||
This fixes two problems:
1. We're triggering a compartment GC when we're not supposed to.
2. We're triggering a GC when malloc bytes and trigger malloc bytes are both 0.
Updated•13 years ago
|
Attachment #605011 -
Flags: review?(anygregor) → review+
| Assignee | ||
Comment 2•13 years ago
|
||
Target Milestone: --- → mozilla13
Comment 3•13 years ago
|
||
Status: ASSIGNED → RESOLVED
Closed: 13 years ago
Resolution: --- → FIXED
You need to log in
before you can comment on or make changes to this bug.
Description
•