If you think a bug might affect users in the 57 release, please set the correct tracking and status flags for Release Management.

Intermittent ASAN test_bug732665.xul | nearNativeStackLimit callback threw: InternalError: too much recursion | uncaught exception - ReferenceError: nestedLimit is not defined

RESOLVED FIXED in Firefox 26

Status

()

Core
XPConnect
RESOLVED FIXED
4 years ago
4 years ago

People

(Reporter: RyanVM, Assigned: bholley)

Tracking

({intermittent-failure})

Trunk
mozilla26
x86_64
Linux
intermittent-failure
Points:
---

Firefox Tracking Flags

(firefox24 unaffected, firefox25 unaffected, firefox26 fixed)

Details

Attachments

(1 attachment)

(Reporter)

Description

4 years ago
https://tbpl.mozilla.org/php/getParsedLog.php?id=27584288&tree=Mozilla-Inbound

Ubuntu ASAN VM 12.04 x64 mozilla-inbound opt test mochitest-other on 2013-09-09 09:22:15 PDT for push 7cadc7e21f56
slave: tst-linux64-ec2-353

09:36:45     INFO -  12298 INFO TEST-START | chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul
09:36:45     INFO -  12299 INFO TEST-PASS | chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul | Chrome should be able to have at least 10 heavy frames more stack than content: 152, 140
09:36:45     INFO -  12300 ERROR TEST-UNEXPECTED-FAIL | chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul | nearNativeStackLimit callback threw: InternalError: too much recursion
09:36:45     INFO -  12301 ERROR TEST-UNEXPECTED-FAIL | chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul | uncaught exception - ReferenceError: nestedLimit is not defined at chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul:63
09:36:45     INFO -  JavaScript error: chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul, line 63: nestedLimit is not defined
09:36:45     INFO -  12302 INFO TEST-END | chrome://mochitests/content/chrome/js/xpconnect/tests/chrome/test_bug732665.xul | finished in 107ms
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Needinfo from bholley who wrote the test. Does this test make assumptions that would be invalid for a build that consumes a lot more stack space per frame? The test isn't crashing but failing on TMR so I assume that it's expecting that it can recurse more than it actually can with ASan builds.
Flags: needinfo?(bobbyholley+bmo)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
(In reply to Christian Holler (:decoder) from comment #10)
> Needinfo from bholley who wrote the test. Does this test make assumptions
> that would be invalid for a build that consumes a lot more stack space per
> frame? The test isn't crashing but failing on TMR so I assume that it's
> expecting that it can recurse more than it actually can with ASan builds.

Yes. I recently fiddled a bunch with the stack limits, and I didn't have ASAN builds handy to take measurements. I got some numbers from Jesse that seemed to corroborate my guess, so I went with it.

http://mxr.mozilla.org/mozilla-central/source/js/xpconnect/src/XPCJSRuntime.cpp#2973

The path forward here is for someone to collect ASAN numbers on windows and linux and bump kTrustedScriptBuffer appropriately for that case. We could also bump kStackQuota if that makes sense.

decoder, is this something you can do?
Flags: needinfo?(bobbyholley+bmo) → needinfo?(choller)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Created attachment 803290 [details] [diff] [review]
Leave ourselves some room to make the call into chrome. v1
Attachment #803290 - Flags: review?(luke)
Comment on attachment 803290 [details] [diff] [review]
Leave ourselves some room to make the call into chrome. v1

Decoder - assuming you were able to reproduce reliably enough, can you verify that this fixes the problem? Sorry for misdiagnosing it at first.
Attachment #803290 - Flags: feedback?(choller)

Updated

4 years ago
Attachment #803290 - Flags: review?(luke) → review+
Comment on attachment 803290 [details] [diff] [review]
Leave ourselves some room to make the call into chrome. v1

Ran this 20 times without failures, so that seems to work. Thanks for taking care of this :)
Attachment #803290 - Flags: feedback?(choller) → feedback+
https://hg.mozilla.org/integration/mozilla-inbound/rev/0be65c045f13
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
Comment hidden (Treeherder Robot)
https://hg.mozilla.org/mozilla-central/rev/0be65c045f13
Status: NEW → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla26

Updated

4 years ago
Flags: needinfo?(choller)
(Reporter)

Updated

4 years ago
Assignee: nobody → bobbyholley+bmo
status-firefox24: --- → unaffected
status-firefox25: --- → unaffected
status-firefox26: --- → fixed
You need to log in before you can comment on or make changes to this bug.