Closed Bug 975695 Opened 11 years ago Closed 11 years ago

"Assertion failure: !NS_HasPendingEvents(currentThread)" with many SharedWorkers

Categories

(Core :: DOM: Workers, defect)

Product:
Core
Component:
DOM: Workers
Platform:
x86_64
macOS
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

()

Status:
VERIFIED FIXED
Milestone:
mozilla30
Tracking Flags:
Tracking Status
firefox28 --- unaffected
firefox29 + verified
firefox30 + verified
firefox-esr24 --- unaffected
b2g18 --- unaffected
b2g-v1.1hd --- unaffected
b2g-v1.2 --- unaffected
b2g-v1.3 --- unaffected
b2g-v1.3T --- unaffected
b2g-v1.4 --- fixed

People

(Reporter: jruderman, Assigned: bent.mozilla)

References

Details

(4 keywords)

Attachments

(3 files)

testcase
(deleted), text/html
Details
stack
(deleted), text/plain
Details
Fix, v1
(deleted), patch
khuey
: review+
Sylvestre
: approval-mozilla-aurora+
abillings
: sec-approval+
Details | Diff | Splinter Review
Attached file testcase (deleted) —
1. Load the testcase in a Firefox debug build 2. Immediately quit Firefox Assertion failure: !NS_HasPendingEvents(currentThread), at dom/workers/WorkerPrivate.cpp:4499
Attached file stack (deleted) —
This is bad. We're violating the threading invariants here by calling ScheduleDeletion on the main thread. I don't understand why those assertions aren't firing though ...
Group: core-security
Keywords: sec-high
Attached patch Fix, v1 (deleted) — — Splinter Review
Attachment #8385628 - Flags: review?(khuey)
Comment on attachment 8385628 [details] [diff] [review] Fix, v1 Review of attachment 8385628 [details] [diff] [review]: ----------------------------------------------------------------- ::: dom/workers/WorkerPrivate.cpp @@ +4481,4 @@ > MOZ_ASSERT(!mCancelAllPendingRunnables); > mCancelAllPendingRunnables = true; > > + if (WorkerNeverRan == aRanOrNot) { Yoda you are not.
Attachment #8385628 - Flags: review?(khuey) → review+
Comment on attachment 8385628 [details] [diff] [review] Fix, v1 [Security approval request comment] How easily could an exploit be constructed based on the patch? Probably easily, it allows us to spin the event loop during GC. Do comments in the patch, the check-in comment, or tests included in the patch paint a bulls-eye on the security problem? Yes. Which older supported branches are affected by this flaw? Just aurora. If not all supported branches, which bug introduced the flaw? Bug 914762 Do you have backports for the affected branches? If not, how different, hard to create, and risky will they be? Yes, identical. How likely is this patch to cause regressions; how much testing does it need? Unlikely to cause regressions.
Attachment #8385628 - Flags: sec-approval?
Comment on attachment 8385628 [details] [diff] [review] Fix, v1 sec-approval+ for trunk. Once it is in, nominate for Aurora and I'll approve it there.
Attachment #8385628 - Flags: sec-approval? → sec-approval+
landed on central - https://hg.mozilla.org/mozilla-central/rev/3fd013ca2273
Status: NEW → RESOLVED
Closed: 11 years ago
status-firefox30: affected → fixed
Flags: in-testsuite?
Resolution: --- → FIXED
Target Milestone: --- → mozilla30
Comment on attachment 8385628 [details] [diff] [review] Fix, v1 [Approval Request Comment] Bug caused by (feature/regressing bug #): 914762 User impact if declined: Untold s-s badness Testing completed (on m-c, etc.): m-c Risk to taking this patch (and alternatives if risky): Low risk String or IDL/UUID changes made by this patch: None
Attachment #8385628 - Flags: approval-mozilla-aurora?
Attachment #8385628 - Flags: approval-mozilla-aurora? → approval-mozilla-aurora+
Confirmed assertion on Fx29, 2014-02-10. Verified fixed on Fx29, Fx30, 2014-04-21.
Status: RESOLVED → VERIFIED
status-firefox29: fixed → verified
status-firefox30: fixed → verified
Group: core-security
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: