669975 - websocket ASSERTION: nsLoadGroup not thread-safe

Status: RESOLVED FIXED

(Core :: Networking: WebSockets, defect)

Description

[Patrick McManus [:mcmanus]]

over in bug 669819 a stack with this assertion was reported. It turns out to be a distraction from the real issue of that bug, so I am splitting them apart.

From https://bugzilla.mozilla.org/show_bug.cgi?id=669819#c7
"the real cause is a regression from 537787, mLoadGroup was supposed to be moved to a base class but was instead duplicated in both the base and derived classes.

It should also be proxy released as I mentioned in comment 6, but that's apparently only relevant on a corner case that isn't being tested."

From https://bugzilla.mozilla.org/show_bug.cgi?id=669819#c10

"r=jdm"

From https://bugzilla.mozilla.org/show_bug.cgi?id=669819#c11

jduell says a? aurora

"The e10s patch landed in aurora, and this fixes a relatively important yet simple bug in it."

Comment 1

[Patrick McManus [:mcmanus]]

Attachment: patch v1

Comment 2

[Christopher Blizzard (:blizzard)]

Comment on attachment 544586 [details] [diff] [review]
patch v1

Approved for Aurora.

Comment 3

[Christopher Blizzard (:blizzard)]

Tracking for Aurora / Firefox 7 because it's a new(ish) feature.

Comment 4

[Marco Bonardo [:mak]]

http://hg.mozilla.org/mozilla-central/rev/82d7d9c93a07

Comment 5

[Patrick McManus [:mcmanus]]

http://hg.mozilla.org/releases/mozilla-aurora/rev/8f58f46715a2

Comment 6

[Virgil Dicu [:virgil] [QA]]

Mozilla/5.0 (X11; Linux x86_64; rv:7.0) Gecko/20100101 Firefox/7.0

Could you provide a test case or some clear STRs in order to get this issue verified on QA side?

Comment 7

[Patrick McManus [:mcmanus]]

the STR was a mochitest orange failure involving test_websocket_basic.html that has cleared up. comment 0 points to 669819 for the particular failure.

Comment 8

[u279076]

qa- as no QA fix verification needed