910159 - Assertion failure "Should not have an exception set here!" in OperationCallback

Status: RESOLVED WORKSFORME

(Core :: DOM: Workers, defect)

Description

[David Teller [:Yoric] - still alive but not very active]

See https://tbpl.mozilla.org/php/getParsedLog.php?id=27082009&tree=Try&full=1#error1

Note that the error message is not standard, I have replaced it with something else while attempting to pinpoint the error. The original assertion failure was "Should not have an exception set here!"

STR: Attempting to land bug 891300 makes this a frequent error in xpcshell tests that use OS.File.

From the point of view of JavaScript, the error seems to be sent by the following sequence:

        let xhr = new XMLHttpRequest();
        xhr.open("GET", uri, false);
        xhr.responseType = "text";
        xhr.send(); // Assertion fails here

The code is executed on a worker.

Comment 1

[David Teller [:Yoric] - still alive but not very active]

Note: I suspect that the issue is actually caused by XHR and not by webrtc. Cc-ing bent.

Comment 2

[Paul Adenot (:padenot)]

This is probably caused by some buffer overflow in js land that cause VoEMediaProcess to appear here. Having VoEMediaProcess in this code makes absolutely no sense anyways. Changing component to JS engine.

Comment 3

[David Teller [:Yoric] - still alive but not very active]

I strongly suspect that this is a bug in the implementation of XHR for workers. Re-classifying in Workers.

Comment 4

[David Teller [:Yoric] - still alive but not very active]

Marking as blocker, as this bug prevents me from proceeding with the OS.File refactoring.

Comment 5

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

"blocks something" does not mean the severity is a blocker.

Comment 6

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Looks like the operation callback is firing during the middle of exception handling?

Comment 7

[David Teller [:Yoric] - still alive but not very active]

(In reply to Kyle Huey [:khuey] ([e-mail]) from comment #5)
> "blocks something" does not mean the severity is a blocker.

True. This bug is still quite annoying for me as it prevents a pile of refactorings from landing.

Comment 8

[David Teller [:Yoric] - still alive but not very active]

The STR cannot be reproduced anymore, so I guess something was fixed somewhere.
Closing.