CRASH: Performance tool crashes with nsIMessageSender.sendAsyncMessage with memory enabled

RESOLVED INVALID

Status

P1
normal
RESOLVED INVALID
4 years ago
6 months ago

People

(Reporter: jsantell, Unassigned)

Tracking

(Blocks: 1 bug)

37 Branch

Firefox Tracking Flags

(Not tracked)

Details

(Whiteboard: [polish-backlog])

I've gotten this a few times -- pulled down anew and rebuilt and still came across this. Seems to happen irregularly, but running an octane test generally causes it to occur immediately. Only seems to occur with memory enabled.

Occurs in both e10s and non e10s.

Handler function threw an exception: [Exception... "Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIMessageSender.sendAsyncMessage]"  nsresult: "0x80004005 (NS_ERROR_FAILURE)"  location: "JS frame :: resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/devtools/server/main.js :: DebuggerServer.connectToChild/destroy< :: line 900"  data: no]
Stack: DebuggerServer.connectToChild/destroy<@resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/devtools/server/main.js:900:9
makeInfallible/<@resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/devtools/DevToolsUtils.js:82:14
DebuggerServer.connectToChild/onMessageManagerClose@resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/devtools/server/main.js:936:9
Blocks: 1075567
Priority: -- → P1
Whiteboard: [devedition-40]
Is this a crash or an error in the console? If it is a crash, do you have the stack from about:crashes, or can you repro with a debug build + debuggger?
Crash: https://crash-stats.mozilla.com/report/index/0df0bdd5-9789-4251-8fb5-531822150428

This is a full crash, exiting in non-e10s, and crashing all tabs in e10s.
Blocks: 1160812
No longer blocks: 1075567
(In reply to Jordan Santell [:jsantell] [@jsantell] from comment #2)
> Crash:
> https://crash-stats.mozilla.com/report/index/0df0bdd5-9789-4251-8fb5-
> 531822150428
> 
> This is a full crash, exiting in non-e10s, and crashing all tabs in e10s.

Any chance you could reproduce this with a debug build?

Does this happen with just allocations and not memory turned on? (Maybe we can stop blocking perf-allocations?)
Haven't seen this in awhile, but I've also had memory and allocations off. With them separate, I'll try allocations always on to see if old or new issues shake up
Whiteboard: [devedition-40] → [polish-backlog]
Haven't seen this.
Status: NEW → RESOLVED
Last Resolved: 3 years ago
Resolution: --- → INVALID

Updated

6 months ago
Product: Firefox → DevTools
You need to log in before you can comment on or make changes to this bug.