1464521 - Perform browser error reporting processing during idle

Status: RESOLVED FIXED

(Firefox :: General, enhancement)

Description

[Osmose [:osmose, :mkelly]]

As per bug 1446817 comment 16, we should schedule the processing necessary for browser error reporting to happen when the browser is idle and has cycles to spare, as it's not vital to the user experience. John-Galt recommends using ChromeUtils.idleDispatch for this purpose.

Comment 1

[Osmose [:osmose, :mkelly]]

Attachment: Bug 1464521: Process collected browser JS errors during idle.

testFetchArguments and the add-on ID mangling tests already cover error
collection end-to-end, so no new tests are needed to cover testing the idle
callback.

Review commit: https://reviewboard.mozilla.org/r/249328/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/249328/

Comment 2

[Kris Maglione [:kmag]]

Comment on attachment 8983464 [details]
Bug 1464521: Process collected browser JS errors during idle.

https://reviewboard.mozilla.org/r/249328/#review256004

Thanks

::: browser/modules/BrowserErrorReporter.jsm:202
(Diff revision 1)
>  
>      return "FILTERED";
>    }
>  
> -  async observe(message) {
> +  observe(message) {
> +    ChromeUtils.idleDispatch(() => this.handleMessage(message));

Maybe do `if (message instanceof Ci.nsIScriptError) { idleDispatch(); }` (and remove the QI call in handleMessage) just so we don't wind up creating idle tasks for messages that don't need to be processed

Comment 3

[Osmose [:osmose, :mkelly]]

Comment on attachment 8983464 [details]
Bug 1464521: Process collected browser JS errors during idle.

Review request updated; see interdiff: https://reviewboard.mozilla.org/r/249328/diff/1-2/

Comment 4

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/e6d89c09bc10
Process collected browser JS errors during idle. r=kmag

Comment 5

[Andrei Ciure[:aciure]]

https://hg.mozilla.org/mozilla-central/rev/e6d89c09bc10