1294029 - window.open makes code called by setTimeout to run in parallel

Status: RESOLVED WORKSFORME

(Core :: DOM: Core & HTML, defect, P3)

Description

[Alex]

User Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:48.0) Gecko/20100101 Firefox/48.0
Build ID: 20160726073904

Steps to reproduce:


1. window.open code executed through a handler function, which is called on a DOM event (click)
2. some arbitrary code executed through the same handler function, which is called repetitively by setTimeout (at very short intervals)

See sample code at http://jsfiddle.net/d9rHQ/3/
Handler function in the sample code is `entry0`




Actual results:

the handler function is executed in parallel

in the specified fiddle, the console outputs a lot of lines - because the calls were executed in parallel the variable `entryDepth` is not 0 anymore


Expected results:

all executions of handler function should be sequential - this would make `entryDepth` always be 0 and would yield no console output

Chrome, Internet Explorer and Edge yields to console output whatsoever.

Comment 1

[Alex]

Debuggin Info:

Nume: Firefox
Versiune: 48.0
Identificator versiune: 20160726073904
Sursă de actualizare: release
Agent de utilizator: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:48.0) Gecko/20100101 Firefox/48.0
SO: Windows_NT 10.0 x86
Ferestre multiproces: 0/10 (Disabled)
Safe mode: false

Comment 2

[Alex]

More or less related bug reports:
https://bugzilla.mozilla.org/show_bug.cgi?id=321880
https://bugzilla.mozilla.org/show_bug.cgi?id=1180005

This issue occurs frequently with GWT. Details here:
https://github.com/gwtproject/gwt/issues/7923
https://github.com/gwtproject/gwt/issues/9152

Comment 3

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Hi Bevis, are you interested in taking a look here?

Comment 4

[Hsin-Yi Tsai (she/her) [:hsinyi]]

I reproduced this issue on FF48, but I didn't see this happening on Nightly 51.

Comment 5

[Bevis Tseng[:bevis][:btseng](Exited)]

(In reply to Hsin-Yi Tsai [:hsinyi] from comment #4)
> I reproduced this issue on FF48, but I didn't see this happening on Nightly
> 51.
After further testing, this can be reproduced from FF48 to FF51 with e10s disabled. :)

Comment 6

[Hsin-Yi Tsai (she/her) [:hsinyi]]

(In reply to Bevis Tseng[:bevistseng][:btseng] from comment #5)
> (In reply to Hsin-Yi Tsai [:hsinyi] from comment #4)
> > I reproduced this issue on FF48, but I didn't see this happening on Nightly
> > 51.
> After further testing, this can be reproduced from FF48 to FF51 with e10s
> disabled. :)

Oh, right, e10s ... :)

Comment 7

[Bevis Tseng[:bevis][:btseng](Exited)]

sorry, still don't have bandwidth to look into this one...

Comment 8

[Hsin-Yi Tsai (she/her) [:hsinyi]]

(In reply to Bevis Tseng[:bevistseng][:btseng] from comment #7)
> sorry, still don't have bandwidth to look into this one...

Let's me explore other options :)

Comment 9

[Kagami Rosylight [:saschanaz] (they/them)]

The attached repro step here doesn't seem to cause problem anymore. Per comment #5 it was seemingly pre-e10s specific issue.