732117 - Fennec debug intent is broken

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[Benoit Girard (:BenWa)]

This is caused by the patches in bug 727960, unapplying them fixes the problem.

Comment 1

[Benoit Girard (:BenWa)]

I should be more specific:

The native front end starts up but Gecko does not.

Comment 2

[Mike Hommey [:glandium]]

(In reply to Benoit Girard (:BenWa) from comment #0)
> This is caused by the patches in bug 727960, unapplying them fixes the
> problem.

Did you try unapplying them one by one?

Comment 3

[Benoit Girard (:BenWa)]

No, both together, do you have a guess and I'll try it?

Comment 4

[Mike Hommey [:glandium]]

1419f0146049 is a likely suspect

Comment 5

[Joe Drew (not getting mail)]

Is this actually maple-specific? Presuming not, I'll remove the whiteboard tag.

Comment 6

[Benoit Girard (:BenWa)]

With 1419f0146049 unapplied I receive a corrupted stack crash. I do receive some log from Gecko so it is starting up:
E/Gecko   (26551): ### Acquire()
E/Gecko   (26551): Registered Compositor
I/Gecko   (26551): ... requesting window surface from bridge
I/Gecko   (26551): got surface 0x3c6e70
I/Gecko   (26551):  -- creating basic, not accelerated
E/Gecko   (26551): ### Basic layers drawing
E/Gecko   (26551): ### Calling EndDrawing()!

Comment 7

[Mike Hommey [:glandium]]

From a quick test, it looks like GeckoThread.run is not even called.

Comment 8

[Mike Hommey [:glandium]]

This is actually a regression from bug 728240. Reverting it fixes debug intent for me.

Comment 9

[Sriram Ramasubramanian [:sriram]]

I wouldnt recommend reverting the patch. Instead, could you please add a condition there to do debug intent?
Also, onResume() is called whenever Fennec comes to foreground. So, this intent should be processed only once in initialize()

Comment 10

[Sriram Ramasubramanian [:sriram]]

Attachment: Patch

This patch adds the debug intent processing to the startup path.

Comment 11

[Mike Hommey [:glandium]]

Comment on attachment 603335 [details] [diff] [review]
Patch

Review of attachment 603335 [details] [diff] [review]:
-----------------------------------------------------------------

It looks good to me (and, more importantly, works), but I won't pretend I know this code to review it.

Comment 12

[Sriram Ramasubramanian [:sriram]]

http://hg.mozilla.org/integration/mozilla-inbound/rev/f1a5fd7d0baf

Comment 13

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/f1a5fd7d0baf

(Please set the assignee when attaching and check when landing, thanks :-))