1048250 - disappearing elements in GUI with OMTC turned on

Status: RESOLVED FIXED

(Core :: Graphics, defect)

Description

[Eimantas]

Attachment: report bug.png

User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:34.0) Gecko/20100101 Firefox/34.0 (Beta/Release)
Build ID: 20140803030203

Steps to reproduce:

Do nothing, it comes always.


Actual results:

If I hover with a mouse over an gui element (Bookmar on Toolbar, or an element in a website, like a textbox) it disappears. Screenshots below.  


Expected results:

The elemts stays the same in GUI (or maybe highlighted)

Comment 1

[Eimantas]

Additional : on Firefox official release it works OK, and nightly from (2014-03-16) it works ok too.

Comment 2

[:Gijs (out for now; he/him)]

Can you check when this broke using mozregression ( http://mozilla.github.io/mozregression/ ) ?

Comment 3

[Eimantas]

Ok, ill check it, but I can do only tomorrow. I'll check in wich build this gone bad.

Comment 4

[Eimantas]

I made the regression tests, and I found out that the last good build was on 2014-05-20, and on the build on 2014-05-21 it gone bad.

Comment 5

[Paul Silaghi, QA [:pauly]]

Did you use mozregression? If yes, could you go forward with the inbound builds ?

Comment 6

[:Gijs (out for now; he/him)]

(In reply to Eimantas from comment #4)
> I made the regression tests, and I found out that the last good build was on
> 2014-05-20, and on the build on 2014-05-21 it gone bad.

http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=cb9f34f73ebe&tochange=9d8d16695f6a

This corresponds to switching on OMTC on Windows. Pinging Milan about this. Eimantas, can you copy the contents of the graphics section of about:support on your machine into the bug?

Comment 7

[Eimantas]

I just done further regression test. Soo, it turns out that the last good revision was "66fe0b3237d6" and the first bad revision was "718a9852b60d"
And here are my Graphics section : 
Adapter Description	NVIDIA NVS 4200M
Adapter Description (GPU #2)	Intel(R) HD Graphics Family
Adapter Drivers	nvd3dumx,nvwgf2umx,nvwgf2umx nvd3dum,nvwgf2um,nvwgf2um
Adapter Drivers (GPU #2)	igdumd64 igd10umd64 igd10umd64 igdumdx32 igd10umd32 igd10umd32
Adapter RAM	512
Adapter RAM (GPU #2)	Unknown
Device ID	0x1056
Device ID (GPU #2)	0x0126
Direct2D Enabled	true
DirectWrite Enabled	true (6.1.7601.18245)
Driver Date	6-5-2011
Driver Date (GPU #2)	6-10-2011
Driver Version	8.17.12.6883
Driver Version (GPU #2)	8.15.10.2418
GPU #2 Active	false
GPU Accelerated Windows	1/1 Direct3D 10
Vendor ID	0x10de
Vendor ID (GPU #2)	0x8086
WebGL Renderer	Google Inc. -- ANGLE (NVIDIA NVS 4200M Direct3D9Ex vs_3_0 ps_3_0)
windowLayerManagerRemote	false
AzureCanvasBackend	direct2d
AzureContentBackend	direct2d
AzureFallbackCanvasBackend	cairo
AzureSkiaAccelerated	0

Comment 8

[Paul Silaghi, QA [:pauly]]

Please post the pushlog link from mozregression

Comment 9

[:Gijs (out for now; he/him)]

(In reply to Eimantas from comment #7)
> I just done further regression test. Soo, it turns out that the last good
> revision was "66fe0b3237d6" and the first bad revision was "718a9852b60d"
> And here are my Graphics section : 
> <snip>

Thanks!

(In reply to Paul Silaghi, QA [:pauly] from comment #8)
> Please post the pushlog link from mozregression

Paul, this is trivial to obtain from what Eimantas already posted in comment #7...:

http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=66fe0b3237d6&tochange=718a9852b60d

and matches what I said earlier.

Comment 10

[Eimantas]

Here you go: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=66fe0b3237d6&tochange=718a9852b60d

Comment 11

[Paul Silaghi, QA [:pauly]]

Thanks Eimantas!

(In reply to :Gijs Kruitbosch (Bugmail catchup, needinfo if urgent) from comment #9)
> Paul, this is trivial to obtain from what Eimantas already posted in comment
> #7...:
It was more like a future reference...

Comment 12

[Bas Schouten (:bas.schouten)]

Could you test with OMTC enabled but the pref gfx.direct2d.disabled set to true?

Comment 13

[Eimantas]

(In reply to Bas Schouten (:bas.schouten) from comment #12)
> Could you test with OMTC enabled but the pref gfx.direct2d.disabled set to
> true?

It helped! I am not sure anymore if its a bug at all. Anyway thanks!

Comment 14

[Matt Woodrow (:mattwoodrow)]

I can't reproduce this either. It's a really old nvidia driver in use, we could probably blacklist it easily enough.

Comment 15

[Bas Schouten (:bas.schouten)]

I don't think this needs to block Aurora at this point considering the low impact it seems to have. It may become relevant for Beta, but I'm not 100% sure yet. Fwiw, if we can confirm that we should blacklist then we should still probably uplift the blacklist patch. On the bright side, we can always use the downloadable blacklist to do this later.

Comment 16

[Nicolas Silva [:nical]]

Lots of fixes related to OMTC on windows have landed the past few weeks and no report of this issue for a while so I am closing this bug. Feel free to reopen if the problems shows up again.