Alice0775 White Reporter
	Description • 10 years ago

Steps to Reproduce:
0. Start with e10s window
1. Open http://webglsamples.org/aquarium/aquarium.html


Actual Results
Canvas become gray

Regression window(m-c)
Good:
https://hg.mozilla.org/mozilla-central/rev/67c42c076393
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:37.0) Gecko/20100101 Firefox/37.0 ID:20141227032230
Bad:
https://hg.mozilla.org/mozilla-central/rev/6d5a1e68f248
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:37.0) Gecko/20100101 Firefox/37.0 ID:20141228104931
Pushlog:
https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=67c42c076393&tochange=6d5a1e68f248

Suspect:
	8d5da151b0df	Jeff Muizelaar — Bug 1112780. Recreate the RenderTargetView on demand. r=jgilbert This avoids reusing the RenderTargetView across AcquireSync/ReleaseSync pairs and fixes bug 1109718. It adds a ReleaseSync call to ANGLE that causes it to invalidate the RenderTargetView.

	Alice0775 White Reporter
	Comment 1 • 10 years ago

Graphics
--------

Adapter Description: AMD Radeon HD 6450
Adapter Drivers: aticfx64 aticfx64 aticfx64 aticfx32 aticfx32 aticfx32 atiumd64 atidxx64 atidxx64 atiumdag atidxx32 atidxx32 atiumdva atiumd6a atitmm64
Adapter RAM: 1024
ClearType Parameters: Gamma: 2200 Pixel Structure: R ClearType Level: 100 Enhanced Contrast: 200
Device ID: 0x6779
Direct2D Enabled: true
DirectWrite Enabled: true (6.2.9200.16571)
Driver Date: 11-20-2014
Driver Version: 14.501.1003.0
GPU #2 Active: false
GPU Accelerated Windows: 2/2 Direct3D 11 (OMTC)
Subsys ID: 23111787
Vendor ID: 0x1002
WebGL Renderer: Google Inc. -- ANGLE (AMD Radeon HD 6450 Direct3D11 vs_5_0 ps_5_0)
windowLayerManagerRemote: true
AzureCanvasBackend: direct2d 1.1
AzureContentBackend: direct2d 1.1
AzureFallbackCanvasBackend: cairo
AzureSkiaAccelerated: 0

	Jeff Muizelaar [:jrmuizel] Assignee
	Comment 2 • 10 years ago

Does this only happen with e10s?

	Alice0775 White Reporter
	Comment 3 • 10 years ago

(In reply to Jeff Muizelaar [:jrmuizel] from comment #2)
> Does this only happen with e10s?

Yes, it only happen with e10s window.
It works OK with non-e10s window.

	Kelsey Gilbert [:jgilbert]
	Comment 4 • 10 years ago

It looks like it's displaying a texture/renderbuffer from inside the demo.

	Jeff Muizelaar [:jrmuizel] Assignee
	Comment 5 • 10 years ago

It looks this might be caused by readback being broken. Maybe because we're throwing away the rendertarget view.

	Jeff Muizelaar [:jrmuizel] Assignee
	Comment 6 • 10 years ago

Looks like it's broken even with out throwing away the render target view. My guess is that may have to do with us no AcquireSyncing on the surface before trying to read from it...

	Kelsey Gilbert [:jgilbert]
	Comment 8 • 10 years ago

Comment on attachment 8542689 [details] [diff] [review]
Make sure we grab the keyed mutex when we read from the surface

Review of attachment 8542689 [details] [diff] [review]:
-----------------------------------------------------------------

::: gfx/gl/GLScreenBuffer.cpp
@@ +525,4 @@
>    }
>  
>    {
> +      src->ProducerAcquire();

It seems like this should be ConsumerAcquire, or does it not matter?

	Ryan VanderMeulen [:RyanVM]
	Comment 9 • 9 years ago

https://hg.mozilla.org/mozilla-central/rev/ed186e5af4a1