New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 752143 link

Starred by 2 users

Issue metadata

Status: Duplicate
Merged: issue 751206
Owner:
OOO until 2019-01-24
Closed: Aug 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 2
Type: Bug-Regression



Sign in to add a comment

WebGL error (glBlitFramebufferCHROMIUM: src and dst formats differ for color), only on chrome

Reported by yaoli...@gmail.com, Aug 3 2017

Issue description

Chrome Version       : 60.0.3112.90
OS Version: OS X 10.12.6
URLs (if applicable) :http://media.hhmi.org/biointeractive/biointeractive_3d_viewer/3D_viewer.html?load=./content/rhinovirus_14/rhinovirus_14.json&autorotate
Other browsers tested:
  Add OK or FAIL after other browsers where you have tested this issue:
     Safari 5:OK
  Firefox 4.x:OK
     IE 11:OK
     Opera:OK

What steps will reproduce the problem?
1. open it


What is the expected result?
This is a WebGL app and you are supposed to see a 3D model in canvas. 

What happens instead of that?
I've tested multiple computers (>20), iPads and cell phones. 3 of the Macs showed black screen. Other macs and devices worked fine. Other browsers worked fine on these 3 Macs as well.

Please provide any additional information below. Attach a screenshot if
possible.
Here is the console message on these 3 comptuers (also attached a screenshot):

[.Offscreen-For-WebGL-0x7fb48fa39800]GL ERROR :GL_INVALID_OPERATION : glBlitFramebufferCHROMIUM: src and dst formats differ for color

UserAgentString: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.90 Safari/537.36



 
Screen Shot 2017-08-03 at 8.13.21 AM.png
33.9 KB View Download
Cc: vmi...@chromium.org ericrk@chromium.org
Components: Blink>WebGL
Labels: -Pri-3 Pri-2
Owner: manoranj...@chromium.org
Status: Untriaged (was: Unconfirmed)
Able to reproduce this issue on reported version of chrome#60.0.3112.90. This seems to be a recent regression and working fine in M56.
Labels: -Type-Bug ReleaseBlock-Stable M-60 Type-Bug-Regression
Owner: kbr@chromium.org
Status: Assigned (was: Untriaged)
Here is the narrow bisect result.

https://chromium.googlesource.com/chromium/src/+log/9a47d86f8e4000d54119e7f860b99684bf72b5ca..3d4d5d7a9198bea6f0018a36fc6825637f9a12c4

Good# 59.0.3071.115
Bad# 60.0.3112.78

kbr@, can you please look into this change (https://chromium.googlesource.com/chromium/src/+/3d4d5d7a9198bea6f0018a36fc6825637f9a12c4) ?

This is a recent regression in M60 stable, so please let us if this is a real blocker for stable release.

Thank you!
Cc: abdulsyed@chromium.org bustamante@chromium.org
Labels: hasbisect

Comment 5 by kbr@chromium.org, Aug 7 2017

Mergedinto: 751206
Status: Duplicate (was: Assigned)
Thanks for the report and sorry for the trouble. This is a WebGL 2.0- and Mac-specific regression and has been fixed in  Issue 751206 . The fix will be in Chrome 61. I'm sorry, but despite the fact that the fix is tiny, I can't justify merging it back to Chrome Stable. It doesn't affect a large enough user base.

In the future, please do test with the early access channels -- in particular Chrome Canary or Dev -- and file bugs (and feel free to email me the bug IDs directly) if you see breakage of your WebGL content.

BTW, you can work around this bug by allocating your back buffer with alpha:true instead of alpha:false. While you'll need to make sure your alpha channel is cleared to 1.0, this should be a relatively small workaround.

Comment 6 by kbr@chromium.org, Aug 7 2017

Labels: -ReleaseBlock-Stable

Comment 7 by yaoli...@gmail.com, Aug 8 2017

Great!Thank you for your response.

Sign in to add a comment