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

Issue 613206 link

Starred by 2 users
Status: Fixed
Owner:
cwallez@chromium.org
Closed: Nov 2016
Cc:
█ kbr@chromium.org
█ qiankun....@intel.com
zmo@chromium.org
yunchao...@intel.com
Components:
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 3
Type: Bug



Sign in to add a comment

deqp referencerenderer rasterization algorithm differ in js and c++

Project Member Reported by zmo@chromium.org, May 19 2016 
In c++, we always do in 2x2 grid for the purpose of texture lod selection, whereas in js it's just simple rasterization per pixel through baricentric coordinates.

This caused rendering mipmapped textures to size 1 (either width or height) textures differ from GL drivers, and therefore a bunch of deqp js tests fail.

For now, we will bypass verifying rendering to size 1 textures to make these tests passing. It would be nice to match the js referencerenderer with the c++ version in the future and remove such hack.

Comment 1 by kbr@chromium.org, May 19 2016 

Thanks for tracking this down and filing it Mo. Could you also file an issue on https://github.com/KhronosGroup/WebGL since the bug's really in that repository? Thanks.

Comment 2 by zmo@chromium.org, May 19 2016 

This only happens to 3D textures. In js, the z dimension is not being considered because it doesn't have the 2x2 grid to get diff on z.  This might be a possibility of the root cause.

Comment 4 by zmo@chromium.org, Nov 5 2016

Blocking: -295792 662644
Owner: cwallez@chromium.org
Status: Assigned (was: Available)
cwallez brought the js implementation much closer to c++.

Not sure if we ever will do more on this.

Comment 5 by cwallez@chromium.org, Nov 7 2016

Status: Fixed (was: Assigned)
Yeah, especially since we are not going to make more dEQP tests and that they are all passing on some implementation.

Comment 6 by jdarpinian@chromium.org, Nov 27

Blocking: -662644
►

Sign in to add a comment