Add optional feature dual_source_blending - #4621
Conversation
|
Previews, as seen when this build job started (b98a4b0): |
kainino0x
left a comment
There was a problem hiding this comment.
Still some issues in the validation algorithm; it was too complicated to try to explain what should be done so I've pushed a commit. Please review my changes!
|
Thanks @kainino0x and @alan-baker ! I've merged the main branch into the PR and fixed several nits in Kai's comment. PTAL ,thanks! |
Thanks for the fixes! API text LGTM (except that I wrote part of it) but I haven't checked if it matches the validation in the backends. Will need tests. |
toji
left a comment
There was a problem hiding this comment.
One request for the WebGPU spec, but overall it's looking good! I'm not in a good position to evaluate the WGSL spec changes.
|
Thanks for adding the feature column to the blend factor table! LGTM, though I'll wait for someone on the WGSL spec side to give final approval. |
As is required by Metal Shading Language: ``` 5.2.3.5 Fragment Function Output Attributes Multiple elements in the fragment function return type that use the same color attachment index for blending needs to be declared with the same data type. ```
|
Thanks @alan-baker! @kainino0x could you take a look? |
dneto0
left a comment
There was a problem hiding this comment.
Overall looks good.
I have one suggestion: forbid use of blend_src without also having a location attribute.
teoxoy
left a comment
There was a problem hiding this comment.
Looks consistent with what we previously agreed on. 👍
WGSL 2024-06-04 Minutes
|
dneto0
left a comment
There was a problem hiding this comment.
Thanks! Sorry for the delay.
|
@kainino0x can we merge this PR right now? |
|
Yes we can merge. |
GPU Web WG 2024-05-29 Atlantic-time
|
This patch adds
dual_source_blendingto WebGPU and WGSL SPEC.dual_source_blendingas an optional feature in both WebGPU and WGSL.@blend_srcin WGSL whendual_source_blendingis enabled.src1to WebGPU SPEC.Fixed: #4283