Skip to content

Can fragment output location be of vecN<f16> type? #5551

@teoxoy

Description

@teoxoy

I couldn't find a mention of this in the f16 proposal. The PR (#2696) also doesn't seem to touch the "validating GPUFragmentState" steps but the problem is that the existing validation talks about vector types without specifying their inner scalar types.

For context, Firefox user bug report: Bug 2013404

Metadata

Metadata

Assignees

No one assigned

    Labels

    apiWebGPU APIcopyeditingPure editorial stuff (copyediting, *.bs file syntax, etc.)

    Type

    No type

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions