MeshPhysicalMaterial: Fix shader error for clear coat/anisotropy when normal map is missing.

[repalash]

Related PR discussion: #25580 (comment)

Description

Add support for using vUv or vClearcoatMapUv for anisotropy when normal map is not present.
This is the smallest fix. It's also possible to set USE_UV in javascript instead of checking for USE_ANISTROPY while defining the variable.

Fixes loading of model from: https://discourse.threejs.org/t/shader-error-0-validate-status-false/53195/2
Model: https://sketchfab.com/3d-models/free-1975-porsche-911-930-turbo-8568d9d14a994b9cae59499f0dbed21e#download

[github-actions]

📦 Bundle size

Full ESM build, minified and gzipped.

Filesize dev	Filesize PR	Diff
642.9 kB (159.4 kB)	643.1 kB (159.5 kB)	+266 B

🌳 Bundle size after tree-shaking

Minimal build including a renderer, camera, empty scene, and dependencies.

Filesize dev	Filesize PR	Diff
436.1 kB (105.6 kB)	436.4 kB (105.6 kB)	+266 B

[elalish]

Thanks, that looks reasonable to me.

[Mugen87]

I can confirm this fixes the issues for me as well! Merging for now so we have a hotfix for r154.

[WestLangley]

If this line is required, then USE_UV is not properly defined.

It's also possible to set USE_UV in javascript instead of checking for USE_ANISOTROPY while defining the variable.

Right. Why didn't you do that?

[repalash]

I was not sure what the plan was for USE_UV after #25740 and a hotfix was required.
Do you recommend setting parameters.vertexUvs when anisotropy is present or creating the define either in WebGLProgram or WebGLRenderer.
Also, there is vAnisotropyMapUv that could be considered or defined cleanly when no normal map is present.

[WestLangley]

@repalash This is not your fault. Thank you for your contribution.

I think USE_UV should be properly defined. Where it is best to define it at this point I would leave to @Mugen87 and yourself.

[Mugen87]

Also, there is vAnisotropyMapUv that could be considered or defined cleanly when no normal map is present.

I would favor this approach (if feasible) since anisotropy would work similar to clear coat then.