Add SCALE_DTYPE and ZP_DTYPE support for quantization shaders

[ahmtox]

Summary:
This change adds support for parameterized SCALE_DTYPE and ZP_DTYPE to the quantization and dequantization shaders. This is necessary as when exporting llama with "8da4w" you might have different affine calls with various scale and zero point dtypes. I've also added functionality to automatically populate optional parameters.

NOTE: Disable the fusion for linear_qta8a_qga4w as the bug for why it doesn't work with exporting llama is being resolved.

Key Changes:

(1) YAML Configuration Updates:

• Added SCALE_DTYPE and ZP_DTYPE parameters to quantize_texture.yaml and dequantize_texture.yaml
• Added generate_variant_forall entries for SCALE_DTYPE (float) and ZP_DTYPE (int8, int32, float)
• This enables shader variants for different scale and zero_point data types

(2) GLSL Shader Updates:

• Added SCALE_T and ZP_T type definitions using the new parameters
• Updated tensor declarations to use parameterized types instead of hardcoded "float" and "int"
• Added proper type casting (float() and int()) for all scale and zero_point accesses
• Added required extensions for SCALE_DTYPE and ZP_DTYPE

(3) C++ Implementation Updates:

• Added dtype suffixes for scale and zero_point in all quantize/dequantize node functions
• Added comprehensive data type validation in all implementation functions:
  • Scale tensors: fp32 only (for now)
  • Zero point tensors: int32, int8, fp32
• Updated Quantize.cpp, Dequantize.cpp, and ChooseQParams.cpp with consistent validation

This change resolves shader compilation errors and enables more flexible quantization strategies by supporting multiple data types for quantization parameters.

Differential Revision: D79835267

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/13225

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 2 Cancelled Jobs, 1 Unrelated Failure

As of commit aaad9b6 with merge base f7ddbde ():

NEW FAILURES - The following jobs have failed:

• Build documentation / build (buck2) / Build doc (gh)
  At least one of the pre-conditions you specified did not hold
• pull / test-moshi-linux / linux-job (gh)
  RuntimeError: Command docker exec -t ea1e1aab94d05894799b8275deec7e687cf068e34c8abbab34541b28a91879f5 /exec failed with exit code 1

CANCELLED JOBS - The following jobs were cancelled. Please retry:

• pull / test-static-llama-qnn-linux / linux-job (gh)
  ##[error]The operation was canceled.
• pull / unittest-arm-backend-with-no-fvp (test_pytest_models) / linux-job (gh)
  ##[error]The operation was canceled.

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / test-openvino-linux / linux-job (gh) (trunk failure)
  AttributeError: '_OpNamespace' 'quantized_decomposed' object has no attribute 'convert_element_type'

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[facebook-github-bot]

This pull request was exported from Phabricator. Differential Revision: D79835267

[github-actions]

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

[facebook-github-bot]

This pull request was exported from Phabricator. Differential Revision: D79835267

[facebook-github-bot]

This pull request was exported from Phabricator. Differential Revision: D79835267