Update on "[ET-VK] Implemement linear_dq8ta_q4gsw"

Title says it all!

Build upon the support for quantized linear introduced in the previous diffs to enable dynamically quantized linear.

Also included in this diff is a cleanup of the glslh files used across quantized linear implementations.

Differential Revision: [D81931060](https://our.internmc.facebook.com/intern/diff/D81931060/)

[ghstack-poisoned]

Author: ssjia

.github/workflows/pull.yml

@@ -939,6 +939,7 @@ jobs:
         # Run e2e testing for selected operators. More operators will be tested via this
         # route in the future.
         python -m unittest backends/vulkan/test/test_vulkan_delegate.py -k "*pt2e*"
+        python -m unittest backends/vulkan/test/test_vulkan_delegate.py -k "*torchao*"
 
   nxp-build-test:
     name: nxp-build-test

backends/vulkan/test/test_vulkan_delegate.py

@@ -2651,6 +2651,7 @@ def forward(self, x):
             rtol=1e-1,
         )
 
+    @unittest.skip("Cannot run on swiftshader due to no 8-bit int support")
     def test_vulkan_backend_torchao_8da4w_quantized_linear(self):
         """
         Test TorchAO 8da4w quantization (int8 dynamic activation + int4 weight) with Vulkan backend.