[Mlir-commits] [mlir] [mlir][spirv] Add support for VectorAnyINTEL capability (PR #68034)

[Jakub Kuderski]

================
@@ -2,7 +2,7 @@
 
 module attributes {
   gpu.container_module,
-  spirv.target_env = #spirv.target_env<#spirv.vce<v1.3, [Kernel, Addresses, Groups, GroupNonUniformArithmetic, GroupUniformArithmeticKHR], []>, #spirv.resource_limits<>>
+  spirv.target_env = #spirv.target_env<#spirv.vce<v1.3, [Kernel, Addresses, Groups, GroupNonUniformArithmetic, GroupUniformArithmeticKHR], [SPV_KHR_uniform_group_instructions]>, #spirv.resource_limits<>>
----------------
kuhar wrote:

Why is this capability added everywhere in this test? If this is necessary to pass validation, we should land it separately from this PR.

https://github.com/llvm/llvm-project/pull/68034