[llvm] AMDGPU: Remove "gws" feature from generic targets (PR #148122)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu

Author: Changpeng Fang (changpeng)

<details>
<summary>Changes</summary>

  Here "generic targets" means when no target is specified. This is because gfx12+ does not support this feature, and thus it is no longer universally available.

Fixes: SWDEV-541399

---
Full diff: https://github.com/llvm/llvm-project/pull/148122.diff


1 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/GCNProcessors.td (+2-2) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/GCNProcessors.td b/llvm/lib/Target/AMDGPU/GCNProcessors.td
index b5ffa64c3a4b4..217942c7e539c 100644
--- a/llvm/lib/Target/AMDGPU/GCNProcessors.td
+++ b/llvm/lib/Target/AMDGPU/GCNProcessors.td
@@ -9,11 +9,11 @@
 // The code produced for "generic" is only useful for tests and cannot
 // reasonably be expected to execute on any particular target.
 def : ProcessorModel<"generic", NoSchedModel,
-  [FeatureGDS, FeatureGWS]
+  [FeatureGDS]
 >;
 
 def : ProcessorModel<"generic-hsa", NoSchedModel,
-  [FeatureGDS, FeatureGWS, FeatureFlatAddressSpace]
+  [FeatureGDS, FeatureFlatAddressSpace]
 >;
 
 //===------------------------------------------------------------===//

``````````

</details>


https://github.com/llvm/llvm-project/pull/148122