[Mlir-commits] [clang] [llvm] [mlir] [AMDGPU] Introduce builtin, intrinsic and ROCDL instruction for buffer_inv (PR #228497)

[Matt Arsenault]

================
@@ -0,0 +1,35 @@
+// REQUIRES: amdgpu-registered-target
+// RUN: %clang_cc1 -cl-std=CL2.0 -triple amdgpu9.4-amd-amdhsa -DTEST_VALID -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -cl-std=CL2.0 -triple amdgpu9.4-amd-amdhsa -DTEST_INVALID -fsyntax-only -verify %s
+// RUN: %clang_cc1 -cl-std=CL2.0 -triple amdgpu9.0a-amd-amdhsa -DTEST_UNSUPPORTED -emit-llvm -o /dev/null -verify %s
----------------
arsenm wrote:

The pure sema tests should go in test/SemaOpenCL

https://github.com/llvm/llvm-project/pull/228497