[llvm] [amdgpu] Add HasBufferInv AMDGPUSubtargetFeature (PR #226254)

[Dan Zimmerman via llvm-commits]

================
@@ -0,0 +1,10 @@
+// RUN: llvm-mc -triple=amdgpu9.42 -show-encoding %s | FileCheck --check-prefix=SUPPORTED %s
+// RUN: llvm-mc -triple=amdgpu9.50 -show-encoding %s | FileCheck --check-prefix=SUPPORTED %s
+// RUN: llvm-mc -triple=amdgcn-amd-amdhsa -mcpu=gfx9-4-generic -show-encoding %s | FileCheck --check-prefix=SUPPORTED %s
+// RUN: not llvm-mc -triple=amdgpu9.0a -filetype=null %s 2>&1 | FileCheck --check-prefix=GFX90A %s
+// RUN: not llvm-mc -triple=amdgpu9.42 -mattr=-buffer-inv-inst -filetype=null %s 2>&1 | FileCheck --check-prefix=DISABLED %s
+
+buffer_inv sc0 sc1
----------------
danzimm wrote:

Yeah, fair- I added this just to cover my bases. Was thinking I should come to the table with more tests than needed rather than look like I didn't think about which test cases to have.

https://github.com/llvm/llvm-project/pull/226254