[llvm] r336374 - AMDGPU: Don't use spir_kernel in a test

[Matt Arsenault via llvm-commits]

Author: arsenm
Date: Thu Jul  5 10:01:29 2018
New Revision: 336374

URL: http://llvm.org/viewvc/llvm-project?rev=336374&view=rev
Log:
AMDGPU: Don't use spir_kernel in a test

Also use verify-machineinstrs.

Modified:
    llvm/trunk/test/CodeGen/AMDGPU/coalescer_distribute.ll

Modified: llvm/trunk/test/CodeGen/AMDGPU/coalescer_distribute.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AMDGPU/coalescer_distribute.ll?rev=336374&r1=336373&r2=336374&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/coalescer_distribute.ll (original)
+++ llvm/trunk/test/CodeGen/AMDGPU/coalescer_distribute.ll Thu Jul  5 10:01:29 2018
@@ -1,9 +1,8 @@
-; RUN: llc -o /dev/null %s
+; RUN: llc -mtriple=amdgcn-- -verify-machineinstrs -o /dev/null %s
 ; This testcase produces a situation with unused value numbers in subregister
 ; liveranges that get distributed by ConnectedVNInfoEqClasses.
-target triple = "amdgcn--"
 
-define spir_kernel void @hoge() {
+define amdgpu_kernel void @hoge() {
 bb:
   %tmp = tail call i32 @llvm.amdgcn.workitem.id.x()
   br i1 undef, label %bb2, label %bb23