1 files changed, 12 insertions, 12 deletions
diff --git a/llvm/test/Analysis/CostModel/AMDGPU/extractelement.ll b/llvm/test/Analysis/CostModel/AMDGPU/extractelement.ll
index c328d768646..1efbb5873ac 100644
--- a/llvm/test/Analysis/CostModel/AMDGPU/extractelement.ll
+++ b/llvm/test/Analysis/CostModel/AMDGPU/extractelement.ll
@@ -2,7 +2,7 @@
 
 ; CHECK: 'extractelement_v2i32'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <2 x i32>
-define void @extractelement_v2i32(i32 addrspace(1)* %out, <2 x i32> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v2i32(i32 addrspace(1)* %out, <2 x i32> addrspace(1)* %vaddr) {
   %vec = load <2 x i32>, <2 x i32> addrspace(1)* %vaddr
   %elt = extractelement <2 x i32> %vec, i32 1
   store i32 %elt, i32 addrspace(1)* %out
@@ -11,7 +11,7 @@ define void @extractelement_v2i32(i32 addrspace(1)* %out, <2 x i32> addrspace(1)
 
 ; CHECK: 'extractelement_v2f32'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <2 x float>
-define void @extractelement_v2f32(float addrspace(1)* %out, <2 x float> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v2f32(float addrspace(1)* %out, <2 x float> addrspace(1)* %vaddr) {
   %vec = load <2 x float>, <2 x float> addrspace(1)* %vaddr
   %elt = extractelement <2 x float> %vec, i32 1
   store float %elt, float addrspace(1)* %out
@@ -20,7 +20,7 @@ define void @extractelement_v2f32(float addrspace(1)* %out, <2 x float> addrspac
 
 ; CHECK: 'extractelement_v3i32'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <3 x i32>
-define void @extractelement_v3i32(i32 addrspace(1)* %out, <3 x i32> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v3i32(i32 addrspace(1)* %out, <3 x i32> addrspace(1)* %vaddr) {
   %vec = load <3 x i32>, <3 x i32> addrspace(1)* %vaddr
   %elt = extractelement <3 x i32> %vec, i32 1
   store i32 %elt, i32 addrspace(1)* %out
@@ -29,7 +29,7 @@ define void @extractelement_v3i32(i32 addrspace(1)* %out, <3 x i32> addrspace(1)
 
 ; CHECK: 'extractelement_v4i32'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <4 x i32>
-define void @extractelement_v4i32(i32 addrspace(1)* %out, <4 x i32> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v4i32(i32 addrspace(1)* %out, <4 x i32> addrspace(1)* %vaddr) {
   %vec = load <4 x i32>, <4 x i32> addrspace(1)* %vaddr
   %elt = extractelement <4 x i32> %vec, i32 1
   store i32 %elt, i32 addrspace(1)* %out
@@ -38,7 +38,7 @@ define void @extractelement_v4i32(i32 addrspace(1)* %out, <4 x i32> addrspace(1)
 
 ; CHECK: 'extractelement_v8i32'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <8 x i32>
-define void @extractelement_v8i32(i32 addrspace(1)* %out, <8 x i32> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v8i32(i32 addrspace(1)* %out, <8 x i32> addrspace(1)* %vaddr) {
   %vec = load <8 x i32>, <8 x i32> addrspace(1)* %vaddr
   %elt = extractelement <8 x i32> %vec, i32 1
   store i32 %elt, i32 addrspace(1)* %out
@@ -48,7 +48,7 @@ define void @extractelement_v8i32(i32 addrspace(1)* %out, <8 x i32> addrspace(1)
 ; FIXME: Should be non-0
 ; CHECK: 'extractelement_v8i32_dynindex'
 ; CHECK: estimated cost of 2 for {{.*}} extractelement <8 x i32>
-define void @extractelement_v8i32_dynindex(i32 addrspace(1)* %out, <8 x i32> addrspace(1)* %vaddr, i32 %idx) {
+define amdgpu_kernel void @extractelement_v8i32_dynindex(i32 addrspace(1)* %out, <8 x i32> addrspace(1)* %vaddr, i32 %idx) {
   %vec = load <8 x i32>, <8 x i32> addrspace(1)* %vaddr
   %elt = extractelement <8 x i32> %vec, i32 %idx
   store i32 %elt, i32 addrspace(1)* %out
@@ -57,7 +57,7 @@ define void @extractelement_v8i32_dynindex(i32 addrspace(1)* %out, <8 x i32> add
 
 ; CHECK: 'extractelement_v2i64'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <2 x i64>
-define void @extractelement_v2i64(i64 addrspace(1)* %out, <2 x i64> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v2i64(i64 addrspace(1)* %out, <2 x i64> addrspace(1)* %vaddr) {
   %vec = load <2 x i64>, <2 x i64> addrspace(1)* %vaddr
   %elt = extractelement <2 x i64> %vec, i64 1
   store i64 %elt, i64 addrspace(1)* %out
@@ -66,7 +66,7 @@ define void @extractelement_v2i64(i64 addrspace(1)* %out, <2 x i64> addrspace(1)
 
 ; CHECK: 'extractelement_v3i64'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <3 x i64>
-define void @extractelement_v3i64(i64 addrspace(1)* %out, <3 x i64> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v3i64(i64 addrspace(1)* %out, <3 x i64> addrspace(1)* %vaddr) {
   %vec = load <3 x i64>, <3 x i64> addrspace(1)* %vaddr
   %elt = extractelement <3 x i64> %vec, i64 1
   store i64 %elt, i64 addrspace(1)* %out
@@ -75,7 +75,7 @@ define void @extractelement_v3i64(i64 addrspace(1)* %out, <3 x i64> addrspace(1)
 
 ; CHECK: 'extractelement_v4i64'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <4 x i64>
-define void @extractelement_v4i64(i64 addrspace(1)* %out, <4 x i64> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v4i64(i64 addrspace(1)* %out, <4 x i64> addrspace(1)* %vaddr) {
   %vec = load <4 x i64>, <4 x i64> addrspace(1)* %vaddr
   %elt = extractelement <4 x i64> %vec, i64 1
   store i64 %elt, i64 addrspace(1)* %out
@@ -84,7 +84,7 @@ define void @extractelement_v4i64(i64 addrspace(1)* %out, <4 x i64> addrspace(1)
 
 ; CHECK: 'extractelement_v8i64'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <8 x i64>
-define void @extractelement_v8i64(i64 addrspace(1)* %out, <8 x i64> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v8i64(i64 addrspace(1)* %out, <8 x i64> addrspace(1)* %vaddr) {
   %vec = load <8 x i64>, <8 x i64> addrspace(1)* %vaddr
   %elt = extractelement <8 x i64> %vec, i64 1
   store i64 %elt, i64 addrspace(1)* %out
@@ -93,7 +93,7 @@ define void @extractelement_v8i64(i64 addrspace(1)* %out, <8 x i64> addrspace(1)
 
 ; CHECK: 'extractelement_v4i8'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <4 x i8>
-define void @extractelement_v4i8(i8 addrspace(1)* %out, <4 x i8> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v4i8(i8 addrspace(1)* %out, <4 x i8> addrspace(1)* %vaddr) {
   %vec = load <4 x i8>, <4 x i8> addrspace(1)* %vaddr
   %elt = extractelement <4 x i8> %vec, i8 1
   store i8 %elt, i8 addrspace(1)* %out
@@ -102,7 +102,7 @@ define void @extractelement_v4i8(i8 addrspace(1)* %out, <4 x i8> addrspace(1)* %
 
 ; CHECK: 'extractelement_v2i16'
 ; CHECK: estimated cost of 0 for {{.*}} extractelement <2 x i16>
-define void @extractelement_v2i16(i16 addrspace(1)* %out, <2 x i16> addrspace(1)* %vaddr) {
+define amdgpu_kernel void @extractelement_v2i16(i16 addrspace(1)* %out, <2 x i16> addrspace(1)* %vaddr) {
   %vec = load <2 x i16>, <2 x i16> addrspace(1)* %vaddr
   %elt = extractelement <2 x i16> %vec, i16 1
   store i16 %elt, i16 addrspace(1)* %out