[llvm] b0e8a6f - AMDGPU: Remove volatile from some atomic tests (#223528)
via llvm-commits
llvm-commits at lists.llvm.org
Mon Sep 14 23:49:55 PDT 2026
Author: Matt Arsenault
Date: 2026-09-15T08:49:49+02:00
New Revision: b0e8a6f3d77950082ad13dfc8666a8994188b3fd
URL: https://github.com/llvm/llvm-project/commit/b0e8a6f3d77950082ad13dfc8666a8994188b3fd
DIFF: https://github.com/llvm/llvm-project/commit/b0e8a6f3d77950082ad13dfc8666a8994188b3fd.diff
LOG: AMDGPU: Remove volatile from some atomic tests (#223528)
Added:
Modified:
llvm/test/CodeGen/AMDGPU/flat_atomics_i64.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-agent.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-cluster.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-singlethread.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-system.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-wavefront.ll
llvm/test/CodeGen/AMDGPU/memory-legalizer-private-workgroup.ll
Removed:
################################################################################
diff --git a/llvm/test/CodeGen/AMDGPU/flat_atomics_i64.ll b/llvm/test/CodeGen/AMDGPU/flat_atomics_i64.ll
index 5559c3e84e9b7..bd863dd197c9d 100644
--- a/llvm/test/CodeGen/AMDGPU/flat_atomics_i64.ll
+++ b/llvm/test/CodeGen/AMDGPU/flat_atomics_i64.ll
@@ -143,7 +143,7 @@ define amdgpu_kernel void @atomic_add_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile add ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %gep, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -308,7 +308,7 @@ define amdgpu_kernel void @atomic_add_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile add ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %gep, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -468,7 +468,7 @@ define amdgpu_kernel void @atomic_add_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile add ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %gep, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -639,7 +639,7 @@ define amdgpu_kernel void @atomic_add_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile add ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %gep, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -778,7 +778,7 @@ define amdgpu_kernel void @atomic_add_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB4_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile add ptr %out, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %out, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -937,7 +937,7 @@ define amdgpu_kernel void @atomic_add_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile add ptr %out, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %out, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -1090,7 +1090,7 @@ define amdgpu_kernel void @atomic_add_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile add ptr %ptr, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %ptr, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -1255,7 +1255,7 @@ define amdgpu_kernel void @atomic_add_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile add ptr %ptr, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw add ptr %ptr, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -1397,7 +1397,7 @@ define amdgpu_kernel void @atomic_and_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -1559,7 +1559,7 @@ define amdgpu_kernel void @atomic_and_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -1716,7 +1716,7 @@ define amdgpu_kernel void @atomic_and_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -1884,7 +1884,7 @@ define amdgpu_kernel void @atomic_and_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -2020,7 +2020,7 @@ define amdgpu_kernel void @atomic_and_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB12_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile and ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -2176,7 +2176,7 @@ define amdgpu_kernel void @atomic_and_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile and ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -2326,7 +2326,7 @@ define amdgpu_kernel void @atomic_and_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile and ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -2488,7 +2488,7 @@ define amdgpu_kernel void @atomic_and_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile and ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw and ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -2633,7 +2633,7 @@ define amdgpu_kernel void @atomic_sub_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -2798,7 +2798,7 @@ define amdgpu_kernel void @atomic_sub_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -2958,7 +2958,7 @@ define amdgpu_kernel void @atomic_sub_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -3129,7 +3129,7 @@ define amdgpu_kernel void @atomic_sub_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -3268,7 +3268,7 @@ define amdgpu_kernel void @atomic_sub_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB20_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile sub ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -3427,7 +3427,7 @@ define amdgpu_kernel void @atomic_sub_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile sub ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -3580,7 +3580,7 @@ define amdgpu_kernel void @atomic_sub_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile sub ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -3745,7 +3745,7 @@ define amdgpu_kernel void @atomic_sub_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile sub ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw sub ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -3890,7 +3890,7 @@ define amdgpu_kernel void @atomic_max_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -4057,7 +4057,7 @@ define amdgpu_kernel void @atomic_max_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -4217,7 +4217,7 @@ define amdgpu_kernel void @atomic_max_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -4390,7 +4390,7 @@ define amdgpu_kernel void @atomic_max_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -4529,7 +4529,7 @@ define amdgpu_kernel void @atomic_max_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB28_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile max ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -4690,7 +4690,7 @@ define amdgpu_kernel void @atomic_max_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile max ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -4843,7 +4843,7 @@ define amdgpu_kernel void @atomic_max_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile max ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -5010,7 +5010,7 @@ define amdgpu_kernel void @atomic_max_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile max ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw max ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -5155,7 +5155,7 @@ define amdgpu_kernel void @atomic_umax_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -5322,7 +5322,7 @@ define amdgpu_kernel void @atomic_umax_i64_ret_offset(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -5482,7 +5482,7 @@ define amdgpu_kernel void @atomic_umax_i64_addr64_offset(ptr %out, i64 %in, i64
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -5655,7 +5655,7 @@ define amdgpu_kernel void @atomic_umax_i64_ret_addr64_offset(ptr %out, ptr %out2
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -5794,7 +5794,7 @@ define amdgpu_kernel void @atomic_umax_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB36_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile umax ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -5955,7 +5955,7 @@ define amdgpu_kernel void @atomic_umax_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile umax ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -6108,7 +6108,7 @@ define amdgpu_kernel void @atomic_umax_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile umax ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -6275,7 +6275,7 @@ define amdgpu_kernel void @atomic_umax_i64_ret_addr64(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile umax ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umax ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -6420,7 +6420,7 @@ define amdgpu_kernel void @atomic_min_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -6587,7 +6587,7 @@ define amdgpu_kernel void @atomic_min_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -6747,7 +6747,7 @@ define amdgpu_kernel void @atomic_min_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -6920,7 +6920,7 @@ define amdgpu_kernel void @atomic_min_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -7059,7 +7059,7 @@ define amdgpu_kernel void @atomic_min_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB44_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile min ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -7220,7 +7220,7 @@ define amdgpu_kernel void @atomic_min_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile min ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -7373,7 +7373,7 @@ define amdgpu_kernel void @atomic_min_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile min ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -7540,7 +7540,7 @@ define amdgpu_kernel void @atomic_min_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile min ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw min ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -7685,7 +7685,7 @@ define amdgpu_kernel void @atomic_umin_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -7852,7 +7852,7 @@ define amdgpu_kernel void @atomic_umin_i64_ret_offset(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -8012,7 +8012,7 @@ define amdgpu_kernel void @atomic_umin_i64_addr64_offset(ptr %out, i64 %in, i64
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -8185,7 +8185,7 @@ define amdgpu_kernel void @atomic_umin_i64_ret_addr64_offset(ptr %out, ptr %out2
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %gep, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -8324,7 +8324,7 @@ define amdgpu_kernel void @atomic_umin_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB52_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile umin ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -8485,7 +8485,7 @@ define amdgpu_kernel void @atomic_umin_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile umin ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %out, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -8638,7 +8638,7 @@ define amdgpu_kernel void @atomic_umin_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile umin ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -8805,7 +8805,7 @@ define amdgpu_kernel void @atomic_umin_i64_ret_addr64(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile umin ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw umin ptr %ptr, i64 %in syncscope("workgroup") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -8947,7 +8947,7 @@ define amdgpu_kernel void @atomic_or_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -9109,7 +9109,7 @@ define amdgpu_kernel void @atomic_or_i64_ret_offset(ptr %out, ptr %out2, i64 %in
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -9266,7 +9266,7 @@ define amdgpu_kernel void @atomic_or_i64_addr64_offset(ptr %out, i64 %in, i64 %i
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -9434,7 +9434,7 @@ define amdgpu_kernel void @atomic_or_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -9570,7 +9570,7 @@ define amdgpu_kernel void @atomic_or_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB60_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile or ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -9726,7 +9726,7 @@ define amdgpu_kernel void @atomic_or_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile or ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -9876,7 +9876,7 @@ define amdgpu_kernel void @atomic_or_i64_addr64(ptr %out, i64 %in, i64 %index) {
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile or ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -10038,7 +10038,7 @@ define amdgpu_kernel void @atomic_or_i64_ret_addr64(ptr %out, ptr %out2, i64 %in
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile or ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw or ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -10169,7 +10169,7 @@ define amdgpu_kernel void @atomic_xchg_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile xchg ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %gep, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -10299,7 +10299,7 @@ define amdgpu_kernel void @atomic_xchg_f64_offset(ptr %out, double %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds double, ptr %out, i64 4
- %tmp0 = atomicrmw volatile xchg ptr %gep, double %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %gep, double %in syncscope("agent") seq_cst
ret void
}
@@ -10429,7 +10429,7 @@ define amdgpu_kernel void @atomic_xchg_pointer_offset(ptr %out, ptr %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds ptr, ptr %out, i32 4
- %val = atomicrmw volatile xchg ptr %gep, ptr %in syncscope("agent") seq_cst
+ %val = atomicrmw xchg ptr %gep, ptr %in syncscope("agent") seq_cst
ret void
}
@@ -10588,7 +10588,7 @@ define amdgpu_kernel void @atomic_xchg_i64_ret_offset(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile xchg ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %gep, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -10734,7 +10734,7 @@ define amdgpu_kernel void @atomic_xchg_i64_addr64_offset(ptr %out, i64 %in, i64
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile xchg ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %gep, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -10899,7 +10899,7 @@ define amdgpu_kernel void @atomic_xchg_i64_ret_addr64_offset(ptr %out, ptr %out2
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile xchg ptr %gep, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %gep, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -11024,7 +11024,7 @@ define amdgpu_kernel void @atomic_xchg_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB70_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile xchg ptr %out, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %out, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -11177,7 +11177,7 @@ define amdgpu_kernel void @atomic_xchg_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile xchg ptr %out, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %out, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -11316,7 +11316,7 @@ define amdgpu_kernel void @atomic_xchg_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile xchg ptr %ptr, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %ptr, i64 %in syncscope("agent") seq_cst
ret void
}
@@ -11475,7 +11475,7 @@ define amdgpu_kernel void @atomic_xchg_i64_ret_addr64(ptr %out, ptr %out2, i64 %
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile xchg ptr %ptr, i64 %in syncscope("agent") seq_cst
+ %tmp0 = atomicrmw xchg ptr %ptr, i64 %in syncscope("agent") seq_cst
store i64 %tmp0, ptr %out2
ret void
}
@@ -11617,7 +11617,7 @@ define amdgpu_kernel void @atomic_xor_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -11779,7 +11779,7 @@ define amdgpu_kernel void @atomic_xor_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -11936,7 +11936,7 @@ define amdgpu_kernel void @atomic_xor_i64_addr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -12104,7 +12104,7 @@ define amdgpu_kernel void @atomic_xor_i64_ret_addr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -12240,7 +12240,7 @@ define amdgpu_kernel void @atomic_xor_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB78_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile xor ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -12396,7 +12396,7 @@ define amdgpu_kernel void @atomic_xor_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile xor ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -12546,7 +12546,7 @@ define amdgpu_kernel void @atomic_xor_i64_addr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile xor ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -12708,7 +12708,7 @@ define amdgpu_kernel void @atomic_xor_i64_ret_addr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile xor ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw xor ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -13276,7 +13276,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_offset(ptr %out, i64 %in, i64 %old
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %val = cmpxchg volatile ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -13431,7 +13431,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_soffset(ptr %out, i64 %in, i64 %ol
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 9000
- %val = cmpxchg volatile ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -13599,7 +13599,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_ret_offset(ptr %out, ptr %out2, i6
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %val = cmpxchg volatile ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
%extract0 = extractvalue { i64, i1 } %val, 0
store i64 %extract0, ptr %out2
ret void
@@ -13762,7 +13762,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_addr64_offset(ptr %out, i64 %in, i
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %val = cmpxchg volatile ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -13944,7 +13944,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_ret_addr64_offset(ptr %out, ptr %o
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %val = cmpxchg volatile ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %gep, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
%extract0 = extractvalue { i64, i1 } %val, 0
store i64 %extract0, ptr %out2
ret void
@@ -14095,7 +14095,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64(ptr %out, i64 %in, i64 %old) {
; GFX12-NEXT: .LBB95_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %val = cmpxchg volatile ptr %out, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %out, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -14257,7 +14257,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_ret(ptr %out, ptr %out2, i64 %in,
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %val = cmpxchg volatile ptr %out, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %out, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
%extract0 = extractvalue { i64, i1 } %val, 0
store i64 %extract0, ptr %out2
ret void
@@ -14414,7 +14414,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_addr64(ptr %out, i64 %in, i64 %ind
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %val = cmpxchg volatile ptr %ptr, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %ptr, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -14590,7 +14590,7 @@ define amdgpu_kernel void @atomic_cmpxchg_i64_ret_addr64(ptr %out, ptr %out2, i6
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %val = cmpxchg volatile ptr %ptr, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr %ptr, i64 %old, i64 %in syncscope("agent") seq_cst seq_cst
%extract0 = extractvalue { i64, i1 } %val, 0
store i64 %extract0, ptr %out2
ret void
@@ -15155,7 +15155,7 @@ define amdgpu_kernel void @atomic_inc_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -15327,7 +15327,7 @@ define amdgpu_kernel void @atomic_inc_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -15494,7 +15494,7 @@ define amdgpu_kernel void @atomic_inc_i64_incr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -15672,7 +15672,7 @@ define amdgpu_kernel void @atomic_inc_i64_ret_incr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -15818,7 +15818,7 @@ define amdgpu_kernel void @atomic_inc_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB111_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile uinc_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -15984,7 +15984,7 @@ define amdgpu_kernel void @atomic_inc_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile uinc_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -16144,7 +16144,7 @@ define amdgpu_kernel void @atomic_inc_i64_incr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile uinc_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -16316,7 +16316,7 @@ define amdgpu_kernel void @atomic_inc_i64_ret_incr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile uinc_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw uinc_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -16474,7 +16474,7 @@ define amdgpu_kernel void @atomic_dec_i64_offset(ptr %out, i64 %in) {
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -16654,7 +16654,7 @@ define amdgpu_kernel void @atomic_dec_i64_ret_offset(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%gep = getelementptr inbounds i64, ptr %out, i64 4
- %tmp0 = atomicrmw volatile udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -16827,7 +16827,7 @@ define amdgpu_kernel void @atomic_dec_i64_decr64_offset(ptr %out, i64 %in, i64 %
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -17013,7 +17013,7 @@ define amdgpu_kernel void @atomic_dec_i64_ret_decr64_offset(ptr %out, ptr %out2,
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
%gep = getelementptr inbounds i64, ptr %ptr, i64 4
- %tmp0 = atomicrmw volatile udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %gep, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -17165,7 +17165,7 @@ define amdgpu_kernel void @atomic_dec_i64(ptr %out, i64 %in) {
; GFX12-NEXT: .LBB119_4: ; %atomicrmw.phi
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile udec_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -17339,7 +17339,7 @@ define amdgpu_kernel void @atomic_dec_i64_ret(ptr %out, ptr %out2, i64 %in) {
; GFX12-NEXT: flat_store_b64 v[2:3], v[0:1]
; GFX12-NEXT: s_endpgm
entry:
- %tmp0 = atomicrmw volatile udec_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %out, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
@@ -17505,7 +17505,7 @@ define amdgpu_kernel void @atomic_dec_i64_decr64(ptr %out, i64 %in, i64 %index)
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile udec_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
ret void
}
@@ -17685,7 +17685,7 @@ define amdgpu_kernel void @atomic_dec_i64_ret_decr64(ptr %out, ptr %out2, i64 %i
; GFX12-NEXT: s_endpgm
entry:
%ptr = getelementptr inbounds i64, ptr %out, i64 %index
- %tmp0 = atomicrmw volatile udec_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
+ %tmp0 = atomicrmw udec_wrap ptr %ptr, i64 %in syncscope("agent") seq_cst, !amdgpu.no.fine.grained.memory !0
store i64 %tmp0, ptr %out2
ret void
}
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-agent.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-agent.ll
index 549c91e3859df..1226707a69432 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-agent.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-agent.ll
@@ -2274,7 +2274,7 @@ define amdgpu_kernel void @private_agent_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") monotonic
ret void
}
@@ -2549,7 +2549,7 @@ define amdgpu_kernel void @private_agent_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acquire
ret void
}
@@ -2826,7 +2826,7 @@ define amdgpu_kernel void @private_agent_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") release
ret void
}
@@ -3106,7 +3106,7 @@ define amdgpu_kernel void @private_agent_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acq_rel
ret void
}
@@ -3386,7 +3386,7 @@ define amdgpu_kernel void @private_agent_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") seq_cst
ret void
}
@@ -3682,7 +3682,7 @@ define amdgpu_kernel void @private_agent_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3984,7 +3984,7 @@ define amdgpu_kernel void @private_agent_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4286,7 +4286,7 @@ define amdgpu_kernel void @private_agent_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4645,7 +4645,7 @@ define amdgpu_kernel void @private_agent_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic monotonic
ret void
}
@@ -5006,7 +5006,7 @@ define amdgpu_kernel void @private_agent_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire monotonic
ret void
}
@@ -5369,7 +5369,7 @@ define amdgpu_kernel void @private_agent_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release monotonic
ret void
}
@@ -5735,7 +5735,7 @@ define amdgpu_kernel void @private_agent_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel monotonic
ret void
}
@@ -6101,7 +6101,7 @@ define amdgpu_kernel void @private_agent_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst monotonic
ret void
}
@@ -6462,7 +6462,7 @@ define amdgpu_kernel void @private_agent_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic acquire
ret void
}
@@ -6823,7 +6823,7 @@ define amdgpu_kernel void @private_agent_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire acquire
ret void
}
@@ -7189,7 +7189,7 @@ define amdgpu_kernel void @private_agent_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release acquire
ret void
}
@@ -7555,7 +7555,7 @@ define amdgpu_kernel void @private_agent_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel acquire
ret void
}
@@ -7921,7 +7921,7 @@ define amdgpu_kernel void @private_agent_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst acquire
ret void
}
@@ -8287,7 +8287,7 @@ define amdgpu_kernel void @private_agent_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic seq_cst
ret void
}
@@ -8653,7 +8653,7 @@ define amdgpu_kernel void @private_agent_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire seq_cst
ret void
}
@@ -9019,7 +9019,7 @@ define amdgpu_kernel void @private_agent_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release seq_cst
ret void
}
@@ -9385,7 +9385,7 @@ define amdgpu_kernel void @private_agent_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel seq_cst
ret void
}
@@ -9751,7 +9751,7 @@ define amdgpu_kernel void @private_agent_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst seq_cst
ret void
}
@@ -10138,7 +10138,7 @@ define amdgpu_kernel void @private_agent_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10529,7 +10529,7 @@ define amdgpu_kernel void @private_agent_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10923,7 +10923,7 @@ define amdgpu_kernel void @private_agent_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11319,7 +11319,7 @@ define amdgpu_kernel void @private_agent_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11715,7 +11715,7 @@ define amdgpu_kernel void @private_agent_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12106,7 +12106,7 @@ define amdgpu_kernel void @private_agent_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12497,7 +12497,7 @@ define amdgpu_kernel void @private_agent_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12893,7 +12893,7 @@ define amdgpu_kernel void @private_agent_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13289,7 +13289,7 @@ define amdgpu_kernel void @private_agent_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13685,7 +13685,7 @@ define amdgpu_kernel void @private_agent_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14081,7 +14081,7 @@ define amdgpu_kernel void @private_agent_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14477,7 +14477,7 @@ define amdgpu_kernel void @private_agent_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14873,7 +14873,7 @@ define amdgpu_kernel void @private_agent_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15269,7 +15269,7 @@ define amdgpu_kernel void @private_agent_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15665,7 +15665,7 @@ define amdgpu_kernel void @private_agent_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -17931,7 +17931,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") monotonic
ret void
}
@@ -18206,7 +18206,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acquire
ret void
}
@@ -18483,7 +18483,7 @@ define amdgpu_kernel void @private_agent_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") release
ret void
}
@@ -18763,7 +18763,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acq_rel
ret void
}
@@ -19043,7 +19043,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") seq_cst
ret void
}
@@ -19339,7 +19339,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19641,7 +19641,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19943,7 +19943,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("agent-one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -20302,7 +20302,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic monotonic
ret void
}
@@ -20663,7 +20663,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire monotonic
ret void
}
@@ -21026,7 +21026,7 @@ define amdgpu_kernel void @private_agent_one_as_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release monotonic
ret void
}
@@ -21392,7 +21392,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel monotonic
ret void
}
@@ -21758,7 +21758,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst monotonic
ret void
}
@@ -22119,7 +22119,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic acquire
ret void
}
@@ -22480,7 +22480,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire acquire
ret void
}
@@ -22846,7 +22846,7 @@ define amdgpu_kernel void @private_agent_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release acquire
ret void
}
@@ -23212,7 +23212,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel acquire
ret void
}
@@ -23578,7 +23578,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst acquire
ret void
}
@@ -23944,7 +23944,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic seq_cst
ret void
}
@@ -24310,7 +24310,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire seq_cst
ret void
}
@@ -24676,7 +24676,7 @@ define amdgpu_kernel void @private_agent_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release seq_cst
ret void
}
@@ -25042,7 +25042,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel seq_cst
ret void
}
@@ -25408,7 +25408,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst seq_cst
ret void
}
@@ -25795,7 +25795,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26186,7 +26186,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26582,7 +26582,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26978,7 +26978,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27369,7 +27369,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27760,7 +27760,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28156,7 +28156,7 @@ define amdgpu_kernel void @private_agent_one_as_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28552,7 +28552,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28948,7 +28948,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29344,7 +29344,7 @@ define amdgpu_kernel void @private_agent_one_as_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29740,7 +29740,7 @@ define amdgpu_kernel void @private_agent_one_as_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30136,7 +30136,7 @@ define amdgpu_kernel void @private_agent_one_as_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30532,7 +30532,7 @@ define amdgpu_kernel void @private_agent_one_as_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30928,7 +30928,7 @@ define amdgpu_kernel void @private_agent_one_as_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("agent-one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-cluster.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-cluster.ll
index ecfb348b0e894..5aac9a888ca88 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-cluster.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-cluster.ll
@@ -2266,7 +2266,7 @@ define amdgpu_kernel void @private_cluster_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") monotonic
ret void
}
@@ -2540,7 +2540,7 @@ define amdgpu_kernel void @private_cluster_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acquire
ret void
}
@@ -2814,7 +2814,7 @@ define amdgpu_kernel void @private_cluster_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") release
ret void
}
@@ -3090,7 +3090,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acq_rel
ret void
}
@@ -3366,7 +3366,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") seq_cst
ret void
}
@@ -3661,7 +3661,7 @@ define amdgpu_kernel void @private_cluster_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3959,7 +3959,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4257,7 +4257,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4616,7 +4616,7 @@ define amdgpu_kernel void @private_cluster_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic monotonic
ret void
}
@@ -4976,7 +4976,7 @@ define amdgpu_kernel void @private_cluster_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire monotonic
ret void
}
@@ -5336,7 +5336,7 @@ define amdgpu_kernel void @private_cluster_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release monotonic
ret void
}
@@ -5698,7 +5698,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel monotonic
ret void
}
@@ -6060,7 +6060,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst monotonic
ret void
}
@@ -6420,7 +6420,7 @@ define amdgpu_kernel void @private_cluster_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic acquire
ret void
}
@@ -6780,7 +6780,7 @@ define amdgpu_kernel void @private_cluster_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire acquire
ret void
}
@@ -7142,7 +7142,7 @@ define amdgpu_kernel void @private_cluster_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release acquire
ret void
}
@@ -7504,7 +7504,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel acquire
ret void
}
@@ -7866,7 +7866,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst acquire
ret void
}
@@ -8228,7 +8228,7 @@ define amdgpu_kernel void @private_cluster_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic seq_cst
ret void
}
@@ -8590,7 +8590,7 @@ define amdgpu_kernel void @private_cluster_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire seq_cst
ret void
}
@@ -8952,7 +8952,7 @@ define amdgpu_kernel void @private_cluster_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release seq_cst
ret void
}
@@ -9314,7 +9314,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel seq_cst
ret void
}
@@ -9676,7 +9676,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst seq_cst
ret void
}
@@ -10063,7 +10063,7 @@ define amdgpu_kernel void @private_cluster_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10453,7 +10453,7 @@ define amdgpu_kernel void @private_cluster_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10844,7 +10844,7 @@ define amdgpu_kernel void @private_cluster_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11236,7 +11236,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11628,7 +11628,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12018,7 +12018,7 @@ define amdgpu_kernel void @private_cluster_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12408,7 +12408,7 @@ define amdgpu_kernel void @private_cluster_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12800,7 +12800,7 @@ define amdgpu_kernel void @private_cluster_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13192,7 +13192,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13584,7 +13584,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13976,7 +13976,7 @@ define amdgpu_kernel void @private_cluster_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14368,7 +14368,7 @@ define amdgpu_kernel void @private_cluster_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14760,7 +14760,7 @@ define amdgpu_kernel void @private_cluster_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15152,7 +15152,7 @@ define amdgpu_kernel void @private_cluster_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15544,7 +15544,7 @@ define amdgpu_kernel void @private_cluster_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -17804,7 +17804,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") monotonic
ret void
}
@@ -18078,7 +18078,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acquire
ret void
}
@@ -18352,7 +18352,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") release
ret void
}
@@ -18628,7 +18628,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acq_rel
ret void
}
@@ -18904,7 +18904,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") seq_cst
ret void
}
@@ -19200,7 +19200,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19499,7 +19499,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19798,7 +19798,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("cluster-one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -20157,7 +20157,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic monotonic
ret void
}
@@ -20517,7 +20517,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire monotonic
ret void
}
@@ -20877,7 +20877,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release monotonic
ret void
}
@@ -21239,7 +21239,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel monotonic
ret void
}
@@ -21601,7 +21601,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst monotonic
ret void
}
@@ -21961,7 +21961,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic acquire
ret void
}
@@ -22321,7 +22321,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire acquire
ret void
}
@@ -22683,7 +22683,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release acquire
ret void
}
@@ -23045,7 +23045,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel acquire
ret void
}
@@ -23407,7 +23407,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst acquire
ret void
}
@@ -23769,7 +23769,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic seq_cst
ret void
}
@@ -24131,7 +24131,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire seq_cst
ret void
}
@@ -24493,7 +24493,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release seq_cst
ret void
}
@@ -24855,7 +24855,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel seq_cst
ret void
}
@@ -25217,7 +25217,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst seq_cst
ret void
}
@@ -25604,7 +25604,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25995,7 +25995,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26388,7 +26388,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26781,7 +26781,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27172,7 +27172,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27563,7 +27563,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27956,7 +27956,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28349,7 +28349,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28742,7 +28742,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29135,7 +29135,7 @@ define amdgpu_kernel void @private_cluster_one_as_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29528,7 +29528,7 @@ define amdgpu_kernel void @private_cluster_one_as_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29921,7 +29921,7 @@ define amdgpu_kernel void @private_cluster_one_as_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30314,7 +30314,7 @@ define amdgpu_kernel void @private_cluster_one_as_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30707,7 +30707,7 @@ define amdgpu_kernel void @private_cluster_one_as_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("cluster-one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-singlethread.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-singlethread.ll
index 0b15aacbf3727..5b9b61365b9a7 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-singlethread.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-singlethread.ll
@@ -2258,7 +2258,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") monotonic
ret void
}
@@ -2530,7 +2530,7 @@ define amdgpu_kernel void @private_singlethread_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acquire
ret void
}
@@ -2802,7 +2802,7 @@ define amdgpu_kernel void @private_singlethread_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") release
ret void
}
@@ -3074,7 +3074,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acq_rel
ret void
}
@@ -3346,7 +3346,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") seq_cst
ret void
}
@@ -3640,7 +3640,7 @@ define amdgpu_kernel void @private_singlethread_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3935,7 +3935,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4230,7 +4230,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4589,7 +4589,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic monotonic
ret void
}
@@ -4947,7 +4947,7 @@ define amdgpu_kernel void @private_singlethread_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire monotonic
ret void
}
@@ -5305,7 +5305,7 @@ define amdgpu_kernel void @private_singlethread_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release monotonic
ret void
}
@@ -5663,7 +5663,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel monotonic
ret void
}
@@ -6021,7 +6021,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst monotonic
ret void
}
@@ -6379,7 +6379,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic acquire
ret void
}
@@ -6737,7 +6737,7 @@ define amdgpu_kernel void @private_singlethread_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire acquire
ret void
}
@@ -7095,7 +7095,7 @@ define amdgpu_kernel void @private_singlethread_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release acquire
ret void
}
@@ -7453,7 +7453,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel acquire
ret void
}
@@ -7811,7 +7811,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst acquire
ret void
}
@@ -8169,7 +8169,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic seq_cst
ret void
}
@@ -8527,7 +8527,7 @@ define amdgpu_kernel void @private_singlethread_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire seq_cst
ret void
}
@@ -8885,7 +8885,7 @@ define amdgpu_kernel void @private_singlethread_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release seq_cst
ret void
}
@@ -9243,7 +9243,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel seq_cst
ret void
}
@@ -9601,7 +9601,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst seq_cst
ret void
}
@@ -9988,7 +9988,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10377,7 +10377,7 @@ define amdgpu_kernel void @private_singlethread_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10766,7 +10766,7 @@ define amdgpu_kernel void @private_singlethread_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11155,7 +11155,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11544,7 +11544,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11933,7 +11933,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12322,7 +12322,7 @@ define amdgpu_kernel void @private_singlethread_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12711,7 +12711,7 @@ define amdgpu_kernel void @private_singlethread_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13100,7 +13100,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13489,7 +13489,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13878,7 +13878,7 @@ define amdgpu_kernel void @private_singlethread_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14267,7 +14267,7 @@ define amdgpu_kernel void @private_singlethread_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14656,7 +14656,7 @@ define amdgpu_kernel void @private_singlethread_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15045,7 +15045,7 @@ define amdgpu_kernel void @private_singlethread_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15434,7 +15434,7 @@ define amdgpu_kernel void @private_singlethread_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -17684,7 +17684,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") monotonic
ret void
}
@@ -17956,7 +17956,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acquire
ret void
}
@@ -18228,7 +18228,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") release
ret void
}
@@ -18500,7 +18500,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acq_rel
ret void
}
@@ -18772,7 +18772,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") seq_cst
ret void
}
@@ -19066,7 +19066,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19361,7 +19361,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19656,7 +19656,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("singlethread-one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -20015,7 +20015,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_monotonic_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic monotonic
ret void
}
@@ -20373,7 +20373,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_monotonic_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire monotonic
ret void
}
@@ -20731,7 +20731,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_monotonic_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release monotonic
ret void
}
@@ -21089,7 +21089,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_monotonic_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel monotonic
ret void
}
@@ -21447,7 +21447,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_monotonic_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst monotonic
ret void
}
@@ -21805,7 +21805,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_acquire_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic acquire
ret void
}
@@ -22163,7 +22163,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire acquire
ret void
}
@@ -22521,7 +22521,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release acquire
ret void
}
@@ -22879,7 +22879,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel acquire
ret void
}
@@ -23237,7 +23237,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst acquire
ret void
}
@@ -23595,7 +23595,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_seq_cst_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic seq_cst
ret void
}
@@ -23953,7 +23953,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire seq_cst
ret void
}
@@ -24311,7 +24311,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release seq_cst
ret void
}
@@ -24669,7 +24669,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel seq_cst
ret void
}
@@ -25027,7 +25027,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst seq_cst
ret void
}
@@ -25414,7 +25414,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_monotonic_ret_c
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25803,7 +25803,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_monotonic_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26192,7 +26192,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_monotonic_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26581,7 +26581,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_monotonic_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26970,7 +26970,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_monotonic_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27359,7 +27359,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_acquire_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27748,7 +27748,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_acquire_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28137,7 +28137,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_acquire_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28526,7 +28526,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_acquire_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28915,7 +28915,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_acquire_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29304,7 +29304,7 @@ define amdgpu_kernel void @private_singlethread_one_as_monotonic_seq_cst_ret_cmp
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29693,7 +29693,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acquire_seq_cst_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30082,7 +30082,7 @@ define amdgpu_kernel void @private_singlethread_one_as_release_seq_cst_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30471,7 +30471,7 @@ define amdgpu_kernel void @private_singlethread_one_as_acq_rel_seq_cst_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30860,7 +30860,7 @@ define amdgpu_kernel void @private_singlethread_one_as_seq_cst_seq_cst_ret_cmpxc
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("singlethread-one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-system.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-system.ll
index 4cc3625fe8d54..e2772a4480061 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-system.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-system.ll
@@ -2274,7 +2274,7 @@ define amdgpu_kernel void @private_system_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in monotonic
ret void
}
@@ -2549,7 +2549,7 @@ define amdgpu_kernel void @private_system_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in acquire
ret void
}
@@ -2826,7 +2826,7 @@ define amdgpu_kernel void @private_system_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in release
ret void
}
@@ -3106,7 +3106,7 @@ define amdgpu_kernel void @private_system_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in acq_rel
ret void
}
@@ -3386,7 +3386,7 @@ define amdgpu_kernel void @private_system_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in seq_cst
ret void
}
@@ -3682,7 +3682,7 @@ define amdgpu_kernel void @private_system_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3984,7 +3984,7 @@ define amdgpu_kernel void @private_system_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4286,7 +4286,7 @@ define amdgpu_kernel void @private_system_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4645,7 +4645,7 @@ define amdgpu_kernel void @private_system_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in monotonic monotonic
ret void
}
@@ -5006,7 +5006,7 @@ define amdgpu_kernel void @private_system_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acquire monotonic
ret void
}
@@ -5369,7 +5369,7 @@ define amdgpu_kernel void @private_system_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in release monotonic
ret void
}
@@ -5735,7 +5735,7 @@ define amdgpu_kernel void @private_system_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel monotonic
ret void
}
@@ -6101,7 +6101,7 @@ define amdgpu_kernel void @private_system_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst monotonic
ret void
}
@@ -6462,7 +6462,7 @@ define amdgpu_kernel void @private_system_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in monotonic acquire
ret void
}
@@ -6823,7 +6823,7 @@ define amdgpu_kernel void @private_system_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acquire acquire
ret void
}
@@ -7189,7 +7189,7 @@ define amdgpu_kernel void @private_system_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in release acquire
ret void
}
@@ -7555,7 +7555,7 @@ define amdgpu_kernel void @private_system_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel acquire
ret void
}
@@ -7921,7 +7921,7 @@ define amdgpu_kernel void @private_system_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst acquire
ret void
}
@@ -8287,7 +8287,7 @@ define amdgpu_kernel void @private_system_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst seq_cst
ret void
}
@@ -8674,7 +8674,7 @@ define amdgpu_kernel void @private_system_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -9065,7 +9065,7 @@ define amdgpu_kernel void @private_system_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -9461,7 +9461,7 @@ define amdgpu_kernel void @private_system_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -9857,7 +9857,7 @@ define amdgpu_kernel void @private_system_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10248,7 +10248,7 @@ define amdgpu_kernel void @private_system_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10639,7 +10639,7 @@ define amdgpu_kernel void @private_system_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11035,7 +11035,7 @@ define amdgpu_kernel void @private_system_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11431,7 +11431,7 @@ define amdgpu_kernel void @private_system_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11827,7 +11827,7 @@ define amdgpu_kernel void @private_system_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12223,7 +12223,7 @@ define amdgpu_kernel void @private_system_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12619,7 +12619,7 @@ define amdgpu_kernel void @private_system_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13015,7 +13015,7 @@ define amdgpu_kernel void @private_system_relese_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13411,7 +13411,7 @@ define amdgpu_kernel void @private_system_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13807,7 +13807,7 @@ define amdgpu_kernel void @private_system_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -16073,7 +16073,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") monotonic
ret void
}
@@ -16348,7 +16348,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acquire
ret void
}
@@ -16625,7 +16625,7 @@ define amdgpu_kernel void @private_system_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") release
ret void
}
@@ -16905,7 +16905,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acq_rel
ret void
}
@@ -17185,7 +17185,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") seq_cst
ret void
}
@@ -17481,7 +17481,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -17783,7 +17783,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -18085,7 +18085,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -18444,7 +18444,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic monotonic
ret void
}
@@ -18805,7 +18805,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire monotonic
ret void
}
@@ -19168,7 +19168,7 @@ define amdgpu_kernel void @private_system_one_as_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release monotonic
ret void
}
@@ -19534,7 +19534,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel monotonic
ret void
}
@@ -19900,7 +19900,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst monotonic
ret void
}
@@ -20261,7 +20261,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic acquire
ret void
}
@@ -20622,7 +20622,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire acquire
ret void
}
@@ -20988,7 +20988,7 @@ define amdgpu_kernel void @private_system_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release acquire
ret void
}
@@ -21354,7 +21354,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel acquire
ret void
}
@@ -21720,7 +21720,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst acquire
ret void
}
@@ -22086,7 +22086,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic seq_cst
ret void
}
@@ -22452,7 +22452,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire seq_cst
ret void
}
@@ -22818,7 +22818,7 @@ define amdgpu_kernel void @private_system_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release seq_cst
ret void
}
@@ -23184,7 +23184,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel seq_cst
ret void
}
@@ -23550,7 +23550,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst seq_cst
ret void
}
@@ -23937,7 +23937,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_monotonic_ret_cmpxchg
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -24328,7 +24328,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -24722,7 +24722,7 @@ define amdgpu_kernel void @private_system_one_as_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25118,7 +25118,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25514,7 +25514,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25905,7 +25905,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26296,7 +26296,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26692,7 +26692,7 @@ define amdgpu_kernel void @private_system_one_as_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27088,7 +27088,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27484,7 +27484,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27880,7 +27880,7 @@ define amdgpu_kernel void @private_system_one_as_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28276,7 +28276,7 @@ define amdgpu_kernel void @private_system_one_as_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28672,7 +28672,7 @@ define amdgpu_kernel void @private_system_one_as_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29068,7 +29068,7 @@ define amdgpu_kernel void @private_system_one_as_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29464,7 +29464,7 @@ define amdgpu_kernel void @private_system_one_as_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-wavefront.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-wavefront.ll
index c9cdf41b26eff..b9a9f62bb3c37 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-wavefront.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-wavefront.ll
@@ -2258,7 +2258,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") monotonic
ret void
}
@@ -2530,7 +2530,7 @@ define amdgpu_kernel void @private_wavefront_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acquire
ret void
}
@@ -2802,7 +2802,7 @@ define amdgpu_kernel void @private_wavefront_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") release
ret void
}
@@ -3074,7 +3074,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acq_rel
ret void
}
@@ -3346,7 +3346,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") seq_cst
ret void
}
@@ -3640,7 +3640,7 @@ define amdgpu_kernel void @private_wavefront_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3935,7 +3935,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4230,7 +4230,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4589,7 +4589,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic monotonic
ret void
}
@@ -4947,7 +4947,7 @@ define amdgpu_kernel void @private_wavefront_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire monotonic
ret void
}
@@ -5305,7 +5305,7 @@ define amdgpu_kernel void @private_wavefront_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release monotonic
ret void
}
@@ -5663,7 +5663,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel monotonic
ret void
}
@@ -6021,7 +6021,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst monotonic
ret void
}
@@ -6379,7 +6379,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic acquire
ret void
}
@@ -6737,7 +6737,7 @@ define amdgpu_kernel void @private_wavefront_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire acquire
ret void
}
@@ -7095,7 +7095,7 @@ define amdgpu_kernel void @private_wavefront_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release acquire
ret void
}
@@ -7453,7 +7453,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel acquire
ret void
}
@@ -7811,7 +7811,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst acquire
ret void
}
@@ -8169,7 +8169,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic seq_cst
ret void
}
@@ -8527,7 +8527,7 @@ define amdgpu_kernel void @private_wavefront_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire seq_cst
ret void
}
@@ -8885,7 +8885,7 @@ define amdgpu_kernel void @private_wavefront_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release seq_cst
ret void
}
@@ -9243,7 +9243,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel seq_cst
ret void
}
@@ -9601,7 +9601,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst seq_cst
ret void
}
@@ -9988,7 +9988,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10377,7 +10377,7 @@ define amdgpu_kernel void @private_wavefront_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10766,7 +10766,7 @@ define amdgpu_kernel void @private_wavefront_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11155,7 +11155,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11544,7 +11544,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11933,7 +11933,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12322,7 +12322,7 @@ define amdgpu_kernel void @private_wavefront_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12711,7 +12711,7 @@ define amdgpu_kernel void @private_wavefront_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13100,7 +13100,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13489,7 +13489,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13878,7 +13878,7 @@ define amdgpu_kernel void @private_wavefront_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14267,7 +14267,7 @@ define amdgpu_kernel void @private_wavefront_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14656,7 +14656,7 @@ define amdgpu_kernel void @private_wavefront_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15045,7 +15045,7 @@ define amdgpu_kernel void @private_wavefront_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15434,7 +15434,7 @@ define amdgpu_kernel void @private_wavefront_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -17684,7 +17684,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") monotonic
ret void
}
@@ -17956,7 +17956,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acquire
ret void
}
@@ -18228,7 +18228,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") release
ret void
}
@@ -18500,7 +18500,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acq_rel
ret void
}
@@ -18772,7 +18772,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") seq_cst
ret void
}
@@ -19066,7 +19066,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19361,7 +19361,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19656,7 +19656,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("wavefront-one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -20015,7 +20015,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic monotonic
ret void
}
@@ -20373,7 +20373,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire monotonic
ret void
}
@@ -20731,7 +20731,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release monotonic
ret void
}
@@ -21089,7 +21089,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel monotonic
ret void
}
@@ -21447,7 +21447,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst monotonic
ret void
}
@@ -21805,7 +21805,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic acquire
ret void
}
@@ -22163,7 +22163,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire acquire
ret void
}
@@ -22521,7 +22521,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release acquire
ret void
}
@@ -22879,7 +22879,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel acquire
ret void
}
@@ -23237,7 +23237,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst acquire
ret void
}
@@ -23595,7 +23595,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic seq_cst
ret void
}
@@ -23953,7 +23953,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire seq_cst
ret void
}
@@ -24311,7 +24311,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release seq_cst
ret void
}
@@ -24669,7 +24669,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel seq_cst
ret void
}
@@ -25027,7 +25027,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst seq_cst
ret void
}
@@ -25414,7 +25414,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_monotonic_ret_cmpx
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25803,7 +25803,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26192,7 +26192,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26581,7 +26581,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26970,7 +26970,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27359,7 +27359,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_acquire_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27748,7 +27748,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28137,7 +28137,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28526,7 +28526,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28915,7 +28915,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29304,7 +29304,7 @@ define amdgpu_kernel void @private_wavefront_one_as_monotonic_seq_cst_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29693,7 +29693,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30082,7 +30082,7 @@ define amdgpu_kernel void @private_wavefront_one_as_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30471,7 +30471,7 @@ define amdgpu_kernel void @private_wavefront_one_as_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30860,7 +30860,7 @@ define amdgpu_kernel void @private_wavefront_one_as_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("wavefront-one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-workgroup.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-workgroup.ll
index a2ac7b88ceddb..86619939e4c50 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-workgroup.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-private-workgroup.ll
@@ -2264,7 +2264,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") monotonic
ret void
}
@@ -2537,7 +2537,7 @@ define amdgpu_kernel void @private_workgroup_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acquire
ret void
}
@@ -2811,7 +2811,7 @@ define amdgpu_kernel void @private_workgroup_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") release
ret void
}
@@ -3086,7 +3086,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acq_rel
ret void
}
@@ -3361,7 +3361,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") seq_cst
ret void
}
@@ -3655,7 +3655,7 @@ define amdgpu_kernel void @private_workgroup_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -3952,7 +3952,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4249,7 +4249,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -4608,7 +4608,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic monotonic
ret void
}
@@ -4967,7 +4967,7 @@ define amdgpu_kernel void @private_workgroup_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire monotonic
ret void
}
@@ -5327,7 +5327,7 @@ define amdgpu_kernel void @private_workgroup_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release monotonic
ret void
}
@@ -5688,7 +5688,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel monotonic
ret void
}
@@ -6049,7 +6049,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst monotonic
ret void
}
@@ -6408,7 +6408,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic acquire
ret void
}
@@ -6767,7 +6767,7 @@ define amdgpu_kernel void @private_workgroup_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire acquire
ret void
}
@@ -7128,7 +7128,7 @@ define amdgpu_kernel void @private_workgroup_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release acquire
ret void
}
@@ -7489,7 +7489,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel acquire
ret void
}
@@ -7850,7 +7850,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst acquire
ret void
}
@@ -8211,7 +8211,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic seq_cst
ret void
}
@@ -8572,7 +8572,7 @@ define amdgpu_kernel void @private_workgroup_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire seq_cst
ret void
}
@@ -8933,7 +8933,7 @@ define amdgpu_kernel void @private_workgroup_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release seq_cst
ret void
}
@@ -9294,7 +9294,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel seq_cst
ret void
}
@@ -9655,7 +9655,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst seq_cst
ret void
}
@@ -10042,7 +10042,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10431,7 +10431,7 @@ define amdgpu_kernel void @private_workgroup_acquire_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -10822,7 +10822,7 @@ define amdgpu_kernel void @private_workgroup_release_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11213,7 +11213,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11604,7 +11604,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_monotonic_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -11993,7 +11993,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12382,7 +12382,7 @@ define amdgpu_kernel void @private_workgroup_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -12773,7 +12773,7 @@ define amdgpu_kernel void @private_workgroup_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13164,7 +13164,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13555,7 +13555,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -13946,7 +13946,7 @@ define amdgpu_kernel void @private_workgroup_monotonic_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14337,7 +14337,7 @@ define amdgpu_kernel void @private_workgroup_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -14728,7 +14728,7 @@ define amdgpu_kernel void @private_workgroup_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15119,7 +15119,7 @@ define amdgpu_kernel void @private_workgroup_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -15510,7 +15510,7 @@ define amdgpu_kernel void @private_workgroup_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -17766,7 +17766,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") monotonic
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") monotonic
ret void
}
@@ -18039,7 +18039,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acquire
ret void
}
@@ -18313,7 +18313,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") release
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") release
ret void
}
@@ -18588,7 +18588,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acq_rel
ret void
}
@@ -18863,7 +18863,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") seq_cst
ret void
}
@@ -19157,7 +19157,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acquire
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acquire
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19454,7 +19454,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acq_rel
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") acq_rel
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -19751,7 +19751,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_ret_atomicrmw(
; GFX1250-NEXT: s_endpgm
ptr addrspace(5) %out, i32 %in) #0 {
entry:
- %val = atomicrmw volatile xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") seq_cst
+ %val = atomicrmw xchg ptr addrspace(5) %out, i32 %in syncscope("workgroup-one-as") seq_cst
store i32 %val, ptr addrspace(5) %out, align 4
ret void
}
@@ -20110,7 +20110,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic monotonic
ret void
}
@@ -20469,7 +20469,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire monotonic
ret void
}
@@ -20829,7 +20829,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release monotonic
ret void
}
@@ -21190,7 +21190,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel monotonic
ret void
}
@@ -21551,7 +21551,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_monotonic_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst monotonic
ret void
}
@@ -21910,7 +21910,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic acquire
ret void
}
@@ -22269,7 +22269,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire acquire
ret void
}
@@ -22630,7 +22630,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release acquire
ret void
}
@@ -22991,7 +22991,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel acquire
ret void
}
@@ -23352,7 +23352,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_acquire_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst acquire
ret void
}
@@ -23713,7 +23713,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic seq_cst
ret void
}
@@ -24074,7 +24074,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire seq_cst
ret void
}
@@ -24435,7 +24435,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release seq_cst
ret void
}
@@ -24796,7 +24796,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel seq_cst
ret void
}
@@ -25157,7 +25157,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_seq_cst_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst seq_cst
ret void
}
@@ -25544,7 +25544,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_monotonic_ret_cmpx
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -25933,7 +25933,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26324,7 +26324,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -26715,7 +26715,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27106,7 +27106,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_monotonic_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst monotonic
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst monotonic
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27495,7 +27495,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_acquire_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -27884,7 +27884,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28275,7 +28275,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -28666,7 +28666,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29057,7 +29057,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_acquire_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst acquire
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst acquire
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29448,7 +29448,7 @@ define amdgpu_kernel void @private_workgroup_one_as_monotonic_seq_cst_ret_cmpxch
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") monotonic seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -29839,7 +29839,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acquire_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acquire seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30230,7 +30230,7 @@ define amdgpu_kernel void @private_workgroup_one_as_release_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") release seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -30621,7 +30621,7 @@ define amdgpu_kernel void @private_workgroup_one_as_acq_rel_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") acq_rel seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
@@ -31012,7 +31012,7 @@ define amdgpu_kernel void @private_workgroup_one_as_seq_cst_seq_cst_ret_cmpxchg(
ptr addrspace(5) %out, i32 %in, i32 %old) #0 {
entry:
%gep = getelementptr i32, ptr addrspace(5) %out, i32 4
- %val = cmpxchg volatile ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst seq_cst
+ %val = cmpxchg ptr addrspace(5) %gep, i32 %old, i32 %in syncscope("workgroup-one-as") seq_cst seq_cst
%val0 = extractvalue { i32, i1 } %val, 0
store i32 %val0, ptr addrspace(5) %out, align 4
ret void
More information about the llvm-commits
mailing list