diff options
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll | 205 |
1 files changed, 115 insertions, 90 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll b/llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll index 0cb2b0b7df3d..6484c2f82ff9 100644 --- a/llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll +++ b/llvm/test/CodeGen/AMDGPU/fp64-atomics-gfx90a.ll @@ -24,9 +24,9 @@ define amdgpu_kernel void @raw_buffer_atomic_add_noret_f64(<4 x i32> %rsrc, doub ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_buffer_atomic_add_noret_f64: @@ -35,9 +35,9 @@ define amdgpu_kernel void @raw_buffer_atomic_add_noret_f64(<4 x i32> %rsrc, doub ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_buffer_atomic_add_noret_f64: @@ -74,7 +74,7 @@ define amdgpu_ps void @raw_buffer_atomic_add_rtn_f64(<4 x i32> inreg %rsrc, doub ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0) @@ -143,9 +143,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_add_noret_f64(ptr addrspace(8) ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_ptr_buffer_atomic_add_noret_f64: @@ -154,9 +154,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_add_noret_f64(ptr addrspace(8) ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_ptr_buffer_atomic_add_noret_f64: @@ -193,7 +193,7 @@ define amdgpu_ps void @raw_ptr_buffer_atomic_add_rtn_f64(ptr addrspace(8) inreg ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0) @@ -262,9 +262,9 @@ define amdgpu_kernel void @struct_buffer_atomic_add_noret_f64(<4 x i32> %rsrc, d ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_buffer_atomic_add_noret_f64: @@ -273,9 +273,9 @@ define amdgpu_kernel void @struct_buffer_atomic_add_noret_f64(<4 x i32> %rsrc, d ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_buffer_atomic_add_noret_f64: @@ -312,7 +312,7 @@ define amdgpu_ps void @struct_buffer_atomic_add_rtn_f64(<4 x i32> inreg %rsrc, d ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -380,9 +380,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_add_noret_f64(ptr addrspace( ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_ptr_buffer_atomic_add_noret_f64: @@ -391,9 +391,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_add_noret_f64(ptr addrspace( ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_add_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_ptr_buffer_atomic_add_noret_f64: @@ -430,7 +430,7 @@ define amdgpu_ps void @struct_ptr_buffer_atomic_add_rtn_f64(ptr addrspace(8) inr ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_add_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -498,9 +498,9 @@ define amdgpu_kernel void @raw_buffer_atomic_min_noret_f64(<4 x i32> %rsrc, doub ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_buffer_atomic_min_noret_f64: @@ -509,9 +509,9 @@ define amdgpu_kernel void @raw_buffer_atomic_min_noret_f64(<4 x i32> %rsrc, doub ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_buffer_atomic_min_noret_f64: @@ -548,7 +548,7 @@ define amdgpu_ps void @raw_buffer_atomic_min_rtn_f64(<4 x i32> inreg %rsrc, doub ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_min_num_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.buffer.atomic.fmin.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0) @@ -617,9 +617,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_min_noret_f64(ptr addrspace(8) ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_ptr_buffer_atomic_min_noret_f64: @@ -628,9 +628,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_min_noret_f64(ptr addrspace(8) ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_ptr_buffer_atomic_min_noret_f64: @@ -667,7 +667,7 @@ define amdgpu_ps void @raw_ptr_buffer_atomic_min_rtn_f64(ptr addrspace(8) inreg ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_min_num_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fmin.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0) @@ -736,9 +736,9 @@ define amdgpu_kernel void @struct_buffer_atomic_min_noret_f64(<4 x i32> %rsrc, d ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_buffer_atomic_min_noret_f64: @@ -747,9 +747,9 @@ define amdgpu_kernel void @struct_buffer_atomic_min_noret_f64(<4 x i32> %rsrc, d ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_buffer_atomic_min_noret_f64: @@ -786,7 +786,7 @@ define amdgpu_ps void @struct_buffer_atomic_min_rtn_f64(<4 x i32> inreg %rsrc, d ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_min_num_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.buffer.atomic.fmin.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -854,9 +854,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_min_noret_f64(ptr addrspace( ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_ptr_buffer_atomic_min_noret_f64: @@ -865,9 +865,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_min_noret_f64(ptr addrspace( ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_min_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_min_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_ptr_buffer_atomic_min_noret_f64: @@ -904,7 +904,7 @@ define amdgpu_ps void @struct_ptr_buffer_atomic_min_rtn_f64(ptr addrspace(8) inr ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_min_num_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fmin.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -972,9 +972,9 @@ define amdgpu_kernel void @raw_buffer_atomic_max_noret_f64(<4 x i32> %rsrc, doub ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_buffer_atomic_max_noret_f64: @@ -983,9 +983,9 @@ define amdgpu_kernel void @raw_buffer_atomic_max_noret_f64(<4 x i32> %rsrc, doub ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_buffer_atomic_max_noret_f64: @@ -1022,7 +1022,7 @@ define amdgpu_ps void @raw_buffer_atomic_max_rtn_f64(<4 x i32> inreg %rsrc, doub ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_max_num_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.buffer.atomic.fmax.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0) @@ -1091,9 +1091,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_max_noret_f64(ptr addrspace(8) ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 offen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 offen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: raw_ptr_buffer_atomic_max_noret_f64: @@ -1102,9 +1102,9 @@ define amdgpu_kernel void @raw_ptr_buffer_atomic_max_noret_f64(ptr addrspace(8) ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 offen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 offen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: raw_ptr_buffer_atomic_max_noret_f64: @@ -1141,7 +1141,7 @@ define amdgpu_ps void @raw_ptr_buffer_atomic_max_rtn_f64(ptr addrspace(8) inreg ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_max_num_f64 v[0:1], v2, s[0:3], null offen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fmax.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0) @@ -1210,9 +1210,9 @@ define amdgpu_kernel void @struct_buffer_atomic_max_noret_f64(<4 x i32> %rsrc, d ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_buffer_atomic_max_noret_f64: @@ -1221,9 +1221,9 @@ define amdgpu_kernel void @struct_buffer_atomic_max_noret_f64(<4 x i32> %rsrc, d ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_buffer_atomic_max_noret_f64: @@ -1260,7 +1260,7 @@ define amdgpu_ps void @struct_buffer_atomic_max_rtn_f64(<4 x i32> inreg %rsrc, d ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_max_num_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.buffer.atomic.fmax.f64(double %data, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -1328,9 +1328,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_max_noret_f64(ptr addrspace( ; GFX90A-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX90A-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[6:7], s[6:7] op_sel:[0,1] -; GFX90A-NEXT: v_mov_b32_e32 v2, s8 -; GFX90A-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[6:7], s[6:7] op_sel:[0,1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s8 +; GFX90A-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX90A-NEXT: s_endpgm ; ; GFX942-LABEL: struct_ptr_buffer_atomic_max_noret_f64: @@ -1339,9 +1339,9 @@ define amdgpu_kernel void @struct_ptr_buffer_atomic_max_noret_f64(ptr addrspace( ; GFX942-NEXT: s_load_dword s8, s[4:5], 0x3c ; GFX942-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x24 ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[6:7] -; GFX942-NEXT: v_mov_b32_e32 v2, s8 -; GFX942-NEXT: buffer_atomic_max_f64 v[0:1], v2, s[0:3], 0 idxen +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[6:7] +; GFX942-NEXT: v_mov_b32_e32 v0, s8 +; GFX942-NEXT: buffer_atomic_max_f64 v[2:3], v0, s[0:3], 0 idxen ; GFX942-NEXT: s_endpgm ; ; GFX1250-LABEL: struct_ptr_buffer_atomic_max_noret_f64: @@ -1378,7 +1378,7 @@ define amdgpu_ps void @struct_ptr_buffer_atomic_max_rtn_f64(ptr addrspace(8) inr ; GFX1250: ; %bb.0: ; %main_body ; GFX1250-NEXT: buffer_atomic_max_num_f64 v[0:1], v2, s[0:3], null idxen th:TH_ATOMIC_RETURN ; GFX1250-NEXT: s_wait_loadcnt 0x0 -; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] scope:SCOPE_SE +; GFX1250-NEXT: flat_store_b64 v[0:1], v[0:1] ; GFX1250-NEXT: s_endpgm main_body: %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fmax.f64(double %data, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 0, i32 0) @@ -1473,6 +1473,7 @@ define amdgpu_kernel void @global_atomic_fadd_f64_noret_pat(ptr addrspace(1) %pt ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 ; GFX1250-NEXT: global_wb scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1513,6 +1514,9 @@ define amdgpu_kernel void @global_atomic_fadd_f64_noret_pat_agent(ptr addrspace( ; GFX1250-NEXT: s_load_b64 s[0:1], s[4:5], 0x24 ; GFX1250-NEXT: v_mov_b64_e32 v[0:1], 4.0 ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV +; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1557,6 +1561,7 @@ define amdgpu_kernel void @global_atomic_fadd_f64_noret_pat_system(ptr addrspace ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 ; GFX1250-NEXT: global_wb scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1597,6 +1602,9 @@ define amdgpu_kernel void @global_atomic_fadd_f64_noret_pat_flush(ptr addrspace( ; GFX1250-NEXT: s_load_b64 s[0:1], s[4:5], 0x24 ; GFX1250-NEXT: v_mov_b64_e32 v[0:1], 4.0 ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV +; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1673,6 +1681,7 @@ define double @global_atomic_fadd_f64_rtn_pat_agent(ptr addrspace(1) %ptr, doubl ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_mov_b64_e32 v[2:3], 4.0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v[0:1], v[0:1], v[2:3], off th:TH_ATOMIC_RETURN scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_loadcnt 0x0 @@ -1765,6 +1774,9 @@ define amdgpu_kernel void @global_atomic_fadd_f64_noret_pat_agent_safe(ptr addrs ; GFX1250-NEXT: s_load_b64 s[0:1], s[4:5], 0x24 ; GFX1250-NEXT: v_mov_b64_e32 v[0:1], 4.0 ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV +; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: global_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1809,6 +1821,7 @@ define amdgpu_kernel void @flat_atomic_fadd_f64_noret_pat(ptr %ptr) #1 { ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 ; GFX1250-NEXT: global_wb scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: flat_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt_dscnt 0x0 @@ -1849,6 +1862,9 @@ define amdgpu_kernel void @flat_atomic_fadd_f64_noret_pat_agent(ptr %ptr) #1 { ; GFX1250-NEXT: s_load_b64 s[0:1], s[4:5], 0x24 ; GFX1250-NEXT: v_mov_b64_e32 v[0:1], 4.0 ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV +; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: flat_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt_dscnt 0x0 @@ -1893,6 +1909,7 @@ define amdgpu_kernel void @flat_atomic_fadd_f64_noret_pat_system(ptr %ptr) #1 { ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 ; GFX1250-NEXT: global_wb scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: flat_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_SYS ; GFX1250-NEXT: s_wait_storecnt 0x0 @@ -1969,6 +1986,7 @@ define double @flat_atomic_fadd_f64_rtn_pat_agent(ptr %ptr) #1 { ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_mov_b64_e32 v[2:3], 4.0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: flat_atomic_add_f64 v[0:1], v[0:1], v[2:3] th:TH_ATOMIC_RETURN scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 @@ -2063,6 +2081,9 @@ define amdgpu_kernel void @flat_atomic_fadd_f64_noret_pat_agent_safe(ptr %ptr) { ; GFX1250-NEXT: s_load_b64 s[0:1], s[4:5], 0x24 ; GFX1250-NEXT: v_mov_b64_e32 v[0:1], 4.0 ; GFX1250-NEXT: v_mov_b32_e32 v2, 0 +; GFX1250-NEXT: global_wb scope:SCOPE_DEV +; GFX1250-NEXT: s_wait_storecnt 0x0 +; GFX1250-NEXT: s_wait_xcnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: flat_atomic_add_f64 v2, v[0:1], s[0:1] scope:SCOPE_DEV ; GFX1250-NEXT: s_wait_storecnt_dscnt 0x0 @@ -2079,9 +2100,9 @@ define amdgpu_kernel void @local_atomic_fadd_f64_noret(ptr addrspace(3) %ptr, do ; GFX90A-NEXT: s_load_dword s2, s[4:5], 0x24 ; GFX90A-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x2c ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) -; GFX90A-NEXT: v_mov_b32_e32 v2, s2 -; GFX90A-NEXT: v_pk_mov_b32 v[0:1], s[0:1], s[0:1] op_sel:[0,1] -; GFX90A-NEXT: ds_add_f64 v2, v[0:1] +; GFX90A-NEXT: v_mov_b32_e32 v0, s2 +; GFX90A-NEXT: v_pk_mov_b32 v[2:3], s[0:1], s[0:1] op_sel:[0,1] +; GFX90A-NEXT: ds_add_f64 v0, v[2:3] ; GFX90A-NEXT: s_waitcnt lgkmcnt(0) ; GFX90A-NEXT: s_endpgm ; @@ -2090,9 +2111,9 @@ define amdgpu_kernel void @local_atomic_fadd_f64_noret(ptr addrspace(3) %ptr, do ; GFX942-NEXT: s_load_dword s2, s[4:5], 0x24 ; GFX942-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x2c ; GFX942-NEXT: s_waitcnt lgkmcnt(0) -; GFX942-NEXT: v_mov_b32_e32 v2, s2 -; GFX942-NEXT: v_mov_b64_e32 v[0:1], s[0:1] -; GFX942-NEXT: ds_add_f64 v2, v[0:1] +; GFX942-NEXT: v_mov_b32_e32 v0, s2 +; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[0:1] +; GFX942-NEXT: ds_add_f64 v0, v[2:3] ; GFX942-NEXT: s_waitcnt lgkmcnt(0) ; GFX942-NEXT: s_endpgm ; @@ -2136,6 +2157,7 @@ define double @local_atomic_fadd_f64_rtn(ptr addrspace(3) %ptr, double %data) { ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_dual_mov_b32 v3, v2 :: v_dual_mov_b32 v2, v1 +; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: ds_add_rtn_f64 v[0:1], v0, v[2:3] ; GFX1250-NEXT: s_wait_dscnt 0x0 ; GFX1250-NEXT: s_set_pc_i64 s[30:31] @@ -2275,6 +2297,7 @@ define double @local_atomic_fadd_f64_rtn_pat(ptr addrspace(3) %ptr, double %data ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_mov_b64_e32 v[2:3], 4.0 +; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: ds_add_rtn_f64 v[0:1], v0, v[2:3] ; GFX1250-NEXT: s_wait_dscnt 0x0 ; GFX1250-NEXT: s_set_pc_i64 s[30:31] @@ -2307,6 +2330,7 @@ define double @local_atomic_fadd_f64_rtn_ieee_unsafe(ptr addrspace(3) %ptr, doub ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_dual_mov_b32 v3, v2 :: v_dual_mov_b32 v2, v1 +; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: ds_add_rtn_f64 v[0:1], v0, v[2:3] ; GFX1250-NEXT: s_wait_dscnt 0x0 ; GFX1250-NEXT: s_set_pc_i64 s[30:31] @@ -2339,6 +2363,7 @@ define double @local_atomic_fadd_f64_rtn_ieee_safe(ptr addrspace(3) %ptr, double ; GFX1250-NEXT: s_wait_loadcnt_dscnt 0x0 ; GFX1250-NEXT: s_wait_kmcnt 0x0 ; GFX1250-NEXT: v_dual_mov_b32 v3, v2 :: v_dual_mov_b32 v2, v1 +; GFX1250-NEXT: s_wait_storecnt 0x0 ; GFX1250-NEXT: ds_add_rtn_f64 v[0:1], v0, v[2:3] ; GFX1250-NEXT: s_wait_dscnt 0x0 ; GFX1250-NEXT: s_set_pc_i64 s[30:31] |