diff options
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll | 72 |
1 files changed, 36 insertions, 36 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll b/llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll index 1ab63762ecbd..9f3596359a66 100644 --- a/llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll +++ b/llvm/test/CodeGen/AMDGPU/si-annotate-cf.ll @@ -5,22 +5,22 @@ define amdgpu_kernel void @break_inserted_outside_of_loop(ptr addrspace(1) %out, i32 %a) { ; SI-LABEL: break_inserted_outside_of_loop: ; SI: ; %bb.0: ; %main_body -; SI-NEXT: s_load_dword s2, s[0:1], 0xb +; SI-NEXT: s_load_dword s0, s[2:3], 0xb ; SI-NEXT: v_mbcnt_lo_u32_b32_e64 v0, -1, 0 ; SI-NEXT: s_waitcnt lgkmcnt(0) -; SI-NEXT: v_and_b32_e32 v0, s2, v0 +; SI-NEXT: v_and_b32_e32 v0, s0, v0 ; SI-NEXT: v_and_b32_e32 v0, 1, v0 ; SI-NEXT: v_cmp_eq_u32_e32 vcc, 1, v0 -; SI-NEXT: s_mov_b64 s[2:3], 0 +; SI-NEXT: s_mov_b64 s[0:1], 0 ; SI-NEXT: .LBB0_1: ; %ENDIF ; SI-NEXT: ; =>This Inner Loop Header: Depth=1 ; SI-NEXT: s_and_b64 s[4:5], exec, vcc -; SI-NEXT: s_or_b64 s[2:3], s[4:5], s[2:3] -; SI-NEXT: s_andn2_b64 exec, exec, s[2:3] +; SI-NEXT: s_or_b64 s[0:1], s[4:5], s[0:1] +; SI-NEXT: s_andn2_b64 exec, exec, s[0:1] ; SI-NEXT: s_cbranch_execnz .LBB0_1 ; SI-NEXT: ; %bb.2: ; %ENDLOOP -; SI-NEXT: s_or_b64 exec, exec, s[2:3] -; SI-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x9 +; SI-NEXT: s_or_b64 exec, exec, s[0:1] +; SI-NEXT: s_load_dwordx2 s[0:1], s[2:3], 0x9 ; SI-NEXT: s_mov_b32 s3, 0xf000 ; SI-NEXT: s_mov_b32 s2, -1 ; SI-NEXT: v_mov_b32_e32 v0, 0 @@ -30,22 +30,22 @@ define amdgpu_kernel void @break_inserted_outside_of_loop(ptr addrspace(1) %out, ; ; FLAT-LABEL: break_inserted_outside_of_loop: ; FLAT: ; %bb.0: ; %main_body -; FLAT-NEXT: s_load_dword s2, s[0:1], 0x2c +; FLAT-NEXT: s_load_dword s0, s[2:3], 0x2c ; FLAT-NEXT: v_mbcnt_lo_u32_b32 v0, -1, 0 ; FLAT-NEXT: s_waitcnt lgkmcnt(0) -; FLAT-NEXT: v_and_b32_e32 v0, s2, v0 +; FLAT-NEXT: v_and_b32_e32 v0, s0, v0 ; FLAT-NEXT: v_and_b32_e32 v0, 1, v0 ; FLAT-NEXT: v_cmp_eq_u32_e32 vcc, 1, v0 -; FLAT-NEXT: s_mov_b64 s[2:3], 0 +; FLAT-NEXT: s_mov_b64 s[0:1], 0 ; FLAT-NEXT: .LBB0_1: ; %ENDIF ; FLAT-NEXT: ; =>This Inner Loop Header: Depth=1 ; FLAT-NEXT: s_and_b64 s[4:5], exec, vcc -; FLAT-NEXT: s_or_b64 s[2:3], s[4:5], s[2:3] -; FLAT-NEXT: s_andn2_b64 exec, exec, s[2:3] +; FLAT-NEXT: s_or_b64 s[0:1], s[4:5], s[0:1] +; FLAT-NEXT: s_andn2_b64 exec, exec, s[0:1] ; FLAT-NEXT: s_cbranch_execnz .LBB0_1 ; FLAT-NEXT: ; %bb.2: ; %ENDLOOP -; FLAT-NEXT: s_or_b64 exec, exec, s[2:3] -; FLAT-NEXT: s_load_dwordx2 s[0:1], s[0:1], 0x24 +; FLAT-NEXT: s_or_b64 exec, exec, s[0:1] +; FLAT-NEXT: s_load_dwordx2 s[0:1], s[2:3], 0x24 ; FLAT-NEXT: s_mov_b32 s3, 0xf000 ; FLAT-NEXT: s_mov_b32 s2, -1 ; FLAT-NEXT: v_mov_b32_e32 v0, 0 @@ -71,23 +71,23 @@ define amdgpu_kernel void @phi_cond_outside_loop(i32 %b) { ; SI: ; %bb.0: ; %entry ; SI-NEXT: v_mbcnt_lo_u32_b32_e64 v0, -1, 0 ; SI-NEXT: v_cmp_ne_u32_e32 vcc, 0, v0 -; SI-NEXT: s_mov_b64 s[2:3], 0 +; SI-NEXT: s_mov_b64 s[0:1], 0 ; SI-NEXT: s_mov_b64 s[4:5], 0 ; SI-NEXT: s_and_saveexec_b64 s[6:7], vcc ; SI-NEXT: s_cbranch_execz .LBB1_2 ; SI-NEXT: ; %bb.1: ; %else -; SI-NEXT: s_load_dword s0, s[0:1], 0x9 +; SI-NEXT: s_load_dword s2, s[2:3], 0x9 ; SI-NEXT: s_waitcnt lgkmcnt(0) -; SI-NEXT: s_cmp_eq_u32 s0, 0 -; SI-NEXT: s_cselect_b64 s[0:1], -1, 0 -; SI-NEXT: s_and_b64 s[4:5], s[0:1], exec +; SI-NEXT: s_cmp_eq_u32 s2, 0 +; SI-NEXT: s_cselect_b64 s[2:3], -1, 0 +; SI-NEXT: s_and_b64 s[4:5], s[2:3], exec ; SI-NEXT: .LBB1_2: ; %endif ; SI-NEXT: s_or_b64 exec, exec, s[6:7] ; SI-NEXT: .LBB1_3: ; %loop ; SI-NEXT: ; =>This Inner Loop Header: Depth=1 -; SI-NEXT: s_and_b64 s[0:1], exec, s[4:5] -; SI-NEXT: s_or_b64 s[2:3], s[0:1], s[2:3] -; SI-NEXT: s_andn2_b64 exec, exec, s[2:3] +; SI-NEXT: s_and_b64 s[2:3], exec, s[4:5] +; SI-NEXT: s_or_b64 s[0:1], s[2:3], s[0:1] +; SI-NEXT: s_andn2_b64 exec, exec, s[0:1] ; SI-NEXT: s_cbranch_execnz .LBB1_3 ; SI-NEXT: ; %bb.4: ; %exit ; SI-NEXT: s_endpgm @@ -96,23 +96,23 @@ define amdgpu_kernel void @phi_cond_outside_loop(i32 %b) { ; FLAT: ; %bb.0: ; %entry ; FLAT-NEXT: v_mbcnt_lo_u32_b32 v0, -1, 0 ; FLAT-NEXT: v_cmp_ne_u32_e32 vcc, 0, v0 -; FLAT-NEXT: s_mov_b64 s[2:3], 0 +; FLAT-NEXT: s_mov_b64 s[0:1], 0 ; FLAT-NEXT: s_mov_b64 s[4:5], 0 ; FLAT-NEXT: s_and_saveexec_b64 s[6:7], vcc ; FLAT-NEXT: s_cbranch_execz .LBB1_2 ; FLAT-NEXT: ; %bb.1: ; %else -; FLAT-NEXT: s_load_dword s0, s[0:1], 0x24 +; FLAT-NEXT: s_load_dword s2, s[2:3], 0x24 ; FLAT-NEXT: s_waitcnt lgkmcnt(0) -; FLAT-NEXT: s_cmp_eq_u32 s0, 0 -; FLAT-NEXT: s_cselect_b64 s[0:1], -1, 0 -; FLAT-NEXT: s_and_b64 s[4:5], s[0:1], exec +; FLAT-NEXT: s_cmp_eq_u32 s2, 0 +; FLAT-NEXT: s_cselect_b64 s[2:3], -1, 0 +; FLAT-NEXT: s_and_b64 s[4:5], s[2:3], exec ; FLAT-NEXT: .LBB1_2: ; %endif ; FLAT-NEXT: s_or_b64 exec, exec, s[6:7] ; FLAT-NEXT: .LBB1_3: ; %loop ; FLAT-NEXT: ; =>This Inner Loop Header: Depth=1 -; FLAT-NEXT: s_and_b64 s[0:1], exec, s[4:5] -; FLAT-NEXT: s_or_b64 s[2:3], s[0:1], s[2:3] -; FLAT-NEXT: s_andn2_b64 exec, exec, s[2:3] +; FLAT-NEXT: s_and_b64 s[2:3], exec, s[4:5] +; FLAT-NEXT: s_or_b64 s[0:1], s[2:3], s[0:1] +; FLAT-NEXT: s_andn2_b64 exec, exec, s[0:1] ; FLAT-NEXT: s_cbranch_execnz .LBB1_3 ; FLAT-NEXT: ; %bb.4: ; %exit ; FLAT-NEXT: s_endpgm @@ -166,12 +166,12 @@ declare float @llvm.fabs.f32(float) nounwind readnone define amdgpu_kernel void @loop_land_info_assert(i32 %c0, i32 %c1, i32 %c2, i32 %c3, i32 %x, i32 %y, i1 %arg) nounwind { ; SI-LABEL: loop_land_info_assert: ; SI: ; %bb.0: ; %entry -; SI-NEXT: s_load_dword s2, s[0:1], 0xa +; SI-NEXT: s_load_dword s0, s[2:3], 0xa ; SI-NEXT: s_waitcnt lgkmcnt(0) -; SI-NEXT: s_cmp_lt_i32 s2, 4 +; SI-NEXT: s_cmp_lt_i32 s0, 4 ; SI-NEXT: s_cbranch_scc1 .LBB3_4 ; SI-NEXT: ; %bb.1: ; %for.cond.preheader -; SI-NEXT: s_load_dword s0, s[0:1], 0xc +; SI-NEXT: s_load_dword s0, s[2:3], 0xc ; SI-NEXT: s_waitcnt lgkmcnt(0) ; SI-NEXT: s_cmpk_lt_i32 s0, 0x3e8 ; SI-NEXT: s_cbranch_scc0 .LBB3_4 @@ -186,12 +186,12 @@ define amdgpu_kernel void @loop_land_info_assert(i32 %c0, i32 %c1, i32 %c2, i32 ; ; FLAT-LABEL: loop_land_info_assert: ; FLAT: ; %bb.0: ; %entry -; FLAT-NEXT: s_load_dword s2, s[0:1], 0x28 +; FLAT-NEXT: s_load_dword s0, s[2:3], 0x28 ; FLAT-NEXT: s_waitcnt lgkmcnt(0) -; FLAT-NEXT: s_cmp_lt_i32 s2, 4 +; FLAT-NEXT: s_cmp_lt_i32 s0, 4 ; FLAT-NEXT: s_cbranch_scc1 .LBB3_4 ; FLAT-NEXT: ; %bb.1: ; %for.cond.preheader -; FLAT-NEXT: s_load_dword s0, s[0:1], 0x30 +; FLAT-NEXT: s_load_dword s0, s[2:3], 0x30 ; FLAT-NEXT: s_waitcnt lgkmcnt(0) ; FLAT-NEXT: s_cmpk_lt_i32 s0, 0x3e8 ; FLAT-NEXT: s_cbranch_scc0 .LBB3_4 |
