Rebase, update tests

Author: gandhi56

llvm/test/CodeGen/AMDGPU/agpr-copy-no-free-registers.ll

@@ -525,29 +525,29 @@ define amdgpu_kernel void @introduced_copy_to_sgpr(i64 %arg, i32 %arg1, i32 %arg
 ; GFX908-NEXT:    v_rcp_iflag_f32_e32 v0, v0
 ; GFX908-NEXT:    v_mul_f32_e32 v0, 0x4f7ffffe, v0
 ; GFX908-NEXT:    v_cvt_u32_f32_e32 v0, v0
-; GFX908-NEXT:    v_readfirstlane_b32 s2, v0
-; GFX908-NEXT:    s_mul_i32 s1, s1, s2
-; GFX908-NEXT:    s_mul_hi_u32 s1, s2, s1
-; GFX908-NEXT:    s_add_i32 s2, s2, s1
-; GFX908-NEXT:    s_mul_hi_u32 s1, s6, s2
-; GFX908-NEXT:    s_mul_i32 s2, s1, s7
-; GFX908-NEXT:    s_sub_i32 s2, s6, s2
-; GFX908-NEXT:    s_add_i32 s3, s1, 1
-; GFX908-NEXT:    s_sub_i32 s6, s2, s7
-; GFX908-NEXT:    s_cmp_ge_u32 s2, s7
-; GFX908-NEXT:    s_cselect_b32 s1, s3, s1
-; GFX908-NEXT:    s_cselect_b32 s2, s6, s2
-; GFX908-NEXT:    s_add_i32 s3, s1, 1
-; GFX908-NEXT:    s_cmp_ge_u32 s2, s7
-; GFX908-NEXT:    s_cselect_b32 s8, s3, s1
-; GFX908-NEXT:    s_lshr_b32 s2, s0, 16
-; GFX908-NEXT:    v_cvt_f32_f16_e32 v19, s2
-; GFX908-NEXT:    s_lshl_b64 s[6:7], s[4:5], 5
-; GFX908-NEXT:    s_lshl_b64 s[14:15], s[10:11], 5
+; GFX908-NEXT:    v_readfirstlane_b32 s10, v0
+; GFX908-NEXT:    s_mul_i32 s8, s8, s10
+; GFX908-NEXT:    s_mul_hi_u32 s8, s10, s8
+; GFX908-NEXT:    s_add_i32 s10, s10, s8
+; GFX908-NEXT:    s_mul_hi_u32 s8, s0, s10
+; GFX908-NEXT:    s_mul_i32 s10, s8, s1
+; GFX908-NEXT:    s_sub_i32 s0, s0, s10
+; GFX908-NEXT:    s_add_i32 s11, s8, 1
+; GFX908-NEXT:    s_sub_i32 s10, s0, s1
+; GFX908-NEXT:    s_cmp_ge_u32 s0, s1
+; GFX908-NEXT:    s_cselect_b32 s8, s11, s8
+; GFX908-NEXT:    s_cselect_b32 s0, s10, s0
+; GFX908-NEXT:    s_add_i32 s10, s8, 1
+; GFX908-NEXT:    s_cmp_ge_u32 s0, s1
+; GFX908-NEXT:    s_cselect_b32 s8, s10, s8
+; GFX908-NEXT:    s_lshr_b32 s5, s5, 16
+; GFX908-NEXT:    v_cvt_f32_f16_e32 v19, s5
+; GFX908-NEXT:    s_lshl_b64 s[10:11], s[2:3], 5
+; GFX908-NEXT:    s_lshl_b64 s[14:15], s[8:9], 5
+; GFX908-NEXT:    s_lshl_b64 s[12:13], s[6:7], 5
 ; GFX908-NEXT:    v_mov_b32_e32 v0, 0
 ; GFX908-NEXT:    s_and_b64 s[0:1], exec, s[0:1]
-; GFX908-NEXT:    s_or_b32 s14, s14, 28
-; GFX908-NEXT:    s_lshl_b64 s[16:17], s[8:9], 5
+; GFX908-NEXT:    s_or_b32 s12, s12, 28
 ; GFX908-NEXT:    v_mov_b32_e32 v1, 0
 ; GFX908-NEXT:    s_waitcnt vmcnt(0)
 ; GFX908-NEXT:    v_readfirstlane_b32 s5, v16
@@ -610,8 +610,8 @@ define amdgpu_kernel void @introduced_copy_to_sgpr(i64 %arg, i32 %arg1, i32 %arg
 ; GFX908-NEXT:  .LBB3_5: ; %bb16
 ; GFX908-NEXT:    ; Parent Loop BB3_2 Depth=1
 ; GFX908-NEXT:    ; => This Inner Loop Header: Depth=2
-; GFX908-NEXT:    s_add_u32 s22, s20, s9
-; GFX908-NEXT:    s_addc_u32 s23, s21, s13
+; GFX908-NEXT:    s_add_u32 s22, s20, s5
+; GFX908-NEXT:    s_addc_u32 s23, s21, s9
 ; GFX908-NEXT:    global_load_dword v21, v17, s[22:23] offset:-12 glc
 ; GFX908-NEXT:    s_waitcnt vmcnt(0)
 ; GFX908-NEXT:    global_load_dword v20, v17, s[22:23] offset:-8 glc
@@ -715,8 +715,7 @@ define amdgpu_kernel void @introduced_copy_to_sgpr(i64 %arg, i32 %arg1, i32 %arg
 ; GFX90A-NEXT:    s_lshl_b64 s[14:15], s[8:9], 5
 ; GFX90A-NEXT:    s_lshl_b64 s[12:13], s[6:7], 5
 ; GFX90A-NEXT:    s_and_b64 s[0:1], exec, s[0:1]
-; GFX90A-NEXT:    s_or_b32 s14, s14, 28
-; GFX90A-NEXT:    s_lshl_b64 s[16:17], s[8:9], 5
+; GFX90A-NEXT:    s_or_b32 s12, s12, 28
 ; GFX90A-NEXT:    s_waitcnt vmcnt(0)
 ; GFX90A-NEXT:    v_readfirstlane_b32 s5, v18
 ; GFX90A-NEXT:    s_and_b32 s5, 0xffff, s5
@@ -774,8 +773,8 @@ define amdgpu_kernel void @introduced_copy_to_sgpr(i64 %arg, i32 %arg1, i32 %arg
 ; GFX90A-NEXT:  .LBB3_5: ; %bb16
 ; GFX90A-NEXT:    ; Parent Loop BB3_2 Depth=1
 ; GFX90A-NEXT:    ; => This Inner Loop Header: Depth=2
-; GFX90A-NEXT:    s_add_u32 s22, s20, s9
-; GFX90A-NEXT:    s_addc_u32 s23, s21, s13
+; GFX90A-NEXT:    s_add_u32 s22, s20, s5
+; GFX90A-NEXT:    s_addc_u32 s23, s21, s9
 ; GFX90A-NEXT:    global_load_dword v21, v19, s[22:23] offset:-12 glc
 ; GFX90A-NEXT:    s_waitcnt vmcnt(0)
 ; GFX90A-NEXT:    global_load_dword v20, v19, s[22:23] offset:-8 glc

llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-idiv.ll

@@ -5512,15 +5512,16 @@ define amdgpu_kernel void @udiv_v2i32_pow2k_denom(ptr addrspace(1) %out, <2 x i3
 ; GFX6-LABEL: udiv_v2i32_pow2k_denom:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    s_lshr_b32 s4, s4, 12
-; GFX6-NEXT:    s_lshr_b32 s5, s5, 12
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    v_mov_b32_e32 v1, s5
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    s_lshr_b32 s0, s2, 12
+; GFX6-NEXT:    s_lshr_b32 s1, s3, 12
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    v_mov_b32_e32 v1, s1
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: udiv_v2i32_pow2k_denom:
@@ -5554,18 +5555,19 @@ define amdgpu_kernel void @udiv_v2i32_mixed_pow2k_denom(ptr addrspace(1) %out, <
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
 ; GFX6-NEXT:    v_mov_b32_e32 v0, 0x100101
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    v_mul_hi_u32 v0, s5, v0
-; GFX6-NEXT:    s_lshr_b32 s4, s4, 12
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_sub_i32_e32 v1, vcc, s5, v0
+; GFX6-NEXT:    v_mul_hi_u32 v0, s3, v0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_lshr_b32 s0, s2, 12
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    v_sub_i32_e32 v1, vcc, s3, v0
 ; GFX6-NEXT:    v_lshrrev_b32_e32 v1, 1, v1
 ; GFX6-NEXT:    v_add_i32_e32 v0, vcc, v1, v0
 ; GFX6-NEXT:    v_lshrrev_b32_e32 v1, 11, v0
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: udiv_v2i32_mixed_pow2k_denom:
@@ -5906,15 +5908,16 @@ define amdgpu_kernel void @urem_v2i32_pow2k_denom(ptr addrspace(1) %out, <2 x i3
 ; GFX6-LABEL: urem_v2i32_pow2k_denom:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    s_and_b32 s4, s4, 0xfff
-; GFX6-NEXT:    s_and_b32 s5, s5, 0xfff
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    v_mov_b32_e32 v1, s5
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    s_and_b32 s0, s2, 0xfff
+; GFX6-NEXT:    s_and_b32 s1, s3, 0xfff
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    v_mov_b32_e32 v1, s1
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: urem_v2i32_pow2k_denom:
@@ -6288,21 +6291,22 @@ define amdgpu_kernel void @sdiv_v2i32_pow2k_denom(ptr addrspace(1) %out, <2 x i3
 ; GFX6-LABEL: sdiv_v2i32_pow2k_denom:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    s_ashr_i32 s6, s4, 31
-; GFX6-NEXT:    s_ashr_i32 s7, s5, 31
-; GFX6-NEXT:    s_lshr_b32 s6, s6, 20
-; GFX6-NEXT:    s_lshr_b32 s7, s7, 20
-; GFX6-NEXT:    s_add_i32 s4, s4, s6
-; GFX6-NEXT:    s_add_i32 s5, s5, s7
-; GFX6-NEXT:    s_ashr_i32 s4, s4, 12
-; GFX6-NEXT:    s_ashr_i32 s5, s5, 12
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    v_mov_b32_e32 v1, s5
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    s_ashr_i32 s0, s2, 31
+; GFX6-NEXT:    s_ashr_i32 s1, s3, 31
+; GFX6-NEXT:    s_lshr_b32 s0, s0, 20
+; GFX6-NEXT:    s_lshr_b32 s1, s1, 20
+; GFX6-NEXT:    s_add_i32 s0, s2, s0
+; GFX6-NEXT:    s_add_i32 s1, s3, s1
+; GFX6-NEXT:    s_ashr_i32 s0, s0, 12
+; GFX6-NEXT:    s_ashr_i32 s1, s1, 12
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    v_mov_b32_e32 v1, s1
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: sdiv_v2i32_pow2k_denom:
@@ -6342,21 +6346,22 @@ define amdgpu_kernel void @ssdiv_v2i32_mixed_pow2k_denom(ptr addrspace(1) %out,
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
 ; GFX6-NEXT:    v_mov_b32_e32 v0, 0x80080081
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    v_mul_hi_i32 v0, s5, v0
-; GFX6-NEXT:    s_ashr_i32 s6, s4, 31
-; GFX6-NEXT:    s_lshr_b32 s6, s6, 20
-; GFX6-NEXT:    s_add_i32 s4, s4, s6
-; GFX6-NEXT:    v_add_i32_e32 v0, vcc, s5, v0
-; GFX6-NEXT:    s_ashr_i32 s4, s4, 12
+; GFX6-NEXT:    v_mul_hi_i32 v0, s3, v0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_ashr_i32 s0, s2, 31
+; GFX6-NEXT:    s_lshr_b32 s0, s0, 20
+; GFX6-NEXT:    s_add_i32 s0, s2, s0
+; GFX6-NEXT:    v_add_i32_e32 v0, vcc, s3, v0
+; GFX6-NEXT:    s_ashr_i32 s0, s0, 12
 ; GFX6-NEXT:    v_lshrrev_b32_e32 v1, 31, v0
 ; GFX6-NEXT:    v_ashrrev_i32_e32 v0, 11, v0
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
+; GFX6-NEXT:    s_mov_b32 s5, s1
 ; GFX6-NEXT:    v_add_i32_e32 v1, vcc, v0, v1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: ssdiv_v2i32_mixed_pow2k_denom:
@@ -6793,23 +6798,24 @@ define amdgpu_kernel void @srem_v2i32_pow2k_denom(ptr addrspace(1) %out, <2 x i3
 ; GFX6-LABEL: srem_v2i32_pow2k_denom:
 ; GFX6:       ; %bb.0:
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    s_ashr_i32 s6, s4, 31
-; GFX6-NEXT:    s_lshr_b32 s6, s6, 20
-; GFX6-NEXT:    s_ashr_i32 s7, s5, 31
-; GFX6-NEXT:    s_add_i32 s6, s4, s6
-; GFX6-NEXT:    s_lshr_b32 s7, s7, 20
-; GFX6-NEXT:    s_and_b32 s6, s6, 0xfffff000
-; GFX6-NEXT:    s_sub_i32 s4, s4, s6
-; GFX6-NEXT:    s_add_i32 s6, s5, s7
-; GFX6-NEXT:    s_and_b32 s6, s6, 0xfffff000
-; GFX6-NEXT:    s_sub_i32 s5, s5, s6
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    v_mov_b32_e32 v1, s5
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    s_ashr_i32 s0, s2, 31
+; GFX6-NEXT:    s_ashr_i32 s1, s3, 31
+; GFX6-NEXT:    s_lshr_b32 s0, s0, 20
+; GFX6-NEXT:    s_lshr_b32 s1, s1, 20
+; GFX6-NEXT:    s_add_i32 s0, s2, s0
+; GFX6-NEXT:    s_add_i32 s1, s3, s1
+; GFX6-NEXT:    s_and_b32 s0, s0, 0xfffff000
+; GFX6-NEXT:    s_and_b32 s1, s1, 0xfffff000
+; GFX6-NEXT:    s_sub_i32 s0, s2, s0
+; GFX6-NEXT:    s_sub_i32 s1, s3, s1
+; GFX6-NEXT:    v_mov_b32_e32 v0, s0
+; GFX6-NEXT:    v_mov_b32_e32 v1, s1
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX9-LABEL: srem_v2i32_pow2k_denom:

llvm/test/CodeGen/AMDGPU/build_vector.ll

@@ -257,15 +257,16 @@ define amdgpu_kernel void @build_v2i32_from_v4i16_shuffle(ptr addrspace(1) %out,
 ; GFX6-LABEL: build_v2i32_from_v4i16_shuffle:
 ; GFX6:       ; %bb.0: ; %entry
 ; GFX6-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; GFX6-NEXT:    s_mov_b32 s7, 0xf000
+; GFX6-NEXT:    s_mov_b32 s6, -1
 ; GFX6-NEXT:    s_waitcnt lgkmcnt(0)
-; GFX6-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; GFX6-NEXT:    s_lshl_b32 s5, s5, 16
-; GFX6-NEXT:    s_lshl_b32 s4, s4, 16
-; GFX6-NEXT:    s_mov_b32 s3, 0xf000
-; GFX6-NEXT:    s_mov_b32 s2, -1
-; GFX6-NEXT:    v_mov_b32_e32 v0, s4
-; GFX6-NEXT:    v_mov_b32_e32 v1, s5
-; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; GFX6-NEXT:    s_mov_b32 s4, s0
+; GFX6-NEXT:    s_mov_b32 s5, s1
+; GFX6-NEXT:    s_lshl_b32 s0, s3, 16
+; GFX6-NEXT:    s_lshl_b32 s1, s2, 16
+; GFX6-NEXT:    v_mov_b32_e32 v0, s1
+; GFX6-NEXT:    v_mov_b32_e32 v1, s0
+; GFX6-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; GFX6-NEXT:    s_endpgm
 ;
 ; GFX8-LABEL: build_v2i32_from_v4i16_shuffle:

llvm/test/CodeGen/AMDGPU/fabs.ll

@@ -99,15 +99,16 @@ define amdgpu_kernel void @fabs_v2f32(ptr addrspace(1) %out, <2 x float> %in) {
 ; SI-LABEL: fabs_v2f32:
 ; SI:       ; %bb.0:
 ; SI-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; SI-NEXT:    s_mov_b32 s7, 0xf000
+; SI-NEXT:    s_mov_b32 s6, -1
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
-; SI-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; SI-NEXT:    s_mov_b32 s3, 0xf000
-; SI-NEXT:    s_mov_b32 s2, -1
-; SI-NEXT:    s_bitset0_b32 s5, 31
-; SI-NEXT:    s_bitset0_b32 s4, 31
-; SI-NEXT:    v_mov_b32_e32 v0, s4
-; SI-NEXT:    v_mov_b32_e32 v1, s5
-; SI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; SI-NEXT:    s_mov_b32 s4, s0
+; SI-NEXT:    s_mov_b32 s5, s1
+; SI-NEXT:    s_and_b32 s0, s3, 0x7fffffff
+; SI-NEXT:    s_and_b32 s1, s2, 0x7fffffff
+; SI-NEXT:    v_mov_b32_e32 v0, s1
+; SI-NEXT:    v_mov_b32_e32 v1, s0
+; SI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; SI-NEXT:    s_endpgm
 ;
 ; VI-LABEL: fabs_v2f32:

llvm/test/CodeGen/AMDGPU/fdiv.ll

@@ -1242,8 +1242,8 @@ define amdgpu_kernel void @s_fdiv_ulp25_v2f32(ptr addrspace(1) %out, <2 x float>
 ; GFX11-NEXT:    v_rcp_f32_e32 v1, s7
 ; GFX11-NEXT:    v_mov_b32_e32 v2, 0
 ; GFX11-NEXT:    s_waitcnt_depctr depctr_va_vdst(0)
-; GFX11-NEXT:    v_dual_mul_f32 v0, s0, v0 :: v_dual_mul_f32 v1, s1, v1
-; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[4:5]
+; GFX11-NEXT:    v_dual_mul_f32 v0, s2, v0 :: v_dual_mul_f32 v1, s3, v1
+; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
 ; GFX11-NEXT:    s_endpgm
 ;
 ; EG-LABEL: s_fdiv_ulp25_v2f32:
@@ -1320,8 +1320,8 @@ define amdgpu_kernel void @s_fdiv_v2f32_fast_math(ptr addrspace(1) %out, <2 x fl
 ; GFX11-NEXT:    v_rcp_f32_e32 v2, s6
 ; GFX11-NEXT:    v_mov_b32_e32 v3, 0
 ; GFX11-NEXT:    s_waitcnt_depctr depctr_va_vdst(0)
-; GFX11-NEXT:    v_dual_mul_f32 v1, s1, v0 :: v_dual_mul_f32 v0, s0, v2
-; GFX11-NEXT:    global_store_b64 v3, v[0:1], s[4:5]
+; GFX11-NEXT:    v_dual_mul_f32 v1, s3, v0 :: v_dual_mul_f32 v0, s2, v2
+; GFX11-NEXT:    global_store_b64 v3, v[0:1], s[0:1]
 ; GFX11-NEXT:    s_endpgm
 ;
 ; EG-LABEL: s_fdiv_v2f32_fast_math:
@@ -1398,8 +1398,8 @@ define amdgpu_kernel void @s_fdiv_v2f32_arcp_math(ptr addrspace(1) %out, <2 x fl
 ; GFX11-NEXT:    v_rcp_f32_e32 v2, s6
 ; GFX11-NEXT:    v_mov_b32_e32 v3, 0
 ; GFX11-NEXT:    s_waitcnt_depctr depctr_va_vdst(0)
-; GFX11-NEXT:    v_dual_mul_f32 v1, s1, v0 :: v_dual_mul_f32 v0, s0, v2
-; GFX11-NEXT:    global_store_b64 v3, v[0:1], s[4:5]
+; GFX11-NEXT:    v_dual_mul_f32 v1, s3, v0 :: v_dual_mul_f32 v0, s2, v2
+; GFX11-NEXT:    global_store_b64 v3, v[0:1], s[0:1]
 ; GFX11-NEXT:    s_endpgm
 ;
 ; EG-LABEL: s_fdiv_v2f32_arcp_math:

llvm/test/CodeGen/AMDGPU/fnearbyint.ll

@@ -121,13 +121,14 @@ define amdgpu_kernel void @fnearbyint_v2f32(ptr addrspace(1) %out, <2 x float> %
 ; SICI-LABEL: fnearbyint_v2f32:
 ; SICI:       ; %bb.0: ; %entry
 ; SICI-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x9
+; SICI-NEXT:    s_mov_b32 s7, 0xf000
+; SICI-NEXT:    s_mov_b32 s6, -1
 ; SICI-NEXT:    s_waitcnt lgkmcnt(0)
-; SICI-NEXT:    s_mov_b64 s[4:5], s[2:3]
-; SICI-NEXT:    s_mov_b32 s3, 0xf000
-; SICI-NEXT:    s_mov_b32 s2, -1
-; SICI-NEXT:    v_rndne_f32_e32 v1, s5
-; SICI-NEXT:    v_rndne_f32_e32 v0, s4
-; SICI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
+; SICI-NEXT:    s_mov_b32 s4, s0
+; SICI-NEXT:    s_mov_b32 s5, s1
+; SICI-NEXT:    v_rndne_f32_e32 v1, s3
+; SICI-NEXT:    v_rndne_f32_e32 v0, s2
+; SICI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[4:7], 0
 ; SICI-NEXT:    s_endpgm
 ;
 ; VI-LABEL: fnearbyint_v2f32: