| ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -global-isel -new-reg-bank-select -mattr=-promote-alloca -mattr=+enable-flat-scratch < %s | FileCheck -check-prefix=GFX9 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1030 -global-isel -new-reg-bank-select -mattr=-promote-alloca -mattr=+enable-flat-scratch < %s | FileCheck -check-prefix=GFX10 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx942 -global-isel -new-reg-bank-select -mattr=-promote-alloca < %s | FileCheck -check-prefix=GFX942 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -global-isel -new-reg-bank-select -mattr=-promote-alloca < %s | FileCheck -check-prefix=GFX11 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 -global-isel -new-reg-bank-select -mattr=-promote-alloca < %s | FileCheck -check-prefix=GFX12 %s |
| |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -global-isel -new-reg-bank-select -mattr=-unaligned-access-mode -mattr=-promote-alloca -mattr=+enable-flat-scratch < %s | FileCheck -check-prefixes=UNALIGNED_GFX9 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1030 -global-isel -new-reg-bank-select -mattr=-unaligned-access-mode -mattr=-promote-alloca -mattr=+enable-flat-scratch < %s | FileCheck -check-prefixes=UNALIGNED_GFX10 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx942 -global-isel -new-reg-bank-select -mattr=-unaligned-access-mode -mattr=-promote-alloca < %s | FileCheck -check-prefixes=UNALIGNED_GFX942 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -global-isel -new-reg-bank-select -mattr=-unaligned-access-mode -mattr=-promote-alloca < %s | FileCheck -check-prefixes=UNALIGNED_GFX11 %s |
| ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 -global-isel -new-reg-bank-select -mattr=-unaligned-access-mode -mattr=-promote-alloca < %s | FileCheck -check-prefixes=UNALIGNED_GFX12 %s |
| |
| define amdgpu_kernel void @store_load_sindex_kernel(i32 %idx) { |
| ; GFX9-LABEL: store_load_sindex_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_sindex_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_sindex_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_sindex_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_sindex_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_sindex_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_sindex_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_sindex_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_sindex_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_sindex_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_vindex_kernel(i32 %n) { |
| ; GFX9-LABEL: store_load_vindex_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v2, 2, v0 |
| ; GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: v_add_u32_e32 v2, s0, v2 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: scratch_store_dword v2, v1, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_vindex_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_vindex_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v2, 2, v0 |
| ; GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX942-NEXT: v_add_u32_e32 v2, s0, v2 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: scratch_store_dword v2, v1, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_vindex_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) |
| ; GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instid1(SALU_CYCLE_1) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX11-NEXT: scratch_store_b32 v0, v2, off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_vindex_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_3) |
| ; GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, s0 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v2, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v2, s0, v2 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v2, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v2, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v2, s0, v2 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v2, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) |
| ; UNALIGNED_GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v2, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_3) |
| ; UNALIGNED_GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, s0 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %i.alloca = alloca [2 x [32 x float]], align 4, addrspace(5) |
| %i = getelementptr inbounds [2 x [32 x float]], ptr addrspace (5) %i.alloca, i32 0, i32 %n |
| %i2 = tail call i32 @llvm.amdgcn.workitem.id.x() |
| %i3 = zext i32 %i2 to i64 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i2 |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = sub nsw i32 0, %i2 |
| %i9.1 = getelementptr [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i9.1, i32 0, i32 31 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define void @store_load_vindex_foo(i32 %idx) { |
| ; GFX9-LABEL: store_load_vindex_foo: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_vindex_foo: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_vindex_foo: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, s32 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_vindex_foo: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v0, s32 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_vindex_foo: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, s32 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s32 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_foo: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_foo: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_foo: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, s32 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_foo: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v0, s32 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_foo: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, s32 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s32 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define void @private_ptr_foo(ptr addrspace(5) nocapture %arg) { |
| ; GFX9-LABEL: private_ptr_foo: |
| ; GFX9: ; %bb.0: |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; GFX9-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: private_ptr_foo: |
| ; GFX10: ; %bb.0: |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; GFX10-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: private_ptr_foo: |
| ; GFX942: ; %bb.0: |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; GFX942-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: private_ptr_foo: |
| ; GFX11: ; %bb.0: |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; GFX11-NEXT: scratch_store_b32 v0, v1, off offset:4 |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: private_ptr_foo: |
| ; GFX12: ; %bb.0: |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; GFX12-NEXT: scratch_store_b32 v0, v1, off offset:4 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: private_ptr_foo: |
| ; UNALIGNED_GFX9: ; %bb.0: |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: private_ptr_foo: |
| ; UNALIGNED_GFX10: ; %bb.0: |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: private_ptr_foo: |
| ; UNALIGNED_GFX942: ; %bb.0: |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v0, v1, off offset:4 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: private_ptr_foo: |
| ; UNALIGNED_GFX11: ; %bb.0: |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v1, off offset:4 |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: private_ptr_foo: |
| ; UNALIGNED_GFX12: ; %bb.0: |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v1, 0x41200000 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v1, off offset:4 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| %gep = getelementptr inbounds float, ptr addrspace(5) %arg, i32 1 |
| store float 1.000000e+01, ptr addrspace(5) %gep, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_sindex_small_offset_kernel(i32 %idx) { |
| ; GFX9-LABEL: store_load_sindex_small_offset_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: s_mov_b32 s1, 0 |
| ; GFX9-NEXT: scratch_load_dword v0, off, s1 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX9-NEXT: s_addk_i32 s1, 0x100 |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_addk_i32 s0, 0x100 |
| ; GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_sindex_small_offset_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX10-NEXT: s_addk_i32 s0, 0x100 |
| ; GFX10-NEXT: s_addk_i32 s1, 0x100 |
| ; GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_sindex_small_offset_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: scratch_load_dword v0, off, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX942-NEXT: s_addk_i32 s1, 0x100 |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_addk_i32 s0, 0x100 |
| ; GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_sindex_small_offset_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX11-NEXT: s_addk_i32 s0, 0x100 |
| ; GFX11-NEXT: s_addk_i32 s1, 0x100 |
| ; GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_sindex_small_offset_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX12-NEXT: s_addk_co_i32 s0, 0x100 |
| ; GFX12-NEXT: s_addk_co_i32 s1, 0x100 |
| ; GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_sindex_small_offset_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: s_mov_b32 s1, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s1 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_addk_i32 s1, 0x100 |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_addk_i32 s0, 0x100 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_sindex_small_offset_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX10-NEXT: s_addk_i32 s0, 0x100 |
| ; UNALIGNED_GFX10-NEXT: s_addk_i32 s1, 0x100 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_sindex_small_offset_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_addk_i32 s1, 0x100 |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_addk_i32 s0, 0x100 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_sindex_small_offset_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX11-NEXT: s_addk_i32 s0, 0x100 |
| ; UNALIGNED_GFX11-NEXT: s_addk_i32 s1, 0x100 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_sindex_small_offset_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX12-NEXT: s_addk_co_i32 s0, 0x100 |
| ; UNALIGNED_GFX12-NEXT: s_addk_co_i32 s1, 0x100 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %padding = alloca [64 x i32], align 4, addrspace(5) |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %pad_gep = getelementptr inbounds [64 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_vindex_small_offset_kernel(i32 %n) { |
| ; GFX9-LABEL: store_load_vindex_small_offset_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: s_mov_b32 s1, 0 |
| ; GFX9-NEXT: scratch_load_dword v1, off, s1 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; GFX9-NEXT: s_add_u32 s0, 0x100, s0 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: scratch_store_dword v1, v2, off offset:128 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_vindex_small_offset_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: scratch_load_dword v3, off, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, 0x100, v0 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off offset:128 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX10-NEXT: s_add_u32 s0, 0x100, s0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_vindex_small_offset_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: scratch_load_dword v1, off, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX942-NEXT: s_add_u32 s0, 0x100, s0 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: scratch_store_dword v1, v2, off offset:384 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_vindex_small_offset_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX11-NEXT: scratch_load_b32 v3, off, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX11-NEXT: scratch_store_b32 v0, v2, off offset:384 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX11-NEXT: s_add_u32 s0, 0x100, s0 |
| ; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_vindex_small_offset_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX12-NEXT: scratch_load_b32 v3, off, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, off offset:384 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX12-NEXT: s_add_co_u32 s0, 0x100, s0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_small_offset_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: s_mov_b32 s1, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v1, off, s1 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 s0, 0x100, s0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v1, v2, off offset:128 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_small_offset_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v3, off, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, 0x100, v0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off offset:128 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, 0x100, s0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_small_offset_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v1, off, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX942-NEXT: s_add_u32 s0, 0x100, s0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v1, v2, off offset:384 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_small_offset_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v3, off, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v2, off offset:384 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX11-NEXT: s_add_u32 s0, 0x100, s0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_small_offset_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v3, off, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, off offset:384 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX12-NEXT: s_add_co_u32 s0, 0x100, s0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %padding = alloca [64 x i32], align 4, addrspace(5) |
| %i.alloca = alloca [2 x [32 x float]], align 4, addrspace(5) |
| %i = getelementptr inbounds [2 x [32 x float]], ptr addrspace (5) %i.alloca, i32 0, i32 %n |
| %i.1 = getelementptr inbounds [2 x [32 x float]], ptr addrspace (5) %i.alloca, i32 0, i32 1 |
| %pad_gep = getelementptr inbounds [64 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i2 = tail call i32 @llvm.amdgcn.workitem.id.x() |
| %i3 = zext i32 %i2 to i64 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i.1, i32 0, i32 %i2 |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = sub nsw i32 0, %i2 |
| %i9.1 = getelementptr [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i9.1, i32 0, i32 31 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define void @store_load_vindex_small_offset_foo(i32 %idx) { |
| ; GFX9-LABEL: store_load_vindex_small_offset_foo: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v1, off, s32 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, 0x100, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_vindex_small_offset_foo: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: scratch_load_dword v3, off, s32 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, 0x100, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, 0x100, v1 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_vindex_small_offset_foo: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v1, off, s32 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX942-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, s32 offset:256 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_vindex_small_offset_foo: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX11-NEXT: scratch_load_b32 v3, off, s32 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, 0x100, v1 |
| ; GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v0, s32 offset:256 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_vindex_small_offset_foo: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: scratch_load_b32 v3, off, s32 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, s32 offset:256 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s32 offset:256 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_small_offset_foo: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v1, off, s32 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, 0x100, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_small_offset_foo: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v3, off, s32 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, 0x100, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, 0x100, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_small_offset_foo: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v1, off, s32 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v1, 0x100, v1 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, s32 offset:256 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_small_offset_foo: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v3, off, s32 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, 0x100, v1 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v0, s32 offset:256 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_small_offset_foo: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v3, off, s32 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, s32 offset:256 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s32 offset:256 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| %padding = alloca [64 x i32], align 4, addrspace(5) |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %pad_gep = getelementptr inbounds [64 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_sindex_large_offset_kernel(i32 %idx) { |
| ; GFX9-LABEL: store_load_sindex_large_offset_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: s_mov_b32 s1, 0 |
| ; GFX9-NEXT: scratch_load_dword v0, off, s1 offset:4 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX9-NEXT: s_addk_i32 s1, 0x4004 |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_addk_i32 s0, 0x4004 |
| ; GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_sindex_large_offset_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, off offset:4 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX10-NEXT: s_addk_i32 s0, 0x4004 |
| ; GFX10-NEXT: s_addk_i32 s1, 0x4004 |
| ; GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_sindex_large_offset_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: scratch_load_dword v0, off, off offset:4 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; GFX942-NEXT: s_addk_i32 s1, 0x4004 |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_addk_i32 s0, 0x4004 |
| ; GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_sindex_large_offset_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, off offset:4 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX11-NEXT: s_addk_i32 s0, 0x4004 |
| ; GFX11-NEXT: s_addk_i32 s1, 0x4004 |
| ; GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_sindex_large_offset_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; GFX12-NEXT: s_addk_co_i32 s0, 0x4000 |
| ; GFX12-NEXT: s_addk_co_i32 s1, 0x4000 |
| ; GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_sindex_large_offset_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: s_mov_b32 s1, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s1 offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_addk_i32 s1, 0x4004 |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_addk_i32 s0, 0x4004 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_sindex_large_offset_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, off offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX10-NEXT: s_addk_i32 s0, 0x4004 |
| ; UNALIGNED_GFX10-NEXT: s_addk_i32 s1, 0x4004 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v0, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, s1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_sindex_large_offset_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s1, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: s_and_b32 s0, s0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_addk_i32 s1, 0x4004 |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_addk_i32 s0, 0x4004 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_sindex_large_offset_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, off offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX11-NEXT: s_addk_i32 s0, 0x4004 |
| ; UNALIGNED_GFX11-NEXT: s_addk_i32 s1, 0x4004 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v0, s0 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, s1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_sindex_large_offset_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_and_b32 s1, s0, 15 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 2 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s1, s1, 2 |
| ; UNALIGNED_GFX12-NEXT: s_addk_co_i32 s0, 0x4000 |
| ; UNALIGNED_GFX12-NEXT: s_addk_co_i32 s1, 0x4000 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v0, s0 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, s1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %padding = alloca [4096 x i32], align 4, addrspace(5) |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %pad_gep = getelementptr inbounds [4096 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_vindex_large_offset_kernel(i32 %n) { |
| ; GFX9-LABEL: store_load_vindex_large_offset_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: s_mov_b32 s1, 0 |
| ; GFX9-NEXT: scratch_load_dword v1, off, s1 offset:4 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; GFX9-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: scratch_store_dword v1, v2, off offset:128 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_vindex_large_offset_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: scratch_load_dword v3, off, off offset:4 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, 0x4004, v0 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off offset:128 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX10-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_vindex_large_offset_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: scratch_load_dword v1, off, off offset:4 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX942-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX942-NEXT: s_movk_i32 s1, 0x4004 |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: scratch_store_dword v1, v2, s1 offset:128 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_vindex_large_offset_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX11-NEXT: scratch_load_b32 v3, off, off offset:4 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(SALU_CYCLE_1) |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX11-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; GFX11-NEXT: s_movk_i32 s0, 0x4004 |
| ; GFX11-NEXT: scratch_store_b32 v0, v2, s0 offset:128 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_vindex_large_offset_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX12-NEXT: scratch_load_b32 v3, off, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, off offset:16512 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; GFX12-NEXT: s_add_co_u32 s0, 0x4000, s0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_large_offset_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: s_mov_b32 s1, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v1, off, s1 offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX9-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v1, v2, off offset:128 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off offset:124 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_large_offset_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v3, off, off offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, 0x4004, v0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off offset:128 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_large_offset_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v1, off, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_sub_u32_e32 v0, 0, v0 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX942-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX942-NEXT: s_movk_i32 s1, 0x4004 |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v1, v2, s1 offset:128 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, off offset:124 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_large_offset_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v3, off, off offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX11-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX11-NEXT: s_add_u32 s0, 0x4004, s0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s0, v1 |
| ; UNALIGNED_GFX11-NEXT: s_movk_i32 s0, 0x4004 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v2, s0 offset:128 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v1, off offset:124 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_large_offset_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v3, off, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_sub_nc_u32_e32 v1, 0, v0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, off offset:16512 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_lshl_b32 s0, s0, 7 |
| ; UNALIGNED_GFX12-NEXT: s_add_co_u32 s0, 0x4000, s0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s0 offset:124 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %padding = alloca [4096 x i32], align 4, addrspace(5) |
| %i.alloca = alloca [2 x [32 x float]], align 4, addrspace(5) |
| %i = getelementptr inbounds [2 x [32 x float]], ptr addrspace (5) %i.alloca, i32 0, i32 %n |
| %i.1 = getelementptr inbounds [2 x [32 x float]], ptr addrspace (5) %i.alloca, i32 0, i32 1 |
| %pad_gep = getelementptr inbounds [4096 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i2 = tail call i32 @llvm.amdgcn.workitem.id.x() |
| %i3 = zext i32 %i2 to i64 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i.1, i32 0, i32 %i2 |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = sub nsw i32 0, %i2 |
| %i9.1 = getelementptr [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i9.1, i32 0, i32 31 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define void @store_load_vindex_large_offset_foo(i32 %idx) { |
| ; GFX9-LABEL: store_load_vindex_large_offset_foo: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v1, off, s32 offset:4 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX9-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_add_u32_e32 v0, 0x4004, v0 |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_vindex_large_offset_foo: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX10-NEXT: scratch_load_dword v3, off, s32 offset:4 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, 0x4004, v0 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v1, 0x4004, v1 |
| ; GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_vindex_large_offset_foo: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v1, off, s32 offset:4 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; GFX942-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX942-NEXT: s_add_i32 s0, s32, 0x4004 |
| ; GFX942-NEXT: scratch_load_dword v0, v0, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_vindex_large_offset_foo: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; GFX11-NEXT: s_add_i32 s0, s32, 0x4004 |
| ; GFX11-NEXT: scratch_load_b32 v3, off, s32 offset:4 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v1, 0x4004, v1 |
| ; GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v0, s0 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_vindex_large_offset_foo: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; GFX12-NEXT: scratch_load_b32 v3, off, s32 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v2, s32 offset:16384 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v1, s32 offset:16384 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vindex_large_offset_foo: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v1, off, s32 offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX9-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v1, v2, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, 0x4004, v0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vindex_large_offset_foo: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_and_b32_e32 v1, 15, v0 |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v3, off, s32 offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s32, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, 0x4004, v0 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v1, 0x4004, v1 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v2, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v1, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vindex_large_offset_foo: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v1, off, s32 offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v1, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v1, 0x4004, v1 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 15 |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v1, v2, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX942-NEXT: s_add_i32 s0, s32, 0x4004 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vindex_large_offset_foo: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_lshlrev_b32 v1, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: v_and_b32_e32 v0, 15, v0 |
| ; UNALIGNED_GFX11-NEXT: s_add_i32 s0, s32, 0x4004 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v3, off, s32 offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, s32, v1 |
| ; UNALIGNED_GFX11-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v1, 0x4004, v1 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v1, v2, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v0, s0 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vindex_large_offset_foo: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v2, 15 :: v_dual_and_b32 v1, 15, v0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v0, 2, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v3, off, s32 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_lshlrev_b32_e32 v1, 2, v1 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v2, s32 offset:16384 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v1, s32 offset:16384 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| %padding = alloca [4096 x i32], align 4, addrspace(5) |
| %i = alloca [32 x float], align 4, addrspace(5) |
| %pad_gep = getelementptr inbounds [4096 x i32], ptr addrspace(5) %padding, i32 0, i32 0 |
| %pad_load = load volatile i32, ptr addrspace(5) %pad_gep, align 4 |
| %i7 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %idx |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i9 = and i32 %idx, 15 |
| %i10 = getelementptr inbounds [32 x float], ptr addrspace(5) %i, i32 0, i32 %i9 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_large_imm_offset_kernel() { |
| ; GFX9-LABEL: store_load_large_imm_offset_kernel: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX9-NEXT: s_mov_b32 s0, 0 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s0 offset:4 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX9-NEXT: s_add_i32 s0, s0, 4 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s0 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_large_imm_offset_kernel: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX10-NEXT: s_add_i32 s0, s0, 4 |
| ; GFX10-NEXT: scratch_store_dword off, v0, off offset:4 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_store_dword off, v1, s0 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, s0 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_large_imm_offset_kernel: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX942-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX942-NEXT: scratch_store_dword off, v0, off offset:4 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX942-NEXT: s_add_i32 s0, s0, 4 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_large_imm_offset_kernel: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; GFX11-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; GFX11-NEXT: s_add_i32 s0, s0, 4 |
| ; GFX11-NEXT: scratch_store_b32 off, v0, off offset:4 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_store_b32 off, v1, s0 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, s0 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_large_imm_offset_kernel: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; GFX12-NEXT: scratch_store_b32 off, v0, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 off, v1, off offset:16000 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, off offset:16000 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_large_imm_offset_kernel: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX9-NEXT: s_mov_b32 s0, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s0 offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_add_i32 s0, s0, 4 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_large_imm_offset_kernel: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX10-NEXT: s_add_i32 s0, s0, 4 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v0, off offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v1, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, s0 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_large_imm_offset_kernel: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX942-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_add_i32 s0, s0, 4 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_large_imm_offset_kernel: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; UNALIGNED_GFX11-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: s_add_i32 s0, s0, 4 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v0, off offset:4 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v1, s0 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, s0 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_large_imm_offset_kernel: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v0, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v1, off offset:16000 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, off offset:16000 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %i = alloca [4096 x i32], align 4, addrspace(5) |
| %i1 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 0 |
| store volatile i32 13, ptr addrspace(5) %i1, align 4 |
| %i7 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 4000 |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i10 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 4000 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define void @store_load_large_imm_offset_foo() { |
| ; GFX9-LABEL: store_load_large_imm_offset_foo: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX9-NEXT: s_add_i32 s1, s32, s0 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s32 offset:4 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX9-NEXT: s_add_i32 s0, s1, 4 |
| ; GFX9-NEXT: scratch_store_dword off, v0, s0 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_large_imm_offset_foo: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX10-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: s_add_i32 s1, s32, s0 |
| ; GFX10-NEXT: s_add_i32 s0, s1, 4 |
| ; GFX10-NEXT: scratch_store_dword off, v0, s32 offset:4 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_store_dword off, v1, s0 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, off, s0 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_large_imm_offset_foo: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 13 |
| ; GFX942-NEXT: s_add_i32 s1, s32, s0 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s32 offset:4 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; GFX942-NEXT: s_add_i32 s0, s1, 4 |
| ; GFX942-NEXT: scratch_store_dword off, v0, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_large_imm_offset_foo: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; GFX11-NEXT: s_movk_i32 s0, 0x3e80 |
| ; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) |
| ; GFX11-NEXT: s_add_i32 s1, s32, s0 |
| ; GFX11-NEXT: s_add_i32 s0, s1, 4 |
| ; GFX11-NEXT: scratch_store_b32 off, v0, s32 offset:4 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_store_b32 off, v1, s0 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, off, s0 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_large_imm_offset_foo: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 off, v0, s32 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b32 off, v1, s32 offset:16000 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, off, s32 offset:16000 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_large_imm_offset_foo: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX9-NEXT: s_add_i32 s1, s32, s0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s32 offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX9-NEXT: s_add_i32 s0, s1, 4 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword off, v0, s0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_large_imm_offset_foo: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX10-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: s_add_i32 s1, s32, s0 |
| ; UNALIGNED_GFX10-NEXT: s_add_i32 s0, s1, 4 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v0, s32 offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword off, v1, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, off, s0 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_large_imm_offset_foo: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 13 |
| ; UNALIGNED_GFX942-NEXT: s_add_i32 s1, s32, s0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s32 offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v0, 15 |
| ; UNALIGNED_GFX942-NEXT: s_add_i32 s0, s1, 4 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword off, v0, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_large_imm_offset_foo: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; UNALIGNED_GFX11-NEXT: s_movk_i32 s0, 0x3e80 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: s_add_i32 s1, s32, s0 |
| ; UNALIGNED_GFX11-NEXT: s_add_i32 s0, s1, 4 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v0, s32 offset:4 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 off, v1, s0 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, off, s0 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_large_imm_offset_foo: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v0, 13 :: v_dual_mov_b32 v1, 15 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v0, s32 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 off, v1, s32 offset:16000 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, off, s32 offset:16000 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| %i = alloca [4096 x i32], align 4, addrspace(5) |
| %i1 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 0 |
| store volatile i32 13, ptr addrspace(5) %i1, align 4 |
| %i7 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 4000 |
| store volatile i32 15, ptr addrspace(5) %i7, align 4 |
| %i10 = getelementptr inbounds [4096 x i32], ptr addrspace(5) %i, i32 0, i32 4000 |
| %i12 = load volatile i32, ptr addrspace(5) %i10, align 4 |
| ret void |
| } |
| |
| define amdgpu_kernel void @store_load_vidx_sidx_offset(i32 %sidx) { |
| ; GFX9-LABEL: store_load_vidx_sidx_offset: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX9-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; GFX9-NEXT: scratch_store_dword v0, v1, off offset:1024 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dword v0, v0, off offset:1024 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: store_load_vidx_sidx_offset: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX10-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; GFX10-NEXT: scratch_store_dword v0, v1, off offset:1024 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dword v0, v0, off offset:1024 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: store_load_vidx_sidx_offset: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX942-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; GFX942-NEXT: scratch_store_dword v0, v1, off offset:1024 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dword v0, v0, off offset:1024 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: store_load_vidx_sidx_offset: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_and_b32 v0, 0x3ff, v0 |
| ; GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; GFX11-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; GFX11-NEXT: scratch_store_b32 v0, v1, off offset:1024 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b32 v0, v0, off offset:1024 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: store_load_vidx_sidx_offset: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_and_b32 v0, 0x3ff, v0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; GFX12-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; GFX12-NEXT: scratch_store_b32 v0, v1, off offset:1024 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b32 v0, v0, off offset:1024 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_vidx_sidx_offset: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s8, s13 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s9, 0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v0, v1, off offset:1024 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v0, v0, off offset:1024 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_vidx_sidx_offset: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s8, s8, s13 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s9, s9, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s8 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s9 |
| ; UNALIGNED_GFX10-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v1, off offset:1024 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v0, v0, off offset:1024 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_vidx_sidx_offset: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_load_dword s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX942-NEXT: v_and_b32_e32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v0, v1, off offset:1024 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v0, v0, off offset:1024 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_vidx_sidx_offset: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_and_b32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; UNALIGNED_GFX11-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v1, off offset:1024 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v0, v0, off offset:1024 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_vidx_sidx_offset: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_load_b32 s0, s[4:5], 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_and_b32 v0, 0x3ff, v0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; UNALIGNED_GFX12-NEXT: v_add_lshl_u32 v0, s0, v0, 2 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v1, off offset:1024 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v0, v0, off offset:1024 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %alloca = alloca [32 x i32], align 4, addrspace(5) |
| %vidx = tail call i32 @llvm.amdgcn.workitem.id.x() |
| %add1 = add nsw i32 %sidx, %vidx |
| %gep1 = getelementptr [32 x float], ptr addrspace(5) %alloca, i32 0, i32 %add1 |
| %gep = getelementptr inbounds [32 x float], ptr addrspace(5) %gep1, i32 0, i32 256 |
| store volatile i32 15, ptr addrspace(5) %gep, align 4 |
| %load = load volatile i32, ptr addrspace(5) %gep, align 4 |
| ret void |
| } |
| |
| define void @store_load_i64_aligned(ptr addrspace(5) nocapture %arg) { |
| ; GFX9-LABEL: store_load_i64_aligned: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX9-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_i64_aligned: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX10-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_i64_aligned: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: v_mov_b64_e32 v[2:3], 15 |
| ; GFX942-NEXT: scratch_store_dwordx2 v0, v[2:3], off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dwordx2 v[0:1], v0, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_i64_aligned: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX11-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX11-NEXT: scratch_store_b64 v0, v[1:2], off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b64 v[0:1], v0, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_i64_aligned: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX12-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b64 v0, v[1:2], off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b64 v[0:1], v0, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_i64_aligned: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_i64_aligned: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_i64_aligned: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b64_e32 v[2:3], 15 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dwordx2 v0, v[2:3], off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dwordx2 v[0:1], v0, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_i64_aligned: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b64 v0, v[1:2], off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b64 v[0:1], v0, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_i64_aligned: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b64 v0, v[1:2], off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b64 v[0:1], v0, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| store volatile i64 15, ptr addrspace(5) %arg, align 8 |
| %load = load volatile i64, ptr addrspace(5) %arg, align 8 |
| ret void |
| } |
| |
| define void @store_load_i64_unaligned(ptr addrspace(5) nocapture %arg) { |
| ; GFX9-LABEL: store_load_i64_unaligned: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX9-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_i64_unaligned: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX10-NEXT: scratch_store_dwordx2 v0, v[1:2], off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dwordx2 v[0:1], v0, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_i64_unaligned: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: v_mov_b64_e32 v[2:3], 15 |
| ; GFX942-NEXT: scratch_store_dwordx2 v0, v[2:3], off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dwordx2 v[0:1], v0, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_i64_unaligned: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX11-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX11-NEXT: scratch_store_b64 v0, v[1:2], off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b64 v[0:1], v0, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_i64_unaligned: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX12-NEXT: v_mov_b32_e32 v2, 0 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b64 v0, v[1:2], off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b64 v[0:1], v0, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_i64_unaligned: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:2 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:3 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:5 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:6 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:7 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: ; kill: killed $vgpr0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:7 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_i64_unaligned: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:1 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:2 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:3 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:5 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:6 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:7 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v0, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_i64_unaligned: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: ; kill: killed $vgpr0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_i64_unaligned: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v1, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:1 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:2 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:3 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:4 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:5 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:6 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:7 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v0, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_i64_unaligned: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v1, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v0, v0, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| store volatile i64 15, ptr addrspace(5) %arg, align 1 |
| %load = load volatile i64, ptr addrspace(5) %arg, align 1 |
| ret void |
| } |
| |
| define void @store_load_v3i32_unaligned(ptr addrspace(5) nocapture %arg) { |
| ; GFX9-LABEL: store_load_v3i32_unaligned: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: s_mov_b32 s2, 3 |
| ; GFX9-NEXT: s_mov_b32 s1, 2 |
| ; GFX9-NEXT: s_mov_b32 s0, 1 |
| ; GFX9-NEXT: v_mov_b32_e32 v3, s2 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, s1 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX9-NEXT: scratch_store_dwordx3 v0, v[1:3], off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dwordx3 v[0:2], v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_v3i32_unaligned: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: s_mov_b32 s2, 3 |
| ; GFX10-NEXT: s_mov_b32 s1, 2 |
| ; GFX10-NEXT: s_mov_b32 s0, 1 |
| ; GFX10-NEXT: v_mov_b32_e32 v3, s2 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, s1 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX10-NEXT: scratch_store_dwordx3 v0, v[1:3], off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dwordx3 v[0:2], v0, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_v3i32_unaligned: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: s_mov_b32 s2, 3 |
| ; GFX942-NEXT: s_mov_b32 s1, 2 |
| ; GFX942-NEXT: s_mov_b32 s0, 1 |
| ; GFX942-NEXT: v_mov_b32_e32 v4, s2 |
| ; GFX942-NEXT: v_mov_b32_e32 v3, s1 |
| ; GFX942-NEXT: v_mov_b32_e32 v2, s0 |
| ; GFX942-NEXT: scratch_store_dwordx3 v0, v[2:4], off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dwordx3 v[0:2], v0, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_v3i32_unaligned: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: s_mov_b32 s2, 3 |
| ; GFX11-NEXT: s_mov_b32 s1, 2 |
| ; GFX11-NEXT: s_mov_b32 s0, 1 |
| ; GFX11-NEXT: v_dual_mov_b32 v3, s2 :: v_dual_mov_b32 v2, s1 |
| ; GFX11-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX11-NEXT: scratch_store_b96 v0, v[1:3], off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b96 v[0:2], v0, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_v3i32_unaligned: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_mov_b32 s2, 3 |
| ; GFX12-NEXT: s_mov_b32 s1, 2 |
| ; GFX12-NEXT: s_mov_b32 s0, 1 |
| ; GFX12-NEXT: s_wait_alu 0xfffe |
| ; GFX12-NEXT: v_dual_mov_b32 v3, s2 :: v_dual_mov_b32 v2, s1 |
| ; GFX12-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b96 v0, v[1:3], off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b96 v[0:2], v0, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_v3i32_unaligned: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:2 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:3 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v2, off offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:5 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:6 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:7 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 3 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v2, off offset:8 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:9 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:10 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:11 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: ; kill: killed $vgpr0 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:7 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:8 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:9 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:10 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:11 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_v3i32_unaligned: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:1 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:2 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:3 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v3, off offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:5 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:6 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:7 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v1, off offset:8 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:9 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:10 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:11 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:8 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:9 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:10 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v0, v0, off offset:11 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_v3i32_unaligned: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v2, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 3 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v2, off offset:8 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:9 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:10 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:11 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: ; kill: killed $vgpr0 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:8 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:9 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:10 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:11 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_v3i32_unaligned: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 1 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v1, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:1 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:2 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:3 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v3, off offset:4 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:5 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:6 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:7 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v1, off offset:8 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:9 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:10 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:11 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:8 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:9 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:10 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v0, v0, off offset:11 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_v3i32_unaligned: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 1 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v1, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v3, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v1, off offset:8 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:9 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:10 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:11 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:8 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:9 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:10 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v0, v0, off offset:11 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| store volatile <3 x i32> <i32 1, i32 2, i32 3>, ptr addrspace(5) %arg, align 1 |
| %load = load volatile <3 x i32>, ptr addrspace(5) %arg, align 1 |
| ret void |
| } |
| |
| define void @store_load_v4i32_unaligned(ptr addrspace(5) nocapture %arg) { |
| ; GFX9-LABEL: store_load_v4i32_unaligned: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX9-NEXT: s_mov_b32 s3, 4 |
| ; GFX9-NEXT: s_mov_b32 s2, 3 |
| ; GFX9-NEXT: s_mov_b32 s1, 2 |
| ; GFX9-NEXT: s_mov_b32 s0, 1 |
| ; GFX9-NEXT: v_mov_b32_e32 v4, s3 |
| ; GFX9-NEXT: v_mov_b32_e32 v3, s2 |
| ; GFX9-NEXT: v_mov_b32_e32 v2, s1 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX9-NEXT: scratch_store_dwordx4 v0, v[1:4], off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: scratch_load_dwordx4 v[0:3], v0, off glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX10-LABEL: store_load_v4i32_unaligned: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX10-NEXT: s_mov_b32 s3, 4 |
| ; GFX10-NEXT: s_mov_b32 s2, 3 |
| ; GFX10-NEXT: s_mov_b32 s1, 2 |
| ; GFX10-NEXT: s_mov_b32 s0, 1 |
| ; GFX10-NEXT: v_mov_b32_e32 v4, s3 |
| ; GFX10-NEXT: v_mov_b32_e32 v3, s2 |
| ; GFX10-NEXT: v_mov_b32_e32 v2, s1 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, s0 |
| ; GFX10-NEXT: scratch_store_dwordx4 v0, v[1:4], off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: scratch_load_dwordx4 v[0:3], v0, off glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX942-LABEL: store_load_v4i32_unaligned: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX942-NEXT: s_mov_b32 s3, 4 |
| ; GFX942-NEXT: s_mov_b32 s2, 3 |
| ; GFX942-NEXT: s_mov_b32 s1, 2 |
| ; GFX942-NEXT: s_mov_b32 s0, 1 |
| ; GFX942-NEXT: v_mov_b64_e32 v[4:5], s[2:3] |
| ; GFX942-NEXT: v_mov_b64_e32 v[2:3], s[0:1] |
| ; GFX942-NEXT: scratch_store_dwordx4 v0, v[2:5], off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: scratch_load_dwordx4 v[0:3], v0, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX11-LABEL: store_load_v4i32_unaligned: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; GFX11-NEXT: s_mov_b32 s3, 4 |
| ; GFX11-NEXT: s_mov_b32 s2, 3 |
| ; GFX11-NEXT: s_mov_b32 s1, 2 |
| ; GFX11-NEXT: s_mov_b32 s0, 1 |
| ; GFX11-NEXT: v_dual_mov_b32 v4, s3 :: v_dual_mov_b32 v3, s2 |
| ; GFX11-NEXT: v_dual_mov_b32 v2, s1 :: v_dual_mov_b32 v1, s0 |
| ; GFX11-NEXT: scratch_store_b128 v0, v[1:4], off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: scratch_load_b128 v[0:3], v0, off glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; GFX12-LABEL: store_load_v4i32_unaligned: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; GFX12-NEXT: s_wait_expcnt 0x0 |
| ; GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; GFX12-NEXT: s_mov_b32 s3, 4 |
| ; GFX12-NEXT: s_mov_b32 s2, 3 |
| ; GFX12-NEXT: s_mov_b32 s1, 2 |
| ; GFX12-NEXT: s_mov_b32 s0, 1 |
| ; GFX12-NEXT: s_wait_alu 0xfffe |
| ; GFX12-NEXT: v_dual_mov_b32 v4, s3 :: v_dual_mov_b32 v3, s2 |
| ; GFX12-NEXT: v_dual_mov_b32 v2, s1 :: v_dual_mov_b32 v1, s0 |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_store_b128 v0, v[1:4], off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: scratch_load_b128 v[0:3], v0, off scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX9-LABEL: store_load_v4i32_unaligned: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 2 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:1 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:2 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:3 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v2, off offset:4 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:5 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:6 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:7 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 3 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v2, off offset:8 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:9 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:10 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:11 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v2, 4 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v2, off offset:12 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:13 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:14 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_store_byte v0, v1, off offset:15 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:7 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:8 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:9 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:10 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:11 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:12 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:13 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v1, v0, off offset:14 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: scratch_load_ubyte v0, v0, off offset:15 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX10-LABEL: store_load_v4i32_unaligned: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v2, 0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v1, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:1 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:2 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:3 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v3, off offset:4 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v3, 4 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:5 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:6 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:7 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v1, off offset:8 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:9 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:10 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:11 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v3, off offset:12 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:13 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:14 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_byte v0, v2, off offset:15 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:8 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:9 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:10 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:11 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:12 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:13 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v1, v0, off offset:14 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: scratch_load_ubyte v0, v0, off offset:15 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX942-LABEL: store_load_v4i32_unaligned: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 1 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 2 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v2, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 3 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v2, off offset:8 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:9 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:10 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:11 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v2, 4 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v2, off offset:12 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:13 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:14 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_store_byte v0, v1, off offset:15 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:1 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:2 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:3 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:4 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:5 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:6 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:7 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:8 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:9 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:10 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:11 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:12 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:13 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v1, v0, off offset:14 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: scratch_load_ubyte v0, v0, off offset:15 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX11-LABEL: store_load_v4i32_unaligned: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 1 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v1, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:1 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:2 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:3 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v3, off offset:4 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX11-NEXT: v_mov_b32_e32 v3, 4 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:5 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:6 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:7 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v1, off offset:8 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:9 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:10 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:11 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v3, off offset:12 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:13 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:14 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b8 v0, v2, off offset:15 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:1 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:2 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:3 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:4 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:5 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:6 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:7 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:8 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:9 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:10 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:11 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:12 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:13 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v1, v0, off offset:14 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: scratch_load_u8 v0, v0, off offset:15 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: s_setpc_b64 s[30:31] |
| ; |
| ; UNALIGNED_GFX12-LABEL: store_load_v4i32_unaligned: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt_dscnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_expcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_samplecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_bvhcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_wait_kmcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 1 :: v_dual_mov_b32 v2, 0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v3, 2 |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v1, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v3, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v1, 3 |
| ; UNALIGNED_GFX12-NEXT: v_mov_b32_e32 v3, 4 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v1, off offset:8 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:9 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:10 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:11 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v3, off offset:12 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:13 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:14 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b8 v0, v2, off offset:15 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:1 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:2 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:3 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:4 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:5 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:6 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:7 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:8 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:9 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:10 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:11 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:12 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:13 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v1, v0, off offset:14 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_u8 v0, v0, off offset:15 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_setpc_b64 s[30:31] |
| bb: |
| store volatile <4 x i32> <i32 1, i32 2, i32 3, i32 4>, ptr addrspace(5) %arg, align 1 |
| %load = load volatile <4 x i32>, ptr addrspace(5) %arg, align 1 |
| ret void |
| } |
| |
| define amdgpu_gs void @sgpr_base_large_offset(ptr addrspace(1) %out, ptr addrspace(5) inreg %sgpr_base) { |
| ; GFX9-LABEL: sgpr_base_large_offset: |
| ; GFX9: ; %bb.0: ; %entry |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; GFX9-NEXT: s_add_u32 s0, s2, 0xffe8 |
| ; GFX9-NEXT: scratch_load_dword v2, off, s0 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: sgpr_base_large_offset: |
| ; GFX10: ; %bb.0: ; %entry |
| ; GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; GFX10-NEXT: s_add_u32 s0, s2, 0xffe8 |
| ; GFX10-NEXT: scratch_load_dword v2, off, s0 |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: sgpr_base_large_offset: |
| ; GFX942: ; %bb.0: ; %entry |
| ; GFX942-NEXT: s_add_u32 s0, s0, 0xffe8 |
| ; GFX942-NEXT: scratch_load_dword v2, off, s0 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: sgpr_base_large_offset: |
| ; GFX11: ; %bb.0: ; %entry |
| ; GFX11-NEXT: s_add_u32 s0, s0, 0xffe8 |
| ; GFX11-NEXT: scratch_load_b32 v2, off, s0 |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: sgpr_base_large_offset: |
| ; GFX12: ; %bb.0: ; %entry |
| ; GFX12-NEXT: scratch_load_b32 v2, off, s0 offset:65512 |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: sgpr_base_large_offset: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 s0, s2, 0xffe8 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v2, off, s0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: sgpr_base_large_offset: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s2, 0xffe8 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v2, off, s0 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: sgpr_base_large_offset: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX942-NEXT: s_add_u32 s0, s0, 0xffe8 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v2, off, s0 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: sgpr_base_large_offset: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX11-NEXT: s_add_u32 s0, s0, 0xffe8 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v2, off, s0 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: sgpr_base_large_offset: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v2, off, s0 offset:65512 |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| entry: |
| %large_offset = getelementptr i8, ptr addrspace(5) %sgpr_base, i32 65512 |
| %load = load i32, ptr addrspace(5) %large_offset, align 4 |
| store i32 %load, ptr addrspace(1) %out |
| ret void |
| } |
| |
| define amdgpu_gs void @sgpr_base_large_offset_split(ptr addrspace(1) %out, ptr addrspace(5) inreg %sgpr_base) { |
| ; GFX9-LABEL: sgpr_base_large_offset_split: |
| ; GFX9: ; %bb.0: ; %entry |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; GFX9-NEXT: s_and_b32 s0, s2, -4 |
| ; GFX9-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; GFX9-NEXT: scratch_load_dword v2, off, s0 glc |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: sgpr_base_large_offset_split: |
| ; GFX10: ; %bb.0: ; %entry |
| ; GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; GFX10-NEXT: s_and_b32 s0, s2, -4 |
| ; GFX10-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; GFX10-NEXT: scratch_load_dword v2, off, s0 glc dlc |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: sgpr_base_large_offset_split: |
| ; GFX942: ; %bb.0: ; %entry |
| ; GFX942-NEXT: s_and_b32 s0, s0, -4 |
| ; GFX942-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; GFX942-NEXT: scratch_load_dword v2, off, s0 sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: sgpr_base_large_offset_split: |
| ; GFX11: ; %bb.0: ; %entry |
| ; GFX11-NEXT: s_and_b32 s0, s0, -4 |
| ; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; GFX11-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; GFX11-NEXT: scratch_load_b32 v2, off, s0 glc dlc |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: sgpr_base_large_offset_split: |
| ; GFX12: ; %bb.0: ; %entry |
| ; GFX12-NEXT: s_and_b32 s0, s0, -4 |
| ; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; GFX12-NEXT: s_add_co_u32 s0, s0, 0x100ffe8 |
| ; GFX12-NEXT: scratch_load_b32 v2, off, s0 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: sgpr_base_large_offset_split: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; UNALIGNED_GFX9-NEXT: s_and_b32 s0, s2, -4 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v2, off, s0 glc |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: sgpr_base_large_offset_split: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; UNALIGNED_GFX10-NEXT: s_and_b32 s0, s2, -4 |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v2, off, s0 glc dlc |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: sgpr_base_large_offset_split: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX942-NEXT: s_and_b32 s0, s0, -4 |
| ; UNALIGNED_GFX942-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v2, off, s0 sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: sgpr_base_large_offset_split: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX11-NEXT: s_and_b32 s0, s0, -4 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; UNALIGNED_GFX11-NEXT: s_add_u32 s0, s0, 0x100ffe8 |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v2, off, s0 glc dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: sgpr_base_large_offset_split: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX12-NEXT: s_and_b32 s0, s0, -4 |
| ; UNALIGNED_GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1) |
| ; UNALIGNED_GFX12-NEXT: s_add_co_u32 s0, s0, 0x100ffe8 |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v2, off, s0 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| entry: |
| ;%allignedBase = alloca [33554432 x i8], align 4, addrspace(5) |
| %sgpr_base_i32 = ptrtoint ptr addrspace(5) %sgpr_base to i32 |
| %sgpr_base_i32_align4 = and i32 %sgpr_base_i32, 4294967292 |
| %sgpr_base_align4 = inttoptr i32 %sgpr_base_i32_align4 to ptr addrspace(5) |
| %split_offset = getelementptr inbounds [33554432 x i8], ptr addrspace(5) %sgpr_base_align4, i32 0, i32 16842728 |
| %load = load volatile i32, ptr addrspace(5) %split_offset, align 4 |
| store i32 %load, ptr addrspace(1) %out |
| ret void |
| } |
| |
| define amdgpu_gs void @sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset(ptr addrspace(5) inreg %sgpr_base, i32 inreg %sidx, i32 %vidx) { |
| ; GFX9-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; GFX9-NEXT: v_add_u32_e32 v0, s3, v0 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 0xffe8 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; GFX9-NEXT: v_add3_u32 v0, s2, v0, v1 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: scratch_store_dword v0, v1, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s3, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: v_add3_u32 v0, s2, v0, 0xffe8 |
| ; GFX10-NEXT: scratch_store_dword v0, v1, off |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: v_add_u32_e32 v0, s1, v0 |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 0xffe8 |
| ; GFX942-NEXT: v_add3_u32 v0, s0, v0, v1 |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX942-NEXT: scratch_store_dword v0, v1, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; GFX11-NEXT: v_add3_u32 v0, s0, v0, 0xffe8 |
| ; GFX11-NEXT: scratch_store_b32 v0, v1, off dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v1, s0 offset:65512 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s3, v0 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 0xffe8 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; UNALIGNED_GFX9-NEXT: v_add3_u32 v0, s2, v0, v1 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v0, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s3, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: v_add3_u32 v0, s2, v0, 0xffe8 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v1, off |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v0, s1, v0 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 0xffe8 |
| ; UNALIGNED_GFX942-NEXT: v_add3_u32 v0, s0, v0, v1 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v0, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; UNALIGNED_GFX11-NEXT: v_add3_u32 v0, s0, v0, 0xffe8 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v1, off dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_large_imm_offset: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v1, s0 offset:65512 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %add1 = add nsw i32 %sidx, %vidx |
| %add2 = add nsw i32 %add1, 65512 |
| %gep = getelementptr inbounds [33554432 x i8], ptr addrspace(5) %sgpr_base, i32 0, i32 %add2 |
| store volatile i32 15, ptr addrspace(5) %gep, align 4 |
| ret void |
| } |
| |
| define amdgpu_gs void @sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset(ptr addrspace(5) inreg %sgpr_base, i32 inreg %sidx, i32 %vidx) { |
| ; GFX9-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; GFX9: ; %bb.0: ; %bb |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; GFX9-NEXT: v_add_u32_e32 v0, s3, v0 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; GFX9-NEXT: v_add3_u32 v0, s2, v0, -16 |
| ; GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX9-NEXT: scratch_store_dword v0, v1, off |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; GFX10: ; %bb.0: ; %bb |
| ; GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s3, v0 |
| ; GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX10-NEXT: v_add_nc_u32_e32 v0, s2, v0 |
| ; GFX10-NEXT: scratch_store_dword v0, v1, off offset:-16 |
| ; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; GFX942: ; %bb.0: ; %bb |
| ; GFX942-NEXT: v_add_u32_e32 v0, s1, v0 |
| ; GFX942-NEXT: v_add3_u32 v0, s0, v0, -16 |
| ; GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; GFX942-NEXT: scratch_store_dword v0, v1, off sc0 sc1 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; GFX11: ; %bb.0: ; %bb |
| ; GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; GFX11-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; GFX11-NEXT: scratch_store_b32 v0, v1, off offset:-16 dlc |
| ; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; GFX12: ; %bb.0: ; %bb |
| ; GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; GFX12-NEXT: scratch_store_b32 v0, v1, s0 offset:-16 scope:SCOPE_SYS |
| ; GFX12-NEXT: s_wait_storecnt 0x0 |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; UNALIGNED_GFX9-NEXT: v_add_u32_e32 v0, s3, v0 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; UNALIGNED_GFX9-NEXT: v_add3_u32 v0, s2, v0, -16 |
| ; UNALIGNED_GFX9-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX9-NEXT: scratch_store_dword v0, v1, off |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s3, v0 |
| ; UNALIGNED_GFX10-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX10-NEXT: v_add_nc_u32_e32 v0, s2, v0 |
| ; UNALIGNED_GFX10-NEXT: scratch_store_dword v0, v1, off offset:-16 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX942-NEXT: v_add_u32_e32 v0, s1, v0 |
| ; UNALIGNED_GFX942-NEXT: v_add3_u32 v0, s0, v0, -16 |
| ; UNALIGNED_GFX942-NEXT: v_mov_b32_e32 v1, 15 |
| ; UNALIGNED_GFX942-NEXT: scratch_store_dword v0, v1, off sc0 sc1 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX11-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; UNALIGNED_GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) |
| ; UNALIGNED_GFX11-NEXT: v_add_nc_u32_e32 v0, s0, v0 |
| ; UNALIGNED_GFX11-NEXT: scratch_store_b32 v0, v1, off offset:-16 dlc |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt_vscnt null, 0x0 |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: sgpr_base_plus_sgpr_plus_vgpr_plus_negative_imm_offset: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %bb |
| ; UNALIGNED_GFX12-NEXT: v_dual_mov_b32 v1, 15 :: v_dual_add_nc_u32 v0, s1, v0 |
| ; UNALIGNED_GFX12-NEXT: scratch_store_b32 v0, v1, s0 offset:-16 scope:SCOPE_SYS |
| ; UNALIGNED_GFX12-NEXT: s_wait_storecnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| bb: |
| %add1 = add nsw i32 %sidx, %vidx |
| %add2 = add nsw i32 %add1, -16 |
| %gep = getelementptr inbounds [16 x i8], ptr addrspace(5) %sgpr_base, i32 0, i32 %add2 |
| store volatile i32 15, ptr addrspace(5) %gep, align 4 |
| ret void |
| } |
| |
| define amdgpu_gs void @sgpr_base_negative_offset(ptr addrspace(1) %out, ptr addrspace(5) inreg %scevgep) { |
| ; GFX9-LABEL: sgpr_base_negative_offset: |
| ; GFX9: ; %bb.0: ; %entry |
| ; GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; GFX9-NEXT: s_add_u32 s0, s2, 0xffffffe8 |
| ; GFX9-NEXT: scratch_load_dword v2, off, s0 |
| ; GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX9-NEXT: s_endpgm |
| ; |
| ; GFX10-LABEL: sgpr_base_negative_offset: |
| ; GFX10: ; %bb.0: ; %entry |
| ; GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; GFX10-NEXT: scratch_load_dword v2, off, s2 offset:-24 |
| ; GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX10-NEXT: s_endpgm |
| ; |
| ; GFX942-LABEL: sgpr_base_negative_offset: |
| ; GFX942: ; %bb.0: ; %entry |
| ; GFX942-NEXT: s_add_u32 s0, s0, 0xffffffe8 |
| ; GFX942-NEXT: scratch_load_dword v2, off, s0 |
| ; GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; GFX942-NEXT: s_endpgm |
| ; |
| ; GFX11-LABEL: sgpr_base_negative_offset: |
| ; GFX11: ; %bb.0: ; %entry |
| ; GFX11-NEXT: scratch_load_b32 v2, off, s0 offset:-24 |
| ; GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX11-NEXT: s_endpgm |
| ; |
| ; GFX12-LABEL: sgpr_base_negative_offset: |
| ; GFX12: ; %bb.0: ; %entry |
| ; GFX12-NEXT: scratch_load_b32 v2, off, s0 offset:-24 |
| ; GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; GFX12-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX9-LABEL: sgpr_base_negative_offset: |
| ; UNALIGNED_GFX9: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 flat_scratch_lo, s0, s5 |
| ; UNALIGNED_GFX9-NEXT: s_addc_u32 flat_scratch_hi, s1, 0 |
| ; UNALIGNED_GFX9-NEXT: s_add_u32 s0, s2, 0xffffffe8 |
| ; UNALIGNED_GFX9-NEXT: scratch_load_dword v2, off, s0 |
| ; UNALIGNED_GFX9-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX9-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX9-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX10-LABEL: sgpr_base_negative_offset: |
| ; UNALIGNED_GFX10: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX10-NEXT: s_add_u32 s0, s0, s5 |
| ; UNALIGNED_GFX10-NEXT: s_addc_u32 s1, s1, 0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_LO), s0 |
| ; UNALIGNED_GFX10-NEXT: s_setreg_b32 hwreg(HW_REG_FLAT_SCR_HI), s1 |
| ; UNALIGNED_GFX10-NEXT: scratch_load_dword v2, off, s2 offset:-24 |
| ; UNALIGNED_GFX10-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX10-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX10-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX942-LABEL: sgpr_base_negative_offset: |
| ; UNALIGNED_GFX942: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX942-NEXT: s_add_u32 s0, s0, 0xffffffe8 |
| ; UNALIGNED_GFX942-NEXT: scratch_load_dword v2, off, s0 |
| ; UNALIGNED_GFX942-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX942-NEXT: global_store_dword v[0:1], v2, off |
| ; UNALIGNED_GFX942-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX11-LABEL: sgpr_base_negative_offset: |
| ; UNALIGNED_GFX11: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX11-NEXT: scratch_load_b32 v2, off, s0 offset:-24 |
| ; UNALIGNED_GFX11-NEXT: s_waitcnt vmcnt(0) |
| ; UNALIGNED_GFX11-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX11-NEXT: s_endpgm |
| ; |
| ; UNALIGNED_GFX12-LABEL: sgpr_base_negative_offset: |
| ; UNALIGNED_GFX12: ; %bb.0: ; %entry |
| ; UNALIGNED_GFX12-NEXT: scratch_load_b32 v2, off, s0 offset:-24 |
| ; UNALIGNED_GFX12-NEXT: s_wait_loadcnt 0x0 |
| ; UNALIGNED_GFX12-NEXT: global_store_b32 v[0:1], v2, off |
| ; UNALIGNED_GFX12-NEXT: s_endpgm |
| entry: |
| %scevgep28 = getelementptr i8, ptr addrspace(5) %scevgep, i32 -24 |
| %0 = load i32, ptr addrspace(5) %scevgep28, align 4 |
| store i32 %0, ptr addrspace(1) %out |
| ret void |
| } |
| |
| declare i32 @llvm.amdgcn.workitem.id.x() |