1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-22 20:43:44 +02:00
llvm-mirror/test/CodeGen/AMDGPU/sdwa-gfx9.mir
Sam Kolton 076a1edc25 [AMDGPU] SDWA: add support for GFX9 in peephole pass
Summary:
Added support based on merged SDWA pseudo instructions. Now peephole allow one scalar operand, omod and clamp modifiers.
Added several subtarget features for GFX9 SDWA.
This diff also contains changes from D34026.
Depends D34026

Reviewers: vpykhtin, rampitec, arsenm

Subscribers: kzhuravl, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye

Differential Revision: https://reviews.llvm.org/D34241

llvm-svn: 305986
2017-06-22 06:26:41 +00:00

89 lines
3.3 KiB
YAML

# RUN: llc -march=amdgcn -mcpu=kaveri -run-pass=si-peephole-sdwa -o - %s | FileCheck -check-prefix=CI -check-prefix=GCN %s
# RUN: llc -march=amdgcn -mcpu=fiji -run-pass=si-peephole-sdwa -o - %s | FileCheck -check-prefix=VI -check-prefix=GCN %s
# RUN: llc -march=amdgcn -mcpu=gfx900 -run-pass=si-peephole-sdwa -o - %s | FileCheck -check-prefix=GFX9 -check-prefix=GCN %s
# GCN-LABEL: {{^}}name: add_shr_i32
# GCN: [[SMOV:%[0-9]+]] = S_MOV_B32 123
# CI: [[SHIFT:%[0-9]+]] = V_LSHRREV_B32_e64 16, %{{[0-9]+}}, implicit %exec
# CI: %{{[0-9]+}} = V_ADD_I32_e32 [[SMOV]], killed [[SHIFT]], implicit-def %vcc, implicit %exec
# VI: [[VMOV:%[0-9]+]] = V_MOV_B32_e32 [[SMOV]], implicit %exec
# VI: %{{[0-9]+}} = V_ADD_I32_sdwa 0, [[VMOV]], 0, %{{[0-9]+}}, 0, 6, 0, 6, 5, implicit-def %vcc, implicit %exec
# GFX9: %{{[0-9]+}} = V_ADD_I32_sdwa 0, [[SMOV]], 0, %{{[0-9]+}}, 0, 6, 0, 6, 5, implicit-def %vcc, implicit %exec
---
name: add_shr_i32
tracksRegLiveness: true
registers:
- { id: 0, class: vreg_64 }
- { id: 1, class: vreg_64 }
- { id: 2, class: sreg_64 }
- { id: 3, class: vgpr_32 }
- { id: 4, class: sreg_32_xm0 }
- { id: 5, class: sreg_32_xm0 }
- { id: 6, class: sreg_32 }
- { id: 7, class: sreg_32_xm0 }
- { id: 8, class: sreg_32 }
- { id: 9, class: vgpr_32 }
- { id: 10, class: vgpr_32 }
- { id: 11, class: vgpr_32 }
- { id: 12, class: sreg_32_xm0 }
body: |
bb.0:
liveins: %vgpr0_vgpr1, %vgpr2_vgpr3, %sgpr30_sgpr31
%2 = COPY %sgpr30_sgpr31
%1 = COPY %vgpr2_vgpr3
%0 = COPY %vgpr0_vgpr1
%3 = FLAT_LOAD_DWORD %1, 0, 0, 0, implicit %exec, implicit %flat_scr :: (load 4)
%12 = S_MOV_B32 123
%10 = V_LSHRREV_B32_e64 16, %3, implicit %exec
%11 = V_ADD_I32_e32 %12, killed %10, implicit-def %vcc, implicit %exec
FLAT_STORE_DWORD %0, %11, 0, 0, 0, implicit %exec, implicit %flat_scr :: (store 4)
%sgpr30_sgpr31 = COPY %2
S_SETPC_B64_return %sgpr30_sgpr31
...
# GCN-LABEL: {{^}}name: trunc_shr_f32
# CI: [[SHIFT:%[0-9]+]] = V_LSHRREV_B32_e64 16, %{{[0-9]+}}, implicit %exec
# CI: %{{[0-9]+}} = V_TRUNC_F32_e64 0, killed [[SHIFT]], 1, 2, implicit-def %vcc, implicit %exec
# VI: [[SHIFT:%[0-9]+]] = V_LSHRREV_B32_e64 16, %{{[0-9]+}}, implicit %exec
# VI: %{{[0-9]+}} = V_TRUNC_F32_e64 0, killed [[SHIFT]], 1, 2, implicit-def %vcc, implicit %exec
#GFX9: %{{[0-9]+}} = V_TRUNC_F32_sdwa 0, %{{[0-9]+}}, 1, 2, 6, 0, 5, implicit %exec
---
name: trunc_shr_f32
tracksRegLiveness: true
registers:
- { id: 0, class: vreg_64 }
- { id: 1, class: vreg_64 }
- { id: 2, class: sreg_64 }
- { id: 3, class: vgpr_32 }
- { id: 4, class: sreg_32_xm0 }
- { id: 5, class: sreg_32_xm0 }
- { id: 6, class: sreg_32 }
- { id: 7, class: sreg_32_xm0 }
- { id: 8, class: sreg_32 }
- { id: 9, class: vgpr_32 }
- { id: 10, class: vgpr_32 }
- { id: 11, class: vgpr_32 }
body: |
bb.0:
liveins: %vgpr0_vgpr1, %vgpr2_vgpr3, %sgpr30_sgpr31
%2 = COPY %sgpr30_sgpr31
%1 = COPY %vgpr2_vgpr3
%0 = COPY %vgpr0_vgpr1
%3 = FLAT_LOAD_DWORD %1, 0, 0, 0, implicit %exec, implicit %flat_scr :: (load 4)
%10 = V_LSHRREV_B32_e64 16, %3, implicit %exec
%11 = V_TRUNC_F32_e64 0, killed %10, 1, 2, implicit-def %vcc, implicit %exec
FLAT_STORE_DWORD %0, %11, 0, 0, 0, implicit %exec, implicit %flat_scr :: (store 4)
%sgpr30_sgpr31 = COPY %2
S_SETPC_B64_return %sgpr30_sgpr31