Files
clang-p2996/llvm/test/CodeGen/AMDGPU/wait-before-stores-with-scope_sys.mir
Petar Avramovic 3e35ba53e2 AMDGPU/GFX12: Insert waitcnts before stores with scope_sys (#82996)
Insert waitcnts for loads and atomics before stores with system scope.
Scope is field in instruction encoding and corresponds to desired
coherence level in cache hierarchy.
Intrinsic stores can set scope in cache policy operand.
If volatile keyword is used on generic stores memory legalizer will set
scope to system. Generic stores, by default, get lowest scope level.
Waitcnts are not required if it is guaranteed that memory is cached.
For example vulkan shaders can guarantee this.
TODO: implement flag for frontends to give us a hint not to insert
waits.
Expecting vulkan flag to be implemented as vulkan:private MMRA.
2024-02-28 16:18:04 +01:00

44 lines
2.0 KiB
YAML

# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 4
# RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -run-pass=si-memory-legalizer %s -o - | FileCheck -check-prefix=GFX12 %s
---
name: intrinsic_store_system_scope
body: |
bb.0:
liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4, $vgpr0, $vgpr1, $vgpr2
; GFX12-LABEL: name: intrinsic_store_system_scope
; GFX12: liveins: $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4, $vgpr0, $vgpr1, $vgpr2
; GFX12-NEXT: {{ $}}
; GFX12-NEXT: S_WAIT_LOADCNT_soft 0
; GFX12-NEXT: S_WAIT_SAMPLECNT_soft 0
; GFX12-NEXT: S_WAIT_BVHCNT_soft 0
; GFX12-NEXT: S_WAIT_KMCNT_soft 0
; GFX12-NEXT: S_WAIT_STORECNT_soft 0
; GFX12-NEXT: BUFFER_STORE_DWORD_VBUFFER_BOTHEN_exact killed renamable $vgpr0, killed renamable $vgpr1_vgpr2, killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, killed renamable $sgpr4, 0, 24, 0, implicit $exec :: (dereferenceable store (s32), align 1, addrspace 8)
; GFX12-NEXT: S_ENDPGM 0
BUFFER_STORE_DWORD_VBUFFER_BOTHEN_exact killed renamable $vgpr0, killed renamable $vgpr1_vgpr2, killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, killed renamable $sgpr4, 0, 24, 0, implicit $exec :: (dereferenceable store (s32), align 1, addrspace 8)
S_ENDPGM 0
...
---
name: generic_store_volatile
body: |
bb.0:
liveins: $vgpr0, $vgpr1, $vgpr2
; GFX12-LABEL: name: generic_store_volatile
; GFX12: liveins: $vgpr0, $vgpr1, $vgpr2
; GFX12-NEXT: {{ $}}
; GFX12-NEXT: S_WAIT_LOADCNT_soft 0
; GFX12-NEXT: S_WAIT_SAMPLECNT_soft 0
; GFX12-NEXT: S_WAIT_BVHCNT_soft 0
; GFX12-NEXT: S_WAIT_KMCNT_soft 0
; GFX12-NEXT: S_WAIT_STORECNT_soft 0
; GFX12-NEXT: GLOBAL_STORE_DWORD killed renamable $vgpr1_vgpr2, killed renamable $vgpr0, 0, 24, implicit $exec :: (volatile store (s32), addrspace 1)
; GFX12-NEXT: S_WAIT_STORECNT_soft 0
; GFX12-NEXT: S_ENDPGM 0
GLOBAL_STORE_DWORD killed renamable $vgpr1_vgpr2, killed renamable $vgpr0, 0, 0, implicit $exec :: (volatile store (s32), addrspace 1)
S_ENDPGM 0
...