llvm.org GIT mirror llvm / release_90 test / CodeGen / AMDGPU / llvm.amdgcn.ds.gws.sema.br.ll
release_90

Tree @release_90 (Download .tar.gz)

llvm.amdgcn.ds.gws.sema.br.ll @release_90raw · history · blame

; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=tahiti -o - -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,LOOP %s
; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=hawaii -o - -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,LOOP %s
; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -o - -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,LOOP %s
; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -o - -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,NOLOOP %s
; RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -o - -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,NOLOOP,GFX10 %s

; GCN-LABEL: {{^}}gws_sema_br_offset0:
; NOLOOP-DAG: s_load_dword [[BAR_NUM:s[0-9]+]]
; NOLOOP-DAG: s_mov_b32 m0, -1{{$}}
; NOLOOP: v_mov_b32_e32 v0, [[BAR_NUM]]
; NOLOOP: ds_gws_sema_br v0 offset:1 gds{{$}}

; LOOP: [[LOOP:BB[0-9]+_[0-9]+]]:
; LOOP-NEXT: s_setreg_imm32_b32 hwreg(HW_REG_TRAPSTS, 8, 1), 0
; LOOP-NEXT: ds_gws_sema_br v0 offset:1 gds
; LOOP-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; LOOP-NEXT: s_getreg_b32 [[GETREG:s[0-9]+]], hwreg(HW_REG_TRAPSTS, 8, 1)
; LOOP-NEXT: s_cmp_lg_u32 [[GETREG]], 0
; LOOP-NEXT: s_cbranch_scc1 [[LOOP]]
define amdgpu_kernel void @gws_sema_br_offset0(i32 %val) #0 {
  call void @llvm.amdgcn.ds.gws.sema.br(i32 %val, i32 0)
  ret void
}

declare void @llvm.amdgcn.ds.gws.sema.br(i32, i32) #0

attributes #0 = { convergent inaccessiblememonly nounwind }