caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
be1294de9dd93ae02002bd1aa4b4a8d62e562c21
clang-p2996/llvm/test/CodeGen/AMDGPU/sched-assert-onlydbg-value-empty-region.mir
Austin Kerbow da067ed569 [AMDGPU] Set most sched model resource's BufferSize to one 
Using a BufferSize of one for memory ProcResources will result in better
ILP since it more accurately models the dependencies between memory ops
and their consumers on an in-order processor. After this change, the
scheduler will treat the data edges from loads as blocking so that
stalls are guaranteed when waiting for data to be retreaved from memory.
Since we don't actually track waitcnt here, this should do a better job
at modeling their behavior.

Practically, this means that the scheduler will trigger the 'STALL'
heuristic more often.

This type of change needs to be evaluated experimentally. Preliminary
results are positive.

Fixes: SWDEV-282962

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D114777
2021-12-01 22:31:28 -08:00

120 lines
6.0 KiB
YAML
Raw Blame History

# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -run-pass=machine-scheduler -verify-machineinstrs %s -o - | FileCheck %s
# The sequence of DBG_VALUEs forms a scheduling region with 0 real
# instructions. The RegPressure tracker would end up skipping over any
# debug instructions, so it would point to the instruction
# before/outside of the region, hitting this assert:
# assert((BotRPTracker.getPos() == RegionEnd ||
# (RegionEnd->isDebugInstr() &&
# BotRPTracker.getPos() == priorNonDebug(RegionEnd, RegionBegin))) &&
# "Can't find the region bottom");
---
name: only_dbg_value_sched_region
tracksRegLiveness: true
machineFunctionInfo:
isEntryFunction: true
waveLimiter: true
body: |
; CHECK-LABEL: name: only_dbg_value_sched_region
; CHECK: bb.0:
; CHECK-NEXT: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $vgpr0
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
; CHECK-NEXT: [[DEF:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: [[GLOBAL_LOAD_DWORDX2_:%[0-9]+]]:vreg_64 = GLOBAL_LOAD_DWORDX2 [[DEF]], 0, 0, implicit $exec
; CHECK-NEXT: [[GLOBAL_LOAD_DWORD:%[0-9]+]]:vgpr_32 = GLOBAL_LOAD_DWORD [[DEF]], 8, 0, implicit $exec
; CHECK-NEXT: [[COPY1:%[0-9]+]]:vreg_64 = COPY [[GLOBAL_LOAD_DWORDX2_]]
; CHECK-NEXT: undef %6.sub0:vreg_64 = V_ADD_F32_e32 [[DEF]].sub0, [[COPY1]].sub0, implicit $mode, implicit $exec
; CHECK-NEXT: dead undef %6.sub1:vreg_64 = V_ADD_F32_e32 [[DEF]].sub1, [[COPY1]].sub0, implicit $mode, implicit $exec
; CHECK-NEXT: [[GLOBAL_LOAD_DWORD1:%[0-9]+]]:vgpr_32 = GLOBAL_LOAD_DWORD [[COPY1]], 0, 0, implicit $exec
; CHECK-NEXT: undef %4.sub0:vreg_64 = V_MOV_B32_e32 111, implicit $exec
; CHECK-NEXT: [[DEF1:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: %4.sub1:vreg_64 = V_ADD_U32_e32 [[COPY]], [[COPY]], implicit $exec
; CHECK-NEXT: undef %19.sub1:vreg_64 = V_ADD_F32_e32 [[GLOBAL_LOAD_DWORD]], [[GLOBAL_LOAD_DWORD]], implicit $mode, implicit $exec
; CHECK-NEXT: %19.sub0:vreg_64 = V_ADD_F32_e32 [[GLOBAL_LOAD_DWORD1]], [[GLOBAL_LOAD_DWORDX2_]].sub0, implicit $mode, implicit $exec
; CHECK-NEXT: GLOBAL_STORE_DWORDX2 %19, %4, 32, 0, implicit $exec
; CHECK-NEXT: undef %11.sub0:vreg_64 = GLOBAL_LOAD_DWORD [[DEF1]], 0, 0, implicit $exec
; CHECK-NEXT: [[DEF2:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF3:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF2]].sub0:vreg_64 = GLOBAL_LOAD_DWORD [[DEF3]], 0, 0, implicit $exec
; CHECK-NEXT: %11.sub1:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF4:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF5:%[0-9]+]]:vreg_64 = IMPLICIT_DEF
; CHECK-NEXT: dead %20:vgpr_32 = GLOBAL_LOAD_DWORD %11, 0, 0, implicit $exec
; CHECK-NEXT: dead %21:vgpr_32 = GLOBAL_LOAD_DWORD [[DEF4]], 0, 0, implicit $exec
; CHECK-NEXT: dead %22:vgpr_32 = GLOBAL_LOAD_DWORD [[DEF5]], 0, 0, implicit $exec
; CHECK-NEXT: [[DEF6:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF7:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
; CHECK-NEXT: [[DEF8:%[0-9]+]]:vgpr_32 = IMPLICIT_DEF
; CHECK-NEXT: [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
; CHECK-NEXT: [[V_LSHLREV_B64_e64_:%[0-9]+]]:vreg_64 = V_LSHLREV_B64_e64 2, [[DEF2]], implicit $exec
; CHECK-NEXT: [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
; CHECK-NEXT: S_NOP 0, implicit [[DEF7]], implicit [[V_LSHLREV_B64_e64_]].sub0, implicit [[DEF6]], implicit [[V_MOV_B32_e32_]]
; CHECK-NEXT: GLOBAL_STORE_DWORD [[DEF5]], [[V_MOV_B32_e32_1]], 0, 0, implicit $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.1:
; CHECK-NEXT: successors: %bb.2(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: S_SETREG_IMM32_B32 0, 1, implicit-def $mode, implicit $mode
; CHECK-NEXT: DBG_VALUE
; CHECK-NEXT: DBG_VALUE
; CHECK-NEXT: DBG_VALUE
; CHECK-NEXT: S_SETREG_IMM32_B32 0, 1, implicit-def $mode, implicit $mode
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: bb.2:
; CHECK-NEXT: S_NOP 0, implicit [[COPY]]
; CHECK-NEXT: S_NOP 0, implicit [[DEF8]]
; CHECK-NEXT: S_ENDPGM 0
bb.0:
liveins: $vgpr0
%0:vgpr_32 = COPY $vgpr0
%1:vreg_64 = IMPLICIT_DEF
%2:vreg_64 = GLOBAL_LOAD_DWORDX2 %1, 0, 0, implicit $exec
%3:vgpr_32 = GLOBAL_LOAD_DWORD %1, 8, 0, implicit $exec
undef %4.sub1:vreg_64 = V_ADD_U32_e32 %0, %0, implicit $exec
%4.sub0:vreg_64 = V_MOV_B32_e32 111, implicit $exec
%5:vreg_64 = COPY %2
undef %6.sub0:vreg_64 = V_ADD_F32_e32 %1.sub0, %5.sub0, implicit $mode, implicit $exec
%6.sub1:vreg_64 = V_ADD_F32_e32 %1.sub1, %5.sub0, implicit $mode, implicit $exec
%7:vgpr_32 = GLOBAL_LOAD_DWORD %5, 0, 0, implicit $exec
%8:vreg_64 = IMPLICIT_DEF
%9:vreg_64 = IMPLICIT_DEF
%10:vreg_64 = IMPLICIT_DEF
undef %11.sub1:vreg_64 = IMPLICIT_DEF
%12:vgpr_32 = IMPLICIT_DEF
%13:vgpr_32 = IMPLICIT_DEF
%14:vreg_64 = IMPLICIT_DEF
%15:vreg_64 = IMPLICIT_DEF
%16:vgpr_32 = IMPLICIT_DEF
%17:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
%18:vgpr_32 = V_MOV_B32_e32 0, implicit $exec
undef %19.sub0:vreg_64 = V_ADD_F32_e32 %7, %2.sub0, implicit $mode, implicit $exec
%19.sub1:vreg_64 = V_ADD_F32_e32 %3, %3, implicit $mode, implicit $exec
GLOBAL_STORE_DWORDX2 %19, %4, 32, 0, implicit $exec
%11.sub0:vreg_64 = GLOBAL_LOAD_DWORD %9, 0, 0, implicit $exec
%8.sub0:vreg_64 = GLOBAL_LOAD_DWORD %10, 0, 0, implicit $exec
%20:vgpr_32 = GLOBAL_LOAD_DWORD %11, 0, 0, implicit $exec
%21:vgpr_32 = GLOBAL_LOAD_DWORD %14, 0, 0, implicit $exec
%22:vgpr_32 = GLOBAL_LOAD_DWORD %15, 0, 0, implicit $exec
%23:vreg_64 = V_LSHLREV_B64_e64 2, %8, implicit $exec
S_NOP 0, implicit %13, implicit %23.sub0, implicit %12, implicit %17
GLOBAL_STORE_DWORD %15, %18, 0, 0, implicit $exec
bb.1:
S_SETREG_IMM32_B32 0, 1, implicit-def $mode, implicit $mode
DBG_VALUE
DBG_VALUE
DBG_VALUE
S_SETREG_IMM32_B32 0, 1, implicit-def $mode, implicit $mode
bb.2:
S_NOP 0, implicit %0
S_NOP 0, implicit %16
S_ENDPGM 0
...
Reference in New Issue View Git Blame Copy Permalink