[llvm][MIRVRegNamerUtils] Adding hashing on memoperands.

No more hash collisions for memoperands. Now the MIRCanonicalization
pass shouldn't hit hash collisions when dealing with nearly identical
memory accessing instructions when their memoperands are in fact different.

Differential Revision: https://reviews.llvm.org/D71328
This commit is contained in:
Puyan Lotfi 2019-12-11 01:39:17 -05:00
parent 7e271adb95
commit d952037c89
2 changed files with 53 additions and 0 deletions

View File

@ -72,6 +72,17 @@ std::string VRegRenamer::getInstructionOpcodeHash(MachineInstr &MI) {
SmallVector<unsigned, 16> MIOperands = {MI.getOpcode(), MI.getFlags()};
llvm::transform(MI.uses(), std::back_inserter(MIOperands), GetHashableMO);
for (const auto *Op : MI.memoperands()) {
MIOperands.push_back((unsigned)Op->getSize());
MIOperands.push_back((unsigned)Op->getFlags());
MIOperands.push_back((unsigned)Op->getOffset());
MIOperands.push_back((unsigned)Op->getOrdering());
MIOperands.push_back((unsigned)Op->getAddrSpace());
MIOperands.push_back((unsigned)Op->getSyncScopeID());
MIOperands.push_back((unsigned)Op->getBaseAlignment());
MIOperands.push_back((unsigned)Op->getFailureOrdering());
}
auto HashMI = hash_combine_range(MIOperands.begin(), MIOperands.end());
return std::to_string(HashMI).substr(0, 5);
}

View File

@ -0,0 +1,42 @@
# RUN: llc -march=amdgcn -mcpu=tahiti -run-pass mir-canonicalizer -o - %s | FileCheck %s
--- |
target datalayout = "e-p:32:32-p1:64:64-p2:64:64-p3:32:32-p4:64:64-p5:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64"
define amdgpu_kernel void @f(i32 addrspace(1)* nocapture %arg) {
unreachable
}
...
---
name: f
alignment: 1
registers:
- { id: 0, class: sgpr_64 }
- { id: 1, class: sreg_64_xexec }
- { id: 2, class: sreg_64_xexec }
- { id: 3, class: sreg_64_xexec }
- { id: 4, class: sreg_64_xexec }
- { id: 5, class: sreg_64_xexec }
- { id: 6, class: sreg_64_xexec }
liveins:
- { reg: '$sgpr4_sgpr5', virtual-reg: '%4' }
body: |
bb.0:
liveins: $sgpr4_sgpr5
; CHECK: COPY
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
; CHECK-NEXT: %bb0_{{[0-9]+}}__1:sreg_64_xexec = S_LOAD_DWORDX2_IMM
%0 = COPY $sgpr4_sgpr5
%1 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: (non-temporal dereferenceable invariant load 8 from `i64 addrspace(4)* undef`)
%2 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: ( dereferenceable invariant load 8 from `i64 addrspace(4)* undef`)
%3 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: ( invariant load 8 from `i64 addrspace(4)* undef`)
%4 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: ( load 8 from `i64 addrspace(4)* undef`)
%5 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: ( load 8 from `i64 addrspace(2)* undef`)
%6 = S_LOAD_DWORDX2_IMM %0, 0, 0, 0 :: ( load 8 from `i64 addrspace(1)* undef`)
...