RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-31 20:51:52 +01:00
Code Issues Projects Releases Wiki Activity

[AArch64][GlobalISel]: Fix a crash in GlobalIsel in dealing with 16bit uadd.with.overflow.

Browse Source 
Summary: AArch64 doesn't support uadd.with.overflow.i16 natively. This change adds a legalization rule to convert the 32bit add result to 16bit. This should fix PR43981.

Reviewers: arsenm, qcolombet, paquette, aemerson

Reviewed By: paquette

Subscribers: wdng, rovka, kristof.beyls, hiraditya, Petar.Avramovic, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D71587
This commit is contained in:
Xiaoqing Wu 2019-12-17 15:41:06 -08:00 committed by Mark Lacey
parent c108d18eaa
commit c4f9702f3a
2 changed files with 38 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
lib/Target/AArch64/AArch64LegalizerInfo.cpp
View File

@ -143,7 +143,8 @@ AArch64LegalizerInfo::AArch64LegalizerInfo(const AArch64Subtarget &ST) {
getActionDefinitionsBuilder({G_SMULH, G_UMULH}).legalFor({s32, s64});
getActionDefinitionsBuilder({G_UADDE, G_USUBE, G_SADDO, G_SSUBO, G_UADDO})
.legalFor({{s32, s1}, {s64, s1}});
.legalFor({{s32, s1}, {s64, s1}})
.minScalar(0, s32);
getActionDefinitionsBuilder({G_FADD, G_FSUB, G_FMUL, G_FDIV, G_FNEG})
.legalFor({s32, s64, v2s64, v4s32, v2s32});

36
test/CodeGen/AArch64/legalize-uaddo.mir Normal file
View File

@ -0,0 +1,36 @@
# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc %s -run-pass legalizer -march=arm64 -global-isel=1 -verify-machineinstrs -o - | FileCheck %s
...
---
name: legalize-uaddo
alignment: 4
tracksRegLiveness: true
body: |
bb.1:
liveins: $x0
; CHECK-LABEL: name: legalize-uaddo
; CHECK: liveins: $x0
; CHECK: [[COPY:%[0-9]+]]:_(s64) = COPY $x0
; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
; CHECK: [[TRUNC:%[0-9]+]]:_(s32) = G_TRUNC [[COPY]](s64)
; CHECK: [[AND:%[0-9]+]]:_(s32) = G_AND [[TRUNC]], [[C]]
; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 2
; CHECK: [[ADD:%[0-9]+]]:_(s32) = G_ADD [[AND]], [[C1]]
; CHECK: [[AND1:%[0-9]+]]:_(s32) = G_AND [[ADD]], [[C]]
; CHECK: [[ICMP:%[0-9]+]]:_(s32) = G_ICMP intpred(ne), [[ADD]](s32), [[AND1]]
; CHECK: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 1
; CHECK: [[ANYEXT:%[0-9]+]]:_(s64) = G_ANYEXT [[ICMP]](s32)
; CHECK: [[AND2:%[0-9]+]]:_(s64) = G_AND [[ANYEXT]], [[C2]]
; CHECK: $x0 = COPY [[AND2]](s64)
; CHECK: RET_ReallyLR implicit $x0
%0:_(s64) = COPY $x0
%4:_(s16) = G_CONSTANT i16 2
%1:_(s16) = G_TRUNC %0(s64)
%2:_(s16), %3:_(s1) = G_UADDO %1, %4
%5:_(s64) = G_ZEXT %3(s1)
$x0 = COPY %5(s64)
RET_ReallyLR implicit $x0
...