[PATCH] D100885: [AArch64][GlobalISel] Fix regbankselect for G_FCMP with vector destinations
Jessica Paquette via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue Apr 20 13:16:09 PDT 2021
paquette created this revision.
paquette added a reviewer: aemerson.
Herald added subscribers: danielkiss, hiraditya, kristof.beyls, rovka.
paquette requested review of this revision.
Herald added a project: LLVM.
These should always go to a FPR, since they always use the vector registers.
https://reviews.llvm.org/D100885
Files:
llvm/lib/Target/AArch64/GISel/AArch64RegisterBankInfo.cpp
llvm/test/CodeGen/AArch64/GlobalISel/regbank-fcmp.mir
Index: llvm/test/CodeGen/AArch64/GlobalISel/regbank-fcmp.mir
===================================================================
--- /dev/null
+++ llvm/test/CodeGen/AArch64/GlobalISel/regbank-fcmp.mir
@@ -0,0 +1,26 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=aarch64 -run-pass=regbankselect -verify-machineinstrs %s -o - | FileCheck %s
+
+...
+---
+name: vector
+legalized: true
+tracksRegLiveness: true
+body: |
+ bb.0:
+ liveins: $q0, $q1
+ ; Vectors should always end up on a FPR.
+
+ ; CHECK-LABEL: name: vector
+ ; CHECK: liveins: $q0, $q1
+ ; CHECK: %x:fpr(<2 x s64>) = COPY $q0
+ ; CHECK: %y:fpr(<2 x s64>) = COPY $q1
+ ; CHECK: %fcmp:fpr(<2 x s64>) = G_FCMP floatpred(olt), %x(<2 x s64>), %y
+ ; CHECK: $q0 = COPY %fcmp(<2 x s64>)
+ ; CHECK: RET_ReallyLR implicit $q0
+ %x:_(<2 x s64>) = COPY $q0
+ %y:_(<2 x s64>) = COPY $q1
+ %fcmp:_(<2 x s64>) = G_FCMP floatpred(olt), %x:_(<2 x s64>), %y:_
+ $q0 = COPY %fcmp
+ RET_ReallyLR implicit $q0
+...
Index: llvm/lib/Target/AArch64/GISel/AArch64RegisterBankInfo.cpp
===================================================================
--- llvm/lib/Target/AArch64/GISel/AArch64RegisterBankInfo.cpp
+++ llvm/lib/Target/AArch64/GISel/AArch64RegisterBankInfo.cpp
@@ -719,10 +719,15 @@
break;
OpRegBankIdx = {PMI_FirstGPR, PMI_FirstFPR};
break;
- case TargetOpcode::G_FCMP:
- OpRegBankIdx = {PMI_FirstGPR,
+ case TargetOpcode::G_FCMP: {
+ // If the result is a vector, it must use a FPR.
+ AArch64GenRegisterBankInfo::PartialMappingIdx Idx0 =
+ MRI.getType(MI.getOperand(0).getReg()).isVector() ? PMI_FirstFPR
+ : PMI_FirstGPR;
+ OpRegBankIdx = {Idx0,
/* Predicate */ PMI_None, PMI_FirstFPR, PMI_FirstFPR};
break;
+ }
case TargetOpcode::G_BITCAST:
// This is going to be a cross register bank copy and this is expensive.
if (OpRegBankIdx[0] != OpRegBankIdx[1])
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D100885.338974.patch
Type: text/x-patch
Size: 2066 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210420/32d1488d/attachment.bin>
More information about the llvm-commits
mailing list