[llvm] r364794 - [ARM] MVE: support QQPRRegClass and QQQQPRRegClass
Mikhail Maltsev via llvm-commits
llvm-commits at lists.llvm.org
Mon Jul 1 09:05:24 PDT 2019
Author: miyuki
Date: Mon Jul 1 09:05:23 2019
New Revision: 364794
URL: http://llvm.org/viewvc/llvm-project?rev=364794&view=rev
Log:
[ARM] MVE: support QQPRRegClass and QQQQPRRegClass
Summary:
QQPRRegClass and QQQQPRRegClass are used by the
interleaving/deinterleaving loads/stores to represent sequences of
consecutive SIMD registers.
Reviewers: ostannard, simon_tatham, dmgreen
Reviewed By: simon_tatham
Subscribers: javed.absar, kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D64009
Modified:
llvm/trunk/lib/Target/ARM/ARMISelLowering.cpp
Modified: llvm/trunk/lib/Target/ARM/ARMISelLowering.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/ARM/ARMISelLowering.cpp?rev=364794&r1=364793&r2=364794&view=diff
==============================================================================
--- llvm/trunk/lib/Target/ARM/ARMISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/ARM/ARMISelLowering.cpp Mon Jul 1 09:05:23 2019
@@ -1565,8 +1565,9 @@ ARMTargetLowering::getRegClassFor(MVT VT
(void)isDivergent;
// Map v4i64 to QQ registers but do not make the type legal. Similarly map
// v8i64 to QQQQ registers. v4i64 and v8i64 are only used for REG_SEQUENCE to
- // load / store 4 to 8 consecutive D registers.
- if (Subtarget->hasNEON()) {
+ // load / store 4 to 8 consecutive NEON D registers, or 2 to 4 consecutive
+ // MVE Q registers.
+ if (Subtarget->hasNEON() || Subtarget->hasMVEIntegerOps()) {
if (VT == MVT::v4i64)
return &ARM::QQPRRegClass;
if (VT == MVT::v8i64)
More information about the llvm-commits
mailing list