maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

Do some peephole optimizations to remove pointless VMOVs from Neon to integer 

registers that arise from argument shuffling with the soft float ABI. These
instructions are particularly slow on Cortex A8. This fixes one half of
<rdar://problem/8674845>.

llvm-svn: 128759
This commit is contained in:
Cameron Zwarich 2011-04-02 02:40:43 +00:00
parent 8c7bbc09e2
commit 6fe5c29430
2 changed files with 42 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
llvm/lib/Target/ARM/ARMISelLowering.cpp
View File

@ -5561,6 +5561,37 @@ static SDValue PerformVMOVRRDCombine(SDNode *N,
SDValue InDouble = N->getOperand(0);
if (InDouble.getOpcode() == ARMISD::VMOVDRR)
return DCI.CombineTo(N, InDouble.getOperand(0), InDouble.getOperand(1));
// vmovrrd(load f64) -> (load i32), (load i32)
SDNode *InNode = InDouble.getNode();
if (ISD::isNormalLoad(InNode) && InNode->hasOneUse() &&
InNode->getValueType(0) == MVT::f64 &&
InNode->getOperand(1).getOpcode() == ISD::FrameIndex &&
!cast<LoadSDNode>(InNode)->isVolatile()) {
// TODO: Should this be done for non-FrameIndex operands?
LoadSDNode *LD = cast<LoadSDNode>(InNode);
SelectionDAG &DAG = DCI.DAG;
DebugLoc DL = LD->getDebugLoc();
SDValue BasePtr = LD->getBasePtr();
SDValue NewLD1 = DAG.getLoad(MVT::i32, DL, LD->getChain(), BasePtr,
LD->getPointerInfo(), LD->isVolatile(),
LD->isNonTemporal(), LD->getAlignment());
SDValue OffsetPtr = DAG.getNode(ISD::ADD, DL, MVT::i32, BasePtr,
DAG.getConstant(4, MVT::i32));
SDValue NewLD2 = DAG.getLoad(MVT::i32, DL, NewLD1.getValue(1), OffsetPtr,
LD->getPointerInfo(), LD->isVolatile(),
LD->isNonTemporal(),
std::min(4U, LD->getAlignment() / 2));
DAG.ReplaceAllUsesOfValueWith(SDValue(LD, 1), NewLD2.getValue(1));
SDValue Result = DCI.CombineTo(N, NewLD1, NewLD2);
DCI.RemoveFromWorklist(LD);
DAG.DeleteNode(LD);
return Result;
}
return SDValue();
}

11
llvm/test/CodeGen/ARM/fp-arg-shuffle.ll Normal file
View File

@ -0,0 +1,11 @@
; RUN: llc < %s -march=arm -mattr=+neon -float-abi=soft | FileCheck %s
; CHECK: function1
; CHECK-NOT: vmov r
define double @function1(double %a, double %b, double %c, double %d, double %e, double %f) nounwind noinline ssp {
entry:
%call = tail call double @function2(double %f, double %e, double %d, double %c, double %b, double %a) nounwind
ret double %call
}
declare double @function2(double, double, double, double, double, double)