6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-12-14 11:32:34 +00:00
Code Issues Projects Releases Wiki Activity

Make ARM and X86 LowerMEMCPY identical by moving the isThumb check into getMaxInlineSizeThreshold

Browse Source 
and by restructuring the X86 version.

New I just have to move this to a common place :-)



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@43554 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Rafael Espindola 2007-10-31 14:39:58 +00:00
parent a24b294e74
commit e0703c84dd
4 changed files with 16 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
lib/Target/ARM/ARMISelLowering.cpp
View File

@ -1287,8 +1287,7 @@ static SDOperand LowerSRx(SDOperand Op, SelectionDAG &DAG,
return DAG.getNode(ISD::BUILD_PAIR, MVT::i64, Lo, Hi);
}
SDOperand ARMTargetLowering::LowerMEMCPY(SDOperand Op, SelectionDAG &DAG,
const ARMSubtarget *ST) {
SDOperand ARMTargetLowering::LowerMEMCPY(SDOperand Op, SelectionDAG &DAG) {
SDOperand ChainOp = Op.getOperand(0);
SDOperand DestOp = Op.getOperand(1);
SDOperand SourceOp = Op.getOperand(2);
@ -1311,11 +1310,9 @@ SDOperand ARMTargetLowering::LowerMEMCPY(SDOperand Op, SelectionDAG &DAG,
// The libc version is likely to be faster for the these cases. It can
// use the address value and run time information about the CPU.
// With glibc 2.6.1 on a core 2, coping an array of 100M longs was 30% faster
// FIXME: For now, we don't lower memcpy's to loads / stores for Thumb. Change
// this once Thumb ldmia / stmia support is added.
unsigned Size = I->getValue();
if (AlwaysInline ||
(!ST->isThumb() && Size <= Subtarget->getMaxInlineSizeThreshold() &&
(Size <= Subtarget->getMaxInlineSizeThreshold() &&
(Align & 3) == 0))
return LowerMEMCPYInline(ChainOp, DestOp, SourceOp, Size, Align, DAG);
return LowerMEMCPYCall(ChainOp, DestOp, SourceOp, CountOp, DAG);
@ -1461,7 +1458,7 @@ SDOperand ARMTargetLowering::LowerOperation(SDOperand Op, SelectionDAG &DAG) {
case ISD::RETURNADDR: break;
case ISD::FRAMEADDR: break;
case ISD::GLOBAL_OFFSET_TABLE: return LowerGLOBAL_OFFSET_TABLE(Op, DAG);
case ISD::MEMCPY: return LowerMEMCPY(Op, DAG, Subtarget);
case ISD::MEMCPY: return LowerMEMCPY(Op, DAG);
}
return SDOperand();
}

3
lib/Target/ARM/ARMISelLowering.h
View File

@ -134,8 +134,7 @@ namespace llvm {
SDOperand LowerGLOBAL_OFFSET_TABLE(SDOperand Op, SelectionDAG &DAG);
SDOperand LowerFORMAL_ARGUMENTS(SDOperand Op, SelectionDAG &DAG);
SDOperand LowerBR_JT(SDOperand Op, SelectionDAG &DAG);
SDOperand LowerMEMCPY(SDOperand Op, SelectionDAG &DAG,
const ARMSubtarget *ST);
SDOperand LowerMEMCPY(SDOperand Op, SelectionDAG &DAG);
SDOperand LowerMEMCPYCall(SDOperand Chain, SDOperand Dest,
SDOperand Source, SDOperand Count,
SelectionDAG &DAG);

6
lib/Target/ARM/ARMSubtarget.h
View File

@ -62,7 +62,11 @@ protected:
///
ARMSubtarget(const Module &M, const std::string &FS, bool thumb);
unsigned getMaxInlineSizeThreshold() const { return 64; }
unsigned getMaxInlineSizeThreshold() const {
// FIXME: For now, we don't lower memcpy's to loads / stores for Thumb.
// Change this once Thumb ldmia / stmia support is added.
return isThumb() ? 0 : 64;
}
/// ParseSubtargetFeatures - Parses features string setting specified
/// subtarget options. Definition of function is auto generated by tblgen.
void ParseSubtargetFeatures(const std::string &FS, const std::string &CPU);

21
lib/Target/X86/X86ISelLowering.cpp
View File

@ -4496,24 +4496,17 @@ SDOperand X86TargetLowering::LowerMEMCPY(SDOperand Op, SelectionDAG &DAG) {
assert(!AlwaysInline && "Cannot inline copy of unknown size");
return LowerMEMCPYCall(ChainOp, DestOp, SourceOp, CountOp, DAG);
}
unsigned Size = I->getValue();
if (AlwaysInline)
return LowerMEMCPYInline(ChainOp, DestOp, SourceOp, Size, Align, DAG);
// If not DWORD aligned or if size is more than threshold, then call memcpy.
// The libc version is likely to be faster for the following cases. It can
// use the address value and run time information about the CPU.
// With glibc 2.6.1 on a core 2, coping an array of 100M longs was 30% faster
// If not DWORD aligned, call memcpy.
if ((Align & 3) != 0)
return LowerMEMCPYCall(ChainOp, DestOp, SourceOp, CountOp, DAG);
// If size is more than the threshold, call memcpy.
if (Size > Subtarget->getMaxInlineSizeThreshold())
return LowerMEMCPYCall(ChainOp, DestOp, SourceOp, CountOp, DAG);
return LowerMEMCPYInline(ChainOp, DestOp, SourceOp, Size, Align, DAG);
unsigned Size = I->getValue();
if (AlwaysInline ||
(Size <= Subtarget->getMaxInlineSizeThreshold() &&
(Align & 3) == 0))
return LowerMEMCPYInline(ChainOp, DestOp, SourceOp, Size, Align, DAG);
return LowerMEMCPYCall(ChainOp, DestOp, SourceOp, CountOp, DAG);
}
SDOperand X86TargetLowering::LowerMEMCPYCall(SDOperand Chain,