[PATCH] D35014: [X86] PR32755 : Improvement in CodeGen instruction selection for LEAs.
Simon Pilgrim via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Mon Sep 4 08:59:33 PDT 2017
RKSimon added a comment.
Some style comments, but @lsaba 's comments need to be dealt with first.
================
Comment at: include/llvm/CodeGen/MachineInstr.h:1296
MachineRegisterInfo *getRegInfo();
+private:
----------------
(style) newline before private
================
Comment at: include/llvm/CodeGen/SelectionDAG.h:305
+ bool IsDAGPartOfLoop = false;
private:
/// DAGUpdateListener is a friend so it can manipulate the listener stack.
----------------
(style) newline before private
================
Comment at: lib/Target/X86/X86ISelDAGToDAG.cpp:171
+ bool OptForAggressingFolding;
public:
explicit X86DAGToDAGISel(X86TargetMachine &tm, CodeGenOpt::Level OptLevel)
----------------
(style) newline
================
Comment at: lib/Target/X86/X86ISelDAGToDAG.cpp:1169
+ X86ISelAddressMode &AM,
unsigned Depth) {
SDLoc dl(N);
----------------
clang-format? If so, commit it as an NFC change
================
Comment at: lib/Target/X86/X86ISelDAGToDAG.cpp:1178
+ if ((!EnableAggressiveFolding && Depth > 5) ||
+ (EnableAggressiveFolding && Depth > 8) )
return matchAddressBase(N, AM);
----------------
Two hard coded depths like this is weird - better to have a getMaxOperandFoldingDepth() helper?
================
Comment at: lib/Target/X86/X86OptimizeLEAs.cpp:224
};
-}
+} // namespace llvm
----------------
NFC change - just commit it if you want, but don't pollute a patch with it
================
Comment at: lib/Target/X86/X86OptimizeLEAs.cpp:528
char OptimizeLEAPass::ID = 0;
-}
+} // namespace
----------------
NFC change
================
Comment at: lib/Target/X86/X86OptimizeLEAs.cpp:561
+ int MemOpNo =
+ X86II::getMemoryOperandNo(Desc.TSFlags) + X86II::getOperandBias(Desc);
----------------
clang-format? If so, commit it as an NFC change
================
Comment at: lib/Target/X86/X86OptimizeLEAs.cpp:854
const MCInstrDesc &Desc = MI.getDesc();
- int MemOpNo =
- X86II::getMemoryOperandNo(Desc.TSFlags) +
- X86II::getOperandBias(Desc);
+ int MemOpNo = X86II::getMemoryOperandNo(Desc.TSFlags) +
+ X86II::getOperandBias(Desc);
----------------
clang-format? If so, commit it as an NFC change
================
Comment at: lib/Target/X86/X86OptimizeLEAs.cpp:969
+ List.sort(CompareFn);
+ }
+ // Loop over all LEA pairs.
----------------
(style) Remove braces
https://reviews.llvm.org/D35014
More information about the llvm-commits
mailing list