6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2025-09-27 16:17:17 +00:00
Code Issues Projects Releases Wiki Activity

Added option to enable generating less precise mad (multiply addition)

Browse Source 
for those architectures that support the instruction.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@67363 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Mon P Wang
2009-03-20 05:06:58 +00:00
parent 654c98c4f3
commit bc65ca8de5
2 changed files with 21 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
include/llvm/Target/TargetOptions.h
View File

@@ -26,6 +26,14 @@ namespace llvm {
/// elimination optimization, this option should disable it. /// elimination optimization, this option should disable it.
extern bool NoFramePointerElim; extern bool NoFramePointerElim;
/// LessPreciseFPMAD - This flag is enabled when the
/// -enable-fp-mad is specified on the command line. When this flag is off
/// (the default), the code generator is not allowed to generate mad
/// (multiply add) if the result is "less precise" than doing those operations
/// individually.
extern bool LessPreciseFPMADOption;
extern bool LessPreciseFPMAD();
/// NoExcessFPPrecision - This flag is enabled when the /// NoExcessFPPrecision - This flag is enabled when the
/// -disable-excess-fp-precision flag is specified on the command line. When /// -disable-excess-fp-precision flag is specified on the command line. When
/// this flag is off (the default), the code generator is allowed to produce /// this flag is off (the default), the code generator is allowed to produce
@@ -39,7 +47,7 @@ namespace llvm {
/// this flag is off (the default), the code generator is not allowed to /// this flag is off (the default), the code generator is not allowed to
/// produce results that are "less precise" than IEEE allows. This includes /// produce results that are "less precise" than IEEE allows. This includes
/// use of X86 instructions like FSIN and FCOS instead of libcalls. /// use of X86 instructions like FSIN and FCOS instead of libcalls.
/// UnsafeFPMath implies FiniteOnlyFPMath. /// UnsafeFPMath implies FiniteOnlyFPMath and LessPreciseFPMAD.
extern bool UnsafeFPMath; extern bool UnsafeFPMath;
/// FiniteOnlyFPMath - This returns true when the -enable-finite-only-fp-math /// FiniteOnlyFPMath - This returns true when the -enable-finite-only-fp-math

12
lib/Target/TargetMachine.cpp
View File

@@ -22,6 +22,7 @@ using namespace llvm;
// //
namespace llvm { namespace llvm {
bool LessPreciseFPMADOption;
bool PrintMachineCode; bool PrintMachineCode;
bool NoFramePointerElim; bool NoFramePointerElim;
bool NoExcessFPPrecision; bool NoExcessFPPrecision;
@@ -59,6 +60,11 @@ DisableExcessPrecision("disable-excess-fp-precision",
cl::location(NoExcessFPPrecision), cl::location(NoExcessFPPrecision),
cl::init(false)); cl::init(false));
static cl::opt<bool, true> static cl::opt<bool, true>
EnableFPMAD("enable-fp-mad",
cl::desc("Enable less precise MAD instructions to be generated"),
cl::location(LessPreciseFPMADOption),
cl::init(false));
static cl::opt<bool, true>
EnableUnsafeFPMath("enable-unsafe-fp-math", EnableUnsafeFPMath("enable-unsafe-fp-math",
cl::desc("Enable optimizations that may decrease FP precision"), cl::desc("Enable optimizations that may decrease FP precision"),
cl::location(UnsafeFPMath), cl::location(UnsafeFPMath),
@@ -198,6 +204,12 @@ void TargetMachine::setCodeModel(CodeModel::Model Model) {
} }
namespace llvm { namespace llvm {
/// LessPreciseFPMAD - This flag return true when -enable-fp-mad option
/// is specified on the command line. When this flag is off(default), the
/// code generator is not allowed to generate mad (multiply add) if the
/// result is "less precise" than doing those operations individually.
bool LessPreciseFPMAD() { return UnsafeFPMath || LessPreciseFPMADOption; }
/// FiniteOnlyFPMath - This returns true when the -enable-finite-only-fp-math /// FiniteOnlyFPMath - This returns true when the -enable-finite-only-fp-math
/// option is specified on the command line. If this returns false (default), /// option is specified on the command line. If this returns false (default),
/// the code generator is not allowed to assume that FP arithmetic arguments /// the code generator is not allowed to assume that FP arithmetic arguments