diff options
| author | NAKAMURA Takumi <geek4civic@gmail.com> | 2016-08-05 11:59:49 +0000 |
|---|---|---|
| committer | NAKAMURA Takumi <geek4civic@gmail.com> | 2016-08-05 11:59:49 +0000 |
| commit | f72c663ac56e361e7b5808f64cd0c16d116db526 (patch) | |
| tree | 3bed2867893cfc3931f06d7743630a1e8ee2d131 /llvm/lib/AsmParser/LLLexer.cpp | |
| parent | 2b8c774ce75e3b8e168889a718f8c90776ba589a (diff) | |
| download | llvm-f72c663ac56e361e7b5808f64cd0c16d116db526.zip llvm-f72c663ac56e361e7b5808f64cd0c16d116db526.tar.gz llvm-f72c663ac56e361e7b5808f64cd0c16d116db526.tar.bz2 | |
LLLexer.cpp: Avoid using BitsToDouble() to preserve SNaN like "double 0x7FF4000000000000".
We should not use double (or float) in the LLVM, unless it is really needed. x87 FP register doesn't preserve SNaN to move the value.
FIXME: APFloat() may have the constructor by raw bit.
llvm-svn: 277813
Diffstat (limited to 'llvm/lib/AsmParser/LLLexer.cpp')
| -rw-r--r-- | llvm/lib/AsmParser/LLLexer.cpp | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/llvm/lib/AsmParser/LLLexer.cpp b/llvm/lib/AsmParser/LLLexer.cpp index 980ae1ec..c118ecf 100644 --- a/llvm/lib/AsmParser/LLLexer.cpp +++ b/llvm/lib/AsmParser/LLLexer.cpp @@ -871,7 +871,8 @@ lltok::Kind LLLexer::Lex0x() { // HexFPConstant - Floating point constant represented in IEEE format as a // hexadecimal number for when exponential notation is not precise enough. // Half, Float, and double only. - APFloatVal = APFloat(BitsToDouble(HexIntToVal(TokStart + 2, CurPtr))); + APFloatVal = APFloat(APFloat::IEEEdouble, + APInt(64, HexIntToVal(TokStart + 2, CurPtr))); return lltok::APFloat; } |