forked from OSchip/llvm-project
Clarified operator precedence.
Silenced VC++ warning. llvm-svn: 43372
This commit is contained in:
parent
d57aca25fc
commit
fc69d322f2
|
@ -1230,7 +1230,7 @@ APFloat::addOrSubtractSpecials(const APFloat &rhs, bool subtract)
|
|||
case convolve(fcInfinity, fcInfinity):
|
||||
/* Differently signed infinities can only be validly
|
||||
subtracted. */
|
||||
if(sign ^ rhs.sign != subtract) {
|
||||
if((sign ^ rhs.sign) != subtract) {
|
||||
makeNaN();
|
||||
return opInvalidOp;
|
||||
}
|
||||
|
@ -1252,7 +1252,7 @@ APFloat::addOrSubtractSignificand(const APFloat &rhs, bool subtract)
|
|||
|
||||
/* Determine if the operation on the absolute values is effectively
|
||||
an addition or subtraction. */
|
||||
subtract ^= (sign ^ rhs.sign);
|
||||
subtract ^= (sign ^ rhs.sign) ? true : false;
|
||||
|
||||
/* Are we bigger exponent-wise than the RHS? */
|
||||
bits = exponent - rhs.exponent;
|
||||
|
|
Loading…
Reference in New Issue