Intel ITANIUM ARCHITECTURE - SOFTWARE DEVELOPERS MANUAL VOLUME 1 REV 2.3 Manual page 1485

FSUB/FSUBP/FISUB—Subtract (Continued)
Table 2-9.
DEST
Notes:
Fmeans finite-real number.
Imeans integer.
*indicates floating-point invalid-arithmetic-operand (#IA) exception.
Operation
IF instruction is FISUB
THEN
DEST  DEST  ConvertExtendedReal(SRC);
ELSE (* source operand is real number *)
DEST  DEST  SRC;
FI;
IF instruction = FSUBP
THEN
PopRegisterStack
FI;
FPU Flags Affected
C1
C0, C2, C3
Floating-point Exceptions
#IS
#IA
#D
#U
#O
#P
Volume 4: Base IA-32 Instruction Reference
FSUB Zeros and NaNs
  F or  I
 
*
 F  F or  0
+
 0  SRC
+
 SRC
+ 0
+
+ F + + F
+  + 
+
NaN NaN NaN
Set to 0 if stack underflow occurred.
Indicates rounding direction if the inexact-result exception (#P) fault
is generated: 0 = not roundup; 1 = roundup.
Undefined.
Stack underflow occurred.
Operand is an SNaN value or unsupported format.
Operands are infinities of like sign.
Source operand is a denormal value.
Result is too small for destination format.
Result is too large for destination format.
Value cannot be represented exactly in destination format.
SRC
 0
+ 0 + F or +I
  
 F
DEST DEST
 0  0  SRC
 0  SRC
+0
 F or  0
DEST DEST
+ + +
NaN NaN NaN
+ 
NaN

NaN

NaN

NaN

NaN

NaN
* NaN
NaN NaN
4:183