6-80
FLT(P),DFLT(P)
(3) Due to the fact that 32-bit floating decimal point type real numbers are processed by simple
32-bit processing, the number of significant digits is 24 bits if the display is binary and
approximately 7 digits if the display is decimal.
For this reason, if the integer exceeds the range of 16777216 to 16777215 (24-bit BIN
value), errors can be generated in the conversion value.
As for the conversion result, the 25th bit from the upper bit of the integer is always filled with
1 and 26th bit and later bits are truncated.
Operation Error
(1) There are no errors associated with the FLT,DFLT instruction.
Integer
b31 b24 b23 b16
222030030
0 0 001 10100111 011
b15 b8
b7
b0
1 1 101 00011001 110
Result of operation
is 222030032.
Truncation
Always filled with 1
After conversion
372588919
0 0 010 11000110 101 0 1 000 00101110 111
Result of operation
is 372588928.
Truncation
Always filled with 1
b31 b24 b23 b16 b15 b8
b7
b0
To Next Page
Loading...
Need help?
General
