CL-PS7500FE 데이터 시트보기 (PDF) - Cirrus Logic
부품명
상세내역
제조사
CL-PS7500FE Datasheet PDF : 251 Pages
| |||
CL-PS7500FE
System-on-a-Chip for Internet Appliance
ef Suffix Destination Rounding Precision Note
00 S
IEEE Single precision
2
01 D
IEEE Double precision
2
10 E
IEEE Double Extended precision 2
11
trap: undefined instruction
gh suffix
Rounding Mode
00
Round to Nearest (default)
01 P
Round towards Plus Infinity
10 M
Round towards Minus Infinity
11 Z
Round towards Zero
NOTES:
1) Deprecated instruction: included for backwards com-
patibility only.
2) The precision must be specified;
there is no default.
3) These are specified when i = 1.
i Fm
1000
1001
1010
1011
1100
1101
1110
1111
Value Assigned
0.0
1.0
2.0
3.0
4.0
5.0
0.5
10.0
Note
3
3
3
3
3
3
3
3
NOTES:
1) FML, FRD, and FDV are only defined to work with single precision operands. It is not guaranteed that any par-
ticular implementation executes the ‘fast’ instructions any quicker than their respective ‘normal’ versions (MUF,
DVF, and RDF).
2) Directed rounding is done only at the last stage of a SIN, COS, and so on; the intermediate calculations to com-
pute the value are done with round-to-nearest using the full working precision.
3) The URD instruction performs the IEEE round-to-integer-value operation, but can leave its result in an abnormal
unnormalized form. The NRM instruction converts this abnormal result into a proper floating-point value.
4) Direct use of the result of a URD instruction by any instruction other than NRM can produce unexpected results
and should therefore not be done. However, there is an exception to this rule, where a URD result may safely
be preserved and restored by STFE/LDFE or SFM/LFM before being processed by NRM. So there is no need,
for instance, to disable interrupts around a URD/NRM instruction sequence.
5) Similarly, the NRM instruction should only be used on an URD result. Again, use of it on other values can pro-
duce unexpected results.
June 1997
ADVANCE DATA BOOK v2.0
FLOATING-POINT INSTRUCTION SET
177
Share Link: 