Intel Itanium 9150M, Manual

Page: 146 / 420

Volume 3: Instruction Reference 3:137
fswap
fswap — Floating-point Swap
Format: (
qp
) fswap
f
1
=
f
2
,
f
3
swap_form F9
(
qp
) fswap.nl
f
1
=
f
2
,
f
3
swap_nl_form F9
(
qp
) fswap.nr
f
1
=
f
2
,
f
3
swap_nr_form F9
Description:
For the swap_form, the left single precision value in FR
f
2
is concatenated with the right
single precision value in FR
f
3
. The concatenated pair is then swapped.
For the swap_nl_form, the left single precision value in FR
f
2
is concatenated with the
right single precision value in FR
f
3
. The concatenated pair is then swapped, and the left
single precision value is negated.
For the swap_nr_form, the left single precision value in FR
f
2
is concatenated with the
right single precision value in FR
f
3
. The concatenated pair is then swapped, and the
right single precision value is negated.
For all forms, the exponent field of FR
f
1
is set to the biased exponent for 2.0
63
(0x1003E) and the sign field of FR
f
1
is set to positive (0).
For all forms, if either FR
f
2
or FR
f
3
is a NaTVal, FR
f
1
is set to NaTVal instead of the
computed result.
Figure 2-18. Floating-point Swap
Figure 2-19. Floating-point Swap Negate Left
81 0 80 64 63 81 0 80 64 63
81 0 80 64 63
FR
f
2
FR
f
3
FR
f
1
31 32
0
1003E
32 31
32 31
81 0 80 64 63 81 0 80 64 63
81 0 80 64 63
FR
f
2
FR
f
3
FR
f
1
31 32
0
1003E
32 31
32 31
Negated Sign Bit
30
62