How does the BSWAP instruction "speed up decimal arithmetic"?

The Intel Software Development Guide, Volume 1, Section 7.3.1.2, states that the BSWAP team "speeds up decimal arithmetic." He does not explain how this happens, and Google does not help either.

Can someone enlighten me on this?

+4
source share
2 answers

This is a little discarded comment, isn't it?

BSWAP (byte swap) 32- . 0-7 24 31, 8 15 16 23. , . BSWAP "big-endian" "little-endian". . ( XCHG .)

, , , . , .

, doco Intel, , : -)


, , (, , , ).

, , , , , . , , ( ) {0x01, 0x02, 0x03, 0x04} 32- (dword) 0x04030201.

bswap dword, bswap, .

, ( (a)), Intel , . 1, 2a, 2b, 2c, 3a, 3b 3c bswap .


(a) Intel "" , , daa/das BCD ( ) , -BCD-. , - , .

, , Intel IDFPL, , IEEE754, . .

+6

, , , . 4 , BSWAP . , , , , , 8 , . BSWAP , . , - 32- , , BSWAP .

+2

All Articles