Aucbvax.6972
fa.info-vax
utzoo!decvax!ucbvax!info-vax
Wed Apr 28 12:01:45 1982
VAX Timing Question
>From PRJohnson@MIT-XX Wed Apr 28 11:39:15 1982
Does anyone know why the first of the two following code sequences
is slower on both the 780 and the 750?

Sequence 1:
       addl3   r1,r1,r3
       movw    r1,r3
       bicl2   $0x10000,r3
       addl3   r2,r2,r4
       movw    r2,r4
       bicl2   $0x10000,r4

Sequence 2 (a shuffle of sequence 1):
       addl3   r1,r1,r3
       addl3   r2,r2,r4
       movw    r1,r3
       movw    r2,r4
       bicl2   $0x10000,r3
       bicl2   $0x10000,r4

Do the Vaxes pipeline???

The percentage difference is not much on the 780 (about 1.5%
speed up), but is larger on the 750 (about 3.5%).  Note that the
differences still exist if the code sequences are long word
aligned, and have NOPs inserted to align all the listed
instructions (i.e.: a NOP after each MOVW and each BICL2).

Thanks for any insights.

---Paul Johnson
-------

-----------------------------------------------------------------
gopher://quux.org/ conversion by John Goerzen <[email protected]>
of http://communication.ucsd.edu/A-News/


This Usenet Oldnews Archive
article may be copied and distributed freely, provided:

1. There is no money collected for the text(s) of the articles.

2. The following notice remains appended to each copy:

The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996
Bruce Jones, Henry Spencer, David Wiseman.

You are viewing proxied material from gopher.quux.org. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.