lkml.org 
[lkml]   [1996]   [Nov]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: IP Checksumming

On 21 Nov 1996, Tom May wrote:

>
> Explain why my implementation of your code is not as fast as you say
> it should be. Also note that the timings in the manual are very slow:
> lodsw takes 5 cycles on a 486 + one cycle for the size prefix.

I'm not sure that this is true.

> The risciness of the 486 architecture is responsible for much of the
> performance improvements over the 386. Combinations of simple single
> cycle instructions execute more quickly with greater flexibility than
> the complex instructions which use dedicated registers. Intel
> themselves point this out in Appendix G of the i486 Microprocessor
> Programmer's Reference Manual.

Yes but they don't show or imply that there would be the difference
that you show.

>
> Just because you've used your code for many years is no reason to hold
> it above suspicion.

This is one of those "ya...but". Situations. The code to which I refer
transfers image data on a CAT-Scanner via TCP/IP Ethernet at 90 percent
of the theoretical 10 megabytes/second bandwidth. The TCP/IP implimentation
is complete, i.e., no short-cuts, with the top-level interface being
Berkeley Sockets. The machines are 486/DX-66. I tested most of the
routines on a 586 so I could tune cycle-counts.

Of course there IS a difference. This TCP/IP implementation only has
ONE Client and ONE Server so there are never any collisions. There are,
however, the usual number of retransmissions because the SNICS (serial
network interface chips) are usually faster than the software so some
packets do get lost.

>
> Would you consider performance testing the Linux code with your
> hardware if I converted it to Intel format?
>

I certainly would do anything I could to help. I have available
private fpt, etc., so we could send things back and forth without
involving the whole world. I can write/rewrite Intel format stuff
in my sleep, do cycle-count testing, send you both the results and
the actual code that produces the results. I am an Engineer so I
don't care what the results are. I think we both should KNOW whatever
the true results of the tests are and let the chips fall where they
are.

I have certainly been wrong a few times.... Maybe even ... naw <grin>.

Anon ftp on boneserver.analogic.com is a good "mailbox". You can send
me private mail to let me know what's there.....

Cheers,
Dick Johnson
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Richard B. Johnson
Project Engineer
Analogic Corporation
Voice : (508) 977-3000 ext. 3754
Fax : (508) 532-6097
Modem : (508) 977-6870
Ftp : ftp@boneserver.analogic.com
Email : rjohnson@analogic.com, johnson@analogic.com
Penguin : Linux version 2.1.11 on an i586 machine.
Warning : It's hard to remain at the trailing edge of technology.
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


\
 
 \ /
  Last update: 2005-03-22 13:38    [W:0.192 / U:1.016 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site