18 messages in org.freebsd.freebsd-currentRe: Pentium optimizations| From | Sent On | Attachments |
|---|---|---|
| Alex | Dec 16, 1997 6:59 pm | |
| Tim Liddelow | Dec 16, 1997 8:05 pm | |
| John S. Dyson | Dec 16, 1997 8:37 pm | |
| Alex | Dec 16, 1997 9:17 pm | |
| Tim Liddelow | Dec 16, 1997 9:36 pm | |
| Scott Michel | Dec 16, 1997 10:02 pm | |
| John S. Dyson | Dec 16, 1997 10:23 pm | |
| Brian Handy | Dec 16, 1997 10:47 pm | |
| John S. Dyson | Dec 16, 1997 11:04 pm | |
| Warner Losh | Dec 16, 1997 11:49 pm | |
| John S. Dyson | Dec 17, 1997 12:04 am | |
| Poul-Henning Kamp | Dec 17, 1997 2:55 am | |
| Warner Losh | Dec 17, 1997 7:09 am | |
| Russell L. Carter | Dec 17, 1997 7:42 am | |
| Eivind Eklund | Dec 17, 1997 10:13 am | |
| Tim Liddelow | Dec 17, 1997 2:26 pm | |
| Doug Rabson | Dec 18, 1997 12:35 pm | |
| John Polstra | Dec 21, 1997 1:35 pm |
| Subject: | Re: Pentium optimizations | |
|---|---|---|
| From: | Eivind Eklund (perh...@yes.no) | |
| Date: | Dec 17, 1997 10:13:03 am | |
| List: | org.freebsd.freebsd-current | |
"John S. Dyson" <dys...@FreeBSD.ORG> writes:
I do have some PPro mods, and they appear to help performance on average. The PPro is a really wierd creature (like the K6.) The darned processor does so much optimization, it appears to be insensitive to code mods. There are areas of reasonable payoffs, and lots of "obvious" optimizations that end up being neutral.
I was working with optimizing assembly code for the PPro a year ago. My experience was that modifying code seldom mattered, except for alignement. Making the tight loops hit 16-byte boundaries roughly doubled the speed.
No other approach made a significant difference, AFAIR (I just supplied ideas and had another programmer implement them). All pairing happened automatically, and touching the cache to make it pre-fetch etc didn't help at all.
Eivind.