lkml.org 
[lkml]   [1999]   [Sep]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch] hashtable sizes for icache and dcache
From
Date
>>>>> "A" == Alan Cox <alan@lxorguk.ukuu.org.uk> writes:

>> to the stock kernel to get better performances. It increase the ihash size
>> from 256 buckets to 16000 buckets that means 131k for the ihash. I enalrged also
>> the dhash from 1024 buckets to 16000 buckets as there are really lots of
>> hash entries.

A> That'll be nice on a 4Mb palmtop.

Yep -- Hard-coding a value suitable for a large config is just as
bad (maybe worse) for a small box than the reverse, hard-coding a
value suitable for a small config if for a large config. (uh, did
that come out right?)

Sounds like you found a new configurable(s)... auto-config would be
good, if you can come up with a good heuristic, but manual config
should be possible in any case.

>> I believe it doesn't worth to save 200k of RAM and to go slow (note
>> the binary image won't bloat as the hash is allocated in the .bss
>> section).

A> I don't believe we can go around throwing 200K of ram away. A few of those
A> and its starting to bite 16-32Mb boxes.

The nice thing about making such things configurable, is that you
can also make it smaller than it is now on a tiny config, like a 4MB
palmtop.

--
Scott Henry <scotth@sgi.com> / Help! My disclaimer is missing!
IRIX MTS, / GIGO *really* means: Garbage in, Gospel Out
Silicon Graphics, Inc / http://reality.sgi.com/scotth/

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:54    [W:0.025 / U:1.344 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site