| 1 |
On Thursday 05 May 2005 19:36, Ned Ludd wrote: |
| 2 |
> How about for 1 telling us about the patch what it does/where to get |
| 3 |
> it/what series/. Why it's beneficial for security/other. |
| 4 |
|
| 5 |
It has no security benefit, and I have no idea why the request was posted here |
| 6 |
instead of gentoo-dev or someplace more general. |
| 7 |
|
| 8 |
It allows a full 1GB of system RAM to be used without HIGHMEM support for up |
| 9 |
to 4GB (or more) enabled in the kernel (the usual limit is around 960MB). |
| 10 |
Supposedly this slightly increases memory usage speed for systems with |
| 11 |
exactly 1GB of RAM as the kernel doesn't have to think about high memory |
| 12 |
paging. |
| 13 |
|
| 14 |
This patch has benefit ONLY to systems with exactly 1GB of RAM. Systems with |
| 15 |
any less won't hit high memory anyways, and systems with any more are going |
| 16 |
to have to use high memory support anyways. |
| 17 |
|
| 18 |
Regardless of the reasons that the LOWMEM limit is where it is with or without |
| 19 |
the patch, this patch only lends miniscule performance increase to a very |
| 20 |
small subset of the user base, and I do not think it should be included in |
| 21 |
ANY production-level kernel until when/if it is in the vanilla sources. |
| 22 |
-love, -nitro, and other ricer kernels are the proper place for testing such |
| 23 |
things. |
| 24 |
|
| 25 |
Cheers, |
| 26 |
-- |
| 27 |
Casey Allen Shobe | SeattleServer, Inc. |
| 28 |
cshobe@×××××××××××××.com | cell 425-443-4653 |
| 29 |
http://www.seattleserver.com |
| 30 |
-- |
| 31 |
gentoo-hardened@g.o mailing list |
Archives