All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mark Zealey <mark@zealos.org>
To: linux-kernel@vger.kernel.org
Subject: Re: Best kernel config for exactly 1GB ram
Date: Sun, 11 Nov 2001 22:33:31 +0000	[thread overview]
Message-ID: <20011111223330.B24030@itsolve.co.uk> (raw)
In-Reply-To: <3BEEE61A.6050002@uhura.rueb.com>
In-Reply-To: <3BEEE61A.6050002@uhura.rueb.com>; from steve@uhura.rueb.com on Sun, Nov 11, 2001 at 02:56:58PM -0600

On Sun, Nov 11, 2001 at 02:56:58PM -0600, Steve Bergman wrote:

> Hi,
> 
> 
> I have just upgraded my athlon 1200 system to 1GB ram.  I am unclear as 
> to how I should configure the kernel for this box.  The config.help says 
>    to say no to "high memory support" if the kernel will not run on a 
> machne with more than 1GB.  When I do this I notice that my available 
> memory as reported by top is ~ 120MB less than if I say I want 4GB 
> support.  I recall that linux reserves some of the address space for 
> itself (I thought it was just 64MB).
> 
> What are the trade offs involved here?  Am I better off sacrificing a 
> bit of the physical memory for reasons of efficiency elsewhere?  When I 
> request support for up to 4GB, what exactly changes with respect to the 
> visible virtual address space that apps see, etc?
> 
> This is a desktop machine, so it's not running Oracle or anything like 
> that.  I seem to recall Linus mentioning that big databases tend to like 
> the large (3GB) virtual address space.

Personally, I'd just leave it at the default no high-mem option.

The kernel will then be able to 'see' about 960MB of the memory, so you loose
about 64MB of it, but it's not worth the kernel using bounce-buffers etc just so
you can get 64MB more memory.

IIRC there was a 2GB patch that just redefined PAGE_OFFSET or something similar,
this means that you could see all the memory, but the max virtual memory a
process could see would be 2 gig (as opposed to 3 gig with the default).

-- 

Mark Zealey
mark@zealos.org
mark@itsolve.co.uk

UL++++>$ G!>(GCM/GCS/GS/GM) dpu? s:-@ a16! C++++>$ P++++>+++++$ L+++>+++++$
!E---? W+++>$ N- !o? !w--- O? !M? !V? !PS !PE--@ PGP+? r++ !t---?@ !X---?
!R- b+ !tv b+ DI+ D+? G+++ e>+++++ !h++* r!-- y--

(www.geekcode.com)

  parent reply	other threads:[~2001-11-11 22:34 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-11-11 20:56 Best kernel config for exactly 1GB ram Steve Bergman
2001-11-11 22:13 ` Xavier Bestel
2001-11-11 22:33 ` Mark Zealey [this message]
  -- strict thread matches above, loose matches on Subject: below --
2001-11-12  6:38 Zwane Mwaikambo
2001-11-12  6:59 ` David Lang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20011111223330.B24030@itsolve.co.uk \
    --to=mark@zealos.org \
    --cc=linux-kernel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.