From: Jack Steiner <steiner@sgi.com>
To: Christoph Lameter <clameter@sgi.com>
Cc: Jes Sorensen <jes@sgi.com>,
Arjan van de Ven <arjan@infradead.org>,
linux-mm@kvack.org, Nick Piggin <nickpiggin@yahoo.com.au>,
Christoph Hellwig <hch@infradead.org>,
linux-ia64@vger.kernel.org, Marcelo Tosatti <marcelo@kvack.org>,
Martin Bligh <mbligh@google.com>,
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>,
Andi Kleen <ak@suse.de>
Subject: Re: [PATCH 0/8] Optional ZONE_DMA V1
Date: Wed, 13 Sep 2006 17:49:48 +0000 [thread overview]
Message-ID: <20060913174948.GA6533@sgi.com> (raw)
In-Reply-To: <Pine.LNX.4.64.0609131015360.17927@schroedinger.engr.sgi.com>
On Wed, Sep 13, 2006 at 10:23:10AM -0700, Christoph Lameter wrote:
> On Wed, 13 Sep 2006, Jes Sorensen wrote:
>
> > > There was the floppy driver and one type of USB stick that I noticed
> > > during the work on the project. But other drivers may depend also depend
> > > indirectly on DMA functionality and may also be disabled.
> >
> > Ok, USB should ring alarm bells, floppy I think is less relevant these
> > days :)
>
> If you want all drivers then you must of course have ZONE_DMA.
> Distributions that want to cover all drivers will have it on by default
> and ZONE_DMA is available by default.
>
> However, if you want to create a lean and mean kernel then you can switch
> ZONE_DMA off and if there is just one zone left then the VM can
> optimized much better because loops are avoided and some macros
> become constant etc etc.
>
> Some architectures never need ZONE_DMA because all hardware supports DMA
> to all of memory. SGI Altix is one example. Carrying an additional
> useless zone around unecessarily bloats the kernel both in term of code
> and data. Data is a particular issue since zones contain per cpu elements.
> For a 1k cpu 1k node configuration this saves around 1 million per cpu
> structures (one zone per node with 1k per cpu pagesets).
>
Most distros release GENERIC kernels for IA64. If _any_ IA64 platform requires
ZONE_DMA, then it must be configured ON.
Two questions:
- will any IA64 platform require that ZONE_DMA be enabled (I think
the answer is "yes")
- if ZONE_DMA is enabled, ALTIX will still use only 1 zone. In your
statement above, you say that disabling ZONE_DMA save 1M cpu
structures. If ZONE_DMA is enabled, will these 1M structure be allocated
on SN even though they are not needed?
-- jack
next prev parent reply other threads:[~2006-09-13 17:49 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20060918183614.19679.50359.sendpatchset@schroedinger.engr.sgi.com>
2006-09-11 22:27 ` [PATCH 0/8] Optional ZONE_DMA V1 Christoph Lameter
2006-09-11 22:27 ` [PATCH 2/6] Introduce CONFIG_ZONE_DMA Christoph Lameter
2006-09-18 13:55 ` Christoph Hellwig
2006-09-18 14:08 ` Martin Schwidefsky
2006-09-18 17:28 ` Christoph Lameter
2006-09-19 8:03 ` Martin Schwidefsky
2006-09-18 14:42 ` Russell King
2006-09-18 14:58 ` James Bottomley
2006-09-18 17:30 ` Christoph Lameter
2006-09-18 15:22 ` Paul Mundt
2006-09-18 17:33 ` Christoph Lameter
2006-09-18 22:45 ` Paul Mundt
2006-09-18 22:58 ` Christoph Lameter
2006-09-18 23:25 ` Paul Mundt
2006-09-11 22:27 ` [PATCH 3/6] Optional ZONE_DMA in the VM Christoph Lameter
2006-09-12 0:35 ` Nick Piggin
2006-09-12 1:40 ` Christoph Lameter
2006-09-12 2:40 ` Nick Piggin
2006-09-12 7:22 ` Andi Kleen
2006-09-12 8:30 ` Jes Sorensen
2006-09-12 9:12 ` Jes Sorensen
2006-09-12 18:05 ` Christoph Lameter
2006-09-12 7:30 ` [PATCH 0/8] Optional ZONE_DMA V1 Arjan van de Ven
2006-09-12 13:34 ` Jack Steiner
2006-09-12 17:25 ` Christoph Lameter
2006-09-12 17:33 ` Christoph Lameter
2006-09-12 17:47 ` Martin Bligh
2006-09-12 17:53 ` Christoph Lameter
2006-09-12 17:40 ` Martin Bligh
2006-09-13 2:41 ` Nick Piggin
2006-09-13 7:55 ` Jes Sorensen
2006-09-13 7:57 ` Jes Sorensen
2006-09-13 8:10 ` Christoph Lameter
2006-09-13 9:52 ` Jes Sorensen
2006-09-13 17:23 ` Christoph Lameter
2006-09-13 17:49 ` Jack Steiner [this message]
2006-09-13 18:00 ` Christoph Lameter
2006-09-14 8:52 ` Jes Sorensen
2006-09-14 16:55 ` Christoph Lameter
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=20060913174948.GA6533@sgi.com \
--to=steiner@sgi.com \
--cc=ak@suse.de \
--cc=arjan@infradead.org \
--cc=clameter@sgi.com \
--cc=hch@infradead.org \
--cc=jes@sgi.com \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=marcelo@kvack.org \
--cc=mbligh@google.com \
--cc=nickpiggin@yahoo.com.au \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox