From: Andi Kleen <ak@muc.de>
To: Christoph Hellwig <hch@infradead.org>, Steve Lord <lord@xfs.org>,
"Mukker, Atul" <Atulm@lsil.com>,
"'Marcelo Tosatti'" <marcelo.tosatti@cyclades.com>,
"'Mel Gorman'" <mel@csn.ul.ie>,
"'William Lee Irwin III'" <wli@holomorphy.com>,
"'Linux Memory Management List'" <linux-mm@kvack.org>,
"'Linux Kernel'" <linux-kernel@vger.kernel.org>,
"'Grant Grundler'" <grundler@parisc-linux.org>
Subject: Re: [PATCH] Avoiding fragmentation through different allocator
Date: 25 Jan 2005 15:49:32 +0100
Date: Tue, 25 Jan 2005 15:49:32 +0100 [thread overview]
Message-ID: <20050125144932.GA75109@muc.de> (raw)
In-Reply-To: <20050125142757.GA20442@infradead.org>
On Tue, Jan 25, 2005 at 02:27:57PM +0000, Christoph Hellwig wrote:
> > It is not the driver per se, but the way the memory which is the I/O
> > source/target is presented to the driver. In linux there is a good
> > chance it will have to use more scatter gather elements to represent
> > the same amount of data.
>
> Note that a change made a few month ago after seeing issues with
> aacraid means it's much more likely to see contingous memory,
> there were some numbers on linux-scsi and/or linux-kernel.
But only at the beginning. iirc after a few days of uptime
and memory fragmentation it degenerates back to the old numbers.
Perhaps the recent anti defragmentation work will help more.
-Andi
P.S.: on a AMD x86-64 box the theory can be relatively easily tested:
just run with iommu=force,biomerge that will use the IOMMU to merge
SG elements. I just don't recommend it for production because some errors
are not well handled.
WARNING: multiple messages have this Message-ID (diff)
From: Andi Kleen <ak@muc.de>
To: Christoph Hellwig <hch@infradead.org>, Steve Lord <lord@xfs.org>,
"Mukker, Atul" <Atulm@lsil.com>,
'Marcelo Tosatti' <marcelo.tosatti@cyclades.com>,
'Mel Gorman' <mel@csn.ul.ie>,
'William Lee Irwin III' <wli@holomorphy.com>,
'Linux Memory Management List' <linux-mm@kvack.org>,
'Linux Kernel' <linux-kernel@vger.kernel.org>,
'Grant Grundler' <grundler@parisc-linux.org>
Subject: Re: [PATCH] Avoiding fragmentation through different allocator
Date: 25 Jan 2005 15:49:32 +0100
Date: Tue, 25 Jan 2005 15:49:32 +0100 [thread overview]
Message-ID: <20050125144932.GA75109@muc.de> (raw)
In-Reply-To: <20050125142757.GA20442@infradead.org>
On Tue, Jan 25, 2005 at 02:27:57PM +0000, Christoph Hellwig wrote:
> > It is not the driver per se, but the way the memory which is the I/O
> > source/target is presented to the driver. In linux there is a good
> > chance it will have to use more scatter gather elements to represent
> > the same amount of data.
>
> Note that a change made a few month ago after seeing issues with
> aacraid means it's much more likely to see contingous memory,
> there were some numbers on linux-scsi and/or linux-kernel.
But only at the beginning. iirc after a few days of uptime
and memory fragmentation it degenerates back to the old numbers.
Perhaps the recent anti defragmentation work will help more.
-Andi
P.S.: on a AMD x86-64 box the theory can be relatively easily tested:
just run with iommu=force,biomerge that will use the IOMMU to merge
SG elements. I just don't recommend it for production because some errors
are not well handled.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>
next prev parent reply other threads:[~2005-01-25 14:49 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-01-25 14:02 [PATCH] Avoiding fragmentation through different allocator Mukker, Atul
2005-01-25 14:02 ` Mukker, Atul
2005-01-25 14:17 ` Steve Lord
2005-01-25 14:17 ` Steve Lord
2005-01-25 14:27 ` Christoph Hellwig
2005-01-25 14:27 ` Christoph Hellwig
2005-01-25 14:49 ` Andi Kleen [this message]
2005-01-25 14:49 ` Andi Kleen
2005-01-25 14:56 ` Andi Kleen
2005-01-25 14:56 ` Andi Kleen
2005-01-25 16:12 ` Mel Gorman
2005-01-25 16:12 ` Mel Gorman
2005-01-25 18:50 ` Grant Grundler
2005-01-25 18:50 ` Grant Grundler
-- strict thread matches above, loose matches on Subject: below --
2005-01-20 10:13 Mel Gorman
2005-01-20 10:13 ` Mel Gorman
2005-01-21 14:28 ` Marcelo Tosatti
2005-01-21 14:28 ` Marcelo Tosatti
2005-01-22 21:48 ` Mel Gorman
2005-01-22 21:48 ` Mel Gorman
2005-01-22 21:59 ` Marcelo Tosatti
2005-01-22 21:59 ` Marcelo Tosatti
2005-01-23 13:28 ` Marcelo Tosatti
2005-01-23 13:28 ` Marcelo Tosatti
2005-01-24 13:28 ` Mel Gorman
2005-01-24 13:28 ` Mel Gorman
2005-01-24 12:29 ` Marcelo Tosatti
2005-01-24 12:29 ` Marcelo Tosatti
2005-01-24 16:44 ` James Bottomley
2005-01-24 16:44 ` James Bottomley
2005-01-24 15:49 ` Marcelo Tosatti
2005-01-24 15:49 ` Marcelo Tosatti
2005-01-24 20:36 ` James Bottomley
2005-01-24 20:36 ` James Bottomley
2005-01-24 20:47 ` Steve Lord
2005-01-24 20:47 ` Steve Lord
2005-01-25 7:39 ` Andi Kleen
2005-01-25 7:39 ` Andi Kleen
2005-01-24 19:55 ` Grant Grundler
2005-01-24 19:55 ` Grant Grundler
2005-01-20 10:12 Mel Gorman
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=20050125144932.GA75109@muc.de \
--to=ak@muc.de \
--cc=Atulm@lsil.com \
--cc=grundler@parisc-linux.org \
--cc=hch@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lord@xfs.org \
--cc=marcelo.tosatti@cyclades.com \
--cc=mel@csn.ul.ie \
--cc=wli@holomorphy.com \
/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.