From: Pavel Machek <pavel@suse.cz>
To: "Richard B. Johnson" <root@chaos.analogic.com>
Cc: Alok Mooley <rangdi@yahoo.com>,
linux-kernel <linux-kernel@vger.kernel.org>,
linux-mm <linux-mm@kvack.org>, Dave Hansen <haveblue@us.ibm.com>
Subject: Re: Active Memory Defragmentation: Our implementation & problems
Date: Thu, 5 Feb 2004 20:03:49 +0100 [thread overview]
Message-ID: <20040205190349.GC294@elf.ucw.cz> (raw)
In-Reply-To: <Pine.LNX.4.53.0402041427270.2947@chaos>
Hi!
> > If this is an Intel x86 machine, it is impossible
> > > for pages
> > > to get fragmented in the first place. The hardware
> > > allows any
> > > page, from anywhere in memory, to be concatenated
> > > into linear
> > > virtual address space. Even the kernel address space
> > > is virtual.
> > > The only time you need physically-adjacent pages is
> > > if you
> > > are doing DMA that is more than a page-length at a
> > > time. The
> > > kernel keeps a bunch of those pages around for just
> > > that
> > > purpose.
> > >
> > > So, if you are making a "memory defragmenter", it is
> > > a CPU time-sink.
> > > That's all.
> >
> > What if the external fragmentation increases so much
> > that it is not possible to find a large sized block?
> > Then, is it not better to defragment rather than swap
> > or fail?
> >
> > -Alok
>
> All "blocks" are the same size, i.e., PAGE_SIZE. When RAM
> is tight the content of a page is written to the swap-file
> according to a least-recently-used protocol. This frees
> a page. Pages are allocated to a process only one page at
> a time. This prevents some hog from grabbing all the memory
> in the machine. Memory allocation and physical page allocation
> are two different things, I can malloc() a gigabyte of RAM on
> a machine. It only gets allocated when an attempt is made
> to access a page.
Alok is right. kernel needs to do kmalloc(8K) from time to time. And
notice that kernel uses 4M tables.
Pavel
--
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]
WARNING: multiple messages have this Message-ID (diff)
From: Pavel Machek <pavel@suse.cz>
To: "Richard B. Johnson" <root@chaos.analogic.com>
Cc: Alok Mooley <rangdi@yahoo.com>,
linux-kernel <linux-kernel@vger.kernel.org>,
linux-mm <linux-mm@kvack.org>, Dave Hansen <haveblue@us.ibm.com>
Subject: Re: Active Memory Defragmentation: Our implementation & problems
Date: Thu, 5 Feb 2004 20:03:49 +0100 [thread overview]
Message-ID: <20040205190349.GC294@elf.ucw.cz> (raw)
In-Reply-To: <Pine.LNX.4.53.0402041427270.2947@chaos>
Hi!
> > If this is an Intel x86 machine, it is impossible
> > > for pages
> > > to get fragmented in the first place. The hardware
> > > allows any
> > > page, from anywhere in memory, to be concatenated
> > > into linear
> > > virtual address space. Even the kernel address space
> > > is virtual.
> > > The only time you need physically-adjacent pages is
> > > if you
> > > are doing DMA that is more than a page-length at a
> > > time. The
> > > kernel keeps a bunch of those pages around for just
> > > that
> > > purpose.
> > >
> > > So, if you are making a "memory defragmenter", it is
> > > a CPU time-sink.
> > > That's all.
> >
> > What if the external fragmentation increases so much
> > that it is not possible to find a large sized block?
> > Then, is it not better to defragment rather than swap
> > or fail?
> >
> > -Alok
>
> All "blocks" are the same size, i.e., PAGE_SIZE. When RAM
> is tight the content of a page is written to the swap-file
> according to a least-recently-used protocol. This frees
> a page. Pages are allocated to a process only one page at
> a time. This prevents some hog from grabbing all the memory
> in the machine. Memory allocation and physical page allocation
> are two different things, I can malloc() a gigabyte of RAM on
> a machine. It only gets allocated when an attempt is made
> to access a page.
Alok is right. kernel needs to do kmalloc(8K) from time to time. And
notice that kernel uses 4M tables.
Pavel
--
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]
--
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:[~2004-02-05 19:05 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-02-03 4:46 Active Memory Defragmentation: Our implementation & problems Alok Mooley
2004-02-03 4:46 ` Alok Mooley
2004-02-03 21:26 ` Dave Hansen
2004-02-03 21:26 ` Dave Hansen
2004-02-03 22:26 ` Martin J. Bligh
2004-02-03 22:26 ` Martin J. Bligh
2004-02-04 5:09 ` Alok Mooley
2004-02-04 5:09 ` Alok Mooley
2004-02-04 5:24 ` Mike Fedyk
2004-02-04 5:24 ` Mike Fedyk
2004-02-04 5:54 ` Dave Hansen
2004-02-04 5:54 ` Dave Hansen
2004-02-04 6:05 ` Martin J. Bligh
2004-02-04 6:05 ` Martin J. Bligh
2004-02-04 6:22 ` Dave Hansen
2004-02-04 6:22 ` Dave Hansen
2004-02-04 6:29 ` Martin J. Bligh
2004-02-04 6:29 ` Martin J. Bligh
2004-02-04 6:40 ` Dave Hansen
2004-02-04 6:40 ` Dave Hansen
2004-02-04 7:17 ` Martin J. Bligh
2004-02-04 7:17 ` Martin J. Bligh
2004-02-04 8:30 ` Andrew Morton
2004-02-04 8:30 ` Andrew Morton
2004-02-04 6:53 ` Doubt about statm_pgd_range patch Arunkumar
2004-02-04 6:57 ` Active Memory Defragmentation: Our implementation & problems IWAMOTO Toshihiro
2004-02-04 6:57 ` IWAMOTO Toshihiro
2004-02-04 7:10 ` Dave Hansen
2004-02-04 7:10 ` Dave Hansen
2004-02-04 7:50 ` IWAMOTO Toshihiro
2004-02-04 7:50 ` IWAMOTO Toshihiro
2004-02-04 10:33 ` Hirokazu Takahashi
2004-02-04 10:33 ` Hirokazu Takahashi
2004-02-04 18:33 ` Alok Mooley
2004-02-04 18:46 ` Dave Hansen
2004-02-04 18:54 ` Alok Mooley
2004-02-04 18:54 ` Alok Mooley
2004-02-04 19:07 ` Richard B. Johnson
2004-02-04 19:07 ` Richard B. Johnson
2004-02-04 19:18 ` Alok Mooley
2004-02-04 19:18 ` Alok Mooley
2004-02-04 19:33 ` Richard B. Johnson
2004-02-04 19:33 ` Richard B. Johnson
2004-02-05 5:07 ` Alok Mooley
2004-02-05 5:07 ` Alok Mooley
2004-02-05 19:03 ` Pavel Machek [this message]
2004-02-05 19:03 ` Pavel Machek
2004-02-04 19:35 ` Dave McCracken
2004-02-04 19:35 ` Dave McCracken
2004-02-04 21:59 ` Timothy Miller
2004-02-04 21:59 ` Timothy Miller
2004-02-04 23:24 ` Dave Hansen
2004-02-04 23:24 ` Dave Hansen
2004-02-05 16:32 ` Dave McCracken
2004-02-05 16:32 ` Dave McCracken
2004-02-04 19:37 ` Timothy Miller
2004-02-04 19:37 ` Timothy Miller
2004-02-04 19:43 ` Dave Hansen
2004-02-04 19:43 ` Dave Hansen
2004-02-04 19:59 ` Richard B. Johnson
2004-02-04 19:59 ` Richard B. Johnson
2004-02-04 19:56 ` Dave Hansen
2004-02-04 19:56 ` Dave Hansen
2004-02-05 5:19 ` Alok Mooley
2004-02-05 5:19 ` Alok Mooley
-- strict thread matches above, loose matches on Subject: below --
2004-02-04 20:12 Mark_H_Johnson
2004-02-04 20:12 ` Mark_H_Johnson
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=20040205190349.GC294@elf.ucw.cz \
--to=pavel@suse.cz \
--cc=haveblue@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=rangdi@yahoo.com \
--cc=root@chaos.analogic.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.