From: "Bryn M. Reeves" <breeves@redhat.com>
To: roland <devzero@web.de>
Cc: device-mapper development <dm-devel@redhat.com>
Subject: Re: bug in dm-loop? - was:Re: Re: device mapper integrated loops - and one more year !
Date: Mon, 22 Jan 2007 22:10:29 +0000 [thread overview]
Message-ID: <45B53655.1040209@redhat.com> (raw)
In-Reply-To: <083901c73e70$ccd85f60$eeeea8c0@aldipc>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
roland wrote:
> Hi Bryn,
>
>> I'll also get the version on kernel.org updated - thanks for catching
>> this!
>
> thanks very much for that quick patch - it works as expected and the
> kernel oops went away!
Your very welcome! I'm glad it addressed the problem & it's good to get
some feedback on dm-loop.
> since i want to use dm-loop for having a large number of loop-devices, i
> did some more "hardcore testing". ;)
>
> so, i wrote a little test script which created a lot of dummy-files and
> turning them into a dm-loop target.
>
> this worked very well - must have been around 3000 loop-targets when i
> got this one:
That's useful to know - this is a known limitation of the current code.
It simply allocates one big area of memory to store the lookup table for
the device. These kind of large, contiguous allocations are a "bad
thing" in the kernel, as you get failures of the kind you noted when
memory gets tight/fragmented.
> this was on a system with 256 MB of ram.
>
> i assume, that i ran out of some kernel-buffer memory.
>
> ok, not a real problem for me since i don`t expect having that much
> number of iso images on my cd-roms server to be loopback mounted, but -
> anyway - could you probably give a comment on this ?
>
> is this some "natural limitation" due to dm-loop or device mapper
> consuming kernel-memory, or could this be another bug ?
>
The current development version allocates these structures piecemeal, so
there isn't a single large allocation, rather many small allocations -
this will make problems like this go away (unless you really are out of
memory!).
You're particularly likely to see these problems with large images such
as DVD ISOs, as the number of entries in the lookup table is
proportional to the size of the image and the degree of fragmentation of
the filesystem on which it is stored.
Thanks again for testing!
Bryn.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
iD8DBQFFtTZV6YSQoMYUY94RAhpAAJ46Bqw+qBEVzapdKZh2V2nbx4AzHACePfsV
ttYf/s/ed3z/cBeCHEE09FY=
=ds+L
-----END PGP SIGNATURE-----
next prev parent reply other threads:[~2007-01-22 22:10 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-01-21 17:16 bug in dm-loop? - was:Re: Re: device mapper integrated loops - and one more year ! devzero
2007-01-22 10:00 ` Bryn M. Reeves
2007-01-22 22:00 ` roland
2007-01-22 22:10 ` Bryn M. Reeves [this message]
2007-01-22 22:42 ` roland
2007-01-22 23:20 ` Bryn M. Reeves
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=45B53655.1040209@redhat.com \
--to=breeves@redhat.com \
--cc=devzero@web.de \
--cc=dm-devel@redhat.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.