From: Andreas Dilger <adilger@turbolabs.com>
To: Christoph Hellwig <hch@caldera.de>,
linux-kernel@vger.kernel.org, Alan Cox <alan@lxorguk.ukuu.org.uk>,
torvalds@transmeta.com
Subject: Re: [PATCH] init/main.c/root_dev_names - another one #ifdef
Date: Wed, 31 Oct 2001 12:08:17 -0700 [thread overview]
Message-ID: <20011031120817.L16554@lynx.no> (raw)
In-Reply-To: <20011030182810.B800@lynx.no> <200110311728.f9VHSE207521@ns.caldera.de> <20011031112055.D16554@lynx.no> <20011031193148.A12919@caldera.de>
In-Reply-To: <20011031193148.A12919@caldera.de>; from hch@caldera.de on Wed, Oct 31, 2001 at 07:31:48PM +0100
On Oct 31, 2001 19:31 +0100, Christoph Hellwig wrote:
> On Wed, Oct 31, 2001 at 11:20:55AM -0700, Andreas Dilger wrote:
> > This seems kind of ugly - an array holding each device name? The patch
> > I have rather puts a function to generate the device names when needed
> > (which is probably not very often, unless GFS does something wierd).
>
> *nod*
>
> > I take it your patch is only the "bare bones" part which shows what is
> > changed?
>
> Well, it's a patch that tries to be not intrusive, it just crates the hooks
> the two blockdevice drivers in the OpenGFS tree can use.
Well, given that it's intrusive enough to change the gendisk struct
(which my patch does as well), we may as well go whole hog and remove
all of the partition name cruft from where it should not be. I suppose
the issue is whether you are expecting GFS to go into the kernel, or if
you want to keep the diff as small as possible if it will be outside
the kernel for a long time.
The LVM code had a global function to fix the naming issue for LVM
devices in /proc/partitions, but Linus (AFAIK) didn't like it because
it didn't fix the real problem, as the posted patch does, of localizing
the partition name generation to the driver itself. The only wart in
the current system is that we use an external buffer passed into the
driver, and if the driver likes long names (ala LVM) it will overflow.
As I look through the kernel today, there is even more ugliness spread
around with "#ifdef CONFIG_DEVFS" to create strange partition names in
places where it should probably not be.
Cheers, Andreas
--
Andreas Dilger
http://sourceforge.net/projects/ext2resize/
http://www-mddsp.enel.ucalgary.ca/People/adilger/
next prev parent reply other threads:[~2001-10-31 19:08 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2001-10-30 23:53 [PATCH] init/main.c/root_dev_names - another one #ifdef Denis Zaitsev
2001-10-31 0:15 ` Alan Cox
2001-10-31 0:12 ` Linus Torvalds
2001-10-31 0:28 ` Alan Cox
2001-10-31 1:28 ` Andreas Dilger
2001-10-31 17:28 ` Christoph Hellwig
2001-10-31 18:20 ` Andreas Dilger
2001-10-31 18:31 ` Christoph Hellwig
2001-10-31 19:08 ` Andreas Dilger [this message]
2001-11-01 2:05 ` Denis Zaitsev
2001-10-31 1:36 ` Denis Zaitsev
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=20011031120817.L16554@lynx.no \
--to=adilger@turbolabs.com \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=hch@caldera.de \
--cc=linux-kernel@vger.kernel.org \
--cc=torvalds@transmeta.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox