From: "H. Peter Anvin" <hpa@zytor.com>
To: Kay Sievers <kay.sievers@vrfy.org>
Cc: Paul Parsons <lost.distance@yahoo.com>,
Kirill Smelkov <kirr@mns.spb.ru>,
linux-kernel@vger.kernel.org
Subject: Re: No /dev/root with devtmpfs?
Date: Wed, 08 Feb 2012 09:36:32 -0800 [thread overview]
Message-ID: <4F32B2A0.70405@zytor.com> (raw)
In-Reply-To: <CAPXgP10OcEc1J4VkCfbMowYR=Ou_F+SqKdDA5kaUAUgjfeiBsw@mail.gmail.com>
On 02/08/2012 09:25 AM, Kay Sievers wrote:
> On Wed, Feb 8, 2012 at 18:20, H. Peter Anvin <hpa@zytor.com> wrote:
>> On 02/08/2012 08:44 AM, Paul Parsons wrote:
>>>
>>> Could you simply use /etc/fstab to identify the root partition?
>>>
>>
>> That's not a very good thing, as it is much more likely to be wrong.
>>
>> It would be a good thing to have the /dev/root symlink *IF* a valid root
>> device exists (defined as a device node appearing which has the same
>> device number as reported by stat on the root directory), if nothing
>> else because we have had one available for a very long time and this is
>> needless breakage.
>>
>> Obviously, if such a device doesn't exist (btrfs, NFS, tmpfs) then don't.
>
> Tools should just do the equivalent of:
> $ ls -l /sys/dev/block/$(mountpoint -d /)
> /sys/dev/block/8:1 ->
> ../../devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/sda1
>
> and all is fine. I'm convinced, that determining the root device is a
> job for *running code* not to expect a symlink to be around.
>
> Kay
Okay, first of all, this is a service (you can think of it as such)
which has been provided for a long time. It can be done with minimal
effort as devices appear, so you don't need to scan the /dev directory
to find the device node that corresponds to the above (as you know,
there are deliberately no device nodes in /sys, although *most* of the
time you obviously have the name right there.)
I'm not saying you're wrong, I'm saying you should consider it a
backwards compatibility solution for an interface which has been
provided for a long time already.
-hpa
--
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel. I don't speak on their behalf.
next prev parent reply other threads:[~2012-02-08 17:36 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-06 11:18 No /dev/root with devtmpfs? Kirill Smelkov
2012-02-06 12:10 ` Paul Parsons
2012-02-06 12:52 ` Kirill Smelkov
2012-02-06 19:45 ` Kay Sievers
2012-02-07 7:55 ` Kirill Smelkov
2012-02-07 10:05 ` Kay Sievers
2012-02-08 7:35 ` Kirill Smelkov
2012-02-08 16:44 ` Paul Parsons
2012-02-08 17:20 ` H. Peter Anvin
2012-02-08 17:25 ` Kay Sievers
2012-02-08 17:36 ` H. Peter Anvin [this message]
2012-02-10 9:04 ` Kirill Smelkov
2012-02-10 13:10 ` Paul Parsons
2012-02-13 8:25 ` Kirill Smelkov
2012-02-13 18:48 ` H. Peter Anvin
2012-02-13 19:30 ` Kay Sievers
2012-02-13 19:32 ` H. Peter Anvin
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=4F32B2A0.70405@zytor.com \
--to=hpa@zytor.com \
--cc=kay.sievers@vrfy.org \
--cc=kirr@mns.spb.ru \
--cc=linux-kernel@vger.kernel.org \
--cc=lost.distance@yahoo.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;
as well as URLs for NNTP newsgroup(s).