From: Rich Turner <rich@storix.com>
To: linux-hotplug@vger.kernel.org
Subject: Re: udevstart v udevtrigger
Date: Fri, 14 Jul 2006 16:49:01 +0000 [thread overview]
Message-ID: <200607140949.02146.rich@storix.com> (raw)
In-Reply-To: <200607131359.39003.rich@storix.com>
what would be the effects of running udevtrigger on a system running a kernel
prior to 2.6.15?
will it fail and return a non-zero return code? will it lock up the kernel?
if running udevtrigger on a pre-2.6.15 kernel causes bad things then i will
need to do some kernel level checking before running it.
On Thursday 13 July 2006 16:59, Kay Sievers wrote:
> On Thu, 2006-07-13 at 15:06 -0700, Rich Turner wrote:
> > we use udev when we perform a bare-metal restore of a system from a
> > backup (if the system that we created boot media from has it available).
> > we use udev to create device nodes in our boot media environment only.
> > once the system is rebuilt then the system's original udev support is
> > used.
> >
> > our only support requirement is that the system is running kernel 2.4 or
> > later. granted udev requires 2.6 so those systems are the only ones we
> > are discussing.
> >
> > in short:
> > we create the system installation boot media and include the following
> > udev-related files on the media:
> > udev
> > udevstart
> > udevd (only if udev is not available and there is udevstart)
> > /etc/udev/*
> >
> > we use udev by running udevstart if we have udev command. or we start
> > udevd if we do not have the udev command and then run udevstart. this is
> > creating all necessary device nodes.
> >
> > i have not tested this yet, but my thoughts are to add udevtrigger to the
> > boot media. then if there is not a udevstart command create a link
> > udevstart->udevtrigger. that way i will not have to make changes
> > everywhere else in the software where i do run udevstart.
> >
> > i guess the question is can i treat udevstart and udevtrigger as the same
> > command?
> >
> > will udevtrigger provide me with the same results as udevstart?
>
> Udevstart, just creates devices nodes and runs completely serialized, it
> will return when all currently known kernel devices are found and all
> device nodes are created. It will not cause a coldplug run or trigger
> kernel module loads for events that got lost during boot.
>
> Udevtrigger, just asks the kernel to emit all events for all devices
> again. Everything runs in parallel, also events to load kerenl modules
> are generated. Everything, including device node creation, runs
> asynchronously. If needed, you have to wait until the kernel has sent
> all events and udev is finished handling the events. That's what
> udevsettle is used for.
>
> Udevstart will hang, if one rule execution blocks, udevtrigger will not
> block the whole process.
>
> Udevtrigger replaced the coldplug logic from the old hotplug package,
> udevstart can't do any coldplug. Udevtrigger needs at least kernel
> 2.6.15.
>
> Udevstart and the udev binary are no longer installed by default on
> recent systems.
>
> Kay
--
Rich Turner
Storix, Inc.
619-543-0200 x113
-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid\x120709&bid&3057&dat\x121642
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
next prev parent reply other threads:[~2006-07-14 16:49 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-07-13 20:59 udevstart v udevtrigger Rich Turner
2006-07-13 21:29 ` Greg KH
2006-07-13 22:06 ` Rich Turner
2006-07-13 22:21 ` Greg KH
2006-07-13 23:59 ` Kay Sievers
2006-07-14 16:49 ` Rich Turner [this message]
2006-07-14 17:09 ` Andrey Borzenkov
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=200607140949.02146.rich@storix.com \
--to=rich@storix.com \
--cc=linux-hotplug@vger.kernel.org \
/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).