linux-hotplug.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kay Sievers <kay.sievers@vrfy.org>
To: linux-hotplug@vger.kernel.org
Subject: Re: [PATCH] udev - next round of udev event order daemon
Date: Sat, 24 Jan 2004 01:21:08 +0000	[thread overview]
Message-ID: <20040124012108.GA4701@vrfy.org> (raw)
In-Reply-To: <20040123151425.GA3813@vrfy.org>

On Fri, Jan 23, 2004 at 02:30:48PM -0800, Greg KH wrote:
> On Fri, Jan 23, 2004 at 04:14:25PM +0100, Kay Sievers wrote:
> > Here is the next round of udevd/udevsend:
> > 
> > udevsend - If the IPC message we send is not catched by a receiver we fork
> >            the udevd daemon to process this and the following events
> 
> Problem is we end up forking a lot of udevd instances if the system is
> under load.  That's not good :(
>
> I applied this patch, and then added some logic to udevd to only let one
> instance of it run at a time.  It isn't the nicest, but seems to work
> for now...

Maybe 190 millisec's are too short?
Hmm, we may also look what pid received the last event and check if this pid
is still running. So we prevent forking a new daemon just because of the
timeout.


> Hm, sometimes I still loose an event, I think it's back to the udevsend
> creating too many udevd instances.  Care to look at this?

While we wait for the return of the udev exec, we don't get the messages
out of the queue. That may cause the problem with the timeout.
Do we need a receiver thread? Or fork udev in the background, but then
a "remove" may be faster than a "add"?


> Is using ipc calls too messy for this?  Others have proposed sockets.
> Would that make the startup logic simpler?

Yes and no. We notice that we can't connect so there is a better
indicator, that there is no daemon. But I think the main problem remains the
same. I have less experience with this time critical stuff, so we may try
it out, if nobody comes with a better idea?

thanks,
Kay


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
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

      parent reply	other threads:[~2004-01-24  1:21 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-01-23 15:14 [PATCH] udev - next round of udev event order daemon Kay Sievers
2004-01-23 22:30 ` Greg KH
2004-01-24  1:21 ` Kay Sievers [this message]

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=20040124012108.GA4701@vrfy.org \
    --to=kay.sievers@vrfy.org \
    --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).