Re: Strange udev problem - More Detail of the problem

Re: Strange udev problem - More Detail of the problem

[Jim Gifford]

The fcron developer gave me a patch, but it didn't work. Which he made
figuring it was a umask issue, but the patch didn't work. Patch is available
for viewing at http://ftp.jg555.com/patches/fcron/fcron-2.9.4-udev-1.patch

After talking to a few people and doing some testing I can describe the
problem a lot better now.

(root@server)-(10:50 PM Sat Jun 19)-(/tmp)
# ls -lc /dev/null
crw-------  1 root root 1, 3 2004-06-19 22:49 /dev/null
(root@server)-(10:50 PM Sat Jun 19)-(/tmp)
# cat /var/log/fcron.log | grep 22:49
Jun 19 22:49:20 server fcron[15999]: Job
/etc/firewall/blacklist --current-only started for user root (pid 16000)
(root@server)-(10:51 PM Sat Jun 19)-(/tmp)
#

In the complete file it shows the rest of blacklist update script running.

The script in question is posted at
http://svn.jg555.com/viewcvs.cgi/trunk/blacklist?rev=2&view=markup . Which I
wrote

I enabled udev debugging, and nothing showed up

I run the script without using fron, no problem.

Checked all my logs for events at the same time, nothing except for this
script and fcron.

Which also leaves udev as a potential cause.

I can set this script to 1 minute, and it will cause this issue every time.




-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
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

Re: Strange udev problem - More Detail of the problem

[Kevin P. Fleming]

Jim Gifford wrote:

> The script in question is posted at
> http://svn.jg555.com/viewcvs.cgi/trunk/blacklist?rev=2&view=markup . Which I
> wrote

You're only got on reference to /dev/null in that script, at the very 
beginning. If you trim the script down to just that block, does the 
problem still occur?


-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
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

Re: Strange udev problem - More Detail of the problem

[Jim Gifford]

Still exists Kevin. I worked all last night with Ryan to try to figure it
out. I even took that line out and put into another script and tried running
it with fcron, the problem didn't happen. It's very strange. The fcron
developer is stumped, so what I'm going to do is to revert back to an
earlier version of udev to figure out when the problem gets introduced so
Greg and Kay have a starting point to look at to help this issue.




-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
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

Re: Strange udev problem - More Detail of the problem

[Kay Sievers]

On Sun, 2004-06-20 at 09:23 -0700, Jim Gifford wrote:
> Still exists Kevin. I worked all last night with Ryan to try to figure it
> out. I even took that line out and put into another script and tried running
> it with fcron, the problem didn't happen. It's very strange. The fcron
> developer is stumped, so what I'm going to do is to revert back to an
> earlier version of udev to figure out when the problem gets introduced so
> Greg and Kay have a starting point to look at to help this issue.

Yeah, seems like a really good idea. Sure udev may apply the wrong
permissions to a node, but is seems impossible, that the kernel call
udev anytime later to recreate '/sys/class/mem/null'. So I still don't
think, that it's a udev issue, but we will see :)

You can also add some catch to the udev code, to be notified of a '/dev/
null' creation event, or remove the inode preserving code, so udev will
remove /dev/null before recreating it.

Good luck,
Kay



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
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

Re: Strange udev problem - More Detail of the problem

[Jim Gifford]

In this email I have copied Kay, hotplug list, and the fcron developer.

Affected Version of Udev: 027 026 025 024

Unaffected Version of Udev: 023

Thanx to the fcron developer Thibault Godouet <fcron@free.fr> for you
assistance on this issue.












-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND
_______________________________________________
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