udev_run_devd and udev_run_hotplugd failes while booting
 

hey,

if somebody have seen this before, I would appreciate some help, cause I can't find any word about this error on the net.

I have a toshiba satellite laptop with debian onboard with the standard debian kernel 2.6.15-1-686 (did not compile my self) fully updated.

Now, when I'm booting, I get around a 1500 times this error:


#udevd-event[xxxx]: run_program: exec of program: /lib/udev/udev_run_devd: failed;
#udevd-event[xxxx]: run_program: exec of program: /lib/udev/udev_run_hotplugd: failed;

even though, all hardware and software is working after booting, it's no proper way to start your system with a 1500 error mgs :'( and I can't find any more information about the cause of the errors or what they mean => is not logged in dmesg or in any other log.
even after installing an other kernel, the problem remains so it's not kernel dependent. it's something with udev itself but i'm not a big geek in udev so ... :-s

I have the same problem ... also a laptop ... Also kernel 2.6.15-1-686. I don't know about 1500 times, but it's a bunch.

It does eventually quit, and boots and works fine, but the error messages are curious. My plan was just to wait a while and see if regular upgrades fix the problem, but if someone has a fix, that would be even better.

Edit: I should note a couple other things. I've had the 2.6.15 kernel quite a while ... I'm pretty sure that this problem just started with my most recent dist-upgrade. There was a LOT of stuff on that upgrade including (I think) a new udev. It's only on my 32 bit system ... Pretty much the same upgrade came on the 64-bit system, and it doesn't have the problem.

WE ARE NOT ALONE!

i also have laptop (IBM R52) also running kernel 2.6.15-686, and i also get that error!!

but i have a few more extra clues for you:
the problem also exists on kernel 2.6.14.

and... the errors are coming from initrd!
i am running this laptop now on kernel 2.6.12 with the initrd that came with the dist (knoppix 4.0.2) and i don't get the errors on startup.


i am using here mkinitramfs

are you saying making a new ramdisk with mkinitramfs would fix this problem? or on what ground you are sure it is caused by the initrd? cause the initrd I'm using also came with the precompiled kernel and I guess not everybody with this kernel has this problem ? I guess not a lot of things can be wrong with the ramdiskfile, it's pretty basic no?

anyway, i'm happy i'm not alone with this problem but it seems it hasn't been really discussed before somewhere on the net. probably a new problem since the 2.6.14 kernel series.

1. i am saying the error messages are displayed at the initrd phase of the boot. meaning they might be the problem.

2. mkinitramfs is what i use, and i think its the default for the post-2.6.12 kernels.

(3?) and B.T.W - i might be mistaken, but i think that there is no initrd that comes with the kernel, mkinitramfs is called by a script after an installation of a new kernel, at least in Debian.

you are right about this. I was not really clear.

but I think we have to find more info about the error msgs what they really mean? anybody an idea to make the logging maybe more verbose during 1st stage booting?

You probably have lingering files from hotplug. udev recently started providing hotplug so the separate hotplug package conflicts with it. During the install/upgrade of udev it states to purge the hotplug package.
Try this.
dpkg -P hotplug
dpkg-reconfigure linux-image-`uname -r`

I did apt-get -u --purge remove hotplug ...

Shouldn't that have worked? I didn't reconfigure the kernel, tho. Somebody else try that first. lol

That fixed the error on my machine.

Thanks

dpkg-reconfigure linux-image-`uname -r` basically rebuilds your initrd and updates grub, it doesn't actually do anything to your kernel

Fixed mine, too. Thanks dracae.

didn't work for me :(

i get also the following warning:
did you also get this one?


i also tried since that was also there before i installed udev.

Same here:

# dpkg -P hotplug
dpkg - warning: ignoring request to remove hotplug which isn't installed.

The udev errors still abound...

dpkg -l | grep hotplug
should show any that may have lingering files.
Like guysoft said, it may be hotplug-knoppix or something if you are using some derivative of debian, of course the forum is for debian installs ;)

$ dpkg -l | grep hotplug
ii udev 0.084-3 /dev/ and hotplug management daemon

Is this a normal response?