HDD mounting problems with fbcondecor kernel

sDoky · 2010-09-16 16:48:06

Hi, I have two external hard drives and there's a little problem. If the're connected during boot, I cannot mount them. I have to unplug and plug them back to make them mountable. The weird (well not so much) thing is, that if I run stock kernel, it doesn't happen. It happens only when I'm running fbcondecor kernel, which I compiled from AUR. I suppose there's something I forgot to tick in the config for the fbcondecor kernel, but I cannot put my finger on it. Can anyone help me, please? Thank you

Last edited by sDoky (2010-09-16 17:14:33)

karol · 2010-09-16 16:56:10

Are there any error messages when you try (and fail) to mount?

sDoky · 2010-09-16 17:00:40

karol wrote:

Are there any error messages when you try (and fail) to mount?

that's the weirdest thing, there's nothing. it doesn't respond to SIGKILL, SIGHUP, CTRL+C, just nothing can kill the mount process (the sigkill part makes me worry a lot)

EDIT: if I remember our operating systems class correctly, sigkill signal is one not to be ignored...
EDIT2: if it helps, those hard drives are ext3 fs

Last edited by sDoky (2010-09-16 17:08:07)

djgera · 2010-09-16 17:12:58

sDoky wrote:

EDIT: if I remember our operating systems class correctly, sigkill signal is one not to be ignored...

Because the process is in "Uninterruptible Sleep". A process only can receive signals when is "Runnable"

karol · 2010-09-16 17:13:46

Even root can't kill them with SIGKILL?
Edit: Thank, djgera, for clarification. Makes sense :-)

I think 'HDD mounting problems with fbcondecor kernel' is a more appropriate title, as the drives mount OK with the stock one.

Last edited by karol (2010-09-16 17:14:37)

sDoky · 2010-09-16 17:14:55

djgera wrote:

sDoky wrote:
EDIT: if I remember our operating systems class correctly, sigkill signal is one not to be ignored...
Because the process is in "Uninterruptible Sleep". A process only can receive signals when is "Runnable"

good point

I suspect the fbcondecor config... but where to start looking? I do not know...

Last edited by sDoky (2010-09-16 17:16:06)

sDoky · 2010-09-16 17:17:06

karol wrote:

Even root can't kill them with SIGKILL?
Edit: Thank, djgera, for clarification. Makes sense :-)
I think 'HDD mounting problems with fbcondecor kernel' is a more appropriate title, as the drives mount OK with the stock one.

I think it is irrelevant who kills, you either succeed or fail in sending the signal

EDIT: wait, so what you're saying is that if a process is in uninterruptible sleep, there's nothing that can kill it?
EDIT2: where should I look for any sign of error in logs? is it kernel, dmesg, messages? because to be honest, I am diving in those log files for like an hour and I have found squat

Last edited by sDoky (2010-09-16 17:20:45)

djgera · 2010-09-16 17:28:29

sDoky wrote:

EDIT: wait, so what you're saying is that if a process is in uninterruptible sleep, there's nothing that can kill it?

Exactly. Only a reboot.

Arch Linux

#1 2010-09-16 16:48:06

HDD mounting problems with fbcondecor kernel

#2 2010-09-16 16:56:10

Re: HDD mounting problems with fbcondecor kernel

#3 2010-09-16 17:00:40

Re: HDD mounting problems with fbcondecor kernel

#4 2010-09-16 17:12:58

Re: HDD mounting problems with fbcondecor kernel

#5 2010-09-16 17:13:46

Re: HDD mounting problems with fbcondecor kernel

#6 2010-09-16 17:14:55

Re: HDD mounting problems with fbcondecor kernel

#7 2010-09-16 17:17:06

Re: HDD mounting problems with fbcondecor kernel

#8 2010-09-16 17:28:29

Re: HDD mounting problems with fbcondecor kernel

Board footer