Bug#337045: race condition?: sometimes fails to correctly detect my harddisk

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Bug#337045: race condition?: sometimes fails to correctly detect my harddisk

Jurij Smakov
Hi,

I've seen similar symptoms (missing devices at boot) go away when I've
increased the delay after the call to udevsynthesize in the initrd's init.
You can try increasing the argument to sleep in /usr/share/initramfs-tools/init
(around line 72):

# Populate /dev tree
log_begin_msg "Initializing /dev"
mkdir /dev/.udevdb
UDEVD_EXPECTED_SEQNUM=$(($(cat /sys/kernel/hotplug_seqnum) + 1)) udevd --daemon
udevsynthesize
sleep 2
log_end_msg

Try changing 'sleep 2' to 'sleep 20', regenerate the initrd and reboot.
Let us know if it fixes the problem.

Best regards,

Jurij Smakov                                        [hidden email]
Key: http://www.wooyd.org/pgpkey/                   KeyID: C99E03CC


--
To UNSUBSCRIBE, email to [hidden email]
with a subject of "unsubscribe". Trouble? Contact [hidden email]

Reply | Threaded
Open this post in threaded view
|

Bug#337045: race condition?: sometimes fails to correctly detect my harddisk

Willi Mann

> I've seen similar symptoms (missing devices at boot) go away when I've
> increased the delay after the call to udevsynthesize in the initrd's
> init. You can try increasing the argument to sleep in
> /usr/share/initramfs-tools/init (around line 72):
>
> # Populate /dev tree
> log_begin_msg "Initializing /dev"
> mkdir /dev/.udevdb
> UDEVD_EXPECTED_SEQNUM=$(($(cat /sys/kernel/hotplug_seqnum) + 1)) udevd
> --daemon
> udevsynthesize
> sleep 2
> log_end_msg
>
> Try changing 'sleep 2' to 'sleep 20', regenerate the initrd and reboot.
> Let us know if it fixes the problem.

I changed it to sleep 10, and up to now, it hasn't occured again. thanks.

However, I would ask you not to call that a fix, it's just a quite ugly
workaround, something that should not end up in a stable release. And it
makes another issue more annoying because the fbcon module is currently
loaded after the sleep, so I have for about 12 seconds a blank screen
(with vga=791). With sleep 2, it was about 4 seconds, something that was
more or less acceptable.

The proper fix is to avoid the race condition, by appropriate locking
code in the concerned software.

Willi


--
To UNSUBSCRIBE, email to [hidden email]
with a subject of "unsubscribe". Trouble? Contact [hidden email]