[Solved] Testing & update = no boot. (sort of)

gauthma · 2012-01-07 02:38:19

After updating, and rebooting, I get a message saying that the harddisk was not detected. After that, I get thrown to the busybox prompt. Now being pretty much sure that I had read in arch-dev-public that this should be a seamless upgrade, despite the change to kmod, I boot a live cd and go check the ML archives. Alas, I was right: there were a couple of issues, but they were solved before my update took place.

Grasping at the straws and not knowing what to do, I decide to call it a day, and power down the box. But on a whim, I reboot instead! And guess what: NOW IT BOOTS! Just fine! Just seamless.

Thoughts/ideas/additional information you need to know?

Last edited by gauthma (2012-02-04 20:33:53)

falconindy · 2012-01-07 03:43:09

I'm not sure. I spent a couple hours with someone else trying to debug this exact scenario and we got pretty much nowhere.

Could you install this version of mkinitcpio, downgrade kmod back to module-init-tools, regenerate your initramfs and reboot a few times to try and replicate this?

road2stat · 2012-01-07 10:02:45

It seems I have the same problem here ...

Waiting 10 seconds for /dev/disk/by-uuid/the device uuid ...

then time out, skip to the ramfs prompt.

The default kernel seems to boot randomly, however, the fallback img has a higher chance of successful booting. That's confusing ... Did I miss anything?

afiestas · 2012-01-07 15:06:34

Same issue here, writting this post with the fallback image on which ATM I have had 100% success botting.

fredbezies · 2012-01-08 09:53:23

Opened a bug : https://bugs.archlinux.org/task/27862

Looks like kmod is kinda busted with my laptop. It works flawlessly with my desktop computer.

I had no booting problem until kmod was introduced. I will try to downgrade to good old module-init-tools, recreate initramfs file and see if bug is dead or not.

See also this post on arch-general : http://mailman.archlinux.org/pipermail/ … 23973.html

Kernel problem ? Kmod problem ? Any idea ?

Last edited by fredbezies (2012-01-08 09:55:22)

loafer · 2012-01-08 09:59:17

Based on falconindy's post I installed the version mkinitcpio from the link and reverted to module-init-tools. I have now shutdown and restarted several times without any problems.

Last edited by loafer (2012-01-08 09:59:37)

Circleo · 2012-01-08 10:03:09

Same issues as above, after upgrading to latest testing kernel and switching from module-init-tools to kmod, "Waiting 10 seconds for /dev/sda7..."
Booting as fallback worked more often than not, booting as default did not work at all.
Using falconindy's suggestion, I installed mkinitcpio from his link, reverted to module-init-tools, and rebuilt the initramfs. The issue hasn't returned.

Last edited by Circleo (2012-01-08 10:15:49)

fredbezies · 2012-01-08 10:23:34

I'm having the same software used on both my desktop and my laptop. Besides the kmod to modules-init-tools downgrade and some laptop specific modules loaded on start (like wifi chipset module), kmod = Ok on my desktop, broken on my laptop.

Weird !

Circleo · 2012-01-09 22:11:47

As an update, re-installing the latest kmod has not reproduced this issue. The computer boots fine with kmod now. I have not regenerate initramfs after installing kmod, in case that matters.

falconindy · 2012-01-10 00:01:33

Circleo wrote:

As an update, re-installing the latest kmod has not reproduced this issue. The computer boots fine with kmod now. I have not regenerate initramfs after installing kmod, in case that matters.

If you didn't regenerate your initramfs after installing kmod, then you effectively haven't done anything -- the old kmod is still on the initramfs.

This thread can die -- I'm confident this particular bug is fixed.

gauthma · 2012-01-10 11:26:49

falconindy wrote:

This thread can die -- I'm confident this particular bug is fixed.

I didn't try your initial suggestion because I had been away from the keyboard for a couple of days. However, after coming back and updating, I noticed that there has been an update to kmod, and I have additionally manually re-generated the initramfs afer that. And that seems to have taken care of the problem. I'll make a couple of tests over the next few days, and mark the thread as solved if I can no longer reproduce this issue.

gauthma · 2012-01-14 13:49:52

Ahem, so this happened again. This time the box booted after two bad attempts. Still, this shows the issue isn't quite resolved yet. In the meantime, today's update brought in new versions of kmod and mkinitcpio, so I'll keep testing.

gauthma · 2012-02-04 20:33:39

Never happened again. Marking as solved

Arch Linux

#1 2012-01-07 02:38:19

[Solved] Testing & update = no boot. (sort of)

#2 2012-01-07 03:43:09

Re: [Solved] Testing & update = no boot. (sort of)

#3 2012-01-07 10:02:45

Re: [Solved] Testing & update = no boot. (sort of)

#4 2012-01-07 15:06:34

Re: [Solved] Testing & update = no boot. (sort of)

#5 2012-01-08 09:53:23

Re: [Solved] Testing & update = no boot. (sort of)

#6 2012-01-08 09:59:17

Re: [Solved] Testing & update = no boot. (sort of)

#7 2012-01-08 10:03:09

Re: [Solved] Testing & update = no boot. (sort of)

#8 2012-01-08 10:23:34

Re: [Solved] Testing & update = no boot. (sort of)

#9 2012-01-09 22:11:47

Re: [Solved] Testing & update = no boot. (sort of)

#10 2012-01-10 00:01:33

Re: [Solved] Testing & update = no boot. (sort of)

#11 2012-01-10 11:26:49

Re: [Solved] Testing & update = no boot. (sort of)

#12 2012-01-14 13:49:52

Re: [Solved] Testing & update = no boot. (sort of)

#13 2012-02-04 20:33:39

Re: [Solved] Testing & update = no boot. (sort of)

Board footer