Can't boot without acpi=off: failed resume?

English Install/Boot/Login
1

After a quick update, consisting mostly of a new kernel, my system stalls at boot after scanning the USB drive it resides on (that whole block of lines saying SDx…)

It pauses, then prints (with large pauses, some a minute or more)


udevadm settle - timeout of 30 seconds reached, the event queue contains: /sys/devices/LNXSYSTM:00/LNXCPU:01
Trying manual resume of [path of my swap partition]
udevadm settle - timeout of 30 seconds reached, the event queue contains: /sys/devices/LNXSYSTM:00/LNXCPU:01
Invoking userspace resume from [path of my swap partition]
resume: libgcrypt version: 1.44
udevd[96]:worker[101] unexpectedly returned with status 0x0100
udevd[96]:worker[101] failed while handling 'devices/LNXSYSTM:00/LNXCPU:01

It then pauses for several more minutes, and produces more output. Maybe it would boot if I leave it long enough - don’t have time to check now, so used magic sysrq to reboot. I can boot fine with acpi=off.

It appears, and searches seem to confirm, that this could be a failed attempt to resume - and I did at one point a while ago attempt to hibernate the system, and it didn’t work, so there could be any manner of garbage flags left over somewhere. Suggestion given was to reset said flags, but I’m having no luck working out where they might be. Does anyone have any suggestions?

Thanks.

2

Have a read here :)??

[opensuse-factory] Warning: Don’t update to kernel2.6.31-9.1 from Fact](http://lists.opensuse.org/opensuse-factory/2009-09/msg00494.html)

It may have some connection with the issues you describe :slight_smile:

It’s a good idea to be subscribed with Akregator to mail lists :slight_smile:

3

Ah, thanks. I did have a good search through the bugs - found a few that looked similar to what was happening to me, but none recent. But forum trawls turned up those symptoms as a failed resume, so I figured it could be that. Made no sense it just started after an update, but there you go… Guess I need to remember to check the mailing list don’t I. :slight_smile:

I guess if I can still boot acpi=off there’s nothing stopping me just waiting till they update again, rather than trying to fix it…

4

IIRC they already fixed it BUT it is still waiting for OBS to build it which will happen in couple of hours. Anyway, the response was very fast. Funny thing that most problems have been discovered with M7, is it becaus it so buggy or because more people is testing it with each release :)??

5

It’s a dilemma isn’t it. You need exposure to different hardware, but with that comes numpties like me who don’t read the lists. :slight_smile:

6

Hey, not everybody has to be a maniac which reads mailing lists :wink:

I read it because i like to read if some things that prevent me from installing 11.2 have been fixed :slight_smile:

I reported couple of bugs already some have been fixed but one is really bad and it doesn’t seem as it will be fixed in the nearest time :slight_smile:

7

Which one?

8

http://bugzilla.novell.com/show_bug.cgi?id=525237

mkinitrd will fail if the system is on LVM which sits on top of a RAID 0/1 stripe.

9

Does look nasty, unless you want to repartition a new /boot.

But, on the plus side, you don’t have to deal with all the fun I’m having… Now KDE has decided that if one window manager is fun, two must be even better!

Keeps us on our toes though, eh? :wink:

10

Oh yes, it definitely keeps us busy lol!

But how boring the world would be if there were no BUGS :wink:

Well, about that BUG, i always have separate /boot partition (i keep all initrd files for all my systems in there, i have only one /boot partition for that).

Anyway, i am looking forward on using openSUSE 11.2 as it will be one great release, also i consider running Factory too IF, but only IF there will be packman packages for it (more than codecs though).

11

So you can’t install till it’s fixed, even with a separate boot?

I admit, I don’t understand RAID at all.

But yes, it does seem to be shaping up to be a great release. Unfortunately, it almost seems to be going backwards right now, but I suppose that’s what we’re here for… :slight_smile:

12

I must agree with the “going backwards” part, the BUG i reported didn’t exist for 11.1 :slight_smile:

But BUGS show up IF there are some changes introduced right? So this should be a good indicator :slight_smile:

One thing i noticed is that it boots A LOT faster than 11.1 (i guess it is thanks to the 2.6.31 kernel), there are some inside changes that should also improve performance and i also hope i will see webyast in this release, i will then be able to manage my mothers PC remotely WITHOUT installing anything additional like krdc etc. :slight_smile:

Also i love the new YaST icon rotfl!

P.S. The separate /boot partition doesn’t help, it just doesn’t create initrd after install. No matter if i have separate /boot or not.

13

Webyast sounds like an interesting development. If there’s one thing you can say about SUSE, it’s that they’re on the cutting edge of conservatism. :stuck_out_tongue:

They don’t seem to go for the whizz-bang! that Fedora goes for (not to be taken as belittling Fedora’s remarkable record on core technologies - I guess they’re just more slanted at enthusiast desktop as well), but they’re certainly still inventing stuff. Look at the build service. I haven’t used it myself, but it looks remarkably useful.

You were quite right, anyway - kernel is all mended. Cheers, and good luck with your bug!

14

Well spoken :slight_smile: I guess openSUSE people go for things that have been well tested since openSUSE is just a testbed for SLED and SLES so it has to be stable :slight_smile:

BTW, i also installed M7 onto my system and i already had the “fixed” kernel in the OSS repo (the one with -10 at the end appended) so i am fine now :slight_smile: