Immediately after the new kernel was installed a week or so again, I get the boot time message “ata2 :COMRESET failed”, which repeats twice and then eventually the boot is resumed and successful. There are no symptoms of problems with any of the hard disks. This happens every boot, and must be related to the kernel update as I’ve never had it before and it coincides with the update.
Upon googling the error message, I find that Ubuntu users and others seem to have had this issue a few years back, and it seems to have been solved then with some update patch. Am I the only one, or has anyone else experienced this with the new kernel?
I am a bit surprised. After more then 1000 postings here (and without doubts you read a multiple of them) I still have to encourage you to give precise details when you ask something. >:(
What is the openSUSE level you are using? What is the new kernel? Is it from the Update repo?
Well I was just checking to see if my assumption that this was to do with the kernel update was correct (see question at the end of the message). Using 12.1 as per signature (sorry, yes, desktop machine), and thought “new kernel” was self-explanatory as the kernel update received the other week or so was the first and (so far) only one for 12.1. And I did say “update” which sort of implies where it came from 
Wouldn’t have made much sense to ask if anyone else here had the same last week if I was talking about some homebrew kernel I had hand-woven into openSUSE 11.1…
Not being sarcastic, Henk, appreciate your reply, just explaining why I thought it was all pretty clear! And BTW, the kernel is 3.1.10-1.9-desktop.
Thanks for the explanation. I do not not take them for sarcasm at all. It explains your way of thinking when writing the post. But as you find out yourself allready, your signature has two openSUSE levels. Also “update” is a generic English term and there is no way to know what you thought when you wrote this. Better never assume that what you write calls for the same view of things in others as you have at that moment. This can be avoided by precise descriptions. Also, at least for me, it is very annoying when I start reading a thread and do not find the openSUSE level in the first place. Having to look at the end if it is there is distracting. You might then say that that is a minor annoyance. But people often forget that they should make it as easy as possible for potential helpers and not let them search left, right, up, down for the obvious data. Some wiill simple skip to the next problem and you will even never know they started with good humour to help you, but stopped in distress.
Also, I personaly know you are an old hand here which made many usefull contributions. Not everybody will know that. And after your long experience with the forums you will very well know that some people do post about 11.1. and hand made kernels without telling this before post #11 is reached in the thread after 2 days and considarable pressure to provide facts, not stories :(.
On 2012-05-02 14:36, gminnerup wrote:
> Not being sarcastic, Henk, appreciate your reply, just explaining why I
> thought it was all pretty clear! And BTW, the kernel is
> 3.1.10-1.9-desktop
No, it wasn’t very clear. I also had doubts.
Re your signature, I seldom read them, not trust their content re the
message itself. Signatures are just adornments 
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Hi
Have you tried adding the following to your grub boot options?
libata.force=noncq
–
Cheers Malcolm °¿° (Linux Counter #276890)
openSUSE 12.1 (x86_64) Kernel 3.1.10-1.9-desktop
up 3 days 8:31, 4 users, load average: 0.06, 0.04, 0.05
CPU Intel i5 CPU M520@2.40GHz | Intel Arrandale GPU
Hi
And in this forum we ask users (well I do) to add their system info in
the signature so it doesn’t need to be asked…
–
Cheers Malcolm °¿° (Linux Counter #276890)
openSUSE 12.1 (x86_64) Kernel 3.1.10-1.9-desktop
up 3 days 8:40, 4 users, load average: 0.04, 0.07, 0.05
CPU Intel i5 CPU M520@2.40GHz | Intel Arrandale GPU
Yes, I know, but it is rather useless if people have more then one system with more then one openSUSE level (as the OP demonstrates). And they often forget to update the signature. I rather love it when they provide the information on the spot where it is needed. Then they chance that it is the truth is much greater.
Thanks, Malcolm, will try that and report back!
Yes, that seems to have worked, thanks!
Follow-up: I’m not quite sure what exactly Malcolm’s boot option does (maybe I should have asked…) but it worked. However, I then noticed that shutdown didn’t work either, eventually coming up with a systemd error message that it couldn’t unmount an external network drive. So I then switched to sysvinit and shutdown worked, and then decided to remove Malcolm’s boot option too in case the original issue was also systemd-related, and it seems to have been: I still get some error message relating to ata2 but it flashes past quickly and the boot process, while a little slower overall, is now quite smooth.
Think I’ll stay away from systemd for a while and stick with sysvinit until things have settled down…
Hi
A google on the string “libata.force=noncq” not sure if there is a BIOS
upgrade available for your system, which may help? In the BIOS is there
an NCQ option?
–
Cheers Malcolm °¿° (Linux Counter #276890)
openSUSE 12.1 (x86_64) Kernel 3.1.10-1.9-desktop
up 5 days 10:48, 5 users, load average: 0.02, 0.03, 0.05
CPU Intel i5 CPU M520@2.40GHz | Intel Arrandale GPU
On 2012-05-10 14:46, gminnerup wrote:
> Think I’ll stay away from systemd for a while and stick with sysvinit
> until things have settled down…
Then don’t forget to write the corresponding bugzilla.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)