Strange kernel behaviour since last update [partly solved]

[born]

Hello all,

sorry for my bad English, I'm not a native speaker and it's really late.

Yesterday I upgraded from the 2.6.12-gentoo-r10 to the 2.6.13-gentoo-r3 kernel and after a reboot I got the following kernel error message (I can't copy & paste, sorry):

[dom_cyrus]

Hi,
my problem disappeared until I have installed newest vanilla sources kernel-2.6.14-rc4. But I' am still not sure if the problem really disappeard, because I did'nt use my gentoo box very often. But one thing is for sure, that the nvidia x86_64 driver has some problem with all kernels > 2.6.11, so there are two options:

- Installs latest vanilla

- Install nvidia-zander-patch by hand

Note that latest nvidia-zander-patch is not included in the ebuilds from gentoo, but someone told me, that it should be fixed with the newest gentoo-sources. I hope this helps you a little.

[born]

Yeah, but like I write, I do not think the problem is exactly the same like yours, because I still get the error message after disabling the nvidia driver, I do not load it and I have nv in the xorg.conf and still this messages, so it simply could not be the nvidia driver, because it is disabled. Only the symptoms are the same, but also the error message is another, because you got the error in X, mine is in events/0, every time PID 4.

Of course I can try to install the nvidia drivers with the latest patch, but I do not think it would help, since there is also the error without the drivers.

Update: I will try to compile a kernel with default options and see if the problem still remains, if not I have to check my kernel configuration. The only strange thing is that the old kernel worked perfectly and now it hangs. I have run a memtest and it passed several times, so it couldn't be a memory failure.

[born]

O.K., I think I found out what the problem was. do_dbs_timer has something to do with the CPU frequency scaling, so I disabled it in the kernel (so PowerNow! is also disabled ) and everything seems to be fine for now. I did not test it really long, but I use my Gentoo box quite often, so I think I can reopen this thread if there will be problems.

Is it a bug? Should I post it to bgo? Or is it only a misconfiguration of my kernel? Is it realistic (I mean possible is everything) that some PowerNow! functions in my CPU are defect?

I really do not know why this happened. Strange, strange, really strange.

Greets! And thanks for everything

[dom_cyrus]

Maybe its thisone --> http://bugzilla.kernel.org/show_bug.cgi?id=4851
Da du ja auch deutsch verstehst --> http://www.heise.de/newsticker/foren/go.shtml?read=1&msg_id=9058111&forum_id=86419

[born]

It isn't the randomize_va bug and it can't be the SMP bug, because I have only one core in my CPU. But thanks for the links. My Gentoo box didn't crash randomly but always after about 5 minutes of runtime. I think the kernel wanted to slow down the CPU after this time and this produced the bug. I even don't know if it's a real bug or if it is my failure.

[urcindalo]

I use the nvidia driver on AMD64 and the same kernel version and everything works like a charm. Here you are some of my info:

[born]

Do you have PowerNow! enabled in your kernel?

[urcindalo]

This is my kernel config in that regard:

[born]

O.K., thanks!

I will try a new run with your ACPI kernel configuration and report my results. Possibly I had a wrong kernel configuration.