System freeze woes (warning, semi-long post)

[Shades3D]

SO recently I decided to format my system and this time I decided to dualboot with WinXP and Gentoo. I had used gentoo on my desktop previously and decided to give the 2.6.x line of kernels a try.

After installing XP on one of two partitions on my main hdd, I grabbed the latest livecd (2005.0) and threw it in my cd drive. I booted into gentoo's live cd interface and went ahead with the install. But oh no! It kept randomly freezing on me! Curious, I decided to throw in an older livecd (2004.0) and test that, since I know that one works, I had used it on this system before (well, almost, as you'll see later). Lo and behold, more freezing.

So I decided to head to the gentoo forums. Maybe there'd been a spring of problems coming up recently that I can try solutions for. Did some searches related to random system lockups/freezes/hangs and came across something interesting, the nForce2 APIC bug. I decided to try out some things. I passed noapic acpi=off to the livecd kernel. It seemed things were good, as I got through the install. I setup my kernel (vanilla sources, 2.6.11.10) without APIC (without really realizing since its defaulted to off) and went ahead and booted. Got back into my fresh system and started my emerge of x and xfce4. The system kept hanging at random spots during the install. Once again, frustration ensued.

I once again went on a hunt for what the problem could be. This time I noted the idea of LOCAL APIC being a cause. So I, once again, passed a few args to my kernel (noapic, nolapic, acpi=off) and booted. Hey, look at that, I went all the way through the xfce4 emerge, I emerged firefox and a few other misc things. Seems my system is stable. Hurrah hurrah. I've won the battle, and am now tired of dealing with emerging and system problems for the moment, and since I have much more to emerge, I decided to take a break and boot into XP.

When I, once again, decided to go into gentoo and get back to setting up my desktop, the system freezes once again resurfaced. I was having them pop up randomly. I did more research, found out that there's a fix for it already in the kernel tree. Checked dmesg, found that it was active on my system, so things SHOULD be good.

So to summarize myself in case you guys don't really want to read the entire thing:

- Random system freezes are happening
- Tried both vanilla sources (2.6.11.10) and gentoo-sources (2.6.11-gentoo-r9)
- No pattern, happens randomly
- APIC bugfix is in effect, but seemingly ineffecitve
- Freezes cause everything to halt, no remote access either
- System hardware:
-- ASUS A7N8X-Deluxe rev2.0 nforce2 motherboard (bios: 1008)
-- AMD Athlon T-bird 1.4GHz Processor
-- 1GB Corsair PC3200 RAM (2x512, dual channel)
-- 40GB Western Digital
-- 40GB Maxtor
-- 160GB Hitachi
-- Lite-on SOHW 832S
-- Hercules Game Theater XP Sound card
-- nVidia GeForce FX 5600 Pro
-- I am using onboard networking (3com 3c59x)
- I've tried all I can seem to find related to the nforce2 APIC problem, minus disabling it in the bios since WinXP wouldn't like that, and I'd have to reinstall that, which I'm saving as a last resort.

Any suggestions or ideas?

EDIT: Oh, I frogot to add, I used to be on an EPoX board with a VIA chipset back when I was stable.
_________________
(Insert witty and shady signature here)

[kopfsalat]

Sounds farfetched but it could also be a bad power supply. If you got a spare one lying around you may try that one out.
It may seem strange but I myself had freezes with linux while windows would run fine. I built in a new power supply and all was good afterwards.

[Shades3D]

well, I tested the power supply idea and it didn't work Popped another 350w psu in there and gentoo still froze at random spots. To make sure it wasn't just a gentoo issue, I popped knoppix in there as well. Still had the random freezes. Linux in general seems to freeze up randomly, even as WinXP stays solid. This bugs me Any other ideas?
_________________
(Insert witty and shady signature here)

[Momo_CCCP]

Is there anything interesting left in /var/log/messages after you have a freeze ?

Have you checked your RAM (shouldn't be a problem if XP works fine, but you never know) ?
_________________
Momo_CCCP

[Shades3D]

No errors in /var/log/messages

And I ran memtest for a bit with no errors. This thing just simply baffles me.
_________________
(Insert witty and shady signature here)

[kmare]

I'm having the same problem.. my system freezes randomly. I thought it was the nvidia driver, but I'm not so sure anymore. I have the nforce2 chipset too.. I'm starting to think that it is the motherboard. In the nvidia forums I read that ppl when they changed their motherboard (manufacture, with the same chipset) with everything else exactly the same, the freezes disappeared, everything was stable again.. also updating the mobo bios helped, but unfortunally for me there's no new bios for my MSI
if this thing persist I'm going to buy a new system, even if I really don't like the idea...

edit: I see this problem mostly when I have high cpu usage for a long time (eg when compiling..)

[enkil]

[kmare]

I don't have SATA.. but I'll try the 2.6.12-r1 just to be sure (using 2.6.12 right now). Thank you for your reply though.. I'll monitor the cpu temperature as soon as I get the lm_sensors to work..

[Shades3D]

I've run memtest a ton of times, did several passes with no errors. Swapped PSUs with the same results. Tried a knooppix cd with teh same results. This sucks :/ *sigh*
_________________
(Insert witty and shady signature here)

[jschellhaass]

I had the same issue. Looks like mine was temp related. I bought a couple of new case fans and so far so good. I will not say it is fixed until I have a week or two of running without a lockup.

You may want to also try running cpuburn (emerge cpuburn) and see if casues lockups.

jeff