Help!
My machine seems to be rebooting at random. Actually, it's more like
the screen blanks, and suddenly the BIOS is going through POST. There
may be a reset-button gnome in my case putting a jumper over the reset
pins, but I seriously doubt it. ;-) I recently tried to switch from APM
to ACPI, when i upgraded from ac9 to ac11, so I thought that might be
the cause. So I switched back to APM with ac13 (I had the same ac12
compile problems as the rest of the world). It's still rebooting,
without any aparent cause.
I upgraded the BIOS on this Asus A7V sometime in the past week, but I
honestly don't remember when. From 1005C to 1007. This was released in
march, so I assumed it was pretty stable, but it could be the cause.
I'm going to go downgrade now, but is this more likely to be a kernel
bug, or a hardware bug/new bios bug?
Thanks,
Nathan
On Tue, Apr 24, 2001 at 12:18:42PM -0400, Nathan Walp wrote:
> I upgraded the BIOS on this Asus A7V sometime in the past week, but I
> honestly don't remember when. From 1005C to 1007. This was released in
> march, so I assumed it was pretty stable, but it could be the cause.
> I'm going to go downgrade now, but is this more likely to be a kernel
> bug, or a hardware bug/new bios bug?
No problem here. I'm using 1007 since its release in first half
of March.
Linux version 2.4.3-ac12-amd (root@ppc) (gcc version 3.0 20010402 (Debian prerelease)) #1 Mon Apr 23 02:31:13 CEST 2001
...
Kernel command line: BOOT_IMAGE=Linux ro root=2105 video=matrox:vesa:0x105,fv:85 devfs=nomount
Initializing CPU#0
Detected 1009.013 MHz processor.
...
CPU: AMD Athlon(tm) Processor stepping 02
Enabling fast FPU save and restore... done.
...
apm: BIOS version 1.2 Flags 0x03 (Driver version 1.14)
...
Except that I got 'ide only func: 14' today with my Promise PDC20265
(which looks strange to me - either all Intel, VIA and Promise have
same bug in their UDMA hardware, or there is a bug in Linux IDE
driver...) (I had to reboot because of kernel somehow believed that
it read some garbage instead of MBR from hdg, so I could not run my
repartitioning session, as I found no way to invalidate hdg kernel
cache).
But machine for sure does not spontaneously reboot. And I have
enabled local apic in kernel configuration. All previous kernels
were built with Debian's 2.95.3-something, ac12 was built with
gcc-3.0, as I wanted to update anyway, and ac12 just gave me a reason.
Best regards,
Petr Vandrovec
[email protected]
I've gotten a lot more response to this than I'd ever dreamed, but I
figured out the problem on my own...
I have (had) one of those exhaust fans that fits into a PCI/ISA slot,
and it died. Not only did it die, but it started creating heat of it's
own. I don't think the video card adjacent to it appreciated this a
whole lot. I removed the dead fan, and everything is back to normal, I
just need to go get a new fan to make myself feel better ;-)
The 1007 bios, and ac14 have been perfectly stable, so no worries about
either.
Thanks,
Nathan
Petr Vandrovec wrote:
>
> On Tue, Apr 24, 2001 at 12:18:42PM -0400, Nathan Walp wrote:
> > I upgraded the BIOS on this Asus A7V sometime in the past week, but I
> > honestly don't remember when. From 1005C to 1007. This was released in
> > march, so I assumed it was pretty stable, but it could be the cause.
> > I'm going to go downgrade now, but is this more likely to be a kernel
> > bug, or a hardware bug/new bios bug?
>
> No problem here. I'm using 1007 since its release in first half
> of March.
>
> Linux version 2.4.3-ac12-amd (root@ppc) (gcc version 3.0 20010402 (Debian prerelease)) #1 Mon Apr 23 02:31:13 CEST 2001
> ...
> Kernel command line: BOOT_IMAGE=Linux ro root=2105 video=matrox:vesa:0x105,fv:85 devfs=nomount
> Initializing CPU#0
> Detected 1009.013 MHz processor.
> ...
> CPU: AMD Athlon(tm) Processor stepping 02
> Enabling fast FPU save and restore... done.
> ...
> apm: BIOS version 1.2 Flags 0x03 (Driver version 1.14)
> ...
>
> Except that I got 'ide only func: 14' today with my Promise PDC20265
> (which looks strange to me - either all Intel, VIA and Promise have
> same bug in their UDMA hardware, or there is a bug in Linux IDE
> driver...) (I had to reboot because of kernel somehow believed that
> it read some garbage instead of MBR from hdg, so I could not run my
> repartitioning session, as I found no way to invalidate hdg kernel
> cache).
> But machine for sure does not spontaneously reboot. And I have
> enabled local apic in kernel configuration. All previous kernels
> were built with Debian's 2.95.3-something, ac12 was built with
> gcc-3.0, as I wanted to update anyway, and ac12 just gave me a reason.
> Best regards,
> Petr Vandrovec
> [email protected]