2003-01-19 18:24:06

by Sampson Fung

[permalink] [raw]
Subject: A Flightening and Strange experience compiling 2.5.58

2 days a ago, I am compiling 2.5.58 to test Intel810fb.

After serveal compiles , my machine stop responding with a repeating
message shows up in console:

<iDDDDD> setting is too high, resetting... (where D is a digit)

(I did not record the details, so please don't blame me not providing
details, just want to share my experience).

At this point, my machine is total not responding, ping from other
machine failed, console keyboard not responding, etc.

But the error mesage is repeating.

Then I press reset button.

After reset, my box ask me to insert a boot disk in Drive A. (I do not
see any POST occuring, no memory counting, etc).

After pressing a few <return>s, I unplug the power and turn on the
machine again.

Same problem. After a few retries (unplug power and power on again), my
heart is fearing my mainboard's BIOS is damaged.
(I have a similar experience after flash a BIOS and result to a useless
mainboard.)

Then I unplug my machine and going somewhere.

After about 2 hours, I power on the machine again.

This time, after POST, BIOS reports CMOS data checksum failed. And
allow me to into BIOS setup to config again.

After that, my machine is working normally again.

And I have totally no hints to what happened.

Does anyone has similar problems? That is: H/W POST do not start after
machine hangs.

Sorry for not able to provide more details again.

Sampson Fung
A New Comer to Kernel Testing.



2003-01-19 18:47:25

by Dmitri

[permalink] [raw]
Subject: Re: A Flightening and Strange experience compiling 2.5.58

On Sun, 2003-01-19 at 10:32, Sampson Fung wrote:

> Does anyone has similar problems? That is: H/W POST do not start after
> machine hangs.

Sure. I have one m/b here that:

a) enters POST only if the planets are in favorable position
b) once running, it sometimes stops unpredictably; then goto a;

I suspect a crack in the m/b, since the RAM was tested in other boxes
and appears to be fine. I already retired this m/b, and it is on its way
to the trash.

Messages that you saw are probably related to some hardware that went
belly up while the box was running, and the code mistook the hardware
failure for some valid condition that only asked for a reset.

Generally, if the box does not enter POST after hard reset then the m/b
or other major components such as RAM or CPU are broken. This is by
definition, because RESET is supposed to reinit all the h/w, and any
memory of the past is not supposed to be there.

Dmitri


Attachments:
signature.asc (189.00 B)
This is a digitally signed message part