On Fri, 17 Sep 1999, Chris Rutter wrote:
chris>On Thu, 16 Sep 1999, John Lapeyre wrote:
chris>
chris>> The 2.0.37 and 2.2.x kernels keep hanging on my AMD K6-2.
chris>
chris>This sounds *bad*, BTW; have you checked around to see if anyone
chris>else has had these kinds of freezing problems? Is your machine
chris>unstable in any other way?
chris>
chris>You may find all you need to do is tweak a CPU register or two,
chris>or apply some patch to the kernel to make the machine stable on
chris>any kernel you like -- it's worth checking, because the kernel
chris>*shouldn't* have become randomly unstable in 2.0.37.
I found a few reports on the kernel mailing list. But, somehow,
the search engine did not pick up references to AMD that I remembered. It
is difficult to get controled information on bugs. Some people find
problems and eventually admit that it was a hardware problem. Tweaking a
register would be fine. I really don't know how to look into that,
however.
It really seems that something changed. I built the 200 MB tar
file about 30 times under 2.0.36, and it was fine. Under 2.0.34, the file
was built every night for over a year, w/o crashing. With 2.0.37 and
2.2.x, it is not totally predictible, but the machine hangs on roughly
half the attempts to make the 200 MB file. As I said, I don't know
enough to say to what extent hardware and the compiler are playing a part.
It is, of course, quite time consuming to run these tests. I will post
something to the kernel list.
John Lapeyre <[EMAIL PROTECTED]>
Tucson,AZ http://www.physics.arizona.edu/~lapeyre
