:> Yes, the testing and the cores help considerably. It crashed in the
:> same place as last time. I had hoped that the bug was simply me
:> forgetting to reload the globaldata pointer, but I fixed that and it
:> still occurs.
:mh I'm waiting for patch3, ok? or do you need more vmcores?
So far for these failures one vmcore/kernel is sufficient.
I couldn't find it from looking at the code, so here is another patch
which puts assertions in where I think the problem is occuring.
Hopefully we will get a panic that is closer to the bug.
fetch http://apollo.backplane.com/DFlyMisc/tsleep03.patch
-Matt
Matthew Dillon
<[EMAIL PROTECTED]>
