> > invalid kernel stack, page fault in the page fault handler. I don’t think
The latter does not seem to be correct.
PF is considered to be serviced after the first instruction of the handler
is executed.
After this, PFs in PF handler will just cause it to be entered recursively.
To clarify what I said, and If I read the Intel processor docs on double
fault correctly, if you get a page fault while activating the PF handler,
for example the first instruction of the PF handler is not paged in (it’s
supposed to be page locked), it causes a double fault. For other faults, it
looks like it’s valid for the trap handler to be pageable, and the faults
will happen serially. Note that this info is generic for an Intel
processor, and NT/W2K has much stricter requirements on fault handlers.
- Jan
You are currently subscribed to ntdev as: $subst(‘Recip.EmailAddr’)
To unsubscribe send a blank email to leave-ntdev-$subst(‘Recip.MemberIDChar’)@lists.osr.com