James,
This should never happen. The released LogTM protocols only abort
when a load or store instruction is nacked during a transaction.
Unless the kernel is transactional, exception handlers shouldn't get
nacked. But, if there was an abort in the middle of a logged
exception, you could see any number of bad outcomes including kernel
panics. LogTM does not restore the pstate register or the default
memory context so it's possible that processor would run user code in
kernel mode operating in the kernel's address space.
--Kevin
On Apr 24, 2007, at 5:15 PM, James Wang wrote:
Hi All:
What would happen if LogTM abort a transaction
when a logged exception just started but did not end?
Would LogTM ever encounter that case?
Thanks in advance for any reply.
Regards
James
_______________________________________________
Gems-users mailing list
Gems-users@xxxxxxxxxxx
https://lists.cs.wisc.edu/mailman/listinfo/gems-users
Use Google to search the GEMS Users mailing list by adding
"site:https://lists.cs.wisc.edu/archive/gems-users/" to your search.
|