Unfortunately, I dont think there is a quick/easy way to check that. You
would have to manually check what
code the offending PCs ( eg. 0x116ce58) correspond to.
Like James suggested, I believe that in your case, OS code is accessing
the same cache blocks as some user transaction.
The reason we print out a warning is that released LogTM implementation
does not correctly handle conflict detection
for escaped stores that overlap with another transaction's read/write set.
However, you might ignore this warning until a consistency violation
between two transactions shows up which would cause
an assert failure.
Jayaram
Jim Leek wrote:
How can I check that?
I haven't used any of the "HANDLER" or "ESCAPE" macros or functions, I
assumed the escape action was just to retry the transaction.
Jim
------------------------------------------------------------------------
_______________________________________________
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.
|
![[Computer Systems Lab]](http://www.cs.wisc.edu/pics/csl_logo.gif){kind=logo}