The stat corresponds to the 'workload' transactions (units of work) and
not to TM transactions.
For information on TM transactions look below the line -
Transactional Memory Stats
Jayaram
Jim Leek wrote:
Running logTM with gems2.1 is giving me some odd stats in the dumps.
btree looks good (assuming it's really only supposed to do 96
transactions) but this is what cholesky gives me:
transactions_started: 0 [ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 ]
transactions_ended: 1 [ 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 ]
....
xact_aborts: 3148
Wow, I never ran a single transaction, but I aborted 3148 of them.
Does anyone have any idea what might cause this? Or am I misreading
the stats somehow? (cholesky isn't the only test to give me weird
results like this, raytrace does as well.)
Thanks,
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.
|