- ...Carnival
- This research was supported by NSF grant CCR-9510173.
Wagner Meira, Jr. is supported by CNPq-Brazil, Grant 200.862/93-6.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...multiprogramming.
- Multiprogramming that occurs
while a process is waiting is counted as waiting time.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...time.
- Our discussion is in terms of
pair-wise synchronization, but the techniques extend
to the analysis of waiting time for synchronization
operations between many processors. In particular, we characterize
waiting time in barriers as a set of pair-wise synchronization operations
between each waiting processor and the last processor to enter the barrier.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...speed.
- We
are currently investigating the use of waiting time analysis in
analyzing memory reference behavior
by instrumenting a software coherence protocol.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...time.
-
We chose 95% as our initial level of tolerance for this algorithm,
so we remove blocks only as long as the remaining blocks
account for 95% of the actual waiting time.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...class.
- The representative for an equivalence class is
never changed, ensuring that the equivalence relation is independent
of the order in which paths are processed. The characterization for
a class is updated each time a new path matches the representative for that
class, reflecting the time spent in the new path.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...path.
- We
compute the relative duration to avoid relying on the exact duration
of steps or the precise number of iterations in a loop
in determining equivalence of execution paths.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
- ...Cholesky.
- The implementations of Cholesky on
the SGI and SP2 are slightly different, and cannot be compared directly.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.