Re: [Gems-users] Signature Conflict Detection

["Haris Volos" <hvolos@xxxxxxxxx>]

Hi Richard -
Yes you are right. It is a scope typo; Under if(! PERFECT_SUMMARY_FILTER),  bool imperfect_result should be imperfect_result

-Haris

On Feb 11, 2008 1:21 PM, Richard Yoo <yoo@xxxxxxxxxxxxxx> wrote:

Hi,

I have a question regarding the
SimicsHypervisor::existSummarySignatureConflict() function in
SimicsHypervisor.C.

if(isRead) {
   // check Summary Write signature
   bool perfect_result =
xact_isol_mgr->isInSummaryWriteSetPerfectFilter(0, addr);
   bool imperfect_result = perfect_result;

   if(! PERFECT_SUMMARY_FILTER) {
       bool imperfect_result =
xact_isol_mgr->isInSummaryWriteSetFilter(0, addr);
       // no false negatives
       assert(!(perfect_result && !imperfect_result));
   }

   if (imperfect_result) {
       return CONFLICT_IS_READER | CONFLICT_WITH_SUMMARY_WRITE;
   } else {
       return 0;
   }
}

Following the above code, imperfect_result variable would always be equal
to perfect_result variable; which means we only use perfect signature in
conflict detection, regardless of whether PERFECT_SUMMARY_FILTER is set or
not.

Is this intentional?
Am I missing something?

Thanks in advance.


-Richard



_______________________________________________
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.