Re: [Gems-users] MOESI_CMP_directory protocol : Invalid Transition

[Mike Marty <mikem@xxxxxxxxxxx>]

Possible protocol bug.  So the bug happens when your DRAM module is used,
right?  Can you reproduce the invalid transition?

The tester is completely deterministic and uses the same random seed by
default (unless you specify it with the -r flag)

If so, I would be curious to see the protocol trace leading up to the bug
for the particular block that has the invalid transition.

The invalid transition will say which block address, and at what time the
invalid transition occurred.  If you start a debug trace (with the -s
<time> tester flag) before the invalid transition occurs and grep for the
block address, this should give a good idea of what happened and I can
probably make a fix.

If you get a trace file, or need more info on getting it, please e-mail
mikem@xxxxxxxxxxx

--Mike


>
> I modified the MOESI_CMP_directory protocol to
> incorporate my dram simulator - but did nto add nay
> additional states. I was using the tester to test that
> I did not have any invalid transitions/deadlocks due
> to my modifications and came across an event that the
> original protocol did not handle.
>
> The combinations is the event L1_PUTX arriving when
> the state of the block in the L2 Cache Controller is
> OLSX.
>
> Thanks
> Brinda
>
>
>
>
> ___________________________________________________________
> How much free photo storage do you get? Store your holiday
> snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
>