Re: [Gems-users] failed assertion in MOESI_CMP_token

[Niket <niketa@xxxxxxxxxxxxx>]

Hi Mike,
Yes that was the case.

I was trying to run my virtual channel flow control network. This was 
not respecting the point to point ordering. But when, I modified it to 1 
virtual channel per virtual network, the ordering is automatically 
respected and my simulation works.
Thanks a bunch for the appropriate and quick response.

Best,
Niket


> Is there any chance that you made interconnect modifications such that 
> point-to-point ordering for the persistent request virtual network was 
> violated?
> --Mike
>
> Niket wrote:
>   
> Hi all,
>   
> I am running the MOESI_CMP_token protocol for 16 processors and the 
> NUCA_ topology. After a while, I get the following assertion failure. 
> Can someone throw some light into the problem. failed assertion 
> 'm_map_ptr->exist(address)' at fn void 
> PersistentTable::persistentRequestUnlock(const Address&, MachineID) in 
> system/PersistentTable.C:132
> Thanks,
> Niket
> _______________________________________________
> 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/"; <https://lists.cs.wisc.edu/archive/gems-users/%22>; to your search.
>