Re: [Gems-users] L2_Cache_Controller Identifier

[Mike Marty <mikem@xxxxxxxxxxx>]

The ID used in the network file is from 0 to the number of L2 banks.
However depending on the g_PROCS_PER_CHIP parameter, the particular L2
bank is located differently.  Each Chip object contains a vector of L2
controllers.  Each L2 controller has a "m_version" field to identify which
L2 controller it is in the CHip.

Each L2 controller _instance_ has a wakeup().  The queues are searched in
the order they are specified in the controller.

--Mike


> Hi all. I'm trying to debug ruby by using gdb debugger with
> tester.exec; as ruby configuration, I use MOESI_CMP_NUCA as protocol
> and "Network File Specified" as Network topology. The system is an
> 8-CMP, and L2cache has 32 banks.
> My question is:
>
> What is the relationship between L2_cache_controller identifier and
> the ID used in the network file for the L2_banks?
>
> The problem is that, by using the debugger,  it seems to me that, when
> a block is searched over L2-cache, the first bank to be scanned
> (L2_cache_controller::wakeup() ) is the one considered "local" respect
> to the L1-cache which has forwarded the request, according to
> Ruby_Slicc_Component_mapping functions; but looking inside the Network
> File, the bank with the same ID really does not seem to be local to
> L1_cache.
>
> Sorry to be a nuisance.
> Thank you very much.
> _______________________________________________
> 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.
>