Re: [Gems-users] Simics-Ruby wakeup interface

[Dan Gibson <degibson@xxxxxxxx>]

The Simics/Ruby interface operates as follows.

Simics Issues a Request to Ruby
Case 1: Fast-Path Hit: Ruby return a stall time of zero.
Case 2: Fast-Path disabled: Ruby enqueues one or more events, depending 
on the protocol, to handle the access. Ruby returns a stall time of two 
billion cycles.
As simulation continues, the Simics processor is stalled. For 
(approximately) each SIMICS_RUBY_MULTIPLIER cycles /observed by Simics/, 
Simics makes one call into Ruby's event queue, nominally RunEventQueue().
Each run of the event queue increments the local time (Ruby_cycles), and 
processes events for the current Ruby_cycle. Eventually, the processing 
of these events handles the request, and Ruby un-stalls the affected 
Simics processor with an explicit API call.
Lastly, Simics re-issues the request to Ruby, a peculiarity of the 
interface. This re-issue is captured in SimicsDriver as I recall, and is 
not observed by higher levels of Ruby in normal operation.
Regards,
Dan


divya ramakrishnan wrote:
> Hi,
>
> I have a very basic question about the flow of control in the 
> interface between Simics and Ruby.
> I understand that for every memory access, Simics wakes up Ruby and 
> adds the transaction to the event queue. How exactly is this event 
> queue triggered [using triggerevents(time)] after the Simics processor 
> is stalled?
> Is Ruby triggered for every cycle after that till it returns a latency 
> and unstalls the processor? Is this triggering done by Simics by 
> waking up Ruby every cycle?
> It would be great if I could understand this basic flow because I am 
> really confused about the network simulation.
> Thank you,
> Divya
> University of Cincinnati
>
> ------------------------------------------------------------------------
> Meet people who discuss and share your passions. Join them now. 
> <http://in.rd.yahoo.com/tagline_groups_7/*http://in.promos.yahoo.com/groups> 
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> 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.
>
>