Re: [Gems-users] FILE_SPECIFIED topology

[Mike Marty <mikem@xxxxxxxxxxx>]

These are the parameters in rubyconfig.defaults that controls
FINITE_BUFFERING:

// NOTE: Finite buffering allows us to simulate a realistic virtual
cut-through
// routed network with idealized flow control.
FINITE_BUFFERING: false
// All message buffers within the network (i.e. the switch's input and
// output buffers) are set to the size specified below by the
FINITE_BUFFER_SIZEFINITE_BUFFER_SIZE: 3


I would grep the code for those.

m_strict_fifo means that messages must be placed on the buffers in FIFO
ordering, and I think adaptive routing is disabled.

--Mike


> Hi Mike,
>              Thanks a lot for the prompt reply. Can you please point
> me to the code where the support for finite buffering can be found.
> And also, in ruby/buffers/MessageBuffer.C - I see that  m_strict_fifo
> = true; is being set, so the buffers are strictly FIFO, right? The
> events are organized as a priority heap, right?
>
>
> On 10/19/06, Mike Marty <mikem@xxxxxxxxxxx> wrote:
> > Bi-directional links.  Default is infinite buffer sizes, however there is
> > support for finite buffering for any topology.
> >
> > --Mike
> >
> >
> > > Hi All,
> > >            Is the buffer size infinite for FILE_SPECFIED topology too?
> > > And also are the links bi-directional or uni-directional?
> > >
> > > thanks,
> > > Mishali
> > > _______________________________________________
> > > 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.
> > >
> > _______________________________________________
> > 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.
> >
> >
>