Re: Future of SID

["Frank Ch. Eigler" <fche at redhat dot com>]

Hi -


> I was wondering what are the future plans for SID.
> [...]
> But I need to make more complex designs, an N-way SMP for example [...]

Some degree of SMP modelling is straightforward: just instantiate N
CPU components, and wire them up in a sid configuration file in
whatever way is appropriate.  Some of the flexibility can be expressed
using the C++ configuration frontend's command line options.

> [...] I looked for components such as a TLB that I could plug on the
> bus,

A complex TLB may be pretty hard to model as an isolated sid
component, because of the interconnection with a typical MMU's
page-table-fetching logic.

> or DMA controller that I can use to generate IO traffic, but could
> not find any.

If you're not sure what flavour controller you'd like, then invent a
very simple one, and perhaps script it in tcl.  Or use a secondary
general purpose processor with DMA-only "firmware".

> [...]  Are you planning to add examples for setting various type of
> boards ? [...]

Can you suggest an example to help me see what you mean?  In some
cases, supporting a "board" just means assembling a configuration from
existing components, which is relatively simple.  In others, it
requires writing new components, which can span the whole spectrum of
complexity.

> What about the support for 'bochs' (x86). Are you going to continue
> work on that when new 'bochs' is released? [...]  The bottom
> question is what are you planning for the next couple of years for
> SID. [...]

Red Hat is contracted to continue development of SID-based simulator
ports, some of which are made public on an ongoing basis.  These are
generally focused efforts toward particular customers' requirements,
though occasionally create nice general-purpose components.  There are
no plans at present to dedicate manpower to improving SID for its own
sake.  We would of course be happy to help users help themselves.


- FChE

Attachment: pgp00000.pgp
Description: PGP signature