FW: Multiple writers to sc_signal

John, All-
My email below bounced, I am resending.
Thanks
Stuart

From: Stuart Swan
Sent: Thursday, November 11, 2010 2:50 PM
To: Bishnupriya Bhattacharya; john.aynsley@doulos.com
Cc: Philipp A. Hartmann; P1666 Technical WG
Subject: RE: Multiple writers to sc_signal

John-

I agree with your list below and I agree with Bishnupriya's comment below.

Thanks
Stuart

From: Bishnupriya Bhattacharya
Sent: Thursday, November 11, 2010 8:36 AM
To: john.aynsley@doulos.com
Cc: Philipp A. Hartmann; Stuart Swan; P1666 Technical WG
Subject: RE: Multiple writers to sc_signal

John,

This looks good.

About "remove all other compile-time switches from the PoC simulator (SC_SIGNAL_WRITE_CHECK, DISABLE)", this is upto the OSCI reference simulator and IMO should be left as is for backwards compatibility.

Thanks,
-Bishnupriya

________________________________
From: john.aynsley@doulos.com [mailto:john.aynsley@doulos.com]
Sent: Thursday, November 11, 2010 3:12 PM
To: Bishnupriya Bhattacharya
Cc: Philipp A. Hartmann; Stuart Swan; P1666 Technical WG
Subject: RE: Multiple writers to sc_signal
Stuart, Bishnupriya, Everyone,

Let me restate what I think you are proposing:

"NOT OK" =
   * It is an error for more than one process instance to call sc_signal::write for a given signal instance at any time (excludes elaboration/callbacks/sc_main)
   * It is an error to bind more than one sc_port<sc_signal_inout_if> to a given sc_signal instance
   * The default

"OK" =
   * Multiple process instances may call sc_signal::write for a given signal instance in different deltas
   * It is an error for more than one process instance to call sc_signal::write in the same delta (equivalently, in the same evaluation phase)
   * It is permitted to bind more than one sc_port<sc_signal_inout_if> to a given sc_signal instance

And, presumably, remove all other compile-time switches from the PoC simulator (SC_SIGNAL_WRITE_CHECK, DISABLE) ?

As for the naming, I propose

enum sc_writer_policy {
   SC_ONE_WRITER,
   SC_MULTIPLE_WRITERS
};

virtual sc_writer_policy get_writer_policy() const;

Does that work for everyone?

John A

From:

Bishnupriya Bhattacharya <bpriya@cadence.com>

To:

Stuart Swan <stuart@cadence.com>, "john.aynsley@doulos.com" <john.aynsley@doulos.com>, "Philipp A. Hartmann" <philipp.hartmann@offis.de>

Cc:

P1666 Technical WG <systemc-p1666-technical@eda.org>

Date:

11/11/2010 04:12

Subject:

RE: Multiple writers to sc_signal

________________________________

One addition.

Regarding if multiple sc_inout ports are bound to a sc_signal with MULTIPLE_WRITERS_OK policy, then is it an error or not - we believe it should not be an error.

Thanks,
-Bishnupriya
________________________________

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.