Re: BTF Please review and VOTE on B35

From: Anders Nordstrom (andersn@nortelnetworks.com)
Date: Fri Jun 18 1999 - 09:46:29 PDT

Next message: Stuart Sutherland: "Re: Urgent BTF feedback needed"
Previous message: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
In reply to: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
Next in thread: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

BAD MSG:
<x-html><!doctype html public "-//w3c//dtd html 4.0 transitional//en">
html>
<body text="#000000" bgcolor="#FFFFFF" link="#0000FF" vlink="#FF0000" alink="#000088">
Adam Krolnik wrote:
<blockquote TYPE=CITE>Good morning Anders:
Please review the last BNF line change:
output_terms ::= {netl_value, {regl_value}}
Why is the first one required to be a net, the remainder a reg?</blockquote>
X-Lines: 281
Content-Length: 14152
X-Status: $$$$
X-UID: 0000000967
Status: RO

--=_AAAYRwAAG1E3glcI
Content-Type: text/html; charset="us-ascii"
>From ???@??? Fri Jun 18 10:32:10 1999
Return-Path: <owner-btf@boyd.com>
Received: from wmers00q.ca.nortel.com by zmerd004.bnr.ca
with SMTP (Microsoft Exchange Internet Mail Service Version 5.0.1460.8)
id MS6NQ9XD; Fri, 18 Jun 1999 12:46:29 -0400
X-Authentication-Warning: gw.boyd.com: majordomo set sender to owner-btf@boyd.com using -f
Message-ID: <376A77E5.E7BED38@americasm01.nt.com>
Date: Fri, 18 Jun 1999 12:46:29 -0400
From: "Anders Nordstrom" <andersn@nortelnetworks.com>
Reply-To: "Anders Nordstrom" <andersn@nortelnetworks.com>
Organization: Nortel Networks
X-Mailer: Mozilla 4.51 [en] (X11; I; SunOS 5.6 sun4u)
X-Accept-Language: en
MIME-Version: 1.0
To: Adam Krolnik <adamk@cyrix.com>
CC: btf reflector <btf@boyd.com>
Subject: Re: BTF Please review and VOTE on B35
References: <199906181634.LAA09663@gemini.eng.cyrix.com>
Sender: owner-btf@boyd.com
Precedence: bulk
Status: RO
<x-html><!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
<body text="#000000" bgcolor="#FFFFFF" link="#0000FF" vlink="#FF0000" alink="#000088">
Adam Krolnik wrote:
<blockquote TYPE=CITE>Good morning Anders:
Please review the last BNF line change:
output_terms ::= {netl_value, {regl_value}}
Why is the first one required to be a net, the remainder a reg?</blockquote>
X-Sun-Content-Length: 12726

This is a typo it should be:
 <tt>output_terms ::= {regl_value, {regl_value}}</tt>
 <tt></tt>
<blockquote TYPE=CITE>
 Seems it's dependent on the PLA type!
Why are there required curly braces?</blockquote>

The reason for the curly braces is so that the original syntax is
still legal.
 
<blockquote TYPE=CITE>
Shalom's example:
$async$or$array(pla_or_array, { ip0[199:0] } , { out[0:118] } );
Why not?
$async$or$array(pla_or_array, ip0[199:0] , out[0:118] );
Instead, I propose:
input_terms ::= netl_value
 output_terms ::= netl_value | regl_value ( depending on type.)
I vote for this proposed change.
 Adam Krolnik
 Verification Engineer
 Cyrix - NSC.
 Richardson TX. 75085</blockquote>

<pre>--
+-------------------------------------------------------------------------+
| Anders Nordstrom |
| Manager OC192 Data ASIC Development |
| |
| Nortel Networks Email: andersn@nortelnetworks.com |
| P.O. Box 3511 Station C Phone: 613-763-9186 |
| Ottawa, Ontario K1Y 4H7 Fax: 613-763-2626 |
+-------------------------------------------------------------------------+</pre>

</body>
</html>
</x-html>From ???@??? Sun Jun 20 11:27:55 1999
Return-Path: <owner-btf@boyd.com>
Received: (from majordomo@localhost)
 by gw.boyd.com (8.9.3/8.9.3) id LAA13765
 for btf-list; Sun, 20 Jun 1999 11:22:26 -0700
X-Authentication-Warning: gw.boyd.com: majordomo set sender to owner-btf@boyd.com using -f
Received: from hub1.san-jose.webnexus.net (hub1-20.san-jose.webnexus.net [209.140.224.20])
 by gw.boyd.com (8.9.3/8.9.3) with ESMTP id LAA13733
 for <btf@boyd.com>; Sun, 20 Jun 1999 11:22:14 -0700
Received: from smtprtp.nortel.com (smtprtp.NortelNetworks.com [192.122.117.66])
 by hub1.san-jose.webnexus.net (8.9.1a/8.8.5/WN-1.3) with ESMTP id IAA18678
 for <btf@boyd.com>; Sun, 20 Jun 1999 08:59:32 -0700 (PDT)
Received: from zcard00n.ca.nortel.com (actually zcard00n) by smtprtp.nortel.com;
 Sun, 20 Jun 1999 11:55:57 -0400
Received: from zmerd004.bnr.ca (zmerd004.ca.nortel.com [47.124.0.133])
 by zcard00n.ca.nortel.com
 with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0)
 id N2F4NY0Q; Sun, 20 Jun 1999 11:55:57 -0400
Received: from wmers00q.ca.nortel.com by zmerd004.bnr.ca
 with SMTP (Microsoft Exchange Internet Mail Service Version 5.0.1460.8)
 id MS6NRADV; Sun, 20 Jun 1999 11:55:56 -0400
Message-ID: <376D0F0C.4D269100@americasm01.nt.com>
Date: Sun, 20 Jun 1999 11:55:56 -0400
From: "Anders Nordstrom" <andersn@nortelnetworks.com>
Reply-To: "Anders Nordstrom" <andersn@nortelnetworks.com>
Organization: Nortel Networks
X-Mailer: Mozilla 4.51 [en] (X11; I; SunOS 5.6 sun4u)
X-Accept-Language: en
MIME-Version: 1.0
To: Adam Krolnik <adamk@cyrix.com>
CC: btf@boyd.com
Subject: Re: BTF Please review and VOTE on B35
References: <199906181657.LAA09991@gemini.eng.cyrix.com>
Sender: owner-btf@boyd.com
Precedence: bulk
Status: RO
<x-html><!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
<body text="#000000" bgcolor="#FFFFFF" link="#0000FF" vlink="#FF0000" alink="#000088">
Adam,
OK, I agree. We want to remove the mandatory concatenation.
 I have updated the proposal and also removed a sentence that said you
have to
 have the concatenation.
Regards,
 Anders
 
 
Adam Krolnik wrote:
<blockquote TYPE=CITE>Good afternoon Anders:
I still vote for my proposed syntax.
Check netl_value and regl_value;
Both include as an option concatenation - which allows
 a single element!
netl_value ::=
 net_identifier
 | net_identifier [ expression ]
 | net_identifier [msb_constant_expression : lsb_constant_expression
]
 | net_concatentation
concatenation ::= { expression {, expression} }
The old way still works, you also can specify a single vector
 as an input and an output - without the required concatenation braces.
 Adam Krolnik
 Verification Engineer
 Cyrix - NSC.
 Richardson TX. 75085</blockquote>

</body>
</html>
</x-html><x-html><HTML>
<HEAD>
<TITLE> B35 </TITLE>
</HEAD>
<BODY BGCOLOR="#FFFFFF">

 
<HR SIZE=5 NOSHADE>

<H2> B35 - PLA System Task Enhancement </H2>

<TABLE BORDER COLS=2 WIDTH="75%" >
<TR><TD>
Section: </TD><TD> 14.6 page 197
</TD></TR><TR><TD>
Date Submitted: </TD><TD> 980827
</TD></TR><TR><TD>
Requestor: </TD><TD> Shalom Bresticker (shalom@msil.sps.mot.com)
</TD></TR><TR><TD>
Status: </TD><TD> Proposal
</TD></TR><TR><TD>
Analyzed by: </TD><TD> Anders Nordstrom
</TD></TR><TR><TD>
Synthesizable: </TD><TD> No
</TD></TR>
</TABLE>

<H3> Details </H3>

The PLA system tasks are described in section 14.6.

The arguments of the tasks are in the form:
<PRE>
( memory_name, input_terms, output_terms )
</PRE>
where input_terms and output_terms are defined as:
<PRE>
input_terms ::= { scalar_variables }
output_terms ::= { scalar_variables }
</PRE>
That is, the input and output terms are concatenations of scalar variables,
where a scalar variable is either a scalar identifier or
a bit select of a vector.

But vectors and part-selects are not allowed !
<P
This is a tremendous absurdity!

Why can't
<PRE>
$async$or$array(pla_or_array, { ip0[199],
 ip0[198], ip0[197], ip0[196], ip0[195], ip0[194], ip0[193], ip0[192], ip0[191], ip0[190], ip0[189],
 ip0[188], ip0[187], ip0[186], ip0[185], ip0[184], ip0[183], ip0[182], ip0[181], ip0[180], ip0[179],
 ip0[178], ip0[177], ip0[176], ip0[175], ip0[174], ip0[173], ip0[172], ip0[171], ip0[170], ip0[169],
 ip0[168], ip0[167], ip0[166], ip0[165], ip0[164], ip0[163], ip0[162], ip0[161], ip0[160], ip0[159],
 ip0[158], ip0[157], ip0[156], ip0[155], ip0[154], ip0[153], ip0[152], ip0[151], ip0[150], ip0[149],
 ip0[148], ip0[147], ip0[146], ip0[145], ip0[144], ip0[143], ip0[142], ip0[141], ip0[140], ip0[139],
 ip0[138], ip0[137], ip0[136], ip0[135], ip0[134], ip0[133], ip0[132], ip0[131], ip0[130], ip0[129],
 ip0[128], ip0[127], ip0[126], ip0[125], ip0[124], ip0[123], ip0[122], ip0[121], ip0[120], ip0[119],
 ip0[118], ip0[117], ip0[116], ip0[115], ip0[114], ip0[113], ip0[112], ip0[111], ip0[110], ip0[109],
 ip0[108], ip0[107], ip0[106], ip0[105], ip0[104], ip0[103], ip0[102], ip0[101], ip0[100], ip0[99],
 ip0[98], ip0[97], ip0[96], ip0[95], ip0[94], ip0[93], ip0[92], ip0[91], ip0[90], ip0[89],
 ip0[88], ip0[87], ip0[86], ip0[85], ip0[84], ip0[83], ip0[82], ip0[81], ip0[80], ip0[79],
 ip0[78], ip0[77], ip0[76], ip0[75], ip0[74], ip0[73], ip0[72], ip0[71], ip0[70], ip0[69],
 ip0[68], ip0[67], ip0[66], ip0[65], ip0[64], ip0[63], ip0[62], ip0[61], ip0[60], ip0[59],
 ip0[58], ip0[57], ip0[56], ip0[55], ip0[54], ip0[53], ip0[52], ip0[51], ip0[50], ip0[49],
 ip0[48], ip0[47], ip0[46], ip0[45], ip0[44], ip0[43], ip0[42], ip0[41], ip0[40], ip0[39],
 ip0[38], ip0[37], ip0[36], ip0[35], ip0[34], ip0[33], ip0[32], ip0[31], ip0[30], ip0[29],
 ip0[28], ip0[27], ip0[26], ip0[25], ip0[24], ip0[23], ip0[22], ip0[21], ip0[20], ip0[19],
 ip0[18], ip0[17], ip0[16], ip0[15], ip0[14], ip0[13], ip0[12], ip0[11], ip0[10], ip0[9],
 ip0[8], ip0[7], ip0[6], ip0[5], ip0[4], ip0[3], ip0[2], ip0[1], ip0[0]},
{out[0], out[1], out[2], out[3], out[4], out[5], out[6], out[7], out[8],
 out[9], out[10], out[11], out[12], out[13], out[14], out[15], out[16], out[17], out[18],
 out[19], out[20], out[21], out[22], out[23], out[24], out[25], out[26], out[27], out[28],
 out[29], out[30], out[31], out[32], out[33], out[34], out[35], out[36], out[37], out[38],
 out[39], out[40], out[41], out[42], out[43], out[44], out[45], out[46], out[47], out[48],
 out[49], out[50], out[51], out[52], out[53], out[54], out[55], out[56], out[57], out[58],
 out[59], out[60], out[61], out[62], out[63], out[64], out[65], out[66], out[67], out[68],
 out[69], out[70], out[71], out[72], out[73], out[74], out[75], out[76], out[77], out[78],
 out[79], out[80], out[81], out[82], out[83], out[84], out[85], out[86], out[87], out[88],
 out[89], out[90], out[91], out[92], out[93], out[94], out[95], out[96], out[97], out[98],
 out[99], out[100], out[101], out[102], out[103], out[104], out[105], out[106], out[107], out[108],
 out[109], out[110], out[111], out[112], out[113], out[114], out[115], out[116], out[117], out[118]});
</PRE>
be written simply as
<PRE>
$async$or$array(pla_or_array, { ip0[199:0] } , { out[0:118] } );
</PRE>
(This is just an example. Our real pla is about 8 times larger!)

Compilers are smart enough these days!!
 
In addition, the arguments of the tasks are in the form:
<PRE>
( memory_name, input_terms, output_terms )
</PRE>
where input_terms and output_terms are defined as:
<PRE>
input_terms ::= { scalar_variables }
output_terms ::= { scalar_variables }
</PRE>
However, "scalar_variables" is not defined.

Nor is it written that while input_terms may be wires,
output_terms must be regs.
 
There is neither an explanation nor an example.
 
Without these, it is not possible to even guess how to use them.
 
The truth is, the Cadence VXL RM is not much better,
but it does show this example:
<PRE>
forever @(posedge clk)
$sync$and$array(mem, { ... }, { ... } ) ;
</PRE>
>From this, it is possible to guess that the $async tasks
are like continous assignments: the outputs are re-evaluated
whenever one of the inputs changes.

In contrast, the $sync tasks are like procedural assignments:
the outputs are re-evaluated only when the task is explicitly
called. To re-evaluate the outputs again, you have to call the
task again.
 
This information about how to use the $sync tasks is essential
and currently missing.

Sincerely,
<PRE>
******************************************************************************
Shalom Bresticker email: shalom@msil.sps.mot.com
Motorola Semiconductor Israel, Ltd. Tel #: +972 9 9522268
P.O.B. 2208, Herzlia 46120, ISRAEL Fax #: +972 9 9522444
http://www.motorola-semi.co.il/
******************************************************************************
</PRE>

<H3> Proposal </H3>

Change Syntax 14-9 on page 197 in LRM 1995 from:

<PRE>
input_terms ::= { scalar_variables }
output_terms ::= { scalar_variables }
</PRE>
to
<PRE>
input_terms ::= netl_value
output_terms ::= regl_value
</PRE>
Remove the last sentence in section 14.6.1 on page 197 in the 1995 LRM. 
"Note that the input terms and output terms are always represented as concatenations."

Change the first example in section 14.6.1 to:
<PRE>

wire a1, a2, a3, a4, a5, a6, a7;
reg b1, b2, b3;
wire [7:1] awire;
reg [3:1] breg;

$async$and$array (mem, {a1,a2,a3,a4,a5,a6,a7}, {b1,b2,b3});
or
$async$and$array (mem, awire, breg);

</PRE>

<HR SIZE=5 NOSHADE>
</BODY>
</HTML>
</x-html>
Attachment Converted: "E:\INCOMING\andersn.vcf"
--=_AAAYRwAAG1E3glcI--

Next message: Stuart Sutherland: "Re: Urgent BTF feedback needed"
Previous message: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
In reply to: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
Next in thread: Adam Krolnik: "Re: BTF Please review and VOTE on B35"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by hypermail 2.1.4 : Mon Jul 08 2002 - 12:53:28 PDT and
sponsored by Boyd Technology, Inc.