Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarApr2017

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search

Messages from 26850

Article: 26850
Subject: Reference Design Xapp205.zip
From: ChenSongWei <chen.songwei@mail.zte.com.cn>
Date: Wed, 1 Nov 2000 03:16:21 -0700
Links: << >>  << T >>  << A >>
hello: when i am doing timing simulator(using the reference design xapp205.zip :fifoctlr_ccmw1.v/fifoctlr_ccmw2.v) for a project,the follwoing message occures: "P1/U1/U11/BU0/INTERNAL_BLOCKRAM--READ Violation.Attempt to read from cell that is also being written to." the capacity of the dual block ram i mentioned is 16x1024--64x256,that is,when write_enable is active,read_enable must inactive.But when i replace it with 16x2048---64x512,it is right,and the phenomena of the warning is not . why?

Article: 26851
Subject: Re: Alliance 3.2i
From: Gary Cook <gc@sonyoxford.co.uk>
Date: Wed, 01 Nov 2000 11:02:52 +0000
Links: << >>  << T >>  << A >>
It's the crashing during p&r bug .... recently reintroduced by
popular demand ... coming to you at a p&r run soon!

Fixed by setting VCCGND_OFF=1

Gary.

Ray Andraka wrote:

> I'm using it.  It fixes a number of bugs, but there is one that it created.
> Can't remember off hand which one it was, as I currently have several cases
> open.  It is supposed to be fixed in the service pack due out next week.
>
> "S. Ramirez" wrote:
> >
> > Dear Newsgroupies,
> >      Has anyone here used Xilinx 3.2i?  I am presently using 3.1i, with
> > service_pack_31i_3_2_03i_pc and am fairly happy with it.  Does 3.2i with SP4
> > bring anything great to the table?  Does it have any bugs that anyone has
> > uncovered?
> >      Thanks.
> > -Simon Ramirez, Consultant
> >  Synchronous Design, Inc.
>
> --
> -Ray Andraka, P.E.
> President, the Andraka Consulting Group, Inc.
> 401/884-7930     Fax 401/884-7950
> email ray@andraka.com
> http://www.andraka.com  or http://www.fpga-guru.com

Article: 26852
Subject: Re: Alliance under Linux?
From: steve (Steve Rencontre)
Date: Wed, 1 Nov 2000 15:36 +0000 (GMT Standard Time)
Links: << >>  << T >>  << A >>
In article <SeJL5.25863$68.6416528@typhoon.tampabay.rr.com>, 
sramirez@deleet.cfl.rr.com (S. Ramirez) wrote:

>      That sucks!  I thought all those Unix systems were compatible

"Compatible" in Unix-land has never really meant binary image compatible. 
Frequently, it means "you can recompile the source with only a modest 
amount of aggro (provided you know what you're doing)" :-)

--
Steve Rencontre		http://www.rsn-tech.co.uk
//#include <disclaimer.h>

Article: 26853
Subject: Re: Spartan II ?
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Wed, 01 Nov 2000 17:22:13 +0000
Links: << >>  << T >>  << A >>


Paul Smith wrote:

> Leon Heller wrote:
>
> >
> > I've just been quoted 14-18 weeks lead-time for some XC2S50-5TQ144C by
> > our distributor, Insight Memec, in the UK
> >
>
> I was just quoted 16 weeks for XC2S30-5VQ100C
>
> Anyone able to buy XC2S30 or XC2S50 with reasonable leadtimes?

Maybe somebody should remind Xilinx and/or the disti's of the concept of
time-to-market.
Around here 16 weeks is well beyond the no design-in limit.

Perhaps Xilinx could take some people/fab time off the G-gate devices to
actually get some useful parts out the door.


Article: 26854
Subject: JBits
From: Ivan Leung <khleung@cse.cuhk.edu.hk>
Date: 1 Nov 2000 17:23:33 GMT
Links: << >>  << T >>  << A >>

Hi all,

	I would like to ask anyone knows anything about JBits?

-- 
Cheers,
Ivan

~~~~~~~~~~~~~~~~
Ivan Leung 

Article: 26855
Subject: Re: JBits
From: Phil James-Roxby <phil.james-roxby@xilinx.com>
Date: Wed, 01 Nov 2000 11:02:43 -0700
Links: << >>  << T >>  << A >>
Ivan Leung wrote:
> 
> Hi all,
> 
>         I would like to ask anyone knows anything about JBits?

Yup,
What do you want to know?
Phil

-- 
---------------------------------------------------------------------
 __
/ /\/  Dr Phil James-Roxby         Direct Dial: 303-544-5545
\ \    Staff Software Engineer     Fax: Unreliable use email :-)
/ /    Loki/DARPA                  Email: phil.james-roxby@xilinx.com
\_\/\  Xilinx Boulder                 
---------------------------------------------------------------------

Article: 26856
Subject: Re: Alliance under Linux?
From: Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de>
Date: 1 Nov 2000 18:15:54 GMT
Links: << >>  << T >>  << A >>
Duane <junkmail@junkmail.com> wrote:

...
: ngdanno, ngd2vhdl, etc... The "Design manager" does not work under wine,
: as for as I know. So I have a somewhat elaborate but very easy (I think)
: to use script that runs through the process, creating directories,

I didn't notice you bugging the Wine developpers on
comp.os.ms-windows.emulators  with good bug reports ...
: ...

: I actually purchased VMware, since I still need it to run the synthesis
: tools. But that is all I use it for. All my other tools are either
: native Linux or run under wine.

For purely running Windows application on Linux, I guess one should have a
look at Win4Lin. It requires much less resources than VMWare and runs
faster. The new beta should also be able to access the parallel port.

Bye
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Article: 26857
Subject: Re: JBits
From: nweaver@soda.CSUA.Berkeley.EDU (Nicholas Weaver)
Date: 1 Nov 2000 18:16:24 GMT
Links: << >>  << T >>  << A >>
In article <8tpjil$hkf$1@eng-ser1.erg.cuhk.edu.hk>,
Ivan Leung  <khleung@cse.cuhk.edu.hk> wrote:
>Hi all,
>
>	I would like to ask anyone knows anything about JBits?

	What is there to know?  It is an abstract API, written in
java, designed to directly manipulate the Virtex bitfile.
-- 
Nicholas C. Weaver                                 nweaver@cs.berkeley.edu

Article: 26858
Subject: Re: Alliance under Linux?
From: Andy Peters <"apeters <"@> n o a o [.] e d u>
Date: Wed, 01 Nov 2000 11:31:47 -0700
Links: << >>  << T >>  << A >>
"S. Ramirez" wrote:
> 
> Andy,
>      That sucks!  I thought all those Unix systems were compatible and
> that's why Peter had no problem running his EDA tools under Linux.
> Apparently, I have been mislead by the Linux people!  Good thing I haven't
> switched -- yet.

Different CPUs!  My Sparcstation has (obviously) a Sparc processor;
HP/UX workstations have HP processors, etc etc.

I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can
get source code for things (and get it to compile, which never ever ever
works the first time).  If the source is not available, like say for
Star Office, I can't use it because it's not a x86 box.

I was rarely able to get Linux code to build on an x86 box.  Configure
scripts never work.  Whatta bunch of hooey: "Just modify the source
code!"  Gimme a break.

-- a
----------------------------
Andy Peters
Sr. Electrical Engineer
National Optical Astronomy Observatory
950 N Cherry Ave
Tucson, AZ 85719
apeters (at) n o a o [dot] e d u

"It is better to be silent and thought a fool, 
 than to send an e-mail to the entire company
 and remove all doubt."

Article: 26859
Subject: Re: Alliance under Linux?
From: Duane <junkmail@junkmail.com>
Date: Wed, 01 Nov 2000 12:12:58 -0800
Links: << >>  << T >>  << A >>
Andy Peters wrote:
> 
> "S. Ramirez" wrote:
> >
> > Andy,
> >      That sucks!  I thought all those Unix systems were compatible and
> > that's why Peter had no problem running his EDA tools under Linux.
> > Apparently, I have been mislead by the Linux people!  Good thing I haven't
> > switched -- yet.

Ohh, go ahead and switch! Just expect, contrary to some of the hype, to
spend a fair amount of time (at least initially) messing with things to
get everything to work the way you want. Some Unix sysadmin experience
helps out a whole lot here, but is not essential. I started using Linux
back in April, and have been extremely happy with it as a CAE platform.
It is very nice to be able to use a really high powered platform,
configured exactly the way I want it, at a remarkably low price.

> 
> Different CPUs!  My Sparcstation has (obviously) a Sparc processor;
> HP/UX workstations have HP processors, etc etc.
> 
> I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can
> get source code for things (and get it to compile, which never ever ever
> works the first time).  If the source is not available, like say for
> Star Office, I can't use it because it's not a x86 box.
> 
> I was rarely able to get Linux code to build on an x86 box.  Configure
> scripts never work.  Whatta bunch of hooey: "Just modify the source
> code!"  Gimme a break.

Yep, that's for sure. I much preferred Imake, which had the potential if
correctly configured of truly being portable. But seriously broken
vendor implementations really caused problems here, and it appears its
days are waning.

--
My real email is akamail.com@dclark (or something like that).

Article: 26860
Subject: Re: Alliance under Linux?
From: Duane <junkmail@junkmail.com>
Date: Wed, 01 Nov 2000 12:44:36 -0800
Links: << >>  << T >>  << A >>
Uwe Bonnes wrote:
> 
> Duane <junkmail@junkmail.com> wrote:
> 
> ...
> : ngdanno, ngd2vhdl, etc... The "Design manager" does not work under wine,
> : as for as I know. So I have a somewhat elaborate but very easy (I think)
> : to use script that runs through the process, creating directories,
> 
> I didn't notice you bugging the Wine developpers on
> comp.os.ms-windows.emulators  with good bug reports ...

The design manager and flow engine sure do look nice, at least they did
a couple of years when I last used them. And they were definitely
helpful when I was first learning the tools. But I suspect a significant
percentage of experienced designers abandon them eventually.

I see that I may have given the idea that I wasn't using them because I
couldn't get them to work with wine, but the real reason is because I
prefer the control I get from the script method.

--
My real email is akamail.com@dclark (or something like that).

Article: 26861
Subject: Re: help on a simple ALU
From: Colin Marquardt <colin.marquardt@usa.alcatel.com>
Date: 01 Nov 2000 13:32:30 -0800
Links: << >>  << T >>  << A >>
"news" <apteuk@vsnl.net> writes:

> > type mismatches.  It seems like some operations (add, subtract) what the
> > inputs to be integers while other operations (and, or) what them to be
> > bit_vectors or std_logic_vectors.  How can I do this?  I have tried

> u can try to include:
> library ieee;
> use ieee.std_logic_arith.all;
> use ieee.std_logic_unsigned.all;

No, don't do that. The correct conversion functions (unless your
tools do not accept them) are in ieee.numeric_std. See e.g.

  http://tech-www.informatik.uni-hamburg.de/vhdl/doc/faq/FAQ1.html#4.10
  http://tech-www.informatik.uni-hamburg.de/vhdl/doc/faq/FAQ1.html#integer_bit_vector

Colin

Article: 26862
Subject: Hardware Engineer position in Pittsburgh
From: autumn@osearch.com
Date: Wed, 01 Nov 2000 22:08:49 GMT
Links: << >>  << T >>  << A >>
I represent a mid-sized firm in Pittsburgh, PA who design, manufacture
and market proprietary electronic equipment-Seeking Senior Hardware
Engineers to be responsible for managing the digital hardware team-
Requires a BSEE with more than 5 years experience in analog and digital
circuit design. Experience with Xilinx FPGAs and C and Assembly
programming also required.
Contact Autumn Feldmeier 724)387-2668
or autumn@osearch.com



Sent via Deja.com http://www.deja.com/
Before you buy.

Article: 26863
Subject: Re: JBits
From: "Steven Guccione" <Steven.Guccione@xilinx.com>
Date: Wed, 1 Nov 2000 16:29:00 -0600
Links: << >>  << T >>  << A >>
Have a look at:  http://www.xilinx.com/products/software/jbits/index.htm for
more information on JBits.

-- Steve
-- 11/1/00

"Ivan Leung" <khleung@cse.cuhk.edu.hk> wrote in message
news:8tpjil$hkf$1@eng-ser1.erg.cuhk.edu.hk...
>
> Hi all,
>
> I would like to ask anyone knows anything about JBits?
>
> --
> Cheers,
> Ivan
>
> ~~~~~~~~~~~~~~~~
> Ivan Leung



Article: 26864
Subject: Re: I need some VHDL/Synthesis Design BOOK recommendations!!
From: Manfredo <manfredob@tin.it>
Date: Wed, 01 Nov 2000 23:30:46 +0100
Links: << >>  << T >>  << A >>
On Sun, 29 Oct 2000 21:30:45 GMT, JoeG <jNOgalleSPAM@pacbell.net>
wrote:

>I already have:
>
>1) Ashenden's "Designers guide to VHDL"
>2) Stefan Sjoholm & Lindh's "VHDL for designers"
>3) Douglas Smith's "HDL Chip Design"
>4) Kurup & Abbasi's "Logic synthesis using Synopsys"

try Lee
VHDL coding and logic syntesis with synopsis
ISBN 0-12-440651-3

Article: 26865
Subject: Re: death of rloc ?
From: Satnam Singh <Satnam.Singh@xilinx.com>
Date: Wed, 01 Nov 2000 14:34:34 -0800
Links: << >>  << T >>  << A >>
This is a multi-part message in MIME format.
--------------95A2AFBED30F21EE0EE4BFDD
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I apologise for the size of this paper. Until now no-one has complained.
The paper is about to be removed from the websitre so download time will
no longer be an issue (otherwise I would replace it with a more compact
version).

Kind regards,

Satnam Singh

--------------95A2AFBED30F21EE0EE4BFDD
Content-Type: text/x-vcard; charset=us-ascii;
 name="Satnam.Singh.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Satnam Singh
Content-Disposition: attachment;
 filename="Satnam.Singh.vcf"

begin:vcard 
n:Singh;Satnam
tel;pager:4087182588@messaging.cellone-sf.com
tel;cell:(408) 718 2588
tel;fax:(208) 293-9440
tel;home:(408) 377 9982
tel;work:(408) 879 4693
x-mozilla-html:TRUE
url:http://www.xilinx.com/labs/satnam
org:<br><img src="http://www.xilinx.com/images/xlogoc.gif" alt="Xilinx">;Xilinx Labs
adr:;;2100 Logic Drive;San Jose;CA;95124-3450;USA
version:2.1
email;internet:Satnam.Singh@xilinx.com
title:Researcher
note:UK Cellular +44-7979-648412
x-mozilla-cpt:;-1856
fn:Satnam Singh
end:vcard

--------------95A2AFBED30F21EE0EE4BFDD--

Article: 26866
Subject: Re: death of rloc ?
From: Satnam Singh <Satnam.Singh@xilinx.com>
Date: Wed, 01 Nov 2000 14:40:14 -0800
Links: << >>  << T >>  << A >>
This is a multi-part message in MIME format.
--------------1BED79354097BF48FAFCFF9D
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I don't know if you've read the paper, but it is about why we NEED
RLOCs.
That is why there is a question mark in the title.

There is still a huge advantage to be gained by floor-planning your
design if you happen to have a good enough intuition of the underlying
fabric. I give two examples where my flooor-plan beat the automatic
tools and where where I failed -- but that was because my floor-plan was
not good enough -- had I put in more effort I am sure I could come up
with a floor-plan that beat the automatic tools (in this case for
butterfly networks).

I speculate that in the future we may not need them if the software gets
clever enough, but that is certainly not the case for Xilinx FPGAs at
the moment or the immediate future.

Kind regards,

Satnam Singh

husby@my-deja.com wrote:

> Of course an easier experiment is to remove the rlocs
> from your existing designs and find that all of
> them perform much worse or can't be routed at all.
>
> The same goes for mapping directives.
>
> Muzaffer Kal <muzaffer@dspia.com> wrote:
> > I think this paper would be helpful to some people here and
> > interesting for most. Interesting read. Available at :
> > http://www.xilinx.com/labs/satnam/death_of_the_rloc.pdf
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.

--------------1BED79354097BF48FAFCFF9D
Content-Type: text/x-vcard; charset=us-ascii;
 name="Satnam.Singh.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Satnam Singh
Content-Disposition: attachment;
 filename="Satnam.Singh.vcf"

begin:vcard 
n:Singh;Satnam
tel;pager:4087182588@messaging.cellone-sf.com
tel;cell:(408) 718 2588
tel;fax:(208) 293-9440
tel;home:(408) 377 9982
tel;work:(408) 879 4693
x-mozilla-html:TRUE
url:http://www.xilinx.com/labs/satnam
org:<br><img src="http://www.xilinx.com/images/xlogoc.gif" alt="Xilinx">;Xilinx Labs
adr:;;2100 Logic Drive;San Jose;CA;95124-3450;USA
version:2.1
email;internet:Satnam.Singh@xilinx.com
title:Researcher
note:UK Cellular +44-7979-648412
x-mozilla-cpt:;-1856
fn:Satnam Singh
end:vcard

--------------1BED79354097BF48FAFCFF9D--

Article: 26867
Subject: Re: death of rloc ?
From: Muzaffer Kal <muzaffer@dspia.com>
Date: 01 Nov 2000 23:05:19 GMT
Links: << >>  << T >>  << A >>
Satnam Singh <Satnam.Singh@xilinx.com> wrote:

>I apologise for the size of this paper. Until now no-one has complained.
>The paper is about to be removed from the websitre so download time will
>no longer be an issue (otherwise I would replace it with a more compact
>version).

just curious, why are you removing it ?

Muzaffer
http://www.dspia.com

Article: 26868
Subject: Re: Alliance under Linux?
From: "S. Ramirez" <sramirez@deleet.cfl.rr.com>
Date: Wed, 01 Nov 2000 23:15:20 GMT
Links: << >>  << T >>  << A >>
Steve,
     Correct you are!  What I really meant to say is "Why doesn't Xilinx (or
other FPGA vendor) make a binary for PCs running Linux?"  I thought this was
the next big tsunami.
-Simon Ramirez, Consultant
 Synchronous Design, Inc.


> >      That sucks!  I thought all those Unix systems were compatible
>
> "Compatible" in Unix-land has never really meant binary image compatible.
> Frequently, it means "you can recompile the source with only a modest
> amount of aggro (provided you know what you're doing)" :-)
>
> --
> Steve Rencontre http://www.rsn-tech.co.uk
> //#include <disclaimer.h>
>
>


Article: 26869
Subject: Re: Hardware Engineer position in Pittsburgh
From: "S. Ramirez" <sramirez@deleet.cfl.rr.com>
Date: Wed, 01 Nov 2000 23:22:50 GMT
Links: << >>  << T >>  << A >>
     Now this is what I call a good, honest ad.
-Simon Ramirez, Consultant
 Synchronous Design, Inc.
<autumn@osearch.com> wrote in message news:8tq49a$mnu$1@nnrp1.deja.com...
> I represent a mid-sized firm in Pittsburgh, PA who design, manufacture
> and market proprietary electronic equipment-Seeking Senior Hardware
> Engineers to be responsible for managing the digital hardware team-
> Requires a BSEE with more than 5 years experience in analog and digital
> circuit design. Experience with Xilinx FPGAs and C and Assembly
> programming also required.
> Contact Autumn Feldmeier 724)387-2668
> or autumn@osearch.com
>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>



Article: 26870
Subject: Re: Alliance under Linux?
From: Zoltan Kocsi <root@127.0.0.1>
Date: 02 Nov 2000 10:59:20 +1100
Links: << >>  << T >>  << A >>
Andy Peters <"apeters <"@> n o a o [.] e d u> writes:

> works the first time).  If the source is not available, like say for
> Star Office, I can't use it because it's not a x86 box.

You can find this on Sun's website:

  Sun is making the source code for its StarOffice[tm] software suite 
  freely available under the GNU General Public License (GPL). 
  OpenOffice.org and the OpenOffice.org Foundation will host the CVS 
  tree (source code) and maintain the reference implementation of the 
  software, APIs and file formats. Development will take place in the 
  open, with all contributors working from the same source. According 
  to the principles of free software and the GPL, all improvements to 
  the software are contributed back to the project and are subsequently 
  available to all. 

For further details take a look at

  http://www.sun.com/developers/openoffice/

You can download OpenOffice, the implied future tense in the 
announcement is actually past tense by now.

> I was rarely able to get Linux code to build on an x86 box.  Configure
> scripts never work.  

Umm, I regularly build all sorts of stuff from source on various
x86 boxes, running different versions of Linux from different
distros and most of the time configure scripts work and everything
compiles just as it should. I have to admit, though, that I read the
README's and INSTALL's.

> Whatta bunch of hooey: "Just modify the source code!"  
> Gimme a break.

Well, that's the whole idea of free software, you *can* modify the 
source if you want to use it for something that it was not written 
for. You can, for example, port StarOffice to a PPC machine because
you can have the source. You don't have to, however. If you don't
like the idea, then just pretend you don't have the source and if 
it isn't available for or does not run on the platform of your 
choice, then tough, just like with closed source stuff.

Zoltan

-- 
+------------------------------------------------------------------+
| ** To reach me write to zoltan in the domain of bendor com au ** |
+--------------------------------+---------------------------------+
| Zoltan Kocsi                   |   I don't believe in miracles   |  
| Bendor Research Pty. Ltd.      |   but I rely on them.           |
+--------------------------------+---------------------------------+

Article: 26871
Subject: Re: Alliance under Linux?
From: Zoltan Kocsi <root@127.0.0.1>
Date: 02 Nov 2000 11:47:46 +1100
Links: << >>  << T >>  << A >>
"S. Ramirez" <sramirez@deleet.cfl.rr.com> writes:

> Andy,
>      That sucks!  I thought all those Unix systems were compatible and
> that's why Peter had no problem running his EDA tools under Linux.

It is a little bit more complicated. You see Windows runs on x86 only,
(although NT used to run on Alpha and PPC, but they were dropped) and thus
you tend to forget that different processors have different architectures
and instruction sets. In the unix world, however, we have to face the 
problem that any given bitstream might encode completely different 
instructions for different processors (if it means anything at all for 
a given processor).

Regardless of the operating system, which from a processor's point of 
view is just code as anything else, you can't exchange binaries between
processors. You can emulate a processor on an other but that is a rather
slow process.

Unix compatibility means that the system calls (i.e. the system/application 
program boundary) are more or less the same between different versions of
unix, at least on the source level. 

> Apparently, I have been mislead by the Linux people!  Good thing I haven't
> switched -- yet.

They did not necessary mislead you. Linux is indeed capable of running x86 
Solaris binaries through the iBCS library. However, when a vendor sells
you a Solaris tool it is almost surely for Sparc processor and would
not run even on a native Solaris environment if it was on x86 iron.

It is like if Windows was available for the PowerMac and you tried to
run your x86 Windows application there, it would not run for the PPC
can't understand x86 instructions, regardless whether the OS is MacOS
or Windows.

You should not, however, brush the compatibility claim of free software
away before you try to appreciate what's behind it.

Linux supports (runs on) more processors than any other OS except *BSD.
Windows runs on 1, namely the x86. You can face severe compatibility 
issues between Windows versions. Interestingly enough a lot of free
code runs on all sorts of CPUs each of them running all sorts of
unix variants *and* the free code often runs on Windows as well.
In case of devtools, like gcc, it can be even more mind boggling.
When you create a crosscompiler you can compile gcc on some 
(processor-A,opsys-A) platform to run on (processor-B,opsys-B) and 
generate code for (processor-C,opsys-C). This is not for the faint 
in heart, but you can do it.

As per if it is good or bad that you have not switched, it is
only up to you. If Windows allows you to do your job, you feel
comfortable with it and you are not in the mood of learning a 
new system (and paradigm), then don't switch. Using Linux is not 
compulsory. In fact, because of the lack of support from tool
vendors, AFAIK it is impossible to design an FPGA from scratch
to bitstream using Linux only (sans running a virtual Windows
machine within an x86 Linux). There is increasing demand for 
Linux tools, which the vendors seem to respond (albeit very 
sluggishly) but currently Windows is the way to go.
If Windows can satisfy your needs, then you should not switch 
just because Linux is a buzzword and a cool thing to use. 

Zoltan

-- 
+------------------------------------------------------------------+
| ** To reach me write to zoltan in the domain of bendor com au ** |
+--------------------------------+---------------------------------+
| Zoltan Kocsi                   |   I don't believe in miracles   |  
| Bendor Research Pty. Ltd.      |   but I rely on them.           |
+--------------------------------+---------------------------------+

Article: 26872
Subject: Bits swapping with XC18V02
From: "Jean-Paul GOGLIO" <goglio@getris.com>
Date: Thu, 2 Nov 2000 09:40:30 +0100
Links: << >>  << T >>  << A >>
Hi all

I try to program a Virtex with a 18V02 ISP Prom using SelectMap mode.
But when I made the PCB, I missed something obvious and i connected the pin
D0 of the PROM to the pin D0 of the FPGA and the pin D7 to the pin D7.
Xilinx says that the good connexion is D0 => D7 and D7 => D0.  ;-)))

Now, how can i swap these bits ?
The first solution is to swap them on the board, but it would be more simple
to swap them in the bitstream.
There is a "swap pins" option in the Prom File Formatter, but only for Hex
format, not for Mcs or Exor, and only the 2 last formats are accepted by the
Jtag Programmer.

Does anyone have a solution ?

Thanks.

--
J-P GOGLIO
GETRIS S.A.
13 Chemin des Prés
38240 Meylan
Tel : (+33) 4 76 18 52 10
E-mail : goglio@getris.com
Fax : (+33) 4 76 18 52 01



Article: 26873
Subject: Re: Spartan II ?
From: eml@riverside-machines.com.NOSPAM
Date: Thu, 02 Nov 2000 10:07:55 GMT
Links: << >>  << T >>  << A >>
On Wed, 01 Nov 2000 17:22:13 +0000, Rick Filipkiewicz
<rick@algor.co.uk> wrote:

>
>
>Paul Smith wrote:
>
>> Leon Heller wrote:
>>
>> >
>> > I've just been quoted 14-18 weeks lead-time for some XC2S50-5TQ144C by
>> > our distributor, Insight Memec, in the UK
>> >
>>
>> I was just quoted 16 weeks for XC2S30-5VQ100C
>>
>> Anyone able to buy XC2S30 or XC2S50 with reasonable leadtimes?
>
>Maybe somebody should remind Xilinx and/or the disti's of the concept of
>time-to-market.
>Around here 16 weeks is well beyond the no design-in limit.

If you've got a real design, get the FAE in and ask about
availability. If you're nice to them, they may offer to supply an
equivalent Virtex (make sure there is one first) at the S-II price.
Worked for me.

Evan

Article: 26874
Subject: Re: Alliance under Linux?
From: eml@riverside-machines.com.NOSPAM
Date: Thu, 02 Nov 2000 10:08:44 GMT
Links: << >>  << T >>  << A >>
On Wed, 01 Nov 2000 11:31:47 -0700, Andy Peters <"apeters <"@> n o a o
[.] e d u> wrote:

>I have a G3 Mac running Yellowdog Linux, which is mostly fine if I can
>get source code for things (and get it to compile, which never ever ever
>works the first time).  If the source is not available, like say for
>Star Office, I can't use it because it's not a x86 box.
>
>I was rarely able to get Linux code to build on an x86 box.  Configure
>scripts never work.  Whatta bunch of hooey: "Just modify the source
>code!"  Gimme a break.

Exactly. And how many of us actually care about whether or not we have
source code? As far as I'm concerned, it's just wasting valuable disk
space. Linux developers seem to have a religious objection to
distributing binaries, so we have to put up with second-rate
installation programs and binaries which aren't necessarily the same
as anyone elses.

And what about actually installing Linux itself in the first place? I
spent a day trying to get the latest Debian onto my new Dell portable
recently. I went through the whole process at least six times, and
couldn't get any further than a simple console interface; Debian
doesn't know about AGP. I eventually found a sysadmin who managed to
install X, but she couldn't get networking going. I remember
installing Windows 3.1 from a box of floppies, and it was easier than
this.

The other bad news is that commerical Linux EDA software appears to be
at Unix prices, not Windows prices, which is going to make it a
non-starter.

Evan



Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarApr2017

Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Custom Search