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
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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 141925

Article: 141925
Subject: Re: How to implementa an FSM in block ram
From: glen herrmannsfeldt <gah@ugcs.caltech.edu>
Date: Fri, 17 Jul 2009 09:26:19 +0000 (UTC)
Links: << >>  << T >>  << A >>
Hal Murray <hal-usenet@ip-64-139-1-69.sjc.megapath.net> wrote:
(snip, I wrote)

>>I suppose it makes some sense.  The address decoders go to
>>enable lines that enable the signal coming out of the bit
>>cells onto the sense lines.  If, for example, two enable lines
>>were active at once then two cells would be enabled onto the
>>same sense line.  If the address goes metastable, it seems
>>that could happen.  Presumably this is true for non-FPGA RAMs, too.
 
> Thanks, that's the hint I was fishing for.  It's an unbuffered
> "output".  At that level, the write isn't gated.  The write
> driver for the sense line just overpowers the driver from the bit
> and smashes the bit to the new state.  If two read lines
> are selected they may fight and one may "write" the other.

For an asynchronous static RAM, if one changes the address
inputs, there might be a short time that two decoder outputs
are enabled.  The designer should guarantee that the bits
won't change in that time.  If the address inputs float or
oscillate at a very high frequency, it seems possible that
the bits might not survive.  That should be in an SRAM data sheet.
 
> That doesn't require any metastability.  Just setup/hold can
> explain it all.  

Without metastability, the register outputs should switch
at a known speed, and the decoder should be tested for that
speed.  

-- glen
 

Article: 141926
Subject: Re: FPGA to PC connection
From: glen herrmannsfeldt <gah@ugcs.caltech.edu>
Date: Fri, 17 Jul 2009 09:28:10 +0000 (UTC)
Links: << >>  << T >>  << A >>
Gints <reganireland@gmail.com> wrote:
 
> I have an FPGA capturing binarizing an image and want to connect it to
> a PC for further analysis etc. What do you think would be the best
> protocol to use? The data itself is about 12.8 Mbits/second so was
> leaning toward 100Mbps ethernet implemented on the FPGA. Is this
> possible (plausible) without a soft processor, ie: just streaming the
> images straight from the FPGA RAM? Any links to relevant readings
> would be greatly appreciated so I could learn.

Depending on your definition of 'soft processor.'

You will need a state machine of some kind, which may or 
may not be considered a processor.

-- glen

Article: 141927
Subject: log likelihood ratio
From: "shereen.ahmed" <shereen.ahmed@gmail.com>
Date: Fri, 17 Jul 2009 04:32:26 -0700 (PDT)
Links: << >>  << T >>  << A >>



Dear All
I need your help

how can i calculate log likelihood  ratio (AWGN noise ) for belief
propagation  decoder?


Thanks for you

Article: 141928
Subject: Xilinx Platinum Support - I found it! This is me :)
From: Antti <Antti.Lukats@googlemail.com>
Date: Fri, 17 Jul 2009 06:12:27 -0700 (PDT)
Links: << >>  << T >>  << A >>
smiling as I write, but maybe I should not.

a friend contacted me with an simple question:
how to make a counter with xilinx coregen for Spartan-3A family.

it suprised me.
I checked coregen, and well it did not support S3A

I did go www.xilinx.com
spent 8 seconds
found AR about this issue
send email with the AR number to my friend

case solved.

just received response from my friend:
he had contacted Xilinx support AND opened webcase as well

but Xilinx support and WebCase personel did NOT find this Xilinx AR

so, it seems that if there is such thing as "Xilinx Platinium
support"
then this is me ?

but it doesnt make me happy, as in the cases where I need support
there will be no-one to help. another thing is that Xilinx support
people
do get paid, I usually do not (not for finding Xilinx AR numbers when
Xilinx support fails with search)


hm.. but maybe Xilinx own support people do not get paid?
maybe that's why they do not provide support?


Antti
PS I'm not that arrogant [as i may sound]
well my wifes sure says its arrogance squared...


Article: 141929
Subject: Re: Generating a negated clock
From: Andy <jonesandy@comcast.net>
Date: Fri, 17 Jul 2009 07:04:01 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Jul 16, 11:19=A0pm, hal-use...@ip-64-139-1-69.sjc.megapath.net (Hal
Murray) wrote:
> >The point is that deep enough is often not a power of two and you
> >may not have the block RAM to waste on the unused portion. =A0The
> >control circuitry would obviously be the same size since you
> >haven't reduced the number of address bits.
>
> I'd expect the control circuitry to be slightly larger for the
> non-power-of-two case. =A0You need to wrap the address counters
> at 3/4 or 7/8 rather than just letting them roll over.
>
> Things may get more complicated if you want signals like
> almost-full, almost-empty, and half-full.
>
> --
> These are my opinions, not necessarily my employer's. =A0I hate spam.

If you use pointers that are down counters, not up counters, the logic
difference may not be much. At least synplify optimizes away the roll
over bit logic that is the same result as a normal roll over
(decrement from zero). In other words only the bits that do not become
one will need additional logic to make the non power of two rollover
work. As long as the fifo size is 2^n - 2^m, there is only one bit
that does not rollover with a normal decrement from zero.

Come to thinkg of it, you could also do the same thing with up
counters, starting at 2^m.  When the counter rolls over at 2^n, just
stuff the 2^m bit. This might get more tricky with keeping track of
AF, AE, HF, etc.

Andy

Article: 141930
Subject: Re: FPGA to PC connection
From: doug <xx@xx.com>
Date: Fri, 17 Jul 2009 08:56:36 -0800
Links: << >>  << T >>  << A >>


Gints wrote:

> hey guys,
> 
> I have an FPGA capturing binarizing an image and want to connect it to
> a PC for further analysis etc. What do you think would be the best
> protocol to use? The data itself is about 12.8 Mbits/second so was
> leaning toward 100Mbps ethernet implemented on the FPGA. Is this
> possible (plausible) without a soft processor, ie: just streaming the
> images straight from the FPGA RAM? Any links to relevant readings
> would be greatly appreciated so I could learn.
> 
> G

This is certainly reasonable if you use UDP. If you only implement
ARP and UDP, the interface with a PHY chip is straighforward.
There is an example of a 10Mbps aat www.fpga4fun.com. I did a
100Mbps interface without a processor for a data collection
system.

Article: 141931
Subject: Re: Generating a negated clock
From: "MM" <mbmsv@yahoo.com>
Date: Fri, 17 Jul 2009 13:38:56 -0400
Links: << >>  << T >>  << A >>
"Nemesis" <nemesis@nowhere.invalid> wrote in message 
news:20090716184149.3460.49994.XPN@orion.invalid...
>
> I interfaced my DSP code with base code of the board (an ADC board).
> I don't want to modify the original code. I think they didn't use the
> Xilinx core because the fifo's depth is about 12000, Xilinx core accept
> only power of 2 depths if I remember well.

I am pretty sure it would be easier and an overall better solution to 
replace a weird FIFO than to create a weird clocking scheme. The reason they 
didn't use a Xilinx core could be because they wanted the code to be 
portable between different FPGA vendors (they might have ADC boards with 
different FPGAs available), which is still a bad excuse for requiring 2 
phases of the same clock.

In any case I doubt there's much value in their ADC interface code. You 
could probably interface to the ADC directly yourself.


/Mikhail




Article: 141932
Subject: Re: Do you prefer paper or plastic... er, I mean paper or e-books?
From: AndreasWallner <Andreas.Wallner.etm07@fh-joanneum.at>
Date: Fri, 17 Jul 2009 11:06:05 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Jul 17, 1:24=A0am, rickman <gnu...@gmail.com> wrote:
> I find there are any number of aspects of the VHDL language that I
> just do not remember and I am not going to make up flash cards to help
> me remember. =A0So I drag a half dozen VHDL books around with me when I
> am working on VHDL (or much less frequently, Verilog; one of the books
> covers both).
>
> I am getting tired of heaving the books up into the truck every time I
> go to the lake and am starting to wonder if I should invest in some
> good e-books on HDL.
>
> What do the rest of you prefer? =A0Do you have both? =A0Do you still need
> to rely on your books or do you pretty well have the language down
> pat?
>
> Is the IEEE VHDL standard available in e-book format? =A0Maybe I need to
> invest in the 200x version anyway?
>
> Rick

In my case it really depends.

I really like all three things...paper books and e-books

I think reading is much better on paper than on a PC monitor. (Here
the Kindle sounds interesting, but in Austria it is not available). So
I really prefer a book if I'm about to read a whole lot (e.g. book to
learn a new programming language). It also has the advantage that you
can have the book open, and read there while working on you code (I
have two monitors, and most of the time when I'm programming VHDL or C+
+ I use both monitors, so having a book is a real advantage)

But If I just have to refresh my memory about something (syntax, etc.)
I prefer PDFs because of their searchability. I have the few book I
use the most in both forms (paper and electronic) and choose depending
on mood/what I want to do which to use.

Andreas

Article: 141933
Subject: Re: Do you prefer paper or plastic... er, I mean paper or e-books?
From: Mike Treseler <mtreseler@gmail.com>
Date: Fri, 17 Jul 2009 11:22:31 -0700
Links: << >>  << T >>  << A >>
AndreasWallner wrote:

> But If I just have to refresh my memory about something (syntax, etc.)
> I prefer PDFs because of their searchability.

A syntax-aware editor that proposes completions
sometimes covers these problems.


    -- Mike Treseler

Article: 141934
Subject: Re: Do you prefer paper or plastic... er, I mean paper or e-books?
From: luudee <rudolf.usselmann@gmail.com>
Date: Fri, 17 Jul 2009 12:31:03 -0700 (PDT)
Links: << >>  << T >>  << A >>

In my case I use lots and lots of books that refer to various
standards, (like ATA, SATA, USB, etc). And for that I prefer
eBooks. 1) I can search them with ease; 2) I can create my
own bookmarks; 3) I can carry a bunch of them on a single
USB drive.

One important thing I learned the hard way, is to never buy
"locked" eBooks. Mindshare comes to mind as one of the worst
experiences (ended up buying a hard copy of a book because I
could not open the eBook freely when I needed it).

During development cycles, I tend to copy and paste important
paragraphs from those eBooks, creating my own reference manuals
for specific subjects (e.g. everything about control endpoint
for USB - that information is spread over 1000+ pages of USB
specification).

Best Regards,
rudi

Article: 141935
Subject: Re: How to implementa an FSM in block ram
From: Peter Alfke <peter@xilinx.com>
Date: Fri, 17 Jul 2009 14:08:11 -0700 (PDT)
Links: << >>  << T >>  << A >>

> > That doesn't require any metastability. =A0Just setup/hold can
> > explain it all. =A0

Not true:
Just violating the set-up time specification (without changing inputs
during the tiny metastability-inducing window) would just create
ambiguity in the decoding of the address, perhaps even a composite
mixed-up address. With WE being inactive, this will never contaminate
the data.

The problem we are discussing here occurs only when you change address
bit(s) during the femtosecond timing window that can cause the address
register to go metstable, thus generate duplicate decoder outputs, and
perhaps cause storage bits to kill each other.Might occur once a month
or once a decade.
The likelihood is extremely small, but not zero.
That's why we published the warning, and the very simple and effective
work-around.
Peter Alfke



Article: 141936
Subject: MPMC4.03 DDR1 question
From: "MM" <mbmsv@yahoo.com>
Date: Fri, 17 Jul 2009 17:48:31 -0400
Links: << >>  << T >>  << A >>
Does anyone know if I can run DDR1 type of memory at double the frequency of 
MPMC4.03 (EDK10.1) main clock? Or is this possible for DDR2 only? I am 
having difficult time trying to understand this from the datasheet. So, 
before digging into the MPMC source code I thought I would try asking...


Thanks,
/Mikhail



Article: 141937
Subject: Re: Problem with System ACE, can't get it to work with partitioned
From: AndreasWallner <Andreas.Wallner.etm07@fh-joanneum.at>
Date: Fri, 17 Jul 2009 18:50:26 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Jul 15, 1:04=A0pm, Fridolin <googlegro...@47110815.com> wrote:
> On Jul 14, 10:46=A0pm, AndreasWallner <Andreas.Wallner.et...@fh-
>
>
>
>
>
> joanneum.at> wrote:
> > Hi,
>
> > I have a Problem getting the System ACE Controller to work.
> > I try it on a Spartan 3A DSP 1800 Board, to which I connected a
> > SystemACE Module from AVNet.
>
> > If I format the CF Card completly, and do not create any partitions
> > (the drive then only contains the FAT16).
>
> > If I try to create two partitions (with linux fdisk, one 64MB
> > Partition, and the second spanning the remaining card), format the
> > first one to FAT16 and copy the exact same files to the card the
> > System ACE board only displays an error and does not load the design.
>
> > Are there any special points to take into consideration?
>
> > Regards,
> > Andreas
>
> Hi Andreas,
>
> look onhttp://www.xilinx.com/products/boards/ml403/reference_designs.htm
> near the end of the page there is a image for the CF card with a DOS
> and Linux partition.
> (CF Card Image (DOS and Linux Partition))
> You can download it and view it with a Disk Editor or Uedit in binary
> mode.
>
> Maybe you find the diffs to the Linux fdisk tool.
>
> Have Fun
> Florian

Hi,

I analysed the partition table of both, my self created and the one in
the image. The rest of the MBR also looked fine (as fine as a
bootloader which is not there can look ;-) )
After not finding a difference (except the size-dependent parts) I
copied the partition table from the image onto my CF card, formatted
the first partition, and tried that one out. It also did not work. So
I tried only creating one partition, which also did not work...

The commands I'm using:
mkdosfs -F 16 -R 1 -I /dev/sdc (for the whole disk)
mkdosfs -F 16 -R 1 /dev/sdc1 (for the partition)

If I create the partition table I use linux fdisk, and change the type
of the partition to 0x06.

I also tried axplicitly specifying different sector-per-cluster sizes,
which also did not change anything. (sector-per-cluster size /=3D 1)
-> mkdosfs -F 16 -R 1 -s 4

Using dd (which is btw. also available for windows) I wrote the image
Fridolin mentioned to the CF card. Then I deleted all files from the
card and copied my design onto it. This time it worked, it's a way to
start, but I really don't think it's a way to do it all the time...

Tomorrow I will try out wheter my VHDL implementation for the MPU bus
works or not, if it does I will tell you the current values of the
status/error/fatstat registers. But it would be really great if you
could tell me how you partition/format you CF card (linux would be
fine, but I would also be grateful for a windows way)

Thanks
Andreas

Article: 141938
Subject: Re: How to implementa an FSM in block ram
From: Allan Herriman <allanherriman@hotmail.com>
Date: 18 Jul 2009 03:42:32 GMT
Links: << >>  << T >>  << A >>
On Fri, 17 Jul 2009 14:08:11 -0700, Peter Alfke wrote:

>> > That doesn't require any metastability.  Just setup/hold can explain
>> > it all.
> 
> Not true:
> Just violating the set-up time specification (without changing inputs
> during the tiny metastability-inducing window) would just create
> ambiguity in the decoding of the address, perhaps even a composite
> mixed-up address. With WE being inactive, this will never contaminate
> the data.
> 
> The problem we are discussing here occurs only when you change address
> bit(s) during the femtosecond timing window that can cause the address
> register to go metstable, thus generate duplicate decoder outputs, and
> perhaps cause storage bits to kill each other.Might occur once a month
> or once a decade.
> The likelihood is extremely small, but not zero. That's why we published
> the warning, and the very simple and effective work-around.
> Peter Alfke

"Once a month or once a decade" was more like once every few times I 
toggled the reset input on the DCM, on the particular design I tested.  
The probability changed from part to part.  On some FPGAs, I couldn't 
make it happen at all.

I did have hundreds of BRAMs on that clock though, and a single bit error 
in any of those RAMs resulted in a test failure.

And yes, this was a fully synchronous design.  The glitches from the DCM 
were sufficient to cause the timing violation.

Regards,
Allan

Article: 141939
Subject: Re: How to implementa an FSM in block ram
From: Allan Herriman <allanherriman@hotmail.com>
Date: 18 Jul 2009 03:45:57 GMT
Links: << >>  << T >>  << A >>
On Sat, 18 Jul 2009 03:42:32 +0000, Allan Herriman wrote:

> On Fri, 17 Jul 2009 14:08:11 -0700, Peter Alfke wrote:
> 
>>> > That doesn't require any metastability.  Just setup/hold can explain
>>> > it all.
>> 
>> Not true:
>> Just violating the set-up time specification (without changing inputs
>> during the tiny metastability-inducing window) would just create
>> ambiguity in the decoding of the address, perhaps even a composite
>> mixed-up address. With WE being inactive, this will never contaminate
>> the data.
>> 
>> The problem we are discussing here occurs only when you change address
>> bit(s) during the femtosecond timing window that can cause the address
>> register to go metstable, thus generate duplicate decoder outputs, and
>> perhaps cause storage bits to kill each other.Might occur once a month
>> or once a decade.
>> The likelihood is extremely small, but not zero. That's why we
>> published the warning, and the very simple and effective work-around.
>> Peter Alfke
> 
> "Once a month or once a decade" was more like once every few times I
> toggled the reset input on the DCM, on the particular design I tested.
> The probability changed from part to part.  On some FPGAs, I couldn't
> make it happen at all.
> 
> I did have hundreds of BRAMs on that clock though, and a single bit
> error in any of those RAMs resulted in a test failure.
> 
> And yes, this was a fully synchronous design.  The glitches from the DCM
> were sufficient to cause the timing violation.
> 
> Regards,
> Allan

BTW, this design has since been retired.  I had a workaround which 
involved gating the clock for that product, and it worked ok.

Newer products haven't exhibited this type of problem at all, possibly 
because we know to disable the ram when there is any chance of a glitch 
on the clock line.

Regards,
Allan

Article: 141940
Subject: Re: Generating a negated clock
From: Nemesis <nemesis@nowhere.invalid>
Date: 18 Jul 2009 07:35:01 GMT
Links: << >>  << T >>  << A >>
MM wrote:

> "Nemesis" <nemesis@nowhere.invalid> wrote in message 
> news:20090716184149.3460.49994.XPN@orion.invalid...
>>
>> I interfaced my DSP code with base code of the board (an ADC board).
>> I don't want to modify the original code. I think they didn't use the
>> Xilinx core because the fifo's depth is about 12000, Xilinx core accept
>> only power of 2 depths if I remember well.
>
> I am pretty sure it would be easier and an overall better solution to 
> replace a weird FIFO than to create a weird clocking scheme. The reason they 
> didn't use a Xilinx core could be because they wanted the code to be 
> portable between different FPGA vendors (they might have ADC boards with 
> different FPGAs available), which is still a bad excuse for requiring 2 
> phases of the same clock.
>
> In any case I doubt there's much value in their ADC interface code. You 
> could probably interface to the ADC directly yourself.

At the end they use (smaller) Xilinx Fifo core to build the large one, I
suppose they didn't use directly a 16k fifo core not to waste BRAMs.

Anyway I don't want to play with their implementation because it contains
also interface code for getting data from the FIFO on the PCI bus or
another customized bus.

Unfortunately I don't have the time to study their solution and modify it
in a reliable way.
-- 
It is better to deserve honors and not have them than to have them and
not to deserve them.
 _  _                  _
| \| |___ _ __  ___ __(_)___
| .` / -_) '  \/ -_|_-< (_-<
|_|\_\___|_|_|_\___/__/_/__/ http://xpn.altervista.org


Article: 141941
Subject: Re: FPGA to PC connection
From: Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de>
Date: Sat, 18 Jul 2009 09:57:14 +0000 (UTC)
Links: << >>  << T >>  << A >>
doug <xx@xx.com> wrote:


> Gints wrote:

> > hey guys,
> > 
> > I have an FPGA capturing binarizing an image and want to connect it to
> > a PC for further analysis etc. What do you think would be the best
> > protocol to use? The data itself is about 12.8 Mbits/second so was
> > leaning toward 100Mbps ethernet implemented on the FPGA. Is this
> > possible (plausible) without a soft processor, ie: just streaming the
> > images straight from the FPGA RAM? Any links to relevant readings
> > would be greatly appreciated so I could learn.
> > 
> > G

> This is certainly reasonable if you use UDP. If you only implement
> ARP and UDP, the interface with a PHY chip is straighforward.
> There is an example of a 10Mbps aat www.fpga4fun.com. I did a
> 100Mbps interface without a processor for a data collection
> system.

Are you willing to share?
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

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

Article: 141942
Subject: Re: FPGA editor in Fedora 11 x86_64
From: anonymous anonymous <bitslipper@googlemail.com>
Date: Sat, 18 Jul 2009 03:04:05 -0700 (PDT)
Links: << >>  << T >>  << A >>
I run into the same problem. I'm using Fedora 11 with ISE 10.1.03.

Please write an email or post it here if you found a solution. Thanks.
bitslipper <at> googlemail.com

Article: 141943
Subject: Re: FPGA editor in Fedora 11 x86_64
From: General Schvantzkoph <schvantzkoph@yahoo.com>
Date: 18 Jul 2009 12:42:06 GMT
Links: << >>  << T >>  << A >>
On Wed, 15 Jul 2009 21:00:08 +0200, Jan Pech wrote:

> Is there any way how to make Xilinx FPGA Editor (and other old Wind/U
> based applications) running in Fedora 11, x86_64?
> 
> In RHEL, CentOS, and older versions of Fedora (all x86_64) it is
> sufficient to set the $DISPLAY environment variable to :0 and to follow
> this http://www.xilinx.com/support/answers/22022.htm answer record.
> Unfortunately this does not work in F11/x86_64.
> 
> I get the same error messages as described in the answer record but the
> answer itself does not help. According to "ldd -v ..../fpga_editor" I am
> not missing any required library.
> 
> Thanks for your help,
> Jan

The problem is that Fedora uses lesstif instead of motif. The solution is 
to install a CentOS5.3 VM and run fpga_ed from there. Fedora 11 has KVM 
built in, if you haven't already done it just install the virtualization 
tools and then run virt-manager and install CentOS. You should do an NFS 
export of your working directories on the host machine and then do an NFS 
mount in the CentOS VM, that will allow you to operate on the same 
directories in both the F11 host and the CentOS VM. I ssh into my CentOS 
VM from Xemacs and run fpga editor from the ssh shell so, from a UI 
standpoint, things that run on the VM are indistinguishable from things 
that run on the host. Make sure that the NFS directories appear at the 
same points in the file system on both the host and the client, that way 
Xemacs can still do auto completion of file names and directories when 
you are operating from the ssh shell.

KVM is the best VM option for F11. The performance of KVM is very very 
close to native performance with Linux VMs. I've found that Windows VMs 
are flaky on KVM, hopefully that will get fixed soon. VMware Server 
doesn't run on .29 kernels, there is a patch but it doesn't work very 
well. If you want to use VMware, which runs Windows VMs flawlessly and 
runs Linux VMs just as well as KVM, you will have to stick with Fedora 10 
until VMware adds support for .29 kernels. F10 uses the .27 kernel which 
is fully supported by VMware.

Article: 141944
Subject: Re: Suzaku SZx30 or similar
From: "AstroLad" <AstroLad@cox.net>
Date: Sat, 18 Jul 2009 08:25:09 -0500
Links: << >>  << T >>  << A >>
John,

Can you give me a few details? Will the new board be the same, or roughly
the the same size as the Craignell2? What about Ethernet? We want a 10/100
PHY (best), or MAC/PHY. The reason the PHY is best for us is that I already
developed a simplified MAC tuned to our processor core. It doesn't take
much space and not much support code.

Thanks,

Larry Dingle

>We have a product coming shortly based on our Craignell2 but a
>development board format that might offer an alternative. If you have
>a few weeks then wait and see if suits your application.
>
>John Adair
>Enterpoint Ltd.
>
>On Jul 6, 2:19=A0pm, "AstroLad" <Astro...@cox.net> wrote:
>> Does anyone know of anything similar to the Suzaku SZ030/SZ130? It's
just
>> about a perfect fit for a short production run product I'm helping a
frie=
>nd
>> with. Perfect that is except the price. What we need is an FPGA as good
o=
>r
>> better than a XC3S1000, 1MB or more of RAM (SRAM or SDRAM) and 100MB
>> Ethernet. It does not absolutely have to be a Spartan. An Altera
Cyclone =
>of
>> some flavor would do if the price were right. We already have a lot of
>> development done using Digilent Spartan boards. We don't need the
>> Microblaze as we have a CPU from OpenCores that is adequate.
>
>

Article: 141945
Subject: Re: FPGA editor in Fedora 11 x86_64
From: anonymous anonymous <bitslipper@googlemail.com>
Date: Sat, 18 Jul 2009 06:36:42 -0700 (PDT)
Links: << >>  << T >>  << A >>
Okay, FPGA Editor is starting now under Fedroa 11. First I had to
install the rpm package from www.motifzone.net. Then you have to
create a library link: "ln -s libXm.so.4 libXm.so.3". After that i had
to install libstdc++-4.4.0-4(x86_64). Finally there was a SELinux
issue:

SummarySELinux is preventing _fpga_editor from loading /opt/Xilinx/
10.1/ISE/lib/lin64/libkernel50.so which requires text relocation.
Detailed DescriptionThe _fpga_editor application attempted to load /
opt/Xilinx/10.1/ISE/lib/lin64/libkernel50.so which requires text
relocation. This is a potential security problem. Most libraries do
not need this permission. Libraries are sometimes coded incorrectly
and request this permission. The SELinux Memory Protection Tests web
page explains how to remove this requirement. You can configure
SELinux temporarily to allow /opt/Xilinx/10.1/ISE/lib/lin64/
libkernel50.so to use relocation as a workaround, until the library is
fixed. Please file a bug report against this package. Allowing
AccessIf you trust /opt/Xilinx/10.1/ISE/lib/lin64/libkernel50.so to
run correctly, you can change the file context to textrel_shlib_t.
"chcon -t textrel_shlib_t '/opt/Xilinx/10.1/ISE/lib/lin64/
libkernel50.so'" You must also change the default file context files
on the system in order to preserve them even on a full relabel.
"semanage fcontext -a -t textrel_shlib_t '/opt/Xilinx/10.1/ISE/lib/
lin64/libkernel50.so'"

After that FPGA Editor starts, but with several error messages:

[root@mobile lin64]# fpga_editor
/opt/Xilinx/10.1/ISE/bin/lin64/_fpga_editor: Symbol
`_XtperDisplayList' causes overflow in R_X86_64_PC32 relocation
/opt/Xilinx/10.1/ISE/bin/lin64/_fpga_editor: Symbol
`_XtGetPerDisplayInput' causes overflow in R_X86_64_PC32 relocation
Wind/U Error (248): Failed to connect to the registry on server mobile

Warning!!: XKEYSYMDB environment variable is set to a wrong location


Article: 141946
Subject: Re: Do you prefer paper or plastic... er, I mean paper or e-books?
From: Poojan Wagh <poojanwagh@gmail.com>
Date: Sat, 18 Jul 2009 07:51:13 -0700 (PDT)
Links: << >>  << T >>  << A >>
On Jul 17, 1:22=A0pm, Mike Treseler <mtrese...@gmail.com> wrote:
> AndreasWallner wrote:
> > But If I just have to refresh my memory about something (syntax, etc.)
> > I prefer PDFs because of their searchability.
>
> A syntax-aware editor that proposes completions
> sometimes covers these problems.
>
> =A0 =A0 -- Mike Treseler

Hi, Mike. Any suggestions for an editor that proposes completions?
(I'm using vim right now which does syntax highlighting.)

Article: 141947
Subject: Re: Do you prefer paper or plastic... er, I mean paper or e-books?
From: Mike Treseler <mtreseler@gmail.com>
Date: Sat, 18 Jul 2009 10:17:20 -0700
Links: << >>  << T >>  << A >>
Poojan Wagh wrote:

> Hi, Mike. Any suggestions for an editor that proposes completions?
> (I'm using vim right now which does syntax highlighting.)

I use vhdl-mode.
The feature is called template insertion (electrification)
For example, if I type "if " I am prompted for a boolean expression,
then for a THEN keyword, etc.
Press Enter any time to end the interaction.

Other key features are:
# Port translation (copy/paste)
# Design hierarchy browser (speedbar)
# Makefile generation
# Word/keyword completion

details here:
http://www.iis.ee.ethz.ch/~zimmi/emacs/vhdl-mode.html
http://www.iis.ee.ethz.ch/~zimmi/emacs/vhdl-mode.gif

  -- Mike Treseler

Article: 141948
Subject: Re: FPGA editor in Fedora 11 x86_64
From: glen herrmannsfeldt <gah@ugcs.caltech.edu>
Date: Sat, 18 Jul 2009 17:20:21 +0000 (UTC)
Links: << >>  << T >>  << A >>
anonymous anonymous <bitslipper@googlemail.com> wrote:
(snip)
 
< SummarySELinux is preventing _fpga_editor from loading /opt/Xilinx/
< 10.1/ISE/lib/lin64/libkernel50.so which requires text relocation.
< Detailed DescriptionThe _fpga_editor application attempted to load /
< opt/Xilinx/10.1/ISE/lib/lin64/libkernel50.so which requires text
< relocation. This is a potential security problem. Most libraries do
< not need this permission. Libraries are sometimes coded incorrectly
< and request this permission. The SELinux Memory Protection Tests web
< page explains how to remove this requirement. You can configure
< SELinux temporarily to allow /opt/Xilinx/10.1/ISE/lib/lin64/
< libkernel50.so to use relocation as a workaround, until the library is
< fixed. Please file a bug report against this package. Allowing
< AccessIf you trust /opt/Xilinx/10.1/ISE/lib/lin64/libkernel50.so to
< run correctly, you can change the file context to textrel_shlib_t.
< "chcon -t textrel_shlib_t '/opt/Xilinx/10.1/ISE/lib/lin64/
< libkernel50.so'" You must also change the default file context files
< on the system in order to preserve them even on a full relabel.
< "semanage fcontext -a -t textrel_shlib_t '/opt/Xilinx/10.1/ISE/lib/
< lin64/libkernel50.so'"

Yes, I found this previously and even sent a note to Xilinx
about it.  I haven't heard if there is interest in fixing it.

-- glen

Article: 141949
Subject: Re: FPGA editor in Fedora 11 x86_64
From: Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de>
Date: Sat, 18 Jul 2009 17:27:04 +0000 (UTC)
Links: << >>  << T >>  << A >>
glen herrmannsfeldt <gah@ugcs.caltech.edu> wrote:

> Yes, I found this previously and even sent a note to Xilinx
> about it.  I haven't heard if there is interest in fixing it.

I fear that these note normally get lost in the dungle of first, second and
higher level support. First level support dooesn't understand _anything_
from the technical things we report and perhaps "translates" the report into
some other form, dropping the technical findings. Higher level doesn't
understand that translation and drops probably the report or "fixed" it in
some strange way, but not the real cause.

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

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



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
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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