date:20120827

Re: casper-scm Commit c77818cc4abca3ab849913f7f940e2956b37b48f breaks ADC083000x2 block

2012-08-27 Thread Andrew Martens

Hi Glen

I have readded the blocks required by the adc083000x2 and pushed to the
ska-sa repo on github (https://github.com/ska-sa/mlib_devel.git) as
commit 2e8fc41. Please check that things are ok before I push to the
main casper-astro repo.

Regards
Andrew 

On Tue, 2012-08-21 at 16:20 -0400, G Jones wrote:
 (repost since I forgot which email address I was subscribed with)
 Hi,
 The cleanup commit c77818cc4abca3ab849913f7f940e2956b37b48f to
 casper-astro/mlib_devel breaks the mask script for ADC083000x2 which
 expects adc_sim to exist in the xps_library. It looks like the
 solution is to replace adc_sim with generic_adc. I'll try to generate
 a pull request or something to fix this, but if someone beats me too
 it I won't be sad.
 
 Glenn

Re: [casper] writing/reading qdr

2012-08-27 Thread Henno Kriel

Hi Matt

I had a look at your model file and suspect the issues is the include CPU
interface option is not enabled for your QDR block.
Enabling it will give you direct access to the QDR memory contents from the
PPC.

Regards
Henno

On Mon, Aug 27, 2012 at 9:09 AM, Matt Strader mstra...@physics.ucsb.eduwrote:

 I have a simple model file that writes a constant value to all of
 addresses in a QDR in the firmware and then reads it with katcp in
 python.  I thought I'd try this before putting QDR in more complicated
 designs.  When I try reading the QDR, I get all zeros.  I have
 attached my mdl and python script.  Could someone tell me what I'm
 missing?

 Thanks,
 Matt




-- 
Henno Kriel

DSP Engineer
Digital Back End
meerKAT

SKA South Africa
Third Floor
The Park
Park Road (off Alexandra Road)
Pinelands
7405
Western Cape
South Africa

Latitude: -33.94329 (South); Longitude: 18.48945 (East).

(p) +27 (0)21 506 7300
(p) +27 (0)21 506 7365 (direct)
(f) +27 (0)21 506 7375
(m) +27 (0)84 504 5050

Re: [casper] compiling fft block

2012-08-27 Thread Andrew Martens

Hi Matt

Thanks for the debugging work.

I have fixed the bug you found and committed the fix to the ska-sa git
repo (https://github.com/ska-sa/mlib_devel.git) as commit 03f2922.
Please check out the latest stuff and if that fixes your problem I will
push to the main casper-astro repo.

Regards
Andrew


On Tue, 2012-08-21 at 11:45 -0700, Matt Strader wrote:
 Hi Andrew,
 
 I managed to fix my problem.  Dragging in a new green fft block from
 the simulink casper library to replace my old fft block didn't work.
 The biplex_core inside this new block had the old mask parameters
 including specify_mult.  I replaced this biplex_core block with one
 from the casper library and found that had the up to date init
 parameters.  My model now compiles.
 
 Thanks,
 Matt
 
 On Sun, Aug 19, 2012 at 12:31 AM, Matt Strader
 mstra...@physics.ucsb.edu wrote:
  Hi Andrew,
 
  I tried compiling and running the init scripts on the matlab command
  line again, now with the trace output.  I have the results below.
  I've also attached my mdl file.  I expect there are other old blocks
  in it that need to be replaced, but I only replaced the PFB/fft blocks
  since they gave me trouble first.
 
   The only script I'm using is Startsg, which sets the environment
  variables to look at the right libraries.  I looked up the way to call
  the init script by first dragging in a new fft block from the simulink
  casper library.  Then I right-click and select 'Look under mask' a
  couple times until I find the biplex_core.  I right click this block,
  select 'Edit Mask', and select the initialization tab.  When I do this
  I see a specify_mult parameter in the list of init parameters for the
  biplex_core.
 
  I have commit e9ecb3f7fbe97039f16ec81949e2f282590e06e1 of the
  casper-astro repository.
 
  Thanks,
  Matt
 
 
  When compiling chan_550_packet.mdl with bee_xps:
 
  #
  ##  System Update  ##
  #
  trace: entering fft_init
  trace: entering fft_init
  trace: entering fft_biplex_init
  trace: entering fft_biplex_init
  trace: entering biplex_core_init
  trace: biplex_core_init post same_state
  trace: entering fft_stage_n_init
  trace: entering fft_stage_n_init
  trace: entering fft_stage_n_init
  trace: entering fft_stage_n_init
  trace: entering biplex_core_init
  trace: biplex_core_init post same_state
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  Running mask script for block: parallel_to_serial_converter
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  trace: entering munge_init
  Error using gen_xps_files (line 199)
  Error in 'chan_550_packet/PFB/fft1/fft_biplex0/biplex_core':
  Initialization commands cannot be evaluated.
 
 
  Now running init scripts from Matlab command line:
 
  biplex_core_init(gcb,...
  'FFTSize', 9,...
  'input_bit_width', 18,...
  'coeff_bit_width', 18,...
  'add_latency', 2,...
  'mult_latency', 2,...
  'bram_latency', 3,...
  'conv_latency', 2, ...
  'quantization', 'Truncate',...
  'overflow', 'Wrap',...
  'arch', 'Virtex5', ...
  'opt_target', 'logic', ...
  'coeffs_bit_limit', 8, ...
  'delays_bit_limit', 8, ...
  'specify_mult', 'off', ...
  'mult_spec', 0, ...
  'hardcode_shifts', 'off', ...
  'shift_schedule', 0, ...
  'dsp48_adders', 'on' );
  trace: entering biplex_core_init
  trace: biplex_core_init post same_state
  trace: entering biplex_core_init
  trace: biplex_core_init post same_state
  trace: entering biplex_core_init
  trace: biplex_core_init

Re: [casper] logging in to ROACH PPC through a tunnel

2012-08-27 Thread Tom Kuiper


On 08/27/2012 08:49 AM, Marc Welz wrote:

Apologies, my previous email contains errors, I didn't realise that you
were using two hops to get to the roach.
   

I was just replying to that when I got this :-)

However, it still seems reasonable to use the ssh port, when using separate
commands (all commands running concurrently): Maybe try :

MYHOST:~$ /usr/bin/ssh -p 22 -c 3des -t -Y -l kuiper -o
ServerAliveInterval=60 -o TCPKeepAlive=no -L 60001:ROACHHOST:22
GATEWAY

MYHOST:~$ ssh -p 60001 -c 3des -t -Y -l kuiper -o
ServerAliveInterval=60 -o TCPKeepAlive=no -L 60002:roach1:22 localhost

MYHOST:~$ ssh -p 60002 root@localhost
   
It works!  Eventually I'll have to try to combine it into one command 
that I can put in a script.  For now, it gets me back to work.


Hartelijk bedankt!
(That should be recognizable if you speak Afrikaans.)

You may have to enable -o NoHostAuthenticationForLocalhost=yes to
inhibit key checking, if the system warns you about host key changed
   

No.  I like the key checking.

I suspect there is some agent forwarding magic which may make some of
this easier.
   

If I figure it out I'll post it here.

Thanks again

Tom

Re: [casper] Roach 2 mechanical drawings?

2012-08-27 Thread Jason Ray


Vaughan,

Thanks for the drawing!  That's what we were looking for.

If you could make the cad model available, that would be great!

Thanks again,
Jason



At 09:21 AM 8/24/2012, Vaughan Moss wrote:

Hi

I have prepared this drawing, hope it helps. It is based on a CAD 
model. If you would like that it can be made available.


Regards

On 23 August 2012 17:30, Jason Manley 
mailto:jman...@ska.ac.zajman...@ska.ac.za wrote:
Vaughan Moss (cc'd here) has these drawings. 
https://github.com/ska-sa/roach2_chassis_designhttps://github.com/ska-sa/roach2_chassis_design


I can tell you that it only just fits in a 1U case, with about 1mm 
to spare above the DRAM DIMM after it's mounted on standoffs.


Jason

On 23 Aug 2012, at 16:37, Jason Ray wrote:

 All,

 Does anyone have a mechanical drawing of the roach 2 
board?  Mainly the board dimensions and mounting hole locations.


 I'll also need to know the approximate overall height of the 
board assembly with the tallest mezzanine board (CX4 or SFP+?) installed.


 Thanks in advance!
 Jason






--
Vaughan Moss
DBE
SKA SA
3rd Floor, The Park
Park Road
Pinelands
7405
South Africa
Tel: +27 (0)21 506 7363
Fax: +27 (0)21 506 7375
Email:mailto:vaug...@ska.ac.zavaug...@ska.ac.za
Web:  http://www.ska.ac.zawww.ska.ac.za
Content-Type: application/pdf; name=roach_2_hardware.pdf
Content-Disposition: attachment; filename=roach_2_hardware.pdf
X-Attachment-Id: f_h69b3zj20

Re: casper-scm Commit c77818cc4abca3ab849913f7f940e2956b37b48f breaks ADC083000x2 block

Re: [casper] writing/reading qdr

Re: [casper] compiling fft block

Re: [casper] logging in to ROACH PPC through a tunnel

Re: [casper] Roach 2 mechanical drawings?

5 matches

Site Navigation

Mail list logo

Footer information