Hi Richard
Sorry about that, updates to the mask of the reorder block did not cater
for defaults expected in other scripts. The reorder block mask parameter
defaults should now play properly. The latest ska-sa repo should work
better for you.
Regards
Andrew
On 21/01/2014 19:46, Richard Black wrote:
Andrew,
git log -1 returns the following:
---------------------
commit 895572b2fd727391d501ff1a6f9686313d62d0d1
Author: Paul Prozesky <pa...@ska.ac.za <mailto:pa...@ska.ac.za>>
Date: Tue Jan 14 17:25:53 2014 +0200
Change needed for updated swreg
---------------------
I assume this is what you requested. If not, let me know.
Using a testbench model with only an fft_biplex_real_2x block (and the
requisite sysgen and xsg blocks), the error does not come up when
running the fft_biplex_real_2x_init.m script. However, when running the
bi_real_unscr_2x_init.m script, the aforementioned error appears.
With casper_log_groups = {'reorder_init_debug'}, the
bi_real_unscr_2x_init.m script logs the following:
---------------------
reorder_init_debug, trace: entering reorder_init
Error using bi_real_unscr_2x_init (line 120)
Invalid Simulink object name: reorder_even/1
---------------------
This makes me believe that the reorder_init.m script has an issue.
Should I clone an earlier git revision before the recent changes to the
reorder block?
Thanks,
On Mon, Jan 20, 2014 at 12:20 AM, Andrew Martens <and...@ska.ac.za
<mailto:and...@ska.ac.za>> wrote:
Hi Richard
Which casper repo are you using?
There have been some recent changes to the reorder block in the
ska-sa repo, which require swapping out all reorder blocks in your
design for new ones (or swapping the block containing them of course).
Cheers
Andrew
I just found this same problem today with a fresh install of RHEL 6,
Xilinx 14.5, and Matlab r2012b.
When I place an fft_biplex_real_2x block and run an update, I
get the
prompt: "Invalid Simulink object name: reorder_even/2." I also
see in
the "bi_real_unscr_2x" block under the mask that an uninitialized
reorder block is placed and not connected to anything.
I've tried re-downloading the casper library, restarting
simulink, and
removing the block and reinserting it, all to no avail.
I also noticed that the problem exists in every block that calls
"reuse_block" on the "reorder" block in
casper_library/casper_library___reorder.mdl. I was able to
reproduce the
same problem with all of the FFT green blocks.
I also suspect it is a problem with the init code, but I haven't
been
able to identify the exact problem yet.
On Fri, Jan 17, 2014 at 12:56 PM, Price, Daniel
<dpr...@cfa.harvard.edu <mailto:dpr...@cfa.harvard.edu>
<mailto:dpr...@cfa.harvard.edu
<mailto:dpr...@cfa.harvard.edu>__>> wrote:
Hi all
I have setup a fresh install of Xilinx 14.6, Matlab r2012b
on ubuntu
12.04, ska-sa library, and am getting errors when
generating blocks.
For example:
-->Error in 'untitled/pfb_fir_generic': Initialization commands
cannot be evaluated. -->Invalid Simulink object name:
pfb_fir_coeff_gen/2
which when using casper_log_groups = {'trace'}; to trace, I
get:
trace: entering pfb_fir_generic_init
trace: entering pfb_fir_generic_init
trace: entering bus_create_init
trace: exiting bus_create_init
trace: entering bus_expand_init
trace: exiting bus_expand_init
trace, pfb_fir_coeff_gen_init_debug: entering
pfb_fir_coeff_gen_init
I also get errors for the FFT, so am thinking that it might
be a
config issue, not a library issue:
-->Error in 'untitled/fft_biplex': Initialization commands
cannot be
evaluated. -->Invalid Simulink object name:
biplex_cplx_unscrambler/3
... many lines ...
trace, biplex_core_init_debug: exiting biplex_core_init
trace, barrel_switcher_init_debug: entering
barrel_switcher_init
trace, barrel_switcher_init_debug: exiting barrel_switcher_init
reorder_init_debug, trace: entering reorder_init
reorder_init_debug, trace: exiting reorder_init
reorder_init_debug, trace: entering reorder_init
reorder_init_debug, trace: exiting reorder_init
Has anyone seen this before, and/or have ideas for a fix?
Thanks
Danny
