cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Explorer
Explorer
1,508 Views
Registered: ‎01-15-2008

ECO Flow

Jump to solution

I was very happy to learn about ECO Flow, whereby one can re-assign the ports of an ILA to quickly regenerate a bitfile with different debug outputs.  I watched the Quick Take video and followed the steps there, opening a checkpoint (Viv2017.4, KC705 dev board) and changing the connections to the ILA.  But although the tool accepted my changes and the Incremental Route proceeded to completion, the new schematic did not show those ILA ports connected.  The old connections had been removed but the new ones were not made.

 

It's my first time using this new capability, so maybe I missed something, but it all flowed very smoothly, just like the video, with now warnings or errors.  Just didn't work.

 

Anybody have any ideas?

 

Thanks,

Rick

0 Kudos
1 Solution

Accepted Solutions
Highlighted
Moderator
Moderator
1,401 Views
Registered: ‎11-04-2010

Re: ECO Flow

Jump to solution

Hi, @rikraf ,

The below command can work in 2018.2:

open_checkpoint ./checkpoint_1.dcp

modify_debug_ports -probes [list {u_ila_1/probe1 0 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/m_axis_tready} {u_ila_1/probe1 1 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/m_axis_tvalid} {u_ila_1/probe3 0 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[4]} {u_ila_1/probe3 1 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[5]} {u_ila_1/probe3 3 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[7]} {u_ila_1/probe3 4 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[8]} {u_ila_1/probe1 2 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/s_axis_tready} {u_ila_1/probe1 3 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/s_axis_tvalid} {u_ila_1/probe3 2 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[6]} {u_ila_1/probe3 5 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[9]}]

 

 

-------------------------------------------------------------------------
Don't forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

ECO.png
0 Kudos
6 Replies
Highlighted
Moderator
Moderator
1,487 Views
Registered: ‎11-04-2010

Re: ECO Flow

Jump to solution

Hi, @rikraf ,

You can refer the attached add_probe.tcl and show us what commands you are using.

You don't need to send out your own design and you can take the example design in Vivado to show what you are doing.

-------------------------------------------------------------------------
Don't forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Explorer
Explorer
1,436 Views
Registered: ‎01-15-2008

Re: ECO Flow

Jump to solution

Attached is the set of commands generated by the GUI.  Having done that, the regenerated schematic shows all of the nets from probes 1 and 3 of ila_1 to be disconnected.  These were the ones that were previously connected, and which I tried to connect to new nets.

 

ECO.JPG

route_design seems to run without problem:

INFO: [DRC 23-27] Running DRC with 2 threads
INFO: [Vivado_Tcl 4-198] DRC finished with 0 Errors
INFO: [Vivado_Tcl 4-199] Please refer to the DRC report (report_drc) for more information.
Running DRC as a precondition to command route_design
INFO: [DRC 23-27] Running DRC with 2 threads
INFO: [Vivado_Tcl 4-198] DRC finished with 0 Errors
INFO: [Vivado_Tcl 4-199] Please refer to the DRC report (report_drc) for more information.
INFO: [Common 17-83] Releasing license: Implementation
8 Infos, 0 Warnings, 0 Critical Warnings and 0 Errors encountered.
route_design completed successfully

 

Any help appreciated.

 

Thanks

0 Kudos
Highlighted
Moderator
Moderator
1,421 Views
Registered: ‎11-04-2010

Re: ECO Flow

Jump to solution

Hi, @rikraf ,

Generally, after executing "modify_debug_ports" command, it's no need to run route_design design again.

 

In your design, the new connection for probe is not established. There must be something wrong.

Could you provide your routed.dcp for further analysis? Are you using Vivado 2018.2?

-------------------------------------------------------------------------
Don't forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Explorer
Explorer
1,412 Views
Registered: ‎01-15-2008

Re: ECO Flow

Jump to solution

Yes, here it is.  The QuickTake video seemed pretty clear on having to do an incremental route after replacing the probes

0 Kudos
Highlighted
Moderator
Moderator
1,402 Views
Registered: ‎11-04-2010

Re: ECO Flow

Jump to solution

Hi, @rikraf ,

The below command can work in 2018.2:

open_checkpoint ./checkpoint_1.dcp

modify_debug_ports -probes [list {u_ila_1/probe1 0 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/m_axis_tready} {u_ila_1/probe1 1 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/m_axis_tvalid} {u_ila_1/probe3 0 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[4]} {u_ila_1/probe3 1 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[5]} {u_ila_1/probe3 3 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[7]} {u_ila_1/probe3 4 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[8]} {u_ila_1/probe1 2 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/s_axis_tready} {u_ila_1/probe1 3 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/s_axis_tvalid} {u_ila_1/probe3 2 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[6]} {u_ila_1/probe3 5 base_mb_i/maroc_dc_0/inst/USR_LOGIC/PH_FIFO/axis_data_count[9]}]

 

 

-------------------------------------------------------------------------
Don't forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

ECO.png
0 Kudos
Highlighted
Explorer
Explorer
1,343 Views
Registered: ‎01-15-2008

Re: ECO Flow

Jump to solution

Hello Hong,

Thanks very much for the TCL script.  In fact I was able to use the "ECO flow" GUI successfully on another design, where I reassigned the debug probes. And, as you said, there was no need to run the "route design" process.  Maybe this was my mistake previously.  I'll use the tcl script next time and see how it goes.

Rick

0 Kudos