09-03-2020 06:50 AM
I am using Vivado 2020.1. It goes all the way through implementation and shows no failed routes. But when it comes to opening the implemented design (or opening the routed design checkpoint) I get the following 2 warnings and 1 error:
WARNING: [Designutils 20-1361] Device model changes made it impossible to restore route node: INT_X52Y246/BOUNCE_E_11_FTS because the parent node: INT_X52Y246/INT_NODE_IMUX_12_INT_OUT has no arc that connects to this node.
CRITICAL WARNING: [Designutils 20-205] Net g_endpoints.crth0/dataMUXn/data_out_i_3_n_0 was unrouted during load.
WARNING: [Designutils 20-1361] Device model changes made it impossible to restore route node: INT_X20Y246/BOUNCE_E_11_FTS because the parent node: INT_X20Y246/INT_NODE_IMUX_7_INT_OUT has no arc that connects to this node.
CRITICAL WARNING: [Designutils 20-205] Net g_endpoints.pcie0/regsync0/register_map_decoding_monitor_p1_reg[DECODING_LINK_ALIGNED]_125 was unrouted during load.
Finished XDEF File Restore: Time (s): cpu = 00:00:34 ; elapsed = 00:00:33 . Memory (MB): peak = 31578.199 ; gain = 0.000 ; free physical = 20741 ; free virtual = 25243
ERROR: [Project 1-74] Physical restoration did not complete without warnings and strict mode is enabled.
Any idea what could be the problem?
09-23-2020 04:42 AM
Is this a DFX (PR) design? From the description, it looks to be a bug. Open the implemented design and run report_route_status to check if all the nets are routed and there are no routing error nets.
Can you share the runme.log file present in <project>/<project>.runs/impl_1/runme.log
Can you share the post opt dcp as well? I have sent you email from ezmove (ftp)
09-23-2020 05:03 AM
My post was placed 2 weeks ago, and I have been doing a ton of updates on the project in the mean time. This problem somehow magically disappeared with slight changes in the code.
So unfortunately I don't have the .dcp and the runme.log, so there is no way I can provide them anymore. I might be able to restore that particular version from GIT, but that would take too much of my time for a problem that already solved itself.
About your question to run "report_route_status", I think I described in my original message that I was unable to open the implemented design due to the error, so that would have been impossible.