Hi Laura,
(Strangely, your message does not appear in the RapidWright Google Group home page).
To be clear, RapidWright does not evaluate or apply XDC constraints within the RapidWright framework. It has APIs to read/write them, but there is not an XDC parser or engine to apply those constraints. Likely what you are seeing is that Vivado is not applying the XDC constraints until you force it to by clicking on the dialogs that cause re-evaluation to happen.
Fortunately, there is a way to set the logical/physical pin mapping in RapidWright through some APIs:
These should help you achieve the mapping you are trying to achieve.
Chris
From: rapid...@googlegroups.com <rapid...@googlegroups.com>
On Behalf Of Laura L.
Sent: Tuesday, August 3, 2021 7:20 AM
To: RapidWright <rapid...@googlegroups.com>
Subject: Constraints
Hi,
I am trying to set XDC constraints in RapidWright using the setXDCConstraints method from the Design class. In my design, I am trying to manually route to gain a better understanding of how to route in RapidWright, directly adding the PIPs to the net. However, when I do this, sometimes the logical/physical pin mapping isn't the default that Vivado expects, leading to routing errors, so I want to be able to change this mapping to be consistent with my routing. When I open the RapidWright generated DCP in Vivado, then I see this occurring but right-clicking on the component (i.e. a LUT), then I can use the "Reassign Cell Input Pins" to change the logical/physical mapping. However, I was hoping to be able to do this in RapidWright, so I used the setXDCConstraints method, inputting the same Tcl command as was generated by the manual method in Vivado. When I do this, the DCP generated still has the incorrect logical/physical pin mapping but when I right-click to change the mapping, it has the mapping I want already loaded in. Clicking OK to confirm the mapping then gives me what I expected, but is there any way to have RapidWright output the correctly mapped DCP automatically, without me having to manually confirm it?
Thank you,
Laura
--
You received this message because you are subscribed to the Google Groups "RapidWright" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
rapidwright...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/rapidwright/fee9fb9d-8a8b-4a3f-87d1-1a74458222c2n%40googlegroups.com.