-
Notifications
You must be signed in to change notification settings - Fork 193
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Support a Vivado IP-packaging flow #393
Comments
Appreciate you reaching out. I think this does need some planning and scoping. E.g., is the intention to create RTL from a bd file or are you looking at some kind of automatic packaging as well? It's been a while since I last used this flow myself so my memory of the details are a bit rusty, but I think some people are using FuseSoC/Edalize with their own bd files and adding some custom tcl to do the RTL generation on the fly. Let's iterate on this so that I understand the use case a bit better. |
Our use-case doesn't involve BDs specifically, but rather automatically packaging our IPs so users can generate XCIs. As a specific example: we have a device we communicate with over a high-speed serial link, and we have the top-level for this device and the IP to communicate with it. Both managed by FuseSoC. Workflow-wise:
The IP itself consists of a number of our internal FuseSoC cores. Ideally this is the form we'd package them in, but there are two obstacles to this: (1) our users don't use FuseSoC, and (2) our internal cores are generically named, so we can't easily make sure there won't be a conflict between module names. By default, FuseSoC has most of the information we copy over into the IP packager. Adding the rest as extra properties in the Once we have the archive, our users can put its contents in their IP repo and specialise our parameterised IP using So my rough thinking on how this breaks down is:
Is there anything I've left unclear there? |
It would be useful if Edalize (and FuseSoC) supported Vivado's custom IP workflow (see UG1118).
We have a use-case for this, and we intend to commit engineer time to develop it. Ideally, we'd also contribute this back upstream so it was part of the in-the-box Vivado flows. Really, I'm opening this issue as a request for 'tell me how you want this done' so I can try to align what we produce with what the user-community wants.
The text was updated successfully, but these errors were encountered: