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
Missing support for xc7a35tftg256-1 in archive #1800
Comments
Hi @dnltz , There is a guide for this case -- https://prjxray.readthedocs.io/en/latest/db_dev_process/newpart.html. You would just be adding a new package (256 pins) for a currently supported device (35t). So you'd skip steps 2-4 of that guide. However, we have planned a project to automatically generate the pinmaps for all packages once a device is supported -- f4pga/prjxray#1476 -- but I don't know what the timeline is for that. |
Thanks @tcal-x ! That was exactly what I was looking for :) Yes, would be nice to have all available packages in the example. Do you need more hardware for testing? I could send a board with this package in January. |
Hi @dnltz ,
|
@dnltz We would like someone to do f4pga/prjxray#1475 and f4pga/prjxray#1476. It should just be some Python work, if you feel up for it! |
hey @litghost - I'm very busy right now until end of December. I can do these PRs than, if it's not that urgent. |
Coming back to my original problem, the way to make my part available in the prebuild toolchain is to add it as board to this project, right? |
Next step is to update the prjxray-db package. I'll poke @mithro |
The Zynq build got stuck in some way, so I killed it and kicked it off again. In the meantime, I pushed the Artix + Kintex stuff to f4pga/prjxray-db@master...mithro:master There seems to be a lot of changes in the |
A zynq run completed successfully, see the changes at f4pga/prjxray-db@master...mithro:master If someone confirms that the mask changes for |
I am not sure what are the mask files used at this point, or if they are every used in the flow. As for they are expected or not, the only recent changes to the DSP happened here and the output is already checked-in in the db, so the change seems a bit odd. |
I think I can close this issue. @mithro Can I add a not for sale board with this part? |
@dnltz - Add to were? |
@mithro - My understand is that the flow needs data from |
@dnltz So you remember how we made the mapping from part -> device -> fabric in prjxray? That same distinction exists in arch-defs, but pre-dates the prjxray change. arch-defs simple maps part directly to fabric (and a "board") in https://github.com/SymbiFlow/symbiflow-arch-defs/blob/a3c75b4b7304a29ff3d48b594e5afe922ce00f7e/xc/xc7/boards.cmake The key CMake function is The key portion for expanding part support for everything is generating "pinmap.csv" for every part in prjxray, which is here: https://github.com/SymbiFlow/symbiflow-arch-defs/blob/a3c75b4b7304a29ff3d48b594e5afe922ce00f7e/xc/common/cmake/device_define.cmake#L160-L172 In summary:
Because the naming conventions in arch-defs appear to generally match the naming conventions from prjxray, I believe it should be possible to emit CMake targets to generate the "pinmap.csv" for every part now in prjxray as part of the install targets. Does this make sense?
|
One additional note, the current |
I'm not very familiar with cmake. A target is a function like |
|
Hi,
I downloaded the toolchain as guided in the symbiflow-example README file and wanted to try the xc7 counter example for the xc7a35tftg256-1 part. I saw there is support for it in prjxray-db but unfortunately this part is not deployed to the tar archive.
Error message from the symbiflow_place:
Seems like only the parts from the two development kits get deployed. Since I have a custom board, can I generate this pinmap by myself?
The text was updated successfully, but these errors were encountered: