Why use progmem.py instead of $readmemh or $readmmemb

[cheungxi]

[chili-chips-ba]

$readmemh / $readmmemb are static, i.e. compile-time.
progmem.py is dynamic, i.e. run-time.

The time it takes to compile and reload Artix7-35 is in tens of minutes. It is therefore much, much faster and more productive to only reload the RISC-V IMEM on an already-loaded, live FPGA, than to recompile and reload the entire FPGA.

Does this make sense?
(if it does, please don't shy away from⭐'ing our project 😉)

[cheungxi]

Thanks , I have stared project

But progmem.py is only used to generate progmem.v， progmem.v is compile-time

[chili-chips-ba]

Thank you for starring the project 👍.

We were too quick to provide an answer, and also forgot the details from the past, as we have in the meantime developed a similar, but enhanced scripting for the live IMEM reloading. For that, please check our wireguard-fpga project:

In the context of the original script, you are right to say that the same result could have been achieved with the single-liner $readmemh / $readmmemb. It is only that Python script gives you better control of the process, thus opens the door for doing more...