Bartosz Gołaszewski
Bartosz Gołaszewski
> Here's the strace for gpioinfo: > > [gpioinfo.log](https://github.com/user-attachments/files/21975473/gpioinfo.log) Ok, so there's an EINVAL coming out from the kernel, the user-space part looks fine (as long as you have not...
Oh, I only now noticed how old your kernel is. It probably doesn't even support libgpiod v2. I can't help you with upgrading, this is outside of the scope of...
This board is supported in mainline so you're good with any recent stable release really.
Makes sense, I guess.
I mean, `calloc()` ultimately does the same thing under the hood, no? I prefer to use it where it's clear we're allocating an array of objects.
I mean... Sure, send the patches, I won't reject them. But it's not like it's a really hot path here, is it?
I think what you may really need is the `gpio-manager` that's now part of the package? It exposes a D-Bus interface and already comes with a systemd service file. It...
> We can alias types in python, so instead of littering the codebase with complex types like: > > dict[Union[Iterable[Union[int, str]], int, str], Optional[LineSettings]] > > Or when we move...
`libgpiodbus` is entirely dynamically generated by `gdbus-codegen`. I suggest you create a library of such helpers and we can discuss making it part of `libgpiod` as a separate entity linking...
> [@brgl](https://github.com/brgl) Most, if not all, of the helper functions that I have utilized are directly from https://github.com/brgl/libgpiod/blob/master/dbus/client/common.c It feels a bit counterproductive to implement yet another library that provides...