Re: [PATCH v3] HID: corsair-void: Add Corsair Void headset family driver

[Christophe JAILLET]

Le 22/08/2024 à 03:03, Stuart a écrit :
> > Maybe is_wired could be close to other bools below, to improve avoid
> > holes in the structure?
>
> Possibly, I put `name`, `is_wired` and `sidetone_max` together, since
> they only depend on the device model and are set once.
>
> > Missing newline.
>
> Done
>
> > I'm not familiar with the hid_hw_raw_request() API, but I think that a
> > kfree(send_buf) is missing here.
>
> The `__free(kfree)` on the declaration should take care of that

Sorry, I missed that.
You are right.

> > Nitpick: No need to init.
>
> Thanks, but that `ret` won't be in the next revision anyway
>
> > You could save 2 lines if ret was initialized when declared.
>
> Could I? Wouldn't it get overwritten by `hid_hw_raw_request`?

No, you are obviously right.

BTW, the return value of corsair_void_request_status() is not used.
Does it make sense to change it into avoid function?

> > devm_kasprintf() would simplify this.
>
> Well that's a lot simpler...
>
> Thanks,
> Stuart