On Tue, 14 Jan 2025 23:14:13 +0100, Terry Junge wrote: > > On 1/14/25 4:38 AM, Takashi Iwai wrote: > > On Tue, 14 Jan 2025 00:51:59 +0100, > > Terry Junge wrote: > >> > >> +/* > >> + * Some Plantronics headsets have control names that don't meet ALSA naming > >> + * standards. This function fixes nonstandard source names. By the time > >> + * this function is called the control name should look like one of these: > >> + * "source names Playback Volume" > >> + * "source names Playback Switch" > >> + * "source names Capture Volume" > >> + * "source names Capture Switch" > >> + * If any of the trigger words are found in the name then the name will > >> + * be changed to: > >> + * "Headset Playback Volume" > >> + * "Headset Playback Switch" > >> + * "Headset Capture Volume" > >> + * "Headset Capture Switch" > >> + * depending on the current suffix. > >> + */ > >> +static void snd_fix_plt_name(struct snd_usb_audio *chip, > >> + typeof_member(struct snd_ctl_elem_id, name) * name) > > > > I personally find this style of argument is difficult to use. > > That said, IMO, it's better to stick with the argument > > struct snd_ctl_elem_id *id > > and access via id->name as in your earlier patch; it's more idiomatic, > > and easier to read. > > > So, is the coding of the rest of the function body acceptable if I just pass a > struct snd_ctl_elem_id *id > instead of an > unsigned char[44] *name > ? Likely yes. thanks, Takashi
