Re: [PATCH v4 1/2] iio: imu: inv_icm42600: add WoM support

[Andy Shevchenko <andriy.shevchenko@intel.com>]

On Fri, Jun 13, 2025 at 12:46:46PM +0000, Jean-Baptiste Maneyrol wrote:
> >From: Andy Shevchenko <andriy.shevchenko@intel.com>
> >Sent: Friday, June 13, 2025 10:29
> >On Fri, Jun 13, 2025 at 09:34:26AM +0200, Jean-Baptiste Maneyrol via B4 Relay wrote:

...

> >> +	if (sleep_ms)
> >> +		msleep(sleep_ms);
> >
> >I still wonder if we can get rid of the conditional here.
> >Would the
> >
> >	fsleep(sleep_ms * USEC_PER_MSEC)
> >
> >actually work as expected?
> >
> >Ditto for other case(s) like this.
> 
> fsleep(0) would call udelay(0) which is architecture dependent. It seems like
> it may delay for a very little while, but I'm not able to check that.

Hmm... This is unfortunate. Somebody needs at least make it clear in the
documentation if not yet done.

...

> >Overall, looking to this patch again, I think it would be better to prepend it
> >by replacing *int*_t types by the respective uXX ones. Because in this patch
> >we add dozens of new ones which increases an unneeded churn in the future.
> >
> In my opinion, to respect the rule don't mix *int*_t and uXX types, it is better
> to keep *int*_t types. If it need to be changed, we can change afterward the
> whole driver types with a replace tool and send it in a separate patch.

It will be never ending story, sorry. We need someone to solve this tech debt.
And since this patch adds more than 3 new users of it, I think it's a candidate
to embrace the burden.

> Jonathan,
> what is your statement on this point?

-- 
With Best Regards,
Andy Shevchenko