Re: [PATCH v2 2/2] PTP: add support for one-shot output

[Richard Cochran <richardcochran@gmail.com>]

On Fri, Aug 30, 2019 at 11:00:20AM +0300, Felipe Balbi wrote:
> >> @@ -177,9 +177,8 @@ long ptp_ioctl(struct posix_clock *pc, unsigned int cmd, unsigned long arg)
> >>  			err = -EFAULT;
> >>  			break;
> >>  		}
> >> -		if ((req.perout.flags || req.perout.rsv[0] || req.perout.rsv[1]
> >> -				|| req.perout.rsv[2] || req.perout.rsv[3])
> >> -			&& cmd == PTP_PEROUT_REQUEST2) {
> >> +		if ((req.perout.rsv[0] || req.perout.rsv[1] || req.perout.rsv[2]
> >> +			|| req.perout.rsv[3]) && cmd == PTP_PEROUT_REQUEST2) {
> >
> > Please check that the reserved bits of req.perout.flags, namely
> > ~PTP_PEROUT_ONE_SHOT, are clear.
> 
> Actually, we should check more. PEROUT_FEATURE_ENABLE is still valid
> here, right? So are RISING and FALLING edges, no?

No.  The ptp_extts_request.flags are indeed defined:

struct ptp_extts_request {
	...
	unsigned int flags;  /* Bit field for PTP_xxx flags. */
	...
};

But the ptp_perout_request.flags are reserved:

struct ptp_perout_request {
	...
	unsigned int flags;           /* Reserved for future use. */
	...
};

For this ioctl, the test for enable/disable is
ptp_perout_request.period is zero:

		enable = req.perout.period.sec || req.perout.period.nsec;
		err = ops->enable(ops, &req, enable);

The usage pattern here is taken from timer_settime(2).

Thanks,
Richard