Re: [PATCH net-next 0/3] Document more PTP timestamping known quirks

Vladimir Oltean [off-list ref] writes:

I've tried to collect and summarize the conclusions of these discussions:
https://patchwork.ozlabs.org/project/netdev/patch/20200711120842.2631-1-sorganov@gmail.com/
https://patchwork.ozlabs.org/project/netdev/patch/20200710113611.3398-5-kurt@linutronix.de/
which were a bit surprising to me. Make sure they are present in the
documentation.

By the way, there is another somewhat related issue that needs to be
addressed. I believe kernel needs to free user space from this trick
found even in kernel sources themselves:

tools/testing/selftests/ptp/testptp.c:87:

static clockid_t get_clockid(int fd)
{
#define CLOCKFD 3
        return (((unsigned int) ~fd) << 3) | CLOCKFD;
}

Once upon a time there was a patch for that, but I don't think it
addressed the issue correctly, here is more background:

https://lore.kernel.org/lkml/87y2pxvsbr.fsf@osv.gnss.ru/ (local)

Thanks,
-- Sergey