Messages in this thread | | | From | Felipe Balbi <> | Subject | Re: [RFC PATCH 0/5] PTP: add support for Intel's TGPIO controller | Date | Tue, 13 Aug 2019 10:50:06 +0300 |
| |
Hi,
Andrew Lunn <andrew@lunn.ch> writes: >> Andrew Lunn <andrew@lunn.ch> writes: >> > On Tue, Jul 16, 2019 at 10:20:33AM +0300, Felipe Balbi wrote: >> >> TGPIO is a new IP which allows for time synchronization between systems >> >> without any other means of synchronization such as PTP or NTP. The >> >> driver is implemented as part of the PTP framework since its features >> >> covered most of what this controller can do. >> > >> > Hi Felipe >> > >> > Given the name TGPIO, can it also be used for plain old boring GPIO? >> >> not really, no. This is a misnomer, IMHO :-) We can only assert output >> pulses at specified intervals or capture a timestamp of an external >> signal. > > Hi Felipe > > So i guess Intel Marketing wants to call it a GPIO, but between > engineers can we give it a better name?
If we do that we make it difficult for those reading specification and trying to find the matching driver.
>> > Also, is this always embedded into a SoC? Or could it actually be in a >> > discrete NIC? >> >> Technically, this could be done as a discrete, but it isn't. In any >> case, why does that matter? From a linux-point of view, we have a device >> driver either way. > > I've seen a lot of i210 used with ARM SoCs. How necessary is the tsc > patch? Is there an architecture independent alternative?
Without the TSC patch, we don't get the timestamp we need. One can argue that $this driver could call get_tsc_ns() directly instead of providing a wrapper for it. But that's something else entirely.
-- balbi
| |