On 31/05/2023 01:12, Daniel Palmer wrote: > Hi Krzysztof, > > On Tue, 30 May 2023 at 17:01, Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote: >>> This is >>> exactly the case for rtc-msc313e and it was not an issue. >> >> So that was my question - can it be anything else? There is literally no >> description of the hardware... Neither in commit msg nor in description: >> field in bindings. > > This RTC block is a block inside of the SSD201/SSD202D (they are the But what is SSD201? > same die with different memory attached) and is only found there. > The documentation we have for this is literally one page in a PDF that > says "RTC registers". > It could be an IP block licensed from somewhere and technically have a > better name but right now all we know is this RTC block is the one in > that chip and that chip is the first known instance of it. > > Say we manage to get the ethernet mainlined at some point: That's a > lot easier as we know already it's a hacked up version of the cadence > macb so the compatible can be "macb something". > >>>> What about interrupt line? >>> >>> There is currently no interrupt right now, we have not yet the irqchip >>> code for handling the alarm irq of this rtc block. >> >> So you are going to change the hardware and add the interrupt line? We >> do not talk about drivers, but hardware. Whether your driver handles it >> or not, matters less. >> >> Describe the hardware, not the current implementation of one driver. > > We don't really know how the interrupt is wired up in the hardware properly yet. OK Best regards, Krzysztof
