On Mon, Jan 15, 2024 at 05:49:32PM +0200, Sagi Maimon wrote: > Thanks for your notes, all of them will be done on the next patch (it > will take some time due to work overload). No hurry, glad you are keeping this going... > The only question that I have is: why not implement it as an IOCTL? > It makes more sense to me since it is close to another IOCTL, the > "PTP_SYS_OFFSET" family. I've often needed other clock offsets, like CLOCK_REALTIME - CLOCK_MONOTONIC. Those don't have a character device, and so there is no way to call ioctl() on them. That is why I'd like to have a system call that handles any two clock_t instances, using the most accurate back end based on the kinds of the two clocks. Thanks, Richard
