Hi, On 24/10/15 04:32, Timur Tabi wrote: > Jun Nie wrote: >> I am OK to add a new driver for ZTE UART if no other's objection >> though I do not agree you. Without this last patch, other patches are >> just for minor refactor for register access. > > In that case, my vote is to drop the whole patchset. Just copy/paste > the amba-pl011 driver and make whatever changes you need. That sort of > thing happens all the time. I don't agree here. The driver _is_ very similar (all the semantics and behaviour), it's just the register addresses that are different. In so far I'd go with that approach of supporting it within the PL011 driver - whether this has to be in this very same file is another question. I think by using register accessors (avoiding direct calls to writel/readl in the code) we can hide a lot of these differences in just two functions. Also, as I said earlier, I guess we don't need all those #ifdefs. So in the end the driver wouldn't be too different, it's just that those diffs that look a bit scary (because they touch every readl/writel). I tried to refactor the driver lately to split up SBSA and PL011 support and got something that compiles, though I wasn't fully satisfied and I ran out of time. The refactor idea was to split driver runtime from initialization, so the different probe and init functions can be moved into separate files. There would be one stub file with all the core driver logic (DMA, IRQ handling, buffer handling, communication parameters setup) and one file for each subtype (PL011, SBSA, ZTE, you name it). If people are interested, I can try to clean this up and post it as an RFC. Cheers, Andre. -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
