On Thu, Aug 20, 2020 at 3:27 PM Daniel Palmer <daniel@xxxxxxxx> wrote: > On Thu, 20 Aug 2020 at 03:19, Denis CIOCCA <denis.ciocca@xxxxxx> wrote: > > > I strongly disagree that these parts will be supported by STMicroelectronics driver. > > The alternative seems to be write a totally new driver that does > exactly the same thing, use a lot of people's time to get it ready to > go into the kernel etc when we could come up with a way to allow the > driver to accept a different WAI register value in a non-intrusive way > that can go in the bin at a later date if it turns out supporting > these chips is too much hassle. Which is fine with me, I mean adding an ID is a common practice for competing hardware clones. > > We DO NOT want to find out one day that we need to modify our structure in order to support competition. > > I think Jonathan suggested adding me as a reviewer for these parts > thus making me responsible for them. > I'm fine with that. If it comes to the point where the driver changes > so much that it's not possible to keep them working then you'd have my > blessing to just remove support for them. I encourage you to do this. I guess it's much better if we have someone, than orphaned code in the kernel. > That said I'm not going to force this on anyone. > This patch was a spin-off of a personal project to try to make cheap > Linux capable SoCs available to makers[0]. And again full support from me here. It's a really very strange position of STMicro here. I as a maintainer / reviewer / supporter of a lot of hardware have a good example of collaboration with contributors from other companies when they do improve our drivers. And I have been doing the same for others, for example Synopsys DesignWare SATA driver which had been found on PPC44x platform. > One of the devices I used > to reverse engineer that hardware had one of these Silan parts, I > noticed the registers looked exactly the same as ST parts supported in > the kernel already and upon testing they worked so I thought it was > worth throwing this out there. If you guys don't want to deal with it > I'll just leave it in my tree. > [0] - https://github.com/breadbee/breadbee -- With Best Regards, Andy Shevchenko
