On Mon, Aug 21, 2023 at 05:59:57AM -0700, Guenter Roeck wrote: > On Mon, Aug 21, 2023 at 12:17:34PM +0300, Andy Shevchenko wrote: > > On Sun, Aug 20, 2023 at 08:55:18AM -0700, Guenter Roeck wrote: ... > > Those drivers are hackishly abusing OF ID tables. Those, that have _no_ OF ID > > tables are okay. > > This is what you are saying. That doesn't make it true. With the same logic > I could claim that drivers providing a pointer in i2c_device_id would > hackishly abuse i2c_device_id tables. ...because kernel_ulong_t was poorly chosen type for that to begin with? Abusing pointer is much worse than using long as a pointer holder. Moreover the latter is valid case (okay, not as kernel_ulong_t, but unsigned long) in many places in the kernel, starting from the memory manager, so your argument doesn't look like a good one to me. -- With Best Regards, Andy Shevchenko
