On Wed, Jul 13, 2022 at 7:37 PM Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote: > On Wed, Jul 13, 2022 at 03:14:18PM +0200, Nuno Sá wrote: > > This change prepares the gpio core to look at firmware flags and set > > 'FLAG_BIAS_DISABLE' if necessary. It works in similar way to > > 'GPIO_PULL_DOWN' and 'GPIO_PULL_UP'. > > ... > > > GPIO_PULL_UP = (1 << 4), > > GPIO_PULL_DOWN = (1 << 5), > > + GPIO_PULL_DISABLE = (1 << 6), > > To me it seems superfluous. You have already two flags: > PUp > PDown > When none is set --> Pdisable What happens in the pin control case for some drivers at least is that the machine comes up with some pull up/downs enabled (by power-on default or from the boot loader), and some systems need to explicitly disable these pulls. In these (device tree) cases they set bias-disable; in the device tree, and the driver will actively disable any pull up/down. OK this is maybe not the most elegant system engineering. But some of those users are hobbyists and cannot affect what the ASIC or firmware is doing, because vendors are not really listening. Another semantic reason is that pins can also be set to bias-high-impedance; which is what "some people" would assume is the default if you disable both pull up and pull down. (Yeah ... semantics...) Device tree also has bias-pull-pin-default; to make things more complicated. This should *really* leave it at power-on default. Explicitly. I think for Nuno's usecase (using a random pin from userspace) the state of biasing cannot be assumed, the driver will not change bias to disabled just because neither pull up or down is specified, so the driver needs an explicit kick saying "disable any bias". Yours, Linus Walleij
