Re: [PATCH 1/4] gpiolib: add support for bias pull disable

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, 2022-07-14 at 12:20 +0800, Kent Gibson wrote:
> On Wed, Jul 13, 2022 at 08:36:38PM +0300, Andy Shevchenko 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
> > 
> 
> Agree with Andy on this.  The FLAG_BIAS_DISABLE was added, by me, to
> allow the cdev interface to support bias.  cdev requires a "don't
> care"
> state, distinct from an explicit BIAS_DISABLE.
> The FLAG_BIAS_DISABLE allows gpiolib-cdev to communicate that to
> gpiolib, without altering the interpretation of the existing PULL_UP
> and
> PULL_DOWN flags.
> That is not an issue on the machine interface, where the two
> GPIO_PULL
> flags suffice.
> 

I see, but this means we can only disable the pin BIAS through
userspace. I might be wrong but I don't see a reason why it wouldn't be
valid to do it from an in kernel path as we do for PULL-UPS and PULL-
DOWNS 

> If you are looking for the place where FLAG_BIAS_DISABLE is set it is
> in
> gpio_v2_line_config_flags_to_desc_flags() in gpiolib-cdev.c.
> 
> Referring to gpio_set_bias(), the only place in gpiolib the
> FLAG_BIAS_DISABLE is used, if neither FLAG_PULL_UP, FLAG_PULL_DOWN,
> nor FLAG_BIAS_DISABLE are set then the bias configuration remains
> unchanged (the don't care case) - no change is passed to the driver.
> Otherwise the corresponding PIN_CONFIG_BIAS flag is passed to the
> driver.
> 

Exactly, but note FLAG_BIAS_DISABLE can only be set from userspace at
this point (IIUTC). If everyone agrees that should be case, so be it.
But as I said, I just don't see why it's wrong to do it within the
kernel.

> If there are cases of drivers not fully or correctly supporting those
> PIN_CONFIG_BIAS flags, then that is an issue with those drivers.
> 

Look at my reply to Andy in the cover for more details

- Nuno Sá




[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]


  Powered by Linux