gpiolib: of: fix bounds check for 'gpio-reserved-ranges'
authorAndrei Lalaev <andrei.lalaev@emlid.com>
Fri, 15 Apr 2022 07:07:11 +0000 (10:07 +0300)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 12 May 2022 10:30:02 +0000 (12:30 +0200)
commitd46f64fd9172e72714e359daca5e13b1460bf4ce
treefdeb2c5a95edac86f0a9c56945e14364d68e7385
parenta06979edcc47d8bfbe4cb512bbc57502c04955cd
gpiolib: of: fix bounds check for 'gpio-reserved-ranges'

commit e75f88efac05bf4e107e4171d8db6d8c3937252d upstream.

Gpiolib interprets the elements of "gpio-reserved-ranges" as "start,size"
because it clears "size" bits starting from the "start" bit in the according
bitmap. So it has to use "greater" instead of "greater or equal" when performs
bounds check to make sure that GPIOs are in the available range.
Previous implementation skipped ranges that include the last GPIO in
the range.

I wrote the mail to the maintainers
(https://lore.kernel.org/linux-gpio/20220412115554.159435-1-andrei.lalaev@emlid.com/T/#u)
of the questioned DTSes (because I couldn't understand how the maintainers
interpreted this property), but I haven't received a response.
Since the questioned DTSes use "gpio-reserved-ranges = <0 4>"
(i.e., the beginning of the range), this patch doesn't affect these DTSes at all.
TBH this patch doesn't break any existing DTSes because none of them
reserve gpios at the end of range.

Fixes: 726cb3ba4969 ("gpiolib: Support 'gpio-reserved-ranges' property")
Signed-off-by: Andrei Lalaev <andrei.lalaev@emlid.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Cc: stable@vger.kernel.org
Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/gpio/gpiolib-of.c