| Message ID | 20210319152133.28705-1-hhhawa@amazon.com |
|---|---|
| Headers | show |
| Series | Fix pinctrl-single pcs_pin_dbg_show() | expand |
On Fri, Mar 19, 2021 at 05:21:30PM +0200, Hanna Hawa wrote: > These patches fix the pcs_pin_dbg_show() function for the scenario where > a single register controls multiple pins (i.e. bits_per_mux is not zero) > Additionally, the common formula is moved to a separate function to > allow reuse. > > Changes since v3: > ----------------- > - define and set variable 'mux_bytes' in one line > - update commit message > > Changes since v2: > ----------------- > - move read() register to be outside of if condition (as it common > read()). > - Remove extra parentheses > - replace offset variable by direct return statements > > Changes since v1: > ----------------- > - remove unused variable in In function 'pcs_allocate_pin_table' > (Reported-by: kernel test robot <lkp@intel.com>) > > Hanna Hawa (3): > pinctrl: pinctrl-single: remove unused variable > pinctrl: pinctrl-single: remove unused parameter > pinctrl: pinctrl-single: fix pcs_pin_dbg_show() when bits_per_mux is > not zero > > drivers/pinctrl/pinctrl-single.c | 65 ++++++++++++++++++-------------- > 1 file changed, 37 insertions(+), 28 deletions(-) > > -- > 2.17.1 > I'm curious what SoC are you using? It's good to know who has hardware to test bits_per_mux in the future. I pay attention to pinctrl-single as that is the driver used for the TI AM3358 SoC used in a variety of BeagleBone boards. It does not use bits_per_mux, but I can verify that this does not cause any regression for the AM3358 SoC: /sys/kernel/debug/pinctrl/44e10800.pinmux-pinctrl-single# cat pins registered pins: 142 pin 0 (PIN0) 0:? 44e10800 00000027 pinctrl-single pin 1 (PIN1) 0:? 44e10804 00000027 pinctrl-single pin 2 (PIN2) 0:? 44e10808 00000027 pinctrl-single pin 3 (PIN3) 0:? 44e1080c 00000027 pinctrl-single pin 4 (PIN4) 0:? 44e10810 00000027 pinctrl-single pin 5 (PIN5) 0:? 44e10814 00000027 pinctrl-single pin 6 (PIN6) 0:? 44e10818 00000027 pinctrl-single pin 7 (PIN7) 0:? 44e1081c 00000027 pinctrl-single pin 8 (PIN8) 22:gpio-96-127 44e10820 00000027 pinctrl-single pin 9 (PIN9) 23:gpio-96-127 44e10824 00000037 pinctrl-single pin 10 (PIN10) 26:gpio-96-127 44e10828 00000037 pinctrl-single pin 11 (PIN11) 27:gpio-96-127 44e1082c 00000037 pinctrl-single pin 12 (PIN12) 0:? 44e10830 00000037 pinctrl-single <snip> pin 140 (PIN140) 0:? 44e10a30 00000028 pinctrl-single pin 141 (PIN141) 13:gpio-64-95 44e10a34 00000020 pinctrl-single Reviewed-by: Drew Fustini <drew@beagleboard.org> Thanks, Drew
* Hanna Hawa <hhhawa@amazon.com> [700101 02:00]: > Remove unused parameter 'num_pins_in_register' from > pcs_allocate_pin_table(). Reviewed-by: Tony Lindgren <tony@atomide.com>
* Hanna Hawa <hhhawa@amazon.com> [700101 02:00]: > Remove unused parameter 'pin_pos' from pcs_add_pin(). Reviewed-by: Tony Lindgren <tony@atomide.com>
On 3/22/2021 7:56 AM, Drew Fustini wrote: > I'm curious what SoC are you using? I'm working on Amazon Annapurna Labs SoCs (based on ARM cortex processors). That include multiple pins controlled with same register. > > It's good to know who has hardware to test bits_per_mux in the future. > > I pay attention to pinctrl-single as that is the driver used for the TI > AM3358 SoC used in a variety of BeagleBone boards. It does not use > bits_per_mux, but I can verify that this does not cause any regression > for the AM3358 SoC: > > /sys/kernel/debug/pinctrl/44e10800.pinmux-pinctrl-single# cat pins > registered pins: 142 > pin 0 (PIN0) 0:? 44e10800 00000027 pinctrl-single > pin 1 (PIN1) 0:? 44e10804 00000027 pinctrl-single > pin 2 (PIN2) 0:? 44e10808 00000027 pinctrl-single > pin 3 (PIN3) 0:? 44e1080c 00000027 pinctrl-single > pin 4 (PIN4) 0:? 44e10810 00000027 pinctrl-single > pin 5 (PIN5) 0:? 44e10814 00000027 pinctrl-single > pin 6 (PIN6) 0:? 44e10818 00000027 pinctrl-single > pin 7 (PIN7) 0:? 44e1081c 00000027 pinctrl-single > pin 8 (PIN8) 22:gpio-96-127 44e10820 00000027 pinctrl-single > pin 9 (PIN9) 23:gpio-96-127 44e10824 00000037 pinctrl-single > pin 10 (PIN10) 26:gpio-96-127 44e10828 00000037 pinctrl-single > pin 11 (PIN11) 27:gpio-96-127 44e1082c 00000037 pinctrl-single > pin 12 (PIN12) 0:? 44e10830 00000037 pinctrl-single > <snip> > pin 140 (PIN140) 0:? 44e10a30 00000028 pinctrl-single > pin 141 (PIN141) 13:gpio-64-95 44e10a34 00000020 pinctrl-single > > Reviewed-by: Drew Fustini<drew@beagleboard.org> Thanks for review and verify the change. Thanks, Hanna > > Thanks, > Drew
On Fri, Mar 19, 2021 at 4:22 PM Hanna Hawa <hhhawa@amazon.com> wrote: > These patches fix the pcs_pin_dbg_show() function for the scenario where > a single register controls multiple pins (i.e. bits_per_mux is not zero) > Additionally, the common formula is moved to a separate function to > allow reuse. This v4 patch set applied! Yours, Linus Walleij
These patches fix the pcs_pin_dbg_show() function for the scenario where a single register controls multiple pins (i.e. bits_per_mux is not zero) Additionally, the common formula is moved to a separate function to allow reuse. Changes since v3: ----------------- - define and set variable 'mux_bytes' in one line - update commit message Changes since v2: ----------------- - move read() register to be outside of if condition (as it common read()). - Remove extra parentheses - replace offset variable by direct return statements Changes since v1: ----------------- - remove unused variable in In function 'pcs_allocate_pin_table' (Reported-by: kernel test robot <lkp@intel.com>) Hanna Hawa (3): pinctrl: pinctrl-single: remove unused variable pinctrl: pinctrl-single: remove unused parameter pinctrl: pinctrl-single: fix pcs_pin_dbg_show() when bits_per_mux is not zero drivers/pinctrl/pinctrl-single.c | 65 ++++++++++++++++++-------------- 1 file changed, 37 insertions(+), 28 deletions(-)