| Message ID | 20240828184018.3097386-1-andriy.shevchenko@linux.intel.com |
|---|---|
| Headers | show |
| Series | pinctrl: intel: High impedance impl. and cleanups | expand |
+Cc: Heiner On Wed, Aug 28, 2024 at 09:38:33PM +0300, Andy Shevchenko wrote: > We would need a high impedance implementation for a quirk, so here it > is. While doing this series I also noticed a couple of opportunities > to clean up, hence two more patches (1st and 5th). Sorry it took a while to actually start implementing the quirk for your case. Here I'm asking for the following things: 1) what is the marketing name of the device you have problems with? (I believe it's available on the free market, correct?); 2) does it have any BIOS updates and, if it has, does it fix the issue? 3) can you apply patches 2,3,4,5 from this series (the first one is buggy and not needed for you) and replace the hack I mentioned earlier with ret = intel_gpio_set_high_impedance(pctrl, 3); if (ret) return ret; somewhere at the end of intel_pinctrl_probe()? Does it still work as expected?
On 28.08.2024 22:00, Andy Shevchenko wrote: > +Cc: Heiner > > On Wed, Aug 28, 2024 at 09:38:33PM +0300, Andy Shevchenko wrote: >> We would need a high impedance implementation for a quirk, so here it >> is. While doing this series I also noticed a couple of opportunities >> to clean up, hence two more patches (1st and 5th). > > Sorry it took a while to actually start implementing the quirk for your case. > Here I'm asking for the following things: > > 1) what is the marketing name of the device you have problems with? > (I believe it's available on the free market, correct?); > Device is a dirt-cheap mini pc, marketed as Chatreey T9. It's available on the free market, right. Dmesg says: DMI: Default string Default string/Default string, BIOS ADLN.M6.SODIMM.ZB.CY.015 08/08/2023 > 2) does it have any BIOS updates and, if it has, does it fix the issue? > No BIOS updates. > 3) can you apply patches 2,3,4,5 from this series (the first one is buggy and > not needed for you) and replace the hack I mentioned earlier with > > ret = intel_gpio_set_high_impedance(pctrl, 3); > if (ret) > return ret; > > somewhere at the end of intel_pinctrl_probe()? > > Does it still work as expected? > > I will check.
On Fri, Aug 30, 2024 at 07:56:11AM +0200, Heiner Kallweit wrote: > On 28.08.2024 22:00, Andy Shevchenko wrote: > > On Wed, Aug 28, 2024 at 09:38:33PM +0300, Andy Shevchenko wrote: > >> We would need a high impedance implementation for a quirk, so here it > >> is. While doing this series I also noticed a couple of opportunities > >> to clean up, hence two more patches (1st and 5th). > > > > Sorry it took a while to actually start implementing the quirk for your case. > > Here I'm asking for the following things: > > > > 1) what is the marketing name of the device you have problems with? > > (I believe it's available on the free market, correct?); > > Device is a dirt-cheap mini pc, marketed as Chatreey T9. It's available > on the free market, right. Dmesg says: > DMI: Default string Default string/Default string, BIOS ADLN.M6.SODIMM.ZB.CY.015 08/08/2023 > > > 2) does it have any BIOS updates and, if it has, does it fix the issue? > > > No BIOS updates. > > > 3) can you apply patches 2,3,4,5 from this series (the first one is buggy and > > not needed for you) and replace the hack I mentioned earlier with > > > > ret = intel_gpio_set_high_impedance(pctrl, 3); > > if (ret) > > return ret; > > > > somewhere at the end of intel_pinctrl_probe()? > > > > Does it still work as expected? > > > I will check. There is a v2 to test, you can take entire series, or for-next branch of Intel pin control tree https://git.kernel.org/pub/scm/linux/kernel/git/pinctrl/intel.git/log/?h=for-next Thanks!
On 28.08.2024 22:00, Andy Shevchenko wrote: > +Cc: Heiner > > On Wed, Aug 28, 2024 at 09:38:33PM +0300, Andy Shevchenko wrote: >> We would need a high impedance implementation for a quirk, so here it >> is. While doing this series I also noticed a couple of opportunities >> to clean up, hence two more patches (1st and 5th). > > Sorry it took a while to actually start implementing the quirk for your case. > Here I'm asking for the following things: > > 1) what is the marketing name of the device you have problems with? > (I believe it's available on the free market, correct?); > > 2) does it have any BIOS updates and, if it has, does it fix the issue? > > 3) can you apply patches 2,3,4,5 from this series (the first one is buggy and > not needed for you) and replace the hack I mentioned earlier with > > ret = intel_gpio_set_high_impedance(pctrl, 3); > if (ret) > return ret; > > somewhere at the end of intel_pinctrl_probe()? > > Does it still work as expected? > > In latest series the return value of intel_gpio_set_high_impedance() has been removed. With the call to intel_gpio_set_high_impedance() changed accordingly this fixes the problem on my system. Thanks a lot for your efforts!
On Fri, Aug 30, 2024 at 11:24:06PM +0200, Heiner Kallweit wrote: > On 28.08.2024 22:00, Andy Shevchenko wrote: > > On Wed, Aug 28, 2024 at 09:38:33PM +0300, Andy Shevchenko wrote: > >> We would need a high impedance implementation for a quirk, so here it > >> is. While doing this series I also noticed a couple of opportunities > >> to clean up, hence two more patches (1st and 5th). > > > > Sorry it took a while to actually start implementing the quirk for your case. > > Here I'm asking for the following things: > > > > 1) what is the marketing name of the device you have problems with? > > (I believe it's available on the free market, correct?); > > > > 2) does it have any BIOS updates and, if it has, does it fix the issue? > > > > 3) can you apply patches 2,3,4,5 from this series (the first one is buggy and > > not needed for you) and replace the hack I mentioned earlier with > > > > ret = intel_gpio_set_high_impedance(pctrl, 3); > > if (ret) > > return ret; > > > > somewhere at the end of intel_pinctrl_probe()? > > > > Does it still work as expected? > > > > > In latest series the return value of intel_gpio_set_high_impedance() > has been removed. With the call to intel_gpio_set_high_impedance() > changed accordingly this fixes the problem on my system. > Thanks a lot for your efforts! Thank you! I will think now, how to make the real quirk (which I may ask you to test in the future, when it will be ready) that looks and feels not as an ugly hack.
We would need a high impedance implementation for a quirk, so here it is. While doing this series I also noticed a couple of opportunities to clean up, hence two more patches (1st and 5th). Andy Shevchenko (5): pinctrl: intel: Move debounce validation out of the lock pinctrl: intel: Refactor __intel_gpio_set_direction() to be more useful pinctrl: intel: Add __intel_gpio_get_direction() helper pinctrl: intel: Implement high impedance support pinctrl: intel: Introduce for_each_intel_gpio_group() helper drivers/pinctrl/intel/pinctrl-intel.c | 239 ++++++++++++++++---------- 1 file changed, 150 insertions(+), 89 deletions(-)