Messages in this thread | | | Date | Thu, 8 Sep 2022 10:25:44 +0200 | From | Uwe Kleine-König <> | Subject | Re: [PATCH v1 1/9] pwm: lpss: Deduplicate board info data structures |
| |
On Wed, Sep 07, 2022 at 05:27:22PM +0300, Andy Shevchenko wrote: > On Wed, Sep 07, 2022 at 11:04:12AM +0200, Uwe Kleine-König wrote: > > On Tue, Sep 06, 2022 at 10:57:27PM +0300, Andy Shevchenko wrote: > > > With help of __maybe_unused, that allows to avoid compilation warnings, > > > move the board info structures from the C files to the common header > > > and hence deduplicate configuration data. > > > > > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > > > --- > > > drivers/pwm/pwm-lpss-pci.c | 29 ----------------------------- > > > drivers/pwm/pwm-lpss-platform.c | 23 ----------------------- > > > drivers/pwm/pwm-lpss.h | 30 ++++++++++++++++++++++++++++++ > > > > Given that both the pci driver and the platform driver alread depend on > > pwm-lpss.o, I'd prefer something like the patch below to really > > deduplicate the data. > > Why not? I can use yours in v2. Can I get your SoB tag?
Sure:
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> > One thing to note is that the two pwm_lpss_bsw_info are not identical. I > > didn't check how that is relevant. Did you check that? > > Yes, ACPI version should be used. Because switch to ACPI/PCI is done in BIOS > while quite likely the rest of AML code is the same, meaning similar issue > might be observed. The no bug report is due to no PCI enabled device in the > wild, I think, and only reference boards can be tested, so nobody really cares > about PCI Braswell case.
I'm willing to believe that; please mention that in the commit log.
Best regards Uwe
-- Pengutronix e.K. | Uwe Kleine-König | Industrial Linux Solutions | https://www.pengutronix.de/ | [unhandled content-type:application/pgp-signature] | |