Messages in this thread | | | From | Andy Shevchenko <> | Date | Tue, 5 Jul 2022 12:29:27 +0200 | Subject | Re: [PATCH v3] hisi_lpc: Use acpi_dev_for_each_child() |
| |
On Tue, Jul 5, 2022 at 12:16 PM John Garry <john.garry@huawei.com> wrote: > On 05/07/2022 10:39, Andy Shevchenko wrote: > > On Tue, Jul 5, 2022 at 11:38 AM Andy Shevchenko > > <andy.shevchenko@gmail.com> wrote: > >> On Tue, Jul 5, 2022 at 10:37 AM John Garry<john.garry@huawei.com> wrote:
...
> >> John, I believe now you may send a formal clean up to convert to platform_device > > Hit Enter too early:-) > > > > ...to platform_device_register_full(). > > Sure, I can look at that now. But I just found where we previously > mentioned the possibility of factoring out some of the ACPI platform > device creation code: > > https://lore.kernel.org/linux-acpi/CAHp75VfOa5pN4MKT-aQmWBwPGWsOaQupyfrN-weWwfR3vMLtuA@mail.gmail.com/ > > There is actually still a comment in the hisi_lpc driver - I should have > checked there first :) > > So my impression is that the hisi_lpc code is almost the same in > acpi_create_platform_device(), apart from we need do the resource fixup > in hisi_lpc_acpi_set_io_res(). > > So we could factor out by dividing acpi_create_platform_device() into 2x > parts: resource get and then platform dev create. But that does not seem > wise as we have 2x parts which don't make sense on their own. Or else > pass a fixup callback into acpi_create_platform_device(). Any other > ideas if we want to go this way?
I prefer having an ops or so structure where you can pass ->fixup() and/or ->xlate() function, Btw, it looks like the code in hisi_lpc needs a lot of cleanups.
-- With Best Regards, Andy Shevchenko
| |