Re: [PATCH v5 2/4] ACPI: of: match PRP0001 in of_match_device

[Markus Probst <markus.probst@posteo.de>]

[-- Attachment #1: Type: text/plain, Size: 3606 bytes --]

On Mon, 2026-03-30 at 09:00 +0200, Krzysztof Kozlowski wrote:
> On Sun, Mar 29, 2026 at 08:02:16PM +0200, Markus Probst wrote:
> > Export `acpi_of_match_device` function and use it to match for PRP0001
> > in `of_match_device`, if the device does not have a device node.
> > 
> > This fixes the match data being NULL when using ACPI PRP0001, even though
> > the device was matched against an of device table.
> 
> Fixes tag?
> 
> I don't see how this is going to fix !ACPI case - the
> acpi_of_match_device() will just return false.
While trying to argue I found out that there already is
`device_get_match_data`, which takes PRP0001 into account.

I will now instead make a patch, which will make rust use this function
instead of calling `of_match_device` and `acpi_match_device`
individually, which ignores PRP0001.

There are still a lot of drivers only using `of_match_device`, which
makes it impossible to use PRP0001 with them. But this is not relevant
for this driver.

Thanks
- Markus Probst

> 
> 
> > 
> > Signed-off-by: Markus Probst <markus.probst@posteo.de>
> > ---
> >  drivers/acpi/bus.c   |  7 ++++---
> >  drivers/of/device.c  |  9 +++++++--
> >  include/linux/acpi.h | 11 +++++++++++
> >  3 files changed, 22 insertions(+), 5 deletions(-)
> > 
> > diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c
> > index 2ec095e2009e..cd02f04cf685 100644
> > --- a/drivers/acpi/bus.c
> > +++ b/drivers/acpi/bus.c
> > @@ -831,9 +831,9 @@ const struct acpi_device *acpi_companion_match(const struct device *dev)
> >   * identifiers and a _DSD object with the "compatible" property, use that
> >   * property to match against the given list of identifiers.
> >   */
> > -static bool acpi_of_match_device(const struct acpi_device *adev,
> > -				 const struct of_device_id *of_match_table,
> > -				 const struct of_device_id **of_id)
> > +bool acpi_of_match_device(const struct acpi_device *adev,
> > +			  const struct of_device_id *of_match_table,
> > +			  const struct of_device_id **of_id)
> >  {
> >  	const union acpi_object *of_compatible, *obj;
> >  	int i, nval;
> > @@ -866,6 +866,7 @@ static bool acpi_of_match_device(const struct acpi_device *adev,
> >  
> >  	return false;
> >  }
> > +EXPORT_SYMBOL_GPL(acpi_of_match_device);
> >  
> >  static bool acpi_of_modalias(struct acpi_device *adev,
> >  			     char *modalias, size_t len)
> > diff --git a/drivers/of/device.c b/drivers/of/device.c
> > index f7e75e527667..128682390058 100644
> > --- a/drivers/of/device.c
> > +++ b/drivers/of/device.c
> > @@ -11,6 +11,7 @@
> >  #include <linux/mod_devicetable.h>
> >  #include <linux/slab.h>
> >  #include <linux/platform_device.h>
> > +#include <linux/acpi.h>
> >  
> >  #include <asm/errno.h>
> >  #include "of_private.h"
> > @@ -26,8 +27,12 @@
> >  const struct of_device_id *of_match_device(const struct of_device_id *matches,
> >  					   const struct device *dev)
> >  {
> > -	if (!matches || !dev->of_node || dev->of_node_reused)
> > -		return NULL;
> > +	if (!matches || !dev->of_node || dev->of_node_reused) {
> > +		const struct of_device_id *id = NULL;
> > +
> > +		acpi_of_match_device(ACPI_COMPANION(dev), matches, &id);
> 
> I don't think this should be done from of_match_device. Yuo will have
> soon recursive calls, because acpi_of_match_device() will call other
> match, that will call of_match_device() and so on...
> 
> of_match_device() is supposed to match only against OF. Not ACPI. There
> should be no ACPI header or code in this unit file.
> 
> Best regards,
> Krzysztof

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 870 bytes --]