| Message ID | 20241225-scmi-fwdevlink-v1-3-e9a3a5341362@nxp.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | scmi: Bypass set fwnode to address devlink issue | expand |
On Wed, Dec 25, 2024 at 04:20:46PM +0800, Peng Fan (OSS) wrote: > From: Peng Fan <peng.fan@nxp.com> > > For the platform that not compatible with scmi pinctrl device, the > fwnode will not be set, so checking fwnode will make code simpler > and easy to maintain. > > Signed-off-by: Peng Fan <peng.fan@nxp.com> > --- > drivers/pinctrl/pinctrl-scmi.c | 7 +------ > 1 file changed, 1 insertion(+), 6 deletions(-) > > diff --git a/drivers/pinctrl/pinctrl-scmi.c b/drivers/pinctrl/pinctrl-scmi.c > index df4bbcd7d1d59ac2c8ddc320dc10d702ad1ed5b2..aade6df77dbb2c391741e77c0aac3f029991e4bb 100644 > --- a/drivers/pinctrl/pinctrl-scmi.c > +++ b/drivers/pinctrl/pinctrl-scmi.c > @@ -505,11 +505,6 @@ static int pinctrl_scmi_get_pins(struct scmi_pinctrl *pmx, > return 0; > } > > -static const char * const scmi_pinctrl_blocklist[] = { > - "fsl,imx95", > - NULL > -}; > - > static int scmi_pinctrl_probe(struct scmi_device *sdev) > { > int ret; > @@ -521,7 +516,7 @@ static int scmi_pinctrl_probe(struct scmi_device *sdev) > if (!sdev->handle) > return -EINVAL; > > - if (of_machine_compatible_match(scmi_pinctrl_blocklist)) > + if (!dev->fwnode) I would prefer to see the blocklist to be explicit here rather than implicitly hiding it away with this change set.
On Fri, Dec 27, 2024 at 03:30:20PM +0000, Sudeep Holla wrote: > On Wed, Dec 25, 2024 at 04:20:46PM +0800, Peng Fan (OSS) wrote: > > From: Peng Fan <peng.fan@nxp.com> > > > > For the platform that not compatible with scmi pinctrl device, the > > fwnode will not be set, so checking fwnode will make code simpler > > and easy to maintain. > > > > Signed-off-by: Peng Fan <peng.fan@nxp.com> > > --- > > drivers/pinctrl/pinctrl-scmi.c | 7 +------ > > 1 file changed, 1 insertion(+), 6 deletions(-) > > > > diff --git a/drivers/pinctrl/pinctrl-scmi.c b/drivers/pinctrl/pinctrl-scmi.c > > index df4bbcd7d1d59ac2c8ddc320dc10d702ad1ed5b2..aade6df77dbb2c391741e77c0aac3f029991e4bb 100644 > > --- a/drivers/pinctrl/pinctrl-scmi.c > > +++ b/drivers/pinctrl/pinctrl-scmi.c > > @@ -505,11 +505,6 @@ static int pinctrl_scmi_get_pins(struct scmi_pinctrl *pmx, > > return 0; > > } > > > > -static const char * const scmi_pinctrl_blocklist[] = { > > - "fsl,imx95", > > - NULL > > -}; > > - > > static int scmi_pinctrl_probe(struct scmi_device *sdev) > > { > > int ret; > > @@ -521,7 +516,7 @@ static int scmi_pinctrl_probe(struct scmi_device *sdev) > > if (!sdev->handle) > > return -EINVAL; > > > > - if (of_machine_compatible_match(scmi_pinctrl_blocklist)) > > + if (!dev->fwnode) > > I would prefer to see the blocklist to be explicit here rather than > implicitly hiding it away with this change set. Using a flag to inhibit device_link_add as said early in the series this could be dropped and kept as is, I suppose. Thanks, Cristian
> Subject: Re: [PATCH 3/4] pinctrl: scmi: Check fwnode instead of > machine compatible > > On Fri, Dec 27, 2024 at 03:30:20PM +0000, Sudeep Holla wrote: > > On Wed, Dec 25, 2024 at 04:20:46PM +0800, Peng Fan (OSS) wrote: > > > From: Peng Fan <peng.fan@nxp.com> > > > > > > For the platform that not compatible with scmi pinctrl device, the > > > fwnode will not be set, so checking fwnode will make code simpler > > > and easy to maintain. > > > > > > Signed-off-by: Peng Fan <peng.fan@nxp.com> > > > --- > > > drivers/pinctrl/pinctrl-scmi.c | 7 +------ > > > 1 file changed, 1 insertion(+), 6 deletions(-) > > > > > > diff --git a/drivers/pinctrl/pinctrl-scmi.c > > > b/drivers/pinctrl/pinctrl-scmi.c index > > > > df4bbcd7d1d59ac2c8ddc320dc10d702ad1ed5b2..aade6df77dbb2c391 > 741e77c0a > > > ac3f029991e4bb 100644 > > > --- a/drivers/pinctrl/pinctrl-scmi.c > > > +++ b/drivers/pinctrl/pinctrl-scmi.c > > > @@ -505,11 +505,6 @@ static int pinctrl_scmi_get_pins(struct > scmi_pinctrl *pmx, > > > return 0; > > > } > > > > > > -static const char * const scmi_pinctrl_blocklist[] = { > > > - "fsl,imx95", > > > - NULL > > > -}; > > > - > > > static int scmi_pinctrl_probe(struct scmi_device *sdev) { > > > int ret; > > > @@ -521,7 +516,7 @@ static int scmi_pinctrl_probe(struct > scmi_device *sdev) > > > if (!sdev->handle) > > > return -EINVAL; > > > > > > - if (of_machine_compatible_match(scmi_pinctrl_blocklist)) > > > + if (!dev->fwnode) > > > > I would prefer to see the blocklist to be explicit here rather than > > implicitly hiding it away with this change set. > > Using a flag to inhibit device_link_add as said early in the series this > could be dropped and kept as is, I suppose. Since the list will expand to include more i.MX9[X] chips, I was thinking to drop the list to avoid update the list every time we add a new SoC. We may need to use a machine saying "fsl,imx9-sm" or else. I will give a check on a flag. Thanks for suggestion. Thanks, Peng. > > Thanks, > Cristian
diff --git a/drivers/pinctrl/pinctrl-scmi.c b/drivers/pinctrl/pinctrl-scmi.c index df4bbcd7d1d59ac2c8ddc320dc10d702ad1ed5b2..aade6df77dbb2c391741e77c0aac3f029991e4bb 100644 --- a/drivers/pinctrl/pinctrl-scmi.c +++ b/drivers/pinctrl/pinctrl-scmi.c @@ -505,11 +505,6 @@ static int pinctrl_scmi_get_pins(struct scmi_pinctrl *pmx, return 0; } -static const char * const scmi_pinctrl_blocklist[] = { - "fsl,imx95", - NULL -}; - static int scmi_pinctrl_probe(struct scmi_device *sdev) { int ret; @@ -521,7 +516,7 @@ static int scmi_pinctrl_probe(struct scmi_device *sdev) if (!sdev->handle) return -EINVAL; - if (of_machine_compatible_match(scmi_pinctrl_blocklist)) + if (!dev->fwnode) return -ENODEV; handle = sdev->handle;