Hi, Seems I wrongly squashed the commit message.
On 27.10.25 07:53, Ahmad Fatoum wrote: have_genpd_providers is a pre-deep-probe artifact, where SoC support would promise that a power domain driver will be registered. In a deep probe system, we can just force the probe in __genpd_dev_pm_attach. Therefore, move the check for have_genpd_providers, until after the lookup of power domain drivers. This will be further refined in follow-up commits. > Signed-off-by: Ahmad Fatoum <[email protected]> > ---> drivers/base/power.c | 12 ++++++++---- > 1 file changed, 8 insertions(+), 4 deletions(-) > > diff --git a/drivers/base/power.c b/drivers/base/power.c > index d60a00f5735e..0cdece210d94 100644 > --- a/drivers/base/power.c > +++ b/drivers/base/power.c > @@ -371,9 +371,6 @@ static int __genpd_dev_pm_attach(struct device *dev, > struct generic_pm_domain *pd; > int ret; > > - if (!have_genpd_providers) > - return 0; > - > ret = of_parse_phandle_with_args(dev->of_node, "power-domains", > "#power-domain-cells", index, &pd_args); > if (ret < 0) > @@ -384,11 +381,18 @@ static int __genpd_dev_pm_attach(struct device *dev, > ret = PTR_ERR(pd); > dev_dbg(dev, "%s() failed to find PM domain: %d\n", > __func__, ret); > + > + if (ret == -ENOENT) > + ret = -EPROBE_DEFER; > + > + if (!have_genpd_providers && ret == -EPROBE_DEFER) > + return 0; > + > /* > * Assume that missing genpds are unresolved > * dependency are report them as deferred > */ > - return (ret == -ENOENT) ? -EPROBE_DEFER : ret; > + return ret; > } > > dev_dbg(dev, "adding to PM domain %s\n", pd ? pd->name : "dummy"); -- Pengutronix e.K. | | Steuerwalder Str. 21 | http://www.pengutronix.de/ | 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
