Am Freitag, dem 24.02.2023 um 13:08 +0100 schrieb Marek Vasut: > On 2/24/23 10:08, Lucas Stach wrote: > > Hi, > > > > diff --git a/drivers/soc/imx/imx8m-blk-ctrl.c b/drivers/soc/imx/imx8m-blk-ctrl.c > > > index 399cb85105a18..77e7dc4eb8cff 100644 > > > --- a/drivers/soc/imx/imx8m-blk-ctrl.c > > > +++ b/drivers/soc/imx/imx8m-blk-ctrl.c > > > @@ -310,7 +310,7 @@ static int imx8m_blk_ctrl_probe(struct platform_device *pdev) > > > > > > dev_set_drvdata(dev, bc); > > > > > > - return 0; > > > + return devm_of_platform_populate(dev); > > > > You need to handle the return value, not simply pass it through as the > > return value of the probe function. When devm_of_platform_populate > > fails you miss to clean up the genpd provider and detach from the power > > domains. > > Hmmm, but then I cannot use the devm_ variant, can I ? Why not? If everything works okay the devm will just take care of cleaning up the platform devices when the blk-ctrl is removed. If it fails you just need to roll back the non-devm parts of the blk-ctrl probe routine. Regards, Lucas
