On Wed 13 May 22:03 PDT 2020, Rajendra Nayak wrote: > > [].. > > > > spi->bus_num = -1; > > > spi->dev.of_node = dev->of_node; > > > @@ -596,6 +607,9 @@ static int spi_geni_probe(struct platform_device *pdev) > > > spi_geni_probe_runtime_disable: > > > pm_runtime_disable(dev); > > > spi_master_put(spi); > > > + if (mas->se.has_opp_table) > > > > Why do you need has_opp_table? > > > > Afaict if dev_pm_opp_of_add_table() returns -ENODEV there's no attached > > opp-table and dev_pm_opp_of_remove_table() is a nop. > > Apparently its not. Calling dev_pm_opp_of_remove_table() when dev_pm_opp_of_add_table() > failed causes use-count mismatch. > You can see https://lkml.org/lkml/2020/4/15/18 for more details. > Must have missed that when I glanced at the code, thanks for clarifying. Regards, Bjorn > -- > QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member > of Code Aurora Forum, hosted by The Linux Foundation
