| Message ID | 20240212061037.4271-3-raag.jadav@intel.com |
|---|---|
| State | Changes Requested |
| Headers | show |
| Series | LPSS PWM cleanups | expand |
On 2/12/24 08:10, Raag Jadav wrote: > We no longer need empty runtime PM handles for PCI devices after commit > c5eb1190074c ("PCI / PM: Allow runtime PM without callback functions"). > Drop them and let PCI core take care of power state transitions. > > Signed-off-by: Raag Jadav <raag.jadav@intel.com> > Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > --- > drivers/pwm/pwm-lpss-pci.c | 22 ---------------------- > 1 file changed, 22 deletions(-) > Will this patch cause a regression if applied without your another patch to the drivers/pci/pci-driver.c? https://lore.kernel.org/linux-pci/20240212063233.5599-1-raag.jadav@intel.com/
On Wed, Feb 14, 2024 at 09:04:09AM +0200, Jarkko Nikula wrote: > On 2/12/24 08:10, Raag Jadav wrote: > > We no longer need empty runtime PM handles for PCI devices after commit > > c5eb1190074c ("PCI / PM: Allow runtime PM without callback functions"). > > Drop them and let PCI core take care of power state transitions. > > > > Signed-off-by: Raag Jadav <raag.jadav@intel.com> > > Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > > --- > > drivers/pwm/pwm-lpss-pci.c | 22 ---------------------- > > 1 file changed, 22 deletions(-) > > > Will this patch cause a regression if applied without your another patch to > the drivers/pci/pci-driver.c? > > https://lore.kernel.org/linux-pci/20240212063233.5599-1-raag.jadav@intel.com/ Yes, I'll send out a v2 with updated commit message after the PCI one is accepted. Raag
diff --git a/drivers/pwm/pwm-lpss-pci.c b/drivers/pwm/pwm-lpss-pci.c index b4134bee2863..51efe37e23fe 100644 --- a/drivers/pwm/pwm-lpss-pci.c +++ b/drivers/pwm/pwm-lpss-pci.c @@ -48,25 +48,6 @@ static void pwm_lpss_remove_pci(struct pci_dev *pdev) pm_runtime_get_sync(&pdev->dev); } -static int pwm_lpss_runtime_suspend_pci(struct device *dev) -{ - /* - * The PCI core will handle transition to D3 automatically. We only - * need to provide runtime PM hooks for that to happen. - */ - return 0; -} - -static int pwm_lpss_runtime_resume_pci(struct device *dev) -{ - return 0; -} - -static DEFINE_RUNTIME_DEV_PM_OPS(pwm_lpss_pci_pm, - pwm_lpss_runtime_suspend_pci, - pwm_lpss_runtime_resume_pci, - NULL); - static const struct pci_device_id pwm_lpss_pci_ids[] = { { PCI_VDEVICE(INTEL, 0x0ac8), (unsigned long)&pwm_lpss_bxt_info}, { PCI_VDEVICE(INTEL, 0x0f08), (unsigned long)&pwm_lpss_byt_info}, @@ -86,9 +67,6 @@ static struct pci_driver pwm_lpss_driver_pci = { .id_table = pwm_lpss_pci_ids, .probe = pwm_lpss_probe_pci, .remove = pwm_lpss_remove_pci, - .driver = { - .pm = pm_ptr(&pwm_lpss_pci_pm), - }, }; module_pci_driver(pwm_lpss_driver_pci);
We no longer need empty runtime PM handles for PCI devices after commit c5eb1190074c ("PCI / PM: Allow runtime PM without callback functions"). Drop them and let PCI core take care of power state transitions. Signed-off-by: Raag Jadav <raag.jadav@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> --- drivers/pwm/pwm-lpss-pci.c | 22 ---------------------- 1 file changed, 22 deletions(-)