Hi, On 22-10-18 17:42, Emil Velikov wrote:
On Mon, 22 Oct 2018 at 15:58, Hans de Goede <hdegoede@xxxxxxxxxx> wrote:Add #ifdef CONFIG_PM_SLEEP around the suspend/hibernate functions. Remove unnecessary #ifdef CONFIG_COMPAT, the .compat_ioctl member is always available and if CONFIG_COMPAT is not set then drm_compat_ioctl is defined to NULL. Signed-off-by: Hans de Goede <hdegoede@xxxxxxxxxx> --- drivers/staging/vboxvideo/vbox_drv.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/staging/vboxvideo/vbox_drv.c b/drivers/staging/vboxvideo/vbox_drv.c index f9f4c6c2a4e9..b6e6530aa4be 100644 --- a/drivers/staging/vboxvideo/vbox_drv.c +++ b/drivers/staging/vboxvideo/vbox_drv.c @@ -117,6 +117,7 @@ static void vbox_pci_remove(struct pci_dev *pdev) drm_dev_put(&vbox->ddev); } +#ifdef CONFIG_PM_SLEEP static int vbox_pm_suspend(struct device *dev)AFAICT Arnd has been annotating these as __maybe_unused
That only works in combinations with a macro like SET_SYSTEM_SLEEP_PM_OPS(), which makes sure that we do not actually set the .suspend, etc. members of dev_pm_ops when CONFIG_PM_SLEEP is not set, so that the code actually gets compiled out. We cannot use SET_SYSTEM_SLEEP_PM_OPS here, since the suspend and hibernate paths differ. Regards, Hans _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
