* Tony Lindgren <tony@xxxxxxxxxxx> [210503 11:16]:
> ..the use of pm_runtime_put_sync() like you suggested. I did a quick
> test with the minimal change below and that works :) Seems like that's
> probably the best minimal fix for the -rc cycle.
Sorry I was mistaken, the patch below won't help for the omapdrm
PM runtime state on suspend.
I had patch "bus: ti-sysc: Fix am335x resume hang for usb otg module"
applied, and that changes ti-sysc to get rid of the PM runtime calls
during system suspend. The side effect is ti-sysc now ignores the module
PM runtime state on suspend. This is pretty much what _od_suspend_noirq()
was also doing.
I think we still fix the dispc related issue too, otherwise the parent
child_count will just keep increasing on each suspend. I check that
again though.
Regards,
Tony
> 8< ----------------
> diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c
> --- a/drivers/gpu/drm/omapdrm/dss/dispc.c
> +++ b/drivers/gpu/drm/omapdrm/dss/dispc.c
> @@ -664,7 +664,7 @@ void dispc_runtime_put(struct dispc_device *dispc)
>
> DSSDBG("dispc_runtime_put\n");
>
> - r = pm_runtime_put_sync(&dispc->pdev->dev);
> + r = pm_runtime_put(&dispc->pdev->dev);
> WARN_ON(r < 0 && r != -ENOSYS);
> }
>
