Hi
On 10/25/18 4:57 PM, Rafael J. Wysocki wrote:
On Tue, Oct 23, 2018 at 1:46 PM Jarkko Nikula
<jarkko.nikula@xxxxxxxxxxxxxxx> wrote:
Commit a9c8088c7988 ("i2c: i801: Don't restore config registers on
runtime PM") nullified the runtime PM suspend/resume callback pointers
while keeping the runtime PM enabled.
This causes that SMBus PCI device stays in D0 power state and sysfs
/sys/bus/pci/devices/[SMBus PCI ID]/power/runtime_status shows "error"
when the runtime PM framework attempts to autosuspend the device. This
is due PCI bus runtime PM which checks for driver runtime PM callbacks
and returns with -ENOSYS if they are not set.
Since i2c-i801.c don't need to do anything device specific beyond PCI
device power state management Jean Delvare proposed if this can be fixed
in the PCI subsystem core level rather than adding dummy runtime PM
callback functions in the PCI drivers.
Change the pci_pm_runtime_suspend()/pci_pm_runtime_resume() semantics so
that they allow change the PCI device power state during runtime PM
transitions even if no runtime PM callback functions are defined.
This change fixes the runtime PM regression on i2c-i801.c.
It is not obvious why the code had hard requirements for the runtime PM
callbacks. Test has been here since the code was introduced by the
commit 6cbf82148ff2 ("PCI PM: Run-time callbacks for PCI bus type").
On the other hand similar change than this was done to generic runtime
PM callbacks way back in the commit 05aa55dddb9e ("PM / Runtime: Lenient
generic runtime pm callbacks").
Fixes: a9c8088c7988 ("i2c: i801: Don't restore config registers on runtime PM")
Reported-by: Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx> # 4.18+
Signed-off-by: Jarkko Nikula <jarkko.nikula@xxxxxxxxxxxxxxx>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
I guess this patch with Reviewed-by tags from Rafael and Jean got buried
under other list traffic as I don't find this from pci.git or linux-next?
--
Jarkko
