[0/3] PM / wakeirq: fix wake irq arming

Message

Johan Hovold July 13, 2023, 2:57 p.m. UTC

When reviewing the Qualcomm serial-driver suspend implementation I
noticed the odd runtime PM state update which had snuck in. Turns out it
was added to work around a bug in PM core which prevented drivers not
implementing runtime PM from using dedicated wake irqs.

This series fixes the wake irq arming and drops the unused wake irq
enable helpers before dropping the bogus runtime PM state update in the
Qualcomm driver.

I suggest that Rafael takes all of these through his tree.

Johan


Johan Hovold (3):
  PM / wakeirq: fix wake irq arming
  PM / wakeirq: drop unused enable helpers
  serial: qcom-geni: drop bogus runtime pm state update

 drivers/base/power/power.h            |  1 +
 drivers/base/power/wakeirq.c          | 61 ++++-----------------------
 drivers/tty/serial/qcom_geni_serial.c |  7 ---
 include/linux/pm_wakeirq.h            | 10 -----
 4 files changed, 9 insertions(+), 70 deletions(-)

Comments

Rafael J. Wysocki July 20, 2023, 5:49 p.m. UTC | #1

On Thu, Jul 13, 2023 at 5:01 PM Johan Hovold <[e-mail]> wrote:
>
> When reviewing the Qualcomm serial-driver suspend implementation I
> noticed the odd runtime PM state update which had snuck in. Turns out it
> was added to work around a bug in PM core which prevented drivers not
> implementing runtime PM from using dedicated wake irqs.
>
> This series fixes the wake irq arming and drops the unused wake irq
> enable helpers before dropping the bogus runtime PM state update in the
> Qualcomm driver.
>
> I suggest that Rafael takes all of these through his tree.
>
> Johan
>
>
> Johan Hovold (3):
>   PM / wakeirq: fix wake irq arming
>   PM / wakeirq: drop unused enable helpers
>   serial: qcom-geni: drop bogus runtime pm state update

All applied and I'm inclined to push them as fixed for 6.5-rc, thanks!

Greg Kroah-Hartman Aug. 1, 2023, 5:48 a.m. UTC | #2

On Thu, Jul 13, 2023 at 04:57:38PM +0200, Johan Hovold wrote:
> When reviewing the Qualcomm serial-driver suspend implementation I
> noticed the odd runtime PM state update which had snuck in. Turns out it
> was added to work around a bug in PM core which prevented drivers not
> implementing runtime PM from using dedicated wake irqs.
> 
> This series fixes the wake irq arming and drops the unused wake irq
> enable helpers before dropping the bogus runtime PM state update in the
> Qualcomm driver.
> 
> I suggest that Rafael takes all of these through his tree.

I agree:

Acked-by: Greg Kroah-Hartman <[e-mail]>