Re: Linux 5.6-rc2

From: Rafael J. Wysocki
Date: Thu Feb 20 2020 - 19:46:23 EST


On Thursday, February 20, 2020 11:41:22 PM CET Rafael J. Wysocki wrote:
> On Monday, February 17, 2020 10:29:35 PM CET Chris Wilson wrote:
> > Quoting Linus Torvalds (2020-02-17 21:20:27)
> > > On Mon, Feb 17, 2020 at 8:22 AM Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> wrote:
> > > >
> > > > Quoting Linus Torvalds (2020-02-16 21:32:32)
> > > > > Rafael J. Wysocki (4):
> > > > > ACPI: EC: Fix flushing of pending work
> > > > > ACPI: PM: s2idle: Avoid possible race related to the EC GPE
> > > > > ACPICA: Introduce acpi_any_gpe_status_set()
> > > > > ACPI: PM: s2idle: Prevent spurious SCIs from waking up the system
> > > >
> > > > Our S0 testing broke on all platforms, so we've reverted
> > > > e3728b50cd9b ("ACPI: PM: s2idle: Avoid possible race related to the EC GPE")
> > > > fdde0ff8590b ("ACPI: PM: s2idle: Prevent spurious SCIs from waking up the system")
> > > >
> > > > There wasn't much in the logs, for example,
> > > > https://intel-gfx-ci.01.org/tree/drm-tip/IGT_5445/fi-kbl-7500u/igt@gem_exec_suspend@basic-s0.html
> > >
> > > So the machine suspends, but never comes back?
> > >
> > > Do you need to revert both for it to work for you? Or is the revert of
> > > fdde0ff8590b just to avoid the conflict?
> >
> > fdde0ff85 was just to avoid conflicts.
> >
> > > I'm assuming you bisected this, and the bisect indicated e3728b50cd9b,
> > > and then to revert it you reverted the other commit too..
> >
> > Lucky guess based on diff rc1..rc2. Bisect was going to be painful, but
> > could be done if this is not enough clue for Rafael.
>
> Sorry for the delayed response, was away.
>
> I'm guessing that you are using rtcwake for wakeup, in which case reverting
> fdde0ff85 alone should unbreak it.
>
> Can you please double check that?

And below is a patch that should fix it if I'm not mistaken (verified on my
system where I was able to reproduce the issue), so it would suffice to test
this one on top of the -rc2.

---
From: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
Subject: [PATCH] ACPI: PM: s2idle: Check fixed wakeup events in acpi_s2idle_wake()

Commit fdde0ff8590b ("ACPI: PM: s2idle: Prevent spurious SCIs from
waking up the system") overlooked the fact that fixed events can wake
up the system too and broke RTC wakeup from suspend-to-idle as a
result.

Fix this issue by checking the fixed events in acpi_s2idle_wake() in
addition to checking wakeup GPEs and break out of the suspend-to-idle
loop if the status bits of any enabled fixed events are set then.

Fixes: fdde0ff8590b ("ACPI: PM: s2idle: Prevent spurious SCIs from waking up the system")
Reported-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Cc: 5.4+ <stable@xxxxxxxxxxxxxxx> # 5.4+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
---
drivers/acpi/acpica/evevent.c | 45 ++++++++++++++++++++++++++++++++++++++++++
drivers/acpi/sleep.c | 7 ++++++
include/acpi/acpixf.h | 1
3 files changed, 53 insertions(+)

Index: linux-pm/drivers/acpi/acpica/evevent.c
===================================================================
--- linux-pm.orig/drivers/acpi/acpica/evevent.c
+++ linux-pm/drivers/acpi/acpica/evevent.c
@@ -265,4 +265,49 @@ static u32 acpi_ev_fixed_event_dispatch(
handler) (acpi_gbl_fixed_event_handlers[event].context));
}

+/*******************************************************************************
+ *
+ * FUNCTION: acpi_any_fixed_event_status_set
+ *
+ * PARAMETERS: None
+ *
+ * RETURN: TRUE or FALSE
+ *
+ * DESCRIPTION: Checks the PM status register for active fixed events
+ *
+ ******************************************************************************/
+
+u32 acpi_any_fixed_event_status_set(void)
+{
+ acpi_status status;
+ u32 in_status;
+ u32 in_enable;
+ u32 i;
+
+ status = acpi_hw_register_read(ACPI_REGISTER_PM1_ENABLE, &in_enable);
+ if (ACPI_FAILURE(status)) {
+ return (FALSE);
+ }
+
+ status = acpi_hw_register_read(ACPI_REGISTER_PM1_STATUS, &in_status);
+ if (ACPI_FAILURE(status)) {
+ return (FALSE);
+ }
+
+ /*
+ * Check for all possible Fixed Events and dispatch those that are active
+ */
+ for (i = 0; i < ACPI_NUM_FIXED_EVENTS; i++) {
+
+ /* Both the status and enable bits must be on for this event */
+
+ if ((in_status & acpi_gbl_fixed_event_info[i].status_bit_mask) &&
+ (in_enable & acpi_gbl_fixed_event_info[i].enable_bit_mask)) {
+ return (TRUE);
+ }
+ }
+
+ return (FALSE);
+}
+
#endif /* !ACPI_REDUCED_HARDWARE */
Index: linux-pm/include/acpi/acpixf.h
===================================================================
--- linux-pm.orig/include/acpi/acpixf.h
+++ linux-pm/include/acpi/acpixf.h
@@ -753,6 +753,7 @@ ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_sta
ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status acpi_enable_all_runtime_gpes(void))
ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status acpi_enable_all_wakeup_gpes(void))
ACPI_HW_DEPENDENT_RETURN_UINT32(u32 acpi_any_gpe_status_set(void))
+ACPI_HW_DEPENDENT_RETURN_UINT32(u32 acpi_any_fixed_event_status_set(void))

ACPI_HW_DEPENDENT_RETURN_STATUS(acpi_status
acpi_get_gpe_device(u32 gpe_index,
Index: linux-pm/drivers/acpi/sleep.c
===================================================================
--- linux-pm.orig/drivers/acpi/sleep.c
+++ linux-pm/drivers/acpi/sleep.c
@@ -1006,6 +1006,13 @@ static bool acpi_s2idle_wake(void)
return true;

/*
+ * If the status bit of any enabled fixed event is set, the
+ * wakeup is regarded as valid.
+ */
+ if (acpi_any_fixed_event_status_set())
+ return true;
+
+ /*
* If there are no EC events to process and at least one of the
* other enabled GPEs is active, the wakeup is regarded as a
* genuine one.