[PATCH] ACPI: PM: s2idle: Use LPS0 idle if ACPI_FADT_LOW_POWER_S0 is unset

[PATCH] ACPI: PM: s2idle: Use LPS0 idle if ACPI_FADT_LOW_POWER_S0 is unset

[Rafael J. Wysocki]

From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

If the PNP0D80 device is present and its _DSM appears to be valid,
there is no reason to avoid using it even if ACPI_FADT_LOW_POWER_S0
is unset in the FADT, because suspend-to-idle may be the only way to
suspend the system if S3 is not supported by the platform, so do not
return early from lps0_device_attach() in that case.

However, still check ACPI_FADT_LOW_POWER_S0 when deciding whether or
not suspend-to-idle should be the default system suspend method.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
---
 drivers/acpi/x86/s2idle.c |   11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

Index: linux-pm/drivers/acpi/x86/s2idle.c
===================================================================
--- linux-pm.orig/drivers/acpi/x86/s2idle.c
+++ linux-pm/drivers/acpi/x86/s2idle.c
@@ -369,9 +369,6 @@ static int lps0_device_attach(struct acp
 	if (lps0_device_handle)
 		return 0;
 
-	if (!(acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0))
-		return 0;
-
 	if (acpi_s2idle_vendor_amd()) {
 		/* AMD0004, AMD0005, AMDI0005:
 		 * - Should use rev_id 0x0
@@ -421,10 +418,12 @@ static int lps0_device_attach(struct acp
 		lpi_device_get_constraints();
 
 	/*
-	 * Use suspend-to-idle by default if the default suspend mode was not
-	 * set from the command line.
+	 * Use suspend-to-idle by default if ACPI_FADT_LOW_POWER_S0 is set in
+	 * the FADT and the default suspend mode was not set from the command
+	 * line.
 	 */
-	if (mem_sleep_default > PM_SUSPEND_MEM && !acpi_sleep_default_s3)
+	if ((acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0) &&
+	    mem_sleep_default > PM_SUSPEND_MEM && !acpi_sleep_default_s3)
 		mem_sleep_current = PM_SUSPEND_TO_IDLE;
 
 	/*

RE: [PATCH] ACPI: PM: s2idle: Use LPS0 idle if ACPI_FADT_LOW_POWER_S0 is unset

[Limonciello, Mario]

[Public]



> -----Original Message-----
> From: Rafael J. Wysocki <rjw@rjwysocki.net>
> Sent: Wednesday, July 13, 2022 13:51
> To: Linux ACPI <linux-acpi@vger.kernel.org>
> Cc: LKML <linux-kernel@vger.kernel.org>; Linux PM <linux-
> pm@vger.kernel.org>; Srinivas Pandruvada
> <srinivas.pandruvada@linux.intel.com>; Zhang Rui <rui.zhang@intel.com>;
> Limonciello, Mario <Mario.Limonciello@amd.com>
> Subject: [PATCH] ACPI: PM: s2idle: Use LPS0 idle if
> ACPI_FADT_LOW_POWER_S0 is unset
> 
> From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> 
> If the PNP0D80 device is present and its _DSM appears to be valid,
> there is no reason to avoid using it even if ACPI_FADT_LOW_POWER_S0
> is unset in the FADT, because suspend-to-idle may be the only way to
> suspend the system if S3 is not supported by the platform, so do not
> return early from lps0_device_attach() in that case.
> 
> However, still check ACPI_FADT_LOW_POWER_S0 when deciding whether or
> not suspend-to-idle should be the default system suspend method.
> 
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> ---

Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>

>  drivers/acpi/x86/s2idle.c |   11 +++++------
>  1 file changed, 5 insertions(+), 6 deletions(-)
> 
> Index: linux-pm/drivers/acpi/x86/s2idle.c
> ==========================================================
> =========
> --- linux-pm.orig/drivers/acpi/x86/s2idle.c
> +++ linux-pm/drivers/acpi/x86/s2idle.c
> @@ -369,9 +369,6 @@ static int lps0_device_attach(struct acp
>  	if (lps0_device_handle)
>  		return 0;
> 
> -	if (!(acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0))
> -		return 0;
> -
>  	if (acpi_s2idle_vendor_amd()) {
>  		/* AMD0004, AMD0005, AMDI0005:
>  		 * - Should use rev_id 0x0
> @@ -421,10 +418,12 @@ static int lps0_device_attach(struct acp
>  		lpi_device_get_constraints();
> 
>  	/*
> -	 * Use suspend-to-idle by default if the default suspend mode was
> not
> -	 * set from the command line.
> +	 * Use suspend-to-idle by default if ACPI_FADT_LOW_POWER_S0 is
> set in
> +	 * the FADT and the default suspend mode was not set from the
> command
> +	 * line.
>  	 */
> -	if (mem_sleep_default > PM_SUSPEND_MEM &&
> !acpi_sleep_default_s3)
> +	if ((acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0) &&
> +	    mem_sleep_default > PM_SUSPEND_MEM &&
> !acpi_sleep_default_s3)
>  		mem_sleep_current = PM_SUSPEND_TO_IDLE;
> 
>  	/*
> 
>