summaryrefslogtreecommitdiff
path: root/drivers/acpi/power.c
diff options
context:
space:
mode:
authorHans de Goede <hdegoede@redhat.com>2017-04-30 22:54:16 +0200
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>2018-03-24 11:00:20 +0100
commita9b90d80db0e53ce4fe06a1ffe12ac683f7b06c9 (patch)
treefa78bc34eb684d19d3489c2857cd362ea2159df4 /drivers/acpi/power.c
parent2710fb8f0c828656a147870dccc5164c0f0c48ea (diff)
ACPI / power: Delay turning off unused power resources after suspend
[ Upstream commit 8ece1d83346bcc431090d59a2184276192189cdd ] Commit 660b1113e0f3 (ACPI / PM: Fix consistency check for power resources during resume) introduced a check for ACPI power resources which have been turned on by the BIOS during suspend and turns these back off again. This is causing problems on a Dell Venue Pro 11 7130 (i5-4300Y) it causes the following messages to show up in dmesg: [ 131.014605] ACPI: Waking up from system sleep state S3 [ 131.150271] acpi LNXPOWER:07: Turning OFF [ 131.150323] acpi LNXPOWER:06: Turning OFF [ 131.150911] acpi LNXPOWER:00: Turning OFF [ 131.169014] ACPI : EC: interrupt unblocked [ 131.181811] xhci_hcd 0000:00:14.0: System wakeup disabled by ACPI [ 133.535728] pci_raw_set_power_state: 76 callbacks suppressed [ 133.535735] iwlwifi 0000:01:00.0: Refused to change power state, currently in D3 [ 133.597672] PM: noirq resume of devices complete after 2428.891 msecs Followed by a bunch of iwlwifi errors later on and the pcie device dropping from the bus (acpiphp thinks it has been unplugged). Disabling the turning off of unused power resources fixes this. Instead of adding a quirk for this system, this commit fixes this by moving the disabling of unused power resources to later in the resume sequence when the iwlwifi card has been moved out of D3 so the ref_count for its power resource no longer is 0. This new behavior seems to match the intend of the original commit which commit-msg says: "(... which means that no devices are going to need them any time soon) and we should turn them off". This also avoids power resources which we need when bringing devices out of D3 from getting bounced off and then back on again. Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'drivers/acpi/power.c')
-rw-r--r--drivers/acpi/power.c10
1 files changed, 10 insertions, 0 deletions
diff --git a/drivers/acpi/power.c b/drivers/acpi/power.c
index 1c2b846c5776..3a6c9b741b23 100644
--- a/drivers/acpi/power.c
+++ b/drivers/acpi/power.c
@@ -864,6 +864,16 @@ void acpi_resume_power_resources(void)
mutex_unlock(&resource->resource_lock);
}
+
+ mutex_unlock(&power_resource_list_lock);
+}
+
+void acpi_turn_off_unused_power_resources(void)
+{
+ struct acpi_power_resource *resource;
+
+ mutex_lock(&power_resource_list_lock);
+
list_for_each_entry_reverse(resource, &acpi_power_resource_list, list_node) {
int result, state;