| Age | Commit message (Collapse) | Author |
|---|
|
Move this API to the canonical timer_*() namespace.
[ tglx: Redone against pre rc1 ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/aB2X0jCKQO56WdMt@gmail.com
|
|
Commit 0cbef962ce1f ("PM: sleep: Resume children after resuming the
parent") introduced a subtle concurrency issue that may lead to a kernel
crash if system suspend is aborted and may also slow down asynchronous
device resume otherwise.
Namely, the initial list walks in dpm_noirq_resume_devices(),
dpm_resume_early(), and dpm_resume() call dpm_clear_async_state() for
every device and attempt to asynchronously resume it if it has no
children (so it is a "root" device). The asynchronous resume of a
root device triggers an attempt to asynchronously resume its children
which may take place before calling dpm_clear_async_state() for them
due to the lack of synchronization between dpm_async_resume_children()
and the code calling dpm_clear_async_state(). If this happens, the
dpm_clear_async_state() that comes in late, will clear
power.work_in_progress for the given device after it has been set by
__dpm_async(), so the suspend callback will be allowed to run once
again for the same device during the same transition. This leads to
a whole range of interesting breakage.
Fortunately, if the suspend transition is not aborted, power.work_in_progress
is set by it for all devices, so dpm_async_resume_children() will not
schedule asynchronous resume for them until dpm_clear_async_state()
clears that flag, but this means missing an opportunity to start the
resume of those devices earlier.
Address the above issue by adding dpm_list_mtx locking to
dpm_async_resume_children(), so it will wait for the entire initial
list walk and the invocation of dpm_clear_async_state() for all devices
to be completed before scheduling any new asynchronous resume callbacks.
Fixes: 0cbef962ce1f ("PM: sleep: Resume children after resuming the parent")
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/4280
Reported-and-tested-by: Chris Bainbridge <chris.bainbridge@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/13779172.uLZWGnKmhe@rjwysocki.net
|
|
Commit 03f1444016b7 ("PM: sleep: Fix handling devices with direct_complete
set on errors") caused power.is_suspended to be set for devices with
power.direct_complete set, but it forgot to ensure the clearing of that
flag for them in device_resume(), so power.is_suspended is still set for
them during the next system suspend-resume cycle.
If that cycle is aborted in dpm_suspend(), the subsequent invocation of
dpm_resume() will trigger a device_resume() call for every device and
because power.is_suspended is set for the devices in question, they will
not be skipped by device_resume() as expected which causes scary error
messages to be logged (as appropriate).
To address this issue, move the clearing of power.is_suspended in
device_resume() immediately after the power.is_suspended check so it
will be always cleared for all devices processed by that function.
Fixes: 03f1444016b7 ("PM: sleep: Fix handling devices with direct_complete set on errors")
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4280
Reported-and-tested-by: Chris Bainbridge <chris.bainbridge@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/4990586.GXAFRqVoOG@rjwysocki.net
|
|
Commits aa7a9275ab81 ("PM: sleep: Suspend async parents after suspending
children") and 443046d1ad66 ("PM: sleep: Make suspend of devices more
asynchronous") added list splicing to the error paths of dpm_suspend(),
dpm_suspend_late(), and dpm_noirq_suspend_devices(), but they should
have used the list_splice_init() variant because the emptied list is
used going forward in all of these cases.
Replace list_splice() with list_splice_init() in the code in question as
appropriate.
Fixes: aa7a9275ab81 ("PM: sleep: Suspend async parents after suspending children")
Fixes: 443046d1ad66 ("PM: sleep: Make suspend of devices more asynchronous")
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/4280
Reported-and-tested-by: Chris Bainbridge <chris.bainbridge@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://patch.msgid.link/4659282.LvFx2qVVIh@rjwysocki.net
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"Once again, the changes are dominated by cpufreq updates, but this
time the majority of them are cpufreq core changes, mostly related to
the introduction of policy locking guards and __free() usage, and
fixes related to boost handling.
Still, there is also a significant update of the intel_pstate driver
making it register an energy model when running on a hybrid platform
which is used for enabling energy-aware scheduling (EAS) if the driver
operates in the passive mode (and schedutil is used as the cpufreq
governor for all CPUs which is the passive mode default).
There are some amd-pstate driver updates too, for a good measure,
including the "Requested CPU Min frequency" BIOS option support and
new online/offline callbacks.
In the cpuidle space, the most significant change is the addition of a
C1 demotion on/off sysfs knob to intel_idle which should help some
users to configure their systems more precisely. There is also the
conversion of the PSCI cpuidle driver to a faux device one and there
are two small updates of cpuidle governors.
Device power management is also modified quite a bit, especially the
handling of devices with asynchronous suspend and resume enabled
during system transitions. They are now going to be handled more
asynchronously during suspend transitions and somewhat less
aggressively during resume transitions.
Apart from the above, the operating performance points (OPP) library
is now going to use mutex locking guards and scope-based cleanup
helpers and there is the usual bunch of assorted fixes and code
cleanups.
Specifics:
- Fix potential division-by-zero error in em_compute_costs() (Yaxiong
Tian)
- Fix typos in energy model documentation and example driver code
(Moon Hee Lee, Atul Kumar Pant)
- Rearrange the energy model management code and add a new function
for adjusting a CPU energy model after adjusting the capacity of
the given CPU to it (Rafael Wysocki)
- Refactor cpufreq_online(), add and use cpufreq policy locking
guards, use __free() in policy reference counting, and clean up
core cpufreq code on top of that (Rafael Wysocki)
- Fix boost handling on CPU suspend/resume and sysfs updates (Viresh
Kumar)
- Fix des_perf clamping with max_perf in amd_pstate_update()
(Dhananjay Ugwekar)
- Add offline, online and suspend callbacks to the amd-pstate driver,
rename and use the existing amd_pstate_epp callbacks in it
(Dhananjay Ugwekar)
- Add support for the "Requested CPU Min frequency" BIOS option to
the amd-pstate driver (Dhananjay Ugwekar)
- Reset amd-pstate driver mode after running selftests (Swapnil
Sapkal)
- Avoid shadowing ret in amd_pstate_ut_check_driver() (Nathan
Chancellor)
- Add helper for governor checks to the schedutil cpufreq governor
and move cpufreq-specific EAS checks to cpufreq (Rafael Wysocki)
- Populate the cpu_capacity sysfs entries from the intel_pstate
driver after registering asym capacity support (Ricardo Neri)
- Add support for enabling Energy-aware scheduling (EAS) to the
intel_pstate driver when operating in the passive mode on a hybrid
platform (Rafael Wysocki)
- Drop redundant cpus_read_lock() from store_local_boost() in the
cpufreq core (Seyediman Seyedarab)
- Replace sscanf() with kstrtouint() in the cpufreq code and use a
symbol instead of a raw number in it (Bowen Yu)
- Add support for autonomous CPU performance state selection to the
CPPC cpufreq driver (Lifeng Zheng)
- OPP: Add dev_pm_opp_set_level() (Praveen Talari)
- Introduce scope-based cleanup headers and mutex locking guards in
OPP core (Viresh Kumar)
- Switch OPP to use kmemdup_array() (Zhang Enpei)
- Optimize bucket assignment when next_timer_ns equals KTIME_MAX in
the menu cpuidle governor (Zhongqiu Han)
- Convert the cpuidle PSCI driver to a faux device one (Sudeep Holla)
- Add C1 demotion on/off sysfs knob to the intel_idle driver (Artem
Bityutskiy)
- Fix typos in two comments in the teo cpuidle governor (Atul Kumar
Pant)
- Fix denying of auto suspend in pm_suspend_timer_fn() (Charan Teja
Kalla)
- Move debug runtime PM attributes to runtime_attrs[] (Rafael
Wysocki)
- Add new devm_ functions for enabling runtime PM and runtime PM
reference counting (Bence Csókás)
- Remove size arguments from strscpy() calls in the hibernation core
code (Thorsten Blum)
- Adjust the handling of devices with asynchronous suspend enabled
during system suspend and resume to start resuming them immediately
after resuming their parents and to start suspending such a device
immediately after suspending its first child (Rafael Wysocki)
- Adjust messages printed during tasks freezing to avoid using
pr_cont() (Andrew Sayers, Paul Menzel)
- Clean up unnecessary usage of !! in pm_print_times_init() (Zihuan
Zhang)
- Add missing wakeup source attribute relax_count to sysfs and remove
the space character at the end ofi the string produced by
pm_show_wakelocks() (Zijun Hu)
- Add configurable pm_test delay for hibernation (Zihuan Zhang)
- Disable asynchronous suspend in ucsi_ccg_probe() to prevent the
cypd4226 device on Tegra boards from suspending prematurely (Jon
Hunter)
- Unbreak printing PM debug messages during hibernation and clean up
some related code (Rafael Wysocki)
- Add a systemd service to run cpupower and change cpupower binding's
Makefile to use -lcpupower (John B. Wyatt IV, Francesco Poli)"
* tag 'pm-6.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (72 commits)
cpufreq: CPPC: Add support for autonomous selection
cpufreq: Update sscanf() to kstrtouint()
cpufreq: Replace magic number
OPP: switch to use kmemdup_array()
PM: freezer: Rewrite restarting tasks log to remove stray *done.*
PM: runtime: fix denying of auto suspend in pm_suspend_timer_fn()
cpufreq: drop redundant cpus_read_lock() from store_local_boost()
cpupower: do not install files to /etc/default/
cpupower: do not call systemctl at install time
cpupower: do not write DESTDIR to cpupower.service
PM: sleep: Introduce pm_sleep_transition_in_progress()
cpufreq/amd-pstate: Avoid shadowing ret in amd_pstate_ut_check_driver()
cpufreq: intel_pstate: Document hybrid processor support
cpufreq: intel_pstate: EAS: Increase cost for CPUs using L3 cache
cpufreq: intel_pstate: EAS support for hybrid platforms
PM: EM: Introduce em_adjust_cpu_capacity()
PM: EM: Move CPU capacity check to em_adjust_new_capacity()
PM: EM: Documentation: Fix typos in example driver code
cpufreq: Drop policy locking from cpufreq_policy_is_good_for_eas()
PM: sleep: Introduce pm_suspend_in_progress()
...
|
|
Move this API to the canonical timer_*() namespace.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250507175338.672442-10-mingo@kernel.org
|
|
In analogy with previous changes, make device_suspend_late() and
device_suspend_noirq() start the async suspend of the device's parent
after the device itself has been processed and make dpm_suspend_late()
and dpm_noirq_suspend_devices() start processing "async" leaf devices
(that is, devices without children) upfront so they don't need to wait
for the other devices they don't depend on.
This change reduces the total duration of device suspend on some systems
measurably, but not significantly.
Suggested-by: Saravana Kannan <saravanak@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/1924195.CQOukoFCf9@rjwysocki.net
|
|
In analogy with the previous change affecting the resume path,
make device_suspend() start the async suspend of the device's parent
after the device itself has been processed and make dpm_suspend() start
processing "async" leaf devices (that is, devices without children)
upfront so they don't need to wait for the "sync" devices they don't
depend on.
On the Dell XPS13 9360 in my office, this change reduces the total
duration of device suspend by approximately 100 ms (over 20%).
Suggested-by: Saravana Kannan <saravanak@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/3541233.QJadu78ljV@rjwysocki.net
|
|
According to [1], the handling of device suspend and resume, and
particularly the latter, involves unnecessary overhead related to
starting new async work items for devices that cannot make progress
right away because they have to wait for other devices.
To reduce this problem in the resume path, use the observation that
starting the async resume of the children of a device after resuming
the parent is likely to produce less scheduling and memory management
noise than starting it upfront while at the same time it should not
increase the resume duration substantially.
Accordingly, modify the code to start the async resume of the device's
children when the processing of the parent has been completed in each
stage of device resume and only start async resume upfront for devices
without parents.
Also make it check if a given device can be resumed asynchronously
before starting the synchronous resume of it in case it will have to
wait for another that is already resuming asynchronously.
In addition to making the async resume of devices more friendly to
systems with relatively less computing resources, this change is also
preliminary for analogous changes in the suspend path.
On the systems where it has been tested, this change by itself does
not affect the overall system resume duration in a measurable way.
Link: https://lore.kernel.org/linux-pm/20241114220921.2529905-1-saravanak@google.com/ [1]
Suggested-by: Saravana Kannan <saravanak@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/22630663.EfDdHjke4D@rjwysocki.net
|
|
timer_delete[_sync]() replaces del_timer[_sync](). Convert the whole tree
over and remove the historical wrapper inlines.
Conversion was done with coccinelle plus manual fixups where necessary.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
|
|
The mask operation link->flags | DL_FLAG_PM_RUNTIME is always true which
is incorrect. The mask operation should be using the bit-wise &
operator. Fix this.
Fixes: bca84a7b93fd ("PM: sleep: Use DPM_FLAG_SMART_SUSPEND conditionally")
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://patch.msgid.link/20250319114324.791829-1-colin.i.king@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
When dpm_suspend() fails, some devices with power.direct_complete set
may not have been handled by device_suspend() yet, so runtime PM has
not been disabled for them yet even though power.direct_complete is set.
Since device_resume() expects that runtime PM has been disabled for all
devices with power.direct_complete set, it will attempt to reenable
runtime PM for the devices that have not been processed by device_suspend()
which does not make sense. Had those devices had runtime PM disabled
before device_suspend() had run, device_resume() would have inadvertently
enable runtime PM for them, but this is not expected to happen because
it would require ->prepare() callbacks to return positive values for
devices with runtime PM disabled, which would be invalid.
In practice, this issue is most likely benign because pm_runtime_enable()
will not allow the "disable depth" counter to underflow, but it causes a
warning message to be printed for each affected device.
To allow device_resume() to distinguish the "direct complete" devices
that have been processed by device_suspend() from those which have not
been handled by it, make device_suspend() set power.is_suspended for
"direct complete" devices.
Next, move the power.is_suspended check in device_resume() before the
power.direct_complete check in it to make it skip the "direct complete"
devices that have not been handled by device_suspend().
This change is based on a preliminary patch from Saravana Kannan.
Fixes: aae4518b3124 ("PM / sleep: Mechanism to avoid resuming runtime-suspended devices unnecessarily")
Link: https://lore.kernel.org/linux-pm/20241114220921.2529905-2-saravanak@google.com/
Reported-by: Saravana Kannan <saravanak@google.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Saravana Kannan <saravanak@google.com>
Link: https://patch.msgid.link/12627587.O9o76ZdvQC@rjwysocki.net
|
|
The body of dpm_wait_for_children() is indented by 7 spaces instead of a
single TAB.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://patch.msgid.link/9c8ff2b103c3ba7b0d27bdc8248b05e3b1dc9551.1741776430.git.geert+renesas@glider.be
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
In preparation for subsequent changes, move the power.completion
reinitialization along with clearing power.work_in_progress into a
separate function called dpm_clear_async_state() and rearrange
dpm_async_fn() to get rid of unnecessary indentation.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/8494650.T7Z3S40VBb@rjwysocki.net
|
|
Rename the async_in_progress field in struct dev_pm_info to
work_in_progress as after subsequent changes it will mean work in
general rather than just async work.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/3338693.aeNJFYEL58@rjwysocki.net
|
|
Modify pm_runtime_block_if_disabled() to return true when runtime PM
is disabled for the device, regardless of the power.last_status value.
This effectively prevents "smart suspend" from being enabled for
devices with runtime PM disabled in device_prepare(), even transiently,
so update the related comment in that function accordingly.
If a device has runtime PM disabled in device_prepare(), it is not
actually known whether or not runtime PM will be enabled for that
device going forward, so it is more appropriate to postpone the
"smart suspend" optimization for the device in the given system
suspend-resume cycle than to enable it and get confused going
forward.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/13718674.uLZWGnKmhe@rjwysocki.net
|
|
Put the update of the power.smart_suspend device flag under the PM
spinlock of the device in case multiple bit fields in struct dev_pm_info
occupy one memory location which needs to be updated via RMW every time
any of these bit fields is updated.
The lock in question is already held around the power.direct_complete
flag update in device_prepare() for the same reason, so this change does
not add locking-related overhead to the code.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/2368159.ElGaqSPkdT@rjwysocki.net
|
|
The check before setting power.must_resume in device_suspend_noirq()
does not take power.child_count into account, but it should do that, so
use pm_runtime_need_not_resume() in it for this purpose and adjust the
comment next to it accordingly.
Fixes: 107d47b2b95e ("PM: sleep: core: Simplify the SMART_SUSPEND flag handling")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/3353728.44csPzL39Z@rjwysocki.net
|
|
Currently, if power.no_callbacks is set, device_prepare() will also set
power.direct_complete for the device. If power.direct_complete is set
in device_resume(), the clearing of power.is_prepared will be skipped
and if new children appear under the device at that point, a warning
will be printed.
After commit (f76b168b6f11 PM: Rename dev_pm_info.in_suspend to
is_prepared), power.is_prepared is generally cleared in device_resume()
before invoking the resume callback for the device which allows that
callback to add new children without triggering the warning, but this
does not happen for devices with power.direct_complete set.
This problem is visible in USB where usb_set_interface() can be called
before device_complete() clears power.is_prepared for interface devices
and since ep devices are added then, the warning is printed:
usb 1-1: reset high-speed USB device number 3 using ci_hdrc
ep_81: PM: parent 1-1:1.1 should not be sleeping
PM: resume devices took 0.936 seconds
Since it is legitimate to add the ep devices at that point, the
warning above is not particularly useful, so get rid of it by
clearing power.is_prepared in device_resume() for devices with
power.direct_complete set if they have no PM callbacks, in which
case they need not actually resume for the new children to work.
Suggested-by: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Xu Yang <xu.yang_2@nxp.com>
Link: https://patch.msgid.link/20250224070049.3338646-1-xu.yang_2@nxp.com
[ rjw: New subject, changelog edits, rephrased new code comment ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Add an optimization (on top of previous changes) to avoid calling
pm_runtime_blocked(), which involves acquiring the device's PM spinlock,
for devices with no PM callbacks and runtime PM "blocked".
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/2978873.e9J7NaK4W3@rjwysocki.net
|
|
A recent discussion has revealed that using DPM_FLAG_SMART_SUSPEND
unconditionally is generally problematic because it may lead to
situations in which the device's runtime PM information is internally
inconsistent or does not reflect its real state [1].
For this reason, change the handling of DPM_FLAG_SMART_SUSPEND so that
it is only taken into account if it is consistently set by the drivers
of all devices having any PM callbacks throughout dependency graphs in
accordance with the following rules:
- The "smart suspend" feature is only enabled for devices whose drivers
ask for it (that is, set DPM_FLAG_SMART_SUSPEND) and for devices
without PM callbacks unless they have never had runtime PM enabled.
- The "smart suspend" feature is not enabled for a device if it has not
been enabled for the device's parent unless the parent does not take
children into account or it has never had runtime PM enabled.
- The "smart suspend" feature is not enabled for a device if it has not
been enabled for one of the device's suppliers taking runtime PM into
account unless that supplier has never had runtime PM enabled.
Namely, introduce a new device PM flag called smart_suspend that is only
set if the above conditions are met and update all DPM_FLAG_SMART_SUSPEND
users to check power.smart_suspend instead of directly checking the
latter.
At the same time, drop the power.set_active flage introduced recently
in commit 3775fc538f53 ("PM: sleep: core: Synchronize runtime PM status
of parents and children") because it is now sufficient to check
power.smart_suspend along with the dev_pm_skip_resume() return value
to decide whether or not pm_runtime_set_active() needs to be called
for the device.
Link: https://lore.kernel.org/linux-pm/CAPDyKFroyU3YDSfw_Y6k3giVfajg3NQGwNWeteJWqpW29BojhQ@mail.gmail.com/ [1]
Fixes: 7585946243d6 ("PM: sleep: core: Restrict power.set_active propagation")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com> # drivers/pci
Link: https://patch.msgid.link/1914558.tdWV9SEqCh@rjwysocki.net
|
|
If device_prepare() runs on a device that has never had runtime
PM enabled so far, it may reasonably assume that runtime PM will
not be enabled for that device during the system suspend-resume
cycle currently in progress, but this has never been guaranteed.
To verify this assumption, make device_prepare() arrange for
triggering a device warning accompanied by a call trace dump if
runtime PM is enabled for such a device after it has returned.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/6131109.lOV4Wx5bFT@rjwysocki.net
|
|
There are only two callers of __pm_runtime_disable(), one of which is
device_suspend_late() and the other is pm_runtime_disable() that has
its own kerneldoc comment and there are no plans to add any more of
them. Since they use different values of the __pm_runtime_disable()
second parameter, the actual code behavior is different in each case,
but it is all documented in the __pm_runtime_disable() kerneldoc comment
which is not particularly straightforward.
For this reason, move the information from the __pm_runtime_disable()
kerneldoc comment to the pm_runtime_disable() one and into a separate
comment in device_suspend_late() and remove the __pm_runtime_disable()
kerneldoc comment altogether.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Link: https://patch.msgid.link/12617588.O9o76ZdvQC@rjwysocki.net
|
|
Commit 3775fc538f53 ("PM: sleep: core: Synchronize runtime PM status of
parents and children") exposed an issue related to simple_pm_bus_pm_ops
that uses pm_runtime_force_suspend() and pm_runtime_force_resume() as
bus type PM callbacks for the noirq phases of system-wide suspend and
resume.
The problem is that pm_runtime_force_suspend() does not distinguish
runtime-suspended devices from devices for which runtime PM has never
been enabled, so if it sees a device with runtime PM status set to
RPM_ACTIVE, it will assume that runtime PM is enabled for that device
and so it will attempt to suspend it with the help of its runtime PM
callbacks which may not be ready for that. As it turns out, this
causes simple_pm_bus_runtime_suspend() to crash due to a NULL pointer
dereference.
Another problem related to the above commit and simple_pm_bus_pm_ops is
that setting runtime PM status of a device handled by the latter to
RPM_ACTIVE will actually prevent it from being resumed because
pm_runtime_force_resume() only resumes devices with runtime PM status
set to RPM_SUSPENDED.
To mitigate these issues, do not allow power.set_active to propagate
beyond the parent of the device with DPM_FLAG_SMART_SUSPEND set that
will need to be resumed, which should be a sufficient stop-gap for the
time being, but they will need to be properly addressed in the future
because in general during system-wide resume it is necessary to resume
all devices in a dependency chain in which at least one device is going
to be resumed.
Fixes: 3775fc538f53 ("PM: sleep: core: Synchronize runtime PM status of parents and children")
Closes: https://lore.kernel.org/linux-pm/1c2433d4-7e0f-4395-b841-b8eac7c25651@nvidia.com/
Reported-by: Jon Hunter <jonathanh@nvidia.com>
Tested-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/6137505.lOV4Wx5bFT@rjwysocki.net
|
|
Commit 6e176bf8d461 ("PM: sleep: core: Do not skip callbacks in the
resume phase") overlooked the case in which the parent of a device with
DPM_FLAG_SMART_SUSPEND set did not use that flag and could be runtime-
suspended before a transition into a system-wide sleep state. In that
case, if the child is resumed during the subsequent transition from
that state into the working state, its runtime PM status will be set to
RPM_ACTIVE, but the runtime PM status of the parent will not be updated
accordingly, even though the parent will be resumed too, because of the
dev_pm_skip_suspend() check in device_resume_noirq().
Address this problem by tracking the need to set the runtime PM status
to RPM_ACTIVE during system-wide resume transitions for devices with
DPM_FLAG_SMART_SUSPEND set and all of the devices depended on by them.
Fixes: 6e176bf8d461 ("PM: sleep: core: Do not skip callbacks in the resume phase")
Closes: https://lore.kernel.org/linux-pm/Z30p2Etwf3F2AUvD@hovoldconsulting.com/
Reported-by: Johan Hovold <johan@kernel.org>
Tested-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Johan Hovold <johan+linaro@kernel.org>
Tested-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://patch.msgid.link/12619233.O9o76ZdvQC@rjwysocki.net
|
|
Allow configuring the DPM watchdog to warn about slow suspend/resume
functions without causing a system panic(). This allows you to set the
DPM_WATCHDOG_WARNING_TIMEOUT to something like 5 or 10 seconds to get
warnings about slow suspend/resume functions that eventually succeed.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Link: https://patch.msgid.link/20250109125957.v2.1.I4554f931b8da97948f308ecc651b124338ee9603@changeid
[ rjw: Subject edit ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
There is no function called __device_suspend() any more and it is still
mentioned in a comment in device_resume(), so update that comment.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patch.msgid.link/2787627.mvXUDI8C0e@rjwysocki.net
|
|
initcall_debug previous and new output:
...PM: calling pci_pm_suspend+0x0/0x1b0 @ 3233, parent: pci0000:00
...PM: calling pci_pm_suspend @ 3233, parent: pci0000:00
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Simplify the system-wide suspend of devices by invoking dpm_async_fn()
directly from the main loop in each suspend phase instead of using an
additional wrapper function for running it.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
During a system-wide suspend of devices, dpm_noirq_suspend_devices(),
dpm_suspend_late() and dpm_suspend() move devices from one list to
another. They do it with each device after its PM callback in the
given suspend phase has run or has been scheduled for asynchronous
execution, in case it is deleted from the current list in the
meantime.
However, devices can be moved to a new list before invoking their PM
callbacks (which usually is the case for the devices whose callbacks
are executed asynchronously anyway), because doing so does not affect
the ordering of that list. In either case, each device is moved to
the new list after the previous device has been moved to it or gone
away, and if a device is removed, it does not matter which list it is
in at that point, because deleting an entry from a list does not change
the ordering of the other entries in it.
Accordingly, modify the functions mentioned above to move devices to
new lists without waiting for their PM callbacks to run regardless of
whether or not they run asynchronously.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
The async_error and pm_transition variables are set under dpm_list_mtx
in multiple places in the system-wide device PM core code, which is
unnecessary and confusing, so rearrange the code so that the variables
in question are set before acquiring the lock.
While at it, add some empty code lines around locking to improve the
consistency of the code.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
The error logging and failure statistics updates are carried out in two
places in each system-wide device suspend phase, which is unnecessary
code duplication, so do that in one place in each phase, right after
invoking device suspend callbacks.
While at it, add "noirq" or "late" to the "async" string printed when
the failing device callback in the "noirq" or "late" suspend phase,
respectively, was run asynchronously.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
If the handling of two or more devices fails in one suspend-resume
phase, it should be counted once in the statistics which is not
guaranteed to happen during system-wide resume of devices due to
the possible asynchronous execution of device callbacks.
Address this by using the async_error static variable during system-wide
device resume to indicate that there has been a device resume error and
the given suspend-resume phase should be counted as failing.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
It is not necessary to define struct suspend_stats in a header file and the
suspend_stats variable in the core device system-wide PM code. They both
can be defined in kernel/power/main.c, next to the sysfs and debugfs code
accessing suspend_stats, which can be static.
Modify the code in question in accordance with the above observation and
replace the static inline functions manipulating suspend_stats with
regular ones defined in kernel/power/main.c.
While at it, move the enum suspend_stat_step to the end of suspend.h which
is a more suitable place for it.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
Instead of using a set of individual struct suspend_stats fields
representing suspend step failure counters, use an array of counters
indexed by enum suspend_stat_step for this purpose, which allows
dpm_save_failed_step() to increment the appropriate counter
automatically, so that its callers don't need to do that directly.
It also allows suspend_stats_show() to carry out a loop over the
counters array to print their values.
Because the counters cannot become negative, use unsigned int for
representing them.
The only user-observable impact of this change is a different
ordering of entries in the suspend_stats debugfs file which is not
expected to matter.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
Move is_async() and dpm_async_fn() in the PM core to a more suitable
place.
No functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
|
Notice that devices can be moved to dpm_prepared_list before running
their resume callbacks, in analogy with dpm_noirq_resume_devices() and
dpm_resume_early(), because doing so will not affect the final ordering
of that list.
Namely, if a device is the first dpm_suspended_list entry while
dpm_list_mtx is held, it has not been removed so far and it cannot be
removed until dpm_list_mtx is released, so moving it to dpm_prepared_list
at that point is valid. If it is removed later, while its resume
callback is running, it will be deleted from dpm_prepared_list without
changing the ordering of the other devices in that list.
Accordingly, rearrange the while () loop in dpm_resume() to move
devices to dpm_prepared_list before running their resume callbacks and
implify the locking and device reference counting in it.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
|
Before commit 7839d0078e0d ("PM: sleep: Fix possible deadlocks in core
system-wide PM code"), the resume of devices that were allowed to resume
asynchronously was scheduled before starting the resume of the other
devices, so the former did not have to wait for the latter unless
functional dependencies were present.
Commit 7839d0078e0d removed that optimization in order to address a
correctness issue, but it can be restored with the help of a new device
power management flag, so do that now.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
|
It is reported that in low-memory situations the system-wide resume core
code deadlocks, because async_schedule_dev() executes its argument
function synchronously if it cannot allocate memory (and not only in
that case) and that function attempts to acquire a mutex that is already
held. Executing the argument function synchronously from within
dpm_async_fn() may also be problematic for ordering reasons (it may
cause a consumer device's resume callback to be invoked before a
requisite supplier device's one, for example).
Address this by changing the code in question to use
async_schedule_dev_nocall() for scheduling the asynchronous
execution of device suspend and resume functions and to directly
run them synchronously if async_schedule_dev_nocall() returns false.
Link: https://lore.kernel.org/linux-pm/ZYvjiqX6EsL15moe@perf/
Reported-by: Youngmin Nam <youngmin.nam@samsung.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Tested-by: Youngmin Nam <youngmin.nam@samsung.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Cc: 5.7+ <stable@vger.kernel.org> # 5.7+: 6aa09a5bccd8 async: Split async_schedule_node_domain()
Cc: 5.7+ <stable@vger.kernel.org> # 5.7+: 7d4b5d7a37bd async: Introduce async_schedule_dev_nocall()
Cc: 5.7+ <stable@vger.kernel.org> # 5.7+
|
|
Assignments from pointer variables of type (void *) do not require
explicit type casts, so remove such type cases from the code in
drivers/base/power/main.c where applicable.
Signed-off-by: Li zeming <zeming@nfschina.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Merge changes related to system sleep, PM domains changes and power
management documentation changes for 5.18-rc1:
- Fix load_image_and_restore() error path (Ye Bin).
- Fix typos in comments in the system wakeup hadling code (Tom Rix).
- Clean up non-kernel-doc comments in hibernation code (Jiapeng
Chong).
- Fix __setup handler error handling in system-wide suspend and
hibernation core code (Randy Dunlap).
- Add device name to suspend_report_result() (Youngjin Jang).
- Make virtual guests honour ACPI S4 hardware signature by
default (David Woodhouse).
- Block power off of a parent PM domain unless child is in deepest
state (Ulf Hansson).
- Use dev_err_probe() to simplify error handling for generic PM
domains (Ahmad Fatoum).
- Fix sleep-in-atomic bug caused by genpd_debug_remove() (Shawn Guo).
- Document Intel uncore frequency scaling (Srinivas Pandruvada).
* pm-sleep:
PM: hibernate: Honour ACPI hardware signature by default for virtual guests
PM: sleep: Add device name to suspend_report_result()
PM: suspend: fix return value of __setup handler
PM: hibernate: fix __setup handler error handling
PM: hibernate: Clean up non-kernel-doc comments
PM: sleep: wakeup: Fix typos in comments
PM: hibernate: fix load_image_and_restore() error path
* pm-domains:
PM: domains: Fix sleep-in-atomic bug caused by genpd_debug_remove()
PM: domains: use dev_err_probe() to simplify error handling
PM: domains: Prevent power off for parent unless child is in deepest state
* pm-docs:
Documentation: admin-guide: pm: Document uncore frequency scaling
|
|
The function device_pm_check_callbacks() can be called under the spin
lock (in the reported case it happens from genpd_add_device() ->
dev_pm_domain_set(), when the genpd uses spinlocks rather than mutexes.
However this function uncoditionally uses spin_lock_irq() /
spin_unlock_irq(), thus not preserving the CPU flags. Use the
irqsave/irqrestore instead.
The backtrace for the reference:
[ 2.752010] ------------[ cut here ]------------
[ 2.756769] raw_local_irq_restore() called with IRQs enabled
[ 2.762596] WARNING: CPU: 4 PID: 1 at kernel/locking/irqflag-debug.c:10 warn_bogus_irq_restore+0x34/0x50
[ 2.772338] Modules linked in:
[ 2.775487] CPU: 4 PID: 1 Comm: swapper/0 Tainted: G S 5.17.0-rc6-00384-ge330d0d82eff-dirty #684
[ 2.781384] Freeing initrd memory: 46024K
[ 2.785839] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 2.785841] pc : warn_bogus_irq_restore+0x34/0x50
[ 2.785844] lr : warn_bogus_irq_restore+0x34/0x50
[ 2.785846] sp : ffff80000805b7d0
[ 2.785847] x29: ffff80000805b7d0 x28: 0000000000000000 x27: 0000000000000002
[ 2.785850] x26: ffffd40e80930b18 x25: ffff7ee2329192b8 x24: ffff7edfc9f60800
[ 2.785853] x23: ffffd40e80930b18 x22: ffffd40e80930d30 x21: ffff7edfc0dffa00
[ 2.785856] x20: ffff7edfc09e3768 x19: 0000000000000000 x18: ffffffffffffffff
[ 2.845775] x17: 6572206f74206465 x16: 6c696166203a3030 x15: ffff80008805b4f7
[ 2.853108] x14: 0000000000000000 x13: ffffd40e809550b0 x12: 00000000000003d8
[ 2.860441] x11: 0000000000000148 x10: ffffd40e809550b0 x9 : ffffd40e809550b0
[ 2.867774] x8 : 00000000ffffefff x7 : ffffd40e809ad0b0 x6 : ffffd40e809ad0b0
[ 2.875107] x5 : 000000000000bff4 x4 : 0000000000000000 x3 : 0000000000000000
[ 2.882440] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff7edfc03a8000
[ 2.889774] Call trace:
[ 2.892290] warn_bogus_irq_restore+0x34/0x50
[ 2.896770] _raw_spin_unlock_irqrestore+0x94/0xa0
[ 2.901690] genpd_unlock_spin+0x20/0x30
[ 2.905724] genpd_add_device+0x100/0x2d0
[ 2.909850] __genpd_dev_pm_attach+0xa8/0x23c
[ 2.914329] genpd_dev_pm_attach_by_id+0xc4/0x190
[ 2.919167] genpd_dev_pm_attach_by_name+0x3c/0xd0
[ 2.924086] dev_pm_domain_attach_by_name+0x24/0x30
[ 2.929102] psci_dt_attach_cpu+0x24/0x90
[ 2.933230] psci_cpuidle_probe+0x2d4/0x46c
[ 2.937534] platform_probe+0x68/0xe0
[ 2.941304] really_probe.part.0+0x9c/0x2fc
[ 2.945605] __driver_probe_device+0x98/0x144
[ 2.950085] driver_probe_device+0x44/0x15c
[ 2.954385] __device_attach_driver+0xb8/0x120
[ 2.958950] bus_for_each_drv+0x78/0xd0
[ 2.962896] __device_attach+0xd8/0x180
[ 2.966843] device_initial_probe+0x14/0x20
[ 2.971144] bus_probe_device+0x9c/0xa4
[ 2.975092] device_add+0x380/0x88c
[ 2.978679] platform_device_add+0x114/0x234
[ 2.983067] platform_device_register_full+0x100/0x190
[ 2.988344] psci_idle_init+0x6c/0xb0
[ 2.992113] do_one_initcall+0x74/0x3a0
[ 2.996060] kernel_init_freeable+0x2fc/0x384
[ 3.000543] kernel_init+0x28/0x130
[ 3.004132] ret_from_fork+0x10/0x20
[ 3.007817] irq event stamp: 319826
[ 3.011404] hardirqs last enabled at (319825): [<ffffd40e7eda0268>] __up_console_sem+0x78/0x84
[ 3.020332] hardirqs last disabled at (319826): [<ffffd40e7fd6d9d8>] el1_dbg+0x24/0x8c
[ 3.028458] softirqs last enabled at (318312): [<ffffd40e7ec90410>] _stext+0x410/0x588
[ 3.036678] softirqs last disabled at (318299): [<ffffd40e7ed1bf68>] __irq_exit_rcu+0x158/0x174
[ 3.045607] ---[ end trace 0000000000000000 ]---
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Currently, suspend_report_result() prints only function information.
If any driver uses a common PM function, nobody knows who exactly
called the failing function.
A device pinter is needed to recognize the failing device.
For example:
PM: dpm_run_callback(): pnp_bus_suspend+0x0/0x10 returns 0
PM: dpm_run_callback(): pci_pm_suspend+0x0/0x150 returns 0
become after the change:
serial 00:05: PM: dpm_run_callback(): pnp_bus_suspend+0x0/0x10 returns 0
pci 0000:00:01.3: PM: dpm_run_callback(): pci_pm_suspend+0x0/0x150 returns 0
Signed-off-by: Youngjin Jang <yj84.jang@samsung.com>
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Commit 2aa36604e824 ("PM: sleep: Avoid calling put_device() under
dpm_list_mtx") forgot to update the while () loop termination
condition to also break the loop if error is nonzero, which
causes the loop to become infinite if device_prepare() returns
an error for one device.
Add the missing !error check.
Fixes: 2aa36604e824 ("PM: sleep: Avoid calling put_device() under dpm_list_mtx")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reported-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Cc: All applicable <stable@vger.kernel.org>
|
|
It is generally unsafe to call put_device() with dpm_list_mtx held,
because the given device's release routine may carry out an action
depending on that lock which then may deadlock, so modify the
system-wide suspend and resume of devices to always drop dpm_list_mtx
before calling put_device() (and adjust white space somewhat while
at it).
For instance, this prevents the following splat from showing up in
the kernel log after a system resume in certain configurations:
[ 3290.969514] ======================================================
[ 3290.969517] WARNING: possible circular locking dependency detected
[ 3290.969519] 5.15.0+ #2420 Tainted: G S
[ 3290.969523] ------------------------------------------------------
[ 3290.969525] systemd-sleep/4553 is trying to acquire lock:
[ 3290.969529] ffff888117ab1138 ((wq_completion)hci0#2){+.+.}-{0:0}, at: flush_workqueue+0x87/0x4a0
[ 3290.969554]
but task is already holding lock:
[ 3290.969556] ffffffff8280fca8 (dpm_list_mtx){+.+.}-{3:3}, at: dpm_resume+0x12e/0x3e0
[ 3290.969571]
which lock already depends on the new lock.
[ 3290.969573]
the existing dependency chain (in reverse order) is:
[ 3290.969575]
-> #3 (dpm_list_mtx){+.+.}-{3:3}:
[ 3290.969583] __mutex_lock+0x9d/0xa30
[ 3290.969591] device_pm_add+0x2e/0xe0
[ 3290.969597] device_add+0x4d5/0x8f0
[ 3290.969605] hci_conn_add_sysfs+0x43/0xb0 [bluetooth]
[ 3290.969689] hci_conn_complete_evt.isra.71+0x124/0x750 [bluetooth]
[ 3290.969747] hci_event_packet+0xd6c/0x28a0 [bluetooth]
[ 3290.969798] hci_rx_work+0x213/0x640 [bluetooth]
[ 3290.969842] process_one_work+0x2aa/0x650
[ 3290.969851] worker_thread+0x39/0x400
[ 3290.969859] kthread+0x142/0x170
[ 3290.969865] ret_from_fork+0x22/0x30
[ 3290.969872]
-> #2 (&hdev->lock){+.+.}-{3:3}:
[ 3290.969881] __mutex_lock+0x9d/0xa30
[ 3290.969887] hci_event_packet+0xba/0x28a0 [bluetooth]
[ 3290.969935] hci_rx_work+0x213/0x640 [bluetooth]
[ 3290.969978] process_one_work+0x2aa/0x650
[ 3290.969985] worker_thread+0x39/0x400
[ 3290.969993] kthread+0x142/0x170
[ 3290.969999] ret_from_fork+0x22/0x30
[ 3290.970004]
-> #1 ((work_completion)(&hdev->rx_work)){+.+.}-{0:0}:
[ 3290.970013] process_one_work+0x27d/0x650
[ 3290.970020] worker_thread+0x39/0x400
[ 3290.970028] kthread+0x142/0x170
[ 3290.970033] ret_from_fork+0x22/0x30
[ 3290.970038]
-> #0 ((wq_completion)hci0#2){+.+.}-{0:0}:
[ 3290.970047] __lock_acquire+0x15cb/0x1b50
[ 3290.970054] lock_acquire+0x26c/0x300
[ 3290.970059] flush_workqueue+0xae/0x4a0
[ 3290.970066] drain_workqueue+0xa1/0x130
[ 3290.970073] destroy_workqueue+0x34/0x1f0
[ 3290.970081] hci_release_dev+0x49/0x180 [bluetooth]
[ 3290.970130] bt_host_release+0x1d/0x30 [bluetooth]
[ 3290.970195] device_release+0x33/0x90
[ 3290.970201] kobject_release+0x63/0x160
[ 3290.970211] dpm_resume+0x164/0x3e0
[ 3290.970215] dpm_resume_end+0xd/0x20
[ 3290.970220] suspend_devices_and_enter+0x1a4/0xba0
[ 3290.970229] pm_suspend+0x26b/0x310
[ 3290.970236] state_store+0x42/0x90
[ 3290.970243] kernfs_fop_write_iter+0x135/0x1b0
[ 3290.970251] new_sync_write+0x125/0x1c0
[ 3290.970257] vfs_write+0x360/0x3c0
[ 3290.970263] ksys_write+0xa7/0xe0
[ 3290.970269] do_syscall_64+0x3a/0x80
[ 3290.970276] entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 3290.970284]
other info that might help us debug this:
[ 3290.970285] Chain exists of:
(wq_completion)hci0#2 --> &hdev->lock --> dpm_list_mtx
[ 3290.970297] Possible unsafe locking scenario:
[ 3290.970299] CPU0 CPU1
[ 3290.970300] ---- ----
[ 3290.970302] lock(dpm_list_mtx);
[ 3290.970306] lock(&hdev->lock);
[ 3290.970310] lock(dpm_list_mtx);
[ 3290.970314] lock((wq_completion)hci0#2);
[ 3290.970319]
*** DEADLOCK ***
[ 3290.970321] 7 locks held by systemd-sleep/4553:
[ 3290.970325] #0: ffff888103bcd448 (sb_writers#4){.+.+}-{0:0}, at: ksys_write+0xa7/0xe0
[ 3290.970341] #1: ffff888115a14488 (&of->mutex){+.+.}-{3:3}, at: kernfs_fop_write_iter+0x103/0x1b0
[ 3290.970355] #2: ffff888100f719e0 (kn->active#233){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x10c/0x1b0
[ 3290.970369] #3: ffffffff82661048 (autosleep_lock){+.+.}-{3:3}, at: state_store+0x12/0x90
[ 3290.970384] #4: ffffffff82658ac8 (system_transition_mutex){+.+.}-{3:3}, at: pm_suspend+0x9f/0x310
[ 3290.970399] #5: ffffffff827f2a48 (acpi_scan_lock){+.+.}-{3:3}, at: acpi_suspend_begin+0x4c/0x80
[ 3290.970416] #6: ffffffff8280fca8 (dpm_list_mtx){+.+.}-{3:3}, at: dpm_resume+0x12e/0x3e0
[ 3290.970428]
stack backtrace:
[ 3290.970431] CPU: 3 PID: 4553 Comm: systemd-sleep Tainted: G S 5.15.0+ #2420
[ 3290.970438] Hardware name: Dell Inc. XPS 13 9380/0RYJWW, BIOS 1.5.0 06/03/2019
[ 3290.970441] Call Trace:
[ 3290.970446] dump_stack_lvl+0x44/0x57
[ 3290.970454] check_noncircular+0x105/0x120
[ 3290.970468] ? __lock_acquire+0x15cb/0x1b50
[ 3290.970474] __lock_acquire+0x15cb/0x1b50
[ 3290.970487] lock_acquire+0x26c/0x300
[ 3290.970493] ? flush_workqueue+0x87/0x4a0
[ 3290.970503] ? __raw_spin_lock_init+0x3b/0x60
[ 3290.970510] ? lockdep_init_map_type+0x58/0x240
[ 3290.970519] flush_workqueue+0xae/0x4a0
[ 3290.970526] ? flush_workqueue+0x87/0x4a0
[ 3290.970544] ? drain_workqueue+0xa1/0x130
[ 3290.970552] drain_workqueue+0xa1/0x130
[ 3290.970561] destroy_workqueue+0x34/0x1f0
[ 3290.970572] hci_release_dev+0x49/0x180 [bluetooth]
[ 3290.970624] bt_host_release+0x1d/0x30 [bluetooth]
[ 3290.970687] device_release+0x33/0x90
[ 3290.970695] kobject_release+0x63/0x160
[ 3290.970705] dpm_resume+0x164/0x3e0
[ 3290.970710] ? dpm_resume_early+0x251/0x3b0
[ 3290.970718] dpm_resume_end+0xd/0x20
[ 3290.970723] suspend_devices_and_enter+0x1a4/0xba0
[ 3290.970737] pm_suspend+0x26b/0x310
[ 3290.970746] state_store+0x42/0x90
[ 3290.970755] kernfs_fop_write_iter+0x135/0x1b0
[ 3290.970764] new_sync_write+0x125/0x1c0
[ 3290.970777] vfs_write+0x360/0x3c0
[ 3290.970785] ksys_write+0xa7/0xe0
[ 3290.970794] do_syscall_64+0x3a/0x80
[ 3290.970803] entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 3290.970811] RIP: 0033:0x7f41b1328164
[ 3290.970819] Code: 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b7 0f 1f 80 00 00 00 00 8b 05 4a d2 2c 00 48 63 ff 85 c0 75 13 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 54 f3 c3 66 90 55 53 48 89 d5 48 89 f3 48 83
[ 3290.970824] RSP: 002b:00007ffe6ae21b28 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[ 3290.970831] RAX: ffffffffffffffda RBX: 0000000000000004 RCX: 00007f41b1328164
[ 3290.970836] RDX: 0000000000000004 RSI: 000055965e651070 RDI: 0000000000000004
[ 3290.970839] RBP: 000055965e651070 R08: 000055965e64f390 R09: 00007f41b1e3d1c0
[ 3290.970843] R10: 000000000000000a R11: 0000000000000246 R12: 0000000000000004
[ 3290.970846] R13: 0000000000000001 R14: 000055965e64f2b0 R15: 0000000000000004
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
In the cpuidle-psci case, runtime PM in combination with the generic PM
domain (genpd), may be used when entering/exiting a shared idlestate. More
precisely, genpd relies on runtime PM to be enabled for the attached device
(in this case it belongs to a CPU), to properly manage the reference
counting of its PM domain.
This works fine most of the time, but during system suspend in
dpm_suspend_late(), the PM core disables runtime PM for all devices. Beyond
this point, calls to pm_runtime_get_sync() to runtime resume a device may
fail and therefore it could also mess up the reference counting in genpd.
To fix this problem, let's call wake_up_all_idle_cpus() in
dpm_suspend_late(), prior to disabling runtime PM. In this way a device
that belongs to a CPU, becomes runtime resumed through cpuidle-psci and
stays like that, because the runtime PM usage count has been bumped in
device_prepare().
Diagnosed-by: Maulik Shah <mkshah@codeaurora.org>
Suggested-by: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Commit 8651f97bd951 ("PM / cpuidle: System resume hang fix with
cpuidle") that introduced cpuidle pausing during system suspend
did that to work around a platform firmware issue causing systems
to hang during resume if CPUs were allowed to enter idle states
in the system suspend and resume code paths.
However, pausing cpuidle before the last phase of suspending
devices is the source of an otherwise arbitrary difference between
the suspend-to-idle path and other system suspend variants, so it is
cleaner to do that later, before taking secondary CPUs offline (it
is still safer to take secondary CPUs offline with cpuidle paused,
though).
Modify the code accordingly, but in order to avoid code duplication,
introduce new wrapper functions, pm_sleep_disable_secondary_cpus()
and pm_sleep_enable_secondary_cpus(), to combine cpuidle_pause()
and cpuidle_resume(), respectively, with the handling of secondary
CPUs during system-wide transitions to sleep states.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
It is pointless to pause cpuidle in the suspend-to-idle path,
because it is going to be resumed in the same path later and
pausing it does not serve any particular purpose in that case.
Rework the code to avoid doing that.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
transitions
There is no reason to allow "syscore" devices to runtime-suspend
during system-wide PM transitions, because they are subject to the
same possible failure modes as any other devices in that respect.
Accordingly, change device_prepare() and device_complete() to call
pm_runtime_get_noresume() and pm_runtime_put(), respectively, for
"syscore" devices too.
Fixes: 057d51a1268f ("Merge branch 'pm-sleep'")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: 3.10+ <stable@vger.kernel.org> # 3.10+
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
|
|
There are variables(power.may_skip_resume and dev->power.must_resume)
and DPM_FLAG_MAY_SKIP_RESUME flags to control the resume of devices after
a system wide suspend transition.
Setting the DPM_FLAG_MAY_SKIP_RESUME flag means that the driver allows
its "noirq" and "early" resume callbacks to be skipped if the device
can be left in suspend after a system-wide transition into the working
state. PM core determines that the driver's "noirq" and "early" resume
callbacks should be skipped or not with dev_pm_skip_resume() function by
checking power.may_skip_resume variable.
power.must_resume variable is getting set to false in __device_suspend()
function without checking device's DPM_FLAG_MAY_SKIP_RESUME settings.
In problematic scenario, where all the devices in the suspend_late
stage are successful and some device can fail to suspend in
suspend_noirq phase. So some devices successfully suspended in suspend_late
stage are not getting chance to execute __device_suspend_noirq()
to set dev->power.must_resume variable to true and not getting
resumed in early_resume phase.
Add a check for device's DPM_FLAG_MAY_SKIP_RESUME flag before
setting power.must_resume variable in __device_suspend function.
Fixes: 6e176bf8d461 ("PM: sleep: core: Do not skip callbacks in the resume phase")
Signed-off-by: Prasad Sodagudi <psodagud@codeaurora.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
