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>
A lot of drivers create a dev_pm_ops struct with the system sleep
suspend/resume callbacks set to pm_runtime_force_suspend() and
pm_runtime_force_resume().
These drivers can now use the DEFINE_RUNTIME_DEV_PM_OPS() macro, which
will use pm_runtime_force_{suspend,resume}() as the system sleep
callbacks, while having the same dead code removal characteristic that
is already provided by DEFINE_SIMPLE_DEV_PM_OPS().
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
These macros are defined conditionally, according to CONFIG_PM:
- if CONFIG_PM is enabled, these macros resolve to
DEFINE_SIMPLE_DEV_PM_OPS(), and the dev_pm_ops symbol will be
exported.
- if CONFIG_PM is disabled, these macros will result in a dummy static
dev_pm_ops to be created with the __maybe_unused flag. The dev_pm_ops
will then be discarded by the compiler, along with the provided
callback functions if they are not used anywhere else.
In the second case, the symbol is not exported, which should be
perfectly fine - users of the symbol should all use the pm_ptr() or
pm_sleep_ptr() macro, so the dev_pm_ops marked as "extern" in the
client's code will never be accessed.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Keep this macro in line with the other ones. This makes it possible to
use them in the cases where the underlying dev_pm_ops structure is
exported.
Restore the "static" qualifier in the two drivers where the
DEFINE_SIMPLE_DEV_PM_OPS macro was used.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The deprecated UNIVERSAL_DEV_PM_OPS() macro uses the provided callbacks
for both runtime PM and system sleep, which is very likely to be a
mistake, as a system sleep can be triggered while a given device is
already PM-suspended, which would cause the suspend callback to be
called twice.
The amount of users of UNIVERSAL_DEV_PM_OPS() is also tiny (16
occurences) compared to the number of places where
SET_SYSTEM_SLEEP_PM_OPS() is used with pm_runtime_force_suspend() and
pm_runtime_force_resume(), which makes me think that none of these cases
are actually valid.
As the new macro DEFINE_UNIVERSAL_DEV_PM_OPS() which was introduced to
replace UNIVERSAL_DEV_PM_OPS() is currently unused, remove it before
someone starts to use it in yet another invalid case.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In some cases (for example, during system-wide suspend and resume of
devices) it is useful to know whether or not runtime PM has ever been
enabled for a given device and, if so, what the runtime PM status of
it had been right before runtime PM was disabled for it last time.
For this reason, introduce a new struct dev_pm_info field called
last_status that will be used for capturing the runtime PM status of
the device when its power.disable_depth counter changes from 0 to 1.
The new field will be set to RPM_INVALID to start with and whenever
power.disable_depth changes from 1 to 0, so it will be valid only
when runtime PM of the device is currently disabled, but it has been
enabled at least once.
Immediately use power.last_status in rpm_resume() to make it handle
the case when PM runtime is disabled for the device, but its runtime
PM status is RPM_ACTIVE more consistently. Namely, make it return 1
if power.last_status is also equal to RPM_ACTIVE in that case (the
idea being that if the status was RPM_ACTIVE last time when
power.disable_depth was changing from 0 to 1 and it is still
RPM_ACTIVE, it can be assumed to reflect what happened to the device
last time when it was using runtime PM) and -EACCES otherwise.
Update the documentation to provide a description of last_status and
change the description of pm_runtime_resume() in it to reflect the
new behavior of rpm_active().
While at it, rearrange the code in pm_runtime_enable() to be more
straightforward and replace the WARN() macro in it with a pr_warn()
invocation which is less disruptive.
Link: https://lore.kernel.org/linux-pm/20211026222626.39222-1-ulf.hansson@linaro.org/t/#u
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This commit introduces the following macros:
SYSTEM_SLEEP_PM_OPS()
LATE_SYSTEM_SLEEP_PM_OPS()
NOIRQ_SYSTEM_SLEEP_PM_OPS()
RUNTIME_PM_OPS()
These new macros are very similar to their SET_*_PM_OPS() equivalent.
They however differ in the fact that the callbacks they set will always
be seen as referenced by the compiler. This means that the callback
functions don't need to be wrapped with a #ifdef CONFIG_PM guard, or
tagged with __maybe_unused, to prevent the compiler from complaining
about unused static symbols. The compiler will then simply evaluate at
compile time whether or not these symbols are dead code.
The callbacks that are only useful with CONFIG_PM_SLEEP is enabled, are
now also wrapped with a new pm_sleep_ptr() macro, which is inspired from
pm_ptr(). This is needed for drivers that use different callbacks for
sleep and runtime PM, to handle the case where CONFIG_PM is set and
CONFIG_PM_SLEEP is not.
This commit also deprecates the following macros:
SIMPLE_DEV_PM_OPS()
UNIVERSAL_DEV_PM_OPS()
And introduces the following macros:
DEFINE_SIMPLE_DEV_PM_OPS()
DEFINE_UNIVERSAL_DEV_PM_OPS()
These macros are similar to the functions they were created to replace,
with the following differences:
- They use the new macros introduced above, and as such always
reference the provided callback functions.
- They are not tagged with __maybe_unused. They are meant to be used
with pm_ptr() or pm_sleep_ptr() for DEFINE_UNIVERSAL_DEV_PM_OPS()
and DEFINE_SIMPLE_DEV_PM_OPS() respectively.
- They declare the symbol static, since every driver seems to do that
anyway; and if a non-static use-case is needed an indirection pointer
could be used.
The point of this change, is to progressively switch from a code model
where PM callbacks are all protected behind CONFIG_PM guards, to a code
model where the PM callbacks are always seen by the compiler, but
discarded if not used.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The pm_ptr() macro was previously conditionally defined, according to
the value of the CONFIG_PM option. This meant that the pointed structure
was either referenced (if CONFIG_PM was set), or never referenced (if
CONFIG_PM was not set), causing it to be detected as unused by the
compiler.
This worked fine, but required the __maybe_unused compiler attribute to
be used to every symbol pointed to by a pointer wrapped with pm_ptr().
We can do better. With this change, the pm_ptr() is now defined the
same, independently of the value of CONFIG_PM. It now uses the (?:)
ternary operator to conditionally resolve to its argument. Since the
condition is known at compile time, the compiler will then choose to
discard the unused symbols, which won't need to be tagged with
__maybe_unused anymore.
This pm_ptr() macro is usually used with pointers to dev_pm_ops
structures created with SIMPLE_DEV_PM_OPS() or similar macros. These do
use a __maybe_unused flag, which is now useless with this change, so it
later can be removed. However in the meantime it causes no harm, and all
the drivers still compile fine with the new pm_ptr() macro.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
As pm_runtime_need_not_resume() relies also on usage_count, it can return
a different value in pm_runtime_force_suspend() compared to when called in
pm_runtime_force_resume(). Different return values can happen if anything
calls PM runtime functions in between, and causes the parent child_count
to increase on every resume.
So far I've seen the issue only for omapdrm that does complicated things
with PM runtime calls during system suspend for legacy reasons:
omap_atomic_commit_tail() for omapdrm.0
dispc_runtime_get()
wakes up 58000000.dss as it's the dispc parent
dispc_runtime_resume()
rpm_resume() increases parent child_count
dispc_runtime_put() won't idle, PM runtime suspend blocked
pm_runtime_force_suspend() for 58000000.dss, !pm_runtime_need_not_resume()
__update_runtime_status()
system suspended
pm_runtime_force_resume() for 58000000.dss, pm_runtime_need_not_resume()
pm_runtime_enable() only called because of pm_runtime_need_not_resume()
omap_atomic_commit_tail() for omapdrm.0
dispc_runtime_get()
wakes up 58000000.dss as it's the dispc parent
dispc_runtime_resume()
rpm_resume() increases parent child_count
dispc_runtime_put() won't idle, PM runtime suspend blocked
...
rpm_suspend for 58000000.dss but parent child_count is now unbalanced
Let's fix the issue by adding a flag for needs_force_resume and use it in
pm_runtime_force_resume() instead of pm_runtime_need_not_resume().
Additionally omapdrm system suspend could be simplified later on to avoid
lots of unnecessary PM runtime calls and the complexity it adds. The
driver can just use internal functions that are shared between the PM
runtime and system suspend related functions.
Fixes: 4918e1f87c ("PM / runtime: Rework pm_runtime_force_suspend/resume()")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Cc: 4.16+ <stable@vger.kernel.org> # 4.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
struct device is declared twice, so remove the duplicate.
Signed-off-by: Wan Jiabing <wanjiabing@vivo.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The clock API splits its interface into sleepable ant atomic contexts:
- clk_prepare/clk_unprepare for stuff that might sleep
- clk_enable_clk_disable for anything that may be done in atomic context
The code handling runtime PM for clocks only calls clk_disable() on
suspend requests, and clk_enable on resume requests. This means that
runtime PM with clock providers that only have the prepare/unprepare
methods implemented is basically useless.
Many clock implementations can't accommodate atomic contexts. This is
often the case when communication with the clock happens through another
subsystem like I2C or SCMI.
Let's make the clock PM code useful with such clocks by safely invoking
clk_prepare/clk_unprepare upon resume/suspend requests. Of course, when
such clocks are registered with the PM layer then pm_runtime_irq_safe()
can't be used, and neither pm_runtime_suspend() nor pm_runtime_resume()
may be invoked in atomic context.
For clocks that do implement the enable and disable methods then
everything just works as before.
A note on sparse:
According to https://lwn.net/Articles/109066/ there are things
that sparse can't cope with. In particular, pm_clk_op_lock() and
pm_clk_op_unlock() may or may not lock/unlock psd->lock depending on
some runtime condition. To work around that we tell it the lock is
always untaken for the purpose of static analisys.
Thanks to Naresh Kamboju for reporting issues with the initial patch.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Tested-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Commit 8234f6734c ("PM-runtime: Switch autosuspend over to using
hrtimers") switched PM runtime autosuspend to use hrtimers and all
related time accounting in ns, but missed to update the timer_expires
data type in struct dev_pm_info to u64.
This causes the timer_expires value to be truncated on 32-bit
architectures when assignment is done from u64 values:
rpm_suspend()
|- dev->power.timer_expires = expires;
Fix it by changing the timer_expires type to u64.
Fixes: 8234f6734c ("PM-runtime: Switch autosuspend over to using hrtimers")
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Acked-by: Pavel Machek <pavel@ucw.cz>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: 5.0+ <stable@vger.kernel.org> # 5.0+
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This way, when the dev_pm_ops instance is not referenced anywhere, it
will simply be dropped by the compiler without a warning.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This macro is analogous to the infamous of_match_ptr(). If CONFIG_PM
is enabled, this macro will resolve to its argument, otherwise to NULL.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Update the documentation of the driver flags for system-wide power
management to reflect the current code flows and be more consistent.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Rename DPM_FLAG_LEAVE_SUSPENDED to DPM_FLAG_MAY_SKIP_RESUME which
matches its purpose more closely.
No functional impact.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Wolfram Sang <wsa@the-dreams.de> # for I2C
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Rename DPM_FLAG_NEVER_SKIP to DPM_FLAG_NO_DIRECT_COMPLETE which
matches its purpose more closely.
No functional impact.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com> # for PCI parts
Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Because all callers of dev_pm_smart_suspend_and_suspended use it only
for checking whether or not to skip driver suspend callbacks for a
device, rename it to dev_pm_skip_suspend() in analogy with
dev_pm_skip_resume().
No functional impact.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
The name of dev_pm_may_skip_resume() may be easily confused with the
power.may_skip_resume flag which is not checked by that function, so
rename the former as dev_pm_skip_resume().
No functional impact.
Suggested-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
For a subsystem/driver that either doesn't support runtime PM or makes use
of pm_runtime_set_active() during ->probe(), may try to access its device
when probing, even if it may not be fully powered on from the PM domain's
point of view. This may be the case when the used PM domain is a genpd
provider, that implements genpd's ->start|stop() device callbacks.
There are cases where the subsystem/driver managed to avoid the above
problem, simply by calling pm_runtime_enable() and pm_runtime_get_sync()
during ->probe(). However, this approach comes with a drawback, especially
if the subsystem/driver implements a ->runtime_resume() callback.
More precisely, the subsystem/driver then needs to use a device flag, which
is checked in its ->runtime_resume() callback, as to avoid powering on its
resources the first time the callback is invoked. This is needed because
the subsystem/driver has already powered on the resources for the device,
during ->probe() and before it called pm_runtime_get_sync().
In a way to avoid this boilerplate code and the inefficient check for "if
(first_time_suspend)" in the ->runtime_resume() callback for these
subsystems/drivers, let's introduce and export a dev_pm_domain_start()
function, that may be called during ->probe() instead.
Moreover, let the dev_pm_domain_start() invoke an optional ->start()
callback, added to the struct dev_pm_domain, as to allow a PM domain
specific implementation.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Note that after previous changes dpm_noirq_begin() and
dpm_noirq_end() each have only one caller, so move the code from
them to their respective callers and drop them.
Also note that dpm_noirq_resume_devices() and
dpm_noirq_suspend_devices() need not be exported any more, so make
them both static.
This change is not expected to alter functionality.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
After recent hibernation-related changes, there are no more callers
of dev_pm_skip_next_resume_phases() except for the PM core itself
in which it is more straightforward to run the statements from
that function directly, so do that and drop it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Convert the PM documents to ReST, in order to allow them to
build with Sphinx.
The conversion is actually:
- add blank lines and indentation in order to identify paragraphs;
- fix tables markups;
- add some lists markups;
- mark literal blocks;
- adjust title markups.
At its new index.rst, let's add a :orphan: while this is not linked to
the main index.rst file, in order to avoid build warnings.
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Mark Brown <broonie@kernel.org>
Acked-by: Srivatsa S. Bhat (VMware) <srivatsa@csail.mit.edu>
Based on 1 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license as published by
the free software foundation either version 2 of the license or at
your option any later version this program is distributed in the
hope that it will be useful but without any warranty without even
the implied warranty of merchantability or fitness for a particular
purpose see the gnu general public license for more details you
should have received a copy of the gnu general public license along
with this program if not write to the free software foundation inc
59 temple place suite 330 boston ma 02111 1307 usa
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-or-later
has been chosen to replace the boilerplate/reference in 1334 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Richard Fontana <rfontana@redhat.com>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070033.113240726@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Avoid the open-coding of the accounted time acquisition in
runtime_active|suspend_time_show() and make them call
pm_runtime_active|suspended_time() instead.
Note that this change also indirectly avoids holding dev->power.lock
around the do_div() computation and the sprintf() call which is an
additional improvement.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
[ rjw: Changelog ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
All device objects in the driver model contain fields that control the
handling of various power management activities. However, it's not
always useful. There are few instances where pseudo devices are added
to the model just to take advantage of many other features like
kobjects, udev events, and so on. One such example is cpu devices and
their caches.
The sysfs for the cpu caches are managed by adding devices with cpu
as the parent in cpu_device_create() when secondary cpu is brought
online. Generally when the secondary CPUs are hotplugged back in as part
of resume from suspend-to-ram, we call cpu_device_create() from the cpu
hotplug state machine while the cpu device associated with that CPU is
not yet ready to be resumed as the device_resume() call happens bit
later. It's not really needed to set the flag is_prepared for cpu
devices as they are mostly pseudo device and hotplug framework deals
with state machine and not managed through the cpu device.
This often results in annoying warning when resuming:
Enabling non-boot CPUs ...
CPU1: Booted secondary processor
cache: parent cpu1 should not be sleeping
CPU1 is up
CPU2: Booted secondary processor
cache: parent cpu2 should not be sleeping
CPU2 is up
.... and so on.
So in order to fix these kind of errors, we could just completely avoid
doing any power management related initialisations and operations if
they are not used by these devices.
Add no_pm flags to indicate that the device doesn't require any sort of
PM activities and all of them can be completely skipped. We can use the
same flag to also avoid adding not used *power* sysfs entries for these
devices. For now, lets use this for cpu cache devices.
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Tested-by: Eugeniu Rosca <erosca@de.adit-jv.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Replace jiffies-based accounting for runtime_active_time and
runtime_suspended_time with ktime-based accounting. This makes the
runtime debug counters inline with genpd and other PM subsytems which
use ktime-based accounting.
Timekeeping is initialized before driver_init(). It's only at that time
that PM-runtime can be enabled.
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
[switch from ktime to raw nsec]
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
PM-runtime uses the timer infrastructure for autosuspend. This implies
that the minimum time before autosuspending a device is in the range
of 1 tick included to 2 ticks excluded
-On arm64 this means between 4ms and 8ms with default jiffies
configuration
-And on arm, it is between 10ms and 20ms
These values are quite high for embedded systems which sometimes want
the duration to be in the range of 1 ms.
It is possible to switch autosuspend over to using hrtimers to get
finer granularity for short durations and take advantage of slack to
retain some margins and get long timeouts with minimum wakeups.
On an arm64 platform that uses 1ms for autosuspending timeout of its
GPU, idle power is reduced by 10% with hrtimer.
The latency impact on arm64 hikey octo cores is:
- mark_last_busy: from 1.11 us to 1.25 us
- rpm_suspend: from 15.54 us to 15.38 us
[Only the code path of rpm_suspend() that starts hrtimer has been
measured.]
arm64 image (arm64 default defconfig) decreases by around 3KB
with following details:
$ size vmlinux-timer
text data bss dec hex filename
12034646 6869268 386840 19290754 1265a82 vmlinux
$ size vmlinux-hrtimer
text data bss dec hex filename
12030550 6870164 387032 19287746 1264ec2 vmlinux
The latency impact on arm 32bits snowball dual cores is :
- mark_last_busy: from 0.31 us usec to 0.77 us
- rpm_suspend: from 6.83 us to 6.67 usec
The increase of the image for snowball platform that I used for
testing performance impact, is neglictable (244B).
$ size vmlinux-timer
text data bss dec hex filename
7157961 2119580 264120 9541661 91981d build-ux500/vmlinux
size vmlinux-hrtimer
text data bss dec hex filename
7157773 2119884 264248 9541905 919911 vmlinux-hrtimer
And arm 32bits image (multi_v7_defconfig) increases by around 1.7KB
with following details:
$ size vmlinux-timer
text data bss dec hex filename
13304443 6803420 402768 20510631 138f7a7 vmlinux
$ size vmlinux-hrtimer
text data bss dec hex filename
13304299 6805276 402768 20512343 138fe57 vmlinux
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Middle-layer code doing suspend-time optimizations for devices with
the DPM_FLAG_SMART_SUSPEND flag set (currently, the PCI bus type and
the ACPI PM domain) needs to make the core skip ->thaw_early and
->thaw callbacks for those devices in some cases and it sets the
power.direct_complete flag for them for this purpose.
However, it turns out that setting power.direct_complete outside of
the PM core is a bad idea as it triggers an excess invocation of
pm_runtime_enable() in device_resume().
For this reason, provide a helper to clear power.is_late_suspended
and power.is_suspended to be invoked by the middle-layer code in
question instead of setting power.direct_complete and make that code
call the new helper.
Fixes: c4b65157ae (PCI / PM: Take SMART_SUSPEND driver flag into account)
Fixes: 05087360fd (ACPI / PM: Take SMART_SUSPEND driver flag into account)
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>
Define and document a new driver flag, DPM_FLAG_LEAVE_SUSPENDED, to
instruct the PM core and middle-layer (bus type, PM domain, etc.)
code that it is desirable to leave the device in runtime suspend
after system-wide transitions to the working state (for example,
the device may be slow to resume and it may be better to avoid
resuming it right away).
Generally, the middle-layer code involved in the handling of the
device is expected to indicate to the PM core whether or not the
device may be left in suspend with the help of the device's
power.may_skip_resume status bit. That has to happen in the "noirq"
phase of the preceding system suspend (or analogous) transition.
The middle layer is then responsible for handling the device as
appropriate in its "noirq" resume callback which is executed
regardless of whether or not the device may be left suspended, but
the other resume callbacks (except for ->complete) will be skipped
automatically by the core if the device really can be left in
suspend.
The additional power.must_resume status bit introduced for the
implementation of this mechanisn is used internally by the PM core
to track the requirement to resume the device (which may depend on
its children etc).
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Make the PCI bus type take DPM_FLAG_SMART_SUSPEND into account in its
system-wide PM callbacks and make sure that all code that should not
run in parallel with pci_pm_runtime_resume() is executed in the "late"
phases of system suspend, freeze and poweroff transitions.
[Note that the pm_runtime_suspended() check in pci_dev_keep_suspended()
is an optimization, because if is not passed, all of the subsequent
checks may be skipped and some of them are much more overhead in
general.]
Also use the observation that if the device is in runtime suspend
at the beginning of the "late" phase of a system-wide suspend-like
transition, its state cannot change going forward (runtime PM is
disabled for it at that time) until the transition is over and the
subsequent system-wide PM callbacks should be skipped for it (as
they generally assume the device to not be suspended), so add checks
for that in pci_pm_suspend_late/noirq(), pci_pm_freeze_late/noirq()
and pci_pm_poweroff_late/noirq().
Moreover, if pci_pm_resume_noirq() or pci_pm_restore_noirq() is
called during the subsequent system-wide resume transition and if
the device was left in runtime suspend previously, its runtime PM
status needs to be changed to "active" as it is going to be put
into the full-power state, so add checks for that too to these
functions.
In turn, if pci_pm_thaw_noirq() runs after the device has been
left in runtime suspend, the subsequent "thaw" callbacks need
to be skipped for it (as they may not work correctly with a
suspended device), so set the power.direct_complete flag for the
device then to make the PM core skip those callbacks.
In addition to the above add a core helper for checking if
DPM_FLAG_SMART_SUSPEND is set and the device runtime PM status is
"suspended" at the same time, which is done quite often in the new
code (and will be done elsewhere going forward too).
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Define and document a SMART_SUSPEND flag to instruct bus types and PM
domains that the system suspend callbacks provided by the driver can
cope with runtime-suspended devices, so from the driver's perspective
it should be safe to leave devices in runtime suspend during system
suspend.
Setting that flag may also cause middle-layer code (bus types,
PM domains etc.) to skip invocations of the ->suspend_late and
->suspend_noirq callbacks provided by the driver if the device
is in runtime suspend at the beginning of the "late" phase of
the system-wide suspend transition, in which case the driver's
system-wide resume callbacks may be invoked back-to-back with
its ->runtime_suspend callback, so the driver has to be able to
cope with that too.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
The motivation for this change is to provide a way to work around
a problem with the direct-complete mechanism used for avoiding
system suspend/resume handling for devices in runtime suspend.
The problem is that some middle layer code (the PCI bus type and
the ACPI PM domain in particular) returns positive values from its
system suspend ->prepare callbacks regardless of whether the driver's
->prepare returns a positive value or 0, which effectively prevents
drivers from being able to control the direct-complete feature.
Some drivers need that control, however, and the PCI bus type has
grown its own flag to deal with this issue, but since it is not
limited to PCI, it is better to address it by adding driver flags at
the core level.
To that end, add a driver_flags field to struct dev_pm_info for flags
that can be set by device drivers at the probe time to inform the PM
core and/or bus types, PM domains and so on on the capabilities and/or
preferences of device drivers. Also add two static inline helpers
for setting that field and testing it against a given set of flags
and make the driver core clear it automatically on driver remove
and probe failures.
Define and document two PM driver flags related to the direct-
complete feature: NEVER_SKIP and SMART_PREPARE that can be used,
respectively, to indicate to the PM core that the direct-complete
mechanism should never be used for the device and to inform the
middle layer code (bus types, PM domains etc) that it can only
request the PM core to use the direct-complete mechanism for
the device (by returning a positive value from its ->prepare
callback) if it also has been requested by the driver.
While at it, make the core check pm_runtime_suspended() when
setting power.direct_complete so that it doesn't need to be
checked by ->prepare callbacks.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
According to recent changes for ACPI, the are longer any users of
pm_complete_with_resume_check(), thus let's drop it.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Put the device interrupts disabling and enabling as well as
cpuidle_pause() and cpuidle_resume() called during the "noirq"
stages of system suspend into separate functions to allow the
core suspend-to-idle code to be optimized (later).
The only functional difference this makes is that debug facilities
and diagnostic tools will not include the above operations into the
"noirq" device suspend/resume duration measurements.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The run_wake flag in struct dev_pm_info is used to indicate whether
or not the device is capable of generating remote wakeup signals at
run time (or in the system working state), but the distinction
between runtime remote wakeup and system wakeup signaling has always
been rather artificial. The only practical reason for it to exist
at the core level was that ACPI and PCI treated those two cases
differently, but that's not the case any more after recent changes.
For this reason, get rid of the run_wake flag and, when applicable,
use device_set_wakeup_capable() and device_can_wakeup() instead of
device_set_run_wake() and device_run_wake(), respectively.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Move the document describing the system sleep state transitions API
for devices to Documentation/driver-api/pm/, convert it to reST and
update it to use current terminology. Also remove the remaining
reference to the old version of it from pm.h.
The new document still contains references to some documents in the
.txt format that will be converted later.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Refresh the struct dev_pm_ops kerneldoc comment, so that it looks
better and is more readable after processing by Sphinx, and drop
the kerneldoc marker from a few other comments ("PM_EVENT_ messages"
and a couple of enum types declarations) which are not proper
kerneldoc and generally confuse Sphinx.
Also change the comment describing struct dev_pm_domain into
a kerneldoc one.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Here's the new driver core patches for 4.10-rc1.
Big thing here is the nice addition of "functional dependencies" to the
driver core. The idea has been talked about for a very long time, great
job to Rafael for stepping up and implementing it. It's been tested for
longer than the 4.9-rc1 date, we held off on merging it earlier in order
to feel more comfortable about it.
Other than that, it's just a handful of small other patches, some good
cleanups to the mess that is the firmware class code, and we have a test
driver for the deferred probe logic.
All of these have been in linux-next for a while with no reported
issues.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCWFAvPQ8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+ym3NgCgmhFeWEkp9SDt17YGGavmnzQUlBQAoJlUipJp
PHeQkq15ZWw3wWC9FEvM
=91M1
-----END PGP SIGNATURE-----
Merge tag 'driver-core-4.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
Pull driver core updates from Greg KH:
"Here's the new driver core patches for 4.10-rc1.
Big thing here is the nice addition of "functional dependencies" to
the driver core. The idea has been talked about for a very long time,
great job to Rafael for stepping up and implementing it. It's been
tested for longer than the 4.9-rc1 date, we held off on merging it
earlier in order to feel more comfortable about it.
Other than that, it's just a handful of small other patches, some good
cleanups to the mess that is the firmware class code, and we have a
test driver for the deferred probe logic.
All of these have been in linux-next for a while with no reported
issues"
* tag 'driver-core-4.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (30 commits)
firmware: Correct handling of fw_state_wait() return value
driver core: Silence device links sphinx warning
firmware: remove warning at documentation generation time
drivers: base: dma-mapping: Fix typo in dmam_alloc_non_coherent comments
driver core: test_async: fix up typo found by 0-day
firmware: move fw_state_is_done() into UHM section
firmware: do not use fw_lock for fw_state protection
firmware: drop bit ops in favor of simple state machine
firmware: refactor loading status
firmware: fix usermode helper fallback loading
driver core: firmware_class: convert to use class_groups
driver core: devcoredump: convert to use class_groups
driver core: class: add class_groups support
kernfs: Declare two local data structures static
driver-core: fix platform_no_drv_owner.cocci warnings
drivers/base/memory.c: Remove unused 'first_page' variable
driver core: add CLASS_ATTR_WO()
drivers: base: cacheinfo: support DT overrides for cache properties
drivers: base: cacheinfo: add pr_fmt logging
drivers: base: cacheinfo: fix boot error message when acpi is enabled
...
If the device has no links to suppliers that should be used for
runtime PM (links with DEVICE_LINK_PM_RUNTIME set), there is no
reason to walk the list of suppliers for that device during
runtime suspend and resume.
Add a simple mechanism to detect that case and possibly avoid the
extra unnecessary overhead.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, there is a problem with taking functional dependencies
between devices into account.
What I mean by a "functional dependency" is when the driver of device
B needs device A to be functional and (generally) its driver to be
present in order to work properly. This has certain consequences
for power management (suspend/resume and runtime PM ordering) and
shutdown ordering of these devices. In general, it also implies that
the driver of A needs to be working for B to be probed successfully
and it cannot be unbound from the device before the B's driver.
Support for representing those functional dependencies between
devices is added here to allow the driver core to track them and act
on them in certain cases where applicable.
The argument for doing that in the driver core is that there are
quite a few distinct use cases involving device dependencies, they
are relatively hard to get right in a driver (if one wants to
address all of them properly) and it only gets worse if multiplied
by the number of drivers potentially needing to do it. Morever, at
least one case (asynchronous system suspend/resume) cannot be handled
in a single driver at all, because it requires the driver of A to
wait for B to suspend (during system suspend) and the driver of B to
wait for A to resume (during system resume).
For this reason, represent dependencies between devices as "links",
with the help of struct device_link objects each containing pointers
to the "linked" devices, a list node for each of them, status
information, flags, and an RCU head for synchronization.
Also add two new list heads, representing the lists of links to the
devices that depend on the given one (consumers) and to the devices
depended on by it (suppliers), and a "driver presence status" field
(needed for figuring out initial states of device links) to struct
device.
The entire data structure consisting of all of the lists of link
objects for all devices is protected by a mutex (for link object
addition/removal and for list walks during device driver probing
and removal) and by SRCU (for list walking in other case that will
be introduced by subsequent change sets). If CONFIG_SRCU is not
selected, however, an rwsem is used for protecting the entire data
structure.
In addition, each link object has an internal status field whose
value reflects whether or not drivers are bound to the devices
pointed to by the link or probing/removal of their drivers is in
progress etc. That field is only modified under the device links
mutex, but it may be read outside of it in some cases (introduced by
subsequent change sets), so modifications of it are annotated with
WRITE_ONCE().
New links are added by calling device_link_add() which takes three
arguments: pointers to the devices in question and flags. In
particular, if DL_FLAG_STATELESS is set in the flags, the link status
is not to be taken into account for this link and the driver core
will not manage it. In turn, if DL_FLAG_AUTOREMOVE is set in the
flags, the driver core will remove the link automatically when the
consumer device driver unbinds from it.
One of the actions carried out by device_link_add() is to reorder
the lists used for device shutdown and system suspend/resume to
put the consumer device along with all of its children and all of
its consumers (and so on, recursively) to the ends of those lists
in order to ensure the right ordering between all of the supplier
and consumer devices.
For this reason, it is not possible to create a link between two
devices if the would-be supplier device already depends on the
would-be consumer device as either a direct descendant of it or a
consumer of one of its direct descendants or one of its consumers
and so on.
There are two types of link objects, persistent and non-persistent.
The persistent ones stay around until one of the target devices is
deleted, while the non-persistent ones are removed automatically when
the consumer driver unbinds from its device (ie. they are assumed to
be valid only as long as the consumer device has a driver bound to
it). Persistent links are created by default and non-persistent
links are created when the DL_FLAG_AUTOREMOVE flag is passed
to device_link_add().
Both persistent and non-persistent device links can be deleted
with an explicit call to device_link_del().
Links created without the DL_FLAG_STATELESS flag set are managed
by the driver core using a simple state machine. There are 5 states
each link can be in: DORMANT (unused), AVAILABLE (the supplier driver
is present and functional), CONSUMER_PROBE (the consumer driver is
probing), ACTIVE (both supplier and consumer drivers are present and
functional), and SUPPLIER_UNBIND (the supplier driver is unbinding).
The driver core updates the link state automatically depending on
what happens to the linked devices and for each link state specific
actions are taken in addition to that.
For example, if the supplier driver unbinds from its device, the
driver core will also unbind the drivers of all of its consumers
automatically under the assumption that they cannot function
properly without the supplier. Analogously, the driver core will
only allow the consumer driver to bind to its device if the
supplier driver is present and functional (ie. the link is in
the AVAILABLE state). If that's not the case, it will rely on
the existing deferred probing mechanism to wait for the supplier
driver to become available.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The previous patch renamed several files that are cross-referenced
along the Kernel documentation. Adjust the links to point to
the right places.
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The ignore_children flag is used only when CONFIG_PM is set, so let's move
it into that section within the struct dev_pm_info.
Move also the corresponding pm_suspend_ignore_children() API out of
device.h into pm_runtime.h, to be consistent with similar APIs.
Unfortunate this causes the Toshiba PCI SD mmc host driver to fail to
compile as it needs pm_runtime.h, so let's fix this here as well.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
If a suitable prepare callback cannot be found for a given device and
its driver has no PM callbacks at all, assume that it can go direct to
complete when the system goes to sleep.
The reason for this is that there's lots of devices in a system that do
no PM at all and there's no reason for them to prevent their ancestors
to do direct_complete if they can support it.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
There is a concern that if the platform firmware was involved in
the system resume that's being completed, some devices might have
been reset by it and if those devices had the power.direct_complete
flag set during the preceding suspend transition, they may stay
in a reset-power-on state indefinitely (until they are runtime-resumed
and then suspended again). That may not be a big deal from the
individual device's perspective, but if the system is an SoC, it may
be prevented from entering deep SoC-wide low-power states on idle
because of that.
The devices that are most likely to be affected by this issue are
PCI devices and ACPI-enumerated devices using the general ACPI PM
domain, so to prevent it from happening for those devices, force a
runtime resume for them if they have their power.direct_complete
flags set and the platform firmware was involved in the resume
transition currently in progress.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Turns out we can automate the handling for the device_may_wakeup()
quite a bit by using the kernel wakeup source list as suggested
by Rafael J. Wysocki <rjw@rjwysocki.net>.
And as some hardware has separate dedicated wake-up interrupt
in addition to the IO interrupt, we can automate the handling by
adding a generic threaded interrupt handler that just calls the
device PM runtime to wake up the device.
This allows dropping code from device drivers as we currently
are doing it in multiple ways, and often wrong.
For most drivers, we should be able to drop the following
boilerplate code from runtime_suspend and runtime_resume
functions:
...
device_init_wakeup(dev, true);
...
if (device_may_wakeup(dev))
enable_irq_wake(irq);
...
if (device_may_wakeup(dev))
disable_irq_wake(irq);
...
device_init_wakeup(dev, false);
...
We can replace it with just the following init and exit
time code:
...
device_init_wakeup(dev, true);
dev_pm_set_wake_irq(dev, irq);
...
dev_pm_clear_wake_irq(dev);
device_init_wakeup(dev, false);
...
And for hardware with dedicated wake-up interrupts:
...
device_init_wakeup(dev, true);
dev_pm_set_dedicated_wake_irq(dev, irq);
...
dev_pm_clear_wake_irq(dev);
device_init_wakeup(dev, false);
...
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The same approach is used as for the existing SET_SYSTEM_SLEEP_PM_OPS,
but for noirq callbacks.
New SET_NOIRQ_SYSTEM_SLEEP_PM_OPS, defined for CONFIG_PM_SLEEP, will
point ->suspend_noirq, ->freeze_noirq and ->poweroff_noirq to the same
function. Vice versa happens for ->resume_noirq, ->thaw_noirq and
->restore_noirq.
Signed-off-by: Grygorii Strashko <Grygorii.Strashko@linaro.org>
Acked-by: Santosh Shilimkar <ssantosh@kernel.org>
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Kevin Hilman <khilman@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Youngjin Jang