Revert commit fbd401e95e ("ACPI: processor: idle: Redefine two
functions as void") because it depends on a problematic one.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Revert commit 559f2eacc8 ACPI: processor: Do not expose global variable
acpi_idle_driver" because it depends on a problematic one.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
acpi_processor_setup_cstates() and acpi_processor_setup_cpuidle_cx()
are called after successfully obtaining power information. Among other
things, these setup functions check the C-state count against zero.
However, that check is done by acpi_processor_get_power_info_cst()
which will cause acpi_processor_get_power_info() to fail if it does
no pass, so the checks in the two functions mentioned above are
redundant.
Drop those redundant checks.
No intentional functional impact.
Signed-off-by: Huisong Li <lihuisong@huawei.com>
[ rjw: Subject and changelog rewrite ]
Link: https://patch.msgid.link/20251105093647.3557248-1-lihuisong@huawei.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since acpi_processor_setup_cstates() is a more logical place for setting
idle state flags than acpi_processor_setup_cpuidle_cx(), move that code
from the latter to the former.
It also allows direct references to acpi_idle_driver in
acpi_processor_setup_cpuidle_cx() to be avoided.
No intentional functional impact.
Signed-off-by: Huisong Li <lihuisong@huawei.com>
[ rjw: Subject and changelog rewrite ]
Link: https://patch.msgid.link/20250929093754.3998136-5-lihuisong@huawei.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Notice that acpi_processor_power_init() and acpi_processor_power_exit()
don't need to return any values because their callers don't check them
anyway, so redefine those functions as void.
While at it, rearrange the code in acpi_processor_power_init() to
reduce the indentation level, get rid of a redundant local variable
in that function, and rephrase a code comment in it.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Instead of using static variable flat_state_cnt to pass data between
functions involved in the _LPI information processing, pass the current
number of "flattened" idle states to flatten_lpi_states() and make it
return the updated number of those states. At the same time, use a
local variable called state_count to store the number of "flattened"
idle states found so far in acpi_processor_get_lpi_info().
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Tested-by: Sudeep Holla <sudeep.holla@arm.com>
Acked-by: lihuisong@huawei.com
Link: https://patch.msgid.link/10715991.nUPlyArG6x@rafael.j.wysocki
Add a new module import namespace called ACPI_PROCESSOR_IDLE for
functions exported from the non-modular part of the ACPI processor
driver to the modular part of it.
Export acpi_processor_claim_cst_control() and acpi_processor_evaluate_cst()
in that namespace to hide them from unrelated modules.
They are also used by the intel_idle driver, but it is non-modular,
so it can call them regardless of the way the symbols are exported.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/3376499.aeNJFYEL58@rafael.j.wysocki
Currently, the ACPI idle driver is registered from within a CPU
hotplug callback. Although this didn't cause any functional issues,
this is questionable and confusing. And it is better to register
the cpuidle driver when all of the CPUs have been brought up.
So add a new function to initialize acpi_idle_driver based on the
power management information of an available CPU and register cpuidle
driver in acpi_processor_driver_init().
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Link: https://patch.msgid.link/20250728070612.1260859-3-lihuisong@huawei.com
[ rjw: Added missing inline modifiers ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The cpuidle device's memory is leaked when cpuidle device registration
fails in acpi_processor_power_init(). Free it as appropriate.
Fixes: 3d339dcbb5 ("cpuidle / ACPI : move cpuidle_device field out of the acpi_processor_power structure")
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Link: https://patch.msgid.link/20250728070612.1260859-2-lihuisong@huawei.com
[ rjw: Changed the order of the new statements, added empty line after if () ]
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Make acpi_processor_driver_init() call arch_cpu_rescan_dead_smt_siblings(),
via a new wrapper function called acpi_idle_rescan_dead_smt_siblings(),
after successfully initializing the driver, to allow the "dead" SMT
siblings to go into deep idle states, which is necessary for the
processor to be able to reach deep package C-states (like PC10) going
forward, so that power can be reduced sufficiently in suspend-to-idle,
among other things.
However, do it only if the ACPI idle driver is the current cpuidle
driver (otherwise it is assumed that another cpuidle driver will take
care of this) and avoid doing it on architectures other than x86.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Link: https://patch.msgid.link/2005721.PYKUYFuaPT@rjwysocki.net
pr->power.count stands for the number of valid processor cstates. And it
is correctly set with the return value of acpi_processor_power_verify(),
in acpi_processor_get_cstate_info().
Remove the redundant logic that unnecessarily overrides pr->power.count
value.
No functional change intended.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Link: https://patch.msgid.link/20250409005026.1932637-3-rui.zhang@intel.com
[ rjw: Removed unused variable, fixed up white space ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Prior to commit 496121c021 ("ACPI: processor: idle: Allow probing on
platforms with one ACPI C-state"), the acpi_idle driver wouldn't load on
systems without a valid C-State at least as deep as C2.
The behavior was desirable for guests on hypervisors such as VMWare
ESXi, which by default don't have the _CST ACPI method, and set the C2
and C3 latencies to 101 and 1001 microseconds respectively via the FADT,
to signify they're unsupported.
Since the above change though, these virtualized deployments end up
loading acpi_idle, and thus entering the default C1 C-State set by
acpi_processor_get_power_info_default(); this is undesirable for a
system that's communicating to the OS it doesn't want C-States (missing
_CST, and invalid C2/C3 in FADT).
Make acpi_processor_get_power_info_fadt() return -ENODEV in that case,
so that acpi_processor_get_cstate_info() exits early and doesn't set
pr->flags.power = 1.
Fixes: 496121c021 ("ACPI: processor: idle: Allow probing on platforms with one ACPI C-state")
Signed-off-by: Giovanni Gherdovich <ggherdovich@suse.cz>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Link: https://patch.msgid.link/20250328143040.9348-1-ggherdovich@suse.cz
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Pull more ACPI updates from Rafael Wysocki:
"These add a common init function for arch-specific ACPI
initialization, clean up idle states initialization in the ACPI
processor_idle driver and update quirks:
- Introduce acpi_arch_init() for architecture-specific ACPI subsystem
initialization (Miao Wang)
- Clean up Asus quirks in acpi_quirk_skip_dmi_ids[] and add a quirk
to skip I2C clients on Acer Iconia One 8 A1-840 (Hans de Goede)
- Make the ACPI processor_idle driver use acpi_idle_play_dead() for
all idle states regardless of their types (Rafael Wysocki)"
* tag 'acpi-6.13-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI: introduce acpi_arch_init()
ACPI: x86: Clean up Asus entries in acpi_quirk_skip_dmi_ids[]
ACPI: x86: Add skip i2c clients quirk for Acer Iconia One 8 A1-840
ACPI: processor_idle: Use acpi_idle_play_dead() for all C-states
The acpi_cst_latency_cmp() comparison function currently used for
sorting C-state latencies does not satisfy transitivity, causing
incorrect sorting results.
Specifically, if there are two valid acpi_processor_cx elements A and B
and one invalid element C, it may occur that A < B, A = C, and B = C.
Sorting algorithms assume that if A < B and A = C, then C < B, leading
to incorrect ordering.
Given the small size of the array (<=8), we replace the library sort
function with a simple insertion sort that properly ignores invalid
elements and sorts valid ones based on latency. This change ensures
correct ordering of the C-state latencies.
Fixes: 65ea8f2c6e ("ACPI: processor idle: Fix up C-state latency if not ordered")
Reported-by: Julian Sikorski <belegdol@gmail.com>
Closes: https://lore.kernel.org/lkml/70674dc7-5586-4183-8953-8095567e73df@gmail.com
Signed-off-by: Kuan-Wei Chiu <visitorckw@gmail.com>
Tested-by: Julian Sikorski <belegdol@gmail.com>
Cc: All applicable <stable@vger.kernel.org>
Link: https://patch.msgid.link/20240701205639.117194-1-visitorckw@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Xen HVM guests were observed taking triple-faults when attempting to
online a previously offlined vCPU.
Investigation showed that the fault was coming from a failing call
to lockdep_assert_irqs_disabled(), in load_current_idt() which was
too early in the CPU bringup to actually catch the exception and
report the failure cleanly.
This was a false positive, caused by acpi_idle_play_dead() setting
the per-cpu hardirqs_enabled flag by calling safe_halt(). Switch it
to use raw_safe_halt() instead, which doesn't do so.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: 6.6+ <stable@vger.kernel.org> # 6.6+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Arm® Functional Fixed Hardware Specification defines LPI states,
which provide an architectural context loss flags field that can
be used to describe the context that might be lost when an LPI
state is entered.
- Core context Lost
- General purpose registers.
- Floating point and SIMD registers.
- System registers, include the System register based
- generic timer for the core.
- Debug register in the core power domain.
- PMU registers in the core power domain.
- Trace register in the core power domain.
- Trace context loss
- GICR
- GICD
Qualcomm's custom CPUs preserves the architectural state,
including keeping the power domain for local timers active.
when core is power gated, the local timers are sufficient to
wake the core up without needing broadcast timer.
The patch fixes the evaluation of cpuidle arch_flags, and moves only to
broadcast timer if core context lost is defined in ACPI LPI.
Fixes: a36a7fecfe ("ACPI / processor_idle: Add support for Low Power Idle(LPI) states")
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Oza Pawandeep <quic_poza@quicinc.com>
Link: https://lore.kernel.org/r/20231003173333.2865323-1-quic_poza@quicinc.com
Signed-off-by: Will Deacon <will@kernel.org>
This was introduced with commit e1c467e690 ("x86, hotplug: Wake up CPU0
via NMI instead of INIT, SIPI, SIPI") to eventually support physical
hotplug of CPU0:
"We'll change this code in the future to wake up hard offlined CPU0 if
real platform and request are available."
11 years later this has not happened and physical hotplug is not officially
supported. Remove the cruft.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Michael Kelley <mikelley@microsoft.com>
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Tested-by: Helge Deller <deller@gmx.de> # parisc
Tested-by: Guilherme G. Piccoli <gpiccoli@igalia.com> # Steam Deck
Link: https://lore.kernel.org/r/20230512205255.768845190@linutronix.de
Pull ACPI updates from Rafael Wysocki:
"These fix a frequency limit issue in the ACPI processor performance
library code, fix a few issues in the ACPICA code, improve Crystal
Cove support in the ACPI PMIC driver, fix string handling in the ACPI
battery driver, add IRQ override quirks for a few machines more, fix
other assorted problems and clean up code and documentation.
Specifics:
- Drop port I/O validation for some regions to avoid AML failures due
to rejections of legitimate port I/O writes (Mario Limonciello)
- Constify acpi_get_handle() pathname argument to allow its callers
to pass const pathnames to it (Sakari Ailus)
- Prevent acpi_ns_simple_repair() from crashing in some cases when
AE_AML_NO_RETURN_VALUE should be returned (Daniil Tatianin)
- Fix typo in CDAT DSMAS struct definition (Lukas Wunner)
- Drop an unnecessary (void *) conversion from the ACPI processor
driver (Zhou jie)
- Modify the ACPI processor performance library code to use the "no
limit" frequency QoS as appropriate and adjust the intel_pstate
driver accordingly (Rafael Wysocki)
- Add support for NBFT to the ACPI table parser (Stuart Hayes)
- Introduce list of known non-PNP devices to avoid enumerating some
of them as PNP devices (Rafael Wysocki)
- Add x86 ACPI paths to the ACPI entry in MAINTAINERS to allow
scripts to report the actual maintainers information (Rafael
Wysocki)
- Add two more entries to the ACPI IRQ override quirk list (Adam
Niederer, Werner Sembach)
- Add a pmic_i2c_address entry for Intel Bay Trail Crystal Cove to
allow intel_soc_pmic_exec_mipi_pmic_seq_element() to be used with
the Bay Trail Crystal Cove PMIC OpRegion driver (Hans de Goede)
- Add comments with DSDT power OpRegion field names to the ACPI PMIC
driver (Hans de Goede)
- Fix string termination handling in the ACPI battery driver (Armin
Wolf)
- Limit error type to 32-bit width in the ACPI APEI error injection
code (Shuai Xue)
- Fix Lenovo Ideapad Z570 DMI match in the ACPI backlight driver
(Hans de Goede)
- Silence missing prototype warnings in some places in the
ACPI-related code (Ammar Faizi)
- Make kobj_type structures used in the ACPI code constant (Thomas
Weißschuh)
- Correct spelling in firmware-guide/ACPI (Randy Dunlap)
- Clarify the meaning of Explicit and Implicit in the _DSD GPIO
properties documentation (Andy Shevchenko)
- Fix some kernel-doc comments in the ACPI CPPC library code (Yang
Li)"
* tag 'acpi-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (25 commits)
ACPI: make kobj_type structures constant
Documentation: firmware-guide: gpio-properties: Clarify Explicit and Implicit
ACPICA: Fix typo in CDAT DSMAS struct definition
ACPI: resource: Do IRQ override on all TongFang GMxRGxx
ACPI: resource: Add IRQ overrides for MAINGEAR Vector Pro 2 models
ACPI: CPPC: Fix some kernel-doc comments
ACPI: video: Fix Lenovo Ideapad Z570 DMI match
Documentation: firmware-guide/ACPI: correct spelling
ACPI: PMIC: Add comments with DSDT power opregion field names
ACPI: battery: Increase maximum string length
ACPI: battery: Fix buffer overread if not NUL-terminated
ACPI: APEI: EINJ: Limit error type to 32-bit width
MAINTAINERS: Add x86 ACPI paths to the ACPI entry
ACPI: battery: Fix missing NUL-termination with large strings
ACPI: PNP: Introduce list of known non-PNP devices
ACPICA: nsrepair: handle cases without a return value correctly
ACPI: Silence missing prototype warnings
cpufreq: intel_pstate: Drop ACPI _PSS states table patching
ACPI: processor: perflib: Avoid updating frequency QoS unnecessarily
ACPI: processor: perflib: Use the "no limit" frequency QoS
...
The (void *) type pointer does not need to be cast, so don't do
that in lapic_timer_check_state().
Signed-off-by: Zhou jie <zhoujie@nfschina.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Drop a redundant "else", a "return" statement at the end of a void
function and redundant parentheses around an unsigent int variable
name from the ACPI processor idle driver.
No expected functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The return value of acpi_fetch_acpi_dev() could be NULL, which would
cause a NULL pointer dereference to occur in acpi_device_hid().
Signed-off-by: Li Zhong <floridsleeves@gmail.com>
[ rjw: Subject and changelog edits, added empty line after if () ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Pull ACPI updates from Rafael Wysocki:
"ACPI and PNP updates for 6.1-rc1.
These rearrange the ACPI device object initialization code (to get rid
of a redundant parent pointer from struct acpi_device among other
things), unify the _UID handling, drop support for some _OSI strings
that should not be necessary any more, add new IDs to support more
hardware and some more quirks, fix a few issues and clean up code all
over.
Specifics:
- Reimplement acpi_get_pci_dev() using the list of physical devices
associated with the given ACPI device object (Rafael Wysocki)
- Rename ACPI device object reference counting functions (Rafael
Wysocki)
- Rearrange ACPI device object initialization code (Rafael Wysocki)
- Drop parent field from struct acpi_device (Rafael Wysocki)
- Extend the the int3472-tps68470 driver to support multiple
consumers of a single TPS68470 along with the requisite
framework-level support (Daniel Scally)
- Filter out non-memory resources in is_memory(), add a helper
function to find all memory type resources of an ACPI device object
and use that function in 3 places (Heikki Krogerus)
- Add IRQ override quirks for Asus Vivobook K3402ZA/K3502ZA and ASUS
model S5402ZA (Tamim Khan, Kellen Renshaw)
- Fix acpi_dev_state_d0() kerneldoc (Sakari Ailus)
- Fix up suspend-to-idle support on ASUS Rembrandt laptops (Mario
Limonciello)
- Clean up ACPI platform devices support code (Andy Shevchenko, John
Garry)
- Clean up ACPI bus management code (Andy Shevchenko, ye xingchen)
- Add support for multiple DMA windows with different offsets to the
ACPI device enumeration code and use it on LoongArch (Jianmin Lv)
- Clean up the ACPI LPSS (Intel SoC) driver (Andy Shevchenko)
- Add a quirk for Dell Inspiron 14 2-in-1 for StorageD3Enable (Mario
Limonciello)
- Drop unused dev_fmt() and redundant 'HMAT' prefix from the HMAT
parsing code (Liu Shixin)
- Make ACPI FPDT parsing code avoid calling acpi_os_map_memory() on
invalid physical addresses (Hans de Goede)
- Silence missing-declarations warning related to Apple device
properties management (Lukas Wunner)
- Disable frequency invariance in the CPPC library if registers used
by cppc_get_perf_ctrs() are accessed via PCC (Jeremy Linton)
- Add ACPI disabled check to acpi_cpc_valid() (Perry Yuan)
- Fix Tx acknowledge in the PCC address space handler (Huisong Li)
- Use wait_for_completion_timeout() for PCC mailbox operations
(Huisong Li)
- Release resources on PCC address space setup failure path (Rafael
Mendonca)
- Remove unneeded result variables from APEI code (ye xingchen)
- Print total number of records found during BERT log parsing (Dmitry
Monakhov)
- Drop support for 3 _OSI strings that should not be necessary any
more and update documentation on custom _OSI strings so that adding
new ones is not encouraged any more (Mario Limonciello)
- Drop unneeded result variable from ec_write() (ye xingchen)
- Remove the leftover struct acpi_ac_bl from the ACPI AC driver
(Hanjun Guo)
- Reorder symbols to get rid of a few forward declarations in the
ACPI fan driver (Uwe Kleine-König)
- Add Toshiba Satellite/Portege Z830 ACPI backlight quirk (Arvid
Norlander)
- Add ARM DMA-330 controller to the supported list in the ACPI AMBA
driver (Vijayenthiran Subramaniam)
- Drop references to non-functional 01.org/linux-acpi web site from
MAINTAINERS and Kconfig help texts (Rafael Wysocki)
- Replace strlcpy() with unused retval with strscpy() in the ACPI
support code (Wolfram Sang)
- Do not initialize ret in main() in the pfrut utility (Shi junming)
- Drop useless ACPI DSDT override documentation (Rafael Wysocki)
- Fix a few typos and wording mistakes in the ACPI device enumeration
documentation (Jean Delvare)
- Introduce acpi_dev_uid_to_integer() to convert a _UID string into
an integer value (Andy Shevchenko)
- Use acpi_dev_uid_to_integer() in several places to unify _UID
handling (Andy Shevchenko)
- Drop unused pnpid32_to_pnpid() declaration from PNP code (Gaosheng
Cui)"
* tag 'acpi-6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (79 commits)
ACPI: LPSS: Deduplicate skipping device in acpi_lpss_create_device()
ACPI: LPSS: Replace loop with first entry retrieval
ACPI: x86: s2idle: Add another ID to s2idle_dmi_table
ACPI: x86: s2idle: Fix a NULL pointer dereference
MAINTAINERS: Drop records pointing to 01.org/linux-acpi
ACPI: Kconfig: Drop link to https://01.org/linux-acpi
ACPI: docs: Drop useless DSDT override documentation
ACPI: DPTF: Drop stale link from Kconfig help
ACPI: x86: s2idle: Add a quirk for ASUSTeK COMPUTER INC. ROG Flow X13
ACPI: x86: s2idle: Add a quirk for Lenovo Slim 7 Pro 14ARH7
ACPI: x86: s2idle: Add a quirk for ASUS ROG Zephyrus G14
ACPI: x86: s2idle: Add a quirk for ASUS TUF Gaming A17 FA707RE
ACPI: x86: s2idle: Add module parameter to prefer Microsoft GUID
ACPI: x86: s2idle: If a new AMD _HID is missing assume Rembrandt
ACPI: x86: s2idle: Move _HID handling for AMD systems into structures
platform/x86: int3472: Add board data for Surface Go2 IR camera
platform/x86: int3472: Support multiple gpio lookups in board data
platform/x86: int3472: Support multiple clock consumers
ACPI: bus: Add iterator for dependent devices
ACPI: scan: Add acpi_dev_get_next_consumer_dev()
...
Old, circa 2002 chipsets have a bug: they don't go idle when they are
supposed to. So, a workaround was added to slow the CPU down and
ensure that the CPU waits a bit for the chipset to actually go idle.
This workaround is ancient and has been in place in some form since
the original kernel ACPI implementation.
But, this workaround is very painful on modern systems. The "inl()"
can take thousands of cycles (see Link: for some more detailed
numbers and some fun kernel archaeology).
First and foremost, modern systems should not be using this code.
Typical Intel systems have not used it in over a decade because it is
horribly inferior to MWAIT-based idle.
Despite this, people do seem to be tripping over this workaround on
AMD system today.
Limit the "dummy wait" workaround to Intel systems. Keep Modern AMD
systems from tripping over the workaround. Remotely modern Intel
systems use intel_idle instead of this code and will, in practice,
remain unaffected by the dummy wait.
Reported-by: K Prateek Nayak <kprateek.nayak@amd.com>
Suggested-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lore.kernel.org/all/20220921063638.2489-1-kprateek.nayak@amd.com/
Link: https://lkml.kernel.org/r/20220922184745.3252932-1-dave.hansen@intel.com
Pull RCU updates from Paul McKenney:
- Documentation updates
- Miscellaneous fixes
- Callback-offload updates, perhaps most notably a new
RCU_NOCB_CPU_DEFAULT_ALL Kconfig option that causes all CPUs to be
offloaded at boot time, regardless of kernel boot parameters.
This is useful to battery-powered systems such as ChromeOS and
Android. In addition, a new RCU_NOCB_CPU_CB_BOOST kernel boot
parameter prevents offloaded callbacks from interfering with
real-time workloads and with energy-efficiency mechanisms
- Polled grace-period updates, perhaps most notably making these APIs
account for both normal and expedited grace periods
- Tasks RCU updates, perhaps most notably reducing the CPU overhead of
RCU tasks trace grace periods by more than a factor of two on a
system with 15,000 tasks.
The reduction is expected to increase with the number of tasks, so it
seems reasonable to hypothesize that a system with 150,000 tasks
might see a 20-fold reduction in CPU overhead
- Torture-test updates
- Updates that merge RCU's dyntick-idle tracking into context tracking,
thus reducing the overhead of transitioning to kernel mode from
either idle or nohz_full userspace execution for kernels that track
context independently of RCU.
This is expected to be helpful primarily for kernels built with
CONFIG_NO_HZ_FULL=y
* tag 'rcu.2022.07.26a' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu: (98 commits)
rcu: Add irqs-disabled indicator to expedited RCU CPU stall warnings
rcu: Diagnose extended sync_rcu_do_polled_gp() loops
rcu: Put panic_on_rcu_stall() after expedited RCU CPU stall warnings
rcutorture: Test polled expedited grace-period primitives
rcu: Add polled expedited grace-period primitives
rcutorture: Verify that polled GP API sees synchronous grace periods
rcu: Make Tiny RCU grace periods visible to polled APIs
rcu: Make polled grace-period API account for expedited grace periods
rcu: Switch polled grace-period APIs to ->gp_seq_polled
rcu/nocb: Avoid polling when my_rdp->nocb_head_rdp list is empty
rcu/nocb: Add option to opt rcuo kthreads out of RT priority
rcu: Add nocb_cb_kthread check to rcu_is_callbacks_kthread()
rcu/nocb: Add an option to offload all CPUs on boot
rcu/nocb: Fix NOCB kthreads spawn failure with rcu_nocb_rdp_deoffload() direct call
rcu/nocb: Invert rcu_state.barrier_mutex VS hotplug lock locking order
rcu/nocb: Add/del rdp to iterate from rcuog itself
rcu/tree: Add comment to describe GP-done condition in fqs loop
rcu: Initialize first_gp_fqs at declaration in rcu_gp_fqs()
rcu/kvfree: Remove useless monitor_todo flag
rcu: Cleanup RCU urgency state for offline CPU
...
Commit 6727ad9e20 ("nmi_backtrace: generate one-line reports for idle cpus")
introduced a new text section called cpuidle; with that, we have a mechanism
to add idling functions in such section and skip them from nmi_backtrace
output, since they're useless and potentially flooding for such report.
Happens that inlining might cause some real idle functions to end-up
outside of such section; this is currently the case of ACPI processor_idle
driver; the functions acpi_idle_enter_* do inline acpi_idle_do_entry(),
hence they stay out of the cpuidle section.
Fix that by marking such functions to also live in the cpuidle section.
Fixes: 6727ad9e20 ("nmi_backtrace: generate one-line reports for idle cpus")
Signed-off-by: Guilherme G. Piccoli <gpiccoli@igalia.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Merge general ACPI cleanups and processor support updates for 5.19-rc1:
- Rearrange find_child_checks() to simplify code (Rafael Wysocki).
- Use memremap() to map the UCSI mailbox that is always in main memory
and drop acpi_release_memory() that has no more users (Heikki
Krogerus, Dan Carpenter).
- Make max_cstate/nocst/bm_check_disable processor module parameters
visible in sysfs (Yajun Deng).
- Fix typo in the CPPC driver (Julia Lawall).
* acpi-glue:
ACPI: glue: Rearrange find_child_checks()
* acpi-osl:
usb: typec: ucsi: acpi: fix a NULL vs IS_ERR() check in probe
ACPI: OSL: Remove the helper for deactivating memory region
usb: typec: ucsi: acpi: Map the mailbox with memremap()
* acpi-processor:
ACPI: processor: idle: Expose max_cstate/nocst/bm_check_disable read-only in sysfs
* acpi-cppc:
ACPI: CPPC: fix typo in comment
This will allow super users to verify the module parameters in question
when changed via kernel command line.
The parameters "nocst/bm_check_disable" are only used for enable/disable,
so change them from integer to bool.
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The "safe state" index is used by acpi_idle_enter_bm() to avoid
entering a C-state that may require bus mastering to be disabled
on entry in the cases when this is not going to happen. For this
reason, it should not be set to point to C3 type of C-states, because
they may require bus mastering to be disabled on entry in principle.
This was broken by commit d6b88ce2eb ("ACPI: processor idle: Allow
playing dead in C3 state") which inadvertently allowed the "safe
state" index to point to C3 type of C-states.
This results in a machine that won't boot past the point when it first
enters C3. Restore the correct behaviour (either demote to C1/C2, or
use C3 but also set ARB_DIS=1).
I hit this on a Fujitsu Siemens Lifebook S6010 (P3) machine.
Fixes: d6b88ce2eb ("ACPI: processor idle: Allow playing dead in C3 state")
Cc: 5.16+ <stable@vger.kernel.org> # 5.16+
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Tested-by: Woody Suwalski <wsuwalski@gmail.com>
[ rjw: Subject and changelog adjustments ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add an optional callback needed by some PMU features, e.g., AMD
BRS, to give a chance to the perf_events code to change its state before
a CPU goes to low power and after it comes back.
The callback is void when the PERF_NEEDS_LOPWR_CB flag is not set.
This flag must be set in arch specific perf_event.h header whenever needed.
When not set, there is no impact on the ACPI code.
Signed-off-by: Stephane Eranian <eranian@google.com>
[peterz: build fix]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20220322221517.2510440-9-eranian@google.com
Revert commit 87ebbb8c61 ("ACPI: processor: idle: Only flush cache
on entering C3") that broke the assumptions of the acpi_idle_play_dead()
callers.
Namely, the CPU cache must always be flushed in acpi_idle_play_dead(),
regardless of the target C-state that is going to be requested, because
this is likely to be part of a CPU offline procedure or preparation for
entering a system-wide sleep state and the lack of synchronization
between the CPU cache and RAM may lead to problems going forward, for
example when the CPU is brought back online.
In particular, it breaks resume from suspend-to-RAM on Lenovo ThinkPad
C13 which fails occasionally until the problematic commit is reverted.
Signed-off-by: Akihiko Odaki <akihiko.odaki@gmail.com>
[ rjw: Changelog ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When `osc_pc_lpi_support_confirmed` is set through `_OSC` and `_LPI` is
populated then the cpuidle driver assumes that LPI is fully functional.
However currently the kernel only provides architectural support for LPI
on ARM. This leads to high power consumption on X86 platforms that
otherwise try to enable LPI.
So probe whether or not LPI support is implemented before enabling LPI in
the kernel. This is done by overloading `acpi_processor_ffh_lpi_probe` to
check whether it returns `-EOPNOTSUPP`. It also means that all future
implementations of `acpi_processor_ffh_lpi_probe` will need to follow
these semantics as well.
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add and ACPI idle power level limit for 32-bit ThinkPad T40.
There is a regression on T40 introduced by commit d6b88ce2, starting
with kernel 5.16:
commit d6b88ce2eb
Author: Richard Gong <richard.gong@amd.com>
Date: Wed Sep 22 08:31:16 2021 -0500
ACPI: processor idle: Allow playing dead in C3 state
The above patch is trying to enter C3 state during init, what is causing
a T40 system freeze. I have not found a similar issue on any other of my
32-bit machines.
The fix is to add another exception to the processor_power_dmi_table[] list.
As a result the dmesg shows as expected:
[2.155398] ACPI: IBM ThinkPad T40 detected - limiting to C2 max_cstate. Override with "processor.max_cstate=9"
[2.155404] ACPI: processor limited to max C-state 2
The fix is trivial and affects only vintage T40 systems.
Fixes: d6b88ce2eb ("CPI: processor idle: Allow playing dead in C3 state")
Signed-off-by: Woody Suwalski <wsuwalski@gmail.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Cc: 5.16+ <stable@vger.kernel.org> # 5.16+
[ rjw: New subject ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Merge ACPI EC driver updates and ACPI processor driver updates for
5.17-rc1:
- Rework flushing of EC work while suspended to idle and clean up
the handling of events in the ACPI EC driver (Rafael Wysocki).
- Prohibit ec_sys module parameter write_support from being used
when the system is locked down (Hans de Goede).
- Make the ACPI processor thermal driver use cpufreq_cpu_get() to
check for presence of cpufreq policy (Manfred Spraul).
- Avoid unnecessary CPU cache flushing in the ACPI processor idle
driver (Kirill A. Shutemov).
- Replace kernel.h with the necessary inclusions in the ACPI
processor driver (Andy Shevchenko).
- Use swap() instead of open coding it in the ACPI processor idle
driver (Guo Zhengkui).
* acpi-ec:
ACPI: EC: Mark the ec_sys write_support param as module_param_hw()
ACPI: EC: Relocate acpi_ec_create_query() and drop acpi_ec_delete_query()
ACPI: EC: Make the event work state machine visible
ACPI: EC: Avoid queuing unnecessary work in acpi_ec_submit_event()
ACPI: EC: Rename three functions
ACPI: EC: Simplify locking in acpi_ec_event_handler()
ACPI: EC: Rearrange the loop in acpi_ec_event_handler()
ACPI: EC: Fold acpi_ec_check_event() into acpi_ec_event_handler()
ACPI: EC: Pass one argument to acpi_ec_query()
ACPI: EC: Call advance_transaction() from acpi_ec_dispatch_gpe()
ACPI: EC: Rework flushing of EC work while suspended to idle
* acpi-processor:
ACPI: processor: thermal: avoid cpufreq_get_policy()
ACPI: processor: idle: Only flush cache on entering C3
ACPI: processor idle: Use swap() instead of open coding it
ACPI: processor: Replace kernel.h with the necessary inclusions
