Commit bf5835bcdb ("intel_idle: Disable IBRS during long idle")
disables IBRS when the cstate is 6 or lower. However, there are
some use cases where a customer may want to use max_cstate=1 to
lower latency. Such use cases will suffer from the performance
degradation caused by the enabling of IBRS in the sibling idle thread.
Add a "ibrs_off" module parameter to force disable IBRS and the
CPUIDLE_FLAG_IRQ_ENABLE flag if set.
In the case of a Skylake server with max_cstate=1, this new ibrs_off
option will likely increase the IRQ response latency as IRQ will now
be disabled.
When running SPECjbb2015 with cstates set to C1 on a Skylake system.
First test when the kernel is booted with: "intel_idle.ibrs_off":
max-jOPS = 117828, critical-jOPS = 66047
Then retest when the kernel is booted without the "intel_idle.ibrs_off"
added:
max-jOPS = 116408, critical-jOPS = 58958
That means booting with "intel_idle.ibrs_off" improves performance by:
max-jOPS: +1.2%, which could be considered noise range.
critical-jOPS: +12%, which is definitely a solid improvement.
The admin-guide/pm/intel_idle.rst file is updated to add a description
about the new "ibrs_off" module parameter.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/20230727184600.26768-5-longman@redhat.com
When intel_idle_ibrs() is called, it modifies the SPEC_CTRL MSR to 0
in order disable IBRS. However, the new MSR value isn't reflected in
x86_spec_ctrl_current which is at odd with the other code that keep track
of its state in that percpu variable. Use the new __update_spec_ctrl()
to have the x86_spec_ctrl_current percpu value properly updated.
Since spec-ctrl.h includes both msr.h and nospec-branch.h, we can remove
those from the include file list.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/20230727184600.26768-4-longman@redhat.com
Alderlake N is an E-core only product using Gracemont
micro-architecture. It fits the pre-existing naming scheme perfectly
fine, adhere to it.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20230807150405.686834933@infradead.org
This reverts commit 2f3d08f074 ("intel_idle: Add support for using
intel_idle in a VM guest using just hlt"), because it causes functional
issues to appear and it is not really useful without a related commit
that got reverted previously.
Link: https://lore.kernel.org/linux-pm/5c7de6d5-7706-c4a5-7c41-146db1269aff@intel.com
Reported-by: Xiaoyao Li <xiaoyao.li@intel.com>
Requested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This reverts commit b2918089d5 ("intel_idle: Add __init annotation to
matchup_vm_state_with_baremetal()"), because the commit fixed by it will
be reverted.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The caller of (recently added) matchup_vm_state_with_baremetal() is an
__init function and it uses some __initdata data structures, so add the
__init annotation to it for consistency.
This addresses the following build warnings:
WARNING: modpost: vmlinux: section mismatch in reference: matchup_vm_state_with_baremetal+0x51 (section: .text) -> intel_idle_max_cstate_reached (section: .init.text)
WARNING: modpost: vmlinux: section mismatch in reference: matchup_vm_state_with_baremetal+0x62 (section: .text) -> cpuidle_state_table (section: .init.data)
WARNING: modpost: vmlinux: section mismatch in reference: matchup_vm_state_with_baremetal+0x79 (section: .text) -> icpu (section: .init.data)
Fixes: 0fac214bb7 ("intel_idle: Add a "Long HLT" C1 state for the VM guest mode")
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
intel_idle will, for the bare metal case, usually have one or more deep
power states that have the CPUIDLE_FLAG_TLB_FLUSHED flag set. When
a state with this flag is selected by the cpuidle framework, it will also
flush the TLBs as part of entering this state. The benefit of doing this is
that the kernel does not need to wake the cpu out of this deep power state
just to flush the TLBs... for which the latency can be very high due to
the exit latency of deep power states.
In a VM guest currently, this benefit of avoiding the wakeup does not exist,
while the problem (long exit latency) is even more severe. Linux will need
to wake up a vCPU (causing the host to either come out of a deep C state,
or the VMM to have to deschedule something else to schedule the vCPU) which
can take a very long time.. adding a lot of latency to tlb flush operations
(including munmap and others).
To solve this, add a "Long HLT" C state to the state table for the VM guest
case that has the CPUIDLE_FLAG_TLB_FLUSHED flag set. The result of that is
that for long idle periods (where the VMM is likely to do things that cause
large latency) the cpuidle framework will flush the TLBs (and avoid the
wakeups), while for short/quick idle durations, the existing behavior is
retained.
Now, there is still only "hlt" available in the guest, but for long idle,
the host can go to a deeper state (say C6). There is a reasonable debate
one can have to what to set for the exit_latency and break even point for
this "Long HLT" state. The good news is that intel_idle has these values
available for the underlying CPU (even when mwait is not exposed). The
solution thus is to just use the latency and break even of the deepest state
from the bare metal CPU. This is under the assumption that this is a pretty
reasonable estimate of what the VMM would do to cause latency.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In a typical VM guest, the mwait instruction is not available, leaving
only the 'hlt' instruction (which causes a VMEXIT to the host).
So for this common case, intel_idle will detect the lack of mwait, and
fail to initialize (after which another idle method would step in which
will just use hlt always).
Other (non-common) cases exist; the table below shows the before/after
for these:
+------------+--------------------------+-------------------------+
| Hypervisor | Idle method before patch | Idle method after patch |
| exposes | | |
+============+==========================+=========================+
| nothing | default_idle fallback | intel_idle VM table |
| (common) | (straight "hlt") | |
+------------+--------------------------+-------------------------+
| mwait | intel_idle mwait table | intel_idle mwait table |
+------------+--------------------------+-------------------------+
| ACPI | ACPI C1 state ("hlt") | intel_idle VM table |
+------------+--------------------------+-------------------------+
This is only applicable to CPUs known by intel_idle. For the bare metal
case, unknown CPU models will use the ACPI tables (when available) to
get estimates for latency and break even point for longer idle states.
In guests, the common case is that ACPI tables are not available, but
even when they are available, they can't and don't provide the latency
information for the longer (mwait based) states. For this scenario
(unknown CPU model), the default_idle mode (no ACPI) or ACPI C1 (ACPI
avaible) will be used.
By providing capability to do this with the intel_idle driver, we can
do better than the fallback or ACPI table methods. While this current
change only gets us to the existing behavior, later patches in this
series will add new capabilities such as optimized TLB flushing.
In order to do this, a simplified version of the initialization
function for VM guests is created, and this will be called if the CPU
is recognized, but mwait is not supported, and we're in a VM guest.
One thing to note is that the max latency (and break even) of this C1
state is higher than the typical bare metal C1 state. Because hlt causes
a vmexit, and the cost of vmexit + hypervisor overhead + vmenter is
typically in the order of upto 5 microseconds... even if the hypervisor
does not actually goes into a hardware power saving state.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
[ rjw: Dropped redundant checks from should_verify_mwait() ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Now that the logic for state_update_enter_method() is in its own
function, the long if .. else if .. else if .. else if chain
can be simplified by just returning from the function
at the various places. This does not change functionality,
but it makes the logic much simpler to read or modify later.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since the 6.4 kernel, the logic for updating a state's enter method
based on "environmental conditions" (command line options, cpu sidechannel
workarounds etc etc) has gotten pretty complex.
This patch refactors this into a seperate small, self contained function
(no behavior changes) for improved readability and to make future
changes to this logic easier to do and understand.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The intention is to clean up the code and make it look a bit more
consistent.
Mark all unitialized module parameter variables as __read_mostly,
not just one of them. The other parameters are read-mostly too.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This is a cleanup which improves code consistency. Move the force_irq_on
module parameter variable and definition to the same place where we have
variables and definitions for other module parameters.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
By default, all non-POLL C-states are entered with interrupts disabled.
There are 2 ways to make 'intel_idle' enter C-states with interrupts
enabled:
1. Mark the C-state with the CPUIDLE_FLAG_IRQ_ENABLE flag.
2. Use the force_irq_on module parameter.
The former is the "proper" way of doing it, it is per-C-state and
per-platform. The latter is for debugging purposes only.
The problem is that intel_idle prints the "forced intel_idle_irq"
message in both cases, even though the former case does not needed
this message, because nothing is forced there. This patch addresses the
problem.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The following C-state flags are currently mutually-exclusive and should not
be combined:
* IRQ_ENABLE
* IBRS
* XSTATE
There is a warning for the situation when the IRQ_ENABLE flag
is combined with the IBRS flag, but no warnings for other combinations.
This is inconsistent and prone to errors.
Improve the situation by adding warnings for all the unexpected
combinations. Add a couple of helpful commentaries too.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Introduce a temporary 'state' variable for referencing the currently
processed C-state in the intel_idle_init_cstates_icpu() function.
This makes code lines shorter and easier to read.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The intel_idle_init_cstates_icpu() function includes a loop that iterates
over every C-state. Inside the loop, the same C-state data is referenced 2
ways:
1. as cpuidle_state_table[cstate]
2. as drv->states[drv->state_count] (but it is a copy of #1, not the same
object).
Make the code be more consistent and easier to read by using only the 2nd
way. So the code structure would be as follows:
1. Use cpuidle_state_table[cstate]
2. Copy cpuidle_state_table[cstate] to drv->states[drv->state_count]
3. Use only drv->states[drv->state_count] from this point.
Note, this change introduces a checkpatch.pl warning (too long line), but it
will be addressed in the next patch.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Substitute 'printk()' with 'pr_info()', because 'intel_idle' already uses
'pr_debug()', so using 'pr_info()' will be more consistent.
In addition to this, this patch addresses the following checkpatch.pl
warning:
WARNING: printk() should include KERN_<LEVEL> facility level
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- Add EPP support to the AMD P-state cpufreq driver (Perry Yuan, Wyes
Karny, Arnd Bergmann, Bagas Sanjaya).
- Drop the custom cpufreq driver for loongson1 that is not necessary
any more and the corresponding cpufreq platform device (Keguang
Zhang).
- Remove "select SRCU" from system sleep, cpufreq and OPP Kconfig
entries (Paul E. McKenney).
- Enable thermal cooling for Tegra194 (Yi-Wei Wang).
- Register module device table and add missing compatibles for
cpufreq-qcom-hw (Nícolas F. R. A. Prado, Abel Vesa and Luca Weiss).
- Various dt binding updates for qcom-cpufreq-nvmem and opp-v2-kryo-cpu
(Christian Marangi).
- Make kobj_type structure in the cpufreq core constant (Thomas
Weißschuh).
- Make cpufreq_unregister_driver() return void (Uwe Kleine-König).
- Make the TEO cpuidle governor check CPU utilization in order to refine
idle state selection (Kajetan Puchalski).
- Make Kconfig select the haltpoll cpuidle governor when the haltpoll
cpuidle driver is selected and replace a default_idle() call in that
driver with arch_cpu_idle() to allow MWAIT to be used (Li RongQing).
- Add Emerald Rapids Xeon support to the intel_idle driver (Artem
Bityutskiy).
- Add ARCH_SUSPEND_POSSIBLE dependencies for ARMv4 cpuidle drivers to
avoid randconfig build failures (Arnd Bergmann).
- Make kobj_type structures used in the cpuidle sysfs interface
constant (Thomas Weißschuh).
- Make the cpuidle driver registration code update microsecond values
of idle state parameters in accordance with their nanosecond values
if they are provided (Rafael Wysocki).
- Make the PSCI cpuidle driver prevent topology CPUs from being
suspended on PREEMPT_RT (Krzysztof Kozlowski).
- Document that pm_runtime_force_suspend() cannot be used with
DPM_FLAG_SMART_SUSPEND (Richard Fitzgerald).
- Add EXPORT macros for exporting PM functions from drivers (Richard
Fitzgerald).
- Remove /** from non-kernel-doc comments in hibernation code (Randy
Dunlap).
- Fix possible name leak in powercap_register_zone() (Yang Yingliang).
- Add Meteor Lake and Emerald Rapids support to the intel_rapl power
capping driver (Zhang Rui).
- Modify the idle_inject power capping facility to support 100% idle
injection (Srinivas Pandruvada).
- Fix large time windows handling in the intel_rapl power capping
driver (Zhang Rui).
- Fix memory leaks with using debugfs_lookup() in the generic PM
domains and Energy Model code (Greg Kroah-Hartman).
- Add missing 'cache-unified' property in the example for kryo OPP
bindings (Rob Herring).
- Fix error checking in opp_migrate_dentry() (Qi Zheng).
- Let qcom,opp-fuse-level be a 2-long array for qcom SoCs (Konrad
Dybcio).
- Modify some power management utilities to use the canonical ftrace
path (Ross Zwisler).
- Correct spelling problems for Documentation/power/ as reported by
codespell (Randy Dunlap).
-----BEGIN PGP SIGNATURE-----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=lZLx
-----END PGP SIGNATURE-----
Merge tag 'pm-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"These add EPP support to the AMD P-state cpufreq driver, add support
for new platforms to the Intel RAPL power capping driver, intel_idle
and the Qualcomm cpufreq driver, enable thermal cooling for Tegra194,
drop the custom cpufreq driver for loongson1 that is not necessary any
more (and the corresponding cpufreq platform device), fix assorted
issues and clean up code.
Specifics:
- Add EPP support to the AMD P-state cpufreq driver (Perry Yuan, Wyes
Karny, Arnd Bergmann, Bagas Sanjaya)
- Drop the custom cpufreq driver for loongson1 that is not necessary
any more and the corresponding cpufreq platform device (Keguang
Zhang)
- Remove "select SRCU" from system sleep, cpufreq and OPP Kconfig
entries (Paul E. McKenney)
- Enable thermal cooling for Tegra194 (Yi-Wei Wang)
- Register module device table and add missing compatibles for
cpufreq-qcom-hw (Nícolas F. R. A. Prado, Abel Vesa and Luca Weiss)
- Various dt binding updates for qcom-cpufreq-nvmem and
opp-v2-kryo-cpu (Christian Marangi)
- Make kobj_type structure in the cpufreq core constant (Thomas
Weißschuh)
- Make cpufreq_unregister_driver() return void (Uwe Kleine-König)
- Make the TEO cpuidle governor check CPU utilization in order to
refine idle state selection (Kajetan Puchalski)
- Make Kconfig select the haltpoll cpuidle governor when the haltpoll
cpuidle driver is selected and replace a default_idle() call in
that driver with arch_cpu_idle() to allow MWAIT to be used (Li
RongQing)
- Add Emerald Rapids Xeon support to the intel_idle driver (Artem
Bityutskiy)
- Add ARCH_SUSPEND_POSSIBLE dependencies for ARMv4 cpuidle drivers to
avoid randconfig build failures (Arnd Bergmann)
- Make kobj_type structures used in the cpuidle sysfs interface
constant (Thomas Weißschuh)
- Make the cpuidle driver registration code update microsecond values
of idle state parameters in accordance with their nanosecond values
if they are provided (Rafael Wysocki)
- Make the PSCI cpuidle driver prevent topology CPUs from being
suspended on PREEMPT_RT (Krzysztof Kozlowski)
- Document that pm_runtime_force_suspend() cannot be used with
DPM_FLAG_SMART_SUSPEND (Richard Fitzgerald)
- Add EXPORT macros for exporting PM functions from drivers (Richard
Fitzgerald)
- Remove /** from non-kernel-doc comments in hibernation code (Randy
Dunlap)
- Fix possible name leak in powercap_register_zone() (Yang Yingliang)
- Add Meteor Lake and Emerald Rapids support to the intel_rapl power
capping driver (Zhang Rui)
- Modify the idle_inject power capping facility to support 100% idle
injection (Srinivas Pandruvada)
- Fix large time windows handling in the intel_rapl power capping
driver (Zhang Rui)
- Fix memory leaks with using debugfs_lookup() in the generic PM
domains and Energy Model code (Greg Kroah-Hartman)
- Add missing 'cache-unified' property in the example for kryo OPP
bindings (Rob Herring)
- Fix error checking in opp_migrate_dentry() (Qi Zheng)
- Let qcom,opp-fuse-level be a 2-long array for qcom SoCs (Konrad
Dybcio)
- Modify some power management utilities to use the canonical ftrace
path (Ross Zwisler)
- Correct spelling problems for Documentation/power/ as reported by
codespell (Randy Dunlap)"
* tag 'pm-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (53 commits)
Documentation: amd-pstate: disambiguate user space sections
cpufreq: amd-pstate: Fix invalid write to MSR_AMD_CPPC_REQ
dt-bindings: opp: opp-v2-kryo-cpu: enlarge opp-supported-hw maximum
dt-bindings: cpufreq: qcom-cpufreq-nvmem: make cpr bindings optional
dt-bindings: cpufreq: qcom-cpufreq-nvmem: specify supported opp tables
PM: Add EXPORT macros for exporting PM functions
cpuidle: psci: Do not suspend topology CPUs on PREEMPT_RT
MIPS: loongson32: Drop obsolete cpufreq platform device
powercap: intel_rapl: Fix handling for large time window
cpuidle: driver: Update microsecond values of state parameters as needed
cpuidle: sysfs: make kobj_type structures constant
cpuidle: add ARCH_SUSPEND_POSSIBLE dependencies
PM: EM: fix memory leak with using debugfs_lookup()
PM: domains: fix memory leak with using debugfs_lookup()
cpufreq: Make kobj_type structure constant
cpufreq: davinci: Fix clk use after free
cpufreq: amd-pstate: avoid uninitialized variable use
cpufreq: Make cpufreq_unregister_driver() return void
OPP: fix error checking in opp_migrate_dentry()
dt-bindings: cpufreq: cpufreq-qcom-hw: Add SM8550 compatible
...
Emerald Rapids (EMR) is the next Intel Xeon processor after Sapphire
Rapids (SPR).
EMR C-states are the same as SPR C-states, and we expect that EMR
C-state characteristics (latency and target residency) will be the
same as in SPR. Therefore, add EMR support by using SPR C-states table.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
So objtool found this bug:
vmlinux.o: warning: objtool: intel_idle_irq+0x10c: call to trace_hardirqs_off() leaves .noinstr.text section
As per commit 32d4fd5751 ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE"):
"must not have tracing in idle functions"
Clearly people can't read and tinker along until splat dissapears.
This straight up reverts commit d295ad34f2 ("intel_idle: Fix false
positive RCU splats due to incorrect hardirqs state").
It doesn't re-introduce the problem because preceding patches fixed it
properly.
Fixes: d295ad34f2 ("intel_idle: Fix false positive RCU splats due to incorrect hardirqs state")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Tony Lindgren <tony@atomide.com>
Tested-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20230112195540.434302128@infradead.org
Similar to the other other AlderLake platforms, the C1 and C1E states on
ADL-N are mutually exclusive. Only one of them can be enabled at a time.
C1E is preferred on ADL-N for better energy efficiency.
C6S is also supported on this platform. Its latency is far bigger than
C6, but really close to C8 (PC8), thus it is not exposed as a separate
state.
Suggested-by: Baieswara Reddy Sagili <baieswara.reddy.sagili@intel.com>
Suggested-by: Vinay Kumar <vinay.kumar@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Follow the advice of the below link and prefer 'strscpy' in this
subsystem. Conversion is 1:1 because the return value is not used.
Generated by a coccinelle script.
Link: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
- Make cpufreq_show_cpus() more straightforward (Viresh Kumar).
- Drop unnecessary CPU hotplug locking from store() used by cpufreq
sysfs attributes (Viresh Kumar).
- Make the ACPI cpufreq driver support the boost control interface on
Zhaoxin/Centaur processors (Tony W Wang-oc).
- Print a warning message on attempts to free an active cpufreq policy
which should never happen (Viresh Kumar).
- Fix grammar in the Kconfig help text for the loongson2 cpufreq
driver (Randy Dunlap).
- Use cpumask_var_t for an on-stack CPU mask in the ondemand cpufreq
governor (Zhao Liu).
- Add trace points for guest_halt_poll_ns grow/shrink to the haltpoll
cpuidle driver (Eiichi Tsukata).
- Modify intel_idle to treat C1 and C1E as independent idle states on
Sapphire Rapids (Artem Bityutskiy).
- Extend support for wakeirq to callback wrappers used during system
suspend and resume (Ulf Hansson).
- Defer waiting for device probe before loading a hibernation image
till the first actual device access to avoid possible deadlocks
reported by syzbot (Tetsuo Handa).
- Unify device_init_wakeup() for PM_SLEEP and !PM_SLEEP (Bjorn
Helgaas).
- Add Raptor Lake-P to the list of processors supported by the Intel
RAPL driver (George D Sworo).
- Add Alder Lake-N and Raptor Lake-P to the list of processors for
which Power Limit4 is supported in the Intel RAPL driver (Sumeet
Pawnikar).
- Make pm_genpd_remove() check genpd_debugfs_dir against NULL before
attempting to remove it (Hsin-Yi Wang).
- Change the Energy Model code to represent power in micro-Watts and
adjust its users accordingly (Lukasz Luba).
- Add new devfreq driver for Mediatek CCI (Cache Coherent
Interconnect) (Johnson Wang).
- Convert the Samsung Exynos SoC Bus bindings to DT schema of
exynos-bus.c (Krzysztof Kozlowski).
- Address kernel-doc warnings by adding the description for unused
fucntion parameters in devfreq core (Mauro Carvalho Chehab).
- Use NULL to pass a null pointer rather than zero according to the
function propotype in imx-bus.c (Colin Ian King).
- Print error message instead of error interger value in
tegra30-devfreq.c (Dmitry Osipenko).
- Add checks to prevent setting negative frequency QoS limits for
CPUs (Shivnandan Kumar).
- Update the pm-graph suite of utilities to the latest revision 5.9
including multiple improvements (Todd Brandt).
- Drop pme_interrupt reference from the PCI power management
documentation (Mario Limonciello).
-----BEGIN PGP SIGNATURE-----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=XeLV
-----END PGP SIGNATURE-----
Merge tag 'pm-5.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"These are mostly minor improvements all over including new CPU IDs for
the Intel RAPL driver, an Energy Model rework to use micro-Watt as the
power unit, cpufreq fixes and cleanus, cpuidle updates, devfreq
updates, documentation cleanups and a new version of the pm-graph
suite of utilities.
Specifics:
- Make cpufreq_show_cpus() more straightforward (Viresh Kumar).
- Drop unnecessary CPU hotplug locking from store() used by cpufreq
sysfs attributes (Viresh Kumar).
- Make the ACPI cpufreq driver support the boost control interface on
Zhaoxin/Centaur processors (Tony W Wang-oc).
- Print a warning message on attempts to free an active cpufreq
policy which should never happen (Viresh Kumar).
- Fix grammar in the Kconfig help text for the loongson2 cpufreq
driver (Randy Dunlap).
- Use cpumask_var_t for an on-stack CPU mask in the ondemand cpufreq
governor (Zhao Liu).
- Add trace points for guest_halt_poll_ns grow/shrink to the haltpoll
cpuidle driver (Eiichi Tsukata).
- Modify intel_idle to treat C1 and C1E as independent idle states on
Sapphire Rapids (Artem Bityutskiy).
- Extend support for wakeirq to callback wrappers used during system
suspend and resume (Ulf Hansson).
- Defer waiting for device probe before loading a hibernation image
till the first actual device access to avoid possible deadlocks
reported by syzbot (Tetsuo Handa).
- Unify device_init_wakeup() for PM_SLEEP and !PM_SLEEP (Bjorn
Helgaas).
- Add Raptor Lake-P to the list of processors supported by the Intel
RAPL driver (George D Sworo).
- Add Alder Lake-N and Raptor Lake-P to the list of processors for
which Power Limit4 is supported in the Intel RAPL driver (Sumeet
Pawnikar).
- Make pm_genpd_remove() check genpd_debugfs_dir against NULL before
attempting to remove it (Hsin-Yi Wang).
- Change the Energy Model code to represent power in micro-Watts and
adjust its users accordingly (Lukasz Luba).
- Add new devfreq driver for Mediatek CCI (Cache Coherent
Interconnect) (Johnson Wang).
- Convert the Samsung Exynos SoC Bus bindings to DT schema of
exynos-bus.c (Krzysztof Kozlowski).
- Address kernel-doc warnings by adding the description for unused
function parameters in devfreq core (Mauro Carvalho Chehab).
- Use NULL to pass a null pointer rather than zero according to the
function propotype in imx-bus.c (Colin Ian King).
- Print error message instead of error interger value in
tegra30-devfreq.c (Dmitry Osipenko).
- Add checks to prevent setting negative frequency QoS limits for
CPUs (Shivnandan Kumar).
- Update the pm-graph suite of utilities to the latest revision 5.9
including multiple improvements (Todd Brandt).
- Drop pme_interrupt reference from the PCI power management
documentation (Mario Limonciello)"
* tag 'pm-5.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (27 commits)
powercap: RAPL: Add Power Limit4 support for Alder Lake-N and Raptor Lake-P
PM: QoS: Add check to make sure CPU freq is non-negative
PM: hibernate: defer device probing when resuming from hibernation
intel_idle: make SPR C1 and C1E be independent
cpufreq: ondemand: Use cpumask_var_t for on-stack cpu mask
cpufreq: loongson2: fix Kconfig "its" grammar
pm-graph v5.9
cpufreq: Warn users while freeing active policy
cpufreq: scmi: Support the power scale in micro-Watts in SCMI v3.1
firmware: arm_scmi: Get detailed power scale from perf
Documentation: EM: Switch to micro-Watts scale
PM: EM: convert power field to micro-Watts precision and align drivers
PM / devfreq: tegra30: Add error message for devm_devfreq_add_device()
PM / devfreq: imx-bus: use NULL to pass a null pointer rather than zero
PM / devfreq: shut up kernel-doc warnings
dt-bindings: interconnect: samsung,exynos-bus: convert to dtschema
PM / devfreq: mediatek: Introduce MediaTek CCI devfreq driver
dt-bindings: interconnect: Add MediaTek CCI dt-bindings
PM: domains: Ensure genpd_debugfs_dir exists before remove
PM: runtime: Extend support for wakeirq for force_suspend|resume
...
be able to enter deeper low-power state
-----BEGIN PGP SIGNATURE-----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=w/wH
-----END PGP SIGNATURE-----
Merge tag 'x86_fpu_for_v6.0_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fpu update from Borislav Petkov:
- Add machinery to initialize AMX register state in order for
AMX-capable CPUs to be able to enter deeper low-power state
* tag 'x86_fpu_for_v6.0_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
intel_idle: Add a new flag to initialize the AMX state
x86/fpu: Add a helper to prepare AMX state for low-power CPU idle
This patch partially reverts the changes made by the following commit:
da0e58c038 intel_idle: add 'preferred_cstates' module argument
As that commit describes, on early Sapphire Rapids Xeon platforms the C1 and
C1E states were mutually exclusive, so that users could only have either C1 and
C6, or C1E and C6.
However, Intel firmware engineers managed to remove this limitation and make C1
and C1E to be completely independent, just like on previous Xeon platforms.
Therefore, this patch:
* Removes commentary describing the old, and now non-existing SPR C1E
limitation.
* Marks SPR C1E as available by default.
* Removes the 'preferred_cstates' parameter handling for SPR. Both C1 and
C1E will be available regardless of 'preferred_cstates' value.
We expect that all SPR systems are shipping with new firmware, which includes
the C1/C1E improvement.
Cc: v5.18+ <stable@vger.kernel.org> # v5.18+
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Commit 32d4fd5751 ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
uses raw_local_irq_enable/local_irq_disable() around call to
__intel_idle() in intel_idle_irq().
With interrupt enabled, timer tick interrupt can happen and a
subsequently call to __do_softirq() may change the lockdep hardirqs state
of a debug kernel back to 'on'. This will result in a mismatch between
the cpu hardirqs state (off) and the lockdep hardirqs state (on) causing
a number of false positive "WARNING: suspicious RCU usage" splats.
Fix that by using local_irq_disable() to disable interrupt in
intel_idle_irq().
Fixes: 32d4fd5751 ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
Signed-off-by: Waiman Long <longman@redhat.com>
Cc: 5.16+ <stable@vger.kernel.org> # 5.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The non-initialized AMX state can be the cause of C-state demotion from C6
to C1E. This low-power idle state may improve power savings and thus result
in a higher available turbo frequency budget.
This behavior is implementation-specific. Initialize the state for the C6
entrance of Sapphire Rapids as needed.
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Zhang Rui <rui.zhang@intel.com>
Link: https://lkml.kernel.org/r/20220614164116.5196-1-chang.seok.bae@intel.com
Having IBRS enabled while the SMT sibling is idle unnecessarily slows
down the running sibling. OTOH, disabling IBRS around idle takes two
MSR writes, which will increase the idle latency.
Therefore, only disable IBRS around deeper idle states. Shallow idle
states are bounded by the tick in duration, since NOHZ is not allowed
for them by virtue of their short target residency.
Only do this for mwait-driven idle, since that keeps interrupts disabled
across idle, which makes disabling IBRS vs IRQ-entry a non-issue.
Note: C6 is a random threshold, most importantly C1 probably shouldn't
disable IBRS, benchmarking needed.
Suggested-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Josh Poimboeuf <jpoimboe@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Commit c227233ad6 ("intel_idle: enable interrupts before C1 on
Xeons") wrecked intel_idle in two ways:
- must not have tracing in idle functions
- must return with IRQs disabled
Additionally, it added a branch for no good reason.
Fixes: c227233ad6 ("intel_idle: enable interrupts before C1 on Xeons")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[ rjw: Moved the intel_idle() kerneldoc comment next to the function ]
Cc: 5.16+ <stable@vger.kernel.org> # 5.16+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Similar to SPR, the C1 and C1E states on ADL are mutually exclusive.
Only one of them can be enabled at a time.
But contrast to SPR, which usually has a strong latency requirement
as a Xeon processor, C1E is preferred on ADL for better energy
efficiency.
Add custom C-state tables for ADL with both C1 and C1E, and
1. Enable the "C1E promotion" bit in MSR_IA32_POWER_CTL and mark C1
with the CPUIDLE_FLAG_UNUSABLE flag, so C1 is not available by
default.
2. Add support for the "preferred_cstates" module parameter, so that
users can choose to use C1 instead of C1E by booting with
"intel_idle.preferred_cstates=2".
Separate custom C-state tables are introduced for the ADL mobile and
desktop processors, because of the exit latency differences between
these two variants, especially with respect to PC10.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
[ rjw: Changelog edits, code rearrangement ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The Sapphire Rapids (SPR) C6 optimization was added to the end of the
'spr_idle_state_table_update()' function. However, the function has a
'return' which may happen before the optimization has a chance to run.
And this may prevent the optimization from happening.
This is an unlikely scenario, but possible if user boots with, say,
the 'intel_idle.preferred_cstates=6' kernel boot option.
This patch fixes the issue by eliminating the problematic 'return'
statement.
Fixes: 3a9cf77b60 ("intel_idle: add core C6 optimization for SPR")
Suggested-by: Jan Beulich <jbeulich@suse.com>
Reported-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
[ rjw: Minor changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Problem description.
When user boots kernel up with the 'intel_idle.preferred_cstates=4' option,
we enable C1E and disable C1 states on Sapphire Rapids Xeon (SPR). In order
for C1E to work on SPR, we have to enable the C1E promotion bit on all
CPUs. However, we enable it only on one CPU.
Fix description.
The 'intel_idle' driver already has the infrastructure for disabling C1E
promotion on every CPU. This patch uses the same infrastructure for
enabling C1E promotion on every CPU. It changes the boolean
'disable_promotion_to_c1e' variable to a tri-state 'c1e_promotion'
variable.
Tested on a 2-socket SPR system. I verified the following combinations:
* C1E promotion enabled and disabled in BIOS.
* Booted with and without the 'intel_idle.preferred_cstates=4' kernel
argument.
In all 4 cases C1E promotion was correctly set on all CPUs.
Also tested on an old Broadwell system, just to make sure it does not cause
a regression. C1E promotion was correctly disabled on that system, both C1
and C1E were exposed (as expected).
Fixes: da0e58c038 ("intel_idle: add 'preferred_cstates' module argument")
Reported-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
[ rjw: Minor changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Drop a redundant backslash character at the end of a line in the
spr_cstates[] definition.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Commit bf9282dc26 ("cpuidle: Make CPUIDLE_FLAG_TLB_FLUSHED generic")
moved the leave_mm() call away from intel_idle(), but it didn't update
its kerneldoc comment accordingly, so do that now.
Fixes: bf9282dc26 ("cpuidle: Make CPUIDLE_FLAG_TLB_FLUSHED generic")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add a Sapphire Rapids Xeon C6 optimization, similar to what we have for Sky Lake
Xeon: if package C6 is disabled, adjust C6 exit latency and target residency to
match core C6 values, instead of using the default package C6 values.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
On Sapphire Rapids Xeon (SPR) the C1 and C1E states are basically mutually
exclusive - only one of them can be enabled. By default, 'intel_idle' driver
enables C1 and disables C1E. However, some users prefer to use C1E instead of
C1, because it saves more energy.
This patch adds a new module parameter ('preferred_cstates') for enabling C1E
and disabling C1. Here is the idea behind it.
1. This option has effect only for "mutually exclusive" C-states like C1 and
C1E on SPR.
2. It does not have any effect on independent C-states, which do not require
other C-states to be disabled (most states on most platforms as of today).
3. For mutually exclusive C-states, the 'intel_idle' driver always has a
reasonable default, such as enabling C1 on SPR by default. On other
platforms, the default may be different.
4. Users can override the default using the 'preferred_cstates' parameter.
5. The parameter accepts the preferred C-states bit-mask, similarly to the
existing 'states_off' parameter.
6. This parameter is not limited to C1/C1E, and leaves room for supporting
other mutually exclusive C-states, if they come in the future.
Today 'intel_idle' can only be compiled-in, which means that on SPR, in order
to disable C1 and enable C1E, users should boot with the following kernel
argument: intel_idle.preferred_cstates=4
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add Sapphire Rapids Xeon support.
Up until very recently, the C1 and C1E C-states were independent, but this
has changed in some new chips, including Sapphire Rapids Xeon (SPR). In these
chips the C1 and C1E states cannot be enabled at the same time. The "C1E
promotion" bit in 'MSR_IA32_POWER_CTL' also has its semantics changed a bit.
Here are the C1, C1E, and "C1E promotion" bit rules on Xeons before SPR.
1. If C1E promotion bit is disabled.
a. C1 requests end up with C1 C-state.
b. C1E requests end up with C1E C-state.
2. If C1E promotion bit is enabled.
a. C1 requests end up with C1E C-state.
b. C1E requests end up with C1E C-state.
Here are the C1, C1E, and "C1E promotion" bit rules on Sapphire Rapids Xeon.
1. If C1E promotion bit is disabled.
a. C1 requests end up with C1 C-state.
b. C1E requests end up with C1 C-state.
2. If C1E promotion bit is enabled.
a. C1 requests end up with C1E C-state.
b. C1E requests end up with C1E C-state.
Before SPR Xeon, the 'intel_idle' driver was disabling C1E promotion and was
exposing C1 and C1E as independent C-states. But on SPR, C1 and C1E cannot be
enabled at the same time.
This patch adds both C1 and C1E states. However, C1E is marked as with the
"CPUIDLE_FLAG_UNUSABLE" flag, which means that in won't be registered by
default. The C1E promotion bit will be cleared, which means that by default
only C1 and C6 will be registered on SPR.
The next patch will add an option for enabling C1E and disabling C1 on SPR.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Enable local interrupts before requesting C1 on the last two generations
of Intel Xeon platforms: Sky Lake, Cascade Lake, Cooper Lake, Ice Lake.
This decreases average C1 interrupt latency by about 5-10%, as measured
with the 'wult' tool.
The '->enter()' function of the driver enters C-states with local
interrupts disabled by executing the 'monitor' and 'mwait' pair of
instructions. If an interrupt happens, the CPU exits the C-state and
continues executing instructions after 'mwait'. It does not jump to
the interrupt handler, because local interrupts are disabled. The
cpuidle subsystem enables interrupts a bit later, after doing some
housekeeping.
With this patch, we enable local interrupts before requesting C1. In
this case, if the CPU wakes up because of an interrupt, it will jump
to the interrupt handler right away. The cpuidle housekeeping will be
done after the pending interrupt(s) are handled.
Enabling interrupts before entering a C-state has measurable impact
for faster C-states, like C1. Deeper, but slower C-states like C6 do
not really benefit from this sort of change, because their latency is
a lot higher comparing to the delay added by cpuidle housekeeping.
This change was also tested with cyclictest and dbench. In case of Ice
Lake, the average cyclictest latency decreased by 5.1%, and the average
'dbench' throughput increased by about 0.8%. Both tests were run for 4
hours with only C1 enabled (all other idle states, including 'POLL',
were disabled). CPU frequency was pinned to HFM, and uncore frequency
was pinned to the maximum value. The other platforms had similar
single-digit percentage improvements.
It is worth noting that this patch affects 'cpuidle' statistics a tiny
bit. Before this patch, C1 residency did not include the interrupt
handling time, but with this patch, it will include it. This is similar
to what happens in case of the 'POLL' state, which also runs with
interrupts enabled.
Suggested-by: Len Brown <len.brown@intel.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Because cpuidle assumes worst-case C-state parameters, PC6 parameters
are used for describing C6, which is worst-case for requesting CC6.
When PC6 is enabled, this is appropriate. But if PC6 is disabled
in the BIOS, the exit latency and target residency should be adjusted
accordingly.
Exit latency:
Previously the C6 exit latency was measured as the PC6 exit latency.
With PC6 disabled, the C6 exit latency should be the one of CC6.
Target residency:
With PC6 disabled, the idle duration within [CC6, PC6) would make the
idle governor choose C1E over C6. This would cause low energy-efficiency.
We should lower the bar to request C6 when PC6 is disabled.
To fill this gap, check if PC6 is disabled in the BIOS in the
MSR_PKG_CST_CONFIG_CONTROL(0xe2) register. If so, use the CC6 exit latency
for C6 and set target_residency to 3 times of the new exit latency. [This
is consistent with how intel_idle driver uses _CST to calculate the
target_residency.] As a result, the OS would be more likely to choose C6
over C1E when PC6 is disabled, which is reasonable, because if C6 is
enabled, it implies that the user cares about energy, so choosing C6 more
frequently makes sense.
The new CC6 exit latency of 92us was measured with wult[1] on SKX via NIC
wakeup as the 99.99th percentile. Also CLX and CPX both have the same CPU
model number as SkX, but their CC6 exit latencies are similar to the SKX
one, 96us and 89us respectively, so reuse the SKX value for them.
There is a concern that it might be better to use a more generic approach
instead of optimizing every platform. However, if the required code
complexity and different PC6 bit interpretation on different platforms
are taken into account, tuning the code per platform seems to be an
acceptable tradeoff.
Link: https://intel.github.io/wult/ # [1]
Suggested-by: Len Brown <len.brown@intel.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Reviewed-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
This patch adds Icelake Xeon D support to the intel_idle driver.
Since Icelake D and Icelake SP C-state characteristics the same,
we use Icelake SP C-states table for Icelake D as well.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Acked-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Change IceLake Xeon C6 latency from 128 us to 170 us. The latency
was measured with the "wult" tool and corresponds to the 99.99th
percentile when measuring with the "nic" method. Note, the 128 us
figure correspond to the median latency, but in intel_idle we use
the "worst case" latency figure instead.
C6 target residency was increased from 384 us to 600 us, which may
result in less C6 residency in some workloads. This value was tested
and compared to values 384, and 1000. Value 600 is a reasonable
tradeoff between power and performance.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Acked-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Defining DEBUG should only be done in development.
So remove DEBUG.
Signed-off-by: Tom Rix <trix@redhat.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Add C-state table for the SnowRidge SoC which is found on Intel Jacobsville
platforms.
The following has been changed.
1. C1E latency changed from 10us to 15us. It was measured using the
open source "wult" tool (the "nic" method, 15us is the 99.99th
percentile).
2. C1E power break even changed from 20us to 25us, which may result
in less C1E residency in some workloads.
3. C6 latency changed from 50us to 130us. Measured the same way as C1E.
The C6 C-state is supported only by some SnowRidge revisions, so add a C-state
table commentary about this.
On SnowRidge, C6 support is enumerated via the usual mechanism: "mwait" leaf of
the "cpuid" instruction. The 'intel_idle' driver does check this leaf, so even
though C6 is present in the table, the driver will only use it if the CPU does
support it.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpuidle->enter() callbacks should not call into tracing because RCU
has already been disabled. Instead of doing the broadcast thing
itself, simply advertise to the cpuidle core that those states stop
the timer.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/20201123143510.GR3021@hirez.programming.kicks-ass.net
