| Age | Commit message (Collapse) | Author |
|---|
|
Combine the ->stop_cpu() and ->offline() callback routines for
intel_pstate in the active mode so as to avoid setting the
->stop_cpu callback pointer which is going to be dropped from
the framework.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
|
|
One of the previous commits introducing hybrid processor support to
intel_pstate broke build with CONFIG_ACPI unset.
Fix that and while at it make empty stubs of two functions related
to ACPI CPPC static inline and fix a spelling mistake in the name of
one of them.
Fixes: eb3693f0521e ("cpufreq: intel_pstate: hybrid: CPU-specific scaling factor")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
|
|
Users may disable HWP in firmware, in which case intel_pstate wouldn't load
unless the CPU model is explicitly supported.
See also commit d8de7a44e11f ("cpufreq: intel_pstate: Add Skylake servers
support").
Suggested-by: Doug Smythies <dsmythies@telus.net>
Tested-by: Doug Smythies <dsmythies@telus.net>
Signed-off-by: Giovanni Gherdovich <ggherdovich@suse.cz>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Users may disable HWP in firmware, in which case intel_pstate wouldn't load
unless the CPU model is explicitly supported.
Add ICELAKE_X to the list of CPUs that can register intel_pstate while not
advertising the HWP capability. Without this change, an ICELAKE_X in no-HWP
mode could only use the acpi_cpufreq frequency scaling driver.
See also commit d8de7a44e11f ("cpufreq: intel_pstate: Add Skylake servers
support").
Signed-off-by: Giovanni Gherdovich <ggherdovich@suse.cz>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
The scaling factor between HWP performance levels and CPU frequency
may be different for different types of CPUs in a hybrid processor
and in general the HWP performance levels need not correspond to
"P-states" representing values that would be written to
MSR_IA32_PERF_CTL if HWP was disabled.
However, the policy limits control in cpufreq is defined in terms
of CPU frequency, so it is necessary to map the frequency limits set
through that interface to HWP performance levels with reasonable
accuracy and the behavior of that interface on hybrid processors
has to be compatible with its behavior on non-hybrid ones.
To address this problem, use the observations that (1) on hybrid
processors the sysfs interface can operate by mapping frequency
to "P-states" and translating those "P-states" to specific HWP
performance levels of the given CPU and (2) the scaling factor
between the MSR_IA32_PERF_CTL "P-states" and CPU frequency can be
regarded as a known value. Moreover, the mapping between the
HWP performance levels and CPU frequency can be assumed to be
linear and such that HWP performance level 0 correspond to the
frequency value of 0, so it is only necessary to know the
frequency corresponding to one specific HWP performance level
to compute the scaling factor applicable to all of them.
One possibility is to take the nominal performance value from CPPC,
if available, and use cpu_khz as the corresponding frequency. If
the CPPC capabilities interface is not there or the nominal
performance value provided by it is out of range, though, something
else needs to be done.
Namely, the guaranteed performance level either from CPPC or from
MSR_HWP_CAPABILITIES can be used instead, but the corresponding
frequency needs to be determined. That can be done by computing the
product of the (known) scaling factor between the MSR_IA32_PERF_CTL
P-states and CPU frequency (the PERF_CTL scaling factor) and the
P-state value referred to as the "TDP ratio".
If the HWP-to-frequency scaling factor value obtained in one of the
ways above turns out to be euqal to the PERF_CTL scaling factor, it
can be assumed that the number of HWP performance levels is equal to
the number of P-states and the given CPU can be handled as though
this was not a hybrid processor.
Otherwise, one more adjustment may still need to be made, because the
HWP-to-frequency scaling factor computed so far may not be accurate
enough (e.g. because the CPPC information does not match the exact
behavior of the processor). Specifically, in that case the frequency
corresponding to the highest HWP performance value from
MSR_HWP_CAPABILITIES (computed as the product of that value and the
HWP-to-frequency scaling factor) cannot exceed the frequency that
corresponds to the maximum 1-core turbo P-state value from
MSR_TURBO_RATIO_LIMIT (computed as the procuct of that value and the
PERF_CTL scaling factor) and the HWP-to-frequency scaling factor may
need to be adjusted accordingly.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
The turbo_pct and num_pstates sysfs attributes represent CPU
properties that may be different for differenty types of CPUs in
a hybrid processor, so avoid exposing them in that case.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
It turns out that there are systems where HWP is enabled during
initialization by the platform firmware (BIOS), but HWP EPP support
is not advertised.
After commit 7aa1031223bc ("cpufreq: intel_pstate: Avoid enabling HWP
if EPP is not supported") intel_pstate refuses to use HWP on those
systems, but the fallback PERF_CTL interface does not work on them
either because of enabled HWP, and once enabled, HWP cannot be
disabled. Consequently, the users of those systems cannot control
CPU performance scaling.
Address this issue by making intel_pstate use HWP unconditionally if
it is enabled already when the driver starts.
Fixes: 7aa1031223bc ("cpufreq: intel_pstate: Avoid enabling HWP if EPP is not supported")
Reported-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Tested-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: 5.9+ <stable@vger.kernel.org> # 5.9+
|
|
Because pstate.max_freq is always equal to the product of
pstate.max_pstate and pstate.scaling and, analogously,
pstate.turbo_freq is always equal to the product of
pstate.turbo_pstate and pstate.scaling, the result of the
max_policy_perf computation in intel_pstate_update_perf_limits() is
always equal to the quotient of policy_max and pstate.scaling,
regardless of whether or not turbo is disabled. Analogously, the
result of min_policy_perf in intel_pstate_update_perf_limits() is
always equal to the quotient of policy_min and pstate.scaling.
Accordingly, intel_pstate_update_perf_limits() need not check
whether or not turbo is enabled at all and in order to compute
max_policy_perf and min_policy_perf it can always divide policy_max
and policy_min, respectively, by pstate.scaling. Make it do so.
While at it, move the definition and initialization of the
turbo_max local variable to the code branch using it.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Chen Yu <yu.c.chen@intel.com>
|
|
Notice that some computations related to frequency in intel_pstate
can be simplified if (a) intel_pstate_get_hwp_max() updates the
relevant members of struct cpudata by itself and (b) the "turbo
disabled" check is moved from it to its callers, so modify the code
accordingly and while at it rename intel_pstate_get_hwp_max() to
intel_pstate_get_hwp_cap() which better reflects its purpose and
provide a simplified variat of it, __intel_pstate_get_hwp_cap(),
suitable for the initialization path.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Chen Yu <yu.c.chen@intel.com>
|
|
In the comment for trace in passive mode there is an
unnecessary "the". Eradicate it.
Signed-off-by: Nigel Christian <nigel.l.christian@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
available
Currently, when turbo is disabled (either by BIOS or by the user),
the intel_pstate driver reads the max non-turbo frequency from the
package-wide MSR_PLATFORM_INFO(0xce) register.
However, on asymmetric platforms it is possible in theory that small
and big core with HWP enabled might have different max non-turbo CPU
frequency, because MSR_HWP_CAPABILITIES is per-CPU scope according
to Intel Software Developer Manual.
The turbo max freq is already per-CPU in current code, so make
similar change to the max non-turbo frequency as well.
Reported-by: Wendy Wang <wendy.wang@intel.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
[ rjw: Subject and changelog edits ]
Cc: 4.18+ <stable@vger.kernel.org> # 4.18+: a45ee4d4e13b: cpufreq: intel_pstate: Change intel_pstate_get_hwp_max() argument
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Rename intel_cpufreq_adjust_hwp() and intel_cpufreq_adjust_perf_ctl()
to intel_cpufreq_hwp_update() and intel_cpufreq_perf_ctl_update(),
respectively, to avoid possible confusion with the ->adjist_perf()
callback function, intel_cpufreq_adjust_perf().
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Chen Yu <yu.c.chen@intel.com>
|
|
All of the callers of intel_pstate_get_hwp_max() access the struct
cpudata object that corresponds to the given CPU already and the
function itself needs to access that object (in order to update
hwp_cap_cached), so modify the code to pass a struct cpudata pointer
to it instead of the CPU number.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Chen Yu <yu.c.chen@intel.com>
|
|
Because intel_pstate_get_hwp_max() which updates hwp_cap_cached
may run in parallel with the readers of it, annotate all of the
read accesses to it with READ_ONCE().
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Chen Yu <yu.c.chen@intel.com>
|
|
percent_fp() was used in intel_pstate_pid_reset(), which was removed in
commit 9d0ef7af1f2d ("cpufreq: intel_pstate: Do not use PID-based P-state
selection") and hence, percent_fp() is unused since then.
percent_ext_fp() was last used in intel_pstate_update_perf_limits(), which
was refactored in commit 1a4fe38add8b ("cpufreq: intel_pstate: Remove
max/min fractions to limit performance"), and hence, percent_ext_fp() is
unused since then.
make CC=clang W=1 points us those unused functions:
drivers/cpufreq/intel_pstate.c:79:23: warning: unused function 'percent_fp' [-Wunused-function]
static inline int32_t percent_fp(int percent)
^
drivers/cpufreq/intel_pstate.c:94:23: warning: unused function 'percent_ext_fp' [-Wunused-function]
static inline int32_t percent_ext_fp(int percent)
^
Remove those obsolete functions.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Reviewed-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
If turbo P-states cannot be used, either due to the configuration of
the processor, or because intel_pstate is not allowed to used them,
the maximum available P-state with HWP enabled corresponds to the
HWP_CAP.GUARANTEED value which is not static. It can be adjusted by
an out-of-band agent or during an Intel Speed Select performance
level change, so long as it remains less than or equal to
HWP_CAP.MAX.
However, if turbo P-states cannot be used, intel_cpufreq_adjust_perf()
always uses pstate.max_pstate (set during the initialization of the
driver only) as the maximum available P-state, so it may miss a change
of the HWP_CAP.GUARANTEED value.
Prevent that from happening by modifyig intel_cpufreq_adjust_perf()
to always read the "guaranteed" and "maximum turbo" performance
levels from the cached HWP_CAP value.
Fixes: a365ab6b9dfb ("cpufreq: intel_pstate: Implement the ->adjust_perf() callback")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
When sugov_update_single_perf() falls back to the "frequency"
path due to the missing scale-invariance, it will call
cpufreq_driver_fast_switch() via sugov_fast_switch()
and the driver's ->fast_switch() callback will be invoked,
so it must not be NULL.
However, after commit a365ab6b9dfb ("cpufreq: intel_pstate: Implement
the ->adjust_perf() callback") intel_pstate sets ->fast_switch() to
NULL when it is going to use intel_cpufreq_adjust_perf(), which is a
mistake, because on x86 the scale-invariance may be turned off
dynamically, so modify it to retain the original ->adjust_perf()
callback pointer.
Fixes: a365ab6b9dfb ("cpufreq: intel_pstate: Implement the ->adjust_perf() callback")
Reported-by: Kenneth R. Crudup <kenny@panix.com>
Tested-by: Kenneth R. Crudup <kenny@panix.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
When turbo has been disabled by the BIOS, but HWP_CAP.GUARANTEED is
changed later, user space may want to take advantage of this increased
guaranteed performance.
HWP_CAP.GUARANTEED is not a static value. It can be adjusted by an
out-of-band agent or during an Intel Speed Select performance level
change. The HWP_CAP.MAX is still the maximum achievable performance
with turbo disabled by the BIOS, so HWP_CAP.GUARANTEED can still
change as long as it remains less than or equal to HWP_CAP.MAX.
When HWP_CAP.GUARANTEED is changed, the sysfs base_frequency
attribute shows the most recent guaranteed frequency value. This
attribute can be used by user space software to update the scaling
min/max limits of the CPU.
Currently, the ->setpolicy() callback already uses the latest
HWP_CAP values when setting HWP_REQ, but the ->verify() callback will
restrict the user settings to the to old guaranteed performance value
which prevents user space from making use of the extra CPU capacity
theoretically available to it after increasing HWP_CAP.GUARANTEED.
To address this, read HWP_CAP in intel_pstate_verify_cpu_policy()
to obtain the maximum P-state that can be used and use that to
confine the policy max limit instead of using the cached and
possibly stale pstate.max_freq value for this purpose.
For consistency, update intel_pstate_update_perf_limits() to use the
maximum available P-state returned by intel_pstate_get_hwp_max() to
compute the maximum frequency instead of using the return value of
intel_pstate_get_max_freq() which, again, may be stale.
This issue is a side-effect of fixing the scaling frequency limits in
commit eacc9c5a927e ("cpufreq: intel_pstate: Fix intel_pstate_get_hwp_max()
for turbo disabled") which corrected the setting of the reduced scaling
frequency values, but caused stale HWP_CAP.GUARANTEED to be used in
the case at hand.
Fixes: eacc9c5a927e ("cpufreq: intel_pstate: Fix intel_pstate_get_hwp_max() for turbo disabled")
Reported-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Tested-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: 5.8+ <stable@vger.kernel.org> # 5.8+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Make intel_pstate expose the ->adjust_perf() callback when it
operates in the passive mode with HWP enabled which causes the
schedutil governor to use that callback instead of ->fast_switch().
The minimum and target performance-level values passed by the
governor to ->adjust_perf() are converted to HWP.REQ.MIN and
HWP.REQ.DESIRED, respectively, which allows the processor to
adjust its configuration to maximize energy-efficiency while
providing sufficient capacity.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
|
|
Avoid doing the same assignment in both branches of a conditional,
do it after the whole conditional instead.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Make intel_pstate take the new CPUFREQ_GOV_STRICT_TARGET governor
flag into account when it operates in the passive mode with HWP
enabled, so as to fix the "powersave" governor behavior in that
case (currently, HWP is allowed to scale the performance all the
way up to the policy max limit when the "powersave" governor is
used, but it should be constrained to the policy min limit then).
Fixes: f6ebbcf08f37 ("cpufreq: intel_pstate: Implement passive mode with HWP enabled")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Cc: 5.9+ <stable@vger.kernel.org> # 5.9+: 9a2a9ebc0a75 cpufreq: Introduce governor flags
Cc: 5.9+ <stable@vger.kernel.org> # 5.9+: 218f66870181 cpufreq: Introduce CPUFREQ_GOV_STRICT_TARGET
Cc: 5.9+ <stable@vger.kernel.org> # 5.9+: ea9364bbadf1 cpufreq: Add strict_target to struct cpufreq_policy
|
|
If the cpufreq policy max limit is changed when intel_pstate operates
in the passive mode with HWP enabled and the "powersave" governor is
used on top of it, the HWP max limit is not updated as appropriate.
Namely, in the "powersave" governor case, the target P-state
is always equal to the policy min limit, so if the latter does
not change, intel_cpufreq_adjust_hwp() is not invoked to update
the HWP Request MSR due to the "target_pstate != old_pstate" check
in intel_cpufreq_update_pstate(), so the HWP max limit is not
updated as a result.
Also, if the CPUFREQ_NEED_UPDATE_LIMITS flag is not set for the
driver and the target frequency does not change along with the
policy max limit, the "target_freq == policy->cur" check in
__cpufreq_driver_target() prevents the driver's ->target() callback
from being invoked at all, so the HWP max limit is not updated.
To prevent that occurring, set the CPUFREQ_NEED_UPDATE_LIMITS flag
in the intel_cpufreq driver structure if HWP is enabled and modify
intel_cpufreq_update_pstate() to do the "target_pstate != old_pstate"
check only in the non-HWP case and let intel_cpufreq_adjust_hwp()
always run in the HWP case (it will update HWP Request only if the
cached value of the register is different from the new one including
the limits, so if neither the target P-state value nor the max limit
changes, the register write will still be avoided).
Fixes: f6ebbcf08f37 ("cpufreq: intel_pstate: Implement passive mode with HWP enabled")
Reported-by: Zhang Rui <rui.zhang@intel.com>
Cc: 5.9+ <stable@vger.kernel.org> # 5.9+: 1c534352f47f cpufreq: Introduce CPUFREQ_NEED_UPDATE_LIMITS ...
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Zhang Rui <rui.zhang@intel.com>
|
|
driver
There is a corner case that if the intel_pstate driver fails to be
registered (might be due to invalid MSR access) and acpi_cpufreq
takse over, the intel_pstate sysfs interface is still populated
which may confuse user space (turbostat for example):
grep . /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
acpi-cpufreq
grep . /sys/devices/system/cpu/intel_pstate/*
/sys/devices/system/cpu/intel_pstate/max_perf_pct:0
/sys/devices/system/cpu/intel_pstate/min_perf_pct:0
grep: /sys/devices/system/cpu/intel_pstate/no_turbo: Resource temporarily unavailable
grep: /sys/devices/system/cpu/intel_pstate/num_pstates: Resource temporarily unavailable
/sys/devices/system/cpu/intel_pstate/status:off
grep: /sys/devices/system/cpu/intel_pstate/turbo_pct: Resource temporarily unavailable
The mere presence of the intel_pstate sysfs interface does not mean
that intel_pstate is in use (for example, echo "off" to "status"),
but it should not be created in the failing case.
Fix this issue by deleting the intel_pstate sysfs if the driver
registration fails.
Reported-by: Wendy Wang <wendy.wang@intel.com>
Suggested-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com
[ rjw: Refactor code to avoid jumps, change function name, changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Fix missing return statement when writing "off" to intel_pstate status
sysfs I/F.
Fixes: 55671ea3257a ("cpufreq: intel_pstate: Free memory only when turning off")
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
This fixes the behavior of the scaling_max_freq and scaling_min_freq
sysfs files in systems which had turbo disabled by the BIOS.
Caleb noticed that the HWP is programmed to operate in the wrong
P-state range on his system when the CPUFREQ policy min/max frequency
is set via sysfs. This seems to be because in his system
intel_pstate_get_hwp_max() is returning the maximum turbo P-state even
though turbo was disabled by the BIOS, which causes intel_pstate to
scale kHz frequencies incorrectly e.g. setting the maximum turbo
frequency whenever the maximum guaranteed frequency is requested via
sysfs.
Tested-by: Caleb Callaway <caleb.callaway@intel.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
[ rjw: Minor subject edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
When intel_pstate switches the operation mode from "active" to
"passive" or the other way around, freeing its data structures
representing CPUs and allocating them again from scratch is not
necessary and wasteful. Moreover, if these data structures are
preserved, the cached HWP Request MSR value from there may be
written to the MSR to start with to reinitialize it and help to
restore the EPP value set previously (it is set to 0xFF when CPUs
go offline to allow their SMT siblings to use the full range of
EPP values and that also happens when the driver gets unregistered).
Accordingly, modify the driver to only do a full cleanup on driver
object registration errors and when its status is changed to "off"
via sysfs and to write the cached HWP Request MSR value back to
the MSR on CPU init if the data structure representing the given
CPU is still there.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
Add ->offline and ->online driver callbacks to prepare for taking a
CPU offline and to restore its working configuration when it goes
back online, respectively, to avoid invoking the ->init callback on
every CPU online which is quite a bit of unnecessary overhead.
Define ->offline and ->online so that they can be used in the
passive mode as well as in the active mode and because ->offline
will do the majority of ->stop_cpu work, the passive mode does
not need that callback any more, so drop it from there.
Also modify the active mode ->suspend and ->resume callbacks to
prevent them from interfering with the new ->offline and ->online
ones in case the latter are invoked withing the system-wide suspend
and resume code flow and make the passive mode use them too.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
Modify the EPP sysfs interface to reject attempts to change the EPP
to values different from 0 ("performance") in the active mode with
the "performance" policy (ie. scaling_governor set to "performance"),
to avoid situations in which the kernel appears to discard data
passed to it via the EPP sysfs attribute.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
Make intel_pstate update the cached EPP value when setting the EPP
via sysfs in the active mode just like it is the case in the passive
mode, for consistency, but also for the benefit of subsequent
changes.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
After commit f6ebbcf08f37 ("cpufreq: intel_pstate: Implement passive
mode with HWP enabled") it is possible to change the driver status
to "off" via sysfs with HWP enabled, which effectively causes the
driver to unregister itself, but HWP remains active and it forces the
minimum performance, so even if another cpufreq driver is loaded,
it will not be able to control the CPU frequency.
For this reason, make the driver refuse to change the status to
"off" with HWP enabled.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
|
|
Allow intel_pstate to work in the passive mode with HWP enabled and
make it set the HWP minimum performance limit (HWP floor) to the
P-state value given by the target frequency supplied by the cpufreq
governor, so as to prevent the HWP algorithm and the CPU scheduler
from working against each other, at least when the schedutil governor
is in use, and update the intel_pstate documentation accordingly.
Among other things, this allows utilization clamps to be taken
into account, at least to a certain extent, when intel_pstate is
in use and makes it more likely that sufficient capacity for
deadline tasks will be provided.
After this change, the resulting behavior of an HWP system with
intel_pstate in the passive mode should be close to the behavior
of the analogous non-HWP system with intel_pstate in the passive
mode, except that the HWP algorithm is generally allowed to make the
CPU run at a frequency above the floor P-state set by intel_pstate in
the entire available range of P-states, while without HWP a CPU can
run in a P-state above the requested one if the latter falls into the
range of turbo P-states (referred to as the turbo range) or if the
P-states of all CPUs in one package are coordinated with each other
at the hardware level.
[Note that in principle the HWP floor may not be taken into account
by the processor if it falls into the turbo range, in which case the
processor has a license to choose any P-state, either below or above
the HWP floor, just like a non-HWP processor in the case when the
target P-state falls into the turbo range.]
With this change applied, intel_pstate in the passive mode assumes
complete control over the HWP request MSR and concurrent changes of
that MSR (eg. via the direct MSR access interface) are overridden by
it.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
|
|
The MSR_TURBO_RATIO_LIMIT can be 0. This is not an error. User can update
this MSR via BIOS settings on some systems or can use msr tools to update.
Also some systems boot with value = 0.
This results in display of cpufreq/cpuinfo_max_freq wrong. This value
will be equal to cpufreq/base_frequency, even though turbo is enabled.
But platform will still function normally in HWP mode as we get max
1-core frequency from the MSR_HWP_CAPABILITIES. This MSR is already used
to calculate cpu->pstate.turbo_freq, which is used for to set
policy->cpuinfo.max_freq. But some other places cpu->pstate.turbo_pstate
is used. For example to set policy->max.
To fix this, also update cpu->pstate.turbo_pstate when updating
cpu->pstate.turbo_freq.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Because intel_pstate_set_energy_pref_index() reads and writes the
MSR_HWP_REQUEST register without using the cached value of it used by
intel_pstate_hwp_boost_up() and intel_pstate_hwp_boost_down(), those
functions may overwrite the value written by it and so the EPP value
set via sysfs may be lost.
To avoid that, make intel_pstate_set_energy_pref_index() take the
cached value of MSR_HWP_REQUEST just like the other two routines
mentioned above and update it with the new EPP value coming from
user space in addition to updating the MSR.
Note that the MSR itself still needs to be updated too in case
hwp_boost is unset or the boosting mechanism is not active at the
EPP change time.
Fixes: e0efd5be63e8 ("cpufreq: intel_pstate: Add HWP boost utility and sched util hooks")
Reported-by: Francisco Jerez <currojerez@riseup.net>
Cc: 4.18+ <stable@vger.kernel.org> # 4.18+: 3da97d4db8ee cpufreq: intel_pstate: Rearrange ...
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
|
|
Move the locking away from intel_pstate_set_energy_pref_index()
into its only caller and drop the (now redundant) return_pref label
from it.
Also move the "raw" EPP value check into the caller of that function,
so as to do it before acquiring the mutex, and reduce code duplication
related to the "raw" EPP values processing somewhat.
No intentional functional impact.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
|
|
|
|
Although there are processors supporting hardware-managed P-states
(HWP) without the energy-performance preference (EPP) feature, they
are not expected to be run with HWP enabled (the BIOS should disable
HWP on those systems). Missing EPP support generally indicates an
incomplete HWP implementation and so it is better to avoid using
HWP on those systems in production.
However, intel_pstate currently enables HWP on such systems, which
is questionable, so prevent it from doing that by making it check
EPP support before enabling HWP and avoid enabling it if EPP is not
supported by the processor at hand.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
The kerneldoc description of the aperf_mperf_shift field in
struct global_params is unclear and there is a typo in it, so
simplify it and clean it up.
Reported-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Lee Jones <lee.jones@linaro.org>
|
|
get_aperf_mperf_shift()
Fixes the following W=1 kernel build warning(s):
drivers/cpufreq/intel_pstate.c:293: warning: Function parameter or member 'get_aperf_mperf_shift' not described in 'pstate_funcs'
Suggested-by: "Rafael J. Wysocki" <rafael@kernel.org>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
[ rjw: Remove line break ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
If intel_pstate starts in the passive mode by default (that happens
when the processor in the system doesn't support HWP), passing
intel_pstate=active in the kernel command line doesn't work, so
fix that.
Fixes: 33aa46f252c7 ("cpufreq: intel_pstate: Use passive mode by default without HWP")
Reported-by: Doug Smythies <dsmythies@telus.net>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Doug Smythies <dsmythies@telus.net>
|
|
Fix warning for:
drivers/cpufreq/intel_pstate.c:731 store_energy_performance_preference()
error: uninitialized symbol 'epp'.
This warning is for a case, when energy_performance_preference attribute
matches pre defined strings. In this case the value of raw epp will not
be used to set EPP bits in MSR_HWP_REQUEST. So initializing with any
value is fine.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Currently using attribute "energy_performance_preference", user space can
write one of the four per-defined preference string. These preference
strings gets mapped to a hard-coded Energy-Performance Preference (EPP) or
Energy-Performance Bias (EPB) knob.
These four values are supposed to cover broad spectrum of use cases, but
are not uniformly distributed in the range. There are number of cases,
where this is not enough. For example:
Suppose user wants more performance when connected to AC. Instead of using
default "balance performance", the "performance" setting can be used. This
changes EPP value from 0x80 to 0x00. But setting EPP to 0, results in
electrical and thermal issues on some platforms. This results in
aggressive throttling, which causes a drop in performance. But some value
between 0x80 and 0x00 results in better performance. But that value can't
be fixed as the power curve is not linear. In some cases just changing EPP
from 0x80 to 0x75 is enough to get significant performance gain.
Similarly on battery the default "balance_performance" mode can be
aggressive in power consumption. But picking up the next choice
"balance power" results in too much loss of performance, which results in
bad user experience in use cases like "Google Hangout". It was observed
that some value between these two EPP is optimal.
This change allows fine grain EPP tuning for platform like Chromebook or
for users who wants to fine tune power and performance.
Here based on the product and use cases, different EPP values can be set.
This change is similar to the change done for:
/sys/devices/system/cpu/cpu*/power/energy_perf_bias
where user has choice to write a predefined string or raw value.
The change itself is trivial. When user preference doesn't match
predefined string preferences and value is an unsigned integer and in
range, use that value for EPP. When the EPP feature is not present
writing raw value is not supported.
Suggested-by: Len Brown <lenb@kernel.org>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
By default intel_pstate the driver disables energy efficiency by setting
MSR_IA32_POWER_CTL bit 19 for Kaby Lake desktop CPU model in HWP mode.
This CPU model is also shared by Coffee Lake desktop CPUs. This allows
these systems to reach maximum possible frequency. But this adds power
penalty, which some customers don't want. They want some way to enable/
disable dynamically.
So, add an additional attribute "energy_efficiency" under
/sys/devices/system/cpu/intel_pstate/ for these CPU models. This allows
to read and write bit 19 ("Disable Energy Efficiency Optimization") in
the MSR IA32_POWER_CTL.
This attribute is present in both HWP and non-HWP mode as this has an
effect in both modes. Refer to Intel Software Developer's manual for
details.
The scope of this bit is package wide. Also these systems are single
package systems. So read/write MSR on the current CPU is enough.
The energy efficiency (EE) bit setting needs to be preserved during
suspend/resume and CPU offline/online operation. To do this:
- Restoring the EE setting from the cpufreq resume() callback, if there
is change from the system default.
- By default, don't disable EE from cpufreq init() callback for matching
CPU models. Since the scope is package wide and is a single package
system, move the disable EE calls from init() callback to
intel_pstate_init() function, which is called only once.
Suggested-by: Len Brown <lenb@kernel.org>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
Add one more bit for OOB (Out Of Band) enabling of P-states.
If OOB handling of P-states is enabled, intel_pstate shouldn't load.
Currently, only "BIT(8) == 1" of the MSR MSR_MISC_PWR_MGMT is
considered as OOB, but "BIT(18) == 1" needs to be taken into
consideration as OOB condition too.
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
[ rjw: Add an empty code line, edit subject and changelog ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
|
|
Make a note of the first time we discover the turbo mode has been
disabled by the BIOS, as otherwise we complain every time we try to
update the mode.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
After recent changes allowing scale-invariant utilization to be
used on x86, the schedutil governor on top of intel_pstate in the
passive mode should be on par with (or better than) the active mode
"powersave" algorithm of intel_pstate on systems in which
hardware-managed P-states (HWP) are not used, so it should not be
necessary to use the internal scaling algorithm in those cases.
Accordingly, modify intel_pstate to start in the passive mode by
default if the processor at hand does not support HWP of if the driver
is requested to avoid using HWP through the kernel command line.
Among other things, that will allow utilization clamps and the
support for RT/DL tasks in the schedutil governor to be utilized on
systems in which intel_pstate is used.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler updates from Ingo Molnar:
"The main changes in this cycle are:
- Various NUMA scheduling updates: harmonize the load-balancer and
NUMA placement logic to not work against each other. The intended
result is better locality, better utilization and fewer migrations.
- Introduce Thermal Pressure tracking and optimizations, to improve
task placement on thermally overloaded systems.
- Implement frequency invariant scheduler accounting on (some) x86
CPUs. This is done by observing and sampling the 'recent' CPU
frequency average at ~tick boundaries. The CPU provides this data
via the APERF/MPERF MSRs. This hopefully makes our capacity
estimates more precise and keeps tasks on the same CPU better even
if it might seem overloaded at a lower momentary frequency. (As
usual, turbo mode is a complication that we resolve by observing
the maximum frequency and renormalizing to it.)
- Add asymmetric CPU capacity wakeup scan to improve capacity
utilization on asymmetric topologies. (big.LITTLE systems)
- PSI fixes and optimizations.
- RT scheduling capacity awareness fixes & improvements.
- Optimize the CONFIG_RT_GROUP_SCHED constraints code.
- Misc fixes, cleanups and optimizations - see the changelog for
details"
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (62 commits)
threads: Update PID limit comment according to futex UAPI change
sched/fair: Fix condition of avg_load calculation
sched/rt: cpupri_find: Trigger a full search as fallback
kthread: Do not preempt current task if it is going to call schedule()
sched/fair: Improve spreading of utilization
sched: Avoid scale real weight down to zero
psi: Move PF_MEMSTALL out of task->flags
MAINTAINERS: Add maintenance information for psi
psi: Optimize switching tasks inside shared cgroups
psi: Fix cpu.pressure for cpu.max and competing cgroups
sched/core: Distribute tasks within affinity masks
sched/fair: Fix enqueue_task_fair warning
thermal/cpu-cooling, sched/core: Move the arch_set_thermal_pressure() API to generic scheduler code
sched/rt: Remove unnecessary push for unfit tasks
sched/rt: Allow pulling unfitting task
sched/rt: Optimize cpupri_find() on non-heterogenous systems
sched/rt: Re-instate old behavior in select_task_rq_rt()
sched/rt: cpupri_find: Implement fallback mechanism for !fit case
sched/fair: Fix reordering of enqueue/dequeue_task_fair()
sched/fair: Fix runnable_avg for throttled cfs
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull perf updates from Ingo Molnar:
"The main changes in this cycle were:
Kernel side changes:
- A couple of x86/cpu cleanups and changes were grandfathered in due
to patch dependencies. These clean up the set of CPU model/family
matching macros with a consistent namespace and C99 initializer
style.
- A bunch of updates to various low level PMU drivers:
* AMD Family 19h L3 uncore PMU
* Intel Tiger Lake uncore support
* misc fixes to LBR TOS sampling
- optprobe fixes
- perf/cgroup: optimize cgroup event sched-in processing
- misc cleanups and fixes
Tooling side changes are to:
- perf {annotate,expr,record,report,stat,test}
- perl scripting
- libapi, libperf and libtraceevent
- vendor events on Intel and S390, ARM cs-etm
- Intel PT updates
- Documentation changes and updates to core facilities
- misc cleanups, fixes and other enhancements"
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (89 commits)
cpufreq/intel_pstate: Fix wrong macro conversion
x86/cpu: Cleanup the now unused CPU match macros
hwrng: via_rng: Convert to new X86 CPU match macros
crypto: Convert to new CPU match macros
ASoC: Intel: Convert to new X86 CPU match macros
powercap/intel_rapl: Convert to new X86 CPU match macros
PCI: intel-mid: Convert to new X86 CPU match macros
mmc: sdhci-acpi: Convert to new X86 CPU match macros
intel_idle: Convert to new X86 CPU match macros
extcon: axp288: Convert to new X86 CPU match macros
thermal: Convert to new X86 CPU match macros
hwmon: Convert to new X86 CPU match macros
platform/x86: Convert to new CPU match macros
EDAC: Convert to new X86 CPU match macros
cpufreq: Convert to new X86 CPU match macros
ACPI: Convert to new X86 CPU match macros
x86/platform: Convert to new CPU match macros
x86/kernel: Convert to new CPU match macros
x86/kvm: Convert to new CPU match macros
x86/perf/events: Convert to new CPU match macros
...
|
|
The initial policy value set by intel_pstate_cpu_init() depends on
whether or not CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE is set, but
that is not necessary, because the core will set the policy to
"performance" in cpufreq_init_policy() if the default governor is
"performance" anyway.
Accordingly, change intel_pstate_cpu_init() to always set policy
to CPUFREQ_POLICY_POWERSAVE initially to provide a valid fallback
value to cpufreq_init_policy() in case the default cpufreq governor
is neither "powersave" nor "performance".
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
The feature flag hwp_support_ids are supposed to match on is
X86_FEATURE_HWP, not X86_FEATURE_APERFMPERF. Fix it.
[ bp: Write commit message. ]
Fixes: b11d77fa300d ("cpufreq: Convert to new X86 CPU match macros")
Reported-by: kernel test robot <rong.a.chen@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20200324060124.GC11705@shao2-debian
|
