The old code used to ignore CPU load older than 1 tick. Since the WALT
window size can be more than 1 tick (and is configurable), this can cause
an incorrect frequency drop in the middle of a WALT window. Use WALT window
size to ignore CPU load instead.
Change-Id: If42420f595ba9722866174c2cda8b6c2fdf34f92
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Fixed merge conflicts.]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
WALT load works better for UX workloads which are important for a
commercial mobile device.
Change-Id: I33fbb5c0566176288029e3249532f8c4b389add3
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Fix merge conflict]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Blindly ignoring frequency updates because of work_in_progress can leave
the CPUs at the wrong frequency for a long time. It's better to update the
frequency immediately than wait for a future event that might take a long
time to come. The irq_work code already ignores double queuing of work. So,
that part of the code is still safe when the work_in_progress flag is
ignored.
Change-Id: Id0b3711314dfbfa18b5f4bce30a239ee3cf962d6
Signed-off-by: Vikram Mulukutla <markivx@codeaurora.org>
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Port to 4.19]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Sprintf has security issues; replace with scnprintf where appropriate.
Change-Id: I5b4bce0d37e2ede73324dc753f298d8ac236ddd9
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Most of the toplogy related declarations have been moved
out of include/linux/sched.h. Use the new header.
Change-Id: If7b83d986c14dce58a68a3f89d5f539398661d7e
Signed-off-by: Vikram Mulukutla <markivx@codeaurora.org>
Signed-off-by: Satya Durga Srinivasu Prabhala <satyap@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Many subsystems depend on cpufreq API for CPU frequency scaling.
Cpufreq API is expected to fail until cpufreq device registers.
Change pr_debug() to pr_info() so that user could determine when
cpufreq API becomes available during boot from kernel messages. This
is crucial to understand whether a cpufreq API failure is benign
during early boot.
Change-Id: Id2dfa009ae33859ec3efcdb29a3296e891852c6a
Signed-off-by: Junjie Wu <junjiew@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
When use common cpufreq table, fix missing register pm notifier and
cpufreq driver register.
Change-Id: I8432696c81c941c641d9ee1db6385e9e94c3a012
Signed-off-by: Maria Yu <aiquny@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Governor error messages point to important failures in governor or
framework. Output triggering CPU and policy->cpu to help debugging.
Change-Id: I4c5c392ec973b764ec3240bb2eb455c624bcaf63
Signed-off-by: Junjie Wu <junjiew@codeaurora.org>
[sboyd@codeaurora.org: Dropped most hunks as they didn't apply anymore between 4.4 and 4.9]
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Thermal cooling device has to be registered when the policy for a CPU is
ready. Cpufreq will get a callback when a cpufreq policy is ready and
register CPU cooling device as a part of this callback, so that the CPU
can be mitigated immediately if needed.
Ignore cpu cooling device registration when there is platform cooling
device available.
Change-Id: I7cfb8598aa8ead4091a617da3faddf86ff0fe6a8
Signed-off-by: Ram Chandrasekar <rkumbako@codeaurora.org>
Signed-off-by: Manaf Meethalavalappu Pallikunhi <manafm@codeaurora.org>
[avilaj@codeaurora.org: Modify to support 4.19 API changes]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
The topology driver keeps track of the current capacity of a CPU
and make it available via arch_scale_freq_capacity(). The PELT
load tracking in the scheduler and energy-aware task placement
are the clients of this API. Hence notify the new frequency to the
topology driver during frequency transition.
Change-Id: I6dae674ba3b9d4c745b3caff59b165215926d970
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
This driver implements resolved_freq and target methods.
Since target method does not know the resolved frequency
index, the frequency table look up is needed again. We
can optimize this by caching the resolved frequency
index in the resolved_freq method.
Change-Id: Ib303ef57c51ce50de52b759728bc8c162a90dc29
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Fix some merge conflicts]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
As remote cpufreq updates are supported on MSM platforms, set
dvfs_possible_from_any_cpu cpufreq driver flag.
Change-Id: I7d8a1a22a512fe87929a7a59724fec6a577bcb05
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
The resolve_freq method is needed for schedutil to avoid
calling into cpufreq for changing to the same frequency
again and again.
Change-Id: Ia03cc49e3b8bad68fa2fda5688c64233b4b2c48d
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
We don't handle probe defer very well in this driver. If the
platform_driver_register() call doesn't immediately probe the
driver there, or if that driver probe defers for some reason,
then the cpufreq driver registration in msm_cpufreq_register()
will fail due to a missing frequency table. Let's move the
cpufreq driver registration (and platform suspend hook part) into
the platform driver probe path, so we handle probe defer properly
and avoid any issues with order of initializations.
Change-Id: I14b514d46239f52b535563d49fb5c19d06072c3a
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
__cpufreq_driver_target() checks if policy->cur is same as target_freq
without holding any lock. This function is used by governor to
directly set CPU frequency. Governor calling this function can't hold
any CPUfreq framework locks due to deadlock possibility.
However, this results in a race condition where one thread could see
a stale policy->cur while another thread is changing CPU frequency.
Thread A: Governor calls __cpufreq_driver_target(), starts increasing
frequency but hasn't sent out CPUFREQ_POSTCHANGE notification yet.
Thread B: Some other driver (could be thermal mitigation) starts
limiting frequency using cpufreq_update_policy(). Every limits are
applied to policy->min/max and final policy->max happens to be same as
policy->cur. __cpufreq_driver_target() simply returns 0.
Thread A: Governor finish scaling and now policy->cur violates
policy->max and could last forever until next CPU frequency scaling
happens.
Shifting the responsibility of checking policy->cur and target_freq
to CPUfreq device driver would resolve the race as long as the device
driver holds a common mutex.
Change-Id: I6f943228e793a4a4300c58b3ae0143e09ed01d7d
Signed-off-by: Junjie Wu <junjiew@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
This is a snapshot of qcom-cpufreq as of msm-3.10 commit
Commit acdce027751d (Merge "defconfig: arm64: Enable
ONESHOT_SYNC for msm8994").
Change-Id: Idb99a856330566ffad6309c48edabb220cee7917
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
[junjiew@codeaurora.org: resolved conflicts in Kconfig.arm and Makefile. Dropped dependency on ARCH_MSM.]
Signed-off-by: Junjie Wu <junjiew@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[dkeitel@codeaurora.org: skip frequencies that round to same rate]
Signed-off-by: David Keitel <dkeitel@codeaurora.org>
[gsantosh@codeaurora.org: adapt to opensource clock framework]
Signed-off-by: Santosh Mardi <gsantosh@codeaurora.org>
[pkondeti@codeaurora.org: Remove superfluous NULL]
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
[avilaj@codeaurora.org: move documentation to separate patch]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
This driver allows for frequency scaling using the cpufreq framework.
Change-Id: I8fc2aca91cb27d94cd952aee0feb14b0a46f3667
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Introduce DT binding information for the qcom_cpufreq driver.
Change-Id: Iaef099cc66e2c3f1e733fdb55c46671f09d00f2a
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
It is sometimes useful to profile how long CPU frequency switches
take, since they often involve variable overhead (PLL lock times,
voltage increase time, etc.). Add additional traces to to make this
possible.
Since the overhead involved may differ based on the frequencies
being switched between, record both the start and the end frequencies
as part of the trace.
Change-Id: I2de743fc357dad3590fd4980f65f38f6073d426e
Signed-off-by: Matt Wagantall <mattw@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
[abhimany@codeaurora.org: resolve trivial merge conflicts]
Signed-off-by: Abhimanyu Kapur <abhimany@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Scheduler provides an API to force tasks to the big cluster. To
improve performance, use this API to move most/all tasks to the
big cluster for short duration on an input event. On the removal of
frequency boost (after input_boost_ms), this scheduler boost is also
deactivated.
Change-Id: I9d643914ebc75266478cc22260a45862faad6236
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Update sched_boost_on_input to be an unsigned int
so that it can utilize the latest sched_boost APIs.
Change-Id: I42f626c9f0112aee3bb9f37fd62d77709f03ebcf
Signed-off-by: Amir Vajid <avajid@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Change to sysfs from module_param]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Different types of CPUs could have different frequency to satisfy same
input workload. Add support for using different input_boost_freq on
different CPUs.
input_boost_freq now either takes a single number which applies to all
CPUs, or cpuid:freq pairs separated by space for different CPUs.
Change-Id: I20506a9fbdb4d532d94168bbd61744595bebc8e5
Signed-off-by: Junjie Wu <junjiew@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
[avilaj@codeaurora.org: Change to sysfs from module param]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Enable the driver which allows for input boosting.
Change-Id: Id240eaa498e4b36e27b4510e83e8c7438e5160f1
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
When certain bursty and important events take place, it might take a while
for the current cpufreq governor to notice the new load and react to it.
That would result in poor user experience. To alleviate this, the cpu-boost
driver boosts the frequency of a CPU for at least input_boost_ms duration
on incoming input events to maintain good user experience while the
governor catches up. This is accomplished by changing the policy->min of
all the online cpus to input_boost_freq.
Since this driver enforces the boost by hooking into standard cpufreq
ADJUST notifiers, it has several advantages:
- More portable across kernel versions where the cpufreq internals might
have been rewritten.
- Governor agnostic and hence works with multiple governors like
conservative, ondemand, interactive, etc.
- Does not affect the sampling period/logic of existing governors.
- Can have the boost period adjusted independent of governor sampling
period.
Change-Id: Idb0ab75d68ae4ceff259cbbaaec1a9bb3bc871d3
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: Rohit Gupta <rohgup@codeaurora.org>
Signed-off-by: Syed Rameez Mustafa <rameezmustafa@codeaurora.org>
Signed-off-by: Joonwoo Park <joonwoop@codeaurora.org>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
[avilaj@codeaurora.org: Change from module_param to sysfs.]
Signed-off-by: Jonathan Avila <avilaj@codeaurora.org>
Initialization function isn't correctly checking the return value of
sysfs_create_group:
drivers/bus/mhi/controllers/mhi_qcom.c: In function 'mhi_register_controller':
drivers/bus/mhi/controllers/mhi_qcom.c:610:2: warning: ignoring return value of 'sysfs_create_group', declared with attribute warn_unused_result [-Wunused-result]
Fixes: bebf2ba49d ("mhi: add snapshot for MHI driver stack")
Change-Id: Ic0dedbad089272344f85e46c18f6218b511f4877
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Fix setting profile & level control for HW decoders
and encoders.
Change-Id: I3e1816c2be1e6b6cc158da04365ce820bac59065
Signed-off-by: Amit Shekhar <ashekhar@codeaurora.org>
Fix compile errors related to arm32 platforms.
CRs-Fixed: 2378639
Change-Id: I6fd41673ff90a6902514d26d5af9e190d171c7a9
Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
Switch to using the prebuilt "make" in Android's tree if available.
The Android build system has reduced out-of-tree/host system
module dependency.
Change-Id: I40e8592e162f0b607f96dd31cc8785af3093e485
Signed-off-by: David Ng <dave@codeaurora.org>
The PIL loads the Secure Processor (SP) firmware image.
The SP-to-HLOS interrupt, intercepted by PIL, is the only way
to detect that SP firmware was loaded.
The SP loading is done in 2 phases, and therefor the interrupt
is fired twice.
Add 2 messages for verifying SP loaded successfully.
Change-Id: Id4a44871989b6ec8de85caa8fef2f66e31910025
Signed-off-by: Amir Samuelov <amirs@codeaurora.org>
Correct misleading numbering of registers on ipa-hal.
Fix the range of start and end reg_n numbers.
Change-Id: Ib3335b782fd781b032bcfb7598e14157a027f4b6
Acked-by: Michal Amsterdam <mamsterd@qti.qualcomm.com>
Signed-off-by: Amir Levy <alevy@codeaurora.org>
DCC can program multiple link lists. Add driver support to
parse multiple link lists in device tree. Support upto
8 link lists.
Change-Id: Ib8a53253d44f4f8568a7dec079c8713e831c2c4d
Signed-off-by: Shaoqing Liu <shaoqingliu@codeaurora.org>
Signed-off-by: Tingwei Zhang <tingwei@codeaurora.org>
Add a new define for rmnet ioctl for RSC/RSB.
Change-Id: Ie86661bb4bb009c2c5c161e9fd140421e8b877e4
Signed-off-by: Michael Adisumarta <madisuma@codeaurora.org>
Add a new IOCTL that will enable and set the parameter for
coalescing through IPACM.
Change-Id: I5da2e8147422728510f4c8842a5a8263556d087a
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
Signed-off-by: Michael Adisumarta <madisuma@codeaurora.org>
Enable rmtfs module for modem processor, which needs access
to store data onto eMMC/UFS device.
The shared memory size used by modem and apps for this purpose is 2MB.
Change-Id: I4722845cd6edfbb7de66652c5f77781439673b93
Signed-off-by: Ankit Jain <jankit@codeaurora.org>
Assert if coal pipe uses anything other than gran 0 for coal pipe since
IPA 4.5 HW will configure coal pipe with pipe 0 config for gran_sel
which we set to 0 in IPA 4.5 GRAN changes.
Change-Id: Ia8f192279a64a7a20732a41f7defdfa02ea26c53
Signed-off-by: Michael Adisumarta <madisuma@codeaurora.org>
