An error occurred fetching the project authors.
- 27 Oct, 2020 1 commit
-
-
Viresh Kumar authored
The patch missed returning 0 early in case of success and hence the static OPPs got removed by mistake. Fix it. Fixes: 90d46d71 ("opp: Handle multiple calls for same OPP table in _of_add_opp_table_v1()") Reported-by:
Aisheng Dong <aisheng.dong@nxp.com> Signed-off-by:
Viresh Kumar <viresh.kumar@linaro.org> Tested-by:
-
- 16 Sep, 2020 3 commits
-
-
Viresh Kumar authored
The DT bindings don't put such a constraint, nor should the kernel. It is perfectly fine for opp-level to be set to 0, if we need to put the performance state votes for a domain for a particular OPP. Reported-by:
Stephan Gerhold <stephan@gerhold.net> Tested-by:
-
Viresh Kumar authored
The bindings allow multiple versions to be passed to "opp-supported-hw" property, either of which can result in enabling of the OPP. Update code to allow that. Tested-by:
Dmitry Osipenko <digetx@gmail.com> Signed-off-by:
-
Viresh Kumar authored
Until now for V1 OPP bindings we used to call dev_pm_opp_of_cpumask_add_table() first and then dev_pm_opp_set_sharing_cpus() in the cpufreq-dt driver. A later patch will though update the cpufreq-dt driver to optimize the code a bit and we will call dev_pm_opp_set_sharing_cpus() first followed by dev_pm_opp_of_cpumask_add_table(), which doesn't work well today as it tries to re parse the OPP entries. This should work nevertheless for V1 bindings as the same works for V2 bindings. Adapt the same approach from V2 bindings and fix this. Reported-by:
Marek Szyprowski <m.szyprowski@samsung.com> Tested-by:
-
- 25 Aug, 2020 1 commit
-
-
Stephan Gerhold authored
The OPP core manages various resources, e.g. clocks or interconnect paths. These resources are looked up when the OPP table is allocated once dev_pm_opp_get_opp_table() is called the first time (either directly or indirectly through one of the many helper functions). At this point, the resources may not be available yet, i.e. looking them up will result in -EPROBE_DEFER. Unfortunately, dev_pm_opp_get_opp_table() is currently unable to propagate this error code since it only returns the allocated OPP table or NULL. This means that all consumers of the OPP core are required to make sure that all necessary resources are available. Usually this happens by requesting them, checking the result and releasing them immediately after. For example, we have added "dev_pm_opp_of_find_icc_paths(dev, NULL)" to several drivers now just to make sure the interconnect providers are ready before the OPP table is allocated. If this call is missing, the OPP core will only warn about this and then attempt to continue without interconnect. This will eventually fail horribly, e.g.: cpu cpu0: _allocate_opp_table: Error finding interconnect paths: -517 ... later ... of: _read_bw: Mismatch between opp-peak-kBps and paths (1 0) cpu cpu0: _opp_add_static_v2: opp key field not found cpu cpu0: _of_add_opp_table_v2: Failed to add OPP, -22 This example happens when trying to use interconnects for a CPU OPP table together with qcom-cpufreq-nvmem.c. qcom-cpufreq-nvmem calls dev_pm_opp_set_supported_hw(), which ends up allocating the OPP table early. To fix the problem with the current approach we would need to add yet another call to dev_pm_opp_of_find_icc_paths(dev, NULL). But actually qcom-cpufreq-nvmem.c has nothing to do with interconnects... This commit attempts to make this more robust by allowing dev_pm_opp_get_opp_table() to return an error pointer. Fixing all the usages is trivial because the function is usually used indirectly through another helper (e.g. dev_pm_opp_set_supported_hw() above). These other helpers already return an error pointer. The example above then works correctly because set_supported_hw() will return -EPROBE_DEFER, and qcom-cpufreq-nvmem.c already propagates that error. It should also be possible to remove the remaining usages of "dev_pm_opp_of_find_icc_paths(dev, NULL)" from other drivers as well. Note that this commit currently only handles -EPROBE_DEFER for the clock/interconnects within _allocate_opp_table(). Other errors are just ignored as before. Eventually those should be propagated as well. Signed-off-by:Krzysztof Kozlowski <krzk@kernel.org> Reviewed-by:
Ulf Hansson <ulf.hansson@linaro.org> [ Viresh: skip checking return value of dev_pm_opp_get_opp_table() for EPROBE_DEFER in domain.c, fix NULL return value and reorder code a bit in core.c, and update exynos-asv.c ] Signed-off-by:
-
- 24 Aug, 2020 1 commit
-
-
Viresh Kumar authored
Remove the unnecessary wrapper and merge _dev_pm_opp_find_and_remove_table() with dev_pm_opp_remove_table(). Tested-by:
Rajendra Nayak <rnayak@codeaurora.org> Signed-off-by:
-
- 16 Jul, 2020 1 commit
-
-
Walter Lozano authored
Currently, when using _of_add_opp_table_v2 parsed_static_opps is increased and this value is used in _opp_remove_all_static() to check if there are static opp entries that need to be freed. Unfortunately this does not happen when using _of_add_opp_table_v1(), which leads to warnings. This patch increases parsed_static_opps in _of_add_opp_table_v1() in a similar way as in _of_add_opp_table_v2(). Fixes: 03758d60 ("opp: Replace list_kref with a local counter") Cc: v5.6+ <stable@vger.kernel.org> # v5.6+ Signed-off-by:
Walter Lozano <walter.lozano@collabora.com> [ Viresh: Do the operation with lock held and set the value to 1 instead of incrementing it ] Signed-off-by:
-
- 24 Jun, 2020 2 commits
-
-
Lukasz Luba authored
The Energy Model framework supports not only CPU devices. Drop the CPU specific interface with cpumask and add struct device. Add also a return value, user might use it. This new interface provides easy way to create a simple Energy Model, which then might be used by e.g. thermal subsystem. Acked-by:
Daniel Lezcano <daniel.lezcano@linaro.org> Signed-off-by:
Lukasz Luba <lukasz.luba@arm.com> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com>
-
Lukasz Luba authored
The Energy Model framework is going to support devices other that CPUs. In order to make this happen change the callback function and add pointer to a device as an argument. Update the related users to use new function and new callback from the Energy Model. Acked-by:
Quentin Perret <qperret@google.com> Signed-off-by:
-
- 01 Jun, 2020 1 commit
-
-
Sibi Sankar authored
The DT node of the device may contain interconnect paths while the OPP table doesn't have the bandwidth values. There is no need to parse the paths in such cases. Signed-off-by:
Sibi Sankar <sibis@codeaurora.org> Tested-by:
-
- 29 May, 2020 2 commits
-
-
Georgi Djakov authored
When we read the OPP keys, it would be nice to do some sanity checks of the values we get from DT and see if they match with the information that is populated in the OPP table. Let's pass a pointer of the table, so that we can do some validation. Signed-off-by:
Georgi Djakov <georgi.djakov@linaro.org> Reviewed-by:
Matthias Kaehlcke <mka@chromium.org> Reviewed-by:
-
Georgi Djakov authored
The OPP bindings now support bandwidth values, so add support to parse it from device tree and store it into the new dev_pm_opp_icc_bw struct, which is part of the dev_pm_opp. Signed-off-by:
-
- 13 May, 2020 1 commit
-
-
Saravana Kannan authored
The opp-hz DT property is not mandatory and we may use another property as a key in the OPP table. Add helper functions to simplify the reading and comparing the keys. Signed-off-by:
Saravana Kannan <saravanak@google.com> Signed-off-by:
-
- 10 Dec, 2019 2 commits
-
-
Viresh Kumar authored
A kref or refcount isn't the right tool to be used here for counting number of devices that are sharing the static OPPs created for the OPP table. For example, we are reinitializing the kref again, after it reaches a value of 0 and frees the resources, if the static OPPs get added for the same OPP table structure (as the OPP table structure was never freed). That is messy and very unclear. This patch makes parsed_static_opps an unsigned integer and uses it to count the number of users of the static OPPs. The increment and decrement to parsed_static_opps is done under opp_table->lock now to make sure no races are possible if the OPP table is getting added and removed in parallel (which doesn't happen in practice, but can in theory). Signed-off-by: -
Viresh Kumar authored
The static OPPs aren't getting freed properly, if errors occur while adding them. Fix that by calling _put_opp_list_kref() and putting their reference on failures. Fixes: 11e1a164 ("opp: Don't decrement uninitialized list_kref") Signed-off-by:
-
- 23 Oct, 2019 1 commit
-
-
Viresh Kumar authored
The list_kref reaches a count of 0 when all the static OPPs are removed, for example when dev_pm_opp_of_cpumask_remove_table() is called, though the actual OPP table may not get freed as it may still be referenced by other parts of the kernel, like from a call to dev_pm_opp_set_supported_hw(). And if we call dev_pm_opp_of_cpumask_add_table() again at this point, we must reinitialize the list_kref otherwise the kernel will hit a WARN() in kref infrastructure for incrementing a kref with value 0. Fixes: 11e1a164 ("opp: Don't decrement uninitialized list_kref") Reported-by:
-
- 10 Oct, 2019 1 commit
-
-
Viresh Kumar authored
_find_opp_of_np() doesn't traverse the list of OPP tables but instead just the entries within an OPP table and so only requires to lock the OPP table itself. The lockdep_assert_held() was added there by mistake and isn't really required. Fixes: 5d6d106f ("OPP: Populate required opp tables from "required-opps" property") Cc: v5.0+ <stable@vger.kernel.org> # v5.0+ Reported-by:
Niklas Cassel <niklas.cassel@linaro.org> Signed-off-by:
-
- 26 Jul, 2019 2 commits
-
-
Anson Huang authored
With property "opp-supported-hw" introduced, the OPP table in DT could be a large OPP table and ONLY a subset of OPPs are available, based on the version of the hardware running on. That introduces restriction of using "opp-suspend" property to define the suspend OPP, as we are NOT sure if the OPP containing "opp-suspend" property is available for the hardware running on, and the of opp core does NOT allow multiple suspend OPPs defined in DT OPP table. To eliminate this restrition, make of opp core allow multiple suspend OPPs defined in DT, and pick the OPP with highest rate and with "opp-suspend" property present to be suspend OPP, it can speed up the suspend/resume process. Signed-off-by:
Anson Huang <Anson.Huang@nxp.com> Signed-off-by:
-
Viresh Kumar authored
The list_kref was added for static OPPs and to track their users. The kref is initialized while the static OPPs are added, but removed unconditionally even if the static OPPs were never added. This causes refcount mismatch warnings currently. Fix that by always initializing the kref when the OPP table is first initialized. The refcount is later incremented only for the second user onwards. Fixes: d0e8ae6c ("OPP: Create separate kref for static OPPs list") Reported-by:
-
- 19 Jun, 2019 1 commit
-
-
Thomas Gleixner authored
Based on 2 normalized pattern(s): this program is free software you can redistribute it and or modify it under the terms of the gnu general public license version 2 as published by the free software foundation this program is free software you can redistribute it and or modify it under the terms of the gnu general public license version 2 as published by the free software foundation # extracted by the scancode license scanner the SPDX license identifier GPL-2.0-only has been chosen to replace the boilerplate/reference in 4122 file(s). Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Reviewed-by:
Enrico Weigelt <info@metux.net> Reviewed-by:
Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by:
Allison Randal <allison@lohutok.net> Cc: linux-spdx@vger.kernel.org Link: https://lkml.kernel.org/r/20190604081206.933168790@linutronix.deSigned-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
- 20 May, 2019 1 commit
-
-
Viresh Kumar authored
Currently the space for the array of virtual devices is allocated along with the OPP table, but that isn't going to work well from now onwards. For single power domain case, a driver can either use the original device structure for setting the performance state (if genpd attached with dev_pm_domain_attach()) or use the virtual device structure (if genpd attached with dev_pm_domain_attach_by_name(), which returns the virtual device) and so we can't know in advance if we are going to need genpd_virt_devs array or not. Lets delay the allocation a bit and do it along with dev_pm_opp_attach_genpd() rather. The deallocation is done from dev_pm_opp_detach_genpd(). Tested-by:
-
- 11 Mar, 2019 1 commit
-
-
Rajendra Nayak authored
We seem to rely on the number of phandles specified in the 'required-opps' property to identify cases where a device is associated with multiple power domains and hence would have multiple virtual devices that have to be dealt with. In cases where we do have devices with multiple power domains but with only one of them being scalable, this logic seems to fail. Instead read the number of power domains from DT to identify such cases. Signed-off-by:
Stephen Boyd <sboyd@kernel.org> Acked-by:
-
- 07 Feb, 2019 1 commit
-
-
Quentin Perret authored
The Energy Model (EM) framework provides an API to let drivers register the active power of CPUs. The drivers are expected to provide a callback method which estimates the power consumed by a CPU at each available performance levels. How exactly this should be implemented, however, depends on the platform. On some systems, PM_OPP knows the voltage and frequency at which CPUs can run. When coupled with the CPU 'capacitance' (as provided by the 'dynamic-power-coefficient' devicetree binding), it is possible to estimate the dynamic power consumption of a CPU as P = C * V^2 * f, with C its capacitance and V and f respectively the voltage and frequency of the OPP. The Intelligent Power Allocator (IPA) thermal governor already implements that estimation method, in the thermal framework. However, this power estimation method can be applied to any platform where all the parameters are known (C, V and f), and not only those suffering thermal issues. As such, the code implementing this feature can be re-used to also populate the EM framework now used by EAS. As a first step, introduce in PM_OPP a helper function which CPUFreq drivers can use to register into the EM framework. This duplicates the power estimation done in IPA until it can be migrated to using the EM framework. This will be done later, once the EM framework has support for at least all platforms currently supported by IPA. Signed-off-by:
Quentin Perret <quentin.perret@arm.com> Tested-by:
-
- 22 Jan, 2019 1 commit
-
-
Rajendra Nayak authored
Now that the OPP bindings are updated to include an optional 'opp-level' property, add support to parse it from device tree and store it as part of dev_pm_opp structure. Also add and export an helper 'dev_pm_opp_get_level()' that can be used to get the level value read from device tree when present. Reviewed-by:
Stephen Boyd <swboyd@chromium.org> Acked-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by:
Andy Gross <andy.gross@linaro.org>
-
- 14 Dec, 2018 2 commits
-
-
Viresh Kumar authored
of_get_required_opp_performance_state() returns 0 on errors currently and a positive performance state otherwise. Since 0 is a valid performance state (representing off), it would be better if this routine returns negative values on error. That will also make it behave similar to dev_pm_opp_xlate_performance_state(), which also returns performance states and returns negative values on error. Change the return type of the function to "int" in order to return negative values. This doesn't have any users for now and so no other part of the kernel will be impacted with this change. Reviewed-by:
-
Viresh Kumar authored
Make _find_table_of_opp_np() more efficient by using of_get_parent() to find the parent OPP table node. Tested-by:
-
- 11 Dec, 2018 1 commit
-
-
Viresh Kumar authored
There is one case where we may end up with no "supply" directory for the OPPs in debugfs. That happens when the OPP core isn't managing the regulators for the device and the device's OPP do have microvolt property. It happens because the opp_table->regulator_count remains set to 0 and the debugfs routines don't add any supply directory in such a case. This commit fixes that by setting opp_table->regulator_count to 1 in that particular case. But to make everything work nicely and not break other parts of the core, regulator_count is defined as "int" now instead of "unsigned int" and it can have different special values now. It is set to -1 initially to mark it "uninitialized" and later only we set it to 0 or positive values after checking how many supplies are there. This also helps in finding the bugs where only few of the OPPs have the "opp-microvolt" property set and not all. Fixes: 1fae788e ("PM / OPP: Don't create debugfs "supply-0" directory unnecessarily") Reported-by:
-
- 23 Nov, 2018 1 commit
-
-
Viresh Kumar authored
We currently return error if more than one phandle is present in the "operating-points-v2" property, which is incorrect. We only want to check the count of phandles here and set index to 0 if only one phandle is present. Fix it. Fixes: 5ed4cecd ("OPP: Pass OPP table to _of_add_opp_table_v{1|2}()") Signed-off-by:
-
- 05 Nov, 2018 7 commits
-
-
Viresh Kumar authored
This isn't used anymore, remove it. Reviewed-by:
-
Viresh Kumar authored
The OPP core already has the performance state values for each of the genpd's OPPs and there is no need to call the genpd callback again to get the performance state for the case where the end device doesn't have an OPP table and has the "required-opps" property directly in its node. This commit renames of_genpd_opp_to_performance_state() as of_get_required_opp_performance_state() and moves it to the OPP core, as it is all about OPP stuff now. Reviewed-by:
-
Viresh Kumar authored
Now that all the infrastructure is in place to support multiple required OPPs, lets switch over to using it. A new internal routine _set_required_opps() takes care of updating performance state for all the required OPPs. With this the performance state updates are supported even when the end device needs to configure regulators as well, that wasn't the case earlier. The pstates were earlier stored in the end device's OPP structures, that also changes now as those values are stored in the genpd's OPP structures. And so we switch over to using pm_genpd_opp_to_performance_state() instead of of_genpd_opp_to_performance_state() to get performance state for the genpd OPPs. The routine _generic_set_opp_domain() is not required anymore and is removed. On errors we don't try to recover by reverting to old settings as things are really complex now and the calls here should never really fail unless there is a bug. There is no point increasing the complexity, for code which will never be executed. Reviewed-by:
-
Viresh Kumar authored
Multiple generic power domains for a consumer device are supported with the help of virtual devices, which are created for each consumer device - genpd pair. These are the device structures which are attached to the power domain and are required by the OPP core to set the performance state of the genpd. The helpers added by this commit are required to be called once for each of these virtual devices. These are required only if multiple domains are available for a device, otherwise the actual device structure will be used instead by the OPP core. The new helpers also support the complex cases where the consumer device wouldn't always require all the domains. For example, a camera may require only one power domain during normal operations but two during high resolution operations. The consumer driver can call dev_pm_opp_put_genpd_virt_dev(high_resolution_genpd_virt_dev) if it is currently operating in the normal mode and doesn't have any performance requirements from the genpd which manages high resolution power requirements. The consumer driver can later call dev_pm_opp_set_genpd_virt_dev(high_resolution_genpd_virt_dev) once it switches back to the high resolution mode. The new helpers differ from other OPP set/put helpers as the new ones can be called with OPPs initialized for the table as we may need to call them on the fly because of the complex case explained above. For this reason it is possible that the genpd virt_dev structure may be used in parallel while the new helpers are running and a new mutex is added to protect against that. We didn't use the existing opp_table->lock mutex as that is widely used in the OPP core and we will need this lock in the dev_pm_opp_set_rate() helper while changing OPP and we need to make sure there is not much contention while doing that as that's the hotpath. Reviewed-by:
-
Viresh Kumar authored
An earlier commit populated the OPP tables from the "required-opps" property, this commit populates the individual OPPs. This is repeated for each OPP in the OPP table and these populated OPPs will be used by later commits. Reviewed-by:
-
Viresh Kumar authored
The current implementation works only for the case where a single phandle is present in the "required-opps" property, while DT allows multiple phandles to be present there. This patch adds new infrastructure to parse all the phandles present in "required-opps" property and save pointers of the required OPP's OPP tables. These will be used by later commits. Reviewed-by:
-
Viresh Kumar authored
We need to handle genpd OPP tables differently, this is already the case at one location and will be extended going forward. Add another field to the OPP table to check if the table belongs to a genpd or not. Reviewed-by:
-
- 04 Oct, 2018 2 commits
-
-
Dave Gerlach authored
Currently the _of_add_opp_table_v2 call loops through the OPP nodes in the operating-points-v2 table in the device tree and calls _opp_add_static_v2 for each to add them to the table. It counts each iteration through this loop as an added OPP, however there are cases where _opp_add_static_v2() returns 0 but no new OPP is added to the list. This can happen while adding duplicate OPP or if the OPP isn't supported by hardware. Because of this the count variable will contain the number of OPP nodes in the table in device tree but not necessarily the ones that are actually added. As this count value is what is checked to determine if there are any valid OPPs, if a platform has an operating-points-v2 table with all OPP nodes containing opp-supported-hw values that are not currently supported, then _of_add_opp_table_v2 will fail to abort as it should due to an empty table. Additionally, since commit 3ba98324 ("PM / OPP: Get performance state using genpd helper"), the same count variable is compared against the number of OPPs containing performance states and requires that either all or none have pstates set, however in the case of any opp table that has any entries that do not get added by _opp_add_static_v2 due to incompatible opp-supported-hw fields, these numbers will not match and _of_add_opp_table_v2 will incorrectly fail. We need to clearly identify all the three cases (success, failure, unsupported/duplicate OPPs) and then increment count only on success case. Change return type of _opp_add_static_v2() to return the pointer to the newly added OPP instead of an integer. This routine now returns a valid pointer if the OPP is really added, NULL for unsupported or duplicate OPPs, and error value cased as a pointer on errors. Ideally the fixes tag in this commit should point back to the commit that introduced OPP v2 initially, as that's where we started incorrectly accounting for duplicate OPPs: commit 27465902 ("PM / OPP: Add support to parse "operating-points-v2" bindings") But it wasn't a real problem until recently as the count was only used to check if any OPPs are added or not. And so this commit points to a rather recent commit where we added more code that depends on the value of "count". Fixes: 3ba98324 ("PM / OPP: Get performance state using genpd helper") Reported-by:
Dave Gerlach <d-gerlach@ti.com> Reported-by:
-
Viresh Kumar authored
The error handling wasn't appropriate in dev_pm_opp_of_cpumask_add_table(). For example it returns 0 on success and also for the case where cpumask is empty or cpu_device wasn't found for any of the CPUs. It should really return error on such cases, so that the callers can be aware of the outcome. Fix it. Signed-off-by:
-
- 01 Oct, 2018 2 commits
-
-
Viresh Kumar authored
Both _of_add_opp_table_v1() and _of_add_opp_table_v2() contain similar code to get the OPP table and their parent routine also parses the DT to find the OPP table's node pointer. This can be simplified by getting the OPP table in advance and then passing it as argument to these routines. Tested-by:
-
Viresh Kumar authored
When two or more devices are sharing their clock and voltage rails, they share the same OPP table. But there are some corner cases where the OPP core incorrectly creates separate OPP tables for them. For example, CPU 0 and 1 share clock/voltage rails. The platform specific code calls dev_pm_opp_set_regulators() for CPU0 and the OPP core creates an OPP table for it (the individual OPPs aren't initialized as of now). The same is repeated for CPU1 then. Because _opp_get_opp_table() doesn't compare DT node pointers currently, it fails to find the link between CPU0 and CPU1 and so creates a new OPP table. Fix this by calling _managed_opp() from _opp_get_opp_table(). _managed_opp() gain an additional argument (index) to get the right node pointer. This resulted in simplifying code in _of_add_opp_table_v2() as well. Tested-by:
-
- 19 Sep, 2018 1 commit
-
-
Viresh Kumar authored
Currently there are two separate ways to free the OPP table based on how it is created in the first place. We call _dev_pm_opp_remove_table() to free the static and/or dynamic OPP, OPP list devices, etc. This is done for the case where the OPP table is added while initializing the OPPs, like via the path dev_pm_opp_of_add_table(). We also call dev_pm_opp_put_opp_table() in some cases which eventually frees the OPP table structure once the reference count reaches 0. This is used by the first case as well as other cases like dev_pm_opp_set_regulators() where the OPPs aren't necessarily initialized at this point. This whole thing is a bit unclear and messy and obstruct any further cleanup/fixup of OPP core. This patch tries to streamline this by keeping a single path for OPP table destruction, i.e. dev_pm_opp_put_opp_table(). All the cleanup happens in _opp_table_kref_release() now after the reference count reaches 0. _dev_pm_opp_remove_table() is removed as it isn't required anymore. We don't drop the reference to the OPP table after creating it from _of_add_opp_table_v{1|2}() anymore and the same is dropped only when we try to remove them. Tested-by:
-
