linux/drivers/opp/core.c

2949 lines
77 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/*
* Generic OPP Interface
*
* Copyright (C) 2009-2010 Texas Instruments Incorporated.
* Nishanth Menon
* Romit Dasgupta
* Kevin Hilman
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/clk.h>
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
#include <linux/errno.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/device.h>
#include <linux/export.h>
#include <linux/pm_domain.h>
#include <linux/regulator/consumer.h>
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
#include "opp.h"
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/*
* The root of the list of all opp-tables. All opp_table structures branch off
* from here, with each opp_table containing the list of opps it supports in
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* various states of availability.
*/
LIST_HEAD(opp_tables);
/* OPP tables with uninitialized required OPPs */
LIST_HEAD(lazy_opp_tables);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/* Lock to allow exclusive modification to the device and opp lists */
DEFINE_MUTEX(opp_table_lock);
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
/* Flag indicating that opp_tables list is being updated at the moment */
static bool opp_tables_busy;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
static bool _find_opp_dev(const struct device *dev, struct opp_table *opp_table)
{
struct opp_device *opp_dev;
bool found = false;
mutex_lock(&opp_table->lock);
list_for_each_entry(opp_dev, &opp_table->dev_list, node)
if (opp_dev->dev == dev) {
found = true;
break;
}
mutex_unlock(&opp_table->lock);
return found;
}
static struct opp_table *_find_opp_table_unlocked(struct device *dev)
{
struct opp_table *opp_table;
list_for_each_entry(opp_table, &opp_tables, node) {
if (_find_opp_dev(dev, opp_table)) {
_get_opp_table_kref(opp_table);
return opp_table;
}
}
return ERR_PTR(-ENODEV);
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* _find_opp_table() - find opp_table struct using device pointer
* @dev: device pointer used to lookup OPP table
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Search OPP table for one containing matching device.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Return: pointer to 'struct opp_table' if found, otherwise -ENODEV or
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* -EINVAL based on type of error.
*
* The callers must call dev_pm_opp_put_opp_table() after the table is used.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
struct opp_table *_find_opp_table(struct device *dev)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (IS_ERR_OR_NULL(dev)) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
pr_err("%s: Invalid parameters\n", __func__);
return ERR_PTR(-EINVAL);
}
mutex_lock(&opp_table_lock);
opp_table = _find_opp_table_unlocked(dev);
mutex_unlock(&opp_table_lock);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return opp_table;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
/**
* dev_pm_opp_get_voltage() - Gets the voltage corresponding to an opp
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @opp: opp for which voltage has to be returned for
*
* Return: voltage in micro volt corresponding to the opp, else
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* return 0
*
* This is useful only for devices with single power supply.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
unsigned long dev_pm_opp_get_voltage(struct dev_pm_opp *opp)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
if (IS_ERR_OR_NULL(opp)) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
pr_err("%s: Invalid parameters\n", __func__);
return 0;
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return opp->supplies[0].u_volt;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_voltage);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_get_freq() - Gets the frequency corresponding to an available opp
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @opp: opp for which frequency has to be returned for
*
* Return: frequency in hertz corresponding to the opp, else
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* return 0
*/
unsigned long dev_pm_opp_get_freq(struct dev_pm_opp *opp)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
if (IS_ERR_OR_NULL(opp)) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
pr_err("%s: Invalid parameters\n", __func__);
return 0;
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return opp->rate;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_freq);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_get_level() - Gets the level corresponding to an available opp
* @opp: opp for which level value has to be returned for
*
* Return: level read from device tree corresponding to the opp, else
* return 0.
*/
unsigned int dev_pm_opp_get_level(struct dev_pm_opp *opp)
{
if (IS_ERR_OR_NULL(opp) || !opp->available) {
pr_err("%s: Invalid parameters\n", __func__);
return 0;
}
return opp->level;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_level);
/**
* dev_pm_opp_get_required_pstate() - Gets the required performance state
* corresponding to an available opp
* @opp: opp for which performance state has to be returned for
* @index: index of the required opp
*
* Return: performance state read from device tree corresponding to the
* required opp, else return 0.
*/
unsigned int dev_pm_opp_get_required_pstate(struct dev_pm_opp *opp,
unsigned int index)
{
if (IS_ERR_OR_NULL(opp) || !opp->available ||
index >= opp->opp_table->required_opp_count) {
pr_err("%s: Invalid parameters\n", __func__);
return 0;
}
/* required-opps not fully initialized yet */
if (lazy_linking_pending(opp->opp_table))
return 0;
return opp->required_opps[index]->pstate;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_required_pstate);
/**
* dev_pm_opp_is_turbo() - Returns if opp is turbo OPP or not
* @opp: opp for which turbo mode is being verified
*
* Turbo OPPs are not for normal use, and can be enabled (under certain
* conditions) for short duration of times to finish high throughput work
* quickly. Running on them for longer times may overheat the chip.
*
* Return: true if opp is turbo opp, else false.
*/
bool dev_pm_opp_is_turbo(struct dev_pm_opp *opp)
{
if (IS_ERR_OR_NULL(opp) || !opp->available) {
pr_err("%s: Invalid parameters\n", __func__);
return false;
}
return opp->turbo;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_is_turbo);
/**
* dev_pm_opp_get_max_clock_latency() - Get max clock latency in nanoseconds
* @dev: device for which we do this operation
*
* Return: This function returns the max clock latency in nanoseconds.
*/
unsigned long dev_pm_opp_get_max_clock_latency(struct device *dev)
{
struct opp_table *opp_table;
unsigned long clock_latency_ns;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return 0;
clock_latency_ns = opp_table->clock_latency_ns_max;
dev_pm_opp_put_opp_table(opp_table);
return clock_latency_ns;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_max_clock_latency);
/**
* dev_pm_opp_get_max_volt_latency() - Get max voltage latency in nanoseconds
* @dev: device for which we do this operation
*
* Return: This function returns the max voltage latency in nanoseconds.
*/
unsigned long dev_pm_opp_get_max_volt_latency(struct device *dev)
{
struct opp_table *opp_table;
struct dev_pm_opp *opp;
struct regulator *reg;
unsigned long latency_ns = 0;
int ret, i, count;
struct {
unsigned long min;
unsigned long max;
} *uV;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return 0;
/* Regulator may not be required for the device */
if (!opp_table->regulators)
goto put_opp_table;
count = opp_table->regulator_count;
uV = kmalloc_array(count, sizeof(*uV), GFP_KERNEL);
if (!uV)
goto put_opp_table;
mutex_lock(&opp_table->lock);
for (i = 0; i < count; i++) {
uV[i].min = ~0;
uV[i].max = 0;
list_for_each_entry(opp, &opp_table->opp_list, node) {
if (!opp->available)
continue;
if (opp->supplies[i].u_volt_min < uV[i].min)
uV[i].min = opp->supplies[i].u_volt_min;
if (opp->supplies[i].u_volt_max > uV[i].max)
uV[i].max = opp->supplies[i].u_volt_max;
}
}
mutex_unlock(&opp_table->lock);
/*
* The caller needs to ensure that opp_table (and hence the regulator)
* isn't freed, while we are executing this routine.
*/
for (i = 0; i < count; i++) {
reg = opp_table->regulators[i];
ret = regulator_set_voltage_time(reg, uV[i].min, uV[i].max);
if (ret > 0)
latency_ns += ret * 1000;
}
kfree(uV);
put_opp_table:
dev_pm_opp_put_opp_table(opp_table);
return latency_ns;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_max_volt_latency);
/**
* dev_pm_opp_get_max_transition_latency() - Get max transition latency in
* nanoseconds
* @dev: device for which we do this operation
*
* Return: This function returns the max transition latency, in nanoseconds, to
* switch from one OPP to other.
*/
unsigned long dev_pm_opp_get_max_transition_latency(struct device *dev)
{
return dev_pm_opp_get_max_volt_latency(dev) +
dev_pm_opp_get_max_clock_latency(dev);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_max_transition_latency);
/**
* dev_pm_opp_get_suspend_opp_freq() - Get frequency of suspend opp in Hz
* @dev: device for which we do this operation
*
* Return: This function returns the frequency of the OPP marked as suspend_opp
* if one is available, else returns 0;
*/
unsigned long dev_pm_opp_get_suspend_opp_freq(struct device *dev)
{
struct opp_table *opp_table;
unsigned long freq = 0;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return 0;
if (opp_table->suspend_opp && opp_table->suspend_opp->available)
freq = dev_pm_opp_get_freq(opp_table->suspend_opp);
dev_pm_opp_put_opp_table(opp_table);
return freq;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_suspend_opp_freq);
int _get_opp_count(struct opp_table *opp_table)
{
struct dev_pm_opp *opp;
int count = 0;
mutex_lock(&opp_table->lock);
list_for_each_entry(opp, &opp_table->opp_list, node) {
if (opp->available)
count++;
}
mutex_unlock(&opp_table->lock);
return count;
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_get_opp_count() - Get number of opps available in the opp table
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
*
* Return: This function returns the number of available opps if there are any,
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* else returns 0 if none or the corresponding error value.
*/
int dev_pm_opp_get_opp_count(struct device *dev)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
int count;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
count = PTR_ERR(opp_table);
dev_dbg(dev, "%s: OPP table not found (%d)\n",
__func__, count);
return count;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
count = _get_opp_count(opp_table);
dev_pm_opp_put_opp_table(opp_table);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return count;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_opp_count);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_find_freq_exact() - search for an exact frequency
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: frequency to search for
* @available: true/false - match for available opp
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Return: Searches for exact match in the opp table and returns pointer to the
* matching opp if found, else returns ERR_PTR in case of error and should
* be handled using IS_ERR. Error return values can be:
* EINVAL: for bad pointer
* ERANGE: no match found for search
* ENODEV: if device not found in list of registered devices
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Note: available is a modifier for the search. if available=true, then the
* match is for exact matching frequency and is available in the stored OPP
* table. if false, the match is for exact frequency which is not available.
*
* This provides a mechanism to enable an opp which is not available currently
* or the opposite as well.
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
struct dev_pm_opp *dev_pm_opp_find_freq_exact(struct device *dev,
unsigned long freq,
bool available)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
int r = PTR_ERR(opp_table);
dev_err(dev, "%s: OPP table not found (%d)\n", __func__, r);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return ERR_PTR(r);
}
mutex_lock(&opp_table->lock);
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (temp_opp->available == available &&
temp_opp->rate == freq) {
opp = temp_opp;
/* Increment the reference count of OPP */
dev_pm_opp_get(opp);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
break;
}
}
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_exact);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_find_level_exact() - search for an exact level
* @dev: device for which we do this operation
* @level: level to search for
*
* Return: Searches for exact match in the opp table and returns pointer to the
* matching opp if found, else returns ERR_PTR in case of error and should
* be handled using IS_ERR. Error return values can be:
* EINVAL: for bad pointer
* ERANGE: no match found for search
* ENODEV: if device not found in list of registered devices
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
*/
struct dev_pm_opp *dev_pm_opp_find_level_exact(struct device *dev,
unsigned int level)
{
struct opp_table *opp_table;
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
int r = PTR_ERR(opp_table);
dev_err(dev, "%s: OPP table not found (%d)\n", __func__, r);
return ERR_PTR(r);
}
mutex_lock(&opp_table->lock);
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
if (temp_opp->level == level) {
opp = temp_opp;
/* Increment the reference count of OPP */
dev_pm_opp_get(opp);
break;
}
}
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
return opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_level_exact);
/**
* dev_pm_opp_find_level_ceil() - search for an rounded up level
* @dev: device for which we do this operation
* @level: level to search for
*
* Return: Searches for rounded up match in the opp table and returns pointer
* to the matching opp if found, else returns ERR_PTR in case of error and
* should be handled using IS_ERR. Error return values can be:
* EINVAL: for bad pointer
* ERANGE: no match found for search
* ENODEV: if device not found in list of registered devices
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
*/
struct dev_pm_opp *dev_pm_opp_find_level_ceil(struct device *dev,
unsigned int *level)
{
struct opp_table *opp_table;
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
int r = PTR_ERR(opp_table);
dev_err(dev, "%s: OPP table not found (%d)\n", __func__, r);
return ERR_PTR(r);
}
mutex_lock(&opp_table->lock);
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
if (temp_opp->available && temp_opp->level >= *level) {
opp = temp_opp;
*level = opp->level;
/* Increment the reference count of OPP */
dev_pm_opp_get(opp);
break;
}
}
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
return opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_level_ceil);
static noinline struct dev_pm_opp *_find_freq_ceil(struct opp_table *opp_table,
unsigned long *freq)
{
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
mutex_lock(&opp_table->lock);
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
if (temp_opp->available && temp_opp->rate >= *freq) {
opp = temp_opp;
*freq = opp->rate;
/* Increment the reference count of OPP */
dev_pm_opp_get(opp);
break;
}
}
mutex_unlock(&opp_table->lock);
return opp;
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_find_freq_ceil() - Search for an rounded ceil freq
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: Start frequency
*
* Search for the matching ceil *available* OPP from a starting freq
* for a device.
*
* Return: matching *opp and refreshes *freq accordingly, else returns
* ERR_PTR in case of error and should be handled using IS_ERR. Error return
* values can be:
* EINVAL: for bad pointer
* ERANGE: no match found for search
* ENODEV: if device not found in list of registered devices
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
struct dev_pm_opp *dev_pm_opp_find_freq_ceil(struct device *dev,
unsigned long *freq)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
struct dev_pm_opp *opp;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (!dev || !freq) {
dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
return ERR_PTR(-EINVAL);
}
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return ERR_CAST(opp_table);
opp = _find_freq_ceil(opp_table, freq);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
dev_pm_opp_put_opp_table(opp_table);
return opp;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_ceil);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_find_freq_floor() - Search for a rounded floor freq
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: Start frequency
*
* Search for the matching floor *available* OPP from a starting freq
* for a device.
*
* Return: matching *opp and refreshes *freq accordingly, else returns
* ERR_PTR in case of error and should be handled using IS_ERR. Error return
* values can be:
* EINVAL: for bad pointer
* ERANGE: no match found for search
* ENODEV: if device not found in list of registered devices
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
struct dev_pm_opp *dev_pm_opp_find_freq_floor(struct device *dev,
unsigned long *freq)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (!dev || !freq) {
dev_err(dev, "%s: Invalid argument freq=%p\n", __func__, freq);
return ERR_PTR(-EINVAL);
}
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return ERR_CAST(opp_table);
mutex_lock(&opp_table->lock);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (temp_opp->available) {
/* go to the next node, before choosing prev */
if (temp_opp->rate > *freq)
break;
else
opp = temp_opp;
}
}
/* Increment the reference count of OPP */
if (!IS_ERR(opp))
dev_pm_opp_get(opp);
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (!IS_ERR(opp))
*freq = opp->rate;
return opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_floor);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_find_freq_ceil_by_volt() - Find OPP with highest frequency for
* target voltage.
* @dev: Device for which we do this operation.
* @u_volt: Target voltage.
*
* Search for OPP with highest (ceil) frequency and has voltage <= u_volt.
*
* Return: matching *opp, else returns ERR_PTR in case of error which should be
* handled using IS_ERR.
*
* Error return values can be:
* EINVAL: bad parameters
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
*/
struct dev_pm_opp *dev_pm_opp_find_freq_ceil_by_volt(struct device *dev,
unsigned long u_volt)
{
struct opp_table *opp_table;
struct dev_pm_opp *temp_opp, *opp = ERR_PTR(-ERANGE);
if (!dev || !u_volt) {
dev_err(dev, "%s: Invalid argument volt=%lu\n", __func__,
u_volt);
return ERR_PTR(-EINVAL);
}
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return ERR_CAST(opp_table);
mutex_lock(&opp_table->lock);
list_for_each_entry(temp_opp, &opp_table->opp_list, node) {
if (temp_opp->available) {
if (temp_opp->supplies[0].u_volt > u_volt)
break;
opp = temp_opp;
}
}
/* Increment the reference count of OPP */
if (!IS_ERR(opp))
dev_pm_opp_get(opp);
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
return opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_ceil_by_volt);
static int _set_opp_voltage(struct device *dev, struct regulator *reg,
struct dev_pm_opp_supply *supply)
{
int ret;
/* Regulator not available for device */
if (IS_ERR(reg)) {
dev_dbg(dev, "%s: regulator not available: %ld\n", __func__,
PTR_ERR(reg));
return 0;
}
dev_dbg(dev, "%s: voltages (mV): %lu %lu %lu\n", __func__,
supply->u_volt_min, supply->u_volt, supply->u_volt_max);
ret = regulator_set_voltage_triplet(reg, supply->u_volt_min,
supply->u_volt, supply->u_volt_max);
if (ret)
dev_err(dev, "%s: failed to set voltage (%lu %lu %lu mV): %d\n",
__func__, supply->u_volt_min, supply->u_volt,
supply->u_volt_max, ret);
return ret;
}
static inline int _generic_set_opp_clk_only(struct device *dev, struct clk *clk,
unsigned long freq)
{
int ret;
/* We may reach here for devices which don't change frequency */
if (IS_ERR(clk))
return 0;
ret = clk_set_rate(clk, freq);
if (ret) {
dev_err(dev, "%s: failed to set clock rate: %d\n", __func__,
ret);
}
return ret;
}
static int _generic_set_opp_regulator(struct opp_table *opp_table,
struct device *dev,
struct dev_pm_opp *opp,
unsigned long freq,
int scaling_down)
{
struct regulator *reg = opp_table->regulators[0];
struct dev_pm_opp *old_opp = opp_table->current_opp;
int ret;
/* This function only supports single regulator per device */
if (WARN_ON(opp_table->regulator_count > 1)) {
dev_err(dev, "multiple regulators are not supported\n");
return -EINVAL;
}
/* Scaling up? Scale voltage before frequency */
if (!scaling_down) {
ret = _set_opp_voltage(dev, reg, opp->supplies);
if (ret)
goto restore_voltage;
}
/* Change frequency */
ret = _generic_set_opp_clk_only(dev, opp_table->clk, freq);
if (ret)
goto restore_voltage;
/* Scaling down? Scale voltage after frequency */
if (scaling_down) {
ret = _set_opp_voltage(dev, reg, opp->supplies);
if (ret)
goto restore_freq;
}
/*
* Enable the regulator after setting its voltages, otherwise it breaks
* some boot-enabled regulators.
*/
if (unlikely(!opp_table->enabled)) {
ret = regulator_enable(reg);
if (ret < 0)
dev_warn(dev, "Failed to enable regulator: %d", ret);
}
return 0;
restore_freq:
if (_generic_set_opp_clk_only(dev, opp_table->clk, old_opp->rate))
dev_err(dev, "%s: failed to restore old-freq (%lu Hz)\n",
__func__, old_opp->rate);
restore_voltage:
/* This shouldn't harm even if the voltages weren't updated earlier */
_set_opp_voltage(dev, reg, old_opp->supplies);
return ret;
}
static int _set_opp_bw(const struct opp_table *opp_table,
struct dev_pm_opp *opp, struct device *dev)
{
u32 avg, peak;
int i, ret;
if (!opp_table->paths)
return 0;
for (i = 0; i < opp_table->path_count; i++) {
if (!opp) {
avg = 0;
peak = 0;
} else {
avg = opp->bandwidth[i].avg;
peak = opp->bandwidth[i].peak;
}
ret = icc_set_bw(opp_table->paths[i], avg, peak);
if (ret) {
dev_err(dev, "Failed to %s bandwidth[%d]: %d\n",
opp ? "set" : "remove", i, ret);
return ret;
}
}
return 0;
}
static int _set_opp_custom(const struct opp_table *opp_table,
struct device *dev, struct dev_pm_opp *opp,
unsigned long freq)
{
struct dev_pm_set_opp_data *data = opp_table->set_opp_data;
struct dev_pm_opp *old_opp = opp_table->current_opp;
int size;
/*
* We support this only if dev_pm_opp_set_regulators() was called
* earlier.
*/
if (opp_table->sod_supplies) {
size = sizeof(*old_opp->supplies) * opp_table->regulator_count;
memcpy(data->old_opp.supplies, old_opp->supplies, size);
memcpy(data->new_opp.supplies, opp->supplies, size);
data->regulator_count = opp_table->regulator_count;
} else {
data->regulator_count = 0;
}
data->regulators = opp_table->regulators;
data->clk = opp_table->clk;
data->dev = dev;
data->old_opp.rate = old_opp->rate;
data->new_opp.rate = freq;
return opp_table->set_opp(data);
}
static int _set_required_opp(struct device *dev, struct device *pd_dev,
struct dev_pm_opp *opp, int i)
{
unsigned int pstate = likely(opp) ? opp->required_opps[i]->pstate : 0;
int ret;
if (!pd_dev)
return 0;
ret = dev_pm_genpd_set_performance_state(pd_dev, pstate);
if (ret) {
dev_err(dev, "Failed to set performance rate of %s: %d (%d)\n",
dev_name(pd_dev), pstate, ret);
}
return ret;
}
/* This is only called for PM domain for now */
static int _set_required_opps(struct device *dev,
struct opp_table *opp_table,
struct dev_pm_opp *opp, bool up)
{
struct opp_table **required_opp_tables = opp_table->required_opp_tables;
struct device **genpd_virt_devs = opp_table->genpd_virt_devs;
int i, ret = 0;
if (!required_opp_tables)
return 0;
/*
* We only support genpd's OPPs in the "required-opps" for now, as we
* don't know much about other use cases. Error out if the required OPP
* doesn't belong to a genpd.
*/
if (unlikely(!required_opp_tables[0]->is_genpd)) {
dev_err(dev, "required-opps don't belong to a genpd\n");
return -ENOENT;
}
/* required-opps not fully initialized yet */
if (lazy_linking_pending(opp_table))
return -EBUSY;
/* Single genpd case */
if (!genpd_virt_devs)
return _set_required_opp(dev, dev, opp, 0);
/* Multiple genpd case */
/*
* Acquire genpd_virt_dev_lock to make sure we don't use a genpd_dev
* after it is freed from another thread.
*/
mutex_lock(&opp_table->genpd_virt_dev_lock);
/* Scaling up? Set required OPPs in normal order, else reverse */
if (up) {
for (i = 0; i < opp_table->required_opp_count; i++) {
ret = _set_required_opp(dev, genpd_virt_devs[i], opp, i);
if (ret)
break;
}
} else {
for (i = opp_table->required_opp_count - 1; i >= 0; i--) {
ret = _set_required_opp(dev, genpd_virt_devs[i], opp, i);
if (ret)
break;
}
}
mutex_unlock(&opp_table->genpd_virt_dev_lock);
return ret;
}
static void _find_current_opp(struct device *dev, struct opp_table *opp_table)
{
struct dev_pm_opp *opp = ERR_PTR(-ENODEV);
unsigned long freq;
if (!IS_ERR(opp_table->clk)) {
freq = clk_get_rate(opp_table->clk);
opp = _find_freq_ceil(opp_table, &freq);
}
/*
* Unable to find the current OPP ? Pick the first from the list since
* it is in ascending order, otherwise rest of the code will need to
* make special checks to validate current_opp.
*/
if (IS_ERR(opp)) {
mutex_lock(&opp_table->lock);
opp = list_first_entry(&opp_table->opp_list, struct dev_pm_opp, node);
dev_pm_opp_get(opp);
mutex_unlock(&opp_table->lock);
}
opp_table->current_opp = opp;
}
static int _disable_opp_table(struct device *dev, struct opp_table *opp_table)
{
int ret;
if (!opp_table->enabled)
return 0;
/*
* Some drivers need to support cases where some platforms may
* have OPP table for the device, while others don't and
* opp_set_rate() just needs to behave like clk_set_rate().
*/
if (!_get_opp_count(opp_table))
return 0;
ret = _set_opp_bw(opp_table, NULL, dev);
if (ret)
return ret;
if (opp_table->regulators)
regulator_disable(opp_table->regulators[0]);
ret = _set_required_opps(dev, opp_table, NULL, false);
opp_table->enabled = false;
return ret;
}
static int _set_opp(struct device *dev, struct opp_table *opp_table,
struct dev_pm_opp *opp, unsigned long freq)
{
struct dev_pm_opp *old_opp;
int scaling_down, ret;
if (unlikely(!opp))
return _disable_opp_table(dev, opp_table);
/* Find the currently set OPP if we don't know already */
if (unlikely(!opp_table->current_opp))
_find_current_opp(dev, opp_table);
old_opp = opp_table->current_opp;
/* Return early if nothing to do */
if (old_opp == opp && opp_table->current_rate == freq &&
opp_table->enabled) {
dev_dbg(dev, "%s: OPPs are same, nothing to do\n", __func__);
return 0;
}
dev_dbg(dev, "%s: switching OPP: Freq %lu -> %lu Hz, Level %u -> %u, Bw %u -> %u\n",
__func__, opp_table->current_rate, freq, old_opp->level,
opp->level, old_opp->bandwidth ? old_opp->bandwidth[0].peak : 0,
opp->bandwidth ? opp->bandwidth[0].peak : 0);
scaling_down = _opp_compare_key(old_opp, opp);
if (scaling_down == -1)
scaling_down = 0;
/* Scaling up? Configure required OPPs before frequency */
if (!scaling_down) {
ret = _set_required_opps(dev, opp_table, opp, true);
if (ret) {
dev_err(dev, "Failed to set required opps: %d\n", ret);
return ret;
}
ret = _set_opp_bw(opp_table, opp, dev);
if (ret) {
dev_err(dev, "Failed to set bw: %d\n", ret);
return ret;
}
}
if (opp_table->set_opp) {
ret = _set_opp_custom(opp_table, dev, opp, freq);
} else if (opp_table->regulators) {
ret = _generic_set_opp_regulator(opp_table, dev, opp, freq,
scaling_down);
} else {
/* Only frequency scaling */
ret = _generic_set_opp_clk_only(dev, opp_table->clk, freq);
}
if (ret)
return ret;
/* Scaling down? Configure required OPPs after frequency */
if (scaling_down) {
ret = _set_opp_bw(opp_table, opp, dev);
if (ret) {
dev_err(dev, "Failed to set bw: %d\n", ret);
return ret;
}
ret = _set_required_opps(dev, opp_table, opp, false);
if (ret) {
dev_err(dev, "Failed to set required opps: %d\n", ret);
return ret;
}
}
opp_table->enabled = true;
dev_pm_opp_put(old_opp);
/* Make sure current_opp doesn't get freed */
dev_pm_opp_get(opp);
opp_table->current_opp = opp;
opp_table->current_rate = freq;
return ret;
}
/**
* dev_pm_opp_set_rate() - Configure new OPP based on frequency
* @dev: device for which we do this operation
* @target_freq: frequency to achieve
*
* This configures the power-supplies to the levels specified by the OPP
* corresponding to the target_freq, and programs the clock to a value <=
* target_freq, as rounded by clk_round_rate(). Device wanting to run at fmax
* provided by the opp, should have already rounded to the target OPP's
* frequency.
*/
int dev_pm_opp_set_rate(struct device *dev, unsigned long target_freq)
{
struct opp_table *opp_table;
unsigned long freq = 0, temp_freq;
struct dev_pm_opp *opp = NULL;
int ret;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
dev_err(dev, "%s: device's opp table doesn't exist\n", __func__);
return PTR_ERR(opp_table);
}
if (target_freq) {
/*
* For IO devices which require an OPP on some platforms/SoCs
* while just needing to scale the clock on some others
* we look for empty OPP tables with just a clock handle and
* scale only the clk. This makes dev_pm_opp_set_rate()
* equivalent to a clk_set_rate()
*/
if (!_get_opp_count(opp_table)) {
ret = _generic_set_opp_clk_only(dev, opp_table->clk, target_freq);
goto put_opp_table;
}
freq = clk_round_rate(opp_table->clk, target_freq);
if ((long)freq <= 0)
freq = target_freq;
/*
* The clock driver may support finer resolution of the
* frequencies than the OPP table, don't update the frequency we
* pass to clk_set_rate() here.
*/
temp_freq = freq;
opp = _find_freq_ceil(opp_table, &temp_freq);
if (IS_ERR(opp)) {
ret = PTR_ERR(opp);
dev_err(dev, "%s: failed to find OPP for freq %lu (%d)\n",
__func__, freq, ret);
goto put_opp_table;
}
}
ret = _set_opp(dev, opp_table, opp, freq);
if (target_freq)
dev_pm_opp_put(opp);
put_opp_table:
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_rate);
/**
* dev_pm_opp_set_opp() - Configure device for OPP
* @dev: device for which we do this operation
* @opp: OPP to set to
*
* This configures the device based on the properties of the OPP passed to this
* routine.
*
* Return: 0 on success, a negative error number otherwise.
*/
int dev_pm_opp_set_opp(struct device *dev, struct dev_pm_opp *opp)
{
struct opp_table *opp_table;
int ret;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
dev_err(dev, "%s: device opp doesn't exist\n", __func__);
return PTR_ERR(opp_table);
}
ret = _set_opp(dev, opp_table, opp, opp ? opp->rate : 0);
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_opp);
/* OPP-dev Helpers */
static void _remove_opp_dev(struct opp_device *opp_dev,
struct opp_table *opp_table)
{
opp_debug_unregister(opp_dev, opp_table);
list_del(&opp_dev->node);
kfree(opp_dev);
}
struct opp_device *_add_opp_dev(const struct device *dev,
struct opp_table *opp_table)
{
struct opp_device *opp_dev;
opp_dev = kzalloc(sizeof(*opp_dev), GFP_KERNEL);
if (!opp_dev)
return NULL;
/* Initialize opp-dev */
opp_dev->dev = dev;
mutex_lock(&opp_table->lock);
list_add(&opp_dev->node, &opp_table->dev_list);
mutex_unlock(&opp_table->lock);
/* Create debugfs entries for the opp_table */
opp_debug_register(opp_dev, opp_table);
return opp_dev;
}
static struct opp_table *_allocate_opp_table(struct device *dev, int index)
{
struct opp_table *opp_table;
struct opp_device *opp_dev;
int ret;
/*
* Allocate a new OPP table. In the infrequent case where a new
* device is needed to be added, we pay this penalty.
*/
opp_table = kzalloc(sizeof(*opp_table), GFP_KERNEL);
if (!opp_table)
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
return ERR_PTR(-ENOMEM);
mutex_init(&opp_table->lock);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
mutex_init(&opp_table->genpd_virt_dev_lock);
INIT_LIST_HEAD(&opp_table->dev_list);
INIT_LIST_HEAD(&opp_table->lazy);
/* Mark regulator count uninitialized */
opp_table->regulator_count = -1;
opp_dev = _add_opp_dev(dev, opp_table);
if (!opp_dev) {
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
ret = -ENOMEM;
goto err;
}
_of_init_opp_table(opp_table, dev, index);
/* Find interconnect path(s) for the device */
ret = dev_pm_opp_of_find_icc_paths(dev, opp_table);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (ret) {
if (ret == -EPROBE_DEFER)
goto remove_opp_dev;
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
dev_warn(dev, "%s: Error finding interconnect paths: %d\n",
__func__, ret);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
}
BLOCKING_INIT_NOTIFIER_HEAD(&opp_table->head);
INIT_LIST_HEAD(&opp_table->opp_list);
kref_init(&opp_table->kref);
return opp_table;
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
opp: fix memory leak in _allocate_opp_table In function _allocate_opp_table, opp_dev is allocated and referenced by opp_table via _add_opp_dev. But in the case that the subsequent calls return -EPROBE_DEFER, it will jump to err label and opp_table will be freed. Then opp_dev becomes an unreferenced object to cause memory leak. So let's call _remove_opp_dev to do the cleanup. This fixes the following kmemleak report: unreferenced object 0xffff000801524a00 (size 128): comm "swapper/0", pid 1, jiffies 4294892465 (age 84.616s) hex dump (first 32 bytes): 40 00 56 01 08 00 ff ff 40 00 56 01 08 00 ff ff @.V.....@.V..... b8 52 77 7f 08 00 ff ff 00 3c 4c 00 08 00 ff ff .Rw......<L..... backtrace: [<00000000b1289fb1>] kmemleak_alloc+0x30/0x40 [<0000000056da48f0>] kmem_cache_alloc+0x3d4/0x588 [<00000000a84b3b0e>] _add_opp_dev+0x2c/0x88 [<0000000062a380cd>] _add_opp_table_indexed+0x124/0x268 [<000000008b4c8f1f>] dev_pm_opp_of_add_table+0x20/0x1d8 [<00000000e5316798>] dev_pm_opp_of_cpumask_add_table+0x48/0xf0 [<00000000db0a8ec2>] dt_cpufreq_probe+0x20c/0x448 [<0000000030a3a26c>] platform_probe+0x68/0xd8 [<00000000c618e78d>] really_probe+0xd0/0x3a0 [<00000000642e856f>] driver_probe_device+0x58/0xb8 [<00000000f10f5307>] device_driver_attach+0x74/0x80 [<0000000004f254b8>] __driver_attach+0x58/0xe0 [<0000000009d5d19e>] bus_for_each_dev+0x70/0xc8 [<0000000000d22e1c>] driver_attach+0x24/0x30 [<0000000001d4e952>] bus_add_driver+0x14c/0x1f0 [<0000000089928aaa>] driver_register+0x64/0x120 Cc: v5.10 <stable@vger.kernel.org> # v5.10 Fixes: dd461cd9183f ("opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER") Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com> [ Viresh: Added the stable tag ] Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-12-24 18:49:27 +08:00
remove_opp_dev:
_remove_opp_dev(opp_dev, opp_table);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
err:
kfree(opp_table);
return ERR_PTR(ret);
}
void _get_opp_table_kref(struct opp_table *opp_table)
{
kref_get(&opp_table->kref);
}
static struct opp_table *_update_opp_table_clk(struct device *dev,
struct opp_table *opp_table,
bool getclk)
{
int ret;
/*
* Return early if we don't need to get clk or we have already tried it
* earlier.
*/
if (!getclk || IS_ERR(opp_table) || opp_table->clk)
return opp_table;
/* Find clk for the device */
opp_table->clk = clk_get(dev, NULL);
ret = PTR_ERR_OR_ZERO(opp_table->clk);
if (!ret)
return opp_table;
if (ret == -ENOENT) {
dev_dbg(dev, "%s: Couldn't find clock: %d\n", __func__, ret);
return opp_table;
}
dev_pm_opp_put_opp_table(opp_table);
dev_err_probe(dev, ret, "Couldn't find clock\n");
return ERR_PTR(ret);
}
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
/*
* We need to make sure that the OPP table for a device doesn't get added twice,
* if this routine gets called in parallel with the same device pointer.
*
* The simplest way to enforce that is to perform everything (find existing
* table and if not found, create a new one) under the opp_table_lock, so only
* one creator gets access to the same. But that expands the critical section
* under the lock and may end up causing circular dependencies with frameworks
* like debugfs, interconnect or clock framework as they may be direct or
* indirect users of OPP core.
*
* And for that reason we have to go for a bit tricky implementation here, which
* uses the opp_tables_busy flag to indicate if another creator is in the middle
* of adding an OPP table and others should wait for it to finish.
*/
struct opp_table *_add_opp_table_indexed(struct device *dev, int index,
bool getclk)
{
struct opp_table *opp_table;
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
again:
mutex_lock(&opp_table_lock);
opp_table = _find_opp_table_unlocked(dev);
if (!IS_ERR(opp_table))
goto unlock;
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
/*
* The opp_tables list or an OPP table's dev_list is getting updated by
* another user, wait for it to finish.
*/
if (unlikely(opp_tables_busy)) {
mutex_unlock(&opp_table_lock);
cpu_relax();
goto again;
}
opp_tables_busy = true;
opp_table = _managed_opp(dev, index);
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
/* Drop the lock to reduce the size of critical section */
mutex_unlock(&opp_table_lock);
if (opp_table) {
if (!_add_opp_dev(dev, opp_table)) {
dev_pm_opp_put_opp_table(opp_table);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
opp_table = ERR_PTR(-ENOMEM);
}
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
mutex_lock(&opp_table_lock);
} else {
opp_table = _allocate_opp_table(dev, index);
mutex_lock(&opp_table_lock);
if (!IS_ERR(opp_table))
list_add(&opp_table->node, &opp_tables);
}
opp: Allocate the OPP table outside of opp_table_lock There is no critical section which needs protection with locks while allocating an OPP table, except while adding it to the opp_tables list. And taking the opp_table_lock for the entire duration causes circular dependency issues like the one mentioned below. This patch takes another approach to reduce the size of the critical section to avoid such issues, the details of that are present within the patch. ====================================================== WARNING: possible circular locking dependency detected 5.4.72 #14 Not tainted ------------------------------------------------------ chrome/1865 is trying to acquire lock: ffffffdd34921750 (opp_table_lock){+.+.}, at: _find_opp_table+0x34/0x74 but task is already holding lock: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec fscrypt: AES-256-CTS-CBC using implementation "cts-cbc-aes-ce" which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #4 (reservation_ww_class_mutex){+.+.}: __mutex_lock_common+0xec/0xc0c ww_mutex_lock_interruptible+0x5c/0xc4 msm_gem_fault+0x2c/0x124 __do_fault+0x40/0x16c handle_mm_fault+0x7cc/0xd98 do_page_fault+0x230/0x3b4 do_translation_fault+0x5c/0x78 do_mem_abort+0x4c/0xb4 el0_da+0x1c/0x20 -> #3 (&mm->mmap_sem){++++}: __might_fault+0x70/0x98 compat_filldir+0xf8/0x48c dcache_readdir+0x70/0x1dc iterate_dir+0xd4/0x180 __arm64_compat_sys_getdents+0xa0/0x19c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 -> #2 (&sb->s_type->i_mutex_key#3){++++}: down_write+0x54/0x16c start_creating+0x68/0x128 debugfs_create_dir+0x28/0x114 opp_debug_register+0x8c/0xc0 _add_opp_dev_unlocked+0x5c/0x70 _add_opp_dev+0x38/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #1 (&opp_table->lock){+.+.}: __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _add_opp_dev+0x2c/0x58 _opp_get_opp_table+0xdc/0x1ac dev_pm_opp_get_opp_table_indexed+0x24/0x30 dev_pm_opp_of_add_table_indexed+0x48/0x84 of_genpd_add_provider_onecell+0xc0/0x1b8 rpmhpd_probe+0x240/0x268 platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_populate+0x70/0xbc devm_of_platform_populate+0x58/0xa0 rpmh_rsc_probe+0x36c/0x3cc platform_drv_probe+0x90/0xb0 really_probe+0x134/0x2ec driver_probe_device+0x64/0xfc __device_attach_driver+0x8c/0xa4 bus_for_each_drv+0x90/0xd8 __device_attach+0xc0/0x148 device_initial_probe+0x20/0x2c bus_probe_device+0x34/0x94 device_add+0x1fc/0x3b0 of_device_add+0x3c/0x4c of_platform_device_create_pdata+0xb8/0xfc of_platform_bus_create+0x1e4/0x368 of_platform_bus_create+0x230/0x368 of_platform_populate+0x70/0xbc of_platform_default_populate_init+0xa8/0xc0 do_one_initcall+0x1c8/0x3fc do_initcall_level+0xb4/0x10c do_basic_setup+0x30/0x48 kernel_init_freeable+0x124/0x1a4 kernel_init+0x14/0x104 ret_from_fork+0x10/0x18 -> #0 (opp_table_lock){+.+.}: __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 other info that might help us debug this: Chain exists of: opp_table_lock --> &mm->mmap_sem --> reservation_ww_class_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reservation_ww_class_mutex); lock(&mm->mmap_sem); lock(reservation_ww_class_mutex); lock(opp_table_lock); *** DEADLOCK *** 3 locks held by chrome/1865: #0: ffffff81edecc0d8 (&dev->struct_mutex){+.+.}, at: msm_ioctl_gem_submit+0x264/0xb60 #1: ffffff81d0000870 (reservation_ww_class_acquire){+.+.}, at: msm_ioctl_gem_submit+0x8e8/0xb60 #2: ffffff81f0fc71a8 (reservation_ww_class_mutex){+.+.}, at: submit_lock_objects+0x70/0x1ec stack backtrace: CPU: 0 PID: 1865 Comm: chrome Not tainted 5.4.72 #14 Hardware name: Google Lazor (rev1+) with LTE (DT) Call trace: dump_backtrace+0x0/0x158 show_stack+0x20/0x2c dump_stack+0xc8/0x160 print_circular_bug+0x2c4/0x2c8 check_noncircular+0x1a8/0x1b0 __lock_acquire+0xee4/0x2450 lock_acquire+0x1cc/0x210 __mutex_lock_common+0xec/0xc0c mutex_lock_nested+0x40/0x50 _find_opp_table+0x34/0x74 dev_pm_opp_find_freq_exact+0x2c/0xdc a6xx_gmu_resume+0xc8/0xecc a6xx_pm_resume+0x148/0x200 adreno_resume+0x28/0x34 pm_generic_runtime_resume+0x34/0x48 __rpm_callback+0x70/0x10c rpm_callback+0x34/0x8c rpm_resume+0x414/0x550 __pm_runtime_resume+0x7c/0xa0 msm_gpu_submit+0x60/0x1c0 msm_ioctl_gem_submit+0xadc/0xb60 drm_ioctl_kernel+0x9c/0x118 drm_ioctl+0x27c/0x408 drm_compat_ioctl+0xcc/0xdc __se_compat_sys_ioctl+0x100/0x206c __arm64_compat_sys_ioctl+0x20/0x2c el0_svc_common+0xa8/0x178 el0_svc_compat_handler+0x2c/0x40 el0_svc_compat+0x8/0x10 Reported-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2020-10-27 19:23:21 +08:00
opp_tables_busy = false;
unlock:
mutex_unlock(&opp_table_lock);
return _update_opp_table_clk(dev, opp_table, getclk);
}
static struct opp_table *_add_opp_table(struct device *dev, bool getclk)
{
return _add_opp_table_indexed(dev, 0, getclk);
}
struct opp_table *dev_pm_opp_get_opp_table(struct device *dev)
{
return _find_opp_table(dev);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_get_opp_table);
static void _opp_table_kref_release(struct kref *kref)
{
struct opp_table *opp_table = container_of(kref, struct opp_table, kref);
struct opp_device *opp_dev, *temp;
int i;
/* Drop the lock as soon as we can */
list_del(&opp_table->node);
mutex_unlock(&opp_table_lock);
if (opp_table->current_opp)
dev_pm_opp_put(opp_table->current_opp);
_of_clear_opp_table(opp_table);
/* Release clk */
if (!IS_ERR(opp_table->clk))
clk_put(opp_table->clk);
if (opp_table->paths) {
for (i = 0; i < opp_table->path_count; i++)
icc_put(opp_table->paths[i]);
kfree(opp_table->paths);
}
WARN_ON(!list_empty(&opp_table->opp_list));
list_for_each_entry_safe(opp_dev, temp, &opp_table->dev_list, node) {
/*
* The OPP table is getting removed, drop the performance state
* constraints.
*/
if (opp_table->genpd_performance_state)
dev_pm_genpd_set_performance_state((struct device *)(opp_dev->dev), 0);
_remove_opp_dev(opp_dev, opp_table);
}
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
mutex_destroy(&opp_table->genpd_virt_dev_lock);
mutex_destroy(&opp_table->lock);
kfree(opp_table);
}
void dev_pm_opp_put_opp_table(struct opp_table *opp_table)
{
kref_put_mutex(&opp_table->kref, _opp_table_kref_release,
&opp_table_lock);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put_opp_table);
void _opp_free(struct dev_pm_opp *opp)
{
kfree(opp);
}
static void _opp_kref_release(struct kref *kref)
{
struct dev_pm_opp *opp = container_of(kref, struct dev_pm_opp, kref);
struct opp_table *opp_table = opp->opp_table;
list_del(&opp->node);
mutex_unlock(&opp_table->lock);
/*
* Notify the changes in the availability of the operable
* frequency/voltage list.
*/
blocking_notifier_call_chain(&opp_table->head, OPP_EVENT_REMOVE, opp);
_of_opp_free_required_opps(opp_table, opp);
opp_debug_remove_one(opp);
kfree(opp);
}
void dev_pm_opp_get(struct dev_pm_opp *opp)
{
kref_get(&opp->kref);
}
void dev_pm_opp_put(struct dev_pm_opp *opp)
{
kref_put_mutex(&opp->kref, _opp_kref_release, &opp->opp_table->lock);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put);
/**
* dev_pm_opp_remove() - Remove an OPP from OPP table
* @dev: device for which we do this operation
* @freq: OPP to remove with matching 'freq'
*
* This function removes an opp from the opp table.
*/
void dev_pm_opp_remove(struct device *dev, unsigned long freq)
{
struct dev_pm_opp *opp;
struct opp_table *opp_table;
bool found = false;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return;
mutex_lock(&opp_table->lock);
list_for_each_entry(opp, &opp_table->opp_list, node) {
if (opp->rate == freq) {
found = true;
break;
}
}
mutex_unlock(&opp_table->lock);
if (found) {
dev_pm_opp_put(opp);
/* Drop the reference taken by dev_pm_opp_add() */
dev_pm_opp_put_opp_table(opp_table);
} else {
dev_warn(dev, "%s: Couldn't find OPP with freq: %lu\n",
__func__, freq);
}
/* Drop the reference taken by _find_opp_table() */
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_remove);
static struct dev_pm_opp *_opp_get_next(struct opp_table *opp_table,
bool dynamic)
{
struct dev_pm_opp *opp = NULL, *temp;
mutex_lock(&opp_table->lock);
list_for_each_entry(temp, &opp_table->opp_list, node) {
opp: Don't drop extra references to OPPs accidentally We are required to call dev_pm_opp_put() from outside of the opp_table->lock as debugfs removal needs to happen lock-less to avoid circular dependency issues. commit cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") tried to fix that introducing a new routine _opp_get_next() which keeps returning OPPs that can be freed by the callers and this routine shall be called without holding the opp_table->lock. Though the commit overlooked the fact that the OPPs can be referenced by other users as well and this routine will end up dropping references which were taken by other users and hence freeing the OPPs prematurely. In effect, other users of the OPPs will end up having invalid pointers at hand. We didn't see any crash reports earlier as the exact situation never happened, though it is certainly possible. We need a way to mark which OPPs are no longer referenced by the OPP core, so we don't drop extra references to them accidentally. This commit adds another OPP flag, "removed", which is used to track this. And now we should never end up dropping extra references to the OPPs. Cc: v5.11+ <stable@vger.kernel.org> # v5.11+ Fixes: cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") Signed-off-by: Beata Michalska <beata.michalska@arm.com> [ Viresh: Almost rewrote entire patch, added new "removed" field, rewrote commit log and added the correct Fixes tag. ] Co-developed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2021-03-04 23:07:34 +08:00
/*
* Refcount must be dropped only once for each OPP by OPP core,
* do that with help of "removed" flag.
*/
if (!temp->removed && dynamic == temp->dynamic) {
opp = temp;
break;
}
}
mutex_unlock(&opp_table->lock);
return opp;
}
opp: Don't drop extra references to OPPs accidentally We are required to call dev_pm_opp_put() from outside of the opp_table->lock as debugfs removal needs to happen lock-less to avoid circular dependency issues. commit cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") tried to fix that introducing a new routine _opp_get_next() which keeps returning OPPs that can be freed by the callers and this routine shall be called without holding the opp_table->lock. Though the commit overlooked the fact that the OPPs can be referenced by other users as well and this routine will end up dropping references which were taken by other users and hence freeing the OPPs prematurely. In effect, other users of the OPPs will end up having invalid pointers at hand. We didn't see any crash reports earlier as the exact situation never happened, though it is certainly possible. We need a way to mark which OPPs are no longer referenced by the OPP core, so we don't drop extra references to them accidentally. This commit adds another OPP flag, "removed", which is used to track this. And now we should never end up dropping extra references to the OPPs. Cc: v5.11+ <stable@vger.kernel.org> # v5.11+ Fixes: cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") Signed-off-by: Beata Michalska <beata.michalska@arm.com> [ Viresh: Almost rewrote entire patch, added new "removed" field, rewrote commit log and added the correct Fixes tag. ] Co-developed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2021-03-04 23:07:34 +08:00
/*
* Can't call dev_pm_opp_put() from under the lock as debugfs removal needs to
* happen lock less to avoid circular dependency issues. This routine must be
* called without the opp_table->lock held.
*/
static void _opp_remove_all(struct opp_table *opp_table, bool dynamic)
{
struct dev_pm_opp *opp;
opp: Don't drop extra references to OPPs accidentally We are required to call dev_pm_opp_put() from outside of the opp_table->lock as debugfs removal needs to happen lock-less to avoid circular dependency issues. commit cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") tried to fix that introducing a new routine _opp_get_next() which keeps returning OPPs that can be freed by the callers and this routine shall be called without holding the opp_table->lock. Though the commit overlooked the fact that the OPPs can be referenced by other users as well and this routine will end up dropping references which were taken by other users and hence freeing the OPPs prematurely. In effect, other users of the OPPs will end up having invalid pointers at hand. We didn't see any crash reports earlier as the exact situation never happened, though it is certainly possible. We need a way to mark which OPPs are no longer referenced by the OPP core, so we don't drop extra references to them accidentally. This commit adds another OPP flag, "removed", which is used to track this. And now we should never end up dropping extra references to the OPPs. Cc: v5.11+ <stable@vger.kernel.org> # v5.11+ Fixes: cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") Signed-off-by: Beata Michalska <beata.michalska@arm.com> [ Viresh: Almost rewrote entire patch, added new "removed" field, rewrote commit log and added the correct Fixes tag. ] Co-developed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2021-03-04 23:07:34 +08:00
while ((opp = _opp_get_next(opp_table, dynamic))) {
opp->removed = true;
dev_pm_opp_put(opp);
/* Drop the references taken by dev_pm_opp_add() */
if (dynamic)
dev_pm_opp_put_opp_table(opp_table);
}
}
bool _opp_remove_all_static(struct opp_table *opp_table)
{
mutex_lock(&opp_table->lock);
if (!opp_table->parsed_static_opps) {
mutex_unlock(&opp_table->lock);
return false;
}
if (--opp_table->parsed_static_opps) {
mutex_unlock(&opp_table->lock);
return true;
}
mutex_unlock(&opp_table->lock);
opp: Don't drop extra references to OPPs accidentally We are required to call dev_pm_opp_put() from outside of the opp_table->lock as debugfs removal needs to happen lock-less to avoid circular dependency issues. commit cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") tried to fix that introducing a new routine _opp_get_next() which keeps returning OPPs that can be freed by the callers and this routine shall be called without holding the opp_table->lock. Though the commit overlooked the fact that the OPPs can be referenced by other users as well and this routine will end up dropping references which were taken by other users and hence freeing the OPPs prematurely. In effect, other users of the OPPs will end up having invalid pointers at hand. We didn't see any crash reports earlier as the exact situation never happened, though it is certainly possible. We need a way to mark which OPPs are no longer referenced by the OPP core, so we don't drop extra references to them accidentally. This commit adds another OPP flag, "removed", which is used to track this. And now we should never end up dropping extra references to the OPPs. Cc: v5.11+ <stable@vger.kernel.org> # v5.11+ Fixes: cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") Signed-off-by: Beata Michalska <beata.michalska@arm.com> [ Viresh: Almost rewrote entire patch, added new "removed" field, rewrote commit log and added the correct Fixes tag. ] Co-developed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2021-03-04 23:07:34 +08:00
_opp_remove_all(opp_table, false);
return true;
}
/**
* dev_pm_opp_remove_all_dynamic() - Remove all dynamically created OPPs
* @dev: device for which we do this operation
*
* This function removes all dynamically created OPPs from the opp table.
*/
void dev_pm_opp_remove_all_dynamic(struct device *dev)
{
struct opp_table *opp_table;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return;
opp: Don't drop extra references to OPPs accidentally We are required to call dev_pm_opp_put() from outside of the opp_table->lock as debugfs removal needs to happen lock-less to avoid circular dependency issues. commit cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") tried to fix that introducing a new routine _opp_get_next() which keeps returning OPPs that can be freed by the callers and this routine shall be called without holding the opp_table->lock. Though the commit overlooked the fact that the OPPs can be referenced by other users as well and this routine will end up dropping references which were taken by other users and hence freeing the OPPs prematurely. In effect, other users of the OPPs will end up having invalid pointers at hand. We didn't see any crash reports earlier as the exact situation never happened, though it is certainly possible. We need a way to mark which OPPs are no longer referenced by the OPP core, so we don't drop extra references to them accidentally. This commit adds another OPP flag, "removed", which is used to track this. And now we should never end up dropping extra references to the OPPs. Cc: v5.11+ <stable@vger.kernel.org> # v5.11+ Fixes: cf1fac943c63 ("opp: Reduce the size of critical section in _opp_kref_release()") Signed-off-by: Beata Michalska <beata.michalska@arm.com> [ Viresh: Almost rewrote entire patch, added new "removed" field, rewrote commit log and added the correct Fixes tag. ] Co-developed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2021-03-04 23:07:34 +08:00
_opp_remove_all(opp_table, true);
/* Drop the reference taken by _find_opp_table() */
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_remove_all_dynamic);
struct dev_pm_opp *_opp_allocate(struct opp_table *table)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct dev_pm_opp *opp;
int supply_count, supply_size, icc_size;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/* Allocate space for at least one supply */
supply_count = table->regulator_count > 0 ? table->regulator_count : 1;
supply_size = sizeof(*opp->supplies) * supply_count;
icc_size = sizeof(*opp->bandwidth) * table->path_count;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/* allocate new OPP node and supplies structures */
opp = kzalloc(sizeof(*opp) + supply_size + icc_size, GFP_KERNEL);
if (!opp)
return NULL;
/* Put the supplies at the end of the OPP structure as an empty array */
opp->supplies = (struct dev_pm_opp_supply *)(opp + 1);
if (icc_size)
opp->bandwidth = (struct dev_pm_opp_icc_bw *)(opp->supplies + supply_count);
INIT_LIST_HEAD(&opp->node);
return opp;
}
static bool _opp_supported_by_regulators(struct dev_pm_opp *opp,
struct opp_table *opp_table)
{
struct regulator *reg;
int i;
if (!opp_table->regulators)
return true;
for (i = 0; i < opp_table->regulator_count; i++) {
reg = opp_table->regulators[i];
if (!regulator_is_supported_voltage(reg,
opp->supplies[i].u_volt_min,
opp->supplies[i].u_volt_max)) {
pr_warn("%s: OPP minuV: %lu maxuV: %lu, not supported by regulator\n",
__func__, opp->supplies[i].u_volt_min,
opp->supplies[i].u_volt_max);
return false;
}
}
return true;
}
int _opp_compare_key(struct dev_pm_opp *opp1, struct dev_pm_opp *opp2)
{
if (opp1->rate != opp2->rate)
return opp1->rate < opp2->rate ? -1 : 1;
if (opp1->bandwidth && opp2->bandwidth &&
opp1->bandwidth[0].peak != opp2->bandwidth[0].peak)
return opp1->bandwidth[0].peak < opp2->bandwidth[0].peak ? -1 : 1;
if (opp1->level != opp2->level)
return opp1->level < opp2->level ? -1 : 1;
return 0;
}
static int _opp_is_duplicate(struct device *dev, struct dev_pm_opp *new_opp,
struct opp_table *opp_table,
struct list_head **head)
{
struct dev_pm_opp *opp;
int opp_cmp;
/*
* Insert new OPP in order of increasing frequency and discard if
* already present.
*
* Need to use &opp_table->opp_list in the condition part of the 'for'
* loop, don't replace it with head otherwise it will become an infinite
* loop.
*/
list_for_each_entry(opp, &opp_table->opp_list, node) {
opp_cmp = _opp_compare_key(new_opp, opp);
if (opp_cmp > 0) {
*head = &opp->node;
continue;
}
if (opp_cmp < 0)
return 0;
/* Duplicate OPPs */
dev_warn(dev, "%s: duplicate OPPs detected. Existing: freq: %lu, volt: %lu, enabled: %d. New: freq: %lu, volt: %lu, enabled: %d\n",
__func__, opp->rate, opp->supplies[0].u_volt,
opp->available, new_opp->rate,
new_opp->supplies[0].u_volt, new_opp->available);
/* Should we compare voltages for all regulators here ? */
return opp->available &&
new_opp->supplies[0].u_volt == opp->supplies[0].u_volt ? -EBUSY : -EEXIST;
}
return 0;
}
void _required_opps_available(struct dev_pm_opp *opp, int count)
{
int i;
for (i = 0; i < count; i++) {
if (opp->required_opps[i]->available)
continue;
opp->available = false;
pr_warn("%s: OPP not supported by required OPP %pOF (%lu)\n",
__func__, opp->required_opps[i]->np, opp->rate);
return;
}
}
/*
* Returns:
* 0: On success. And appropriate error message for duplicate OPPs.
* -EBUSY: For OPP with same freq/volt and is available. The callers of
* _opp_add() must return 0 if they receive -EBUSY from it. This is to make
* sure we don't print error messages unnecessarily if different parts of
* kernel try to initialize the OPP table.
* -EEXIST: For OPP with same freq but different volt or is unavailable. This
* should be considered an error by the callers of _opp_add().
*/
int _opp_add(struct device *dev, struct dev_pm_opp *new_opp,
struct opp_table *opp_table, bool rate_not_available)
{
struct list_head *head;
int ret;
mutex_lock(&opp_table->lock);
head = &opp_table->opp_list;
ret = _opp_is_duplicate(dev, new_opp, opp_table, &head);
if (ret) {
mutex_unlock(&opp_table->lock);
return ret;
}
list_add(&new_opp->node, head);
mutex_unlock(&opp_table->lock);
new_opp->opp_table = opp_table;
kref_init(&new_opp->kref);
opp_debug_create_one(new_opp, opp_table);
if (!_opp_supported_by_regulators(new_opp, opp_table)) {
new_opp->available = false;
dev_warn(dev, "%s: OPP not supported by regulators (%lu)\n",
__func__, new_opp->rate);
}
/* required-opps not fully initialized yet */
if (lazy_linking_pending(opp_table))
return 0;
_required_opps_available(new_opp, opp_table->required_opp_count);
return 0;
}
/**
* _opp_add_v1() - Allocate a OPP based on v1 bindings.
* @opp_table: OPP table
* @dev: device for which we do this operation
* @freq: Frequency in Hz for this OPP
* @u_volt: Voltage in uVolts for this OPP
* @dynamic: Dynamically added OPPs.
*
* This function adds an opp definition to the opp table and returns status.
* The opp is made available by default and it can be controlled using
* dev_pm_opp_enable/disable functions and may be removed by dev_pm_opp_remove.
*
* NOTE: "dynamic" parameter impacts OPPs added by the dev_pm_opp_of_add_table
* and freed by dev_pm_opp_of_remove_table.
*
* Return:
* 0 On success OR
* Duplicate OPPs (both freq and volt are same) and opp->available
* -EEXIST Freq are same and volt are different OR
* Duplicate OPPs (both freq and volt are same) and !opp->available
* -ENOMEM Memory allocation failure
*/
int _opp_add_v1(struct opp_table *opp_table, struct device *dev,
unsigned long freq, long u_volt, bool dynamic)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct dev_pm_opp *new_opp;
unsigned long tol;
int ret;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
new_opp = _opp_allocate(opp_table);
if (!new_opp)
return -ENOMEM;
/* populate the opp table */
new_opp->rate = freq;
tol = u_volt * opp_table->voltage_tolerance_v1 / 100;
new_opp->supplies[0].u_volt = u_volt;
new_opp->supplies[0].u_volt_min = u_volt - tol;
new_opp->supplies[0].u_volt_max = u_volt + tol;
new_opp->available = true;
new_opp->dynamic = dynamic;
ret = _opp_add(dev, new_opp, opp_table, false);
if (ret) {
/* Don't return error for duplicate OPPs */
if (ret == -EBUSY)
ret = 0;
goto free_opp;
}
/*
* Notify the changes in the availability of the operable
* frequency/voltage list.
*/
blocking_notifier_call_chain(&opp_table->head, OPP_EVENT_ADD, new_opp);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return 0;
free_opp:
_opp_free(new_opp);
return ret;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
/**
* dev_pm_opp_set_supported_hw() - Set supported platforms
* @dev: Device for which supported-hw has to be set.
* @versions: Array of hierarchy of versions to match.
* @count: Number of elements in the array.
*
* This is required only for the V2 bindings, and it enables a platform to
* specify the hierarchy of versions it supports. OPP layer will then enable
* OPPs, which are available for those versions, based on its 'opp-supported-hw'
* property.
*/
struct opp_table *dev_pm_opp_set_supported_hw(struct device *dev,
const u32 *versions, unsigned int count)
{
struct opp_table *opp_table;
opp_table = _add_opp_table(dev, false);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return opp_table;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
/* Another CPU that shares the OPP table has set the property ? */
if (opp_table->supported_hw)
return opp_table;
opp_table->supported_hw = kmemdup(versions, count * sizeof(*versions),
GFP_KERNEL);
if (!opp_table->supported_hw) {
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(-ENOMEM);
}
opp_table->supported_hw_count = count;
return opp_table;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_supported_hw);
/**
* dev_pm_opp_put_supported_hw() - Releases resources blocked for supported hw
* @opp_table: OPP table returned by dev_pm_opp_set_supported_hw().
*
* This is required only for the V2 bindings, and is called for a matching
* dev_pm_opp_set_supported_hw(). Until this is called, the opp_table structure
* will not be freed.
*/
void dev_pm_opp_put_supported_hw(struct opp_table *opp_table)
{
if (unlikely(!opp_table))
return;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
kfree(opp_table->supported_hw);
opp_table->supported_hw = NULL;
opp_table->supported_hw_count = 0;
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put_supported_hw);
static void devm_pm_opp_supported_hw_release(void *data)
{
dev_pm_opp_put_supported_hw(data);
}
/**
* devm_pm_opp_set_supported_hw() - Set supported platforms
* @dev: Device for which supported-hw has to be set.
* @versions: Array of hierarchy of versions to match.
* @count: Number of elements in the array.
*
* This is a resource-managed variant of dev_pm_opp_set_supported_hw().
*
* Return: 0 on success and errorno otherwise.
*/
int devm_pm_opp_set_supported_hw(struct device *dev, const u32 *versions,
unsigned int count)
{
struct opp_table *opp_table;
opp_table = dev_pm_opp_set_supported_hw(dev, versions, count);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
return devm_add_action_or_reset(dev, devm_pm_opp_supported_hw_release,
opp_table);
}
EXPORT_SYMBOL_GPL(devm_pm_opp_set_supported_hw);
/**
* dev_pm_opp_set_prop_name() - Set prop-extn name
* @dev: Device for which the prop-name has to be set.
* @name: name to postfix to properties.
*
* This is required only for the V2 bindings, and it enables a platform to
* specify the extn to be used for certain property names. The properties to
* which the extension will apply are opp-microvolt and opp-microamp. OPP core
* should postfix the property name with -<name> while looking for them.
*/
struct opp_table *dev_pm_opp_set_prop_name(struct device *dev, const char *name)
{
struct opp_table *opp_table;
opp_table = _add_opp_table(dev, false);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return opp_table;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
/* Another CPU that shares the OPP table has set the property ? */
if (opp_table->prop_name)
return opp_table;
opp_table->prop_name = kstrdup(name, GFP_KERNEL);
if (!opp_table->prop_name) {
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(-ENOMEM);
}
return opp_table;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_prop_name);
/**
* dev_pm_opp_put_prop_name() - Releases resources blocked for prop-name
* @opp_table: OPP table returned by dev_pm_opp_set_prop_name().
*
* This is required only for the V2 bindings, and is called for a matching
* dev_pm_opp_set_prop_name(). Until this is called, the opp_table structure
* will not be freed.
*/
void dev_pm_opp_put_prop_name(struct opp_table *opp_table)
{
if (unlikely(!opp_table))
return;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
kfree(opp_table->prop_name);
opp_table->prop_name = NULL;
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put_prop_name);
/**
* dev_pm_opp_set_regulators() - Set regulator names for the device
* @dev: Device for which regulator name is being set.
* @names: Array of pointers to the names of the regulator.
* @count: Number of regulators.
*
* In order to support OPP switching, OPP layer needs to know the name of the
* device's regulators, as the core would be required to switch voltages as
* well.
*
* This must be called before any OPPs are initialized for the device.
*/
struct opp_table *dev_pm_opp_set_regulators(struct device *dev,
const char * const names[],
unsigned int count)
{
struct dev_pm_opp_supply *supplies;
struct opp_table *opp_table;
struct regulator *reg;
int ret, i;
opp_table = _add_opp_table(dev, false);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return opp_table;
/* This should be called before OPPs are initialized */
if (WARN_ON(!list_empty(&opp_table->opp_list))) {
ret = -EBUSY;
goto err;
}
/* Another CPU that shares the OPP table has set the regulators ? */
if (opp_table->regulators)
return opp_table;
opp_table->regulators = kmalloc_array(count,
sizeof(*opp_table->regulators),
GFP_KERNEL);
if (!opp_table->regulators) {
ret = -ENOMEM;
goto err;
}
for (i = 0; i < count; i++) {
reg = regulator_get_optional(dev, names[i]);
if (IS_ERR(reg)) {
ret = PTR_ERR(reg);
if (ret != -EPROBE_DEFER)
dev_err(dev, "%s: no regulator (%s) found: %d\n",
__func__, names[i], ret);
goto free_regulators;
}
opp_table->regulators[i] = reg;
}
opp_table->regulator_count = count;
supplies = kmalloc_array(count * 2, sizeof(*supplies), GFP_KERNEL);
if (!supplies) {
ret = -ENOMEM;
goto free_regulators;
}
mutex_lock(&opp_table->lock);
opp_table->sod_supplies = supplies;
if (opp_table->set_opp_data) {
opp_table->set_opp_data->old_opp.supplies = supplies;
opp_table->set_opp_data->new_opp.supplies = supplies + count;
}
mutex_unlock(&opp_table->lock);
return opp_table;
free_regulators:
while (i != 0)
regulator_put(opp_table->regulators[--i]);
kfree(opp_table->regulators);
opp_table->regulators = NULL;
opp_table->regulator_count = -1;
err:
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(ret);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_regulators);
/**
* dev_pm_opp_put_regulators() - Releases resources blocked for regulator
* @opp_table: OPP table returned from dev_pm_opp_set_regulators().
*/
void dev_pm_opp_put_regulators(struct opp_table *opp_table)
{
int i;
if (unlikely(!opp_table))
return;
if (!opp_table->regulators)
goto put_opp_table;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
if (opp_table->enabled) {
for (i = opp_table->regulator_count - 1; i >= 0; i--)
regulator_disable(opp_table->regulators[i]);
}
for (i = opp_table->regulator_count - 1; i >= 0; i--)
regulator_put(opp_table->regulators[i]);
mutex_lock(&opp_table->lock);
if (opp_table->set_opp_data) {
opp_table->set_opp_data->old_opp.supplies = NULL;
opp_table->set_opp_data->new_opp.supplies = NULL;
}
kfree(opp_table->sod_supplies);
opp_table->sod_supplies = NULL;
mutex_unlock(&opp_table->lock);
kfree(opp_table->regulators);
opp_table->regulators = NULL;
opp_table->regulator_count = -1;
put_opp_table:
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put_regulators);
static void devm_pm_opp_regulators_release(void *data)
{
dev_pm_opp_put_regulators(data);
}
/**
* devm_pm_opp_set_regulators() - Set regulator names for the device
* @dev: Device for which regulator name is being set.
* @names: Array of pointers to the names of the regulator.
* @count: Number of regulators.
*
* This is a resource-managed variant of dev_pm_opp_set_regulators().
*
* Return: 0 on success and errorno otherwise.
*/
int devm_pm_opp_set_regulators(struct device *dev,
const char * const names[],
unsigned int count)
{
struct opp_table *opp_table;
opp_table = dev_pm_opp_set_regulators(dev, names, count);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
return devm_add_action_or_reset(dev, devm_pm_opp_regulators_release,
opp_table);
}
EXPORT_SYMBOL_GPL(devm_pm_opp_set_regulators);
/**
* dev_pm_opp_set_clkname() - Set clk name for the device
* @dev: Device for which clk name is being set.
* @name: Clk name.
*
* In order to support OPP switching, OPP layer needs to get pointer to the
* clock for the device. Simple cases work fine without using this routine (i.e.
* by passing connection-id as NULL), but for a device with multiple clocks
* available, the OPP core needs to know the exact name of the clk to use.
*
* This must be called before any OPPs are initialized for the device.
*/
struct opp_table *dev_pm_opp_set_clkname(struct device *dev, const char *name)
{
struct opp_table *opp_table;
int ret;
opp_table = _add_opp_table(dev, false);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return opp_table;
/* This should be called before OPPs are initialized */
if (WARN_ON(!list_empty(&opp_table->opp_list))) {
ret = -EBUSY;
goto err;
}
/* clk shouldn't be initialized at this point */
if (WARN_ON(opp_table->clk)) {
ret = -EBUSY;
goto err;
}
/* Find clk for the device */
opp_table->clk = clk_get(dev, name);
if (IS_ERR(opp_table->clk)) {
ret = PTR_ERR(opp_table->clk);
if (ret != -EPROBE_DEFER) {
dev_err(dev, "%s: Couldn't find clock: %d\n", __func__,
ret);
}
goto err;
}
return opp_table;
err:
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(ret);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_set_clkname);
/**
* dev_pm_opp_put_clkname() - Releases resources blocked for clk.
* @opp_table: OPP table returned from dev_pm_opp_set_clkname().
*/
void dev_pm_opp_put_clkname(struct opp_table *opp_table)
{
if (unlikely(!opp_table))
return;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
clk_put(opp_table->clk);
opp_table->clk = ERR_PTR(-EINVAL);
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_put_clkname);
static void devm_pm_opp_clkname_release(void *data)
{
dev_pm_opp_put_clkname(data);
}
/**
* devm_pm_opp_set_clkname() - Set clk name for the device
* @dev: Device for which clk name is being set.
* @name: Clk name.
*
* This is a resource-managed variant of dev_pm_opp_set_clkname().
*
* Return: 0 on success and errorno otherwise.
*/
int devm_pm_opp_set_clkname(struct device *dev, const char *name)
{
struct opp_table *opp_table;
opp_table = dev_pm_opp_set_clkname(dev, name);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
return devm_add_action_or_reset(dev, devm_pm_opp_clkname_release,
opp_table);
}
EXPORT_SYMBOL_GPL(devm_pm_opp_set_clkname);
/**
* dev_pm_opp_register_set_opp_helper() - Register custom set OPP helper
* @dev: Device for which the helper is getting registered.
* @set_opp: Custom set OPP helper.
*
* This is useful to support complex platforms (like platforms with multiple
* regulators per device), instead of the generic OPP set rate helper.
*
* This must be called before any OPPs are initialized for the device.
*/
struct opp_table *dev_pm_opp_register_set_opp_helper(struct device *dev,
int (*set_opp)(struct dev_pm_set_opp_data *data))
{
struct dev_pm_set_opp_data *data;
struct opp_table *opp_table;
if (!set_opp)
return ERR_PTR(-EINVAL);
opp_table = _add_opp_table(dev, false);
if (IS_ERR(opp_table))
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
return opp_table;
/* This should be called before OPPs are initialized */
if (WARN_ON(!list_empty(&opp_table->opp_list))) {
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(-EBUSY);
}
/* Another CPU that shares the OPP table has set the helper ? */
if (opp_table->set_opp)
return opp_table;
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (!data)
return ERR_PTR(-ENOMEM);
mutex_lock(&opp_table->lock);
opp_table->set_opp_data = data;
if (opp_table->sod_supplies) {
data->old_opp.supplies = opp_table->sod_supplies;
data->new_opp.supplies = opp_table->sod_supplies +
opp_table->regulator_count;
}
mutex_unlock(&opp_table->lock);
opp_table->set_opp = set_opp;
return opp_table;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_register_set_opp_helper);
/**
* dev_pm_opp_unregister_set_opp_helper() - Releases resources blocked for
* set_opp helper
* @opp_table: OPP table returned from dev_pm_opp_register_set_opp_helper().
*
* Release resources blocked for platform specific set_opp helper.
*/
void dev_pm_opp_unregister_set_opp_helper(struct opp_table *opp_table)
{
if (unlikely(!opp_table))
return;
/* Make sure there are no concurrent readers while updating opp_table */
WARN_ON(!list_empty(&opp_table->opp_list));
opp_table->set_opp = NULL;
mutex_lock(&opp_table->lock);
kfree(opp_table->set_opp_data);
opp_table->set_opp_data = NULL;
mutex_unlock(&opp_table->lock);
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_unregister_set_opp_helper);
static void devm_pm_opp_unregister_set_opp_helper(void *data)
{
dev_pm_opp_unregister_set_opp_helper(data);
}
/**
* devm_pm_opp_register_set_opp_helper() - Register custom set OPP helper
* @dev: Device for which the helper is getting registered.
* @set_opp: Custom set OPP helper.
*
* This is a resource-managed version of dev_pm_opp_register_set_opp_helper().
*
* Return: 0 on success and errorno otherwise.
*/
int devm_pm_opp_register_set_opp_helper(struct device *dev,
int (*set_opp)(struct dev_pm_set_opp_data *data))
{
struct opp_table *opp_table;
opp_table = dev_pm_opp_register_set_opp_helper(dev, set_opp);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
return devm_add_action_or_reset(dev, devm_pm_opp_unregister_set_opp_helper,
opp_table);
}
EXPORT_SYMBOL_GPL(devm_pm_opp_register_set_opp_helper);
static void _opp_detach_genpd(struct opp_table *opp_table)
{
int index;
if (!opp_table->genpd_virt_devs)
return;
for (index = 0; index < opp_table->required_opp_count; index++) {
if (!opp_table->genpd_virt_devs[index])
continue;
dev_pm_domain_detach(opp_table->genpd_virt_devs[index], false);
opp_table->genpd_virt_devs[index] = NULL;
}
kfree(opp_table->genpd_virt_devs);
opp_table->genpd_virt_devs = NULL;
}
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
/**
* dev_pm_opp_attach_genpd - Attach genpd(s) for the device and save virtual device pointer
* @dev: Consumer device for which the genpd is getting attached.
* @names: Null terminated array of pointers containing names of genpd to attach.
* @virt_devs: Pointer to return the array of virtual devices.
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*
* Multiple generic power domains for a device are supported with the help of
* virtual genpd 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 same API also works for the case where single genpd is available and so
* we don't need to support that separately.
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*
* This helper will normally be called by the consumer driver of the device
* "dev", as only that has details of the genpd names.
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*
* This helper needs to be called once with a list of all genpd to attach.
* Otherwise the original device structure will be used instead by the OPP core.
*
* The order of entries in the names array must match the order in which
* "required-opps" are added in DT.
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*/
struct opp_table *dev_pm_opp_attach_genpd(struct device *dev,
const char **names, struct device ***virt_devs)
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
{
struct opp_table *opp_table;
struct device *virt_dev;
int index = 0, ret = -EINVAL;
const char **name = names;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
opp_table = _add_opp_table(dev, false);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return opp_table;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
if (opp_table->genpd_virt_devs)
return opp_table;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
/*
* If the genpd's OPP table isn't already initialized, parsing of the
* required-opps fail for dev. We should retry this after genpd's OPP
* table is added.
*/
if (!opp_table->required_opp_count) {
ret = -EPROBE_DEFER;
goto put_table;
}
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
mutex_lock(&opp_table->genpd_virt_dev_lock);
opp_table->genpd_virt_devs = kcalloc(opp_table->required_opp_count,
sizeof(*opp_table->genpd_virt_devs),
GFP_KERNEL);
if (!opp_table->genpd_virt_devs)
goto unlock;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
while (*name) {
if (index >= opp_table->required_opp_count) {
dev_err(dev, "Index can't be greater than required-opp-count - 1, %s (%d : %d)\n",
*name, opp_table->required_opp_count, index);
goto err;
}
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
virt_dev = dev_pm_domain_attach_by_name(dev, *name);
if (IS_ERR(virt_dev)) {
ret = PTR_ERR(virt_dev);
dev_err(dev, "Couldn't attach to pm_domain: %d\n", ret);
goto err;
}
opp_table->genpd_virt_devs[index] = virt_dev;
index++;
name++;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
}
if (virt_devs)
*virt_devs = opp_table->genpd_virt_devs;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
mutex_unlock(&opp_table->genpd_virt_dev_lock);
return opp_table;
err:
_opp_detach_genpd(opp_table);
unlock:
mutex_unlock(&opp_table->genpd_virt_dev_lock);
put_table:
dev_pm_opp_put_opp_table(opp_table);
return ERR_PTR(ret);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_attach_genpd);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
/**
* dev_pm_opp_detach_genpd() - Detach genpd(s) from the device.
* @opp_table: OPP table returned by dev_pm_opp_attach_genpd().
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*
* This detaches the genpd(s), resets the virtual device pointers, and puts the
* OPP table.
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
*/
void dev_pm_opp_detach_genpd(struct opp_table *opp_table)
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
{
if (unlikely(!opp_table))
return;
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
/*
* Acquire genpd_virt_dev_lock to make sure virt_dev isn't getting
* used in parallel.
*/
mutex_lock(&opp_table->genpd_virt_dev_lock);
_opp_detach_genpd(opp_table);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
mutex_unlock(&opp_table->genpd_virt_dev_lock);
dev_pm_opp_put_opp_table(opp_table);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_detach_genpd);
OPP: Add dev_pm_opp_{set|put}_genpd_virt_dev() helper 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: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
2018-06-26 18:59:34 +08:00
static void devm_pm_opp_detach_genpd(void *data)
{
dev_pm_opp_detach_genpd(data);
}
/**
* devm_pm_opp_attach_genpd - Attach genpd(s) for the device and save virtual
* device pointer
* @dev: Consumer device for which the genpd is getting attached.
* @names: Null terminated array of pointers containing names of genpd to attach.
* @virt_devs: Pointer to return the array of virtual devices.
*
* This is a resource-managed version of dev_pm_opp_attach_genpd().
*
* Return: 0 on success and errorno otherwise.
*/
int devm_pm_opp_attach_genpd(struct device *dev, const char **names,
struct device ***virt_devs)
{
struct opp_table *opp_table;
opp_table = dev_pm_opp_attach_genpd(dev, names, virt_devs);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
return devm_add_action_or_reset(dev, devm_pm_opp_detach_genpd,
opp_table);
}
EXPORT_SYMBOL_GPL(devm_pm_opp_attach_genpd);
/**
* dev_pm_opp_xlate_required_opp() - Find required OPP for @src_table OPP.
* @src_table: OPP table which has @dst_table as one of its required OPP table.
* @dst_table: Required OPP table of the @src_table.
* @src_opp: OPP from the @src_table.
*
* This function returns the OPP (present in @dst_table) pointed out by the
* "required-opps" property of the @src_opp (present in @src_table).
*
* The callers are required to call dev_pm_opp_put() for the returned OPP after
* use.
*
* Return: pointer to 'struct dev_pm_opp' on success and errorno otherwise.
*/
struct dev_pm_opp *dev_pm_opp_xlate_required_opp(struct opp_table *src_table,
struct opp_table *dst_table,
struct dev_pm_opp *src_opp)
{
struct dev_pm_opp *opp, *dest_opp = ERR_PTR(-ENODEV);
int i;
if (!src_table || !dst_table || !src_opp ||
!src_table->required_opp_tables)
return ERR_PTR(-EINVAL);
/* required-opps not fully initialized yet */
if (lazy_linking_pending(src_table))
return ERR_PTR(-EBUSY);
for (i = 0; i < src_table->required_opp_count; i++) {
if (src_table->required_opp_tables[i] == dst_table) {
mutex_lock(&src_table->lock);
list_for_each_entry(opp, &src_table->opp_list, node) {
if (opp == src_opp) {
dest_opp = opp->required_opps[i];
dev_pm_opp_get(dest_opp);
break;
}
}
mutex_unlock(&src_table->lock);
break;
}
}
if (IS_ERR(dest_opp)) {
pr_err("%s: Couldn't find matching OPP (%p: %p)\n", __func__,
src_table, dst_table);
}
return dest_opp;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_xlate_required_opp);
/**
* dev_pm_opp_xlate_performance_state() - Find required OPP's pstate for src_table.
* @src_table: OPP table which has dst_table as one of its required OPP table.
* @dst_table: Required OPP table of the src_table.
* @pstate: Current performance state of the src_table.
*
* This Returns pstate of the OPP (present in @dst_table) pointed out by the
* "required-opps" property of the OPP (present in @src_table) which has
* performance state set to @pstate.
*
* Return: Zero or positive performance state on success, otherwise negative
* value on errors.
*/
int dev_pm_opp_xlate_performance_state(struct opp_table *src_table,
struct opp_table *dst_table,
unsigned int pstate)
{
struct dev_pm_opp *opp;
int dest_pstate = -EINVAL;
int i;
/*
* Normally the src_table will have the "required_opps" property set to
* point to one of the OPPs in the dst_table, but in some cases the
* genpd and its master have one to one mapping of performance states
* and so none of them have the "required-opps" property set. Return the
* pstate of the src_table as it is in such cases.
*/
if (!src_table || !src_table->required_opp_count)
return pstate;
/* required-opps not fully initialized yet */
if (lazy_linking_pending(src_table))
return -EBUSY;
for (i = 0; i < src_table->required_opp_count; i++) {
if (src_table->required_opp_tables[i]->np == dst_table->np)
break;
}
if (unlikely(i == src_table->required_opp_count)) {
pr_err("%s: Couldn't find matching OPP table (%p: %p)\n",
__func__, src_table, dst_table);
return -EINVAL;
}
mutex_lock(&src_table->lock);
list_for_each_entry(opp, &src_table->opp_list, node) {
if (opp->pstate == pstate) {
dest_pstate = opp->required_opps[i]->pstate;
goto unlock;
}
}
pr_err("%s: Couldn't find matching OPP (%p: %p)\n", __func__, src_table,
dst_table);
unlock:
mutex_unlock(&src_table->lock);
return dest_pstate;
}
/**
* dev_pm_opp_add() - Add an OPP table from a table definitions
* @dev: device for which we do this operation
* @freq: Frequency in Hz for this OPP
* @u_volt: Voltage in uVolts for this OPP
*
* This function adds an opp definition to the opp table and returns status.
* The opp is made available by default and it can be controlled using
* dev_pm_opp_enable/disable functions.
*
* Return:
* 0 On success OR
* Duplicate OPPs (both freq and volt are same) and opp->available
* -EEXIST Freq are same and volt are different OR
* Duplicate OPPs (both freq and volt are same) and !opp->available
* -ENOMEM Memory allocation failure
*/
int dev_pm_opp_add(struct device *dev, unsigned long freq, unsigned long u_volt)
{
struct opp_table *opp_table;
int ret;
opp_table = _add_opp_table(dev, true);
opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER 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: Stephan Gerhold <stephan@gerhold.net> Acked-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: Viresh Kumar <viresh.kumar@linaro.org>
2020-07-27 17:30:46 +08:00
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
/* Fix regulator count for dynamic OPPs */
opp_table->regulator_count = 1;
ret = _opp_add_v1(opp_table, dev, freq, u_volt, true);
if (ret)
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_add);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* _opp_set_availability() - helper to set the availability of an opp
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: OPP frequency to modify availability
* @availability_req: availability status requested for this opp
*
* Set the availability of an OPP, opp_{enable,disable} share a common logic
* which is isolated here.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Return: -EINVAL for bad pointers, -ENOMEM if no memory available for the
* copy operation, returns 0 if no modification was done OR modification was
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* successful.
*/
static int _opp_set_availability(struct device *dev, unsigned long freq,
bool availability_req)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
struct opp_table *opp_table;
struct dev_pm_opp *tmp_opp, *opp = ERR_PTR(-ENODEV);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
int r = 0;
/* Find the opp_table */
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
r = PTR_ERR(opp_table);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
dev_warn(dev, "%s: Device OPP not found (%d)\n", __func__, r);
return r;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
mutex_lock(&opp_table->lock);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/* Do we have the frequency? */
list_for_each_entry(tmp_opp, &opp_table->opp_list, node) {
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (tmp_opp->rate == freq) {
opp = tmp_opp;
break;
}
}
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
if (IS_ERR(opp)) {
r = PTR_ERR(opp);
goto unlock;
}
/* Is update really needed? */
if (opp->available == availability_req)
goto unlock;
opp->available = availability_req;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
dev_pm_opp_get(opp);
mutex_unlock(&opp_table->lock);
/* Notify the change of the OPP availability */
if (availability_req)
blocking_notifier_call_chain(&opp_table->head, OPP_EVENT_ENABLE,
opp);
else
blocking_notifier_call_chain(&opp_table->head,
OPP_EVENT_DISABLE, opp);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
dev_pm_opp_put(opp);
goto put_table;
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
unlock:
mutex_unlock(&opp_table->lock);
put_table:
dev_pm_opp_put_opp_table(opp_table);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
return r;
}
/**
* dev_pm_opp_adjust_voltage() - helper to change the voltage of an OPP
* @dev: device for which we do this operation
* @freq: OPP frequency to adjust voltage of
* @u_volt: new OPP target voltage
* @u_volt_min: new OPP min voltage
* @u_volt_max: new OPP max voltage
*
* Return: -EINVAL for bad pointers, -ENOMEM if no memory available for the
* copy operation, returns 0 if no modifcation was done OR modification was
* successful.
*/
int dev_pm_opp_adjust_voltage(struct device *dev, unsigned long freq,
unsigned long u_volt, unsigned long u_volt_min,
unsigned long u_volt_max)
{
struct opp_table *opp_table;
struct dev_pm_opp *tmp_opp, *opp = ERR_PTR(-ENODEV);
int r = 0;
/* Find the opp_table */
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
r = PTR_ERR(opp_table);
dev_warn(dev, "%s: Device OPP not found (%d)\n", __func__, r);
return r;
}
mutex_lock(&opp_table->lock);
/* Do we have the frequency? */
list_for_each_entry(tmp_opp, &opp_table->opp_list, node) {
if (tmp_opp->rate == freq) {
opp = tmp_opp;
break;
}
}
if (IS_ERR(opp)) {
r = PTR_ERR(opp);
goto adjust_unlock;
}
/* Is update really needed? */
if (opp->supplies->u_volt == u_volt)
goto adjust_unlock;
opp->supplies->u_volt = u_volt;
opp->supplies->u_volt_min = u_volt_min;
opp->supplies->u_volt_max = u_volt_max;
dev_pm_opp_get(opp);
mutex_unlock(&opp_table->lock);
/* Notify the voltage change of the OPP */
blocking_notifier_call_chain(&opp_table->head, OPP_EVENT_ADJUST_VOLTAGE,
opp);
dev_pm_opp_put(opp);
goto adjust_put_table;
adjust_unlock:
mutex_unlock(&opp_table->lock);
adjust_put_table:
dev_pm_opp_put_opp_table(opp_table);
return r;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_adjust_voltage);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_enable() - Enable a specific OPP
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: OPP frequency to enable
*
* Enables a provided opp. If the operation is valid, this returns 0, else the
* corresponding error value. It is meant to be used for users an OPP available
* after being temporarily made unavailable with dev_pm_opp_disable.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Return: -EINVAL for bad pointers, -ENOMEM if no memory available for the
* copy operation, returns 0 if no modification was done OR modification was
* successful.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
int dev_pm_opp_enable(struct device *dev, unsigned long freq)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
return _opp_set_availability(dev, freq, true);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_enable);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_disable() - Disable a specific OPP
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
* @dev: device for which we do this operation
* @freq: OPP frequency to disable
*
* Disables a provided opp. If the operation is valid, this returns
* 0, else the corresponding error value. It is meant to be a temporary
* control by users to make this OPP not available until the circumstances are
* right to make it available again (with a call to dev_pm_opp_enable).
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*
* Return: -EINVAL for bad pointers, -ENOMEM if no memory available for the
* copy operation, returns 0 if no modification was done OR modification was
* successful.
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
*/
int dev_pm_opp_disable(struct device *dev, unsigned long freq)
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
{
return _opp_set_availability(dev, freq, false);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
}
EXPORT_SYMBOL_GPL(dev_pm_opp_disable);
PM: Introduce library for device-specific OPPs (v7) SoCs have a standard set of tuples consisting of frequency and voltage pairs that the device will support per voltage domain. These are called Operating Performance Points or OPPs. The actual definitions of OPP varies over silicon versions. For a specific domain, we can have a set of {frequency, voltage} pairs. As the kernel boots and more information is available, a default set of these are activated based on the precise nature of device. Further on operation, based on conditions prevailing in the system (such as temperature), some OPP availability may be temporarily controlled by the SoC frameworks. To implement an OPP, some sort of power management support is necessary hence this library depends on CONFIG_PM. Contributions include: Sanjeev Premi for the initial concept: http://patchwork.kernel.org/patch/50998/ Kevin Hilman for converting original design to device-based. Kevin Hilman and Paul Walmsey for cleaning up many of the function abstractions, improvements and data structure handling. Romit Dasgupta for using enums instead of opp pointers. Thara Gopinath, Eduardo Valentin and Vishwanath BS for fixes and cleanups. Linus Walleij for recommending this layer be made generic for usage in other architectures beyond OMAP and ARM. Mark Brown, Andrew Morton, Rafael J. Wysocki, Paul E. McKenney for valuable improvements. Discussions and comments from: http://marc.info/?l=linux-omap&m=126033945313269&w=2 http://marc.info/?l=linux-omap&m=125482970102327&w=2 http://marc.info/?t=125809247500002&r=1&w=2 http://marc.info/?l=linux-omap&m=126025973426007&w=2 http://marc.info/?t=128152609200064&r=1&w=2 http://marc.info/?t=128468723000002&r=1&w=2 incorporated. v1: http://marc.info/?t=128468723000002&r=1&w=2 Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
2010-10-13 06:13:10 +08:00
/**
* dev_pm_opp_register_notifier() - Register OPP notifier for the device
* @dev: Device for which notifier needs to be registered
* @nb: Notifier block to be registered
*
* Return: 0 on success or a negative error value.
*/
int dev_pm_opp_register_notifier(struct device *dev, struct notifier_block *nb)
{
struct opp_table *opp_table;
int ret;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
ret = blocking_notifier_chain_register(&opp_table->head, nb);
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL(dev_pm_opp_register_notifier);
/**
* dev_pm_opp_unregister_notifier() - Unregister OPP notifier for the device
* @dev: Device for which notifier needs to be unregistered
* @nb: Notifier block to be unregistered
*
* Return: 0 on success or a negative error value.
*/
int dev_pm_opp_unregister_notifier(struct device *dev,
struct notifier_block *nb)
{
struct opp_table *opp_table;
int ret;
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return PTR_ERR(opp_table);
ret = blocking_notifier_chain_unregister(&opp_table->head, nb);
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL(dev_pm_opp_unregister_notifier);
/**
* dev_pm_opp_remove_table() - Free all OPPs associated with the device
* @dev: device pointer used to lookup OPP table.
*
* Free both OPPs created using static entries present in DT and the
* dynamically added entries.
*/
void dev_pm_opp_remove_table(struct device *dev)
{
struct opp_table *opp_table;
/* Check for existing table for 'dev' */
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table)) {
int error = PTR_ERR(opp_table);
if (error != -ENODEV)
WARN(1, "%s: opp_table: %d\n",
IS_ERR_OR_NULL(dev) ?
"Invalid device" : dev_name(dev),
error);
return;
}
/*
* Drop the extra reference only if the OPP table was successfully added
* with dev_pm_opp_of_add_table() earlier.
**/
if (_opp_remove_all_static(opp_table))
dev_pm_opp_put_opp_table(opp_table);
/* Drop reference taken by _find_opp_table() */
dev_pm_opp_put_opp_table(opp_table);
}
EXPORT_SYMBOL_GPL(dev_pm_opp_remove_table);
/**
* dev_pm_opp_sync_regulators() - Sync state of voltage regulators
* @dev: device for which we do this operation
*
* Sync voltage state of the OPP table regulators.
*
* Return: 0 on success or a negative error value.
*/
int dev_pm_opp_sync_regulators(struct device *dev)
{
struct opp_table *opp_table;
struct regulator *reg;
int i, ret = 0;
/* Device may not have OPP table */
opp_table = _find_opp_table(dev);
if (IS_ERR(opp_table))
return 0;
/* Regulator may not be required for the device */
if (unlikely(!opp_table->regulators))
goto put_table;
/* Nothing to sync if voltage wasn't changed */
if (!opp_table->enabled)
goto put_table;
for (i = 0; i < opp_table->regulator_count; i++) {
reg = opp_table->regulators[i];
ret = regulator_sync_voltage(reg);
if (ret)
break;
}
put_table:
/* Drop reference taken by _find_opp_table() */
dev_pm_opp_put_opp_table(opp_table);
return ret;
}
EXPORT_SYMBOL_GPL(dev_pm_opp_sync_regulators);