linux/drivers/thermal/thermal_of.c

623 lines
15 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* of-thermal.c - Generic Thermal Management device tree support.
*
* Copyright (C) 2013 Texas Instruments
* Copyright (C) 2013 Eduardo Valentin <eduardo.valentin@ti.com>
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/err.h>
#include <linux/export.h>
#include <linux/of_device.h>
#include <linux/of_platform.h>
#include <linux/slab.h>
#include <linux/thermal.h>
#include <linux/types.h>
#include <linux/string.h>
#include "thermal_core.h"
/*** functions parsing device tree nodes ***/
static int of_find_trip_id(struct device_node *np, struct device_node *trip)
{
struct device_node *trips;
struct device_node *t;
int i = 0;
trips = of_get_child_by_name(np, "trips");
if (!trips) {
pr_err("Failed to find 'trips' node\n");
return -EINVAL;
}
/*
* Find the trip id point associated with the cooling device map
*/
for_each_child_of_node(trips, t) {
if (t == trip)
goto out;
i++;
}
i = -ENXIO;
out:
of_node_put(trips);
return i;
}
/*
* It maps 'enum thermal_trip_type' found in include/linux/thermal.h
* into the device tree binding of 'trip', property type.
*/
static const char * const trip_types[] = {
[THERMAL_TRIP_ACTIVE] = "active",
[THERMAL_TRIP_PASSIVE] = "passive",
[THERMAL_TRIP_HOT] = "hot",
[THERMAL_TRIP_CRITICAL] = "critical",
};
/**
* thermal_of_get_trip_type - Get phy mode for given device_node
* @np: Pointer to the given device_node
* @type: Pointer to resulting trip type
*
* The function gets trip type string from property 'type',
* and store its index in trip_types table in @type,
*
* Return: 0 on success, or errno in error case.
*/
static int thermal_of_get_trip_type(struct device_node *np,
enum thermal_trip_type *type)
{
const char *t;
int err, i;
err = of_property_read_string(np, "type", &t);
if (err < 0)
return err;
for (i = 0; i < ARRAY_SIZE(trip_types); i++)
if (!strcasecmp(t, trip_types[i])) {
*type = i;
return 0;
}
return -ENODEV;
}
static int thermal_of_populate_trip(struct device_node *np,
struct thermal_trip *trip)
{
int prop;
int ret;
ret = of_property_read_u32(np, "temperature", &prop);
if (ret < 0) {
pr_err("missing temperature property\n");
return ret;
}
trip->temperature = prop;
ret = of_property_read_u32(np, "hysteresis", &prop);
if (ret < 0) {
pr_err("missing hysteresis property\n");
return ret;
}
trip->hysteresis = prop;
ret = thermal_of_get_trip_type(np, &trip->type);
if (ret < 0) {
pr_err("wrong trip type property\n");
return ret;
}
return 0;
}
static struct thermal_trip *thermal_of_trips_init(struct device_node *np, int *ntrips)
{
struct thermal_trip *tt;
struct device_node *trips, *trip;
int ret, count;
trips = of_get_child_by_name(np, "trips");
if (!trips) {
pr_err("Failed to find 'trips' node\n");
return ERR_PTR(-EINVAL);
}
count = of_get_child_count(trips);
if (!count) {
pr_err("No trip point defined\n");
ret = -EINVAL;
goto out_of_node_put;
}
tt = kzalloc(sizeof(*tt) * count, GFP_KERNEL);
if (!tt) {
ret = -ENOMEM;
goto out_of_node_put;
}
*ntrips = count;
count = 0;
for_each_child_of_node(trips, trip) {
ret = thermal_of_populate_trip(trip, &tt[count++]);
if (ret)
goto out_kfree;
}
of_node_put(trips);
return tt;
out_kfree:
kfree(tt);
*ntrips = 0;
out_of_node_put:
of_node_put(trips);
return ERR_PTR(ret);
}
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
static struct device_node *of_thermal_zone_find(struct device_node *sensor, int id)
{
struct device_node *np, *tz;
struct of_phandle_args sensor_specs;
np = of_find_node_by_name(NULL, "thermal-zones");
if (!np) {
pr_debug("No thermal zones description\n");
return ERR_PTR(-ENODEV);
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
}
/*
* Search for each thermal zone, a defined sensor
* corresponding to the one passed as parameter
*/
for_each_available_child_of_node(np, tz) {
int count, i;
count = of_count_phandle_with_args(tz, "thermal-sensors",
"#thermal-sensor-cells");
if (count <= 0) {
pr_err("%pOFn: missing thermal sensor\n", tz);
tz = ERR_PTR(-EINVAL);
goto out;
}
for (i = 0; i < count; i++) {
int ret;
ret = of_parse_phandle_with_args(tz, "thermal-sensors",
"#thermal-sensor-cells",
i, &sensor_specs);
if (ret < 0) {
pr_err("%pOFn: Failed to read thermal-sensors cells: %d\n", tz, ret);
tz = ERR_PTR(ret);
goto out;
}
if ((sensor == sensor_specs.np) && id == (sensor_specs.args_count ?
sensor_specs.args[0] : 0)) {
pr_debug("sensor %pOFn id=%d belongs to %pOFn\n", sensor, id, tz);
goto out;
}
}
}
tz = ERR_PTR(-ENODEV);
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
out:
of_node_put(np);
return tz;
}
static int thermal_of_monitor_init(struct device_node *np, int *delay, int *pdelay)
{
int ret;
ret = of_property_read_u32(np, "polling-delay-passive", pdelay);
if (ret < 0) {
pr_err("%pOFn: missing polling-delay-passive property\n", np);
return ret;
}
ret = of_property_read_u32(np, "polling-delay", delay);
if (ret < 0) {
pr_err("%pOFn: missing polling-delay property\n", np);
return ret;
}
return 0;
}
static struct thermal_zone_params *thermal_of_parameters_init(struct device_node *np)
{
struct thermal_zone_params *tzp;
int coef[2];
int ncoef = ARRAY_SIZE(coef);
int prop, ret;
tzp = kzalloc(sizeof(*tzp), GFP_KERNEL);
if (!tzp)
return ERR_PTR(-ENOMEM);
tzp->no_hwmon = true;
if (!of_property_read_u32(np, "sustainable-power", &prop))
tzp->sustainable_power = prop;
/*
* For now, the thermal framework supports only one sensor per
* thermal zone. Thus, we are considering only the first two
* values as slope and offset.
*/
ret = of_property_read_u32_array(np, "coefficients", coef, ncoef);
if (ret) {
coef[0] = 1;
coef[1] = 0;
}
tzp->slope = coef[0];
tzp->offset = coef[1];
return tzp;
}
static struct device_node *thermal_of_zone_get_by_name(struct thermal_zone_device *tz)
{
struct device_node *np, *tz_np;
np = of_find_node_by_name(NULL, "thermal-zones");
if (!np)
return ERR_PTR(-ENODEV);
tz_np = of_get_child_by_name(np, tz->type);
of_node_put(np);
if (!tz_np)
return ERR_PTR(-ENODEV);
return tz_np;
}
static int __thermal_of_unbind(struct device_node *map_np, int index, int trip_id,
struct thermal_zone_device *tz, struct thermal_cooling_device *cdev)
{
struct of_phandle_args cooling_spec;
int ret;
ret = of_parse_phandle_with_args(map_np, "cooling-device", "#cooling-cells",
index, &cooling_spec);
of_node_put(cooling_spec.np);
if (ret < 0) {
pr_err("Invalid cooling-device entry\n");
return ret;
}
if (cooling_spec.args_count < 2) {
pr_err("wrong reference to cooling device, missing limits\n");
return -EINVAL;
}
if (cooling_spec.np != cdev->np)
return 0;
ret = thermal_zone_unbind_cooling_device(tz, trip_id, cdev);
if (ret)
pr_err("Failed to unbind '%s' with '%s': %d\n", tz->type, cdev->type, ret);
return ret;
}
static int __thermal_of_bind(struct device_node *map_np, int index, int trip_id,
struct thermal_zone_device *tz, struct thermal_cooling_device *cdev)
{
struct of_phandle_args cooling_spec;
int ret, weight = THERMAL_WEIGHT_DEFAULT;
of_property_read_u32(map_np, "contribution", &weight);
ret = of_parse_phandle_with_args(map_np, "cooling-device", "#cooling-cells",
index, &cooling_spec);
of_node_put(cooling_spec.np);
if (ret < 0) {
pr_err("Invalid cooling-device entry\n");
return ret;
}
if (cooling_spec.args_count < 2) {
pr_err("wrong reference to cooling device, missing limits\n");
return -EINVAL;
}
if (cooling_spec.np != cdev->np)
return 0;
ret = thermal_zone_bind_cooling_device(tz, trip_id, cdev, cooling_spec.args[1],
cooling_spec.args[0],
weight);
if (ret)
pr_err("Failed to bind '%s' with '%s': %d\n", tz->type, cdev->type, ret);
return ret;
}
static int thermal_of_for_each_cooling_device(struct device_node *tz_np, struct device_node *map_np,
struct thermal_zone_device *tz, struct thermal_cooling_device *cdev,
int (*action)(struct device_node *, int, int,
struct thermal_zone_device *, struct thermal_cooling_device *))
{
struct device_node *tr_np;
int count, i, trip_id;
tr_np = of_parse_phandle(map_np, "trip", 0);
if (!tr_np)
return -ENODEV;
trip_id = of_find_trip_id(tz_np, tr_np);
if (trip_id < 0)
return trip_id;
count = of_count_phandle_with_args(map_np, "cooling-device", "#cooling-cells");
if (count <= 0) {
pr_err("Add a cooling_device property with at least one device\n");
return -ENOENT;
}
/*
* At this point, we don't want to bail out when there is an
* error, we will try to bind/unbind as many as possible
* cooling devices
*/
for (i = 0; i < count; i++)
action(map_np, i, trip_id, tz, cdev);
return 0;
}
static int thermal_of_for_each_cooling_maps(struct thermal_zone_device *tz,
struct thermal_cooling_device *cdev,
int (*action)(struct device_node *, int, int,
struct thermal_zone_device *, struct thermal_cooling_device *))
{
struct device_node *tz_np, *cm_np, *child;
int ret = 0;
tz_np = thermal_of_zone_get_by_name(tz);
if (IS_ERR(tz_np)) {
pr_err("Failed to get node tz by name\n");
return PTR_ERR(tz_np);
}
cm_np = of_get_child_by_name(tz_np, "cooling-maps");
if (!cm_np)
goto out;
for_each_child_of_node(cm_np, child) {
ret = thermal_of_for_each_cooling_device(tz_np, child, tz, cdev, action);
if (ret)
break;
}
of_node_put(cm_np);
out:
of_node_put(tz_np);
return ret;
}
static int thermal_of_bind(struct thermal_zone_device *tz,
struct thermal_cooling_device *cdev)
{
return thermal_of_for_each_cooling_maps(tz, cdev, __thermal_of_bind);
}
static int thermal_of_unbind(struct thermal_zone_device *tz,
struct thermal_cooling_device *cdev)
{
return thermal_of_for_each_cooling_maps(tz, cdev, __thermal_of_unbind);
}
/**
* thermal_of_zone_unregister - Cleanup the specific allocated ressources
*
* This function disables the thermal zone and frees the different
* ressources allocated specific to the thermal OF.
*
* @tz: a pointer to the thermal zone structure
*/
void thermal_of_zone_unregister(struct thermal_zone_device *tz)
{
thermal/of: Fix free after use in thermal_of_unregister() The thermal zone is freed after being unregistered. The release method devm_thermal_zone_device_register() calls -> thermal_of_zone_device_unregister() This one calls thermal_zone_device_unregister() which frees the thermal zone. However, thermal_of_zone_device_unregister() does access this freed pointer to free different resources allocated by the thermal_of framework which is invalid. It results in a kernel panic: [ 1.915140] thermal_sys: Failed to find thermal zone for tmu id=2 [ 1.921279] qoriq_thermal 1f80000.tmu: Failed to register sensors [ 1.927395] qoriq_thermal: probe of 1f80000.tmu failed with error -22 [ 1.934189] Unable to handle kernel paging request at virtual address 01adadadadadad88 [ 1.942146] Mem abort info: [ 1.944948] ESR = 0x0000000096000004 [ 1.948708] EC = 0x25: DABT (current EL), IL = 32 bits [ 1.954042] SET = 0, FnV = 0 [ 1.957107] EA = 0, S1PTW = 0 [ 1.960253] FSC = 0x04: level 0 translation fault [ 1.965147] Data abort info: [ 1.968030] ISV = 0, ISS = 0x00000004 [ 1.971878] CM = 0, WnR = 0 [ 1.974852] [01adadadadadad88] address between user and kernel address ranges [ 1.982016] Internal error: Oops: 96000004 [#1] SMP [ 1.986907] Modules linked in: [ 1.989969] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.19.0-next-20220808-00080-g1c46f44502e0 #1697 [ 1.999135] Hardware name: Kontron KBox A-230-LS (DT) [ 2.004199] pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 2.011185] pc : kfree+0x5c/0x3c0 [ 2.014516] lr : devm_thermal_of_zone_release+0x38/0x60 [ 2.019761] sp : ffff80000a22bad0 [ 2.023081] x29: ffff80000a22bad0 x28: 0000000000000000 x27: ffff800009960464 [ 2.030245] x26: ffff800009a16960 x25: 0000000000000006 x24: ffff800009f09a40 [ 2.037407] x23: ffff800009ab9008 x22: ffff800008d0eea8 x21: 01adadadadadad80 [ 2.044569] x20: 6b6b6b6b6b6b6b6b x19: ffff00200232b800 x18: 00000000fffffffb [ 2.051731] x17: ffff800008d0eea0 x16: ffff800008d07d44 x15: ffff800008d0d154 [ 2.056647] usb 1-1: new high-speed USB device number 2 using xhci-hcd [ 2.058893] x14: ffff800008d0cddc x13: ffff8000088d1c2c x12: ffff8000088d5034 [ 2.072597] x11: ffff8000088d46d4 x10: 0000000000000000 x9 : ffff800008d0eea8 [ 2.079759] x8 : ffff002000b1a158 x7 : bbbbbbbbbbbbbbbb x6 : ffff80000a0f53b8 [ 2.086921] x5 : ffff80000a22b960 x4 : 0000000000000000 x3 : 0000000000000000 [ 2.094082] x2 : fffffc0000000000 x1 : ffff002000838040 x0 : 01adb1adadadad80 [ 2.101244] Call trace: [ 2.103692] kfree+0x5c/0x3c0 [ 2.106666] devm_thermal_of_zone_release+0x38/0x60 [ 2.111561] release_nodes+0x64/0xd0 [ 2.115146] devres_release_all+0xbc/0x350 [ 2.119253] device_unbind_cleanup+0x20/0x70 [ 2.123536] really_probe+0x1a0/0x2e4 [ 2.127208] __driver_probe_device+0x80/0xec [ 2.131490] driver_probe_device+0x44/0x130 [ 2.135685] __driver_attach+0x104/0x1b4 [ 2.139619] bus_for_each_dev+0x7c/0xe0 [ 2.143465] driver_attach+0x30/0x40 [ 2.147048] bus_add_driver+0x160/0x210 [ 2.150894] driver_register+0x84/0x140 [ 2.154741] __platform_driver_register+0x34/0x40 [ 2.159461] qoriq_tmu_init+0x28/0x34 [ 2.163133] do_one_initcall+0x50/0x250 [ 2.166979] kernel_init_freeable+0x278/0x31c [ 2.171349] kernel_init+0x30/0x140 [ 2.174847] ret_from_fork+0x10/0x20 [ 2.178433] Code: b25657e2 d34cfc00 d37ae400 8b020015 (f94006a1) [ 2.184546] ---[ end trace 0000000000000000 ]--- Store the allocated resource pointers before the thermal zone is free and use them to release the resource after unregistering the thermal zone. Fixes: 3bd52ac87347 ("thermal/of: Rework the thermal device tree initialization") Reported-by: Michael Walle <michael@walle.cc> Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org> Tested-by: Michael Walle <michael@walle.cc> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20220809085629.509116-4-daniel.lezcano@linaro.org
2022-08-09 16:56:29 +08:00
struct thermal_trip *trips = tz->trips;
struct thermal_zone_params *tzp = tz->tzp;
struct thermal_zone_device_ops *ops = tz->ops;
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
thermal_zone_device_disable(tz);
thermal_zone_device_unregister(tz);
thermal/of: Fix free after use in thermal_of_unregister() The thermal zone is freed after being unregistered. The release method devm_thermal_zone_device_register() calls -> thermal_of_zone_device_unregister() This one calls thermal_zone_device_unregister() which frees the thermal zone. However, thermal_of_zone_device_unregister() does access this freed pointer to free different resources allocated by the thermal_of framework which is invalid. It results in a kernel panic: [ 1.915140] thermal_sys: Failed to find thermal zone for tmu id=2 [ 1.921279] qoriq_thermal 1f80000.tmu: Failed to register sensors [ 1.927395] qoriq_thermal: probe of 1f80000.tmu failed with error -22 [ 1.934189] Unable to handle kernel paging request at virtual address 01adadadadadad88 [ 1.942146] Mem abort info: [ 1.944948] ESR = 0x0000000096000004 [ 1.948708] EC = 0x25: DABT (current EL), IL = 32 bits [ 1.954042] SET = 0, FnV = 0 [ 1.957107] EA = 0, S1PTW = 0 [ 1.960253] FSC = 0x04: level 0 translation fault [ 1.965147] Data abort info: [ 1.968030] ISV = 0, ISS = 0x00000004 [ 1.971878] CM = 0, WnR = 0 [ 1.974852] [01adadadadadad88] address between user and kernel address ranges [ 1.982016] Internal error: Oops: 96000004 [#1] SMP [ 1.986907] Modules linked in: [ 1.989969] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.19.0-next-20220808-00080-g1c46f44502e0 #1697 [ 1.999135] Hardware name: Kontron KBox A-230-LS (DT) [ 2.004199] pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 2.011185] pc : kfree+0x5c/0x3c0 [ 2.014516] lr : devm_thermal_of_zone_release+0x38/0x60 [ 2.019761] sp : ffff80000a22bad0 [ 2.023081] x29: ffff80000a22bad0 x28: 0000000000000000 x27: ffff800009960464 [ 2.030245] x26: ffff800009a16960 x25: 0000000000000006 x24: ffff800009f09a40 [ 2.037407] x23: ffff800009ab9008 x22: ffff800008d0eea8 x21: 01adadadadadad80 [ 2.044569] x20: 6b6b6b6b6b6b6b6b x19: ffff00200232b800 x18: 00000000fffffffb [ 2.051731] x17: ffff800008d0eea0 x16: ffff800008d07d44 x15: ffff800008d0d154 [ 2.056647] usb 1-1: new high-speed USB device number 2 using xhci-hcd [ 2.058893] x14: ffff800008d0cddc x13: ffff8000088d1c2c x12: ffff8000088d5034 [ 2.072597] x11: ffff8000088d46d4 x10: 0000000000000000 x9 : ffff800008d0eea8 [ 2.079759] x8 : ffff002000b1a158 x7 : bbbbbbbbbbbbbbbb x6 : ffff80000a0f53b8 [ 2.086921] x5 : ffff80000a22b960 x4 : 0000000000000000 x3 : 0000000000000000 [ 2.094082] x2 : fffffc0000000000 x1 : ffff002000838040 x0 : 01adb1adadadad80 [ 2.101244] Call trace: [ 2.103692] kfree+0x5c/0x3c0 [ 2.106666] devm_thermal_of_zone_release+0x38/0x60 [ 2.111561] release_nodes+0x64/0xd0 [ 2.115146] devres_release_all+0xbc/0x350 [ 2.119253] device_unbind_cleanup+0x20/0x70 [ 2.123536] really_probe+0x1a0/0x2e4 [ 2.127208] __driver_probe_device+0x80/0xec [ 2.131490] driver_probe_device+0x44/0x130 [ 2.135685] __driver_attach+0x104/0x1b4 [ 2.139619] bus_for_each_dev+0x7c/0xe0 [ 2.143465] driver_attach+0x30/0x40 [ 2.147048] bus_add_driver+0x160/0x210 [ 2.150894] driver_register+0x84/0x140 [ 2.154741] __platform_driver_register+0x34/0x40 [ 2.159461] qoriq_tmu_init+0x28/0x34 [ 2.163133] do_one_initcall+0x50/0x250 [ 2.166979] kernel_init_freeable+0x278/0x31c [ 2.171349] kernel_init+0x30/0x140 [ 2.174847] ret_from_fork+0x10/0x20 [ 2.178433] Code: b25657e2 d34cfc00 d37ae400 8b020015 (f94006a1) [ 2.184546] ---[ end trace 0000000000000000 ]--- Store the allocated resource pointers before the thermal zone is free and use them to release the resource after unregistering the thermal zone. Fixes: 3bd52ac87347 ("thermal/of: Rework the thermal device tree initialization") Reported-by: Michael Walle <michael@walle.cc> Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org> Tested-by: Michael Walle <michael@walle.cc> Reviewed-by: Guenter Roeck <linux@roeck-us.net> Link: https://lore.kernel.org/r/20220809085629.509116-4-daniel.lezcano@linaro.org
2022-08-09 16:56:29 +08:00
kfree(trips);
kfree(tzp);
kfree(ops);
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
}
EXPORT_SYMBOL_GPL(thermal_of_zone_unregister);
/**
* thermal_of_zone_register - Register a thermal zone with device node
* sensor
*
* The thermal_of_zone_register() parses a device tree given a device
* node sensor and identifier. It searches for the thermal zone
* associated to the couple sensor/id and retrieves all the thermal
* zone properties and registers new thermal zone with those
* properties.
*
* @sensor: A device node pointer corresponding to the sensor in the device tree
* @id: An integer as sensor identifier
* @data: A private data to be stored in the thermal zone dedicated private area
* @ops: A set of thermal sensor ops
*
* Return: a valid thermal zone structure pointer on success.
* - EINVAL: if the device tree thermal description is malformed
* - ENOMEM: if one structure can not be allocated
* - Other negative errors are returned by the underlying called functions
*/
struct thermal_zone_device *thermal_of_zone_register(struct device_node *sensor, int id, void *data,
const struct thermal_zone_device_ops *ops)
{
struct thermal_zone_device *tz;
struct thermal_trip *trips;
struct thermal_zone_params *tzp;
struct thermal_zone_device_ops *of_ops;
struct device_node *np;
int delay, pdelay;
int ntrips, mask;
int ret;
of_ops = kmemdup(ops, sizeof(*ops), GFP_KERNEL);
if (!of_ops)
return ERR_PTR(-ENOMEM);
np = of_thermal_zone_find(sensor, id);
if (IS_ERR(np)) {
if (PTR_ERR(np) != -ENODEV)
pr_err("Failed to find thermal zone for %pOFn id=%d\n", sensor, id);
thermal/of: Fix memory leak on thermal_of_zone_register() failure The function does not free 'of_ops' upon failure, leading to a memory leak [1]. Fix by freeing 'of_ops' in the error path. [1] unreferenced object 0xffff8ee846198c80 (size 128): comm "swapper/0", pid 1, jiffies 4294699704 (age 70.076s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ d0 3f 6e 8c ff ff ff ff 00 00 00 00 00 00 00 00 .?n............. backtrace: [<00000000d136f562>] __kmalloc_node_track_caller+0x42/0x120 [<0000000063f31678>] kmemdup+0x1d/0x40 [<00000000e6d24096>] thermal_of_zone_register+0x49/0x520 [<000000005e78c755>] devm_thermal_of_zone_register+0x54/0x90 [<00000000ee6b209e>] pmbus_add_sensor+0x1b4/0x1d0 [<00000000896105e3>] pmbus_add_sensor_attrs_one+0x123/0x440 [<0000000049e990a6>] pmbus_add_sensor_attrs+0xfe/0x1d0 [<00000000466b5440>] pmbus_do_probe+0x66b/0x14e0 [<0000000084d42285>] i2c_device_probe+0x13b/0x2f0 [<0000000029e2ae74>] really_probe+0xce/0x2c0 [<00000000692df15c>] driver_probe_device+0x19/0xd0 [<00000000547d9cce>] __device_attach_driver+0x6f/0x100 [<0000000020abd24b>] bus_for_each_drv+0x76/0xc0 [<00000000665d9563>] __device_attach+0xfc/0x180 [<000000008ddd4d6a>] bus_probe_device+0x82/0xa0 [<000000009e61132b>] device_add+0x3fe/0x920 Fixes: 3fd6d6e2b4e8 ("thermal/of: Rework the thermal device tree initialization") Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Vadim Pasternak <vadimp@nvidia.com> Link: https://lore.kernel.org/r/20221020103658.802457-1-idosch@nvidia.com Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-10-20 18:36:58 +08:00
ret = PTR_ERR(np);
goto out_kfree_of_ops;
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
}
trips = thermal_of_trips_init(np, &ntrips);
if (IS_ERR(trips)) {
pr_err("Failed to find trip points for %pOFn id=%d\n", sensor, id);
thermal/of: Fix memory leak on thermal_of_zone_register() failure The function does not free 'of_ops' upon failure, leading to a memory leak [1]. Fix by freeing 'of_ops' in the error path. [1] unreferenced object 0xffff8ee846198c80 (size 128): comm "swapper/0", pid 1, jiffies 4294699704 (age 70.076s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ d0 3f 6e 8c ff ff ff ff 00 00 00 00 00 00 00 00 .?n............. backtrace: [<00000000d136f562>] __kmalloc_node_track_caller+0x42/0x120 [<0000000063f31678>] kmemdup+0x1d/0x40 [<00000000e6d24096>] thermal_of_zone_register+0x49/0x520 [<000000005e78c755>] devm_thermal_of_zone_register+0x54/0x90 [<00000000ee6b209e>] pmbus_add_sensor+0x1b4/0x1d0 [<00000000896105e3>] pmbus_add_sensor_attrs_one+0x123/0x440 [<0000000049e990a6>] pmbus_add_sensor_attrs+0xfe/0x1d0 [<00000000466b5440>] pmbus_do_probe+0x66b/0x14e0 [<0000000084d42285>] i2c_device_probe+0x13b/0x2f0 [<0000000029e2ae74>] really_probe+0xce/0x2c0 [<00000000692df15c>] driver_probe_device+0x19/0xd0 [<00000000547d9cce>] __device_attach_driver+0x6f/0x100 [<0000000020abd24b>] bus_for_each_drv+0x76/0xc0 [<00000000665d9563>] __device_attach+0xfc/0x180 [<000000008ddd4d6a>] bus_probe_device+0x82/0xa0 [<000000009e61132b>] device_add+0x3fe/0x920 Fixes: 3fd6d6e2b4e8 ("thermal/of: Rework the thermal device tree initialization") Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Vadim Pasternak <vadimp@nvidia.com> Link: https://lore.kernel.org/r/20221020103658.802457-1-idosch@nvidia.com Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-10-20 18:36:58 +08:00
ret = PTR_ERR(trips);
goto out_kfree_of_ops;
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
}
ret = thermal_of_monitor_init(np, &delay, &pdelay);
if (ret) {
pr_err("Failed to initialize monitoring delays from %pOFn\n", np);
goto out_kfree_trips;
}
tzp = thermal_of_parameters_init(np);
if (IS_ERR(tzp)) {
ret = PTR_ERR(tzp);
pr_err("Failed to initialize parameter from %pOFn: %d\n", np, ret);
goto out_kfree_trips;
}
of_ops->bind = thermal_of_bind;
of_ops->unbind = thermal_of_unbind;
mask = GENMASK_ULL((ntrips) - 1, 0);
tz = thermal_zone_device_register_with_trips(np->name, trips, ntrips,
mask, data, of_ops, tzp,
pdelay, delay);
if (IS_ERR(tz)) {
ret = PTR_ERR(tz);
pr_err("Failed to register thermal zone %pOFn: %d\n", np, ret);
goto out_kfree_tzp;
}
ret = thermal_zone_device_enable(tz);
if (ret) {
pr_err("Failed to enabled thermal zone '%s', id=%d: %d\n",
tz->type, tz->id, ret);
thermal_of_zone_unregister(tz);
return ERR_PTR(ret);
}
return tz;
out_kfree_tzp:
kfree(tzp);
out_kfree_trips:
kfree(trips);
thermal/of: Fix memory leak on thermal_of_zone_register() failure The function does not free 'of_ops' upon failure, leading to a memory leak [1]. Fix by freeing 'of_ops' in the error path. [1] unreferenced object 0xffff8ee846198c80 (size 128): comm "swapper/0", pid 1, jiffies 4294699704 (age 70.076s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ d0 3f 6e 8c ff ff ff ff 00 00 00 00 00 00 00 00 .?n............. backtrace: [<00000000d136f562>] __kmalloc_node_track_caller+0x42/0x120 [<0000000063f31678>] kmemdup+0x1d/0x40 [<00000000e6d24096>] thermal_of_zone_register+0x49/0x520 [<000000005e78c755>] devm_thermal_of_zone_register+0x54/0x90 [<00000000ee6b209e>] pmbus_add_sensor+0x1b4/0x1d0 [<00000000896105e3>] pmbus_add_sensor_attrs_one+0x123/0x440 [<0000000049e990a6>] pmbus_add_sensor_attrs+0xfe/0x1d0 [<00000000466b5440>] pmbus_do_probe+0x66b/0x14e0 [<0000000084d42285>] i2c_device_probe+0x13b/0x2f0 [<0000000029e2ae74>] really_probe+0xce/0x2c0 [<00000000692df15c>] driver_probe_device+0x19/0xd0 [<00000000547d9cce>] __device_attach_driver+0x6f/0x100 [<0000000020abd24b>] bus_for_each_drv+0x76/0xc0 [<00000000665d9563>] __device_attach+0xfc/0x180 [<000000008ddd4d6a>] bus_probe_device+0x82/0xa0 [<000000009e61132b>] device_add+0x3fe/0x920 Fixes: 3fd6d6e2b4e8 ("thermal/of: Rework the thermal device tree initialization") Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Vadim Pasternak <vadimp@nvidia.com> Link: https://lore.kernel.org/r/20221020103658.802457-1-idosch@nvidia.com Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-10-20 18:36:58 +08:00
out_kfree_of_ops:
kfree(of_ops);
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
return ERR_PTR(ret);
}
EXPORT_SYMBOL_GPL(thermal_of_zone_register);
static void devm_thermal_of_zone_release(struct device *dev, void *res)
{
thermal_of_zone_unregister(*(struct thermal_zone_device **)res);
}
static int devm_thermal_of_zone_match(struct device *dev, void *res,
void *data)
{
struct thermal_zone_device **r = res;
if (WARN_ON(!r || !*r))
return 0;
return *r == data;
}
/**
* devm_thermal_of_zone_register - register a thermal tied with the sensor life cycle
*
* This function is the device version of the thermal_of_zone_register() function.
*
* @dev: a device structure pointer to sensor to be tied with the thermal zone OF life cycle
* @sensor_id: the sensor identifier
* @data: a pointer to a private data to be stored in the thermal zone 'devdata' field
* @ops: a pointer to the ops structure associated with the sensor
*/
struct thermal_zone_device *devm_thermal_of_zone_register(struct device *dev, int sensor_id, void *data,
const struct thermal_zone_device_ops *ops)
{
struct thermal_zone_device **ptr, *tzd;
ptr = devres_alloc(devm_thermal_of_zone_release, sizeof(*ptr),
GFP_KERNEL);
if (!ptr)
return ERR_PTR(-ENOMEM);
tzd = thermal_of_zone_register(dev->of_node, sensor_id, data, ops);
if (IS_ERR(tzd)) {
devres_free(ptr);
return tzd;
}
*ptr = tzd;
devres_add(dev, ptr);
return tzd;
}
EXPORT_SYMBOL_GPL(devm_thermal_of_zone_register);
/**
* devm_thermal_of_zone_unregister - Resource managed version of
* thermal_of_zone_unregister().
* @dev: Device for which which resource was allocated.
* @tz: a pointer to struct thermal_zone where the sensor is registered.
*
* This function removes the sensor callbacks and private data from the
* thermal zone device registered with devm_thermal_zone_of_sensor_register()
* API. It will also silent the zone by remove the .get_temp() and .get_trend()
* thermal zone device callbacks.
* Normally this function will not need to be called and the resource
* management code will ensure that the resource is freed.
*/
void devm_thermal_of_zone_unregister(struct device *dev, struct thermal_zone_device *tz)
{
WARN_ON(devres_release(dev, devm_thermal_of_zone_release,
thermal/of: Rework the thermal device tree initialization The following changes are reworking entirely the thermal device tree initialization. The old version is kept until the different drivers using it are converted to the new API. The old approach creates the different actors independently. This approach is the source of the code duplication in the thermal OF because a thermal zone is created but a sensor is registered after. The thermal zones are created unconditionnaly with a fake sensor at init time, thus forcing to provide fake ops and store all the thermal zone related information in duplicated structures. Then the sensor is initialized and the code looks up the thermal zone name using the device tree. Then the sensor is associated to the thermal zone, and the sensor specific ops are called with a second level of indirection from the thermal zone ops. When a sensor is removed (with a module unload), the thermal zone stays there with the fake sensor. The cooling device associated with a thermal zone and a trip point is stored in a list, again duplicating information, using the node name of the device tree to match afterwards the cooling devices. The new approach is simpler, it creates a thermal zone when the sensor is registered and destroys it when the sensor is removed. All the matching between the cooling device, trip points and thermal zones are done using the device tree, as well as bindings. The ops are no longer specific but uses the generic ones provided by the thermal framework. When the old code won't have any users, it can be removed and the remaining thermal OF code will be much simpler. Signed-off-by: Daniel Lezcano <daniel.lezcano@linexp.org> Link: https://lore.kernel.org/r/20220804224349.1926752-2-daniel.lezcano@linexp.org Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
2022-08-05 06:43:17 +08:00
devm_thermal_of_zone_match, tz));
}
EXPORT_SYMBOL_GPL(devm_thermal_of_zone_unregister);