From patchwork Wed Dec 27 10:26:42 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Daniel Lezcano X-Patchwork-Id: 13505285 Received: from mail-wm1-f52.google.com (mail-wm1-f52.google.com [209.85.128.52]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id E36132C867 for ; Wed, 27 Dec 2023 10:26:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linaro.org Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linaro.org Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=linaro.org header.i=@linaro.org header.b="tS+rQPOz" Received: by mail-wm1-f52.google.com with SMTP id 5b1f17b1804b1-40d5aef534eso10058155e9.0 for ; Wed, 27 Dec 2023 02:26:54 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1703672813; x=1704277613; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:from:to:cc:subject:date:message-id:reply-to; bh=8OcDR+Nega8FhnoPMRAPEOetFIgTrBFAnqGKFJ3ZWsQ=; b=tS+rQPOzfZP2OeRlqTZX5DoLH0HvEmUqdzGnC1NCg+5hgO6nncfw7CFMG/fyu2tHah +MJ6g87TRvjTjBZRkb5He3i3V1MbguNV6E+Ud51N1Wi3d7nqTRVF9OmO9Y690IEB4i/u vDpIh8T5tv8KiKdBU7rT/6BJmLwVFHkzDfnPx4kgmxRI5tefkgeOQOhJetr1Kgtn32wg 8jwtiu2/Zcf5A+IOrYHOs4l7dr42Vv6d5az+2/43RqwH+4WuPW3NWVlglckcP8ABmzcA 4P0rVnNlqmLta0xtuNcuNic+twg4LdaNgV/ASHVhr/rN5b9P4HlNpVvwU4LWSe4PBCqZ AntQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1703672813; x=1704277613; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=8OcDR+Nega8FhnoPMRAPEOetFIgTrBFAnqGKFJ3ZWsQ=; b=sg+JFB2tbwWN1KVaqk6mQmgta/Jeaf5L3MuE+1fdrgV2dPAnmtyVtJUhlrm943rAuB qYbEO8QGAk0va3oiQRfXhiCkB/WRZCEvao6VVHmidEBtkU68OEwBL+GGH9F0O4oX5E7U 00mXakc8QPdE31ZrWX1vH8pr5AY9DPyPWxISJcreHgndyjj5E+7a4xQ5aVCdSc+fmDcM +mxfv8uZ8b/N8xIZatFKFM0F1xm4hWZI+Ib8ihx5+hdDtmFDaoFtcpvUCK+6ZQu/nhdQ 9T+/McDrU6smsqhvhbHTpsutDX7A89S86mh0EkiGwT5Pa8pKaAl1bJ/+lhc3F9Axo0ya V+RQ== X-Gm-Message-State: AOJu0YzfvrBYqWs3A/hoftU1f+dnXiTG7n91z5vTWaQyhOs1D3UrNQZc X9qeOrUK0dHxDQWLlVgbtRUqZz1ZSRwrFQ== X-Google-Smtp-Source: AGHT+IF+ENqN1RkOAx7IyRIky119QRHlqQts9x8RpJWJcmVm8z1fcSTSAYWbv6fF5jnmALSVE/zekg== X-Received: by 2002:a05:600c:1503:b0:40d:355e:ab92 with SMTP id b3-20020a05600c150300b0040d355eab92mr4671008wmg.22.1703672812971; Wed, 27 Dec 2023 02:26:52 -0800 (PST) Received: from mai.. (146725694.box.freepro.com. [130.180.211.218]) by smtp.gmail.com with ESMTPSA id y17-20020adff151000000b0033664ffaf5dsm14556763wro.37.2023.12.27.02.26.52 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 27 Dec 2023 02:26:52 -0800 (PST) From: Daniel Lezcano To: daniel.lezcano@linaro.org, rjw@rjwysocki.net, lukasz.luba@arm.com Cc: rui.zhang@intel.com, linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Subject: [PATCH v4 1/2] thermal/debugfs: Add thermal cooling device debugfs information Date: Wed, 27 Dec 2023 11:26:42 +0100 Message-Id: <20231227102643.2655651-1-daniel.lezcano@linaro.org> X-Mailer: git-send-email 2.34.1 Precedence: bulk X-Mailing-List: linux-pm@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 The thermal framework does not have any debug information except a sysfs stat which is a bit controversial. This one allocates big chunks of memory for every cooling devices with a high number of states and could represent on some systems in production several megabytes of memory for just a portion of it. As the sysfs is limited to a page size, the output is not exploitable with large data array and gets truncated. The patch provides the same information than sysfs except the transitions are dynamically allocated, thus they won't show more events than the ones which actually occurred. There is no longer a size limitation and it opens the field for more debugging information where the debugfs is designed for, not sysfs. The thermal debugfs directory structure tries to stay consistent with the sysfs one but in a very simplified way: thermal/ -- cooling_devices |-- 0 | |-- clear | |-- time_in_state_ms | |-- total_trans | `-- trans_table |-- 1 | |-- clear | |-- time_in_state_ms | |-- total_trans | `-- trans_table |-- 2 | |-- clear | |-- time_in_state_ms | |-- total_trans | `-- trans_table |-- 3 | |-- clear | |-- time_in_state_ms | |-- total_trans | `-- trans_table `-- 4 |-- clear |-- time_in_state_ms |-- total_trans `-- trans_table The content of the files in the cooling devices directory is the same as the sysfs one except for the trans_table which has the following format: Transition Hits 1->0 246 0->1 246 2->1 632 1->2 632 3->2 98 2->3 98 Signed-off-by: Daniel Lezcano --- Changelog: - v4 - Fixed kerneldoc description ordering (Rafael) - Fixed comment (Rafael) - Renamed s/cdev_value/cdev_record/ (Rafael) - Used union instead of a common 'value' field in cdev_record (Rafael) - Renamed s/cdev/cdev_dbg/ for clarity (Rafael) - Renamed s/dfs/thermal_dbg/ for clarity (Rafael) - Renamed s/list/lists/ in the place where there are array of lists (Rafael) - Inverted initialization logic when allocating a cdev_record (Rafael) - Moved now = ktime_get() closer to the place where it is used (Rafael) - Moved two lines down to a condition (Rafael) - Removed strange and unexpected addition of function (Rafael) - v3 - Fixed kerneldoc description (kbuild) - Made some functions static - v2 - Added parameter names to fix kbuild report - Renamed 'reset' to 'clear' to avoid confusion (Rafael) - Fixed several typos and rephrased some sentences (Rafael) - Renamed structure field name s/list/node/ (Rafael) - Documented structures and exported functions (Rafael) - s/trans_list/transitions/ (Rafael) - s/duration_list/durations/ (Rafael) - Folded 'alloc' and 'insert' into a single function (Rafael) - s/list/lists/ as it is an array of lists (Rafael) - s/pos/entry/ (Rafael) - Introduced a locking in the 'clear' callback function (Rafael) - s/to/new_state/ and s/from/old_state/ (Rafael) - Added some comments in thermal_debug_cdev_transition() (Rafael) - Explained why char[11] (Rafael) - s/Hits/Occurrences/ (Rafael) - s/Time/Residency/ (Rafael) - Constified structure pointer passed to thermal_debug_cdev_transition() - s/thermal_debug_cdev_transition()/thermal_debug_cdev_state_update()/ - v1 (from RFC): - Fixed typo "occurred" - Changed Kconfig option name and description - Removed comment in the Makefile - Renamed exported function name s/debugfs/debug/ - Replaced thermal_debug_cdev_[unregister|register] by [add|remove] --- drivers/thermal/Kconfig | 7 + drivers/thermal/Makefile | 2 + drivers/thermal/thermal_core.c | 6 + drivers/thermal/thermal_core.h | 1 + drivers/thermal/thermal_debugfs.c | 445 ++++++++++++++++++++++++++++++ drivers/thermal/thermal_debugfs.h | 14 + drivers/thermal/thermal_helpers.c | 27 +- include/linux/thermal.h | 7 + 8 files changed, 503 insertions(+), 6 deletions(-) create mode 100644 drivers/thermal/thermal_debugfs.c create mode 100644 drivers/thermal/thermal_debugfs.h diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig index c81a00fbca7d..3ff7add3fb7c 100644 --- a/drivers/thermal/Kconfig +++ b/drivers/thermal/Kconfig @@ -33,6 +33,13 @@ config THERMAL_STATISTICS If in doubt, say N. +config THERMAL_DEBUGFS + bool "Thermal subsystem debug support" + depends on DEBUG_FS + help + Say Y to allow the thermal subsystem to collect diagnostic + information that can be accessed via debugfs. + config THERMAL_EMERGENCY_POWEROFF_DELAY_MS int "Emergency poweroff delay in milli-seconds" default 0 diff --git a/drivers/thermal/Makefile b/drivers/thermal/Makefile index c934cab309ae..0f65ae86a9c6 100644 --- a/drivers/thermal/Makefile +++ b/drivers/thermal/Makefile @@ -10,6 +10,8 @@ thermal_sys-y += thermal_trip.o thermal_helpers.o # netlink interface to manage the thermal framework thermal_sys-$(CONFIG_THERMAL_NETLINK) += thermal_netlink.o +thermal_sys-$(CONFIG_THERMAL_DEBUGFS) += thermal_debugfs.o + # interface to/from other layers providing sensors thermal_sys-$(CONFIG_THERMAL_HWMON) += thermal_hwmon.o thermal_sys-$(CONFIG_THERMAL_OF) += thermal_of.o diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c index 625ba07cbe2f..33332d401b13 100644 --- a/drivers/thermal/thermal_core.c +++ b/drivers/thermal/thermal_core.c @@ -945,6 +945,8 @@ __thermal_cooling_device_register(struct device_node *np, mutex_unlock(&thermal_list_lock); + thermal_debug_cdev_add(cdev); + return cdev; out_cooling_dev: @@ -1151,6 +1153,8 @@ void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev) if (!cdev) return; + thermal_debug_cdev_remove(cdev); + mutex_lock(&thermal_list_lock); if (!thermal_cooling_device_present(cdev)) { @@ -1570,6 +1574,8 @@ static int __init thermal_init(void) { int result; + thermal_debug_init(); + result = thermal_netlink_init(); if (result) goto error; diff --git a/drivers/thermal/thermal_core.h b/drivers/thermal/thermal_core.h index 7dfe6c8deb8e..4b3452a65a2f 100644 --- a/drivers/thermal/thermal_core.h +++ b/drivers/thermal/thermal_core.h @@ -13,6 +13,7 @@ #include #include "thermal_netlink.h" +#include "thermal_debugfs.h" /* Default Thermal Governor */ #if defined(CONFIG_THERMAL_DEFAULT_GOV_STEP_WISE) diff --git a/drivers/thermal/thermal_debugfs.c b/drivers/thermal/thermal_debugfs.c new file mode 100644 index 000000000000..84e5ef3dc16c --- /dev/null +++ b/drivers/thermal/thermal_debugfs.c @@ -0,0 +1,445 @@ + +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright 2023 Linaro Limited + * + * Author: Daniel Lezcano + * + * Thermal subsystem debug support + */ +#include +#include +#include +#include +#include +#include + +static struct dentry *d_root; +static struct dentry *d_cdev; + +/* + * Length of the string containing the thermal zone id or the cooling + * device id, including the ending nul character. We can reasonably + * assume there won't be more than 256 thermal zones as the maximum + * observed today is around 32. + */ +#define IDSLENGTH 4 + +/* + * The cooling device transition list is stored in a hash table where + * the size is CDEVSTATS_HASH_SIZE. The majority of cooling devices + * have dozen of states but some can have much more, so a hash table + * is more adequate in this case, because the cost of browsing the entire + * list when storing the transitions may not be negligible. + */ +#define CDEVSTATS_HASH_SIZE 16 + +/** + * struct cdev_debugfs - per cooling device statistics structure + * A cooling device can have a high number of states. Showing the + * transitions on a matrix based representation can be overkill given + * most of the transitions won't happen and we end up with a matrix + * filled with zero. Instead, we show the transitions which actually + * happened. + * + * Every transition updates the current_state and the timestamp. The + * transitions and the durations are stored in lists. + * + * @total: the number of transitions for this cooling device + * @current_state: the current cooling device state + * @timestamp: the state change timestamp + * @transitions: an array of lists containing the state transitions + * @durations: an array of lists containing the residencies of each state + */ +struct cdev_debugfs { + u32 total; + int current_state; + ktime_t timestamp; + struct list_head transitions[CDEVSTATS_HASH_SIZE]; + struct list_head durations[CDEVSTATS_HASH_SIZE]; +}; + +/** + * struct cdev_value - Common structure for cooling device entry + * + * The following common structure allows to store the information + * related to the transitions and to the state residencies. They are + * identified with a id which is associated to a value. It is used as + * nodes for the "transitions" and "durations" above. + * + * @node: node to insert the structure in a list + * @id: identifier of the value which can be a state or a transition + * @residency: a ktime_t representing a state residency duration + * @count: a number of occurrences + */ +struct cdev_record { + struct list_head node; + int id; + union { + ktime_t residency; + u64 count; + }; +}; + +/** + * struct thermal_debugfs - High level structure for a thermal object in debugfs + * + * The thermal_debugfs structure is the common structure used by the + * cooling device to compute the statistics. + * + * @d_top: top directory of the thermal object directory + * @lock: per object lock to protect the internals + * + * @cdev: a cooling device debug structure + */ +struct thermal_debugfs { + struct dentry *d_top; + struct mutex lock; + union { + struct cdev_debugfs cdev_dbg; + }; +}; + +void thermal_debug_init(void) +{ + d_root = debugfs_create_dir("thermal", NULL); + if (!d_root) + return; + + d_cdev = debugfs_create_dir("cooling_devices", d_root); +} + +static struct thermal_debugfs *thermal_debugfs_add_id(struct dentry *d, int id) +{ + struct thermal_debugfs *thermal_dbg; + char ids[IDSLENGTH]; + + thermal_dbg = kzalloc(sizeof(*thermal_dbg), GFP_KERNEL); + if (!thermal_dbg) + return NULL; + + mutex_init(&thermal_dbg->lock); + + snprintf(ids, IDSLENGTH, "%d", id); + + thermal_dbg->d_top = debugfs_create_dir(ids, d); + if (!thermal_dbg->d_top) { + kfree(thermal_dbg); + return NULL; + } + + return thermal_dbg; +} + +static void thermal_debugfs_remove_id(struct thermal_debugfs *thermal_dbg) +{ + if (!thermal_dbg) + return; + + debugfs_remove(thermal_dbg->d_top); + + kfree(thermal_dbg); +} + +static struct cdev_record * +thermal_debugfs_cdev_record_alloc(struct thermal_debugfs *thermal_dbg, + struct list_head *lists, int id) +{ + struct cdev_record *cdev_record; + + cdev_record = kzalloc(sizeof(*cdev_record), GFP_KERNEL); + if (!cdev_record) + return NULL; + + cdev_record->id = id; + INIT_LIST_HEAD(&cdev_record->node); + list_add_tail(&cdev_record->node, + &lists[cdev_record->id % CDEVSTATS_HASH_SIZE]); + + return cdev_record; +} + +static struct cdev_record * +thermal_debugfs_cdev_record_find(struct thermal_debugfs *thermal_dbg, + struct list_head *lists, int id) +{ + struct cdev_record *entry; + + list_for_each_entry(entry, &lists[id % CDEVSTATS_HASH_SIZE], node) + if (entry->id == id) + return entry; + + return NULL; +} + +static struct cdev_record * +thermal_debugfs_cdev_record_get(struct thermal_debugfs *thermal_dbg, + struct list_head *lists, int id) +{ + struct cdev_record *cdev_record; + + cdev_record = thermal_debugfs_cdev_record_find(thermal_dbg, lists, id); + if (cdev_record) + return cdev_record; + + return thermal_debugfs_cdev_record_alloc(thermal_dbg, lists, id); +} + +static void thermal_debugfs_cdev_clear(struct cdev_debugfs *cdev_dbg) +{ + int i; + struct cdev_record *entry, *tmp; + + for (i = 0; i < CDEVSTATS_HASH_SIZE; i++) { + + list_for_each_entry_safe(entry, tmp, + &cdev_dbg->transitions[i], node) { + list_del(&entry->node); + kfree(entry); + } + + list_for_each_entry_safe(entry, tmp, + &cdev_dbg->durations[i], node) { + list_del(&entry->node); + kfree(entry); + } + } + + cdev_dbg->total = 0; +} + +static void *cdev_seq_start(struct seq_file *s, loff_t *pos) +{ + struct thermal_debugfs *thermal_dbg = s->private; + + mutex_lock(&thermal_dbg->lock); + + return (*pos < CDEVSTATS_HASH_SIZE) ? pos : NULL; +} + +static void *cdev_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + + return (*pos < CDEVSTATS_HASH_SIZE) ? pos : NULL; +} + +static void cdev_seq_stop(struct seq_file *s, void *v) +{ + struct thermal_debugfs *thermal_dbg = s->private; + + mutex_unlock(&thermal_dbg->lock); +} + +static int cdev_tt_seq_show(struct seq_file *s, void *v) +{ + struct thermal_debugfs *thermal_dbg = s->private; + struct cdev_debugfs *cdev_dbg = &thermal_dbg->cdev_dbg; + struct list_head *transitions = cdev_dbg->transitions; + struct cdev_record *entry; + int i = *(loff_t *)v; + + if (!i) + seq_puts(s, "Transition\tOccurences\n"); + + list_for_each_entry(entry, &transitions[i], node) { + /* + * Assuming maximum cdev states is 1024, the longer + * string for a transition would be "1024->1024\0" + */ + char buffer[11]; + + snprintf(buffer, ARRAY_SIZE(buffer), "%d->%d", + entry->id >> 16, entry->id & 0xFFFF); + + seq_printf(s, "%-10s\t%-10llu\n", buffer, entry->count); + } + + return 0; +} + +static const struct seq_operations tt_sops = { + .start = cdev_seq_start, + .next = cdev_seq_next, + .stop = cdev_seq_stop, + .show = cdev_tt_seq_show, +}; + +DEFINE_SEQ_ATTRIBUTE(tt); + +static int cdev_dt_seq_show(struct seq_file *s, void *v) +{ + struct thermal_debugfs *thermal_dbg = s->private; + struct cdev_debugfs *cdev_dbg = &thermal_dbg->cdev_dbg; + struct list_head *durations = cdev_dbg->durations; + struct cdev_record *entry; + int i = *(loff_t *)v; + + if (!i) + seq_puts(s, "State\tResidency\n"); + + list_for_each_entry(entry, &durations[i], node) { + s64 duration = ktime_to_ms(entry->residency); + + if (entry->id == cdev_dbg->current_state) + duration += ktime_ms_delta(ktime_get(), + cdev_dbg->timestamp); + + seq_printf(s, "%-5d\t%-10llu\n", entry->id, duration); + } + + return 0; +} + +static const struct seq_operations dt_sops = { + .start = cdev_seq_start, + .next = cdev_seq_next, + .stop = cdev_seq_stop, + .show = cdev_dt_seq_show, +}; + +DEFINE_SEQ_ATTRIBUTE(dt); + +static int cdev_clear_set(void *data, u64 val) +{ + struct thermal_debugfs *thermal_dbg = data; + + if (!val) + return -EINVAL; + + mutex_lock(&thermal_dbg->lock); + + thermal_debugfs_cdev_clear(&thermal_dbg->cdev_dbg); + + mutex_unlock(&thermal_dbg->lock); + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(cdev_clear_fops, NULL, cdev_clear_set, "%llu\n"); + +/** + * thermal_debug_cdev_state_update - Update a cooling device state change + * + * Computes a transition and the duration of the previous state residency. + * + * @cdev : a pointer to a cooling device + * @new_state: an integer corresponding to the new cooling device state + */ +void thermal_debug_cdev_state_update(const struct thermal_cooling_device *cdev, + int new_state) +{ + struct thermal_debugfs *thermal_dbg = cdev->debugfs; + struct cdev_debugfs *cdev_dbg; + struct cdev_record *cdev_record; + int transition, old_state; + + if (!thermal_dbg || (thermal_dbg->cdev_dbg.current_state == new_state)) + return; + + mutex_lock(&thermal_dbg->lock); + + cdev_dbg = &thermal_dbg->cdev_dbg; + + old_state = cdev_dbg->current_state; + + /* + * Get the old state information in the durations list. If + * this one does not exist, a new allocated one will be + * returned. Recompute the total duration in the old state and + * get a new timestamp for the new state. + */ + cdev_record = thermal_debugfs_cdev_record_get(thermal_dbg, + cdev_dbg->durations, + old_state); + if (cdev_record) { + ktime_t now = ktime_get(); + ktime_t delta = ktime_sub(now, cdev_dbg->timestamp); + cdev_record->residency = ktime_add(cdev_record->residency, delta); + cdev_dbg->timestamp = now; + } + + cdev_dbg->current_state = new_state; + transition = (old_state << 16) | new_state; + + /* + * Get the transition in the transitions list. If this one + * does not exist, a new allocated one will be returned. + * Increment the occurrence of this transition which is stored + * in the value field. + */ + cdev_record = thermal_debugfs_cdev_record_get(thermal_dbg, + cdev_dbg->transitions, + transition); + if (cdev_record) + cdev_record->count++; + + cdev_dbg->total++; + + mutex_unlock(&thermal_dbg->lock); +} + +/** + * thermal_debug_cdev_add - Add a cooling device debugfs entry + * + * Allocates a cooling device object for debug, initializes the + * statistics and create the entries in sysfs. + * @cdev: a pointer to a cooling device + */ +void thermal_debug_cdev_add(struct thermal_cooling_device *cdev) +{ + struct thermal_debugfs *thermal_dbg; + struct cdev_debugfs *cdev_dbg; + int i; + + thermal_dbg = thermal_debugfs_add_id(d_cdev, cdev->id); + if (!thermal_dbg) + return; + + cdev_dbg = &thermal_dbg->cdev_dbg; + + for (i = 0; i < CDEVSTATS_HASH_SIZE; i++) { + INIT_LIST_HEAD(&cdev_dbg->transitions[i]); + INIT_LIST_HEAD(&cdev_dbg->durations[i]); + } + + cdev_dbg->current_state = 0; + cdev_dbg->timestamp = ktime_get(); + + debugfs_create_file("trans_table", 0400, thermal_dbg->d_top, + thermal_dbg, &tt_fops); + + debugfs_create_file("time_in_state_ms", 0400, thermal_dbg->d_top, + thermal_dbg, &dt_fops); + + debugfs_create_file("clear", 0200, thermal_dbg->d_top, + thermal_dbg, &cdev_clear_fops); + + debugfs_create_u32("total_trans", 0400, thermal_dbg->d_top, + &cdev_dbg->total); + + cdev->debugfs = thermal_dbg; +} + +/** + * thermal_debug_cdev_remove - Remove a cooling device debugfs entry + * + * Frees the statistics memory data and remove the debugfs entry + * + * @cdev: a pointer to a cooling device + */ +void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev) +{ + struct thermal_debugfs *thermal_dbg = cdev->debugfs; + + if (!thermal_dbg) + return; + + mutex_lock(&thermal_dbg->lock); + + thermal_debugfs_cdev_clear(&thermal_dbg->cdev_dbg); + cdev->debugfs = NULL; + thermal_debugfs_remove_id(thermal_dbg); + + mutex_unlock(&thermal_dbg->lock); +} diff --git a/drivers/thermal/thermal_debugfs.h b/drivers/thermal/thermal_debugfs.h new file mode 100644 index 000000000000..341499388448 --- /dev/null +++ b/drivers/thermal/thermal_debugfs.h @@ -0,0 +1,14 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifdef CONFIG_THERMAL_DEBUGFS +void thermal_debug_init(void); +void thermal_debug_cdev_add(struct thermal_cooling_device *cdev); +void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev); +void thermal_debug_cdev_state_update(const struct thermal_cooling_device *cdev, int state); +#else +static inline void thermal_debug_init(void) {} +static inline void thermal_debug_cdev_add(struct thermal_cooling_device *cdev) {} +static inline void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev) {} +static inline void thermal_debug_cdev_state_update(const struct thermal_cooling_device *cdev, + int state) {} +#endif /* CONFIG_THERMAL_DEBUGFS */ diff --git a/drivers/thermal/thermal_helpers.c b/drivers/thermal/thermal_helpers.c index 69e8ea4aa908..435c123b721b 100644 --- a/drivers/thermal/thermal_helpers.c +++ b/drivers/thermal/thermal_helpers.c @@ -151,14 +151,29 @@ int thermal_zone_get_temp(struct thermal_zone_device *tz, int *temp) } EXPORT_SYMBOL_GPL(thermal_zone_get_temp); -static void thermal_cdev_set_cur_state(struct thermal_cooling_device *cdev, - int target) +void thermal_set_delay_jiffies(unsigned long *delay_jiffies, int delay_ms) { - if (cdev->ops->set_cur_state(cdev, target)) - return; + *delay_jiffies = msecs_to_jiffies(delay_ms); + if (delay_ms > 1000) + *delay_jiffies = round_jiffies(*delay_jiffies); +} + +static int thermal_cdev_set_cur_state(struct thermal_cooling_device *cdev, int state) +{ + int ret; - thermal_notify_cdev_state_update(cdev->id, target); - thermal_cooling_device_stats_update(cdev, target); + /* + * No check is needed for the ops->set_cur_state as the + * registering function checked the ops are correctly set + */ + ret = cdev->ops->set_cur_state(cdev, state); + if (!ret) { + thermal_notify_cdev_state_update(cdev->id, state); + thermal_cooling_device_stats_update(cdev, state); + thermal_debug_cdev_state_update(cdev, state); + } + + return ret; } void __thermal_cdev_update(struct thermal_cooling_device *cdev) diff --git a/include/linux/thermal.h b/include/linux/thermal.h index 0ea99f50d57c..8461f008c3de 100644 --- a/include/linux/thermal.h +++ b/include/linux/thermal.h @@ -32,6 +32,7 @@ struct thermal_zone_device; struct thermal_cooling_device; struct thermal_instance; +struct thermal_debugfs; struct thermal_attr; enum thermal_trend { @@ -110,6 +111,9 @@ struct thermal_cooling_device { struct mutex lock; /* protect thermal_instances list */ struct list_head thermal_instances; struct list_head node; +#ifdef CONFIG_THERMAL_DEBUGFS + struct thermal_debugfs *debugfs; +#endif }; /** @@ -183,6 +187,9 @@ struct thermal_zone_device { struct list_head node; struct delayed_work poll_queue; enum thermal_notify_event notify_event; +#ifdef CONFIG_THERMAL_DEBUGFS + struct thermal_debugfs *debugfs; +#endif }; /** From patchwork Wed Dec 27 10:26:43 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Daniel Lezcano X-Patchwork-Id: 13505286 Received: from mail-lj1-f180.google.com (mail-lj1-f180.google.com [209.85.208.180]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D2B592C868 for ; Wed, 27 Dec 2023 10:26:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linaro.org Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linaro.org Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=linaro.org header.i=@linaro.org header.b="aYHB0cCN" Received: by mail-lj1-f180.google.com with SMTP id 38308e7fff4ca-2cca9a9362aso35780121fa.3 for ; Wed, 27 Dec 2023 02:26:55 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; t=1703672814; x=1704277614; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=5pYdZbFTr6EOagElFAJ5jqcne6NyYjq9KUM2N1ydhjs=; b=aYHB0cCNUANJjgzmot+A20M/MJOl5QInj1GsFhBbv1YEHOVbYtOanuPQNBvKMEPLl3 P/jgnBsJqHm0ufRhdL3HgUK8e+XBoK8Gx5FIw1tOxuimUBT51Ihaf2SUbNJoOCvc6rGw Vl/DurRPl6VCB+5/cJzCIIccGZ4oc+jRyknp5gchWadUEDYuaXcW4g9af2xHo0gnRHaD +kjfqJIkoPGw1bHyWECAazmiIVR3kHpvePo2dLFLmxY1s4nsX9tUt/GsEvpzL+z5hGnD N3m2v1L0oKWH/+ImNXtf1JTJ1nVz7Q9XeeFF+ENUWAlU2479uXuAx/xVjNeqbLjNzoiQ c+Gg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1703672814; x=1704277614; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=5pYdZbFTr6EOagElFAJ5jqcne6NyYjq9KUM2N1ydhjs=; b=u0hJvtVoLTfo7SXdOQ4z2PsZXR32F+/KeaGNo0gt/99FupoiHqExsPiM+HYQHhsdSF syd5nNjbd57W9tzr+TRlQ133bQy9yn9fd2L3HR3F7w/PTtzGIMSqibAbs9Q3OxhOBh13 co8Guv1A0evrKk7FHGeUiqw0JKBlcfTsFlKNpGY4mVdJY2OVn6EuBKq+IAe7cyz6Uthr GuFggxuCcQp+og61TT2Ak2pOu4gU1Cp/XYsq21Rb/Mzyr91MDOfE1ndJyOkw9QGBYtiL WTAp9biIpsomRIWMD3rR58F8P3iDghGr3XNPBlG2kmcPcpjqmFJwP9uXhtWv1NXzPmA7 4hJg== X-Gm-Message-State: AOJu0YyT1ibRt2kZ5msjaFEIv8wayqpOqK2D98KHLGUSeFA57DnIVHHU o3vqCEFGIoVOQTvpwlyZ47CzVqU3Q01Fp5jrsgKSZhZVl/0= X-Google-Smtp-Source: AGHT+IFEP6H0ZGbER0ED4jo46IPV/kLgZJuEcrMS2UzoIcFt7QYyUQ77bOPpfYt3wMSylNRDJNE85A== X-Received: by 2002:a05:651c:168e:b0:2cc:dcb6:5198 with SMTP id bd14-20020a05651c168e00b002ccdcb65198mr152173ljb.110.1703672813669; Wed, 27 Dec 2023 02:26:53 -0800 (PST) Received: from mai.. (146725694.box.freepro.com. [130.180.211.218]) by smtp.gmail.com with ESMTPSA id y17-20020adff151000000b0033664ffaf5dsm14556763wro.37.2023.12.27.02.26.53 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 27 Dec 2023 02:26:53 -0800 (PST) From: Daniel Lezcano To: daniel.lezcano@linaro.org, rjw@rjwysocki.net, lukasz.luba@arm.com Cc: rui.zhang@intel.com, linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Subject: [PATCH v4 2/2] thermal/debugfs: Add thermal debugfs information for mitigation episodes Date: Wed, 27 Dec 2023 11:26:43 +0100 Message-Id: <20231227102643.2655651-2-daniel.lezcano@linaro.org> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231227102643.2655651-1-daniel.lezcano@linaro.org> References: <20231227102643.2655651-1-daniel.lezcano@linaro.org> Precedence: bulk X-Mailing-List: linux-pm@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 The mitigation episodes are recorded. A mitigation episode happens when the first trip point is crossed the way up and then the way down. During this episode other trip points can be crossed also and are accounted for this mitigation episode. The interesting information is the average temperature at the trip point, the undershot and the overshot. The standard deviation of the mitigated temperature will be added later. The thermal debugfs directory structure tries to stay consistent with the sysfs one but in a very simplified way: thermal/ `-- thermal_zones |-- 0 | `-- mitigations `-- 1 `-- mitigations The content of the mitigations file has the following format: ,-Mitigation at 349988258us, duration=130136ms | trip | type | temp(°mC) | hyst(°mC) | duration | avg(°mC) | min(°mC) | max(°mC) | | 0 | passive | 65000 | 2000 | 130136 | 68227 | 62500 | 75625 | | 1 | passive | 75000 | 2000 | 104209 | 74857 | 71666 | 77500 | ,-Mitigation at 272451637us, duration=75000ms | trip | type | temp(°mC) | hyst(°mC) | duration | avg(°mC) | min(°mC) | max(°mC) | | 0 | passive | 65000 | 2000 | 75000 | 68561 | 62500 | 75000 | | 1 | passive | 75000 | 2000 | 60714 | 74820 | 70555 | 77500 | ,-Mitigation at 238184119us, duration=27316ms | trip | type | temp(°mC) | hyst(°mC) | duration | avg(°mC) | min(°mC) | max(°mC) | | 0 | passive | 65000 | 2000 | 27316 | 73377 | 62500 | 75000 | | 1 | passive | 75000 | 2000 | 19468 | 75284 | 69444 | 77500 | ,-Mitigation at 39863713us, duration=136196ms | trip | type | temp(°mC) | hyst(°mC) | duration | avg(°mC) | min(°mC) | max(°mC) | | 0 | passive | 65000 | 2000 | 136196 | 73922 | 62500 | 75000 | | 1 | passive | 75000 | 2000 | 91721 | 74386 | 69444 | 78125 | More information for a better understanding of the thermal behavior will be added after. The idea is to give detailed statistics information about the undershots and overshots, the temperature speed, etc... As all the information in a single file is too much, the idea would be to create a directory named with the mitigation timestamp where all data could be added. Please note this code is immune against trip ordering but not against a trip temperature change while a mitigation is happening. However, this situation should be extremely rare, perhaps not happening and we might question ourselves if something should be done in the core framework for other components first. Signed-off-by: Daniel Lezcano --- Changelog: - v4 - Reworded "trip violation" (Rafael) - Reworded some comments and clarified other ones (Rafael) - Renamed s/tz/tz_dbg/ for clarity (Rafael) - Used flexible array and struct_size for allocation (Rafael) - Renamed s/dfs/thermal_dbg/ for clarity (Rafael) - Used intermediate variable to prevent dereferencing multiple times a structure (Rafael) - Renamed variable s/trip/trip_id/ (Rafael) - Clarified trip_index usage in the comment (Rafael) - v3 - Fixed kerneldoc (kbuild) - Fixed wrong indentation s/// - v2 - Applied changes based on comments from patch 1/2 - Constified structure in function parameters - v1 (from RFC): - Replaced exported function name s/debugfs/debug/ - Used "struct thermal_trip" parameter instead of "trip_id" - Renamed handle_way_[up|down] by tz_trip_[up|down] - Replaced thermal_debug_tz_[unregister|register] by [add|remove] --- drivers/thermal/thermal_core.c | 7 + drivers/thermal/thermal_debugfs.c | 383 +++++++++++++++++++++++++++++- drivers/thermal/thermal_debugfs.h | 14 ++ 3 files changed, 403 insertions(+), 1 deletion(-) diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c index 33332d401b13..a0cbe8d7b945 100644 --- a/drivers/thermal/thermal_core.c +++ b/drivers/thermal/thermal_core.c @@ -367,6 +367,7 @@ static void handle_thermal_trip(struct thermal_zone_device *tz, thermal_notify_tz_trip_up(tz->id, thermal_zone_trip_id(tz, trip), tz->temperature); + thermal_debug_tz_trip_up(tz, trip); trip->threshold = trip->temperature - trip->hysteresis; } else { trip->threshold = trip->temperature; @@ -386,6 +387,7 @@ static void handle_thermal_trip(struct thermal_zone_device *tz, thermal_notify_tz_trip_down(tz->id, thermal_zone_trip_id(tz, trip), tz->temperature); + thermal_debug_tz_trip_down(tz, trip); trip->threshold = trip->temperature; } else { trip->threshold = trip->temperature - trip->hysteresis; @@ -417,6 +419,7 @@ static void update_temperature(struct thermal_zone_device *tz) trace_thermal_temperature(tz); thermal_genl_sampling_temp(tz->id, temp); + thermal_debug_update_temp(tz); } static void thermal_zone_device_init(struct thermal_zone_device *tz) @@ -1396,6 +1399,8 @@ thermal_zone_device_register_with_trips(const char *type, struct thermal_trip *t thermal_notify_tz_create(tz->id, tz->type); + thermal_debug_tz_add(tz); + return tz; unregister: @@ -1461,6 +1466,8 @@ void thermal_zone_device_unregister(struct thermal_zone_device *tz) if (!tz) return; + thermal_debug_tz_remove(tz); + tz_id = tz->id; mutex_lock(&thermal_list_lock); diff --git a/drivers/thermal/thermal_debugfs.c b/drivers/thermal/thermal_debugfs.c index 84e5ef3dc16c..0f0f0e603428 100644 --- a/drivers/thermal/thermal_debugfs.c +++ b/drivers/thermal/thermal_debugfs.c @@ -14,8 +14,11 @@ #include #include +#include "thermal_core.h" + static struct dentry *d_root; static struct dentry *d_cdev; +static struct dentry *d_tz; /* * Length of the string containing the thermal zone id or the cooling @@ -81,22 +84,87 @@ struct cdev_record { }; }; +/** + * struct trip_stats - Thermal trip statistics + * + * The trip_stats structure has the relevant information to show the + * statistics related to temperature going above a trip point. + * + * @timestamp: the trip crossing timestamp + * @duration: total time when the zone temperature was above the trip point + * @count: the number of times the zone temperature was above the trip point + * @max: maximum recorded temperature above the trip point + * @min: minimum recorded temperature above the trip point + * @avg: average temperature above the trip point + */ +struct trip_stats { + ktime_t timestamp; + ktime_t duration; + int count; + int max; + int min; + int avg; +}; + +/** + * struct tz_events - Collection of all mitigation episodes for a thermal zone + * + * The tz_events structure describes a mitigation episode. A + * mitigation episode is when the mitigation begins and ends. During + * this episode we can have multiple trip points crossed the way up + * and down if there are multiple trip describes in the firmware. + * + * @timestamp: First trip point crossed the way up + * @duration: total duration of the mitigation episode + * @node: a list element to be added to the list of tz events + * @trip_stats: per trip point statistics, flexible array + */ +struct tz_events { + ktime_t timestamp; + ktime_t duration; + struct list_head node; + struct trip_stats trip_stats[]; +}; + +/** + * struct tz_debugfs - Store all mitigations episodes for a thermal zone + * + * The tz_debugfs structure contains the list of the mitigation + * episodes and has to track which trip point has been crossed in + * order to handle correctly nested trip point mitigation episodes. + * + * We keep the history of the trip point crossed in an array and as we + * can go back and forth inside this history, eg. trip 0,1,2,1,2,1,0, + * we keep track of the current position in the history array. + * + * @tz_events: a list of thermal mitigation episodes + * @trips_crossed: an array of trip points crossed by id + * @trip_index: the current trip point index in the trip points history + */ +struct tz_debugfs { + struct list_head tz_events; + int *trips_crossed; + int trip_index; +}; + /** * struct thermal_debugfs - High level structure for a thermal object in debugfs * * The thermal_debugfs structure is the common structure used by the - * cooling device to compute the statistics. + * cooling device or the thermal zone to store the statistics. * * @d_top: top directory of the thermal object directory * @lock: per object lock to protect the internals * * @cdev: a cooling device debug structure + * @tz: a thermal zone debug structure */ struct thermal_debugfs { struct dentry *d_top; struct mutex lock; union { struct cdev_debugfs cdev_dbg; + struct tz_debugfs tz_dbg; }; }; @@ -107,6 +175,10 @@ void thermal_debug_init(void) return; d_cdev = debugfs_create_dir("cooling_devices", d_root); + if (!d_cdev) + return; + + d_tz = debugfs_create_dir("thermal_zones", d_root); } static struct thermal_debugfs *thermal_debugfs_add_id(struct dentry *d, int id) @@ -443,3 +515,312 @@ void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev) mutex_unlock(&thermal_dbg->lock); } + +static struct tz_events *thermal_debugfs_tz_event_alloc(struct thermal_zone_device *tz, + ktime_t now) +{ + struct tz_events *tze; + int i; + + tze = kzalloc(struct_size(tze, trip_stats, tz->num_trips), GFP_KERNEL); + if (!tze) + return NULL; + + INIT_LIST_HEAD(&tze->node); + tze->timestamp = now; + + for (i = 0; i < tz->num_trips; i++) { + tze->trip_stats[i].min = INT_MAX; + tze->trip_stats[i].max = INT_MIN; + } + + return tze; +} + +void thermal_debug_tz_trip_up(struct thermal_zone_device *tz, + const struct thermal_trip *trip) +{ + struct tz_events *tze; + struct tz_debugfs *tz_dbg; + struct thermal_debugfs *thermal_dbg = tz->debugfs; + int temperature = tz->temperature; + int trip_id = thermal_zone_trip_id(tz, trip); + ktime_t now = ktime_get(); + + if (!thermal_dbg) + return; + + mutex_lock(&thermal_dbg->lock); + + tz_dbg = &thermal_dbg->tz_dbg; + + /* + * The mitigation is starting. A mitigation can contain + * several episodes where each of them is related to a + * temperature crossing a trip point. The episodes are + * nested. That means when the temperature is crossing the + * first trip point, the duration begins to be measured. If + * the temperature continues to increase and reaches the + * second trip point, the duration of the first trip must be + * also accumulated. + * + * eg. + * + * temp + * ^ + * | -------- + * trip 2 / \ ------ + * | /| |\ /| |\ + * trip 1 / | | `---- | | \ + * | /| | | | | |\ + * trip 0 / | | | | | | \ + * | /| | | | | | | |\ + * | / | | | | | | | | `-- + * | / | | | | | | | | + * |----- | | | | | | | | + * | | | | | | | | | + * --------|-|-|--------|--------|------|-|-|------------------> time + * | | |<--t2-->| |<-t2'>| | | + * | | | | + * | |<------------t1------------>| | + * | | + * |<-------------t0--------------->| + * + */ + if (tz_dbg->trip_index < 0) { + tze = thermal_debugfs_tz_event_alloc(tz, now); + if (!tze) + return; + + list_add(&tze->node, &tz_dbg->tz_events); + } + + /* + * The trip_index points to the current position in the + * trip_crossed[] array. Each time a trip point is crossed the + * way up, the trip_id is stored in the array and the + * trip_index is incremented. A negative trip_index means we + * are entering a mitigation episode. + * + * The trip ids may not be in the ascending order but the + * result in the array trips_crossed will be in the ascending + * temperature order. + */ + tz_dbg->trip_index++; + tz_dbg->trips_crossed[tz_dbg->trip_index] = trip_id; + + tze = list_first_entry(&tz_dbg->tz_events, struct tz_events, node); + tze->trip_stats[trip_id].timestamp = now; + tze->trip_stats[trip_id].max = max(tze->trip_stats[trip_id].max, temperature); + tze->trip_stats[trip_id].min = min(tze->trip_stats[trip_id].min, temperature); + tze->trip_stats[trip_id].avg = tze->trip_stats[trip_id].avg + + (temperature - tze->trip_stats[trip_id].avg) / + tze->trip_stats[trip_id].count; + + mutex_unlock(&thermal_dbg->lock); +} + +void thermal_debug_tz_trip_down(struct thermal_zone_device *tz, + const struct thermal_trip *trip) +{ + struct thermal_debugfs *thermal_dbg = tz->debugfs; + struct tz_events *tze; + struct tz_debugfs *tz_dbg; + + int trip_id = thermal_zone_trip_id(tz, trip); + ktime_t delta, now = ktime_get(); + + if (!thermal_dbg) + return; + + mutex_lock(&thermal_dbg->lock); + + tz_dbg = &thermal_dbg->tz_dbg; + + /* + * The temperature crosses the way down but there was not + * mitigation detected before. That may happen when the + * temperature is greater than a trip point when registering a + * thermal zone, which is a common use case as the kernel has + * no mitigation mechanism yet at boot time. + */ + if (tz_dbg->trip_index < 0) + goto out; + + tze = list_first_entry(&tz_dbg->tz_events, struct tz_events, node); + + delta = ktime_sub(now, tze->trip_stats[trip_id].timestamp); + tze->trip_stats[trip_id].duration = ktime_add(delta, + tze->trip_stats[trip_id].duration); + + tz_dbg->trip_index--; + + /* + * This event closes the mitigation as we are crossing the + * last trip point the way down. + */ + if (tz_dbg->trip_index < 0) + tze->duration = ktime_sub(now, tze->timestamp); + +out: + mutex_unlock(&thermal_dbg->lock); +} + +void thermal_debug_update_temp(struct thermal_zone_device *tz) +{ + struct thermal_debugfs *thermal_dbg = tz->debugfs; + struct tz_events *tze; + struct tz_debugfs *tz_dbg; + int trip_id; + + if (!thermal_dbg) + return; + + mutex_lock(&thermal_dbg->lock); + + tz_dbg = &thermal_dbg->tz_dbg; + + if (tz_dbg->trip_index >= 0) { + trip_id = tz_dbg->trip_index; + tze = list_first_entry(&tz_dbg->tz_events, struct tz_events, node); + tze->trip_stats[trip_id].count++; + tze->trip_stats[trip_id].max = max(tze->trip_stats[trip_id].max, tz->temperature); + tze->trip_stats[trip_id].min = min(tze->trip_stats[trip_id].min, tz->temperature); + tze->trip_stats[trip_id].avg = tze->trip_stats[trip_id].avg + + (tz->temperature - tze->trip_stats[trip_id].avg) / + tze->trip_stats[trip_id].count; + } + + mutex_unlock(&thermal_dbg->lock); +} + +static void *tze_seq_start(struct seq_file *s, loff_t *pos) +{ + struct thermal_zone_device *tz = s->private; + struct thermal_debugfs *thermal_dbg = tz->debugfs; + struct tz_debugfs *tz_dbg = &thermal_dbg->tz_dbg; + + mutex_lock(&thermal_dbg->lock); + + return seq_list_start(&tz_dbg->tz_events, *pos); +} + +static void *tze_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + struct thermal_zone_device *tz = s->private; + struct thermal_debugfs *thermal_dbg = tz->debugfs; + struct tz_debugfs *tz_dbg = &thermal_dbg->tz_dbg; + + return seq_list_next(v, &tz_dbg->tz_events, pos); +} + +static void tze_seq_stop(struct seq_file *s, void *v) +{ + struct thermal_zone_device *tz = s->private; + struct thermal_debugfs *thermal_dbg = tz->debugfs; + + mutex_unlock(&thermal_dbg->lock); +} + +static int tze_seq_show(struct seq_file *s, void *v) +{ + struct thermal_zone_device *tz = s->private; + struct tz_events *tze; + int i; + + tze = list_entry((struct list_head *)v, struct tz_events, node); + + seq_printf(s, ",-Mitigation at %lluus, duration=%llums\n", + ktime_to_us(tze->timestamp), + ktime_to_ms(tze->duration)); + + seq_printf(s, "| trip | type | temp(°mC) | hyst(°mC) | duration | avg(°mC) | min(°mC) | max(°mC) |\n"); + + for (i = 0; i < tz->num_trips; i++) { + + struct thermal_trip trip; + const char *type; + + if (__thermal_zone_get_trip(tz, i, &trip)) + continue; + + /* + * There is no possible mitigation happening at the + * critical trip point, so the stats will be always + * zero, skip this trip point + */ + if (trip.type == THERMAL_TRIP_CRITICAL) + continue; + + if (trip.type == THERMAL_TRIP_PASSIVE) + type = "passive"; + else if (trip.type == THERMAL_TRIP_ACTIVE) + type = "active"; + else + type = "hot"; + + seq_printf(s, "| %*d | %*s | %*d | %*d | %*lld | %*d | %*d | %*d |\n", + 4 , i, + 8, type, + 9, trip.temperature, + 9, trip.hysteresis, + 10, ktime_to_ms(tze->trip_stats[i].duration), + 9, tze->trip_stats[i].avg, + 9, tze->trip_stats[i].min, + 9, tze->trip_stats[i].max); + } + + return 0; +} + +static const struct seq_operations tze_sops = { + .start = tze_seq_start, + .next = tze_seq_next, + .stop = tze_seq_stop, + .show = tze_seq_show, +}; + +DEFINE_SEQ_ATTRIBUTE(tze); + +void thermal_debug_tz_add(struct thermal_zone_device *tz) +{ + struct thermal_debugfs *thermal_dbg; + struct tz_debugfs *tz_dbg; + + thermal_dbg = thermal_debugfs_add_id(d_tz, tz->id); + if (!thermal_dbg) + return; + + tz_dbg = &thermal_dbg->tz_dbg; + + tz_dbg->trips_crossed = kzalloc(sizeof(int) * tz->num_trips, GFP_KERNEL); + if (!tz_dbg->trips_crossed) { + thermal_debugfs_remove_id(thermal_dbg); + return; + } + + /* + * Trip index '-1' means no mitigation + */ + tz_dbg->trip_index = -1; + INIT_LIST_HEAD(&tz_dbg->tz_events); + + debugfs_create_file("mitigations", 0400, thermal_dbg->d_top, tz, &tze_fops); + + tz->debugfs = thermal_dbg; +} + +void thermal_debug_tz_remove(struct thermal_zone_device *tz) +{ + struct thermal_debugfs *thermal_dbg = tz->debugfs; + + if (!thermal_dbg) + return; + + mutex_lock(&thermal_dbg->lock); + + tz->debugfs = NULL; + thermal_debugfs_remove_id(thermal_dbg); + + mutex_unlock(&thermal_dbg->lock); +} diff --git a/drivers/thermal/thermal_debugfs.h b/drivers/thermal/thermal_debugfs.h index 341499388448..155b9af5fe87 100644 --- a/drivers/thermal/thermal_debugfs.h +++ b/drivers/thermal/thermal_debugfs.h @@ -5,10 +5,24 @@ void thermal_debug_init(void); void thermal_debug_cdev_add(struct thermal_cooling_device *cdev); void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev); void thermal_debug_cdev_state_update(const struct thermal_cooling_device *cdev, int state); +void thermal_debug_tz_add(struct thermal_zone_device *tz); +void thermal_debug_tz_remove(struct thermal_zone_device *tz); +void thermal_debug_tz_trip_up(struct thermal_zone_device *tz, + const struct thermal_trip *trip); +void thermal_debug_tz_trip_down(struct thermal_zone_device *tz, + const struct thermal_trip *trip); +void thermal_debug_update_temp(struct thermal_zone_device *tz); #else static inline void thermal_debug_init(void) {} static inline void thermal_debug_cdev_add(struct thermal_cooling_device *cdev) {} static inline void thermal_debug_cdev_remove(struct thermal_cooling_device *cdev) {} static inline void thermal_debug_cdev_state_update(const struct thermal_cooling_device *cdev, int state) {} +static inline void thermal_debug_tz_add(struct thermal_zone_device *tz) {} +static inline void thermal_debug_tz_remove(struct thermal_zone_device *tz) {} +static inline void thermal_debug_tz_trip_up(struct thermal_zone_device *tz, + const struct thermal_trip *trip) {}; +static inline void thermal_debug_tz_trip_down(struct thermal_zone_device *tz, + const struct thermal_trip *trip) {} +static inline void thermal_debug_update_temp(struct thermal_zone_device *tz) {} #endif /* CONFIG_THERMAL_DEBUGFS */