// SPDX-License-Identifier: GPL-2.0 /* * trace_hwlatdetect.c - A simple Hardware Latency detector. * * Use this tracer to detect large system latencies induced by the behavior of * certain underlying system hardware or firmware, independent of Linux itself. * The code was developed originally to detect the presence of SMIs on Intel * and AMD systems, although there is no dependency upon x86 herein. * * The classical example usage of this tracer is in detecting the presence of * SMIs or System Management Interrupts on Intel and AMD systems. An SMI is a * somewhat special form of hardware interrupt spawned from earlier CPU debug * modes in which the (BIOS/EFI/etc.) firmware arranges for the South Bridge * LPC (or other device) to generate a special interrupt under certain * circumstances, for example, upon expiration of a special SMI timer device, * due to certain external thermal readings, on certain I/O address accesses, * and other situations. An SMI hits a special CPU pin, triggers a special * SMI mode (complete with special memory map), and the OS is unaware. * * Although certain hardware-inducing latencies are necessary (for example, * a modern system often requires an SMI handler for correct thermal control * and remote management) they can wreak havoc upon any OS-level performance * guarantees toward low-latency, especially when the OS is not even made * aware of the presence of these interrupts. For this reason, we need a * somewhat brute force mechanism to detect these interrupts. In this case, * we do it by hogging all of the CPU(s) for configurable timer intervals, * sampling the built-in CPU timer, looking for discontiguous readings. * * WARNING: This implementation necessarily introduces latencies. Therefore, * you should NEVER use this tracer while running in a production * environment requiring any kind of low-latency performance * guarantee(s). * * Copyright (C) 2008-2009 Jon Masters, Red Hat, Inc. <jcm@redhat.com> * Copyright (C) 2013-2016 Steven Rostedt, Red Hat, Inc. <srostedt@redhat.com> * * Includes useful feedback from Clark Williams <clark@redhat.com> * */ #include <linux/kthread.h> #include <linux/tracefs.h> #include <linux/uaccess.h> #include <linux/cpumask.h> #include <linux/delay.h> #include <linux/sched/clock.h> #include "trace.h" static struct trace_array *hwlat_trace; #define U64STR_SIZE 22 /* 20 digits max */ #define BANNER "hwlat_detector: " #define DEFAULT_SAMPLE_WINDOW 1000000 /* 1s */ #define DEFAULT_SAMPLE_WIDTH 500000 /* 0.5s */ #define DEFAULT_LAT_THRESHOLD 10 /* 10us */ /* sampling thread*/ static struct task_struct *hwlat_kthread; static struct dentry *hwlat_sample_width; /* sample width us */ static struct dentry *hwlat_sample_window; /* sample window us */ /* Save the previous tracing_thresh value */ static unsigned long save_tracing_thresh; /* NMI timestamp counters */ static u64 nmi_ts_start; static u64 nmi_total_ts; static int nmi_count; static int nmi_cpu; /* Tells NMIs to call back to the hwlat tracer to record timestamps */ bool trace_hwlat_callback_enabled; /* If the user changed threshold, remember it */ static u64 last_tracing_thresh = DEFAULT_LAT_THRESHOLD * NSEC_PER_USEC; /* Individual latency samples are stored here when detected. */ struct hwlat_sample { u64 seqnum; /* unique sequence */ u64 duration; /* delta */ u64 outer_duration; /* delta (outer loop) */ u64 nmi_total_ts; /* Total time spent in NMIs */ struct timespec64 timestamp; /* wall time */ int nmi_count; /* # NMIs during this sample */ }; /* keep the global state somewhere. */ static struct hwlat_data { struct mutex lock; /* protect changes */ u64 count; /* total since reset */ u64 sample_window; /* total sampling window (on+off) */ u64 sample_width; /* active sampling portion of window */ } hwlat_data = { .sample_window = DEFAULT_SAMPLE_WINDOW, .sample_width = DEFAULT_SAMPLE_WIDTH, }; static void trace_hwlat_sample(struct hwlat_sample *sample) { struct trace_array *tr = hwlat_trace; struct trace_event_call *call = &event_hwlat; struct ring_buffer *buffer = tr->trace_buffer.buffer; struct ring_buffer_event *event; struct hwlat_entry *entry; unsigned long flags; int pc; pc = preempt_count(); local_save_flags(flags); event = trace_buffer_lock_reserve(buffer, TRACE_HWLAT, sizeof(*entry), flags, pc); if (!event) return; entry = ring_buffer_event_data(event); entry->seqnum = sample->seqnum; entry->duration = sample->duration; entry->outer_duration = sample->outer_duration; entry->timestamp = sample->timestamp; entry->nmi_total_ts = sample->nmi_total_ts; entry->nmi_count = sample->nmi_count; if (!call_filter_check_discard(call, entry, buffer, event)) trace_buffer_unlock_commit_nostack(buffer, event); } /* Macros to encapsulate the time capturing infrastructure */ #define time_type u64 #define time_get() trace_clock_local() #define time_to_us(x) div_u64(x, 1000) #define time_sub(a, b) ((a) - (b)) #define init_time(a, b) (a = b) #define time_u64(a) a void trace_hwlat_callback(bool enter) { if (smp_processor_id() != nmi_cpu) return; /* * Currently trace_clock_local() calls sched_clock() and the * generic version is not NMI safe. */ if (!IS_ENABLED(CONFIG_GENERIC_SCHED_CLOCK)) { if (enter) nmi_ts_start = time_get(); else nmi_total_ts = time_get() - nmi_ts_start; } if (enter) nmi_count++; } /** * get_sample - sample the CPU TSC and look for likely hardware latencies * * Used to repeatedly capture the CPU TSC (or similar), looking for potential * hardware-induced latency. Called with interrupts disabled and with * hwlat_data.lock held. */ static int get_sample(void) { struct trace_array *tr = hwlat_trace; time_type start, t1, t2, last_t2; s64 diff, total, last_total = 0; u64 sample = 0; u64 thresh = tracing_thresh; u64 outer_sample = 0; int ret = -1; do_div(thresh, NSEC_PER_USEC); /* modifies interval value */ nmi_cpu = smp_processor_id(); nmi_total_ts = 0; nmi_count = 0; /* Make sure NMIs see this first */ barrier(); trace_hwlat_callback_enabled = true; init_time(last_t2, 0); start = time_get(); /* start timestamp */ do { t1 = time_get(); /* we'll look for a discontinuity */ t2 = time_get(); if (time_u64(last_t2)) { /* Check the delta from outer loop (t2 to next t1) */ diff = time_to_us(time_sub(t1, last_t2)); /* This shouldn't happen */ if (diff < 0) { pr_err(BANNER "time running backwards\n"); goto out; } if (diff > outer_sample) outer_sample = diff; } last_t2 = t2; total = time_to_us(time_sub(t2, start)); /* sample width */ /* Check for possible overflows */ if (total < last_total) { pr_err("Time total overflowed\n"); break; } last_total = total; /* This checks the inner loop (t1 to t2) */ diff = time_to_us(time_sub(t2, t1)); /* current diff */ /* This shouldn't happen */ if (diff < 0) { pr_err(BANNER "time running backwards\n"); goto out; } if (diff > sample) sample = diff; /* only want highest value */ } while (total <= hwlat_data.sample_width); barrier(); /* finish the above in the view for NMIs */ trace_hwlat_callback_enabled = false; barrier(); /* Make sure nmi_total_ts is no longer updated */ ret = 0; /* If we exceed the threshold value, we have found a hardware latency */ if (sample > thresh || outer_sample > thresh) { struct hwlat_sample s; ret = 1; /* We read in microseconds */ if (nmi_total_ts) do_div(nmi_total_ts, NSEC_PER_USEC); hwlat_data.count++; s.seqnum = hwlat_data.count; s.duration = sample; s.outer_duration = outer_sample; ktime_get_real_ts64(&s.timestamp); s.nmi_total_ts = nmi_total_ts; s.nmi_count = nmi_count; trace_hwlat_sample(&s); /* Keep a running maximum ever recorded hardware latency */ if (sample > tr->max_latency) tr->max_latency = sample; } out: return ret; } static struct cpumask save_cpumask; static bool disable_migrate; static void move_to_next_cpu(void) { struct cpumask *current_mask = &save_cpumask; int next_cpu; if (disable_migrate) return; /* * If for some reason the user modifies the CPU affinity * of this thread, than stop migrating for the duration * of the current test. */ if (!cpumask_equal(current_mask, current->cpus_ptr)) goto disable; get_online_cpus(); cpumask_and(current_mask, cpu_online_mask, tracing_buffer_mask); next_cpu = cpumask_next(smp_processor_id(), current_mask); put_online_cpus(); if (next_cpu >= nr_cpu_ids) next_cpu = cpumask_first(current_mask); if (next_cpu >= nr_cpu_ids) /* Shouldn't happen! */ goto disable; cpumask_clear(current_mask); cpumask_set_cpu(next_cpu, current_mask); sched_setaffinity(0, current_mask); return; disable: disable_migrate = true; } /* * kthread_fn - The CPU time sampling/hardware latency detection kernel thread * * Used to periodically sample the CPU TSC via a call to get_sample. We * disable interrupts, which does (intentionally) introduce latency since we * need to ensure nothing else might be running (and thus preempting). * Obviously this should never be used in production environments. * * Executes one loop interaction on each CPU in tracing_cpumask sysfs file. */ static int kthread_fn(void *data) { u64 interval; while (!kthread_should_stop()) { move_to_next_cpu(); local_irq_disable(); get_sample(); local_irq_enable(); mutex_lock(&hwlat_data.lock); interval = hwlat_data.sample_window - hwlat_data.sample_width; mutex_unlock(&hwlat_data.lock); do_div(interval, USEC_PER_MSEC); /* modifies interval value */ /* Always sleep for at least 1ms */ if (interval < 1) interval = 1; if (msleep_interruptible(interval)) break; } return 0; } /** * start_kthread - Kick off the hardware latency sampling/detector kthread * * This starts the kernel thread that will sit and sample the CPU timestamp * counter (TSC or similar) and look for potential hardware latencies. */ static int start_kthread(struct trace_array *tr) { struct cpumask *current_mask = &save_cpumask; struct task_struct *kthread; int next_cpu; if (WARN_ON(hwlat_kthread)) return 0; /* Just pick the first CPU on first iteration */ current_mask = &save_cpumask; get_online_cpus(); cpumask_and(current_mask, cpu_online_mask, tracing_buffer_mask); put_online_cpus(); next_cpu = cpumask_first(current_mask); kthread = kthread_create(kthread_fn, NULL, "hwlatd"); if (IS_ERR(kthread)) { pr_err(BANNER "could not start sampling thread\n"); return -ENOMEM; } cpumask_clear(current_mask); cpumask_set_cpu(next_cpu, current_mask); sched_setaffinity(kthread->pid, current_mask); hwlat_kthread = kthread; wake_up_process(kthread); return 0; } /** * stop_kthread - Inform the hardware latency samping/detector kthread to stop * * This kicks the running hardware latency sampling/detector kernel thread and * tells it to stop sampling now. Use this on unload and at system shutdown. */ static void stop_kthread(void) { if (!hwlat_kthread) return; kthread_stop(hwlat_kthread); hwlat_kthread = NULL; } /* * hwlat_read - Wrapper read function for reading both window and width * @filp: The active open file structure * @ubuf: The userspace provided buffer to read value into * @cnt: The maximum number of bytes to read * @ppos: The current "file" position * * This function provides a generic read implementation for the global state * "hwlat_data" structure filesystem entries. */ static ssize_t hwlat_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) { char buf[U64STR_SIZE]; u64 *entry = filp->private_data; u64 val; int len; if (!entry) return -EFAULT; if (cnt > sizeof(buf)) cnt = sizeof(buf); val = *entry; len = snprintf(buf, sizeof(buf), "%llu\n", val); return simple_read_from_buffer(ubuf, cnt, ppos, buf, len); } /** * hwlat_width_write - Write function for "width" entry * @filp: The active open file structure * @ubuf: The user buffer that contains the value to write * @cnt: The maximum number of bytes to write to "file" * @ppos: The current position in @file * * This function provides a write implementation for the "width" interface * to the hardware latency detector. It can be used to configure * for how many us of the total window us we will actively sample for any * hardware-induced latency periods. Obviously, it is not possible to * sample constantly and have the system respond to a sample reader, or, * worse, without having the system appear to have gone out to lunch. It * is enforced that width is less that the total window size. */ static ssize_t hwlat_width_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { u64 val; int err; err = kstrtoull_from_user(ubuf, cnt, 10, &val); if (err) return err; mutex_lock(&hwlat_data.lock); if (val < hwlat_data.sample_window) hwlat_data.sample_width = val; else err = -EINVAL; mutex_unlock(&hwlat_data.lock); if (err) return err; return cnt; } /** * hwlat_window_write - Write function for "window" entry * @filp: The active open file structure * @ubuf: The user buffer that contains the value to write * @cnt: The maximum number of bytes to write to "file" * @ppos: The current position in @file * * This function provides a write implementation for the "window" interface * to the hardware latency detetector. The window is the total time * in us that will be considered one sample period. Conceptually, windows * occur back-to-back and contain a sample width period during which * actual sampling occurs. Can be used to write a new total window size. It * is enfoced that any value written must be greater than the sample width * size, or an error results. */ static ssize_t hwlat_window_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { u64 val; int err; err = kstrtoull_from_user(ubuf, cnt, 10, &val); if (err) return err; mutex_lock(&hwlat_data.lock); if (hwlat_data.sample_width < val) hwlat_data.sample_window = val; else err = -EINVAL; mutex_unlock(&hwlat_data.lock); if (err) return err; return cnt; } static const struct file_operations width_fops = { .open = tracing_open_generic, .read = hwlat_read, .write = hwlat_width_write, }; static const struct file_operations window_fops = { .open = tracing_open_generic, .read = hwlat_read, .write = hwlat_window_write, }; /** * init_tracefs - A function to initialize the tracefs interface files * * This function creates entries in tracefs for "hwlat_detector". * It creates the hwlat_detector directory in the tracing directory, * and within that directory is the count, width and window files to * change and view those values. */ static int init_tracefs(void) { struct dentry *d_tracer; struct dentry *top_dir; d_tracer = tracing_init_dentry(); if (IS_ERR(d_tracer)) return -ENOMEM; top_dir = tracefs_create_dir("hwlat_detector", d_tracer); if (!top_dir) return -ENOMEM; hwlat_sample_window = tracefs_create_file("window", 0640, top_dir, &hwlat_data.sample_window, &window_fops); if (!hwlat_sample_window) goto err; hwlat_sample_width = tracefs_create_file("width", 0644, top_dir, &hwlat_data.sample_width, &width_fops); if (!hwlat_sample_width) goto err; return 0; err: tracefs_remove_recursive(top_dir); return -ENOMEM; } static void hwlat_tracer_start(struct trace_array *tr) { int err; err = start_kthread(tr); if (err) pr_err(BANNER "Cannot start hwlat kthread\n"); } static void hwlat_tracer_stop(struct trace_array *tr) { stop_kthread(); } static bool hwlat_busy; static int hwlat_tracer_init(struct trace_array *tr) { /* Only allow one instance to enable this */ if (hwlat_busy) return -EBUSY; hwlat_trace = tr; disable_migrate = false; hwlat_data.count = 0; tr->max_latency = 0; save_tracing_thresh = tracing_thresh; /* tracing_thresh is in nsecs, we speak in usecs */ if (!tracing_thresh) tracing_thresh = last_tracing_thresh; if (tracer_tracing_is_on(tr)) hwlat_tracer_start(tr); hwlat_busy = true; return 0; } static void hwlat_tracer_reset(struct trace_array *tr) { stop_kthread(); /* the tracing threshold is static between runs */ last_tracing_thresh = tracing_thresh; tracing_thresh = save_tracing_thresh; hwlat_busy = false; } static struct tracer hwlat_tracer __read_mostly = { .name = "hwlat", .init = hwlat_tracer_init, .reset = hwlat_tracer_reset, .start = hwlat_tracer_start, .stop = hwlat_tracer_stop, .allow_instances = true, }; __init static int init_hwlat_tracer(void) { int ret; mutex_init(&hwlat_data.lock); ret = register_tracer(&hwlat_tracer); if (ret) return ret; init_tracefs(); return 0; } late_initcall(init_hwlat_tracer);