~ [ source navigation ] ~ [ diff markup ] ~ [ identifier search ] ~

TOMOYO Linux Cross Reference
Linux/kernel/trace/trace_boot.c

Version: ~ [ linux-6.11.5 ] ~ [ linux-6.10.14 ] ~ [ linux-6.9.12 ] ~ [ linux-6.8.12 ] ~ [ linux-6.7.12 ] ~ [ linux-6.6.58 ] ~ [ linux-6.5.13 ] ~ [ linux-6.4.16 ] ~ [ linux-6.3.13 ] ~ [ linux-6.2.16 ] ~ [ linux-6.1.114 ] ~ [ linux-6.0.19 ] ~ [ linux-5.19.17 ] ~ [ linux-5.18.19 ] ~ [ linux-5.17.15 ] ~ [ linux-5.16.20 ] ~ [ linux-5.15.169 ] ~ [ linux-5.14.21 ] ~ [ linux-5.13.19 ] ~ [ linux-5.12.19 ] ~ [ linux-5.11.22 ] ~ [ linux-5.10.228 ] ~ [ linux-5.9.16 ] ~ [ linux-5.8.18 ] ~ [ linux-5.7.19 ] ~ [ linux-5.6.19 ] ~ [ linux-5.5.19 ] ~ [ linux-5.4.284 ] ~ [ linux-5.3.18 ] ~ [ linux-5.2.21 ] ~ [ linux-5.1.21 ] ~ [ linux-5.0.21 ] ~ [ linux-4.20.17 ] ~ [ linux-4.19.322 ] ~ [ linux-4.18.20 ] ~ [ linux-4.17.19 ] ~ [ linux-4.16.18 ] ~ [ linux-4.15.18 ] ~ [ linux-4.14.336 ] ~ [ linux-4.13.16 ] ~ [ linux-4.12.14 ] ~ [ linux-4.11.12 ] ~ [ linux-4.10.17 ] ~ [ linux-4.9.337 ] ~ [ linux-4.4.302 ] ~ [ linux-3.10.108 ] ~ [ linux-2.6.32.71 ] ~ [ linux-2.6.0 ] ~ [ linux-2.4.37.11 ] ~ [ unix-v6-master ] ~ [ ccs-tools-1.8.9 ] ~ [ policy-sample ] ~
Architecture: ~ [ i386 ] ~ [ alpha ] ~ [ m68k ] ~ [ mips ] ~ [ ppc ] ~ [ sparc ] ~ [ sparc64 ] ~

  1 // SPDX-License-Identifier: GPL-2.0
  2 /*
  3  * trace_boot.c
  4  * Tracing kernel boot-time
  5  */
  6 
  7 #define pr_fmt(fmt)     "trace_boot: " fmt
  8 
  9 #include <linux/bootconfig.h>
 10 #include <linux/cpumask.h>
 11 #include <linux/ftrace.h>
 12 #include <linux/init.h>
 13 #include <linux/kernel.h>
 14 #include <linux/mutex.h>
 15 #include <linux/string.h>
 16 #include <linux/slab.h>
 17 #include <linux/trace.h>
 18 #include <linux/trace_events.h>
 19 
 20 #include "trace.h"
 21 
 22 #define MAX_BUF_LEN 256
 23 
 24 static void __init
 25 trace_boot_set_instance_options(struct trace_array *tr, struct xbc_node *node)
 26 {
 27         struct xbc_node *anode;
 28         const char *p;
 29         char buf[MAX_BUF_LEN];
 30         unsigned long v = 0;
 31 
 32         /* Common ftrace options */
 33         xbc_node_for_each_array_value(node, "options", anode, p) {
 34                 if (strscpy(buf, p, ARRAY_SIZE(buf)) < 0) {
 35                         pr_err("String is too long: %s\n", p);
 36                         continue;
 37                 }
 38 
 39                 if (trace_set_options(tr, buf) < 0)
 40                         pr_err("Failed to set option: %s\n", buf);
 41         }
 42 
 43         p = xbc_node_find_value(node, "tracing_on", NULL);
 44         if (p && *p != '\0') {
 45                 if (kstrtoul(p, 10, &v))
 46                         pr_err("Failed to set tracing on: %s\n", p);
 47                 if (v)
 48                         tracer_tracing_on(tr);
 49                 else
 50                         tracer_tracing_off(tr);
 51         }
 52 
 53         p = xbc_node_find_value(node, "trace_clock", NULL);
 54         if (p && *p != '\0') {
 55                 if (tracing_set_clock(tr, p) < 0)
 56                         pr_err("Failed to set trace clock: %s\n", p);
 57         }
 58 
 59         p = xbc_node_find_value(node, "buffer_size", NULL);
 60         if (p && *p != '\0') {
 61                 v = memparse(p, NULL);
 62                 if (v < PAGE_SIZE)
 63                         pr_err("Buffer size is too small: %s\n", p);
 64                 if (tracing_resize_ring_buffer(tr, v, RING_BUFFER_ALL_CPUS) < 0)
 65                         pr_err("Failed to resize trace buffer to %s\n", p);
 66         }
 67 
 68         p = xbc_node_find_value(node, "cpumask", NULL);
 69         if (p && *p != '\0') {
 70                 cpumask_var_t new_mask;
 71 
 72                 if (alloc_cpumask_var(&new_mask, GFP_KERNEL)) {
 73                         if (cpumask_parse(p, new_mask) < 0 ||
 74                             tracing_set_cpumask(tr, new_mask) < 0)
 75                                 pr_err("Failed to set new CPU mask %s\n", p);
 76                         free_cpumask_var(new_mask);
 77                 }
 78         }
 79 }
 80 
 81 #ifdef CONFIG_EVENT_TRACING
 82 static void __init
 83 trace_boot_enable_events(struct trace_array *tr, struct xbc_node *node)
 84 {
 85         struct xbc_node *anode;
 86         char buf[MAX_BUF_LEN];
 87         const char *p;
 88 
 89         xbc_node_for_each_array_value(node, "events", anode, p) {
 90                 if (strscpy(buf, p, ARRAY_SIZE(buf)) < 0) {
 91                         pr_err("String is too long: %s\n", p);
 92                         continue;
 93                 }
 94 
 95                 if (ftrace_set_clr_event(tr, buf, 1) < 0)
 96                         pr_err("Failed to enable event: %s\n", p);
 97         }
 98 }
 99 
100 #ifdef CONFIG_KPROBE_EVENTS
101 static int __init
102 trace_boot_add_kprobe_event(struct xbc_node *node, const char *event)
103 {
104         struct dynevent_cmd cmd;
105         struct xbc_node *anode;
106         char buf[MAX_BUF_LEN];
107         const char *val;
108         int ret = 0;
109 
110         xbc_node_for_each_array_value(node, "probes", anode, val) {
111                 kprobe_event_cmd_init(&cmd, buf, MAX_BUF_LEN);
112 
113                 ret = kprobe_event_gen_cmd_start(&cmd, event, val);
114                 if (ret) {
115                         pr_err("Failed to generate probe: %s\n", buf);
116                         break;
117                 }
118 
119                 ret = kprobe_event_gen_cmd_end(&cmd);
120                 if (ret) {
121                         pr_err("Failed to add probe: %s\n", buf);
122                         break;
123                 }
124         }
125 
126         return ret;
127 }
128 #else
129 static inline int __init
130 trace_boot_add_kprobe_event(struct xbc_node *node, const char *event)
131 {
132         pr_err("Kprobe event is not supported.\n");
133         return -ENOTSUPP;
134 }
135 #endif
136 
137 #ifdef CONFIG_SYNTH_EVENTS
138 static int __init
139 trace_boot_add_synth_event(struct xbc_node *node, const char *event)
140 {
141         struct dynevent_cmd cmd;
142         struct xbc_node *anode;
143         char buf[MAX_BUF_LEN];
144         const char *p;
145         int ret;
146 
147         synth_event_cmd_init(&cmd, buf, MAX_BUF_LEN);
148 
149         ret = synth_event_gen_cmd_start(&cmd, event, NULL);
150         if (ret)
151                 return ret;
152 
153         xbc_node_for_each_array_value(node, "fields", anode, p) {
154                 ret = synth_event_add_field_str(&cmd, p);
155                 if (ret)
156                         return ret;
157         }
158 
159         ret = synth_event_gen_cmd_end(&cmd);
160         if (ret < 0)
161                 pr_err("Failed to add synthetic event: %s\n", buf);
162 
163         return ret;
164 }
165 #else
166 static inline int __init
167 trace_boot_add_synth_event(struct xbc_node *node, const char *event)
168 {
169         pr_err("Synthetic event is not supported.\n");
170         return -ENOTSUPP;
171 }
172 #endif
173 
174 #ifdef CONFIG_HIST_TRIGGERS
175 static int __init __printf(3, 4)
176 append_printf(char **bufp, char *end, const char *fmt, ...)
177 {
178         va_list args;
179         int ret;
180 
181         if (*bufp == end)
182                 return -ENOSPC;
183 
184         va_start(args, fmt);
185         ret = vsnprintf(*bufp, end - *bufp, fmt, args);
186         if (ret < end - *bufp) {
187                 *bufp += ret;
188         } else {
189                 *bufp = end;
190                 ret = -ERANGE;
191         }
192         va_end(args);
193 
194         return ret;
195 }
196 
197 static int __init
198 append_str_nospace(char **bufp, char *end, const char *str)
199 {
200         char *p = *bufp;
201         int len;
202 
203         while (p < end - 1 && *str != '\0') {
204                 if (!isspace(*str))
205                         *(p++) = *str;
206                 str++;
207         }
208         *p = '\0';
209         if (p == end - 1) {
210                 *bufp = end;
211                 return -ENOSPC;
212         }
213         len = p - *bufp;
214         *bufp = p;
215         return (int)len;
216 }
217 
218 static int __init
219 trace_boot_hist_add_array(struct xbc_node *hnode, char **bufp,
220                           char *end, const char *key)
221 {
222         struct xbc_node *anode;
223         const char *p;
224         char sep;
225 
226         p = xbc_node_find_value(hnode, key, &anode);
227         if (p) {
228                 if (!anode) {
229                         pr_err("hist.%s requires value(s).\n", key);
230                         return -EINVAL;
231                 }
232 
233                 append_printf(bufp, end, ":%s", key);
234                 sep = '=';
235                 xbc_array_for_each_value(anode, p) {
236                         append_printf(bufp, end, "%c%s", sep, p);
237                         if (sep == '=')
238                                 sep = ',';
239                 }
240         } else
241                 return -ENOENT;
242 
243         return 0;
244 }
245 
246 static int __init
247 trace_boot_hist_add_one_handler(struct xbc_node *hnode, char **bufp,
248                                 char *end, const char *handler,
249                                 const char *param)
250 {
251         struct xbc_node *knode, *anode;
252         const char *p;
253         char sep;
254 
255         /* Compose 'handler' parameter */
256         p = xbc_node_find_value(hnode, param, NULL);
257         if (!p) {
258                 pr_err("hist.%s requires '%s' option.\n",
259                        xbc_node_get_data(hnode), param);
260                 return -EINVAL;
261         }
262         append_printf(bufp, end, ":%s(%s)", handler, p);
263 
264         /* Compose 'action' parameter */
265         knode = xbc_node_find_subkey(hnode, "trace");
266         if (!knode)
267                 knode = xbc_node_find_subkey(hnode, "save");
268 
269         if (knode) {
270                 anode = xbc_node_get_child(knode);
271                 if (!anode || !xbc_node_is_value(anode)) {
272                         pr_err("hist.%s.%s requires value(s).\n",
273                                xbc_node_get_data(hnode),
274                                xbc_node_get_data(knode));
275                         return -EINVAL;
276                 }
277 
278                 append_printf(bufp, end, ".%s", xbc_node_get_data(knode));
279                 sep = '(';
280                 xbc_array_for_each_value(anode, p) {
281                         append_printf(bufp, end, "%c%s", sep, p);
282                         if (sep == '(')
283                                 sep = ',';
284                 }
285                 append_printf(bufp, end, ")");
286         } else if (xbc_node_find_subkey(hnode, "snapshot")) {
287                 append_printf(bufp, end, ".snapshot()");
288         } else {
289                 pr_err("hist.%s requires an action.\n",
290                        xbc_node_get_data(hnode));
291                 return -EINVAL;
292         }
293 
294         return 0;
295 }
296 
297 static int __init
298 trace_boot_hist_add_handlers(struct xbc_node *hnode, char **bufp,
299                              char *end, const char *param)
300 {
301         struct xbc_node *node;
302         const char *p, *handler;
303         int ret = 0;
304 
305         handler = xbc_node_get_data(hnode);
306 
307         xbc_node_for_each_subkey(hnode, node) {
308                 p = xbc_node_get_data(node);
309                 if (!isdigit(p[0]))
310                         continue;
311                 /* All digit started node should be instances. */
312                 ret = trace_boot_hist_add_one_handler(node, bufp, end, handler, param);
313                 if (ret < 0)
314                         break;
315         }
316 
317         if (xbc_node_find_subkey(hnode, param))
318                 ret = trace_boot_hist_add_one_handler(hnode, bufp, end, handler, param);
319 
320         return ret;
321 }
322 
323 /*
324  * Histogram boottime tracing syntax.
325  *
326  * ftrace.[instance.INSTANCE.]event.GROUP.EVENT.hist[.N] {
327  *      keys = <KEY>[,...]
328  *      values = <VAL>[,...]
329  *      sort = <SORT-KEY>[,...]
330  *      size = <ENTRIES>
331  *      name = <HISTNAME>
332  *      var { <VAR> = <EXPR> ... }
333  *      pause|continue|clear
334  *      onmax|onchange[.N] { var = <VAR>; <ACTION> [= <PARAM>] }
335  *      onmatch[.N] { event = <EVENT>; <ACTION> [= <PARAM>] }
336  *      filter = <FILTER>
337  * }
338  *
339  * Where <ACTION> are;
340  *
341  *      trace = <EVENT>, <ARG1>[, ...]
342  *      save = <ARG1>[, ...]
343  *      snapshot
344  */
345 static int __init
346 trace_boot_compose_hist_cmd(struct xbc_node *hnode, char *buf, size_t size)
347 {
348         struct xbc_node *node, *knode;
349         char *end = buf + size;
350         const char *p;
351         int ret = 0;
352 
353         append_printf(&buf, end, "hist");
354 
355         ret = trace_boot_hist_add_array(hnode, &buf, end, "keys");
356         if (ret < 0) {
357                 if (ret == -ENOENT)
358                         pr_err("hist requires keys.\n");
359                 return -EINVAL;
360         }
361 
362         ret = trace_boot_hist_add_array(hnode, &buf, end, "values");
363         if (ret == -EINVAL)
364                 return ret;
365         ret = trace_boot_hist_add_array(hnode, &buf, end, "sort");
366         if (ret == -EINVAL)
367                 return ret;
368 
369         p = xbc_node_find_value(hnode, "size", NULL);
370         if (p)
371                 append_printf(&buf, end, ":size=%s", p);
372 
373         p = xbc_node_find_value(hnode, "name", NULL);
374         if (p)
375                 append_printf(&buf, end, ":name=%s", p);
376 
377         node = xbc_node_find_subkey(hnode, "var");
378         if (node) {
379                 xbc_node_for_each_key_value(node, knode, p) {
380                         /* Expression must not include spaces. */
381                         append_printf(&buf, end, ":%s=",
382                                       xbc_node_get_data(knode));
383                         append_str_nospace(&buf, end, p);
384                 }
385         }
386 
387         /* Histogram control attributes (mutual exclusive) */
388         if (xbc_node_find_value(hnode, "pause", NULL))
389                 append_printf(&buf, end, ":pause");
390         else if (xbc_node_find_value(hnode, "continue", NULL))
391                 append_printf(&buf, end, ":continue");
392         else if (xbc_node_find_value(hnode, "clear", NULL))
393                 append_printf(&buf, end, ":clear");
394 
395         /* Histogram handler and actions */
396         node = xbc_node_find_subkey(hnode, "onmax");
397         if (node && trace_boot_hist_add_handlers(node, &buf, end, "var") < 0)
398                 return -EINVAL;
399         node = xbc_node_find_subkey(hnode, "onchange");
400         if (node && trace_boot_hist_add_handlers(node, &buf, end, "var") < 0)
401                 return -EINVAL;
402         node = xbc_node_find_subkey(hnode, "onmatch");
403         if (node && trace_boot_hist_add_handlers(node, &buf, end, "event") < 0)
404                 return -EINVAL;
405 
406         p = xbc_node_find_value(hnode, "filter", NULL);
407         if (p)
408                 append_printf(&buf, end, " if %s", p);
409 
410         if (buf == end) {
411                 pr_err("hist exceeds the max command length.\n");
412                 return -E2BIG;
413         }
414 
415         return 0;
416 }
417 
418 static void __init
419 trace_boot_init_histograms(struct trace_event_file *file,
420                            struct xbc_node *hnode, char *buf, size_t size)
421 {
422         struct xbc_node *node;
423         const char *p;
424         char *tmp;
425 
426         xbc_node_for_each_subkey(hnode, node) {
427                 p = xbc_node_get_data(node);
428                 if (!isdigit(p[0]))
429                         continue;
430                 /* All digit started node should be instances. */
431                 if (trace_boot_compose_hist_cmd(node, buf, size) == 0) {
432                         tmp = kstrdup(buf, GFP_KERNEL);
433                         if (!tmp)
434                                 return;
435                         if (trigger_process_regex(file, buf) < 0)
436                                 pr_err("Failed to apply hist trigger: %s\n", tmp);
437                         kfree(tmp);
438                 }
439         }
440 
441         if (xbc_node_find_subkey(hnode, "keys")) {
442                 if (trace_boot_compose_hist_cmd(hnode, buf, size) == 0) {
443                         tmp = kstrdup(buf, GFP_KERNEL);
444                         if (!tmp)
445                                 return;
446                         if (trigger_process_regex(file, buf) < 0)
447                                 pr_err("Failed to apply hist trigger: %s\n", tmp);
448                         kfree(tmp);
449                 }
450         }
451 }
452 #else
453 static void __init
454 trace_boot_init_histograms(struct trace_event_file *file,
455                            struct xbc_node *hnode, char *buf, size_t size)
456 {
457         /* do nothing */
458 }
459 #endif
460 
461 static void __init
462 trace_boot_init_one_event(struct trace_array *tr, struct xbc_node *gnode,
463                           struct xbc_node *enode)
464 {
465         struct trace_event_file *file;
466         struct xbc_node *anode;
467         char buf[MAX_BUF_LEN];
468         const char *p, *group, *event;
469 
470         group = xbc_node_get_data(gnode);
471         event = xbc_node_get_data(enode);
472 
473         if (!strcmp(group, "kprobes"))
474                 if (trace_boot_add_kprobe_event(enode, event) < 0)
475                         return;
476         if (!strcmp(group, "synthetic"))
477                 if (trace_boot_add_synth_event(enode, event) < 0)
478                         return;
479 
480         mutex_lock(&event_mutex);
481         file = find_event_file(tr, group, event);
482         if (!file) {
483                 pr_err("Failed to find event: %s:%s\n", group, event);
484                 goto out;
485         }
486 
487         p = xbc_node_find_value(enode, "filter", NULL);
488         if (p && *p != '\0') {
489                 if (strscpy(buf, p, ARRAY_SIZE(buf)) < 0)
490                         pr_err("filter string is too long: %s\n", p);
491                 else if (apply_event_filter(file, buf) < 0)
492                         pr_err("Failed to apply filter: %s\n", buf);
493         }
494 
495         if (IS_ENABLED(CONFIG_HIST_TRIGGERS)) {
496                 xbc_node_for_each_array_value(enode, "actions", anode, p) {
497                         if (strscpy(buf, p, ARRAY_SIZE(buf)) < 0)
498                                 pr_err("action string is too long: %s\n", p);
499                         else if (trigger_process_regex(file, buf) < 0)
500                                 pr_err("Failed to apply an action: %s\n", p);
501                 }
502                 anode = xbc_node_find_subkey(enode, "hist");
503                 if (anode)
504                         trace_boot_init_histograms(file, anode, buf, ARRAY_SIZE(buf));
505         } else if (xbc_node_find_value(enode, "actions", NULL))
506                 pr_err("Failed to apply event actions because CONFIG_HIST_TRIGGERS is not set.\n");
507 
508         if (xbc_node_find_value(enode, "enable", NULL)) {
509                 if (trace_event_enable_disable(file, 1, 0) < 0)
510                         pr_err("Failed to enable event node: %s:%s\n",
511                                 group, event);
512         }
513 out:
514         mutex_unlock(&event_mutex);
515 }
516 
517 static void __init
518 trace_boot_init_events(struct trace_array *tr, struct xbc_node *node)
519 {
520         struct xbc_node *gnode, *enode;
521         bool enable, enable_all = false;
522         const char *data;
523 
524         node = xbc_node_find_subkey(node, "event");
525         if (!node)
526                 return;
527         /* per-event key starts with "event.GROUP.EVENT" */
528         xbc_node_for_each_subkey(node, gnode) {
529                 data = xbc_node_get_data(gnode);
530                 if (!strcmp(data, "enable")) {
531                         enable_all = true;
532                         continue;
533                 }
534                 enable = false;
535                 xbc_node_for_each_subkey(gnode, enode) {
536                         data = xbc_node_get_data(enode);
537                         if (!strcmp(data, "enable")) {
538                                 enable = true;
539                                 continue;
540                         }
541                         trace_boot_init_one_event(tr, gnode, enode);
542                 }
543                 /* Event enablement must be done after event settings */
544                 if (enable) {
545                         data = xbc_node_get_data(gnode);
546                         trace_array_set_clr_event(tr, data, NULL, true);
547                 }
548         }
549         /* Ditto */
550         if (enable_all)
551                 trace_array_set_clr_event(tr, NULL, NULL, true);
552 }
553 #else
554 #define trace_boot_enable_events(tr, node) do {} while (0)
555 #define trace_boot_init_events(tr, node) do {} while (0)
556 #endif
557 
558 #ifdef CONFIG_DYNAMIC_FTRACE
559 static void __init
560 trace_boot_set_ftrace_filter(struct trace_array *tr, struct xbc_node *node)
561 {
562         struct xbc_node *anode;
563         const char *p;
564         char *q;
565 
566         xbc_node_for_each_array_value(node, "ftrace.filters", anode, p) {
567                 q = kstrdup(p, GFP_KERNEL);
568                 if (!q)
569                         return;
570                 if (ftrace_set_filter(tr->ops, q, strlen(q), 0) < 0)
571                         pr_err("Failed to add %s to ftrace filter\n", p);
572                 else
573                         ftrace_filter_param = true;
574                 kfree(q);
575         }
576         xbc_node_for_each_array_value(node, "ftrace.notraces", anode, p) {
577                 q = kstrdup(p, GFP_KERNEL);
578                 if (!q)
579                         return;
580                 if (ftrace_set_notrace(tr->ops, q, strlen(q), 0) < 0)
581                         pr_err("Failed to add %s to ftrace filter\n", p);
582                 else
583                         ftrace_filter_param = true;
584                 kfree(q);
585         }
586 }
587 #else
588 #define trace_boot_set_ftrace_filter(tr, node) do {} while (0)
589 #endif
590 
591 static void __init
592 trace_boot_enable_tracer(struct trace_array *tr, struct xbc_node *node)
593 {
594         const char *p;
595 
596         trace_boot_set_ftrace_filter(tr, node);
597 
598         p = xbc_node_find_value(node, "tracer", NULL);
599         if (p && *p != '\0') {
600                 if (tracing_set_tracer(tr, p) < 0)
601                         pr_err("Failed to set given tracer: %s\n", p);
602         }
603 
604         /* Since tracer can free snapshot buffer, allocate snapshot here.*/
605         if (xbc_node_find_value(node, "alloc_snapshot", NULL)) {
606                 if (tracing_alloc_snapshot_instance(tr) < 0)
607                         pr_err("Failed to allocate snapshot buffer\n");
608         }
609 }
610 
611 static void __init
612 trace_boot_init_one_instance(struct trace_array *tr, struct xbc_node *node)
613 {
614         trace_boot_set_instance_options(tr, node);
615         trace_boot_init_events(tr, node);
616         trace_boot_enable_events(tr, node);
617         trace_boot_enable_tracer(tr, node);
618 }
619 
620 static void __init
621 trace_boot_init_instances(struct xbc_node *node)
622 {
623         struct xbc_node *inode;
624         struct trace_array *tr;
625         const char *p;
626 
627         node = xbc_node_find_subkey(node, "instance");
628         if (!node)
629                 return;
630 
631         xbc_node_for_each_subkey(node, inode) {
632                 p = xbc_node_get_data(inode);
633                 if (!p || *p == '\0')
634                         continue;
635 
636                 tr = trace_array_get_by_name(p, NULL);
637                 if (!tr) {
638                         pr_err("Failed to get trace instance %s\n", p);
639                         continue;
640                 }
641                 trace_boot_init_one_instance(tr, inode);
642                 trace_array_put(tr);
643         }
644 }
645 
646 static int __init trace_boot_init(void)
647 {
648         struct xbc_node *trace_node;
649         struct trace_array *tr;
650 
651         trace_node = xbc_find_node("ftrace");
652         if (!trace_node)
653                 return 0;
654 
655         tr = top_trace_array();
656         if (!tr)
657                 return 0;
658 
659         /* Global trace array is also one instance */
660         trace_boot_init_one_instance(tr, trace_node);
661         trace_boot_init_instances(trace_node);
662 
663         disable_tracing_selftest("running boot-time tracing");
664 
665         return 0;
666 }
667 /*
668  * Start tracing at the end of core-initcall, so that it starts tracing
669  * from the beginning of postcore_initcall.
670  */
671 core_initcall_sync(trace_boot_init);
672 

~ [ source navigation ] ~ [ diff markup ] ~ [ identifier search ] ~

kernel.org | git.kernel.org | LWN.net | Project Home | SVN repository | Mail admin

Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.

sflogo.php