2 * block queue tracing parse application
4 * Copyright (C) 2005 Jens Axboe <axboe@suse.de>
5 * Copyright (C) 2006 Jens Axboe <axboe@kernel.dk>
7 * This program is free software; you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation; either version 2 of the License, or
10 * (at your option) any later version.
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with this program; if not, write to the Free Software
19 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
22 #include <sys/types.h>
39 static char blkparse_version[] = "1.2.0";
42 unsigned long start, end;
43 struct skip_info *prev, *next;
51 unsigned long long events;
52 unsigned long long first_reported_time;
53 unsigned long long last_reported_time;
54 unsigned long long last_read_time;
55 struct io_stats io_stats;
57 unsigned long long seq_skips;
58 unsigned int max_depth[2];
59 unsigned int cur_depth[2];
61 struct rb_root rb_track;
66 unsigned long *cpu_map;
67 unsigned int cpu_map_max;
69 struct per_cpu_info *cpus;
73 * some duplicated effort here, we can unify this hash and the ppi hash later
75 struct process_pid_map {
78 struct process_pid_map *hash_next, *list_next;
81 #define PPM_HASH_SHIFT (8)
82 #define PPM_HASH_SIZE (1 << PPM_HASH_SHIFT)
83 #define PPM_HASH_MASK (PPM_HASH_SIZE - 1)
84 static struct process_pid_map *ppm_hash_table[PPM_HASH_SIZE];
86 struct per_process_info {
87 struct process_pid_map *ppm;
88 struct io_stats io_stats;
89 struct per_process_info *hash_next, *list_next;
95 unsigned long long longest_allocation_wait[2];
96 unsigned long long longest_dispatch_wait[2];
97 unsigned long long longest_completion_wait[2];
100 #define PPI_HASH_SHIFT (8)
101 #define PPI_HASH_SIZE (1 << PPI_HASH_SHIFT)
102 #define PPI_HASH_MASK (PPI_HASH_SIZE - 1)
103 static struct per_process_info *ppi_hash_table[PPI_HASH_SIZE];
104 static struct per_process_info *ppi_list;
105 static int ppi_list_entries;
107 static struct option l_opts[] = {
110 .has_arg = required_argument,
116 .has_arg = required_argument,
122 .has_arg = required_argument,
127 .name = "input-directory",
128 .has_arg = required_argument,
133 .name = "dump-binary",
134 .has_arg = required_argument,
140 .has_arg = required_argument,
145 .name = "format-spec",
146 .has_arg = required_argument,
151 .name = "hash-by-name",
152 .has_arg = no_argument,
158 .has_arg = required_argument,
164 .has_arg = no_argument,
170 .has_arg = required_argument,
175 .name = "no-text-output",
176 .has_arg = no_argument,
182 .has_arg = no_argument,
187 .name = "per-program-stats",
188 .has_arg = no_argument,
194 .has_arg = no_argument,
200 .has_arg = required_argument,
206 .has_arg = no_argument,
212 .has_arg = no_argument,
222 * for sorting the displayed output
225 struct blk_io_trace *bit;
226 struct rb_node rb_node;
228 unsigned long read_sequence;
231 static struct rb_root rb_sort_root;
232 static unsigned long rb_sort_entries;
234 static struct trace *trace_list;
239 static struct blk_io_trace *bit_alloc_list;
240 static struct trace *t_alloc_list;
243 * for tracking individual ios
246 struct rb_node rb_node;
248 struct process_pid_map *ppm;
250 unsigned long long allocation_time;
251 unsigned long long queue_time;
252 unsigned long long dispatch_time;
253 unsigned long long completion_time;
257 static struct per_dev_info *devices;
258 static char *get_dev_name(struct per_dev_info *, char *, int);
259 static int trace_rb_insert_last(struct per_dev_info *, struct trace *);
262 static char *output_name;
263 static char *input_dir;
265 static unsigned long long genesis_time;
266 static unsigned long long last_allowed_time;
267 static unsigned long long stopwatch_start; /* start from zero by default */
268 static unsigned long long stopwatch_end = -1ULL; /* "infinity" */
269 static unsigned long read_sequence;
271 static int per_process_stats;
272 static int per_device_and_cpu_stats = 1;
273 static int track_ios;
274 static int ppi_hash_by_pid = 1;
276 static unsigned int act_mask = -1U;
277 static int stats_printed;
278 static int bin_output_msgs = 1;
279 int data_is_native = -1;
281 static FILE *dump_fp;
282 static char *dump_binary;
284 static unsigned int t_alloc_cache;
285 static unsigned int bit_alloc_cache;
287 #define RB_BATCH_DEFAULT (512)
288 static unsigned int rb_batch = RB_BATCH_DEFAULT;
291 static char *pipename;
293 static int text_output = 1;
295 #define is_done() (*(volatile int *)(&done))
296 static volatile int done;
298 struct timespec abs_start_time;
299 static unsigned long long start_timestamp;
301 static int have_drv_data = 0;
303 #define JHASH_RANDOM (0x3af5f2ee)
305 #define CPUS_PER_LONG (8 * sizeof(unsigned long))
306 #define CPU_IDX(cpu) ((cpu) / CPUS_PER_LONG)
307 #define CPU_BIT(cpu) ((cpu) & (CPUS_PER_LONG - 1))
309 static void output_binary(void *buf, int len)
312 size_t n = fwrite(buf, len, 1, dump_fp);
321 static void resize_cpu_info(struct per_dev_info *pdi, int cpu)
323 struct per_cpu_info *cpus = pdi->cpus;
324 int ncpus = pdi->ncpus;
325 int new_count = cpu + 1;
329 size = new_count * sizeof(struct per_cpu_info);
330 cpus = realloc(cpus, size);
333 fprintf(stderr, "Out of memory, CPU info for device %s (%d)\n",
334 get_dev_name(pdi, name, sizeof(name)), size);
338 new_start = (char *)cpus + (ncpus * sizeof(struct per_cpu_info));
339 new_space = (new_count - ncpus) * sizeof(struct per_cpu_info);
340 memset(new_start, 0, new_space);
342 pdi->ncpus = new_count;
345 for (new_count = 0; new_count < pdi->ncpus; new_count++) {
346 struct per_cpu_info *pci = &pdi->cpus[new_count];
350 memset(&pci->rb_last, 0, sizeof(pci->rb_last));
351 pci->rb_last_entries = 0;
352 pci->last_sequence = -1;
357 static struct per_cpu_info *get_cpu_info(struct per_dev_info *pdi, int cpu)
359 struct per_cpu_info *pci;
361 if (cpu >= pdi->ncpus)
362 resize_cpu_info(pdi, cpu);
364 pci = &pdi->cpus[cpu];
370 static int resize_devices(char *name)
372 int size = (ndevices + 1) * sizeof(struct per_dev_info);
374 devices = realloc(devices, size);
376 fprintf(stderr, "Out of memory, device %s (%d)\n", name, size);
379 memset(&devices[ndevices], 0, sizeof(struct per_dev_info));
380 devices[ndevices].name = name;
385 static struct per_dev_info *get_dev_info(dev_t dev)
387 struct per_dev_info *pdi;
390 for (i = 0; i < ndevices; i++) {
392 devices[i].dev = dev;
393 if (devices[i].dev == dev)
397 if (resize_devices(NULL))
400 pdi = &devices[ndevices - 1];
402 pdi->first_reported_time = 0;
403 pdi->last_read_time = 0;
408 static void insert_skip(struct per_cpu_info *pci, unsigned long start,
411 struct skip_info *sip;
413 for (sip = pci->skips_tail; sip != NULL; sip = sip->prev) {
414 if (end == (sip->start - 1)) {
417 } else if (start == (sip->end + 1)) {
423 sip = malloc(sizeof(struct skip_info));
426 sip->prev = sip->next = NULL;
427 if (pci->skips_tail == NULL)
428 pci->skips_head = pci->skips_tail = sip;
430 sip->prev = pci->skips_tail;
431 pci->skips_tail->next = sip;
432 pci->skips_tail = sip;
436 static void remove_sip(struct per_cpu_info *pci, struct skip_info *sip)
438 if (sip->prev == NULL) {
439 if (sip->next == NULL)
440 pci->skips_head = pci->skips_tail = NULL;
442 pci->skips_head = sip->next;
443 sip->next->prev = NULL;
445 } else if (sip->next == NULL) {
446 pci->skips_tail = sip->prev;
447 sip->prev->next = NULL;
449 sip->prev->next = sip->next;
450 sip->next->prev = sip->prev;
453 sip->prev = sip->next = NULL;
457 #define IN_SKIP(sip,seq) (((sip)->start <= (seq)) && ((seq) <= sip->end))
458 static int check_current_skips(struct per_cpu_info *pci, unsigned long seq)
460 struct skip_info *sip;
462 for (sip = pci->skips_tail; sip != NULL; sip = sip->prev) {
463 if (IN_SKIP(sip, seq)) {
464 if (sip->start == seq) {
466 remove_sip(pci, sip);
469 } else if (sip->end == seq)
473 insert_skip(pci, seq + 1, sip->end);
482 static void collect_pdi_skips(struct per_dev_info *pdi)
484 struct skip_info *sip;
490 for (cpu = 0; cpu < pdi->ncpus; cpu++) {
491 struct per_cpu_info *pci = &pdi->cpus[cpu];
493 for (sip = pci->skips_head; sip != NULL; sip = sip->next) {
495 pdi->seq_skips += (sip->end - sip->start + 1);
497 fprintf(stderr,"(%d,%d): skipping %lu -> %lu\n",
498 MAJOR(pdi->dev), MINOR(pdi->dev),
499 sip->start, sip->end);
504 static void cpu_mark_online(struct per_dev_info *pdi, unsigned int cpu)
506 if (cpu >= pdi->cpu_map_max || !pdi->cpu_map) {
507 int new_max = (cpu + CPUS_PER_LONG) & ~(CPUS_PER_LONG - 1);
508 unsigned long *map = malloc(new_max / sizeof(long));
510 memset(map, 0, new_max / sizeof(long));
513 memcpy(map, pdi->cpu_map, pdi->cpu_map_max / sizeof(long));
518 pdi->cpu_map_max = new_max;
521 pdi->cpu_map[CPU_IDX(cpu)] |= (1UL << CPU_BIT(cpu));
524 static inline void cpu_mark_offline(struct per_dev_info *pdi, int cpu)
526 pdi->cpu_map[CPU_IDX(cpu)] &= ~(1UL << CPU_BIT(cpu));
529 static inline int cpu_is_online(struct per_dev_info *pdi, int cpu)
531 return (pdi->cpu_map[CPU_IDX(cpu)] & (1UL << CPU_BIT(cpu))) != 0;
534 static inline int ppm_hash_pid(pid_t pid)
536 return jhash_1word(pid, JHASH_RANDOM) & PPM_HASH_MASK;
539 static struct process_pid_map *find_ppm(pid_t pid)
541 const int hash_idx = ppm_hash_pid(pid);
542 struct process_pid_map *ppm;
544 ppm = ppm_hash_table[hash_idx];
549 ppm = ppm->hash_next;
555 static struct process_pid_map *add_ppm_hash(pid_t pid, const char *name)
557 const int hash_idx = ppm_hash_pid(pid);
558 struct process_pid_map *ppm;
562 ppm = malloc(sizeof(*ppm));
563 memset(ppm, 0, sizeof(*ppm));
565 memset(ppm->comm, 0, sizeof(ppm->comm));
566 strncpy(ppm->comm, name, sizeof(ppm->comm));
567 ppm->comm[sizeof(ppm->comm) - 1] = '\0';
568 ppm->hash_next = ppm_hash_table[hash_idx];
569 ppm_hash_table[hash_idx] = ppm;
575 static void handle_notify(struct blk_io_trace *bit)
577 void *payload = (caddr_t) bit + sizeof(*bit);
580 switch (bit->action) {
582 add_ppm_hash(bit->pid, payload);
585 case BLK_TN_TIMESTAMP:
586 if (bit->pdu_len != sizeof(two32))
588 memcpy(two32, payload, sizeof(two32));
589 if (!data_is_native) {
590 two32[0] = be32_to_cpu(two32[0]);
591 two32[1] = be32_to_cpu(two32[1]);
593 start_timestamp = bit->time;
594 abs_start_time.tv_sec = two32[0];
595 abs_start_time.tv_nsec = two32[1];
596 if (abs_start_time.tv_nsec < 0) {
597 abs_start_time.tv_sec--;
598 abs_start_time.tv_nsec += 1000000000;
604 if (bit->pdu_len > 0) {
605 char msg[bit->pdu_len+1];
607 memcpy(msg, (char *)payload, bit->pdu_len);
608 msg[bit->pdu_len] = '\0';
611 "%3d,%-3d %2d %8s %5d.%09lu %5u %2s %3s %s\n",
612 MAJOR(bit->device), MINOR(bit->device),
613 bit->cpu, "0", (int) SECONDS(bit->time),
614 (unsigned long) NANO_SECONDS(bit->time),
620 /* Ignore unknown notify events */
625 char *find_process_name(pid_t pid)
627 struct process_pid_map *ppm = find_ppm(pid);
635 static inline int ppi_hash_pid(pid_t pid)
637 return jhash_1word(pid, JHASH_RANDOM) & PPI_HASH_MASK;
640 static inline int ppi_hash_name(const char *name)
642 return jhash(name, 16, JHASH_RANDOM) & PPI_HASH_MASK;
645 static inline int ppi_hash(struct per_process_info *ppi)
647 struct process_pid_map *ppm = ppi->ppm;
650 return ppi_hash_pid(ppm->pid);
652 return ppi_hash_name(ppm->comm);
655 static inline void add_ppi_to_hash(struct per_process_info *ppi)
657 const int hash_idx = ppi_hash(ppi);
659 ppi->hash_next = ppi_hash_table[hash_idx];
660 ppi_hash_table[hash_idx] = ppi;
663 static inline void add_ppi_to_list(struct per_process_info *ppi)
665 ppi->list_next = ppi_list;
670 static struct per_process_info *find_ppi_by_name(char *name)
672 const int hash_idx = ppi_hash_name(name);
673 struct per_process_info *ppi;
675 ppi = ppi_hash_table[hash_idx];
677 struct process_pid_map *ppm = ppi->ppm;
679 if (!strcmp(ppm->comm, name))
682 ppi = ppi->hash_next;
688 static struct per_process_info *find_ppi_by_pid(pid_t pid)
690 const int hash_idx = ppi_hash_pid(pid);
691 struct per_process_info *ppi;
693 ppi = ppi_hash_table[hash_idx];
695 struct process_pid_map *ppm = ppi->ppm;
700 ppi = ppi->hash_next;
706 static struct per_process_info *find_ppi(pid_t pid)
708 struct per_process_info *ppi;
712 return find_ppi_by_pid(pid);
714 name = find_process_name(pid);
718 ppi = find_ppi_by_name(name);
719 if (ppi && ppi->ppm->pid != pid)
720 ppi->more_than_one = 1;
726 * struct trace and blktrace allocation cache, we do potentially
727 * millions of mallocs for these structures while only using at most
728 * a few thousand at the time
730 static inline void t_free(struct trace *t)
732 if (t_alloc_cache < 1024) {
733 t->next = t_alloc_list;
740 static inline struct trace *t_alloc(void)
742 struct trace *t = t_alloc_list;
745 t_alloc_list = t->next;
750 return malloc(sizeof(*t));
753 static inline void bit_free(struct blk_io_trace *bit)
755 if (bit_alloc_cache < 1024 && !bit->pdu_len) {
757 * abuse a 64-bit field for a next pointer for the free item
759 bit->time = (__u64) (unsigned long) bit_alloc_list;
760 bit_alloc_list = (struct blk_io_trace *) bit;
766 static inline struct blk_io_trace *bit_alloc(void)
768 struct blk_io_trace *bit = bit_alloc_list;
771 bit_alloc_list = (struct blk_io_trace *) (unsigned long) \
777 return malloc(sizeof(*bit));
780 static inline void __put_trace_last(struct per_dev_info *pdi, struct trace *t)
782 struct per_cpu_info *pci = get_cpu_info(pdi, t->bit->cpu);
784 rb_erase(&t->rb_node, &pci->rb_last);
785 pci->rb_last_entries--;
791 static void put_trace(struct per_dev_info *pdi, struct trace *t)
793 rb_erase(&t->rb_node, &rb_sort_root);
796 trace_rb_insert_last(pdi, t);
799 static inline int trace_rb_insert(struct trace *t, struct rb_root *root)
801 struct rb_node **p = &root->rb_node;
802 struct rb_node *parent = NULL;
808 __t = rb_entry(parent, struct trace, rb_node);
810 if (t->bit->time < __t->bit->time)
812 else if (t->bit->time > __t->bit->time)
814 else if (t->bit->device < __t->bit->device)
816 else if (t->bit->device > __t->bit->device)
818 else if (t->bit->sequence < __t->bit->sequence)
820 else /* >= sequence */
824 rb_link_node(&t->rb_node, parent, p);
825 rb_insert_color(&t->rb_node, root);
829 static inline int trace_rb_insert_sort(struct trace *t)
831 if (!trace_rb_insert(t, &rb_sort_root)) {
839 static int trace_rb_insert_last(struct per_dev_info *pdi, struct trace *t)
841 struct per_cpu_info *pci = get_cpu_info(pdi, t->bit->cpu);
843 if (trace_rb_insert(t, &pci->rb_last))
846 pci->rb_last_entries++;
848 if (pci->rb_last_entries > rb_batch * pdi->nfiles) {
849 struct rb_node *n = rb_first(&pci->rb_last);
851 t = rb_entry(n, struct trace, rb_node);
852 __put_trace_last(pdi, t);
858 static struct trace *trace_rb_find(dev_t device, unsigned long sequence,
859 struct rb_root *root, int order)
861 struct rb_node *n = root->rb_node;
862 struct rb_node *prev = NULL;
866 __t = rb_entry(n, struct trace, rb_node);
869 if (device < __t->bit->device)
871 else if (device > __t->bit->device)
873 else if (sequence < __t->bit->sequence)
875 else if (sequence > __t->bit->sequence)
882 * hack - the list may not be sequence ordered because some
883 * events don't have sequence and time matched. so we end up
884 * being a little off in the rb lookup here, because we don't
885 * know the time we are looking for. compensate by browsing
886 * a little ahead from the last entry to find the match
891 while (((n = rb_next(prev)) != NULL) && max--) {
892 __t = rb_entry(n, struct trace, rb_node);
894 if (__t->bit->device == device &&
895 __t->bit->sequence == sequence)
905 static inline struct trace *trace_rb_find_last(struct per_dev_info *pdi,
906 struct per_cpu_info *pci,
909 return trace_rb_find(pdi->dev, seq, &pci->rb_last, 0);
912 static inline int track_rb_insert(struct per_dev_info *pdi,struct io_track *iot)
914 struct rb_node **p = &pdi->rb_track.rb_node;
915 struct rb_node *parent = NULL;
916 struct io_track *__iot;
920 __iot = rb_entry(parent, struct io_track, rb_node);
922 if (iot->sector < __iot->sector)
924 else if (iot->sector > __iot->sector)
928 "sector alias (%Lu) on device %d,%d!\n",
929 (unsigned long long) iot->sector,
930 MAJOR(pdi->dev), MINOR(pdi->dev));
935 rb_link_node(&iot->rb_node, parent, p);
936 rb_insert_color(&iot->rb_node, &pdi->rb_track);
940 static struct io_track *__find_track(struct per_dev_info *pdi, __u64 sector)
942 struct rb_node *n = pdi->rb_track.rb_node;
943 struct io_track *__iot;
946 __iot = rb_entry(n, struct io_track, rb_node);
948 if (sector < __iot->sector)
950 else if (sector > __iot->sector)
959 static struct io_track *find_track(struct per_dev_info *pdi, pid_t pid,
962 struct io_track *iot;
964 iot = __find_track(pdi, sector);
966 iot = malloc(sizeof(*iot));
967 iot->ppm = find_ppm(pid);
969 iot->ppm = add_ppm_hash(pid, "unknown");
970 iot->sector = sector;
971 track_rb_insert(pdi, iot);
977 static void log_track_frontmerge(struct per_dev_info *pdi,
978 struct blk_io_trace *t)
980 struct io_track *iot;
985 iot = __find_track(pdi, t->sector + t_sec(t));
988 fprintf(stderr, "merge not found for (%d,%d): %llu\n",
989 MAJOR(pdi->dev), MINOR(pdi->dev),
990 (unsigned long long) t->sector + t_sec(t));
994 rb_erase(&iot->rb_node, &pdi->rb_track);
995 iot->sector -= t_sec(t);
996 track_rb_insert(pdi, iot);
999 static void log_track_getrq(struct per_dev_info *pdi, struct blk_io_trace *t)
1001 struct io_track *iot;
1006 iot = find_track(pdi, t->pid, t->sector);
1007 iot->allocation_time = t->time;
1010 static inline int is_remapper(struct per_dev_info *pdi)
1012 int major = MAJOR(pdi->dev);
1014 return (major == 253 || major == 9);
1018 * for md/dm setups, the interesting cycle is Q -> C. So track queueing
1019 * time here, as dispatch time
1021 static void log_track_queue(struct per_dev_info *pdi, struct blk_io_trace *t)
1023 struct io_track *iot;
1027 if (!is_remapper(pdi))
1030 iot = find_track(pdi, t->pid, t->sector);
1031 iot->dispatch_time = t->time;
1035 * return time between rq allocation and insertion
1037 static unsigned long long log_track_insert(struct per_dev_info *pdi,
1038 struct blk_io_trace *t)
1040 unsigned long long elapsed;
1041 struct io_track *iot;
1046 iot = find_track(pdi, t->pid, t->sector);
1047 iot->queue_time = t->time;
1049 if (!iot->allocation_time)
1052 elapsed = iot->queue_time - iot->allocation_time;
1054 if (per_process_stats) {
1055 struct per_process_info *ppi = find_ppi(iot->ppm->pid);
1056 int w = (t->action & BLK_TC_ACT(BLK_TC_WRITE)) != 0;
1058 if (ppi && elapsed > ppi->longest_allocation_wait[w])
1059 ppi->longest_allocation_wait[w] = elapsed;
1066 * return time between queue and issue
1068 static unsigned long long log_track_issue(struct per_dev_info *pdi,
1069 struct blk_io_trace *t)
1071 unsigned long long elapsed;
1072 struct io_track *iot;
1076 if ((t->action & BLK_TC_ACT(BLK_TC_FS)) == 0)
1079 iot = __find_track(pdi, t->sector);
1082 fprintf(stderr, "issue not found for (%d,%d): %llu\n",
1083 MAJOR(pdi->dev), MINOR(pdi->dev),
1084 (unsigned long long) t->sector);
1088 iot->dispatch_time = t->time;
1089 elapsed = iot->dispatch_time - iot->queue_time;
1091 if (per_process_stats) {
1092 struct per_process_info *ppi = find_ppi(iot->ppm->pid);
1093 int w = (t->action & BLK_TC_ACT(BLK_TC_WRITE)) != 0;
1095 if (ppi && elapsed > ppi->longest_dispatch_wait[w])
1096 ppi->longest_dispatch_wait[w] = elapsed;
1103 * return time between dispatch and complete
1105 static unsigned long long log_track_complete(struct per_dev_info *pdi,
1106 struct blk_io_trace *t)
1108 unsigned long long elapsed;
1109 struct io_track *iot;
1114 iot = __find_track(pdi, t->sector);
1117 fprintf(stderr,"complete not found for (%d,%d): %llu\n",
1118 MAJOR(pdi->dev), MINOR(pdi->dev),
1119 (unsigned long long) t->sector);
1123 iot->completion_time = t->time;
1124 elapsed = iot->completion_time - iot->dispatch_time;
1126 if (per_process_stats) {
1127 struct per_process_info *ppi = find_ppi(iot->ppm->pid);
1128 int w = (t->action & BLK_TC_ACT(BLK_TC_WRITE)) != 0;
1130 if (ppi && elapsed > ppi->longest_completion_wait[w])
1131 ppi->longest_completion_wait[w] = elapsed;
1135 * kill the trace, we don't need it after completion
1137 rb_erase(&iot->rb_node, &pdi->rb_track);
1144 static struct io_stats *find_process_io_stats(pid_t pid)
1146 struct per_process_info *ppi = find_ppi(pid);
1149 ppi = malloc(sizeof(*ppi));
1150 memset(ppi, 0, sizeof(*ppi));
1151 ppi->ppm = find_ppm(pid);
1153 ppi->ppm = add_ppm_hash(pid, "unknown");
1154 add_ppi_to_hash(ppi);
1155 add_ppi_to_list(ppi);
1158 return &ppi->io_stats;
1161 static char *get_dev_name(struct per_dev_info *pdi, char *buffer, int size)
1164 snprintf(buffer, size, "%s", pdi->name);
1166 snprintf(buffer, size, "%d,%d",MAJOR(pdi->dev),MINOR(pdi->dev));
1170 static void check_time(struct per_dev_info *pdi, struct blk_io_trace *bit)
1172 unsigned long long this = bit->time;
1173 unsigned long long last = pdi->last_reported_time;
1175 pdi->backwards = (this < last) ? 'B' : ' ';
1176 pdi->last_reported_time = this;
1179 static inline void __account_m(struct io_stats *ios, struct blk_io_trace *t,
1184 ios->mwrite_kb += t_kb(t);
1185 ios->mwrite_b += t_b(t);
1188 ios->mread_kb += t_kb(t);
1189 ios->mread_b += t_b(t);
1193 static inline void account_m(struct blk_io_trace *t, struct per_cpu_info *pci,
1196 __account_m(&pci->io_stats, t, rw);
1198 if (per_process_stats) {
1199 struct io_stats *ios = find_process_io_stats(t->pid);
1201 __account_m(ios, t, rw);
1205 static inline void __account_pc_queue(struct io_stats *ios,
1206 struct blk_io_trace *t, int rw)
1210 ios->qwrite_kb_pc += t_kb(t);
1211 ios->qwrite_b_pc += t_b(t);
1214 ios->qread_kb += t_kb(t);
1215 ios->qread_b_pc += t_b(t);
1219 static inline void account_pc_queue(struct blk_io_trace *t,
1220 struct per_cpu_info *pci, int rw)
1222 __account_pc_queue(&pci->io_stats, t, rw);
1224 if (per_process_stats) {
1225 struct io_stats *ios = find_process_io_stats(t->pid);
1227 __account_pc_queue(ios, t, rw);
1231 static inline void __account_pc_issue(struct io_stats *ios, int rw,
1236 ios->iwrite_kb_pc += bytes >> 10;
1237 ios->iwrite_b_pc += bytes & 1023;
1240 ios->iread_kb_pc += bytes >> 10;
1241 ios->iread_b_pc += bytes & 1023;
1245 static inline void account_pc_issue(struct blk_io_trace *t,
1246 struct per_cpu_info *pci, int rw)
1248 __account_pc_issue(&pci->io_stats, rw, t->bytes);
1250 if (per_process_stats) {
1251 struct io_stats *ios = find_process_io_stats(t->pid);
1253 __account_pc_issue(ios, rw, t->bytes);
1257 static inline void __account_pc_requeue(struct io_stats *ios,
1258 struct blk_io_trace *t, int rw)
1262 ios->iwrite_kb_pc -= t_kb(t);
1263 ios->iwrite_b_pc -= t_b(t);
1266 ios->iread_kb_pc -= t_kb(t);
1267 ios->iread_b_pc -= t_b(t);
1271 static inline void account_pc_requeue(struct blk_io_trace *t,
1272 struct per_cpu_info *pci, int rw)
1274 __account_pc_requeue(&pci->io_stats, t, rw);
1276 if (per_process_stats) {
1277 struct io_stats *ios = find_process_io_stats(t->pid);
1279 __account_pc_requeue(ios, t, rw);
1283 static inline void __account_pc_c(struct io_stats *ios, int rw)
1291 static inline void account_pc_c(struct blk_io_trace *t,
1292 struct per_cpu_info *pci, int rw)
1294 __account_pc_c(&pci->io_stats, rw);
1296 if (per_process_stats) {
1297 struct io_stats *ios = find_process_io_stats(t->pid);
1299 __account_pc_c(ios, rw);
1303 static inline void __account_queue(struct io_stats *ios, struct blk_io_trace *t,
1308 ios->qwrite_kb += t_kb(t);
1309 ios->qwrite_b += t_b(t);
1312 ios->qread_kb += t_kb(t);
1313 ios->qread_b += t_b(t);
1317 static inline void account_queue(struct blk_io_trace *t,
1318 struct per_cpu_info *pci, int rw)
1320 __account_queue(&pci->io_stats, t, rw);
1322 if (per_process_stats) {
1323 struct io_stats *ios = find_process_io_stats(t->pid);
1325 __account_queue(ios, t, rw);
1329 static inline void __account_c(struct io_stats *ios, int rw, int bytes)
1333 ios->cwrite_kb += bytes >> 10;
1334 ios->cwrite_b += bytes & 1023;
1337 ios->cread_kb += bytes >> 10;
1338 ios->cread_b += bytes & 1023;
1342 static inline void account_c(struct blk_io_trace *t, struct per_cpu_info *pci,
1345 __account_c(&pci->io_stats, rw, bytes);
1347 if (per_process_stats) {
1348 struct io_stats *ios = find_process_io_stats(t->pid);
1350 __account_c(ios, rw, bytes);
1354 static inline void __account_issue(struct io_stats *ios, int rw,
1359 ios->iwrite_kb += bytes >> 10;
1360 ios->iwrite_b += bytes & 1023;
1363 ios->iread_kb += bytes >> 10;
1364 ios->iread_b += bytes & 1023;
1368 static inline void account_issue(struct blk_io_trace *t,
1369 struct per_cpu_info *pci, int rw)
1371 __account_issue(&pci->io_stats, rw, t->bytes);
1373 if (per_process_stats) {
1374 struct io_stats *ios = find_process_io_stats(t->pid);
1376 __account_issue(ios, rw, t->bytes);
1380 static inline void __account_unplug(struct io_stats *ios, int timer)
1383 ios->timer_unplugs++;
1388 static inline void account_unplug(struct blk_io_trace *t,
1389 struct per_cpu_info *pci, int timer)
1391 __account_unplug(&pci->io_stats, timer);
1393 if (per_process_stats) {
1394 struct io_stats *ios = find_process_io_stats(t->pid);
1396 __account_unplug(ios, timer);
1400 static inline void __account_requeue(struct io_stats *ios,
1401 struct blk_io_trace *t, int rw)
1405 ios->iwrite_kb -= t_kb(t);
1406 ios->iwrite_b -= t_b(t);
1409 ios->iread_kb -= t_kb(t);
1410 ios->iread_b -= t_b(t);
1414 static inline void account_requeue(struct blk_io_trace *t,
1415 struct per_cpu_info *pci, int rw)
1417 __account_requeue(&pci->io_stats, t, rw);
1419 if (per_process_stats) {
1420 struct io_stats *ios = find_process_io_stats(t->pid);
1422 __account_requeue(ios, t, rw);
1426 static void log_complete(struct per_dev_info *pdi, struct per_cpu_info *pci,
1427 struct blk_io_trace *t, char *act)
1429 process_fmt(act, pci, t, log_track_complete(pdi, t), 0, NULL);
1432 static void log_insert(struct per_dev_info *pdi, struct per_cpu_info *pci,
1433 struct blk_io_trace *t, char *act)
1435 process_fmt(act, pci, t, log_track_insert(pdi, t), 0, NULL);
1438 static void log_queue(struct per_cpu_info *pci, struct blk_io_trace *t,
1441 process_fmt(act, pci, t, -1, 0, NULL);
1444 static void log_issue(struct per_dev_info *pdi, struct per_cpu_info *pci,
1445 struct blk_io_trace *t, char *act)
1447 process_fmt(act, pci, t, log_track_issue(pdi, t), 0, NULL);
1450 static void log_merge(struct per_dev_info *pdi, struct per_cpu_info *pci,
1451 struct blk_io_trace *t, char *act)
1454 log_track_frontmerge(pdi, t);
1456 process_fmt(act, pci, t, -1ULL, 0, NULL);
1459 static void log_action(struct per_cpu_info *pci, struct blk_io_trace *t,
1462 process_fmt(act, pci, t, -1ULL, 0, NULL);
1465 static void log_generic(struct per_cpu_info *pci, struct blk_io_trace *t,
1468 process_fmt(act, pci, t, -1ULL, 0, NULL);
1471 static void log_unplug(struct per_cpu_info *pci, struct blk_io_trace *t,
1474 process_fmt(act, pci, t, -1ULL, 0, NULL);
1477 static void log_split(struct per_cpu_info *pci, struct blk_io_trace *t,
1480 process_fmt(act, pci, t, -1ULL, 0, NULL);
1483 static void log_pc(struct per_cpu_info *pci, struct blk_io_trace *t, char *act)
1485 unsigned char *buf = (unsigned char *) t + sizeof(*t);
1487 process_fmt(act, pci, t, -1ULL, t->pdu_len, buf);
1490 static void dump_trace_pc(struct blk_io_trace *t, struct per_dev_info *pdi,
1491 struct per_cpu_info *pci)
1493 int w = (t->action & BLK_TC_ACT(BLK_TC_WRITE)) != 0;
1494 int act = t->action & 0xffff;
1497 case __BLK_TA_QUEUE:
1498 log_generic(pci, t, "Q");
1499 account_pc_queue(t, pci, w);
1501 case __BLK_TA_GETRQ:
1502 log_generic(pci, t, "G");
1504 case __BLK_TA_SLEEPRQ:
1505 log_generic(pci, t, "S");
1507 case __BLK_TA_REQUEUE:
1509 * can happen if we miss traces, don't let it go
1512 if (pdi->cur_depth[w])
1513 pdi->cur_depth[w]--;
1514 account_pc_requeue(t, pci, w);
1515 log_generic(pci, t, "R");
1517 case __BLK_TA_ISSUE:
1518 account_pc_issue(t, pci, w);
1519 pdi->cur_depth[w]++;
1520 if (pdi->cur_depth[w] > pdi->max_depth[w])
1521 pdi->max_depth[w] = pdi->cur_depth[w];
1522 log_pc(pci, t, "D");
1524 case __BLK_TA_COMPLETE:
1525 if (pdi->cur_depth[w])
1526 pdi->cur_depth[w]--;
1527 log_pc(pci, t, "C");
1528 account_pc_c(t, pci, w);
1530 case __BLK_TA_INSERT:
1531 log_pc(pci, t, "I");
1534 fprintf(stderr, "Bad pc action %x\n", act);
1539 static void dump_trace_fs(struct blk_io_trace *t, struct per_dev_info *pdi,
1540 struct per_cpu_info *pci)
1542 int w = (t->action & BLK_TC_ACT(BLK_TC_WRITE)) != 0;
1543 int act = t->action & 0xffff;
1546 case __BLK_TA_QUEUE:
1547 log_track_queue(pdi, t);
1548 account_queue(t, pci, w);
1549 log_queue(pci, t, "Q");
1551 case __BLK_TA_INSERT:
1552 log_insert(pdi, pci, t, "I");
1554 case __BLK_TA_BACKMERGE:
1555 account_m(t, pci, w);
1556 log_merge(pdi, pci, t, "M");
1558 case __BLK_TA_FRONTMERGE:
1559 account_m(t, pci, w);
1560 log_merge(pdi, pci, t, "F");
1562 case __BLK_TA_GETRQ:
1563 log_track_getrq(pdi, t);
1564 log_generic(pci, t, "G");
1566 case __BLK_TA_SLEEPRQ:
1567 log_generic(pci, t, "S");
1569 case __BLK_TA_REQUEUE:
1571 * can happen if we miss traces, don't let it go
1574 if (pdi->cur_depth[w])
1575 pdi->cur_depth[w]--;
1576 account_requeue(t, pci, w);
1577 log_queue(pci, t, "R");
1579 case __BLK_TA_ISSUE:
1580 account_issue(t, pci, w);
1581 pdi->cur_depth[w]++;
1582 if (pdi->cur_depth[w] > pdi->max_depth[w])
1583 pdi->max_depth[w] = pdi->cur_depth[w];
1584 log_issue(pdi, pci, t, "D");
1586 case __BLK_TA_COMPLETE:
1587 if (pdi->cur_depth[w])
1588 pdi->cur_depth[w]--;
1589 account_c(t, pci, w, t->bytes);
1590 log_complete(pdi, pci, t, "C");
1593 log_action(pci, t, "P");
1595 case __BLK_TA_UNPLUG_IO:
1596 account_unplug(t, pci, 0);
1597 log_unplug(pci, t, "U");
1599 case __BLK_TA_UNPLUG_TIMER:
1600 account_unplug(t, pci, 1);
1601 log_unplug(pci, t, "UT");
1603 case __BLK_TA_SPLIT:
1604 log_split(pci, t, "X");
1606 case __BLK_TA_BOUNCE:
1607 log_generic(pci, t, "B");
1609 case __BLK_TA_REMAP:
1610 log_generic(pci, t, "A");
1612 case __BLK_TA_DRV_DATA:
1614 /* dump to binary file only */
1617 fprintf(stderr, "Bad fs action %x\n", t->action);
1622 static void dump_trace(struct blk_io_trace *t, struct per_cpu_info *pci,
1623 struct per_dev_info *pdi)
1626 if (t->action == BLK_TN_MESSAGE)
1628 else if (t->action & BLK_TC_ACT(BLK_TC_PC))
1629 dump_trace_pc(t, pdi, pci);
1631 dump_trace_fs(t, pdi, pci);
1635 pdi->first_reported_time = t->time;
1639 if (bin_output_msgs ||
1640 !(t->action & BLK_TC_ACT(BLK_TC_NOTIFY) &&
1641 t->action == BLK_TN_MESSAGE))
1642 output_binary(t, sizeof(*t) + t->pdu_len);
1646 * print in a proper way, not too small and not too big. if more than
1647 * 1000,000K, turn into M and so on
1649 static char *size_cnv(char *dst, unsigned long long num, int in_kb)
1651 char suff[] = { '\0', 'K', 'M', 'G', 'P' };
1657 while (num > 1000 * 1000ULL && (i < sizeof(suff) - 1)) {
1662 sprintf(dst, "%'8Lu%c", num, suff[i]);
1666 static void dump_io_stats(struct per_dev_info *pdi, struct io_stats *ios,
1669 static char x[256], y[256];
1671 fprintf(ofp, "%s\n", msg);
1673 fprintf(ofp, " Reads Queued: %s, %siB\t",
1674 size_cnv(x, ios->qreads, 0),
1675 size_cnv(y, ios->qread_kb + (ios->qread_b>>10), 1));
1676 fprintf(ofp, " Writes Queued: %s, %siB\n",
1677 size_cnv(x, ios->qwrites, 0),
1678 size_cnv(y, ios->qwrite_kb + (ios->qwrite_b>>10), 1));
1679 fprintf(ofp, " Read Dispatches: %s, %siB\t",
1680 size_cnv(x, ios->ireads, 0),
1681 size_cnv(y, ios->iread_kb + (ios->iread_b>>10), 1));
1682 fprintf(ofp, " Write Dispatches: %s, %siB\n",
1683 size_cnv(x, ios->iwrites, 0),
1684 size_cnv(y, ios->iwrite_kb + (ios->iwrite_b>>10), 1));
1685 fprintf(ofp, " Reads Requeued: %s\t\t", size_cnv(x, ios->rrqueue, 0));
1686 fprintf(ofp, " Writes Requeued: %s\n", size_cnv(x, ios->wrqueue, 0));
1687 fprintf(ofp, " Reads Completed: %s, %siB\t",
1688 size_cnv(x, ios->creads, 0),
1689 size_cnv(y, ios->cread_kb + (ios->cread_b>>10), 1));
1690 fprintf(ofp, " Writes Completed: %s, %siB\n",
1691 size_cnv(x, ios->cwrites, 0),
1692 size_cnv(y, ios->cwrite_kb + (ios->cwrite_b>>10), 1));
1693 fprintf(ofp, " Read Merges: %s, %siB\t",
1694 size_cnv(x, ios->mreads, 0),
1695 size_cnv(y, ios->mread_kb + (ios->mread_b>>10), 1));
1696 fprintf(ofp, " Write Merges: %s, %siB\n",
1697 size_cnv(x, ios->mwrites, 0),
1698 size_cnv(y, ios->mwrite_kb + (ios->mwrite_b>>10), 1));
1700 fprintf(ofp, " Read depth: %'8u%8c\t", pdi->max_depth[0], ' ');
1701 fprintf(ofp, " Write depth: %'8u\n", pdi->max_depth[1]);
1703 if (ios->qreads_pc || ios->qwrites_pc || ios->ireads_pc || ios->iwrites_pc ||
1704 ios->rrqueue_pc || ios->wrqueue_pc || ios->creads_pc || ios->cwrites_pc) {
1705 fprintf(ofp, " PC Reads Queued: %s, %siB\t",
1706 size_cnv(x, ios->qreads_pc, 0),
1708 ios->qread_kb_pc + (ios->qread_b_pc>>10), 1));
1709 fprintf(ofp, " PC Writes Queued: %s, %siB\n",
1710 size_cnv(x, ios->qwrites_pc, 0),
1712 ios->qwrite_kb_pc + (ios->qwrite_b_pc>>10), 1));
1713 fprintf(ofp, " PC Read Disp.: %s, %siB\t",
1714 size_cnv(x, ios->ireads_pc, 0),
1716 ios->iread_kb_pc + (ios->iread_b_pc>>10), 1));
1717 fprintf(ofp, " PC Write Disp.: %s, %siB\n",
1718 size_cnv(x, ios->iwrites_pc, 0),
1720 ios->iwrite_kb_pc + (ios->iwrite_b_pc>>10),
1722 fprintf(ofp, " PC Reads Req.: %s\t\t", size_cnv(x, ios->rrqueue_pc, 0));
1723 fprintf(ofp, " PC Writes Req.: %s\n", size_cnv(x, ios->wrqueue_pc, 0));
1724 fprintf(ofp, " PC Reads Compl.: %s\t\t", size_cnv(x, ios->creads_pc, 0));
1725 fprintf(ofp, " PC Writes Compl.: %s\n", size_cnv(x, ios->cwrites_pc, 0));
1727 fprintf(ofp, " IO unplugs: %'8lu%8c\t", ios->io_unplugs, ' ');
1728 fprintf(ofp, " Timer unplugs: %'8lu\n", ios->timer_unplugs);
1731 static void dump_wait_stats(struct per_process_info *ppi)
1733 unsigned long rawait = ppi->longest_allocation_wait[0] / 1000;
1734 unsigned long rdwait = ppi->longest_dispatch_wait[0] / 1000;
1735 unsigned long rcwait = ppi->longest_completion_wait[0] / 1000;
1736 unsigned long wawait = ppi->longest_allocation_wait[1] / 1000;
1737 unsigned long wdwait = ppi->longest_dispatch_wait[1] / 1000;
1738 unsigned long wcwait = ppi->longest_completion_wait[1] / 1000;
1740 fprintf(ofp, " Allocation wait: %'8lu%8c\t", rawait, ' ');
1741 fprintf(ofp, " Allocation wait: %'8lu\n", wawait);
1742 fprintf(ofp, " Dispatch wait: %'8lu%8c\t", rdwait, ' ');
1743 fprintf(ofp, " Dispatch wait: %'8lu\n", wdwait);
1744 fprintf(ofp, " Completion wait: %'8lu%8c\t", rcwait, ' ');
1745 fprintf(ofp, " Completion wait: %'8lu\n", wcwait);
1748 static int ppi_name_compare(const void *p1, const void *p2)
1750 struct per_process_info *ppi1 = *((struct per_process_info **) p1);
1751 struct per_process_info *ppi2 = *((struct per_process_info **) p2);
1754 res = strverscmp(ppi1->ppm->comm, ppi2->ppm->comm);
1756 res = ppi1->ppm->pid > ppi2->ppm->pid;
1761 static void sort_process_list(void)
1763 struct per_process_info **ppis;
1764 struct per_process_info *ppi;
1767 ppis = malloc(ppi_list_entries * sizeof(struct per_process_info *));
1772 ppi = ppi->list_next;
1775 qsort(ppis, ppi_list_entries, sizeof(ppi), ppi_name_compare);
1777 i = ppi_list_entries - 1;
1782 ppi->list_next = ppi_list;
1790 static void show_process_stats(void)
1792 struct per_process_info *ppi;
1794 sort_process_list();
1798 struct process_pid_map *ppm = ppi->ppm;
1801 if (ppi->more_than_one)
1802 sprintf(name, "%s (%u, ...)", ppm->comm, ppm->pid);
1804 sprintf(name, "%s (%u)", ppm->comm, ppm->pid);
1806 dump_io_stats(NULL, &ppi->io_stats, name);
1807 dump_wait_stats(ppi);
1808 ppi = ppi->list_next;
1814 static void show_device_and_cpu_stats(void)
1816 struct per_dev_info *pdi;
1817 struct per_cpu_info *pci;
1818 struct io_stats total, *ios;
1819 unsigned long long rrate, wrate, msec;
1820 int i, j, pci_events;
1821 char line[3 + 8/*cpu*/ + 2 + 32/*dev*/ + 3];
1825 for (pdi = devices, i = 0; i < ndevices; i++, pdi++) {
1827 memset(&total, 0, sizeof(total));
1833 for (pci = pdi->cpus, j = 0; j < pdi->ncpus; j++, pci++) {
1837 ios = &pci->io_stats;
1838 total.qreads += ios->qreads;
1839 total.qwrites += ios->qwrites;
1840 total.creads += ios->creads;
1841 total.cwrites += ios->cwrites;
1842 total.mreads += ios->mreads;
1843 total.mwrites += ios->mwrites;
1844 total.ireads += ios->ireads;
1845 total.iwrites += ios->iwrites;
1846 total.rrqueue += ios->rrqueue;
1847 total.wrqueue += ios->wrqueue;
1848 total.qread_kb += ios->qread_kb;
1849 total.qwrite_kb += ios->qwrite_kb;
1850 total.cread_kb += ios->cread_kb;
1851 total.cwrite_kb += ios->cwrite_kb;
1852 total.iread_kb += ios->iread_kb;
1853 total.iwrite_kb += ios->iwrite_kb;
1854 total.mread_kb += ios->mread_kb;
1855 total.mwrite_kb += ios->mwrite_kb;
1856 total.qread_b += ios->qread_b;
1857 total.qwrite_b += ios->qwrite_b;
1858 total.cread_b += ios->cread_b;
1859 total.cwrite_b += ios->cwrite_b;
1860 total.iread_b += ios->iread_b;
1861 total.iwrite_b += ios->iwrite_b;
1862 total.mread_b += ios->mread_b;
1863 total.mwrite_b += ios->mwrite_b;
1865 total.qreads_pc += ios->qreads_pc;
1866 total.qwrites_pc += ios->qwrites_pc;
1867 total.creads_pc += ios->creads_pc;
1868 total.cwrites_pc += ios->cwrites_pc;
1869 total.ireads_pc += ios->ireads_pc;
1870 total.iwrites_pc += ios->iwrites_pc;
1871 total.rrqueue_pc += ios->rrqueue_pc;
1872 total.wrqueue_pc += ios->wrqueue_pc;
1873 total.qread_kb_pc += ios->qread_kb_pc;
1874 total.qwrite_kb_pc += ios->qwrite_kb_pc;
1875 total.iread_kb_pc += ios->iread_kb_pc;
1876 total.iwrite_kb_pc += ios->iwrite_kb_pc;
1877 total.qread_b_pc += ios->qread_b_pc;
1878 total.qwrite_b_pc += ios->qwrite_b_pc;
1879 total.iread_b_pc += ios->iread_b_pc;
1880 total.iwrite_b_pc += ios->iwrite_b_pc;
1882 total.timer_unplugs += ios->timer_unplugs;
1883 total.io_unplugs += ios->io_unplugs;
1885 snprintf(line, sizeof(line) - 1, "CPU%d (%s):",
1886 j, get_dev_name(pdi, name, sizeof(name)));
1887 dump_io_stats(pdi, ios, line);
1891 if (pci_events > 1) {
1893 snprintf(line, sizeof(line) - 1, "Total (%s):",
1894 get_dev_name(pdi, name, sizeof(name)));
1895 dump_io_stats(NULL, &total, line);
1899 msec = (pdi->last_reported_time - pdi->first_reported_time) / 1000000;
1901 rrate = ((1000 * total.cread_kb) + total.cread_b) /
1903 wrate = ((1000 * total.cwrite_kb) + total.cwrite_b) /
1907 fprintf(ofp, "\nThroughput (R/W): %'LuKiB/s / %'LuKiB/s\n",
1909 fprintf(ofp, "Events (%s): %'Lu entries\n",
1910 get_dev_name(pdi, line, sizeof(line)), pdi->events);
1912 collect_pdi_skips(pdi);
1913 if (!pdi->skips && !pdi->events)
1916 ratio = 100.0 * ((double)pdi->seq_skips /
1917 (double)(pdi->events + pdi->seq_skips));
1918 fprintf(ofp, "Skips: %'lu forward (%'llu - %5.1lf%%)\n",
1919 pdi->skips, pdi->seq_skips, ratio);
1923 static void find_genesis(void)
1925 struct trace *t = trace_list;
1927 genesis_time = -1ULL;
1929 if (t->bit->time < genesis_time)
1930 genesis_time = t->bit->time;
1935 /* The time stamp record will usually be the first
1936 * record in the trace, but not always.
1939 && start_timestamp != genesis_time) {
1940 long delta = genesis_time - start_timestamp;
1942 abs_start_time.tv_sec += SECONDS(delta);
1943 abs_start_time.tv_nsec += NANO_SECONDS(delta);
1944 if (abs_start_time.tv_nsec < 0) {
1945 abs_start_time.tv_nsec += 1000000000;
1946 abs_start_time.tv_sec -= 1;
1948 if (abs_start_time.tv_nsec > 1000000000) {
1949 abs_start_time.tv_nsec -= 1000000000;
1950 abs_start_time.tv_sec += 1;
1955 static inline int check_stopwatch(struct blk_io_trace *bit)
1957 if (bit->time < stopwatch_end &&
1958 bit->time >= stopwatch_start)
1965 * return youngest entry read
1967 static int sort_entries(unsigned long long *youngest)
1969 struct per_dev_info *pdi = NULL;
1970 struct per_cpu_info *pci = NULL;
1977 while ((t = trace_list) != NULL) {
1978 struct blk_io_trace *bit = t->bit;
1980 trace_list = t->next;
1982 bit->time -= genesis_time;
1984 if (bit->time < *youngest || !*youngest)
1985 *youngest = bit->time;
1987 if (!pdi || pdi->dev != bit->device) {
1988 pdi = get_dev_info(bit->device);
1992 if (!pci || pci->cpu != bit->cpu)
1993 pci = get_cpu_info(pdi, bit->cpu);
1995 if (bit->sequence < pci->smallest_seq_read)
1996 pci->smallest_seq_read = bit->sequence;
1998 if (check_stopwatch(bit)) {
2004 if (trace_rb_insert_sort(t))
2012 * to continue, we must have traces from all online cpus in the tree
2014 static int check_cpu_map(struct per_dev_info *pdi)
2016 unsigned long *cpu_map;
2023 * create a map of the cpus we have traces for
2025 cpu_map = malloc(pdi->cpu_map_max / sizeof(long));
2026 memset(cpu_map, 0, sizeof(*cpu_map));
2027 n = rb_first(&rb_sort_root);
2029 __t = rb_entry(n, struct trace, rb_node);
2030 cpu = __t->bit->cpu;
2032 cpu_map[CPU_IDX(cpu)] |= (1UL << CPU_BIT(cpu));
2037 * we can't continue if pdi->cpu_map has entries set that we don't
2038 * have in the sort rbtree. the opposite is not a problem, though
2041 for (i = 0; i < pdi->cpu_map_max / CPUS_PER_LONG; i++) {
2042 if (pdi->cpu_map[i] & ~(cpu_map[i])) {
2052 static int check_sequence(struct per_dev_info *pdi, struct trace *t, int force)
2054 struct blk_io_trace *bit = t->bit;
2055 unsigned long expected_sequence;
2056 struct per_cpu_info *pci;
2059 pci = get_cpu_info(pdi, bit->cpu);
2060 expected_sequence = pci->last_sequence + 1;
2062 if (!expected_sequence) {
2064 * 1 should be the first entry, just allow it
2066 if (bit->sequence == 1)
2068 if (bit->sequence == pci->smallest_seq_read)
2071 return check_cpu_map(pdi);
2074 if (bit->sequence == expected_sequence)
2078 * we may not have seen that sequence yet. if we are not doing
2079 * the final run, break and wait for more entries.
2081 if (expected_sequence < pci->smallest_seq_read) {
2082 __t = trace_rb_find_last(pdi, pci, expected_sequence);
2086 __put_trace_last(pdi, __t);
2088 } else if (!force) {
2092 if (check_current_skips(pci, bit->sequence))
2095 if (expected_sequence < bit->sequence)
2096 insert_skip(pci, expected_sequence, bit->sequence - 1);
2101 static void show_entries_rb(int force)
2103 struct per_dev_info *pdi = NULL;
2104 struct per_cpu_info *pci = NULL;
2105 struct blk_io_trace *bit;
2109 while ((n = rb_first(&rb_sort_root)) != NULL) {
2110 if (is_done() && !force && !pipeline)
2113 t = rb_entry(n, struct trace, rb_node);
2116 if (read_sequence - t->read_sequence < 1 && !force)
2119 if (!pdi || pdi->dev != bit->device) {
2120 pdi = get_dev_info(bit->device);
2125 fprintf(stderr, "Unknown device ID? (%d,%d)\n",
2126 MAJOR(bit->device), MINOR(bit->device));
2130 if (!(bit->action == BLK_TN_MESSAGE) &&
2131 check_sequence(pdi, t, force))
2134 if (!force && bit->time > last_allowed_time)
2137 check_time(pdi, bit);
2139 if (!pci || pci->cpu != bit->cpu)
2140 pci = get_cpu_info(pdi, bit->cpu);
2142 if (!(bit->action == BLK_TN_MESSAGE))
2143 pci->last_sequence = bit->sequence;
2147 if (bit->action & (act_mask << BLK_TC_SHIFT))
2148 dump_trace(bit, pci, pdi);
2154 static int read_data(int fd, void *buffer, int bytes, int block, int *fdblock)
2156 int ret, bytes_left, fl;
2159 if (block != *fdblock) {
2160 fl = fcntl(fd, F_GETFL);
2164 fcntl(fd, F_SETFL, fl | O_NONBLOCK);
2167 fcntl(fd, F_SETFL, fl & ~O_NONBLOCK);
2173 while (bytes_left > 0) {
2174 ret = read(fd, p, bytes_left);
2178 if (errno != EAGAIN) {
2184 * never do partial reads. we can return if we
2185 * didn't read anything and we should not block,
2186 * otherwise wait for data
2188 if ((bytes_left == bytes) && !block)
2202 static inline __u16 get_pdulen(struct blk_io_trace *bit)
2205 return bit->pdu_len;
2207 return __bswap_16(bit->pdu_len);
2210 static inline __u32 get_magic(struct blk_io_trace *bit)
2215 return __bswap_32(bit->magic);
2218 static int read_events(int fd, int always_block, int *fdblock)
2220 struct per_dev_info *pdi = NULL;
2221 unsigned int events = 0;
2223 while (!is_done() && events < rb_batch) {
2224 struct blk_io_trace *bit;
2226 int pdu_len, should_block, ret;
2231 should_block = !events || always_block;
2233 ret = read_data(fd, bit, sizeof(*bit), should_block, fdblock);
2236 if (!events && ret < 0)
2242 * look at first trace to check whether we need to convert
2243 * data in the future
2245 if (data_is_native == -1 && check_data_endianness(bit->magic))
2248 magic = get_magic(bit);
2249 if ((magic & 0xffffff00) != BLK_IO_TRACE_MAGIC) {
2250 fprintf(stderr, "Bad magic %x\n", magic);
2254 pdu_len = get_pdulen(bit);
2256 void *ptr = realloc(bit, sizeof(*bit) + pdu_len);
2258 if (read_data(fd, ptr + sizeof(*bit), pdu_len, 1, fdblock)) {
2268 if (verify_trace(bit)) {
2274 * not a real trace, so grab and handle it here
2276 if (bit->action & BLK_TC_ACT(BLK_TC_NOTIFY) && bit->action != BLK_TN_MESSAGE) {
2278 output_binary(bit, sizeof(*bit) + bit->pdu_len);
2283 memset(t, 0, sizeof(*t));
2285 t->read_sequence = read_sequence;
2287 t->next = trace_list;
2290 if (!pdi || pdi->dev != bit->device)
2291 pdi = get_dev_info(bit->device);
2293 if (bit->time > pdi->last_read_time)
2294 pdi->last_read_time = bit->time;
2303 * Managing input streams
2307 struct ms_stream *next;
2308 struct trace *first, *last;
2309 struct per_dev_info *pdi;
2313 #define MS_HASH(d, c) ((MAJOR(d) & 0xff) ^ (MINOR(d) & 0xff) ^ (cpu & 0xff))
2315 struct ms_stream *ms_head;
2316 struct ms_stream *ms_hash[256];
2318 static void ms_sort(struct ms_stream *msp);
2319 static int ms_prime(struct ms_stream *msp);
2321 static inline struct trace *ms_peek(struct ms_stream *msp)
2323 return (msp == NULL) ? NULL : msp->first;
2326 static inline __u64 ms_peek_time(struct ms_stream *msp)
2328 return ms_peek(msp)->bit->time;
2331 static inline void ms_resort(struct ms_stream *msp)
2333 if (msp->next && ms_peek_time(msp) > ms_peek_time(msp->next)) {
2334 ms_head = msp->next;
2340 static inline void ms_deq(struct ms_stream *msp)
2342 msp->first = msp->first->next;
2345 if (!ms_prime(msp)) {
2346 ms_head = msp->next;
2355 static void ms_sort(struct ms_stream *msp)
2357 __u64 msp_t = ms_peek_time(msp);
2358 struct ms_stream *this_msp = ms_head;
2360 if (this_msp == NULL)
2362 else if (msp_t < ms_peek_time(this_msp)) {
2363 msp->next = this_msp;
2367 while (this_msp->next && ms_peek_time(this_msp->next) < msp_t)
2368 this_msp = this_msp->next;
2370 msp->next = this_msp->next;
2371 this_msp->next = msp;
2375 static int ms_prime(struct ms_stream *msp)
2380 struct per_dev_info *pdi = msp->pdi;
2381 struct per_cpu_info *pci = get_cpu_info(pdi, msp->cpu);
2382 struct blk_io_trace *bit = NULL;
2383 int ret, pdu_len, ndone = 0;
2385 for (i = 0; !is_done() && pci->fd >= 0 && i < rb_batch; i++) {
2387 ret = read_data(pci->fd, bit, sizeof(*bit), 1, &pci->fdblock);
2391 if (data_is_native == -1 && check_data_endianness(bit->magic))
2394 magic = get_magic(bit);
2395 if ((magic & 0xffffff00) != BLK_IO_TRACE_MAGIC) {
2396 fprintf(stderr, "Bad magic %x\n", magic);
2401 pdu_len = get_pdulen(bit);
2403 void *ptr = realloc(bit, sizeof(*bit) + pdu_len);
2404 ret = read_data(pci->fd, ptr + sizeof(*bit), pdu_len,
2416 if (verify_trace(bit))
2419 if (bit->cpu != pci->cpu) {
2420 fprintf(stderr, "cpu %d trace info has error cpu %d\n",
2421 pci->cpu, bit->cpu);
2425 if (bit->action & BLK_TC_ACT(BLK_TC_NOTIFY) && bit->action != BLK_TN_MESSAGE) {
2427 output_binary(bit, sizeof(*bit) + bit->pdu_len);
2434 if (bit->time > pdi->last_read_time)
2435 pdi->last_read_time = bit->time;
2438 memset(t, 0, sizeof(*t));
2441 if (msp->first == NULL)
2442 msp->first = msp->last = t;
2444 msp->last->next = t;
2454 if (bit) bit_free(bit);
2456 cpu_mark_offline(pdi, pci->cpu);
2463 static struct ms_stream *ms_alloc(struct per_dev_info *pdi, int cpu)
2465 struct ms_stream *msp = malloc(sizeof(*msp));
2468 msp->first = msp->last = NULL;
2478 static int setup_file(struct per_dev_info *pdi, int cpu)
2483 struct per_cpu_info *pci = get_cpu_info(pdi, cpu);
2488 p = strdup(pdi->name);
2490 if (strcmp(dname, ".")) {
2492 p = strdup(pdi->name);
2493 strcpy(pdi->name, basename(p));
2498 len = sprintf(pci->fname, "%s/", input_dir);
2500 snprintf(pci->fname + len, sizeof(pci->fname)-1-len,
2501 "%s.blktrace.%d", pdi->name, pci->cpu);
2502 if (stat(pci->fname, &st) < 0)
2507 pci->fd = open(pci->fname, O_RDONLY);
2513 printf("Input file %s added\n", pci->fname);
2514 cpu_mark_online(pdi, pci->cpu);
2517 ms_alloc(pdi, pci->cpu);
2522 static int handle(struct ms_stream *msp)
2525 struct per_dev_info *pdi;
2526 struct per_cpu_info *pci;
2527 struct blk_io_trace *bit;
2533 pci = get_cpu_info(pdi, msp->cpu);
2535 bit->time -= genesis_time;
2537 if (t->bit->time > stopwatch_end)
2540 pdi->last_reported_time = bit->time;
2541 if ((bit->action & (act_mask << BLK_TC_SHIFT))&&
2542 t->bit->time >= stopwatch_start)
2543 dump_trace(bit, pci, pdi);
2548 trace_rb_insert_last(pdi, t);
2558 * Check if we need to sanitize the name. We allow 'foo', or if foo.blktrace.X
2559 * is given, then strip back down to 'foo' to avoid missing files.
2561 static int name_fixup(char *name)
2568 b = strstr(name, ".blktrace.");
2575 static int do_file(void)
2578 struct per_dev_info *pdi;
2581 * first prepare all files for reading
2583 for (i = 0; i < ndevices; i++) {
2585 ret = name_fixup(pdi->name);
2589 for (cpu = 0; setup_file(pdi, cpu); cpu++)
2593 fprintf(stderr,"No input files found for %s\n",
2600 * Get the initial time stamp
2603 genesis_time = ms_peek_time(ms_head);
2606 * Keep processing traces while any are left
2608 while (!is_done() && ms_head && handle(ms_head))
2614 static void do_pipe(int fd)
2616 unsigned long long youngest;
2617 int events, fdblock;
2619 last_allowed_time = -1ULL;
2621 while ((events = read_events(fd, 0, &fdblock)) > 0) {
2625 smallest_seq_read = -1U;
2628 if (sort_entries(&youngest))
2631 if (youngest > stopwatch_end)
2637 if (rb_sort_entries)
2641 static int do_fifo(void)
2645 if (!strcmp(pipename, "-"))
2646 fd = dup(STDIN_FILENO);
2648 fd = open(pipename, O_RDONLY);
2651 perror("dup stdin");
2660 static void show_stats(void)
2669 if (per_process_stats)
2670 show_process_stats();
2672 if (per_device_and_cpu_stats)
2673 show_device_and_cpu_stats();
2678 static void handle_sigint(__attribute__((__unused__)) int sig)
2684 * Extract start and duration times from a string, allowing
2685 * us to specify a time interval of interest within a trace.
2686 * Format: "duration" (start is zero) or "start:duration".
2688 static int find_stopwatch_interval(char *string)
2693 value = strtod(string, &sp);
2695 fprintf(stderr,"Invalid stopwatch timer: %s\n", string);
2699 stopwatch_start = DOUBLE_TO_NANO_ULL(value);
2701 value = strtod(string, &sp);
2702 if (sp == string || *sp != '\0') {
2703 fprintf(stderr,"Invalid stopwatch duration time: %s\n",
2707 } else if (*sp != '\0') {
2708 fprintf(stderr,"Invalid stopwatch start timer: %s\n", string);
2711 stopwatch_end = DOUBLE_TO_NANO_ULL(value);
2712 if (stopwatch_end <= stopwatch_start) {
2713 fprintf(stderr, "Invalid stopwatch interval: %Lu -> %Lu\n",
2714 stopwatch_start, stopwatch_end);
2721 static int is_pipe(const char *str)
2725 if (!strcmp(str, "-"))
2727 if (!stat(str, &st) && S_ISFIFO(st.st_mode))
2733 #define S_OPTS "a:A:b:D:d:f:F:hi:o:Oqstw:vVM"
2734 static char usage_str[] = "\n\n" \
2735 "-i <file> | --input=<file>\n" \
2736 "[ -a <action field> | --act-mask=<action field> ]\n" \
2737 "[ -A <action mask> | --set-mask=<action mask> ]\n" \
2738 "[ -b <traces> | --batch=<traces> ]\n" \
2739 "[ -d <file> | --dump-binary=<file> ]\n" \
2740 "[ -D <dir> | --input-directory=<dir> ]\n" \
2741 "[ -f <format> | --format=<format> ]\n" \
2742 "[ -F <spec> | --format-spec=<spec> ]\n" \
2743 "[ -h | --hash-by-name ]\n" \
2744 "[ -o <file> | --output=<file> ]\n" \
2745 "[ -O | --no-text-output ]\n" \
2746 "[ -q | --quiet ]\n" \
2747 "[ -s | --per-program-stats ]\n" \
2748 "[ -t | --track-ios ]\n" \
2749 "[ -w <time> | --stopwatch=<time> ]\n" \
2750 "[ -M | --no-msgs\n" \
2751 "[ -v | --verbose ]\n" \
2752 "[ -V | --version ]\n\n" \
2753 "\t-a Only trace specified actions. See documentation\n" \
2754 "\t-A Give trace mask as a single value. See documentation\n" \
2755 "\t-b stdin read batching\n" \
2756 "\t-d Output file. If specified, binary data is written to file\n" \
2757 "\t-D Directory to prepend to input file names\n" \
2758 "\t-f Output format. Customize the output format. The format field\n" \
2759 "\t identifies can be found in the documentation\n" \
2760 "\t-F Format specification. Can be found in the documentation\n" \
2761 "\t-h Hash processes by name, not pid\n" \
2762 "\t-i Input file containing trace data, or '-' for stdin\n" \
2763 "\t-o Output file. If not given, output is stdout\n" \
2764 "\t-O Do NOT output text data\n" \
2765 "\t-q Quiet. Don't display any stats at the end of the trace\n" \
2766 "\t-s Show per-program io statistics\n" \
2767 "\t-t Track individual ios. Will tell you the time a request took\n" \
2768 "\t to get queued, to get dispatched, and to get completed\n" \
2769 "\t-w Only parse data between the given time interval in seconds.\n" \
2770 "\t If 'start' isn't given, blkparse defaults the start time to 0\n" \
2771 "\t-M Do not output messages to binary file\n" \
2772 "\t-v More verbose for marginal errors\n" \
2773 "\t-V Print program version info\n\n";
2775 static void usage(char *prog)
2777 fprintf(stderr, "Usage: %s %s", prog, usage_str);
2780 int main(int argc, char *argv[])
2782 int i, c, ret, mode;
2783 int act_mask_tmp = 0;
2784 char *ofp_buffer = NULL;
2785 char *bin_ofp_buffer = NULL;
2787 while ((c = getopt_long(argc, argv, S_OPTS, l_opts, NULL)) != -1) {
2790 i = find_mask_map(optarg);
2792 fprintf(stderr,"Invalid action mask %s\n",
2800 if ((sscanf(optarg, "%x", &i) != 1) ||
2801 !valid_act_opt(i)) {
2803 "Invalid set action mask %s/0x%x\n",
2810 if (is_pipe(optarg) && !pipeline) {
2812 pipename = strdup(optarg);
2813 } else if (resize_devices(optarg) != 0)
2820 output_name = optarg;
2826 rb_batch = atoi(optarg);
2828 rb_batch = RB_BATCH_DEFAULT;
2831 per_process_stats = 1;
2837 per_device_and_cpu_stats = 0;
2840 if (find_stopwatch_interval(optarg) != 0)
2844 set_all_format_specs(optarg);
2847 if (add_format_spec(optarg) != 0)
2851 ppi_hash_by_pid = 0;
2857 printf("%s version %s\n", argv[0], blkparse_version);
2860 dump_binary = optarg;
2863 bin_output_msgs = 0;
2871 while (optind < argc) {
2872 if (is_pipe(argv[optind]) && !pipeline) {
2874 pipename = strdup(argv[optind]);
2875 } else if (resize_devices(argv[optind]) != 0)
2880 if (!pipeline && !ndevices) {
2885 if (act_mask_tmp != 0)
2886 act_mask = act_mask_tmp;
2888 memset(&rb_sort_root, 0, sizeof(rb_sort_root));
2890 signal(SIGINT, handle_sigint);
2891 signal(SIGHUP, handle_sigint);
2892 signal(SIGTERM, handle_sigint);
2894 setlocale(LC_NUMERIC, "en_US");
2898 ofp = fdopen(STDOUT_FILENO, "w");
2901 char ofname[PATH_MAX];
2903 snprintf(ofname, sizeof(ofname) - 1, "%s", output_name);
2904 ofp = fopen(ofname, "w");
2913 ofp_buffer = malloc(4096);
2914 if (setvbuf(ofp, ofp_buffer, mode, 4096)) {
2921 if (!strcmp(dump_binary, "-"))
2924 dump_fp = fopen(dump_binary, "w");
2926 perror(dump_binary);
2931 bin_ofp_buffer = malloc(128 * 1024);
2932 if (setvbuf(dump_fp, bin_ofp_buffer, _IOFBF, 128 * 1024)) {
2933 perror("setvbuf binary");
2946 if (have_drv_data && !dump_binary)
2947 printf("\ndiscarded traces containing low-level device driver "
2948 "specific data (only available in binary output)\n");
2954 if (bin_ofp_buffer) {
2956 free(bin_ofp_buffer);