Commit | Line | Data |
---|---|---|
d0ca268b JA |
1 | /* |
2 | * block queue tracing application | |
3 | * | |
d956a2cd | 4 | * Copyright (C) 2005 Jens Axboe <axboe@suse.de> |
46e37c55 | 5 | * Copyright (C) 2006 Jens Axboe <axboe@kernel.dk> |
d956a2cd | 6 | * |
3fe0b570 AB |
7 | * Rewrite to have a single thread per CPU (managing all devices on that CPU) |
8 | * Alan D. Brunelle <alan.brunelle@hp.com> - January 2009 | |
9 | * | |
d956a2cd JA |
10 | * This program is free software; you can redistribute it and/or modify |
11 | * it under the terms of the GNU General Public License as published by | |
12 | * the Free Software Foundation; either version 2 of the License, or | |
13 | * (at your option) any later version. | |
14 | * | |
15 | * This program is distributed in the hope that it will be useful, | |
16 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
17 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
18 | * GNU General Public License for more details. | |
19 | * | |
20 | * You should have received a copy of the GNU General Public License | |
21 | * along with this program; if not, write to the Free Software | |
22 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | |
23 | * | |
d0ca268b | 24 | */ |
3fe0b570 AB |
25 | |
26 | #include <errno.h> | |
27 | #include <stdarg.h> | |
28 | #include <stdio.h> | |
29 | #include <stdlib.h> | |
30 | #include <string.h> | |
31 | #include <fcntl.h> | |
32 | #include <getopt.h> | |
33 | #include <sched.h> | |
d0ca268b | 34 | #include <unistd.h> |
3fe0b570 | 35 | #include <poll.h> |
d0ca268b | 36 | #include <signal.h> |
3fe0b570 AB |
37 | #include <pthread.h> |
38 | #include <locale.h> | |
d0ca268b | 39 | #include <sys/ioctl.h> |
3fe0b570 AB |
40 | #include <sys/types.h> |
41 | #include <sys/stat.h> | |
42 | #include <sys/vfs.h> | |
b7106311 | 43 | #include <sys/mman.h> |
3fe0b570 AB |
44 | #include <sys/param.h> |
45 | #include <sys/time.h> | |
46 | #include <sys/resource.h> | |
8e86c98a | 47 | #include <sys/socket.h> |
8e86c98a JA |
48 | #include <netinet/in.h> |
49 | #include <arpa/inet.h> | |
50 | #include <netdb.h> | |
32f18c48 | 51 | #include <sys/sendfile.h> |
d0ca268b | 52 | |
3fe0b570 | 53 | #include "btt/list.h" |
d0ca268b | 54 | #include "blktrace.h" |
52724a0e | 55 | |
8f551a39 JA |
56 | /* |
57 | * You may want to increase this even more, if you are logging at a high | |
58 | * rate and see skipped/missed events | |
59 | */ | |
3fe0b570 AB |
60 | #define BUF_SIZE (512 * 1024) |
61 | #define BUF_NR (4) | |
62 | ||
63 | #define FILE_VBUF_SIZE (128 * 1024) | |
64 | ||
65 | #define DEBUGFS_TYPE (0x64626720) | |
66 | #define TRACE_NET_PORT (8462) | |
67 | ||
68 | enum { | |
69 | Net_none = 0, | |
70 | Net_server, | |
71 | Net_client, | |
72 | }; | |
73 | ||
df81fdb5 AB |
74 | enum thread_status { |
75 | Th_running, | |
76 | Th_leaving, | |
77 | Th_error | |
78 | }; | |
79 | ||
3fe0b570 AB |
80 | /* |
81 | * Generic stats collected: nevents can be _roughly_ estimated by data_read | |
82 | * (discounting pdu...) | |
83 | * | |
84 | * These fields are updated w/ pdc_dr_update & pdc_nev_update below. | |
85 | */ | |
86 | struct pdc_stats { | |
87 | unsigned long long data_read; | |
88 | unsigned long long nevents; | |
89 | }; | |
90 | ||
91 | struct devpath { | |
92 | struct list_head head; | |
93 | char *path; /* path to device special file */ | |
94 | char *buts_name; /* name returned from bt kernel code */ | |
95 | struct pdc_stats *stats; | |
96 | int fd, idx, ncpus; | |
97 | unsigned long long drops; | |
98 | ||
99 | /* | |
100 | * For piped output only: | |
101 | * | |
102 | * Each tracer will have a tracer_devpath_head that it will add new | |
103 | * data onto. It's list is protected above (tracer_devpath_head.mutex) | |
104 | * and it will signal the processing thread using the dp_cond, | |
105 | * dp_mutex & dp_entries variables above. | |
106 | */ | |
107 | struct tracer_devpath_head *heads; | |
108 | ||
109 | /* | |
110 | * For network server mode only: | |
111 | */ | |
112 | struct cl_host *ch; | |
113 | u32 cl_id; | |
114 | time_t cl_connect_time; | |
115 | struct io_info *ios; | |
116 | }; | |
117 | ||
118 | /* | |
119 | * For piped output to stdout we will have each tracer thread (one per dev) | |
120 | * tack buffers read from the relay queues on a per-device list. | |
121 | * | |
122 | * The main thread will then collect trace buffers from each of lists in turn. | |
123 | * | |
124 | * We will use a mutex to guard each of the trace_buf list. The tracers | |
125 | * can then signal the main thread using <dp_cond,dp_mutex> and | |
126 | * dp_entries. (When dp_entries is 0, and a tracer adds an entry it will | |
127 | * signal. When dp_entries is 0, the main thread will wait for that condition | |
128 | * to be signalled.) | |
129 | * | |
130 | * adb: It may be better just to have a large buffer per tracer per dev, | |
131 | * and then use it as a ring-buffer. This would certainly cut down a lot | |
132 | * of malloc/free thrashing, at the cost of more memory movements (potentially). | |
133 | */ | |
134 | struct trace_buf { | |
135 | struct list_head head; | |
136 | struct devpath *dpp; | |
137 | void *buf; | |
138 | int cpu, len; | |
139 | }; | |
140 | ||
141 | struct tracer_devpath_head { | |
142 | pthread_mutex_t mutex; | |
143 | struct list_head head; | |
144 | struct trace_buf *prev; | |
145 | }; | |
146 | ||
147 | /* | |
148 | * Used to handle the mmap() interfaces for output file (containing traces) | |
149 | */ | |
150 | struct mmap_info { | |
151 | void *fs_buf; | |
152 | unsigned long long fs_size, fs_max_size, fs_off, fs_buf_len; | |
153 | unsigned long buf_size, buf_nr; | |
154 | int pagesize; | |
155 | }; | |
156 | ||
157 | /* | |
158 | * Each thread doing work on a (client) side of blktrace will have one | |
159 | * of these. The ios array contains input/output information, pfds holds | |
160 | * poll() data. The volatile's provide flags to/from the main executing | |
161 | * thread. | |
162 | */ | |
163 | struct tracer { | |
164 | struct list_head head; | |
165 | struct io_info *ios; | |
166 | struct pollfd *pfds; | |
167 | pthread_t thread; | |
3fe0b570 | 168 | int cpu, nios; |
df81fdb5 | 169 | volatile int status, is_done; |
3fe0b570 AB |
170 | }; |
171 | ||
172 | /* | |
173 | * networking stuff follows. we include a magic number so we know whether | |
174 | * to endianness convert or not. | |
175 | * | |
176 | * The len field is overloaded: | |
177 | * 0 - Indicates an "open" - allowing the server to set up for a dev/cpu | |
178 | * 1 - Indicates a "close" - Shut down connection orderly | |
179 | * | |
180 | * The cpu field is overloaded on close: it will contain the number of drops. | |
181 | */ | |
182 | struct blktrace_net_hdr { | |
183 | u32 magic; /* same as trace magic */ | |
184 | char buts_name[32]; /* trace name */ | |
185 | u32 cpu; /* for which cpu */ | |
186 | u32 max_cpus; | |
187 | u32 len; /* length of following trace data */ | |
188 | u32 cl_id; /* id for set of client per-cpu connections */ | |
189 | u32 buf_size; /* client buf_size for this trace */ | |
190 | u32 buf_nr; /* client buf_nr for this trace */ | |
191 | u32 page_size; /* client page_size for this trace */ | |
192 | }; | |
193 | ||
194 | /* | |
195 | * Each host encountered has one of these. The head is used to link this | |
196 | * on to the network server's ch_list. Connections associated with this | |
197 | * host are linked on conn_list, and any devices traced on that host | |
198 | * are connected on the devpaths list. | |
199 | */ | |
200 | struct cl_host { | |
201 | struct list_head head; | |
202 | struct list_head conn_list; | |
203 | struct list_head devpaths; | |
204 | struct net_server_s *ns; | |
205 | char *hostname; | |
206 | struct in_addr cl_in_addr; | |
207 | int connects, ndevs, cl_opens; | |
208 | }; | |
209 | ||
210 | /* | |
211 | * Each connection (client to server socket ('fd')) has one of these. A | |
212 | * back reference to the host ('ch'), and lists headers (for the host | |
213 | * list, and the network server conn_list) are also included. | |
214 | */ | |
215 | struct cl_conn { | |
216 | struct list_head ch_head, ns_head; | |
217 | struct cl_host *ch; | |
218 | int fd, ncpus; | |
219 | time_t connect_time; | |
220 | }; | |
221 | ||
222 | /* | |
223 | * The network server requires some poll structures to be maintained - | |
224 | * one per conection currently on conn_list. The nchs/ch_list values | |
225 | * are for each host connected to this server. The addr field is used | |
226 | * for scratch as new connections are established. | |
227 | */ | |
228 | struct net_server_s { | |
229 | struct list_head conn_list; | |
230 | struct list_head ch_list; | |
231 | struct pollfd *pfds; | |
232 | int listen_fd, connects, nchs; | |
233 | struct sockaddr_in addr; | |
234 | }; | |
235 | ||
236 | /* | |
237 | * This structure is (generically) used to providide information | |
238 | * for a read-to-write set of values. | |
239 | * | |
240 | * ifn & ifd represent input information | |
241 | * | |
242 | * ofn, ofd, ofp, obuf & mmap_info are used for output file (optionally). | |
243 | */ | |
244 | struct io_info { | |
245 | struct devpath *dpp; | |
246 | FILE *ofp; | |
247 | char *obuf; | |
248 | struct cl_conn *nc; /* Server network connection */ | |
249 | ||
250 | /* | |
251 | * mmap controlled output files | |
252 | */ | |
253 | struct mmap_info mmap_info; | |
254 | ||
255 | /* | |
256 | * Client network fields | |
257 | */ | |
258 | unsigned int ready; | |
259 | unsigned long long data_queued; | |
260 | ||
261 | /* | |
262 | * Input/output file descriptors & names | |
263 | */ | |
264 | int ifd, ofd; | |
265 | char ifn[MAXPATHLEN + 64]; | |
266 | char ofn[MAXPATHLEN + 64]; | |
267 | }; | |
268 | ||
269 | static char blktrace_version[] = "2.0.0"; | |
270 | ||
271 | /* | |
272 | * Linkage to blktrace helper routines (trace conversions) | |
273 | */ | |
274 | int data_is_native = -1; | |
275 | ||
055cc3e5 | 276 | static int ndevs; |
3fe0b570 AB |
277 | static int ncpus; |
278 | static int pagesize; | |
279 | static int act_mask = ~0U; | |
055cc3e5 AB |
280 | static int kill_running_trace; |
281 | static int stop_watch; | |
282 | static int piped_output; | |
283 | ||
3fe0b570 AB |
284 | static char *debugfs_path = "/sys/kernel/debug"; |
285 | static char *output_name; | |
286 | static char *output_dir; | |
055cc3e5 | 287 | |
3fe0b570 AB |
288 | static unsigned long buf_size = BUF_SIZE; |
289 | static unsigned long buf_nr = BUF_NR; | |
055cc3e5 AB |
290 | |
291 | static FILE *pfp; | |
292 | ||
3fe0b570 AB |
293 | static LIST_HEAD(devpaths); |
294 | static LIST_HEAD(tracers); | |
055cc3e5 | 295 | |
3fe0b570 | 296 | static volatile int done; |
d0ca268b | 297 | |
6488ca48 AB |
298 | /* |
299 | * tracer threads add entries, the main thread takes them off and processes | |
300 | * them. These protect the dp_entries variable. | |
301 | */ | |
3fe0b570 AB |
302 | static pthread_cond_t dp_cond = PTHREAD_COND_INITIALIZER; |
303 | static pthread_mutex_t dp_mutex = PTHREAD_MUTEX_INITIALIZER; | |
304 | static volatile int dp_entries; | |
305 | ||
6488ca48 | 306 | /* |
df81fdb5 | 307 | * These synchronize master / thread interactions. |
6488ca48 | 308 | */ |
df81fdb5 AB |
309 | static pthread_cond_t mt_cond = PTHREAD_COND_INITIALIZER; |
310 | static pthread_mutex_t mt_mutex = PTHREAD_MUTEX_INITIALIZER; | |
311 | static volatile int nthreads_running; | |
312 | static volatile int nthreads_leaving; | |
313 | static volatile int nthreads_error; | |
314 | static volatile int tracers_run; | |
6488ca48 | 315 | |
3fe0b570 AB |
316 | /* |
317 | * network cmd line params | |
318 | */ | |
e58f3937 | 319 | static struct sockaddr_in hostname_addr; |
3fe0b570 AB |
320 | static char hostname[MAXHOSTNAMELEN]; |
321 | static int net_port = TRACE_NET_PORT; | |
322 | static int net_use_sendfile = 1; | |
323 | static int net_mode; | |
324 | static int *cl_fds; | |
007c233c | 325 | |
3fe0b570 AB |
326 | static int (*handle_pfds)(struct tracer *, int, int); |
327 | static int (*handle_list)(struct tracer_devpath_head *, struct list_head *); | |
e3e74029 | 328 | |
5d4f19d9 | 329 | #define S_OPTS "d:a:A:r:o:kw:vVb:n:D:lh:p:sI:" |
d5396421 | 330 | static struct option l_opts[] = { |
5c86134e | 331 | { |
d39c04ca | 332 | .name = "dev", |
428683db | 333 | .has_arg = required_argument, |
d39c04ca AB |
334 | .flag = NULL, |
335 | .val = 'd' | |
336 | }, | |
cf1edb17 AB |
337 | { |
338 | .name = "input-devs", | |
339 | .has_arg = required_argument, | |
340 | .flag = NULL, | |
341 | .val = 'I' | |
342 | }, | |
5c86134e | 343 | { |
d39c04ca | 344 | .name = "act-mask", |
428683db | 345 | .has_arg = required_argument, |
d39c04ca AB |
346 | .flag = NULL, |
347 | .val = 'a' | |
348 | }, | |
5c86134e | 349 | { |
d39c04ca | 350 | .name = "set-mask", |
428683db | 351 | .has_arg = required_argument, |
d39c04ca AB |
352 | .flag = NULL, |
353 | .val = 'A' | |
354 | }, | |
5c86134e | 355 | { |
5270dddd | 356 | .name = "relay", |
428683db | 357 | .has_arg = required_argument, |
5270dddd JA |
358 | .flag = NULL, |
359 | .val = 'r' | |
360 | }, | |
d5396421 JA |
361 | { |
362 | .name = "output", | |
428683db | 363 | .has_arg = required_argument, |
d5396421 JA |
364 | .flag = NULL, |
365 | .val = 'o' | |
366 | }, | |
bc39777c JA |
367 | { |
368 | .name = "kill", | |
428683db | 369 | .has_arg = no_argument, |
bc39777c JA |
370 | .flag = NULL, |
371 | .val = 'k' | |
372 | }, | |
ece238a6 NS |
373 | { |
374 | .name = "stopwatch", | |
428683db | 375 | .has_arg = required_argument, |
ece238a6 NS |
376 | .flag = NULL, |
377 | .val = 'w' | |
378 | }, | |
5d4f19d9 JA |
379 | { |
380 | .name = "version", | |
381 | .has_arg = no_argument, | |
382 | .flag = NULL, | |
383 | .val = 'v' | |
384 | }, | |
52724a0e JA |
385 | { |
386 | .name = "version", | |
387 | .has_arg = no_argument, | |
388 | .flag = NULL, | |
57ea8602 | 389 | .val = 'V' |
52724a0e | 390 | }, |
129aa440 | 391 | { |
3f65c585 | 392 | .name = "buffer-size", |
129aa440 JA |
393 | .has_arg = required_argument, |
394 | .flag = NULL, | |
395 | .val = 'b' | |
396 | }, | |
397 | { | |
3f65c585 | 398 | .name = "num-sub-buffers", |
129aa440 JA |
399 | .has_arg = required_argument, |
400 | .flag = NULL, | |
401 | .val = 'n' | |
402 | }, | |
d1d7f15f | 403 | { |
3f65c585 | 404 | .name = "output-dir", |
d1d7f15f JA |
405 | .has_arg = required_argument, |
406 | .flag = NULL, | |
407 | .val = 'D' | |
408 | }, | |
8e86c98a JA |
409 | { |
410 | .name = "listen", | |
411 | .has_arg = no_argument, | |
412 | .flag = NULL, | |
413 | .val = 'l' | |
414 | }, | |
415 | { | |
416 | .name = "host", | |
417 | .has_arg = required_argument, | |
418 | .flag = NULL, | |
419 | .val = 'h' | |
420 | }, | |
421 | { | |
422 | .name = "port", | |
423 | .has_arg = required_argument, | |
424 | .flag = NULL, | |
425 | .val = 'p' | |
426 | }, | |
32f18c48 | 427 | { |
79971f43 | 428 | .name = "no-sendfile", |
32f18c48 JA |
429 | .has_arg = no_argument, |
430 | .flag = NULL, | |
431 | .val = 's' | |
432 | }, | |
71ef8b7c JA |
433 | { |
434 | .name = NULL, | |
435 | } | |
d39c04ca AB |
436 | }; |
437 | ||
3fe0b570 AB |
438 | static char usage_str[] = \ |
439 | "-d <dev> [ -r debugfs path ] [ -o <output> ] [-k ] [ -w time ]\n" \ | |
440 | "[ -a action ] [ -A action mask ] [ -I <devs file> ] [ -v ]\n\n" \ | |
441 | "\t-d Use specified device. May also be given last after options\n" \ | |
442 | "\t-r Path to mounted debugfs, defaults to /sys/kernel/debug\n" \ | |
443 | "\t-o File(s) to send output to\n" \ | |
444 | "\t-D Directory to prepend to output file names\n" \ | |
445 | "\t-k Kill a running trace\n" \ | |
446 | "\t-w Stop after defined time, in seconds\n" \ | |
447 | "\t-a Only trace specified actions. See documentation\n" \ | |
448 | "\t-A Give trace mask as a single value. See documentation\n" \ | |
449 | "\t-b Sub buffer size in KiB\n" \ | |
450 | "\t-n Number of sub buffers\n" \ | |
451 | "\t-l Run in network listen mode (blktrace server)\n" \ | |
452 | "\t-h Run in network client mode, connecting to the given host\n" \ | |
453 | "\t-p Network port to use (default 8462)\n" \ | |
454 | "\t-s Make the network client NOT use sendfile() to transfer data\n" \ | |
455 | "\t-I Add devices found in <devs file>\n" \ | |
456 | "\t-V Print program version info\n\n"; | |
9db17354 | 457 | |
3fe0b570 AB |
458 | static void clear_events(struct pollfd *pfd) |
459 | { | |
460 | pfd->events = 0; | |
461 | pfd->revents = 0; | |
462 | } | |
21f55651 | 463 | |
3fe0b570 AB |
464 | static inline int net_client_use_sendfile(void) |
465 | { | |
466 | return net_mode == Net_client && net_use_sendfile; | |
467 | } | |
21f55651 | 468 | |
3fe0b570 AB |
469 | static inline int net_client_use_send(void) |
470 | { | |
471 | return net_mode == Net_client && !net_use_sendfile; | |
472 | } | |
b9d4294e | 473 | |
3fe0b570 AB |
474 | static inline int use_tracer_devpaths(void) |
475 | { | |
476 | return piped_output || net_client_use_send(); | |
477 | } | |
b9d4294e | 478 | |
3fe0b570 AB |
479 | static inline int in_addr_eq(struct in_addr a, struct in_addr b) |
480 | { | |
481 | return a.s_addr == b.s_addr; | |
482 | } | |
007c233c | 483 | |
3fe0b570 AB |
484 | static inline void pdc_dr_update(struct devpath *dpp, int cpu, int data_read) |
485 | { | |
486 | dpp->stats[cpu].data_read += data_read; | |
487 | } | |
0cc7d25e | 488 | |
3fe0b570 AB |
489 | static inline void pdc_nev_update(struct devpath *dpp, int cpu, int nevents) |
490 | { | |
491 | dpp->stats[cpu].nevents += nevents; | |
492 | } | |
9db17354 | 493 | |
3fe0b570 AB |
494 | static void show_usage(char *prog) |
495 | { | |
496 | fprintf(stderr, "Usage: %s %s %s", prog, blktrace_version, usage_str); | |
497 | } | |
9db17354 | 498 | |
df81fdb5 AB |
499 | /* |
500 | * Create a timespec 'msec' milliseconds into the future | |
501 | */ | |
502 | static inline void make_timespec(struct timespec *tsp, long delta_msec) | |
503 | { | |
504 | struct timeval now; | |
505 | ||
506 | gettimeofday(&now, NULL); | |
507 | tsp->tv_sec = now.tv_sec; | |
508 | tsp->tv_nsec = 1000L * now.tv_usec; | |
509 | ||
510 | tsp->tv_nsec += (delta_msec * 1000000L); | |
511 | if (tsp->tv_nsec > 1000000000L) { | |
512 | long secs = tsp->tv_nsec / 1000000000L; | |
513 | ||
514 | tsp->tv_sec += secs; | |
515 | tsp->tv_nsec -= (secs * 1000000000L); | |
516 | } | |
517 | } | |
518 | ||
519 | /* | |
520 | * Add a timer to ensure wait ends | |
521 | */ | |
522 | static void t_pthread_cond_wait(pthread_cond_t *cond, pthread_mutex_t *mutex) | |
523 | { | |
524 | struct timespec ts; | |
525 | ||
526 | make_timespec(&ts, 50); | |
527 | pthread_cond_timedwait(cond, mutex, &ts); | |
528 | } | |
529 | ||
530 | static void unblock_tracers(void) | |
531 | { | |
532 | pthread_mutex_lock(&mt_mutex); | |
533 | tracers_run = 1; | |
534 | pthread_cond_broadcast(&mt_cond); | |
535 | pthread_mutex_unlock(&mt_mutex); | |
536 | } | |
537 | ||
538 | static void tracer_wait_unblock(struct tracer *tp) | |
539 | { | |
540 | pthread_mutex_lock(&mt_mutex); | |
541 | while (!tp->is_done && !tracers_run) | |
542 | pthread_cond_wait(&mt_cond, &mt_mutex); | |
543 | pthread_mutex_unlock(&mt_mutex); | |
544 | } | |
545 | ||
546 | static void tracer_signal_ready(struct tracer *tp, | |
547 | enum thread_status th_status, | |
548 | int status) | |
549 | { | |
550 | pthread_mutex_lock(&mt_mutex); | |
551 | tp->status = status; | |
552 | ||
553 | if (th_status == Th_running) | |
554 | nthreads_running++; | |
555 | else if (th_status == Th_error) | |
556 | nthreads_error++; | |
557 | else | |
558 | nthreads_leaving++; | |
559 | ||
560 | pthread_cond_signal(&mt_cond); | |
561 | pthread_mutex_unlock(&mt_mutex); | |
562 | } | |
563 | ||
564 | static void wait_tracers_ready(int ncpus_started) | |
565 | { | |
566 | pthread_mutex_lock(&mt_mutex); | |
567 | while ((nthreads_running + nthreads_error) < ncpus_started) | |
568 | t_pthread_cond_wait(&mt_cond, &mt_mutex); | |
569 | pthread_mutex_unlock(&mt_mutex); | |
570 | } | |
571 | ||
572 | static void wait_tracers_leaving(void) | |
573 | { | |
574 | pthread_mutex_lock(&mt_mutex); | |
575 | while (nthreads_leaving < nthreads_running) | |
576 | t_pthread_cond_wait(&mt_cond, &mt_mutex); | |
577 | pthread_mutex_unlock(&mt_mutex); | |
578 | } | |
579 | ||
3fe0b570 AB |
580 | static void init_mmap_info(struct mmap_info *mip) |
581 | { | |
582 | mip->buf_size = buf_size; | |
583 | mip->buf_nr = buf_nr; | |
584 | mip->pagesize = pagesize; | |
585 | } | |
b7106311 | 586 | |
3fe0b570 AB |
587 | static void net_close_connection(int *fd) |
588 | { | |
589 | shutdown(*fd, SHUT_RDWR); | |
590 | close(*fd); | |
591 | *fd = -1; | |
592 | } | |
ff11d54c | 593 | |
3fe0b570 AB |
594 | static void dpp_free(struct devpath *dpp) |
595 | { | |
596 | if (dpp->stats) | |
597 | free(dpp->stats); | |
598 | if (dpp->ios) | |
599 | free(dpp->ios); | |
600 | if (dpp->path) | |
601 | free(dpp->path); | |
602 | if (dpp->buts_name) | |
603 | free(dpp->buts_name); | |
604 | free(dpp); | |
605 | } | |
d0ca268b | 606 | |
3fe0b570 AB |
607 | static int lock_on_cpu(int cpu) |
608 | { | |
609 | cpu_set_t cpu_mask; | |
ff11d54c | 610 | |
3fe0b570 AB |
611 | CPU_ZERO(&cpu_mask); |
612 | CPU_SET(cpu, &cpu_mask); | |
f547a39d | 613 | if (sched_setaffinity(0, sizeof(cpu_mask), &cpu_mask) < 0) |
3fe0b570 | 614 | return errno; |
d0ca268b | 615 | |
3fe0b570 AB |
616 | return 0; |
617 | } | |
e7c9f3ff | 618 | |
3fe0b570 AB |
619 | static int increase_limit(int resource, rlim_t increase) |
620 | { | |
621 | struct rlimit rlim; | |
622 | int save_errno = errno; | |
8e86c98a | 623 | |
3fe0b570 AB |
624 | if (!getrlimit(resource, &rlim)) { |
625 | rlim.rlim_cur += increase; | |
626 | if (rlim.rlim_cur >= rlim.rlim_max) | |
627 | rlim.rlim_max = rlim.rlim_cur + increase; | |
72ca8801 | 628 | |
3fe0b570 AB |
629 | if (!setrlimit(resource, &rlim)) |
630 | return 1; | |
631 | } | |
99c1f5ab | 632 | |
3fe0b570 AB |
633 | errno = save_errno; |
634 | return 0; | |
635 | } | |
e0a1988b | 636 | |
3fe0b570 AB |
637 | static int handle_open_failure(void) |
638 | { | |
639 | if (errno == ENFILE || errno == EMFILE) | |
640 | return increase_limit(RLIMIT_NOFILE, 16); | |
641 | return 0; | |
642 | } | |
99c1f5ab | 643 | |
3fe0b570 AB |
644 | static int handle_mem_failure(size_t length) |
645 | { | |
646 | if (errno == ENFILE) | |
647 | return handle_open_failure(); | |
648 | else if (errno == ENOMEM) | |
649 | return increase_limit(RLIMIT_MEMLOCK, 2 * length); | |
650 | return 0; | |
651 | } | |
99c1f5ab | 652 | |
3fe0b570 AB |
653 | static FILE *my_fopen(const char *path, const char *mode) |
654 | { | |
655 | FILE *fp; | |
8e86c98a | 656 | |
3fe0b570 AB |
657 | do { |
658 | fp = fopen(path, mode); | |
659 | } while (fp == NULL && handle_open_failure()); | |
8e86c98a | 660 | |
3fe0b570 AB |
661 | return fp; |
662 | } | |
8e86c98a | 663 | |
3fe0b570 AB |
664 | static int my_open(const char *path, int flags) |
665 | { | |
666 | int fd; | |
8e86c98a | 667 | |
3fe0b570 AB |
668 | do { |
669 | fd = open(path, flags); | |
670 | } while (fd < 0 && handle_open_failure()); | |
e0a1988b | 671 | |
3fe0b570 AB |
672 | return fd; |
673 | } | |
ff11d54c | 674 | |
3fe0b570 AB |
675 | static int my_socket(int domain, int type, int protocol) |
676 | { | |
677 | int fd; | |
ff11d54c | 678 | |
3fe0b570 AB |
679 | do { |
680 | fd = socket(domain, type, protocol); | |
681 | } while (fd < 0 && handle_open_failure()); | |
8e86c98a | 682 | |
3fe0b570 AB |
683 | return fd; |
684 | } | |
685 | ||
d5302b03 AB |
686 | static int my_accept(int sockfd, struct sockaddr *addr, socklen_t *addrlen) |
687 | { | |
688 | int fd; | |
689 | ||
690 | do { | |
691 | fd = accept(sockfd, addr, addrlen); | |
692 | } while (fd < 0 && handle_open_failure()); | |
693 | ||
694 | return fd; | |
695 | } | |
696 | ||
3fe0b570 AB |
697 | static void *my_mmap(void *addr, size_t length, int prot, int flags, int fd, |
698 | off_t offset) | |
69dd57c2 | 699 | { |
3fe0b570 | 700 | void *new; |
69dd57c2 | 701 | |
3fe0b570 AB |
702 | do { |
703 | new = mmap(addr, length, prot, flags, fd, offset); | |
704 | } while (new == MAP_FAILED && handle_mem_failure(length)); | |
705 | ||
706 | return new; | |
707 | } | |
708 | ||
709 | static int my_mlock(const void *addr, size_t len) | |
710 | { | |
711 | int ret; | |
712 | ||
713 | do { | |
714 | ret = mlock(addr, len); | |
715 | } while (ret < 0 && handle_mem_failure(len)); | |
716 | ||
717 | return ret; | |
718 | } | |
719 | ||
055cc3e5 AB |
720 | static int setup_mmap(int fd, unsigned int maxlen, struct mmap_info *mip) |
721 | { | |
722 | if (mip->fs_off + maxlen > mip->fs_buf_len) { | |
723 | unsigned long nr = max(16, mip->buf_nr); | |
724 | ||
725 | if (mip->fs_buf) { | |
726 | munlock(mip->fs_buf, mip->fs_buf_len); | |
727 | munmap(mip->fs_buf, mip->fs_buf_len); | |
728 | mip->fs_buf = NULL; | |
729 | } | |
730 | ||
731 | mip->fs_off = mip->fs_size & (mip->pagesize - 1); | |
732 | mip->fs_buf_len = (nr * mip->buf_size) - mip->fs_off; | |
733 | mip->fs_max_size += mip->fs_buf_len; | |
734 | ||
735 | if (ftruncate(fd, mip->fs_max_size) < 0) { | |
736 | perror("setup_mmap: ftruncate"); | |
737 | return 1; | |
738 | } | |
739 | ||
740 | mip->fs_buf = my_mmap(NULL, mip->fs_buf_len, PROT_WRITE, | |
741 | MAP_SHARED, fd, | |
742 | mip->fs_size - mip->fs_off); | |
743 | if (mip->fs_buf == MAP_FAILED) { | |
744 | perror("setup_mmap: mmap"); | |
745 | return 1; | |
746 | } | |
747 | my_mlock(mip->fs_buf, mip->fs_buf_len); | |
748 | } | |
749 | ||
750 | return 0; | |
751 | } | |
752 | ||
3fe0b570 AB |
753 | static int __stop_trace(int fd) |
754 | { | |
755 | /* | |
756 | * Should be stopped, don't complain if it isn't | |
757 | */ | |
758 | ioctl(fd, BLKTRACESTOP); | |
759 | return ioctl(fd, BLKTRACETEARDOWN); | |
760 | } | |
761 | ||
762 | static int write_data(char *buf, int len) | |
763 | { | |
764 | int ret; | |
765 | ||
766 | rewrite: | |
767 | ret = fwrite(buf, len, 1, pfp); | |
768 | if (ferror(pfp) || ret != 1) { | |
769 | if (errno == EINTR) { | |
770 | clearerr(pfp); | |
771 | goto rewrite; | |
772 | } | |
773 | ||
774 | if (!piped_output || (errno != EPIPE && errno != EBADF)) { | |
775 | fprintf(stderr, "write(%d) failed: %d/%s\n", | |
776 | len, errno, strerror(errno)); | |
69dd57c2 | 777 | } |
3fe0b570 | 778 | goto err; |
69dd57c2 AB |
779 | } |
780 | ||
3fe0b570 | 781 | fflush(pfp); |
69dd57c2 | 782 | return 0; |
3fe0b570 AB |
783 | |
784 | err: | |
785 | clearerr(pfp); | |
786 | return 1; | |
69dd57c2 AB |
787 | } |
788 | ||
789 | /* | |
3fe0b570 | 790 | * Returns the number of bytes read (successfully) |
69dd57c2 | 791 | */ |
3fe0b570 | 792 | static int __net_recv_data(int fd, void *buf, unsigned int len) |
69dd57c2 | 793 | { |
3fe0b570 AB |
794 | unsigned int bytes_left = len; |
795 | ||
796 | while (bytes_left && !done) { | |
797 | int ret = recv(fd, buf, bytes_left, MSG_WAITALL); | |
798 | ||
799 | if (ret == 0) | |
800 | break; | |
801 | else if (ret < 0) { | |
5d65b5e6 AB |
802 | if (errno == EAGAIN) { |
803 | usleep(50); | |
804 | continue; | |
805 | } | |
806 | perror("server: net_recv_data: recv failed"); | |
055cc3e5 | 807 | break; |
3fe0b570 AB |
808 | } else { |
809 | buf += ret; | |
810 | bytes_left -= ret; | |
811 | } | |
812 | } | |
69dd57c2 | 813 | |
3fe0b570 | 814 | return len - bytes_left; |
69dd57c2 AB |
815 | } |
816 | ||
3fe0b570 | 817 | static int net_recv_data(int fd, void *buf, unsigned int len) |
8e86c98a | 818 | { |
3fe0b570 AB |
819 | return __net_recv_data(fd, buf, len); |
820 | } | |
7035d92d | 821 | |
3fe0b570 AB |
822 | /* |
823 | * Returns number of bytes written | |
824 | */ | |
825 | static int net_send_data(int fd, void *buf, unsigned int buf_len) | |
826 | { | |
827 | int ret; | |
828 | unsigned int bytes_left = buf_len; | |
829 | ||
830 | while (bytes_left) { | |
831 | ret = send(fd, buf, bytes_left, 0); | |
832 | if (ret < 0) { | |
833 | perror("send"); | |
834 | break; | |
835 | } | |
836 | ||
837 | buf += ret; | |
838 | bytes_left -= ret; | |
7035d92d JA |
839 | } |
840 | ||
3fe0b570 | 841 | return buf_len - bytes_left; |
8e86c98a JA |
842 | } |
843 | ||
3fe0b570 | 844 | static int net_send_header(int fd, int cpu, char *buts_name, int len) |
eb3c8108 | 845 | { |
3fe0b570 | 846 | struct blktrace_net_hdr hdr; |
eb3c8108 | 847 | |
3fe0b570 | 848 | memset(&hdr, 0, sizeof(hdr)); |
eb3c8108 | 849 | |
3fe0b570 AB |
850 | hdr.magic = BLK_IO_TRACE_MAGIC; |
851 | strncpy(hdr.buts_name, buts_name, sizeof(hdr.buts_name)); | |
852 | hdr.buts_name[sizeof(hdr.buts_name)-1] = '\0'; | |
853 | hdr.cpu = cpu; | |
854 | hdr.max_cpus = ncpus; | |
855 | hdr.len = len; | |
856 | hdr.cl_id = getpid(); | |
857 | hdr.buf_size = buf_size; | |
858 | hdr.buf_nr = buf_nr; | |
859 | hdr.page_size = pagesize; | |
eb3c8108 | 860 | |
3fe0b570 AB |
861 | return net_send_data(fd, &hdr, sizeof(hdr)) != sizeof(hdr); |
862 | } | |
eb3c8108 | 863 | |
3fe0b570 AB |
864 | static void net_send_open_close(int fd, int cpu, char *buts_name, int len) |
865 | { | |
866 | struct blktrace_net_hdr ret_hdr; | |
eb3c8108 | 867 | |
3fe0b570 AB |
868 | net_send_header(fd, cpu, buts_name, len); |
869 | net_recv_data(fd, &ret_hdr, sizeof(ret_hdr)); | |
870 | } | |
eb3c8108 | 871 | |
3fe0b570 AB |
872 | static void net_send_open(int fd, int cpu, char *buts_name) |
873 | { | |
874 | net_send_open_close(fd, cpu, buts_name, 0); | |
eb3c8108 JA |
875 | } |
876 | ||
3fe0b570 | 877 | static void net_send_close(int fd, char *buts_name, int drops) |
d0ca268b | 878 | { |
3fe0b570 AB |
879 | /* |
880 | * Overload CPU w/ number of drops | |
881 | * | |
882 | * XXX: Need to clear/set done around call - done=1 (which | |
883 | * is true here) stops reads from happening... :-( | |
884 | */ | |
885 | done = 0; | |
886 | net_send_open_close(fd, drops, buts_name, 1); | |
887 | done = 1; | |
888 | } | |
d0ca268b | 889 | |
3fe0b570 AB |
890 | static void ack_open_close(int fd, char *buts_name) |
891 | { | |
892 | net_send_header(fd, 0, buts_name, 2); | |
893 | } | |
d0ca268b | 894 | |
3fe0b570 AB |
895 | static void net_send_drops(int fd) |
896 | { | |
897 | struct list_head *p; | |
ed71a31e | 898 | |
3fe0b570 AB |
899 | __list_for_each(p, &devpaths) { |
900 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
901 | ||
902 | net_send_close(fd, dpp->buts_name, dpp->drops); | |
d0ca268b | 903 | } |
3fe0b570 | 904 | } |
d0ca268b | 905 | |
3fe0b570 AB |
906 | /* |
907 | * Returns: | |
055cc3e5 AB |
908 | * 0: "EOF" |
909 | * 1: OK | |
910 | * -1: Error | |
3fe0b570 AB |
911 | */ |
912 | static int net_get_header(struct cl_conn *nc, struct blktrace_net_hdr *bnh) | |
913 | { | |
914 | int bytes_read; | |
915 | int fl = fcntl(nc->fd, F_GETFL); | |
916 | ||
917 | fcntl(nc->fd, F_SETFL, fl | O_NONBLOCK); | |
918 | bytes_read = __net_recv_data(nc->fd, bnh, sizeof(*bnh)); | |
919 | fcntl(nc->fd, F_SETFL, fl & ~O_NONBLOCK); | |
920 | ||
921 | if (bytes_read == sizeof(*bnh)) | |
922 | return 1; | |
923 | else if (bytes_read == 0) | |
924 | return 0; | |
055cc3e5 AB |
925 | else |
926 | return -1; | |
d0ca268b JA |
927 | } |
928 | ||
e58f3937 | 929 | static int net_setup_addr(void) |
d0ca268b | 930 | { |
e58f3937 | 931 | struct sockaddr_in *addr = &hostname_addr; |
cf9208ea | 932 | |
e58f3937 AB |
933 | memset(addr, 0, sizeof(*addr)); |
934 | addr->sin_family = AF_INET; | |
935 | addr->sin_port = htons(net_port); | |
3fe0b570 | 936 | |
e58f3937 AB |
937 | if (inet_aton(hostname, &addr->sin_addr) != 1) { |
938 | struct hostent *hent; | |
939 | retry: | |
940 | hent = gethostbyname(hostname); | |
3fe0b570 | 941 | if (!hent) { |
e58f3937 AB |
942 | if (h_errno == TRY_AGAIN) { |
943 | usleep(100); | |
944 | goto retry; | |
945 | } else if (h_errno == NO_RECOVERY) { | |
946 | fprintf(stderr, "gethostbyname(%s)" | |
947 | "non-recoverable error encountered\n", | |
948 | hostname); | |
949 | } else { | |
950 | /* | |
951 | * HOST_NOT_FOUND, NO_ADDRESS or NO_DATA | |
952 | */ | |
953 | fprintf(stderr, "Host %s not found\n", | |
954 | hostname); | |
955 | } | |
3fe0b570 AB |
956 | return 1; |
957 | } | |
958 | ||
e58f3937 | 959 | memcpy(&addr->sin_addr, hent->h_addr, 4); |
3fe0b570 AB |
960 | strcpy(hostname, hent->h_name); |
961 | } | |
7035d92d | 962 | |
e58f3937 AB |
963 | return 0; |
964 | } | |
965 | ||
966 | static int net_setup_client(void) | |
967 | { | |
968 | int fd; | |
969 | struct sockaddr_in *addr = &hostname_addr; | |
970 | ||
3fe0b570 AB |
971 | fd = my_socket(AF_INET, SOCK_STREAM, 0); |
972 | if (fd < 0) { | |
973 | perror("client: socket"); | |
974 | return -1; | |
975 | } | |
cf9208ea | 976 | |
e58f3937 | 977 | if (connect(fd, (struct sockaddr *)addr, sizeof(*addr)) < 0) { |
3fe0b570 AB |
978 | if (errno == ECONNREFUSED) |
979 | fprintf(stderr, | |
980 | "\nclient: Connection to %s refused, " | |
981 | "perhaps the server is not started?\n\n", | |
982 | hostname); | |
983 | else | |
984 | perror("client: connect"); | |
055cc3e5 | 985 | |
3fe0b570 AB |
986 | close(fd); |
987 | return -1; | |
707b0914 | 988 | } |
3fe0b570 AB |
989 | |
990 | return fd; | |
d0ca268b JA |
991 | } |
992 | ||
3fe0b570 | 993 | static int open_client_connections(void) |
e7c9f3ff | 994 | { |
3fe0b570 | 995 | int cpu; |
e7c9f3ff | 996 | |
3fe0b570 AB |
997 | cl_fds = calloc(ncpus, sizeof(*cl_fds)); |
998 | for (cpu = 0; cpu < ncpus; cpu++) { | |
999 | cl_fds[cpu] = net_setup_client(); | |
1000 | if (cl_fds[cpu] < 0) | |
1001 | goto err; | |
eb3c8108 | 1002 | } |
3fe0b570 AB |
1003 | return 0; |
1004 | ||
1005 | err: | |
1006 | while (cpu > 0) | |
1007 | close(cl_fds[cpu--]); | |
1008 | free(cl_fds); | |
1009 | return 1; | |
e7c9f3ff NS |
1010 | } |
1011 | ||
3fe0b570 | 1012 | static void close_client_connections(void) |
eb3c8108 | 1013 | { |
3fe0b570 AB |
1014 | if (cl_fds) { |
1015 | int cpu, *fdp; | |
eb3c8108 | 1016 | |
3fe0b570 AB |
1017 | for (cpu = 0, fdp = cl_fds; cpu < ncpus; cpu++, fdp++) { |
1018 | if (*fdp >= 0) { | |
1019 | net_send_drops(*fdp); | |
1020 | net_close_connection(fdp); | |
1021 | } | |
7934e668 | 1022 | } |
3fe0b570 | 1023 | free(cl_fds); |
ff11d54c | 1024 | } |
eb3c8108 JA |
1025 | } |
1026 | ||
3fe0b570 | 1027 | static void setup_buts(void) |
d0ca268b | 1028 | { |
3fe0b570 | 1029 | struct list_head *p; |
bbabf03a | 1030 | |
3fe0b570 AB |
1031 | __list_for_each(p, &devpaths) { |
1032 | struct blk_user_trace_setup buts; | |
1033 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
ae9f71b3 | 1034 | |
3fe0b570 AB |
1035 | memset(&buts, 0, sizeof(buts)); |
1036 | buts.buf_size = buf_size; | |
1037 | buts.buf_nr = buf_nr; | |
1038 | buts.act_mask = act_mask; | |
1039 | ||
055cc3e5 | 1040 | if (ioctl(dpp->fd, BLKTRACESETUP, &buts) >= 0) { |
3b552a2d AB |
1041 | dpp->ncpus = ncpus; |
1042 | dpp->buts_name = strdup(buts.name); | |
1043 | if (dpp->stats) | |
1044 | free(dpp->stats); | |
1045 | dpp->stats = calloc(dpp->ncpus, sizeof(*dpp->stats)); | |
1046 | memset(dpp->stats, 0, dpp->ncpus * sizeof(*dpp->stats)); | |
055cc3e5 AB |
1047 | } else |
1048 | fprintf(stderr, "BLKTRACESETUP(2) %s failed: %d/%s\n", | |
1049 | dpp->path, errno, strerror(errno)); | |
3b552a2d AB |
1050 | } |
1051 | } | |
1052 | ||
1053 | static void start_buts(void) | |
1054 | { | |
1055 | struct list_head *p; | |
1056 | ||
1057 | __list_for_each(p, &devpaths) { | |
1058 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
1059 | ||
1060 | if (ioctl(dpp->fd, BLKTRACESTART) < 0) { | |
3fe0b570 AB |
1061 | fprintf(stderr, "BLKTRACESTART %s failed: %d/%s\n", |
1062 | dpp->path, errno, strerror(errno)); | |
bbabf03a | 1063 | } |
3fe0b570 | 1064 | } |
8a43bac5 JA |
1065 | } |
1066 | ||
3fe0b570 | 1067 | static int get_drops(struct devpath *dpp) |
8e86c98a | 1068 | { |
3fe0b570 AB |
1069 | int fd, drops = 0; |
1070 | char fn[MAXPATHLEN + 64], tmp[256]; | |
8e86c98a | 1071 | |
3fe0b570 AB |
1072 | snprintf(fn, sizeof(fn), "%s/block/%s/dropped", debugfs_path, |
1073 | dpp->buts_name); | |
8e86c98a | 1074 | |
3fe0b570 AB |
1075 | fd = my_open(fn, O_RDONLY); |
1076 | if (fd < 0) { | |
1077 | /* | |
1078 | * This may be ok: the kernel may not support | |
1079 | * dropped counts. | |
1080 | */ | |
1081 | if (errno != ENOENT) | |
1082 | fprintf(stderr, "Could not open %s: %d/%s\n", | |
1083 | fn, errno, strerror(errno)); | |
1084 | return 0; | |
1085 | } else if (read(fd, tmp, sizeof(tmp)) < 0) { | |
1086 | fprintf(stderr, "Could not read %s: %d/%s\n", | |
1087 | fn, errno, strerror(errno)); | |
1088 | } else | |
1089 | drops = atoi(tmp); | |
1090 | close(fd); | |
8e86c98a | 1091 | |
3fe0b570 | 1092 | return drops; |
8e86c98a JA |
1093 | } |
1094 | ||
3fe0b570 | 1095 | static void get_all_drops(void) |
a3e4d330 | 1096 | { |
3fe0b570 | 1097 | struct list_head *p; |
21f55651 | 1098 | |
3fe0b570 AB |
1099 | __list_for_each(p, &devpaths) { |
1100 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
055cc3e5 | 1101 | |
3fe0b570 | 1102 | dpp->drops = get_drops(dpp); |
21f55651 | 1103 | } |
9db17354 | 1104 | } |
eb3c8108 | 1105 | |
3fe0b570 | 1106 | static inline struct trace_buf *alloc_trace_buf(int cpu, int bufsize) |
9db17354 | 1107 | { |
3fe0b570 | 1108 | struct trace_buf *tbp; |
21f55651 | 1109 | |
3fe0b570 AB |
1110 | tbp = malloc(sizeof(*tbp) + bufsize); |
1111 | INIT_LIST_HEAD(&tbp->head); | |
1112 | tbp->len = 0; | |
1113 | tbp->buf = (void *)(tbp + 1); | |
1114 | tbp->cpu = cpu; | |
1115 | tbp->dpp = NULL; /* Will be set when tbp is added */ | |
21f55651 | 1116 | |
3fe0b570 | 1117 | return tbp; |
a3e4d330 JA |
1118 | } |
1119 | ||
3fe0b570 | 1120 | static void free_tracer_heads(struct devpath *dpp) |
b7106311 | 1121 | { |
3fe0b570 AB |
1122 | int cpu; |
1123 | struct tracer_devpath_head *hd; | |
b7106311 | 1124 | |
3fe0b570 AB |
1125 | for (cpu = 0, hd = dpp->heads; cpu < ncpus; cpu++, hd++) { |
1126 | if (hd->prev) | |
1127 | free(hd->prev); | |
055cc3e5 | 1128 | |
3fe0b570 AB |
1129 | pthread_mutex_destroy(&hd->mutex); |
1130 | } | |
1131 | free(dpp->heads); | |
1132 | } | |
b7106311 | 1133 | |
3fe0b570 AB |
1134 | static int setup_tracer_devpaths(void) |
1135 | { | |
1136 | struct list_head *p; | |
b7106311 | 1137 | |
3fe0b570 AB |
1138 | if (net_client_use_send()) |
1139 | if (open_client_connections()) | |
1140 | return 1; | |
b7106311 | 1141 | |
3fe0b570 AB |
1142 | __list_for_each(p, &devpaths) { |
1143 | int cpu; | |
1144 | struct tracer_devpath_head *hd; | |
1145 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
b7106311 | 1146 | |
3fe0b570 AB |
1147 | dpp->heads = calloc(ncpus, sizeof(struct tracer_devpath_head)); |
1148 | for (cpu = 0, hd = dpp->heads; cpu < ncpus; cpu++, hd++) { | |
1149 | INIT_LIST_HEAD(&hd->head); | |
1150 | pthread_mutex_init(&hd->mutex, NULL); | |
1151 | hd->prev = NULL; | |
1152 | } | |
b7106311 JA |
1153 | } |
1154 | ||
3fe0b570 | 1155 | return 0; |
b7106311 JA |
1156 | } |
1157 | ||
3fe0b570 AB |
1158 | static inline void add_trace_buf(struct devpath *dpp, int cpu, |
1159 | struct trace_buf **tbpp) | |
18eed2a7 | 1160 | { |
3fe0b570 AB |
1161 | struct trace_buf *tbp = *tbpp; |
1162 | struct tracer_devpath_head *hd = &dpp->heads[cpu]; | |
18eed2a7 | 1163 | |
3fe0b570 | 1164 | tbp->dpp = dpp; |
2f064793 | 1165 | |
3fe0b570 AB |
1166 | pthread_mutex_lock(&hd->mutex); |
1167 | list_add_tail(&tbp->head, &hd->head); | |
1168 | pthread_mutex_unlock(&hd->mutex); | |
18eed2a7 | 1169 | |
3fe0b570 | 1170 | *tbpp = alloc_trace_buf(cpu, buf_size); |
18eed2a7 JA |
1171 | } |
1172 | ||
3fe0b570 | 1173 | static inline void incr_entries(int entries_handled) |
a3e4d330 | 1174 | { |
3fe0b570 AB |
1175 | pthread_mutex_lock(&dp_mutex); |
1176 | if (dp_entries == 0) | |
1177 | pthread_cond_signal(&dp_cond); | |
1178 | dp_entries += entries_handled; | |
1179 | pthread_mutex_unlock(&dp_mutex); | |
a3e4d330 JA |
1180 | } |
1181 | ||
055cc3e5 AB |
1182 | static void decr_entries(int handled) |
1183 | { | |
1184 | pthread_mutex_lock(&dp_mutex); | |
1185 | dp_entries -= handled; | |
1186 | pthread_mutex_unlock(&dp_mutex); | |
1187 | } | |
1188 | ||
1189 | static int wait_empty_entries(void) | |
1190 | { | |
1191 | pthread_mutex_lock(&dp_mutex); | |
1192 | while (!done && dp_entries == 0) | |
1193 | t_pthread_cond_wait(&dp_cond, &dp_mutex); | |
1194 | pthread_mutex_unlock(&dp_mutex); | |
1195 | ||
1196 | return !done; | |
1197 | } | |
1198 | ||
3fe0b570 | 1199 | static int add_devpath(char *path) |
8e86c98a | 1200 | { |
3fe0b570 AB |
1201 | int fd; |
1202 | struct devpath *dpp; | |
1203 | ||
8e86c98a | 1204 | /* |
3fe0b570 | 1205 | * Verify device is valid before going too far |
8e86c98a | 1206 | */ |
3fe0b570 AB |
1207 | fd = my_open(path, O_RDONLY | O_NONBLOCK); |
1208 | if (fd < 0) { | |
1209 | fprintf(stderr, "Invalid path %s specified: %d/%s\n", | |
1210 | path, errno, strerror(errno)); | |
1211 | return 1; | |
1212 | } | |
8e86c98a | 1213 | |
3fe0b570 AB |
1214 | dpp = malloc(sizeof(*dpp)); |
1215 | memset(dpp, 0, sizeof(*dpp)); | |
1216 | dpp->path = strdup(path); | |
1217 | dpp->fd = fd; | |
1218 | dpp->idx = ndevs++; | |
1219 | list_add_tail(&dpp->head, &devpaths); | |
8e86c98a | 1220 | |
3fe0b570 | 1221 | return 0; |
8e86c98a JA |
1222 | } |
1223 | ||
3fe0b570 | 1224 | static void rel_devpaths(void) |
a3e4d330 | 1225 | { |
3fe0b570 | 1226 | struct list_head *p, *q; |
a3e4d330 | 1227 | |
3fe0b570 AB |
1228 | list_for_each_safe(p, q, &devpaths) { |
1229 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
a3e4d330 | 1230 | |
3fe0b570 AB |
1231 | list_del(&dpp->head); |
1232 | __stop_trace(dpp->fd); | |
1233 | close(dpp->fd); | |
a3e4d330 | 1234 | |
3fe0b570 AB |
1235 | if (dpp->heads) |
1236 | free_tracer_heads(dpp); | |
a3e4d330 | 1237 | |
3fe0b570 AB |
1238 | dpp_free(dpp); |
1239 | ndevs--; | |
b7106311 | 1240 | } |
8e86c98a | 1241 | } |
b7106311 | 1242 | |
3fe0b570 | 1243 | static int flush_subbuf_net(struct trace_buf *tbp) |
8e86c98a | 1244 | { |
3fe0b570 AB |
1245 | int fd = cl_fds[tbp->cpu]; |
1246 | struct devpath *dpp = tbp->dpp; | |
b7106311 | 1247 | |
3fe0b570 AB |
1248 | if (net_send_header(fd, tbp->cpu, dpp->buts_name, tbp->len)) |
1249 | return 1; | |
055cc3e5 | 1250 | else if (net_send_data(fd, tbp->buf, tbp->len) != tbp->len) |
3fe0b570 | 1251 | return 1; |
a3e4d330 | 1252 | |
8e86c98a | 1253 | return 0; |
a3e4d330 JA |
1254 | } |
1255 | ||
3fe0b570 AB |
1256 | static int |
1257 | handle_list_net(__attribute__((__unused__))struct tracer_devpath_head *hd, | |
1258 | struct list_head *list) | |
8e86c98a | 1259 | { |
3fe0b570 AB |
1260 | struct trace_buf *tbp; |
1261 | struct list_head *p, *q; | |
1262 | int entries_handled = 0; | |
8e86c98a | 1263 | |
3fe0b570 AB |
1264 | list_for_each_safe(p, q, list) { |
1265 | tbp = list_entry(p, struct trace_buf, head); | |
8e86c98a | 1266 | |
3fe0b570 AB |
1267 | list_del(&tbp->head); |
1268 | entries_handled++; | |
6a752c90 | 1269 | |
3fe0b570 AB |
1270 | if (cl_fds[tbp->cpu] >= 0) { |
1271 | if (flush_subbuf_net(tbp)) { | |
1272 | close(cl_fds[tbp->cpu]); | |
1273 | cl_fds[tbp->cpu] = -1; | |
1274 | } | |
1275 | } | |
7ab2f837 | 1276 | |
3fe0b570 | 1277 | free(tbp); |
7934e668 JA |
1278 | } |
1279 | ||
3fe0b570 | 1280 | return entries_handled; |
6a752c90 JA |
1281 | } |
1282 | ||
055cc3e5 AB |
1283 | /* |
1284 | * Tack 'tbp's buf onto the tail of 'prev's buf | |
1285 | */ | |
1286 | static struct trace_buf *tb_combine(struct trace_buf *prev, | |
1287 | struct trace_buf *tbp) | |
1288 | { | |
1289 | unsigned long tot_len; | |
1290 | ||
1291 | tot_len = prev->len + tbp->len; | |
1292 | if (tot_len > buf_size) { | |
1293 | /* | |
1294 | * tbp->head isn't connected (it was 'prev' | |
1295 | * so it had been taken off of the list | |
1296 | * before). Therefore, we can realloc | |
1297 | * the whole structures, as the other fields | |
1298 | * are "static". | |
1299 | */ | |
1300 | prev = realloc(prev->buf, sizeof(*prev) + tot_len); | |
1301 | prev->buf = (void *)(prev + 1); | |
1302 | } | |
1303 | ||
1304 | memcpy(prev->buf + prev->len, tbp->buf, tbp->len); | |
1305 | prev->len = tot_len; | |
1306 | ||
1307 | free(tbp); | |
1308 | return prev; | |
1309 | } | |
1310 | ||
3fe0b570 AB |
1311 | static int handle_list_file(struct tracer_devpath_head *hd, |
1312 | struct list_head *list) | |
f6fead25 | 1313 | { |
3fe0b570 AB |
1314 | int off, t_len, nevents; |
1315 | struct blk_io_trace *t; | |
1316 | struct list_head *p, *q; | |
1317 | int entries_handled = 0; | |
1318 | struct trace_buf *tbp, *prev; | |
11629347 | 1319 | |
3fe0b570 AB |
1320 | prev = hd->prev; |
1321 | list_for_each_safe(p, q, list) { | |
1322 | tbp = list_entry(p, struct trace_buf, head); | |
1323 | list_del(&tbp->head); | |
1324 | entries_handled++; | |
18eed2a7 | 1325 | |
3fe0b570 AB |
1326 | /* |
1327 | * If there was some leftover before, tack this new | |
1328 | * entry onto the tail of the previous one. | |
1329 | */ | |
055cc3e5 AB |
1330 | if (prev) |
1331 | tbp = tb_combine(prev, tbp); | |
ff11d54c | 1332 | |
3fe0b570 AB |
1333 | /* |
1334 | * See how many whole traces there are - send them | |
1335 | * all out in one go. | |
1336 | */ | |
1337 | off = 0; | |
1338 | nevents = 0; | |
1339 | while (off + (int)sizeof(*t) <= tbp->len) { | |
1340 | t = (struct blk_io_trace *)(tbp->buf + off); | |
1341 | t_len = sizeof(*t) + t->pdu_len; | |
1342 | if (off + t_len > tbp->len) | |
1343 | break; | |
ff11d54c | 1344 | |
3fe0b570 AB |
1345 | off += t_len; |
1346 | nevents++; | |
1347 | } | |
1348 | if (nevents) | |
1349 | pdc_nev_update(tbp->dpp, tbp->cpu, nevents); | |
4aeec019 | 1350 | |
3fe0b570 AB |
1351 | /* |
1352 | * Write any full set of traces, any remaining data is kept | |
1353 | * for the next pass. | |
1354 | */ | |
1355 | if (off) { | |
055cc3e5 | 1356 | if (write_data(tbp->buf, off) || off == tbp->len) { |
3fe0b570 | 1357 | free(tbp); |
055cc3e5 AB |
1358 | prev = NULL; |
1359 | } | |
3fe0b570 AB |
1360 | else { |
1361 | /* | |
1362 | * Move valid data to beginning of buffer | |
1363 | */ | |
1364 | tbp->len -= off; | |
1365 | memmove(tbp->buf, tbp->buf + off, tbp->len); | |
1366 | prev = tbp; | |
1367 | } | |
1368 | } else | |
1369 | prev = tbp; | |
ff11d54c | 1370 | } |
3fe0b570 | 1371 | hd->prev = prev; |
ff11d54c | 1372 | |
3fe0b570 | 1373 | return entries_handled; |
ff11d54c TZ |
1374 | } |
1375 | ||
3fe0b570 | 1376 | static void __process_trace_bufs(void) |
8a43bac5 | 1377 | { |
3fe0b570 AB |
1378 | int cpu; |
1379 | struct list_head *p; | |
1380 | struct list_head list; | |
1381 | int handled = 0; | |
1382 | ||
1383 | __list_for_each(p, &devpaths) { | |
1384 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
1385 | struct tracer_devpath_head *hd = dpp->heads; | |
1386 | ||
1387 | for (cpu = 0; cpu < ncpus; cpu++, hd++) { | |
1388 | pthread_mutex_lock(&hd->mutex); | |
1389 | if (list_empty(&hd->head)) { | |
1390 | pthread_mutex_unlock(&hd->mutex); | |
1391 | continue; | |
1392 | } | |
8a43bac5 | 1393 | |
3fe0b570 AB |
1394 | list_replace_init(&hd->head, &list); |
1395 | pthread_mutex_unlock(&hd->mutex); | |
6480258a | 1396 | |
3fe0b570 AB |
1397 | handled += handle_list(hd, &list); |
1398 | } | |
d0ca268b JA |
1399 | } |
1400 | ||
055cc3e5 AB |
1401 | if (handled) |
1402 | decr_entries(handled); | |
8a43bac5 JA |
1403 | } |
1404 | ||
3fe0b570 | 1405 | static void process_trace_bufs(void) |
8a43bac5 | 1406 | { |
055cc3e5 | 1407 | while (wait_empty_entries()) |
3fe0b570 | 1408 | __process_trace_bufs(); |
3fe0b570 | 1409 | } |
3a9d6c13 | 1410 | |
3fe0b570 AB |
1411 | static void clean_trace_bufs(void) |
1412 | { | |
3a9d6c13 | 1413 | /* |
3fe0b570 AB |
1414 | * No mutex needed here: we're only reading from the lists, |
1415 | * tracers are done | |
3a9d6c13 | 1416 | */ |
3fe0b570 AB |
1417 | while (dp_entries) |
1418 | __process_trace_bufs(); | |
1419 | } | |
4b5db44a | 1420 | |
3fe0b570 AB |
1421 | static inline void read_err(int cpu, char *ifn) |
1422 | { | |
1423 | if (errno != EAGAIN) | |
1424 | fprintf(stderr, "Thread %d failed read of %s: %d/%s\n", | |
1425 | cpu, ifn, errno, strerror(errno)); | |
4b5db44a JA |
1426 | } |
1427 | ||
3fe0b570 | 1428 | static int net_sendfile(struct io_info *iop) |
d5396421 | 1429 | { |
3fe0b570 | 1430 | int ret; |
d5396421 | 1431 | |
3fe0b570 AB |
1432 | ret = sendfile(iop->ofd, iop->ifd, NULL, iop->ready); |
1433 | if (ret < 0) { | |
1434 | perror("sendfile"); | |
1435 | return 1; | |
1436 | } else if (ret < (int)iop->ready) { | |
1437 | fprintf(stderr, "short sendfile send (%d of %d)\n", | |
1438 | ret, iop->ready); | |
1439 | return 1; | |
1440 | } | |
91816d54 | 1441 | |
9db17354 | 1442 | return 0; |
91816d54 JA |
1443 | } |
1444 | ||
3fe0b570 | 1445 | static inline int net_sendfile_data(struct tracer *tp, struct io_info *iop) |
d0ca268b | 1446 | { |
3fe0b570 | 1447 | struct devpath *dpp = iop->dpp; |
d0ca268b | 1448 | |
3fe0b570 AB |
1449 | if (net_send_header(iop->ofd, tp->cpu, dpp->buts_name, iop->ready)) |
1450 | return 1; | |
1451 | return net_sendfile(iop); | |
1452 | } | |
d0ca268b | 1453 | |
3fe0b570 | 1454 | static int fill_ofname(struct io_info *iop, int cpu) |
8e86c98a | 1455 | { |
3fe0b570 | 1456 | int len; |
e3bf54d8 | 1457 | struct stat sb; |
3fe0b570 | 1458 | char *dst = iop->ofn; |
8e86c98a JA |
1459 | |
1460 | if (output_dir) | |
3fe0b570 | 1461 | len = snprintf(iop->ofn, sizeof(iop->ofn), "%s/", output_dir); |
dd870ef6 | 1462 | else |
3fe0b570 | 1463 | len = snprintf(iop->ofn, sizeof(iop->ofn), "./"); |
8e86c98a | 1464 | |
e3bf54d8 | 1465 | if (net_mode == Net_server) { |
3fe0b570 | 1466 | struct cl_conn *nc = iop->nc; |
e0a1988b | 1467 | |
3fe0b570 AB |
1468 | len += sprintf(dst + len, "%s-", nc->ch->hostname); |
1469 | len += strftime(dst + len, 64, "%F-%T/", | |
1470 | gmtime(&iop->dpp->cl_connect_time)); | |
e3bf54d8 JA |
1471 | } |
1472 | ||
3fe0b570 | 1473 | if (stat(iop->ofn, &sb) < 0) { |
e3bf54d8 | 1474 | if (errno != ENOENT) { |
3fe0b570 AB |
1475 | fprintf(stderr, |
1476 | "Destination dir %s stat failed: %d/%s\n", | |
1477 | iop->ofn, errno, strerror(errno)); | |
e3bf54d8 JA |
1478 | return 1; |
1479 | } | |
60886290 JM |
1480 | /* |
1481 | * There is no synchronization between multiple threads | |
1482 | * trying to create the directory at once. It's harmless | |
1483 | * to let them try, so just detect the problem and move on. | |
1484 | */ | |
1485 | if (mkdir(iop->ofn, 0755) < 0 && errno != EEXIST) { | |
3fe0b570 AB |
1486 | fprintf(stderr, |
1487 | "Destination dir %s can't be made: %d/%s\n", | |
1488 | iop->ofn, errno, strerror(errno)); | |
e3bf54d8 JA |
1489 | return 1; |
1490 | } | |
1491 | } | |
1492 | ||
8e86c98a | 1493 | if (output_name) |
3fe0b570 AB |
1494 | snprintf(iop->ofn + len, sizeof(iop->ofn), "%s.blktrace.%d", |
1495 | output_name, cpu); | |
8e86c98a | 1496 | else |
3fe0b570 AB |
1497 | snprintf(iop->ofn + len, sizeof(iop->ofn), "%s.blktrace.%d", |
1498 | iop->dpp->buts_name, cpu); | |
e3bf54d8 JA |
1499 | |
1500 | return 0; | |
8e86c98a JA |
1501 | } |
1502 | ||
3fe0b570 | 1503 | static int set_vbuf(struct io_info *iop, int mode, size_t size) |
0cc7d25e | 1504 | { |
3fe0b570 AB |
1505 | iop->obuf = malloc(size); |
1506 | if (setvbuf(iop->ofp, iop->obuf, mode, size) < 0) { | |
1507 | fprintf(stderr, "setvbuf(%s, %d) failed: %d/%s\n", | |
1508 | iop->dpp->path, (int)size, errno, | |
1509 | strerror(errno)); | |
1510 | free(iop->obuf); | |
ddf22842 JA |
1511 | return 1; |
1512 | } | |
d5396421 | 1513 | |
ddf22842 JA |
1514 | return 0; |
1515 | } | |
007c233c | 1516 | |
3fe0b570 | 1517 | static int iop_open(struct io_info *iop, int cpu) |
ddf22842 | 1518 | { |
3fe0b570 AB |
1519 | iop->ofd = -1; |
1520 | if (fill_ofname(iop, cpu)) | |
1521 | return 1; | |
0cc7d25e | 1522 | |
3fe0b570 AB |
1523 | iop->ofp = my_fopen(iop->ofn, "w+"); |
1524 | if (iop->ofp == NULL) { | |
1525 | fprintf(stderr, "Open output file %s failed: %d/%s\n", | |
1526 | iop->ofn, errno, strerror(errno)); | |
1527 | return 1; | |
1528 | } | |
055cc3e5 | 1529 | |
3fe0b570 AB |
1530 | if (set_vbuf(iop, _IOLBF, FILE_VBUF_SIZE)) { |
1531 | fprintf(stderr, "set_vbuf for file %s failed: %d/%s\n", | |
1532 | iop->ofn, errno, strerror(errno)); | |
1533 | fclose(iop->ofp); | |
1534 | return 1; | |
d0ca268b JA |
1535 | } |
1536 | ||
3fe0b570 | 1537 | iop->ofd = fileno(iop->ofp); |
e7c9f3ff | 1538 | return 0; |
d0ca268b JA |
1539 | } |
1540 | ||
df81fdb5 AB |
1541 | static void close_iop(struct io_info *iop) |
1542 | { | |
1543 | struct mmap_info *mip = &iop->mmap_info; | |
1544 | ||
1545 | if (mip->fs_buf) | |
1546 | munmap(mip->fs_buf, mip->fs_buf_len); | |
1547 | ||
1548 | if (!piped_output) { | |
1549 | if (ftruncate(fileno(iop->ofp), mip->fs_size) < 0) { | |
1550 | fprintf(stderr, | |
1551 | "Ignoring err: ftruncate(%s): %d/%s\n", | |
1552 | iop->ofn, errno, strerror(errno)); | |
1553 | } | |
1554 | } | |
1555 | ||
1556 | if (iop->ofp) | |
1557 | fclose(iop->ofp); | |
1558 | if (iop->obuf) | |
1559 | free(iop->obuf); | |
1560 | } | |
1561 | ||
1562 | static void close_ios(struct tracer *tp) | |
1563 | { | |
1564 | while (tp->nios > 0) { | |
1565 | struct io_info *iop = &tp->ios[--tp->nios]; | |
1566 | ||
1567 | iop->dpp->drops = get_drops(iop->dpp); | |
1568 | if (iop->ifd >= 0) | |
1569 | close(iop->ifd); | |
1570 | ||
1571 | if (iop->ofp) | |
1572 | close_iop(iop); | |
1573 | else if (iop->ofd >= 0) { | |
1574 | struct devpath *dpp = iop->dpp; | |
1575 | ||
1576 | net_send_close(iop->ofd, dpp->buts_name, dpp->drops); | |
1577 | net_close_connection(&iop->ofd); | |
1578 | } | |
1579 | } | |
1580 | ||
1581 | free(tp->ios); | |
1582 | free(tp->pfds); | |
1583 | } | |
1584 | ||
3fe0b570 | 1585 | static int open_ios(struct tracer *tp) |
3aabcd89 | 1586 | { |
3fe0b570 AB |
1587 | struct pollfd *pfd; |
1588 | struct io_info *iop; | |
1589 | struct list_head *p; | |
1590 | ||
1591 | tp->ios = calloc(ndevs, sizeof(struct io_info)); | |
3fe0b570 | 1592 | memset(tp->ios, 0, ndevs * sizeof(struct io_info)); |
055cc3e5 AB |
1593 | |
1594 | tp->pfds = calloc(ndevs, sizeof(struct pollfd)); | |
3fe0b570 AB |
1595 | memset(tp->pfds, 0, ndevs * sizeof(struct pollfd)); |
1596 | ||
1597 | tp->nios = 0; | |
1598 | iop = tp->ios; | |
1599 | pfd = tp->pfds; | |
1600 | __list_for_each(p, &devpaths) { | |
1601 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
1602 | ||
1603 | iop->dpp = dpp; | |
1604 | iop->ofd = -1; | |
1605 | snprintf(iop->ifn, sizeof(iop->ifn), "%s/block/%s/trace%d", | |
1606 | debugfs_path, dpp->buts_name, tp->cpu); | |
1607 | ||
1608 | iop->ifd = my_open(iop->ifn, O_RDONLY | O_NONBLOCK); | |
1609 | if (iop->ifd < 0) { | |
1610 | fprintf(stderr, "Thread %d failed open %s: %d/%s\n", | |
1611 | tp->cpu, iop->ifn, errno, strerror(errno)); | |
1612 | return 1; | |
1613 | } | |
1614 | ||
1615 | init_mmap_info(&iop->mmap_info); | |
1616 | ||
1617 | pfd->fd = iop->ifd; | |
1618 | pfd->events = POLLIN; | |
1619 | ||
1620 | if (piped_output) | |
1621 | ; | |
1622 | else if (net_client_use_sendfile()) { | |
1623 | iop->ofd = net_setup_client(); | |
1624 | if (iop->ofd < 0) | |
1625 | goto err; | |
1626 | net_send_open(iop->ofd, tp->cpu, dpp->buts_name); | |
1627 | } else if (net_mode == Net_none) { | |
1628 | if (iop_open(iop, tp->cpu)) | |
1629 | goto err; | |
1630 | } else { | |
1631 | /* | |
1632 | * This ensures that the server knows about all | |
1633 | * connections & devices before _any_ closes | |
1634 | */ | |
1635 | net_send_open(cl_fds[tp->cpu], tp->cpu, dpp->buts_name); | |
1636 | } | |
007c233c | 1637 | |
3fe0b570 AB |
1638 | pfd++; |
1639 | iop++; | |
1640 | tp->nios++; | |
9db17354 | 1641 | } |
3aabcd89 | 1642 | |
3fe0b570 | 1643 | return 0; |
72ca8801 | 1644 | |
3fe0b570 AB |
1645 | err: |
1646 | close(iop->ifd); /* tp->nios _not_ bumped */ | |
df81fdb5 | 1647 | close_ios(tp); |
3fe0b570 | 1648 | return 1; |
e7c9f3ff NS |
1649 | } |
1650 | ||
3fe0b570 | 1651 | static int handle_pfds_file(struct tracer *tp, int nevs, int force_read) |
e7c9f3ff | 1652 | { |
3fe0b570 AB |
1653 | struct mmap_info *mip; |
1654 | int i, ret, nentries = 0; | |
1655 | struct pollfd *pfd = tp->pfds; | |
1656 | struct io_info *iop = tp->ios; | |
1657 | ||
1658 | for (i = 0; nevs > 0 && i < ndevs; i++, pfd++, iop++) { | |
1659 | if (pfd->revents & POLLIN || force_read) { | |
1660 | mip = &iop->mmap_info; | |
1661 | ||
1662 | ret = setup_mmap(iop->ofd, buf_size, mip); | |
1663 | if (ret < 0) { | |
1664 | pfd->events = 0; | |
1665 | break; | |
1666 | } | |
428683db | 1667 | |
3fe0b570 AB |
1668 | ret = read(iop->ifd, mip->fs_buf + mip->fs_off, |
1669 | buf_size); | |
1670 | if (ret > 0) { | |
1671 | pdc_dr_update(iop->dpp, tp->cpu, ret); | |
1672 | mip->fs_size += ret; | |
1673 | mip->fs_off += ret; | |
1674 | nentries++; | |
1675 | } else if (ret == 0) { | |
1676 | /* | |
1677 | * Short reads after we're done stop us | |
1678 | * from trying reads. | |
1679 | */ | |
1680 | if (tp->is_done) | |
1681 | clear_events(pfd); | |
1682 | } else { | |
1683 | read_err(tp->cpu, iop->ifn); | |
1684 | if (errno != EAGAIN || tp->is_done) | |
1685 | clear_events(pfd); | |
1686 | } | |
1687 | nevs--; | |
e7c9f3ff | 1688 | } |
e7c9f3ff | 1689 | } |
56070ea4 | 1690 | |
3fe0b570 | 1691 | return nentries; |
e7c9f3ff | 1692 | } |
52724a0e | 1693 | |
055cc3e5 AB |
1694 | static int handle_pfds_netclient(struct tracer *tp, int nevs, int force_read) |
1695 | { | |
1696 | struct stat sb; | |
1697 | int i, nentries = 0; | |
1698 | struct pdc_stats *sp; | |
1699 | struct pollfd *pfd = tp->pfds; | |
1700 | struct io_info *iop = tp->ios; | |
1701 | ||
1702 | for (i = 0; i < ndevs; i++, pfd++, iop++, sp++) { | |
1703 | if (pfd->revents & POLLIN || force_read) { | |
1704 | if (fstat(iop->ifd, &sb) < 0) { | |
1705 | perror(iop->ifn); | |
1706 | pfd->events = 0; | |
1707 | } else if (sb.st_size > (off_t)iop->data_queued) { | |
1708 | iop->ready = sb.st_size - iop->data_queued; | |
1709 | iop->data_queued = sb.st_size; | |
1710 | ||
1711 | if (!net_sendfile_data(tp, iop)) { | |
1712 | pdc_dr_update(iop->dpp, tp->cpu, | |
1713 | iop->ready); | |
1714 | nentries++; | |
1715 | } else | |
1716 | clear_events(pfd); | |
1717 | } | |
1718 | if (--nevs == 0) | |
1719 | break; | |
1720 | } | |
1721 | } | |
1722 | ||
1723 | if (nentries) | |
1724 | incr_entries(nentries); | |
1725 | ||
1726 | return nentries; | |
1727 | } | |
1728 | ||
1729 | static int handle_pfds_entries(struct tracer *tp, int nevs, int force_read) | |
1730 | { | |
1731 | int i, nentries = 0; | |
1732 | struct trace_buf *tbp; | |
1733 | struct pollfd *pfd = tp->pfds; | |
1734 | struct io_info *iop = tp->ios; | |
1735 | ||
1736 | tbp = alloc_trace_buf(tp->cpu, buf_size); | |
1737 | for (i = 0; i < ndevs; i++, pfd++, iop++) { | |
1738 | if (pfd->revents & POLLIN || force_read) { | |
1739 | tbp->len = read(iop->ifd, tbp->buf, buf_size); | |
1740 | if (tbp->len > 0) { | |
1741 | pdc_dr_update(iop->dpp, tp->cpu, tbp->len); | |
1742 | add_trace_buf(iop->dpp, tp->cpu, &tbp); | |
1743 | nentries++; | |
1744 | } else if (tbp->len == 0) { | |
1745 | /* | |
1746 | * Short reads after we're done stop us | |
1747 | * from trying reads. | |
1748 | */ | |
1749 | if (tp->is_done) | |
1750 | clear_events(pfd); | |
1751 | } else { | |
1752 | read_err(tp->cpu, iop->ifn); | |
1753 | if (errno != EAGAIN || tp->is_done) | |
1754 | clear_events(pfd); | |
1755 | } | |
1756 | if (!piped_output && --nevs == 0) | |
1757 | break; | |
1758 | } | |
1759 | } | |
1760 | free(tbp); | |
1761 | ||
1762 | if (nentries) | |
1763 | incr_entries(nentries); | |
1764 | ||
1765 | return nentries; | |
1766 | } | |
1767 | ||
3fe0b570 | 1768 | static void *thread_main(void *arg) |
8e86c98a | 1769 | { |
df81fdb5 | 1770 | int ret, ndone, to_val; |
3fe0b570 | 1771 | struct tracer *tp = arg; |
8e86c98a | 1772 | |
3fe0b570 AB |
1773 | ret = lock_on_cpu(tp->cpu); |
1774 | if (ret) | |
1775 | goto err; | |
ff11d54c | 1776 | |
3fe0b570 | 1777 | ret = open_ios(tp); |
df81fdb5 | 1778 | if (ret) |
3fe0b570 | 1779 | goto err; |
6a6d3f0f | 1780 | |
3fe0b570 AB |
1781 | if (piped_output) |
1782 | to_val = 50; /* Frequent partial handles */ | |
ff11d54c | 1783 | else |
3fe0b570 AB |
1784 | to_val = 500; /* 1/2 second intervals */ |
1785 | ||
df81fdb5 AB |
1786 | |
1787 | tracer_signal_ready(tp, Th_running, 0); | |
1788 | tracer_wait_unblock(tp); | |
6488ca48 | 1789 | |
3fe0b570 AB |
1790 | while (!tp->is_done) { |
1791 | ndone = poll(tp->pfds, ndevs, to_val); | |
1792 | if (ndone || piped_output) | |
1793 | (void)handle_pfds(tp, ndone, piped_output); | |
1794 | else if (ndone < 0 && errno != EINTR) | |
1795 | fprintf(stderr, "Thread %d poll failed: %d/%s\n", | |
1796 | tp->cpu, errno, strerror(errno)); | |
1797 | } | |
22cd0c02 JA |
1798 | |
1799 | /* | |
3fe0b570 | 1800 | * Trace is stopped, pull data until we get a short read |
22cd0c02 | 1801 | */ |
3fe0b570 AB |
1802 | while (handle_pfds(tp, ndevs, 1) > 0) |
1803 | ; | |
055cc3e5 | 1804 | |
3fe0b570 | 1805 | close_ios(tp); |
df81fdb5 AB |
1806 | tracer_signal_ready(tp, Th_leaving, 0); |
1807 | return NULL; | |
8e86c98a | 1808 | |
3fe0b570 | 1809 | err: |
df81fdb5 | 1810 | tracer_signal_ready(tp, Th_error, ret); |
3fe0b570 | 1811 | return NULL; |
22cd0c02 JA |
1812 | } |
1813 | ||
3fe0b570 | 1814 | static int start_tracer(int cpu) |
22cd0c02 | 1815 | { |
3fe0b570 | 1816 | struct tracer *tp; |
22cd0c02 | 1817 | |
3fe0b570 AB |
1818 | tp = malloc(sizeof(*tp)); |
1819 | memset(tp, 0, sizeof(*tp)); | |
7ab2f837 | 1820 | |
3fe0b570 | 1821 | INIT_LIST_HEAD(&tp->head); |
3fe0b570 AB |
1822 | tp->status = 0; |
1823 | tp->cpu = cpu; | |
8e86c98a | 1824 | |
3fe0b570 AB |
1825 | if (pthread_create(&tp->thread, NULL, thread_main, tp)) { |
1826 | fprintf(stderr, "FAILED to start thread on CPU %d: %d/%s\n", | |
1827 | cpu, errno, strerror(errno)); | |
df81fdb5 AB |
1828 | free(tp); |
1829 | return 1; | |
8e86c98a | 1830 | } |
3fe0b570 | 1831 | |
df81fdb5 AB |
1832 | list_add_tail(&tp->head, &tracers); |
1833 | return 0; | |
8e86c98a JA |
1834 | } |
1835 | ||
df81fdb5 | 1836 | static void start_tracers(void) |
e0a1988b | 1837 | { |
3fe0b570 | 1838 | int cpu; |
df81fdb5 | 1839 | struct list_head *p; |
3fe0b570 AB |
1840 | |
1841 | for (cpu = 0; cpu < ncpus; cpu++) | |
1842 | if (start_tracer(cpu)) | |
1843 | break; | |
e0a1988b | 1844 | |
df81fdb5 AB |
1845 | wait_tracers_ready(cpu); |
1846 | ||
1847 | __list_for_each(p, &tracers) { | |
1848 | struct tracer *tp = list_entry(p, struct tracer, head); | |
1849 | if (tp->status) | |
1850 | fprintf(stderr, | |
1851 | "FAILED to start thread on CPU %d: %d/%s\n", | |
1852 | tp->cpu, tp->status, strerror(tp->status)); | |
1853 | } | |
3fe0b570 | 1854 | } |
e0a1988b | 1855 | |
3fe0b570 AB |
1856 | static void stop_tracers(void) |
1857 | { | |
1858 | struct list_head *p; | |
e0a1988b JA |
1859 | |
1860 | /* | |
3fe0b570 | 1861 | * Stop the tracing - makes the tracer threads clean up quicker. |
e0a1988b | 1862 | */ |
3fe0b570 AB |
1863 | __list_for_each(p, &devpaths) { |
1864 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
1865 | (void)ioctl(dpp->fd, BLKTRACESTOP); | |
e0a1988b JA |
1866 | } |
1867 | ||
3fe0b570 AB |
1868 | /* |
1869 | * Tell each tracer to quit | |
1870 | */ | |
1871 | __list_for_each(p, &tracers) { | |
1872 | struct tracer *tp = list_entry(p, struct tracer, head); | |
1873 | tp->is_done = 1; | |
1874 | } | |
ff11d54c | 1875 | } |
e0a1988b | 1876 | |
3fe0b570 | 1877 | static void del_tracers(void) |
ff11d54c | 1878 | { |
3fe0b570 | 1879 | struct list_head *p, *q; |
ff11d54c | 1880 | |
3fe0b570 AB |
1881 | list_for_each_safe(p, q, &tracers) { |
1882 | struct tracer *tp = list_entry(p, struct tracer, head); | |
ff11d54c | 1883 | |
3fe0b570 AB |
1884 | list_del(&tp->head); |
1885 | free(tp); | |
e0a1988b | 1886 | } |
ff11d54c | 1887 | } |
e0a1988b | 1888 | |
3fe0b570 | 1889 | static void wait_tracers(void) |
ff11d54c | 1890 | { |
3fe0b570 | 1891 | struct list_head *p; |
ff11d54c | 1892 | |
3fe0b570 AB |
1893 | if (use_tracer_devpaths()) |
1894 | process_trace_bufs(); | |
1895 | ||
df81fdb5 AB |
1896 | wait_tracers_leaving(); |
1897 | ||
3fe0b570 AB |
1898 | __list_for_each(p, &tracers) { |
1899 | int ret; | |
1900 | struct tracer *tp = list_entry(p, struct tracer, head); | |
1901 | ||
3fe0b570 AB |
1902 | ret = pthread_join(tp->thread, NULL); |
1903 | if (ret) | |
1904 | fprintf(stderr, "Thread join %d failed %d\n", | |
1905 | tp->cpu, ret); | |
ff11d54c TZ |
1906 | } |
1907 | ||
3fe0b570 AB |
1908 | if (use_tracer_devpaths()) |
1909 | clean_trace_bufs(); | |
1910 | ||
1911 | get_all_drops(); | |
ff11d54c TZ |
1912 | } |
1913 | ||
3fe0b570 | 1914 | static void exit_tracing(void) |
ff11d54c | 1915 | { |
3fe0b570 AB |
1916 | signal(SIGINT, SIG_IGN); |
1917 | signal(SIGHUP, SIG_IGN); | |
1918 | signal(SIGTERM, SIG_IGN); | |
1919 | signal(SIGALRM, SIG_IGN); | |
1920 | ||
1921 | stop_tracers(); | |
1922 | wait_tracers(); | |
1923 | del_tracers(); | |
1924 | rel_devpaths(); | |
e0a1988b JA |
1925 | } |
1926 | ||
3fe0b570 | 1927 | static void handle_sigint(__attribute__((__unused__)) int sig) |
8e86c98a | 1928 | { |
3fe0b570 AB |
1929 | done = 1; |
1930 | stop_tracers(); | |
8e86c98a JA |
1931 | } |
1932 | ||
3fe0b570 | 1933 | static void show_stats(struct list_head *devpaths) |
659bcc3f | 1934 | { |
3fe0b570 AB |
1935 | FILE *ofp; |
1936 | struct list_head *p; | |
1937 | unsigned long long nevents, data_read; | |
1938 | unsigned long long total_drops = 0; | |
1939 | unsigned long long total_events = 0; | |
1940 | ||
1941 | if (piped_output) | |
1942 | ofp = my_fopen("/dev/null", "w"); | |
1943 | else | |
1944 | ofp = stdout; | |
ff11d54c | 1945 | |
3fe0b570 AB |
1946 | __list_for_each(p, devpaths) { |
1947 | int cpu; | |
1948 | struct pdc_stats *sp; | |
1949 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
e0a1988b | 1950 | |
3fe0b570 AB |
1951 | if (net_mode == Net_server) |
1952 | printf("server: end of run for %s:%s\n", | |
1953 | dpp->ch->hostname, dpp->buts_name); | |
e0a1988b | 1954 | |
3fe0b570 AB |
1955 | data_read = 0; |
1956 | nevents = 0; | |
1957 | ||
1958 | fprintf(ofp, "=== %s ===\n", dpp->buts_name); | |
1959 | for (cpu = 0, sp = dpp->stats; cpu < dpp->ncpus; cpu++, sp++) { | |
1960 | /* | |
1961 | * Estimate events if not known... | |
1962 | */ | |
1963 | if (sp->nevents == 0) { | |
1964 | sp->nevents = sp->data_read / | |
1965 | sizeof(struct blk_io_trace); | |
ff11d54c | 1966 | } |
e0a1988b | 1967 | |
3fe0b570 AB |
1968 | fprintf(ofp, |
1969 | " CPU%3d: %20llu events, %8llu KiB data\n", | |
1970 | cpu, sp->nevents, (sp->data_read + 1023) >> 10); | |
e0a1988b | 1971 | |
3fe0b570 AB |
1972 | data_read += sp->data_read; |
1973 | nevents += sp->nevents; | |
e0a1988b JA |
1974 | } |
1975 | ||
3fe0b570 AB |
1976 | fprintf(ofp, " Total: %20llu events (dropped %llu)," |
1977 | " %8llu KiB data\n", nevents, | |
1978 | dpp->drops, (data_read + 1024) >> 10); | |
8e86c98a | 1979 | |
3fe0b570 AB |
1980 | total_drops += dpp->drops; |
1981 | total_events += (nevents + dpp->drops); | |
8e86c98a JA |
1982 | } |
1983 | ||
3fe0b570 AB |
1984 | fflush(ofp); |
1985 | if (piped_output) | |
1986 | fclose(ofp); | |
8e86c98a | 1987 | |
3fe0b570 AB |
1988 | if (total_drops) { |
1989 | double drops_ratio = 1.0; | |
8e86c98a | 1990 | |
3fe0b570 AB |
1991 | if (total_events) |
1992 | drops_ratio = (double)total_drops/(double)total_events; | |
8e86c98a | 1993 | |
3fe0b570 AB |
1994 | fprintf(stderr, "\nYou have %llu (%5.1lf%%) dropped events\n" |
1995 | "Consider using a larger buffer size (-b) " | |
1996 | "and/or more buffers (-n)\n", | |
1997 | total_drops, 100.0 * drops_ratio); | |
8e86c98a | 1998 | } |
8e86c98a JA |
1999 | } |
2000 | ||
3fe0b570 | 2001 | static int handle_args(int argc, char *argv[]) |
8e86c98a | 2002 | { |
3fe0b570 | 2003 | int c, i; |
e3e74029 | 2004 | struct statfs st; |
d39c04ca AB |
2005 | int act_mask_tmp = 0; |
2006 | ||
2007 | while ((c = getopt_long(argc, argv, S_OPTS, l_opts, NULL)) >= 0) { | |
2008 | switch (c) { | |
2009 | case 'a': | |
2010 | i = find_mask_map(optarg); | |
2011 | if (i < 0) { | |
3fe0b570 | 2012 | fprintf(stderr, "Invalid action mask %s\n", |
d39c04ca | 2013 | optarg); |
7425d456 | 2014 | return 1; |
d39c04ca AB |
2015 | } |
2016 | act_mask_tmp |= i; | |
2017 | break; | |
2018 | ||
2019 | case 'A': | |
3fe0b570 | 2020 | if ((sscanf(optarg, "%x", &i) != 1) || |
98f8386b | 2021 | !valid_act_opt(i)) { |
d39c04ca | 2022 | fprintf(stderr, |
ab197ca7 | 2023 | "Invalid set action mask %s/0x%x\n", |
d39c04ca | 2024 | optarg, i); |
7425d456 | 2025 | return 1; |
d39c04ca AB |
2026 | } |
2027 | act_mask_tmp = i; | |
2028 | break; | |
d0ca268b | 2029 | |
d39c04ca | 2030 | case 'd': |
3fe0b570 | 2031 | if (add_devpath(optarg) != 0) |
e7c9f3ff | 2032 | return 1; |
d39c04ca AB |
2033 | break; |
2034 | ||
cf1edb17 AB |
2035 | case 'I': { |
2036 | char dev_line[256]; | |
3fe0b570 | 2037 | FILE *ifp = my_fopen(optarg, "r"); |
cf1edb17 AB |
2038 | |
2039 | if (!ifp) { | |
3fe0b570 AB |
2040 | fprintf(stderr, |
2041 | "Invalid file for devices %s\n", | |
cf1edb17 AB |
2042 | optarg); |
2043 | return 1; | |
2044 | } | |
2045 | ||
2046 | while (fscanf(ifp, "%s\n", dev_line) == 1) | |
3fe0b570 | 2047 | if (add_devpath(dev_line) != 0) |
cf1edb17 AB |
2048 | return 1; |
2049 | break; | |
2050 | } | |
cf1edb17 | 2051 | |
5270dddd | 2052 | case 'r': |
3d06efea | 2053 | debugfs_path = optarg; |
5270dddd JA |
2054 | break; |
2055 | ||
d5396421 | 2056 | case 'o': |
66efebf8 | 2057 | output_name = optarg; |
d5396421 | 2058 | break; |
bc39777c JA |
2059 | case 'k': |
2060 | kill_running_trace = 1; | |
2061 | break; | |
ece238a6 NS |
2062 | case 'w': |
2063 | stop_watch = atoi(optarg); | |
2064 | if (stop_watch <= 0) { | |
2065 | fprintf(stderr, | |
2066 | "Invalid stopwatch value (%d secs)\n", | |
2067 | stop_watch); | |
2068 | return 1; | |
2069 | } | |
2070 | break; | |
57ea8602 | 2071 | case 'V': |
5d4f19d9 | 2072 | case 'v': |
52724a0e | 2073 | printf("%s version %s\n", argv[0], blktrace_version); |
3fe0b570 AB |
2074 | exit(0); |
2075 | /*NOTREACHED*/ | |
129aa440 | 2076 | case 'b': |
eb3c8108 | 2077 | buf_size = strtoul(optarg, NULL, 10); |
183a0855 | 2078 | if (buf_size <= 0 || buf_size > 16*1024) { |
3fe0b570 AB |
2079 | fprintf(stderr, "Invalid buffer size (%lu)\n", |
2080 | buf_size); | |
129aa440 JA |
2081 | return 1; |
2082 | } | |
2083 | buf_size <<= 10; | |
2084 | break; | |
2085 | case 'n': | |
eb3c8108 | 2086 | buf_nr = strtoul(optarg, NULL, 10); |
129aa440 JA |
2087 | if (buf_nr <= 0) { |
2088 | fprintf(stderr, | |
eb3c8108 | 2089 | "Invalid buffer nr (%lu)\n", buf_nr); |
129aa440 JA |
2090 | return 1; |
2091 | } | |
2092 | break; | |
d1d7f15f JA |
2093 | case 'D': |
2094 | output_dir = optarg; | |
2095 | break; | |
8e86c98a JA |
2096 | case 'h': |
2097 | net_mode = Net_client; | |
2098 | strcpy(hostname, optarg); | |
2099 | break; | |
2100 | case 'l': | |
2101 | net_mode = Net_server; | |
2102 | break; | |
2103 | case 'p': | |
2104 | net_port = atoi(optarg); | |
2105 | break; | |
32f18c48 | 2106 | case 's': |
79971f43 | 2107 | net_use_sendfile = 0; |
32f18c48 | 2108 | break; |
d39c04ca | 2109 | default: |
ee1f4158 | 2110 | show_usage(argv[0]); |
3fe0b570 AB |
2111 | exit(1); |
2112 | /*NOTREACHED*/ | |
d39c04ca AB |
2113 | } |
2114 | } | |
2115 | ||
3fe0b570 AB |
2116 | while (optind < argc) |
2117 | if (add_devpath(argv[optind++]) != 0) | |
2118 | return 1; | |
8e86c98a | 2119 | |
3fe0b570 AB |
2120 | if (net_mode != Net_server && ndevs == 0) { |
2121 | show_usage(argv[0]); | |
2122 | return 1; | |
2123 | } | |
8e86c98a | 2124 | |
3fe0b570 AB |
2125 | if (statfs(debugfs_path, &st) < 0 || st.f_type != (long)DEBUGFS_TYPE) { |
2126 | fprintf(stderr, "Invalid debug path %s: %d/%s\n", | |
2127 | debugfs_path, errno, strerror(errno)); | |
2128 | return 1; | |
2129 | } | |
2130 | ||
2131 | if (act_mask_tmp != 0) | |
2132 | act_mask = act_mask_tmp; | |
2133 | ||
e58f3937 AB |
2134 | if (net_mode == Net_client && net_setup_addr()) |
2135 | return 1; | |
2136 | ||
3fe0b570 AB |
2137 | /* |
2138 | * Set up for appropriate PFD handler based upon output name. | |
2139 | */ | |
2140 | if (net_client_use_sendfile()) | |
2141 | handle_pfds = handle_pfds_netclient; | |
2142 | else if (net_client_use_send()) | |
2143 | handle_pfds = handle_pfds_entries; | |
2144 | else if (output_name && (strcmp(output_name, "-") == 0)) { | |
2145 | piped_output = 1; | |
2146 | handle_pfds = handle_pfds_entries; | |
2147 | pfp = stdout; | |
2148 | setvbuf(pfp, NULL, _IONBF, 0); | |
2149 | } else | |
2150 | handle_pfds = handle_pfds_file; | |
2151 | return 0; | |
2152 | } | |
2153 | ||
2154 | static void ch_add_connection(struct net_server_s *ns, struct cl_host *ch, | |
2155 | int fd) | |
2156 | { | |
2157 | struct cl_conn *nc; | |
2158 | ||
2159 | nc = malloc(sizeof(*nc)); | |
2160 | memset(nc, 0, sizeof(*nc)); | |
2161 | ||
2162 | time(&nc->connect_time); | |
2163 | nc->ch = ch; | |
2164 | nc->fd = fd; | |
2165 | nc->ncpus = -1; | |
2166 | ||
2167 | list_add_tail(&nc->ch_head, &ch->conn_list); | |
2168 | ch->connects++; | |
2169 | ||
2170 | list_add_tail(&nc->ns_head, &ns->conn_list); | |
2171 | ns->connects++; | |
2172 | ns->pfds = realloc(ns->pfds, (ns->connects+1) * sizeof(struct pollfd)); | |
2173 | } | |
2174 | ||
2175 | static void ch_rem_connection(struct net_server_s *ns, struct cl_host *ch, | |
2176 | struct cl_conn *nc) | |
2177 | { | |
2178 | net_close_connection(&nc->fd); | |
2179 | ||
2180 | list_del(&nc->ch_head); | |
2181 | ch->connects--; | |
2182 | ||
2183 | list_del(&nc->ns_head); | |
2184 | ns->connects--; | |
2185 | ns->pfds = realloc(ns->pfds, (ns->connects+1) * sizeof(struct pollfd)); | |
2186 | ||
2187 | free(nc); | |
2188 | } | |
2189 | ||
2190 | static struct cl_host *net_find_client_host(struct net_server_s *ns, | |
2191 | struct in_addr cl_in_addr) | |
2192 | { | |
2193 | struct list_head *p; | |
2194 | ||
2195 | __list_for_each(p, &ns->ch_list) { | |
2196 | struct cl_host *ch = list_entry(p, struct cl_host, head); | |
2197 | ||
2198 | if (in_addr_eq(ch->cl_in_addr, cl_in_addr)) | |
2199 | return ch; | |
2200 | } | |
2201 | ||
2202 | return NULL; | |
2203 | } | |
2204 | ||
2205 | static struct cl_host *net_add_client_host(struct net_server_s *ns, | |
2206 | struct sockaddr_in *addr) | |
2207 | { | |
2208 | struct cl_host *ch; | |
2209 | ||
2210 | ch = malloc(sizeof(*ch)); | |
2211 | memset(ch, 0, sizeof(*ch)); | |
2212 | ||
2213 | ch->ns = ns; | |
2214 | ch->cl_in_addr = addr->sin_addr; | |
2215 | list_add_tail(&ch->head, &ns->ch_list); | |
2216 | ns->nchs++; | |
ec685dd2 | 2217 | |
3fe0b570 AB |
2218 | ch->hostname = strdup(inet_ntoa(addr->sin_addr)); |
2219 | printf("server: connection from %s\n", ch->hostname); | |
2220 | ||
2221 | INIT_LIST_HEAD(&ch->conn_list); | |
2222 | INIT_LIST_HEAD(&ch->devpaths); | |
2223 | ||
2224 | return ch; | |
2225 | } | |
2226 | ||
2227 | static void device_done(struct devpath *dpp, int ncpus) | |
2228 | { | |
2229 | int cpu; | |
2230 | struct io_info *iop; | |
2231 | ||
2232 | for (cpu = 0, iop = dpp->ios; cpu < ncpus; cpu++, iop++) | |
2233 | close_iop(iop); | |
2234 | ||
2235 | list_del(&dpp->head); | |
2236 | dpp_free(dpp); | |
2237 | } | |
2238 | ||
2239 | static void net_ch_remove(struct cl_host *ch, int ncpus) | |
2240 | { | |
2241 | struct list_head *p, *q; | |
2242 | struct net_server_s *ns = ch->ns; | |
2243 | ||
2244 | list_for_each_safe(p, q, &ch->devpaths) { | |
2245 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
2246 | device_done(dpp, ncpus); | |
ec685dd2 | 2247 | } |
8e86c98a | 2248 | |
3fe0b570 AB |
2249 | list_for_each_safe(p, q, &ch->conn_list) { |
2250 | struct cl_conn *nc = list_entry(p, struct cl_conn, ch_head); | |
2251 | ||
2252 | ch_rem_connection(ns, ch, nc); | |
22cd0c02 JA |
2253 | } |
2254 | ||
3fe0b570 AB |
2255 | list_del(&ch->head); |
2256 | ns->nchs--; | |
2257 | ||
2258 | if (ch->hostname) | |
2259 | free(ch->hostname); | |
2260 | free(ch); | |
2261 | } | |
2262 | ||
2263 | static void net_add_connection(struct net_server_s *ns) | |
2264 | { | |
2265 | int fd; | |
2266 | struct cl_host *ch; | |
2267 | socklen_t socklen = sizeof(ns->addr); | |
2268 | ||
d5302b03 | 2269 | fd = my_accept(ns->listen_fd, (struct sockaddr *)&ns->addr, &socklen); |
3fe0b570 AB |
2270 | if (fd < 0) { |
2271 | /* | |
2272 | * This is OK: we just won't accept this connection, | |
2273 | * nothing fatal. | |
2274 | */ | |
2275 | perror("accept"); | |
2276 | } else { | |
2277 | ch = net_find_client_host(ns, ns->addr.sin_addr); | |
2278 | if (!ch) | |
2279 | ch = net_add_client_host(ns, &ns->addr); | |
2280 | ||
2281 | ch_add_connection(ns, ch, fd); | |
d39c04ca | 2282 | } |
3fe0b570 | 2283 | } |
d39c04ca | 2284 | |
3fe0b570 AB |
2285 | static struct devpath *nc_add_dpp(struct cl_conn *nc, |
2286 | struct blktrace_net_hdr *bnh, | |
2287 | time_t connect_time) | |
2288 | { | |
2289 | int cpu; | |
2290 | struct io_info *iop; | |
2291 | struct devpath *dpp; | |
2292 | ||
2293 | dpp = malloc(sizeof(*dpp)); | |
2294 | memset(dpp, 0, sizeof(*dpp)); | |
2295 | ||
2296 | dpp->buts_name = strdup(bnh->buts_name); | |
2297 | dpp->path = strdup(bnh->buts_name); | |
2298 | dpp->fd = -1; | |
2299 | dpp->ch = nc->ch; | |
2300 | dpp->cl_id = bnh->cl_id; | |
2301 | dpp->cl_connect_time = connect_time; | |
2302 | dpp->ncpus = nc->ncpus; | |
2303 | dpp->stats = calloc(dpp->ncpus, sizeof(*dpp->stats)); | |
2304 | memset(dpp->stats, 0, dpp->ncpus * sizeof(*dpp->stats)); | |
2305 | ||
2306 | list_add_tail(&dpp->head, &nc->ch->devpaths); | |
2307 | nc->ch->ndevs++; | |
2308 | ||
2309 | dpp->ios = calloc(nc->ncpus, sizeof(*iop)); | |
2310 | memset(dpp->ios, 0, ndevs * sizeof(*iop)); | |
2311 | ||
2312 | for (cpu = 0, iop = dpp->ios; cpu < nc->ncpus; cpu++, iop++) { | |
2313 | iop->dpp = dpp; | |
2314 | iop->nc = nc; | |
2315 | init_mmap_info(&iop->mmap_info); | |
2316 | ||
2317 | if (iop_open(iop, cpu)) | |
2318 | goto err; | |
69dd57c2 AB |
2319 | } |
2320 | ||
3fe0b570 | 2321 | return dpp; |
69dd57c2 | 2322 | |
3fe0b570 AB |
2323 | err: |
2324 | /* | |
2325 | * Need to unravel what's been done... | |
2326 | */ | |
2327 | while (cpu >= 0) | |
2328 | close_iop(&dpp->ios[cpu--]); | |
2329 | dpp_free(dpp); | |
2330 | ||
2331 | return NULL; | |
2332 | } | |
d0ca268b | 2333 | |
3fe0b570 AB |
2334 | static struct devpath *nc_find_dpp(struct cl_conn *nc, |
2335 | struct blktrace_net_hdr *bnh) | |
2336 | { | |
2337 | struct list_head *p; | |
2338 | time_t connect_time = nc->connect_time; | |
3d06efea | 2339 | |
3fe0b570 AB |
2340 | __list_for_each(p, &nc->ch->devpaths) { |
2341 | struct devpath *dpp = list_entry(p, struct devpath, head); | |
2342 | ||
2343 | if (!strcmp(dpp->buts_name, bnh->buts_name)) | |
2344 | return dpp; | |
2345 | ||
2346 | if (dpp->cl_id == bnh->cl_id) | |
2347 | connect_time = dpp->cl_connect_time; | |
d0ca268b JA |
2348 | } |
2349 | ||
3fe0b570 AB |
2350 | return nc_add_dpp(nc, bnh, connect_time); |
2351 | } | |
bc39777c | 2352 | |
3fe0b570 AB |
2353 | static void net_client_read_data(struct cl_conn *nc, struct devpath *dpp, |
2354 | struct blktrace_net_hdr *bnh) | |
2355 | { | |
2356 | int ret; | |
2357 | struct io_info *iop = &dpp->ios[bnh->cpu]; | |
2358 | struct mmap_info *mip = &iop->mmap_info; | |
2359 | ||
2360 | if (setup_mmap(iop->ofd, bnh->len, &iop->mmap_info)) { | |
2361 | fprintf(stderr, "ncd(%s:%d): mmap failed\n", | |
2362 | nc->ch->hostname, nc->fd); | |
2363 | exit(1); | |
2364 | } | |
2365 | ||
2366 | ret = net_recv_data(nc->fd, mip->fs_buf + mip->fs_off, bnh->len); | |
2367 | if (ret > 0) { | |
2368 | pdc_dr_update(dpp, bnh->cpu, ret); | |
2369 | mip->fs_size += ret; | |
2370 | mip->fs_off += ret; | |
2371 | } else if (ret < 0) | |
2372 | exit(1); | |
2373 | } | |
2374 | ||
2375 | /* | |
2376 | * Returns 1 if we closed a host - invalidates other polling information | |
2377 | * that may be present. | |
2378 | */ | |
2379 | static int net_client_data(struct cl_conn *nc) | |
2380 | { | |
2381 | int ret; | |
2382 | struct devpath *dpp; | |
2383 | struct blktrace_net_hdr bnh; | |
2384 | ||
2385 | ret = net_get_header(nc, &bnh); | |
2386 | if (ret == 0) | |
7425d456 | 2387 | return 0; |
3fe0b570 AB |
2388 | |
2389 | if (ret < 0) { | |
2390 | fprintf(stderr, "ncd(%d): header read failed\n", nc->fd); | |
2391 | exit(1); | |
2392 | } | |
2393 | ||
2394 | if (data_is_native == -1 && check_data_endianness(bnh.magic)) { | |
2395 | fprintf(stderr, "ncd(%d): received data is bad\n", nc->fd); | |
2396 | exit(1); | |
2397 | } | |
2398 | ||
2399 | if (!data_is_native) { | |
2400 | bnh.magic = be32_to_cpu(bnh.magic); | |
2401 | bnh.cpu = be32_to_cpu(bnh.cpu); | |
2402 | bnh.max_cpus = be32_to_cpu(bnh.max_cpus); | |
2403 | bnh.len = be32_to_cpu(bnh.len); | |
2404 | bnh.cl_id = be32_to_cpu(bnh.cl_id); | |
2405 | bnh.buf_size = be32_to_cpu(bnh.buf_size); | |
2406 | bnh.buf_nr = be32_to_cpu(bnh.buf_nr); | |
2407 | bnh.page_size = be32_to_cpu(bnh.page_size); | |
2408 | } | |
2409 | ||
2410 | if ((bnh.magic & 0xffffff00) != BLK_IO_TRACE_MAGIC) { | |
2411 | fprintf(stderr, "ncd(%s:%d): bad data magic\n", | |
2412 | nc->ch->hostname, nc->fd); | |
2413 | exit(1); | |
2414 | } | |
2415 | ||
2416 | if (nc->ncpus == -1) | |
2417 | nc->ncpus = bnh.max_cpus; | |
2418 | ||
2419 | /* | |
2420 | * len == 0 means the other end is sending us a new connection/dpp | |
2421 | * len == 1 means that the other end signalled end-of-run | |
2422 | */ | |
2423 | dpp = nc_find_dpp(nc, &bnh); | |
2424 | if (bnh.len == 0) { | |
2425 | /* | |
2426 | * Just adding in the dpp above is enough | |
2427 | */ | |
2428 | ack_open_close(nc->fd, dpp->buts_name); | |
2429 | nc->ch->cl_opens++; | |
2430 | } else if (bnh.len == 1) { | |
2431 | /* | |
2432 | * overload cpu count with dropped events | |
2433 | */ | |
2434 | dpp->drops = bnh.cpu; | |
2435 | ||
2436 | ack_open_close(nc->fd, dpp->buts_name); | |
2437 | if (--nc->ch->cl_opens == 0) { | |
2438 | show_stats(&nc->ch->devpaths); | |
2439 | net_ch_remove(nc->ch, nc->ncpus); | |
2440 | return 1; | |
2441 | } | |
2442 | } else | |
2443 | net_client_read_data(nc, dpp, &bnh); | |
2444 | ||
2445 | return 0; | |
2446 | } | |
2447 | ||
2448 | static void handle_client_data(struct net_server_s *ns, int events) | |
2449 | { | |
2450 | struct cl_conn *nc; | |
2451 | struct pollfd *pfd; | |
2452 | struct list_head *p, *q; | |
2453 | ||
2454 | pfd = &ns->pfds[1]; | |
2455 | list_for_each_safe(p, q, &ns->conn_list) { | |
2456 | if (pfd->revents & POLLIN) { | |
2457 | nc = list_entry(p, struct cl_conn, ns_head); | |
2458 | ||
2459 | if (net_client_data(nc) || --events == 0) | |
2460 | break; | |
2461 | } | |
2462 | pfd++; | |
2463 | } | |
2464 | } | |
2465 | ||
2466 | static void net_setup_pfds(struct net_server_s *ns) | |
2467 | { | |
2468 | struct pollfd *pfd; | |
2469 | struct list_head *p; | |
2470 | ||
2471 | ns->pfds[0].fd = ns->listen_fd; | |
2472 | ns->pfds[0].events = POLLIN; | |
2473 | ||
2474 | pfd = &ns->pfds[1]; | |
2475 | __list_for_each(p, &ns->conn_list) { | |
2476 | struct cl_conn *nc = list_entry(p, struct cl_conn, ns_head); | |
2477 | ||
2478 | pfd->fd = nc->fd; | |
2479 | pfd->events = POLLIN; | |
2480 | pfd++; | |
2481 | } | |
2482 | } | |
2483 | ||
2484 | static int net_server_handle_connections(struct net_server_s *ns) | |
2485 | { | |
2486 | int events; | |
2487 | ||
2488 | printf("server: waiting for connections...\n"); | |
2489 | ||
2490 | while (!done) { | |
2491 | net_setup_pfds(ns); | |
2492 | events = poll(ns->pfds, ns->connects + 1, -1); | |
2493 | if (events < 0) { | |
2494 | if (errno != EINTR) { | |
2495 | perror("FATAL: poll error"); | |
2496 | return 1; | |
2497 | } | |
2498 | } else if (events > 0) { | |
2499 | if (ns->pfds[0].revents & POLLIN) { | |
2500 | net_add_connection(ns); | |
2501 | events--; | |
2502 | } | |
2503 | ||
2504 | if (events) | |
2505 | handle_client_data(ns, events); | |
2506 | } | |
2507 | } | |
2508 | ||
2509 | return 0; | |
2510 | } | |
2511 | ||
2512 | static int net_server(void) | |
2513 | { | |
2514 | int fd, opt; | |
2515 | int ret = 1; | |
2516 | struct net_server_s net_server; | |
2517 | struct net_server_s *ns = &net_server; | |
2518 | ||
2519 | memset(ns, 0, sizeof(*ns)); | |
2520 | INIT_LIST_HEAD(&ns->ch_list); | |
2521 | INIT_LIST_HEAD(&ns->conn_list); | |
2522 | ns->pfds = malloc(sizeof(struct pollfd)); | |
2523 | ||
2524 | fd = my_socket(AF_INET, SOCK_STREAM, 0); | |
2525 | if (fd < 0) { | |
2526 | perror("server: socket"); | |
2527 | goto out; | |
2528 | } | |
2529 | ||
2530 | opt = 1; | |
2531 | if (setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, &opt, sizeof(opt)) < 0) { | |
2532 | perror("setsockopt"); | |
2533 | goto out; | |
2534 | } | |
2535 | ||
2536 | memset(&ns->addr, 0, sizeof(ns->addr)); | |
2537 | ns->addr.sin_family = AF_INET; | |
2538 | ns->addr.sin_addr.s_addr = htonl(INADDR_ANY); | |
2539 | ns->addr.sin_port = htons(net_port); | |
2540 | ||
2541 | if (bind(fd, (struct sockaddr *) &ns->addr, sizeof(ns->addr)) < 0) { | |
2542 | perror("bind"); | |
2543 | goto out; | |
2544 | } | |
2545 | ||
2546 | if (listen(fd, 1) < 0) { | |
2547 | perror("listen"); | |
2548 | goto out; | |
2549 | } | |
2550 | ||
2551 | /* | |
2552 | * The actual server looping is done here: | |
2553 | */ | |
2554 | ns->listen_fd = fd; | |
2555 | ret = net_server_handle_connections(ns); | |
2556 | ||
2557 | /* | |
2558 | * Clean up and return... | |
2559 | */ | |
2560 | out: | |
2561 | free(ns->pfds); | |
2562 | return ret; | |
2563 | } | |
2564 | ||
055cc3e5 AB |
2565 | static int run_tracers(void) |
2566 | { | |
2567 | atexit(exit_tracing); | |
2568 | if (net_mode == Net_client) | |
2569 | printf("blktrace: connecting to %s\n", hostname); | |
2570 | ||
2571 | setup_buts(); | |
2572 | ||
2573 | if (use_tracer_devpaths()) { | |
2574 | if (setup_tracer_devpaths()) | |
2575 | return 1; | |
2576 | ||
2577 | if (piped_output) | |
2578 | handle_list = handle_list_file; | |
2579 | else | |
2580 | handle_list = handle_list_net; | |
2581 | } | |
2582 | ||
2583 | start_tracers(); | |
2584 | if (nthreads_running == ncpus) { | |
2585 | unblock_tracers(); | |
2586 | start_buts(); | |
2587 | if (net_mode == Net_client) | |
2588 | printf("blktrace: connected!\n"); | |
2589 | if (stop_watch) | |
2590 | alarm(stop_watch); | |
2591 | } else | |
2592 | stop_tracers(); | |
2593 | ||
2594 | wait_tracers(); | |
2595 | if (nthreads_running == ncpus) | |
2596 | show_stats(&devpaths); | |
2597 | if (net_client_use_send()) | |
2598 | close_client_connections(); | |
2599 | del_tracers(); | |
2600 | ||
2601 | return 0; | |
2602 | } | |
2603 | ||
3fe0b570 AB |
2604 | int main(int argc, char *argv[]) |
2605 | { | |
2606 | int ret = 0; | |
2607 | ||
2608 | setlocale(LC_NUMERIC, "en_US"); | |
2609 | pagesize = getpagesize(); | |
2610 | ncpus = sysconf(_SC_NPROCESSORS_ONLN); | |
2611 | if (ncpus < 0) { | |
2612 | fprintf(stderr, "sysconf(_SC_NPROCESSORS_ONLN) failed %d/%s\n", | |
2613 | errno, strerror(errno)); | |
2614 | ret = 1; | |
2615 | goto out; | |
055cc3e5 | 2616 | } else if (handle_args(argc, argv)) { |
3fe0b570 AB |
2617 | ret = 1; |
2618 | goto out; | |
bc39777c JA |
2619 | } |
2620 | ||
d0ca268b JA |
2621 | signal(SIGINT, handle_sigint); |
2622 | signal(SIGHUP, handle_sigint); | |
2623 | signal(SIGTERM, handle_sigint); | |
ece238a6 | 2624 | signal(SIGALRM, handle_sigint); |
38e1f0c6 | 2625 | signal(SIGPIPE, SIG_IGN); |
d0ca268b | 2626 | |
3fe0b570 AB |
2627 | if (kill_running_trace) { |
2628 | struct devpath *dpp; | |
2629 | struct list_head *p; | |
8e86c98a | 2630 | |
3fe0b570 AB |
2631 | __list_for_each(p, &devpaths) { |
2632 | dpp = list_entry(p, struct devpath, head); | |
2633 | if (__stop_trace(dpp->fd)) { | |
2634 | fprintf(stderr, | |
2635 | "BLKTRACETEARDOWN %s failed: %d/%s\n", | |
2636 | dpp->path, errno, strerror(errno)); | |
2637 | } | |
2638 | } | |
2639 | } else if (net_mode == Net_server) { | |
2640 | if (output_name) { | |
2641 | fprintf(stderr, "-o ignored in server mode\n"); | |
2642 | output_name = NULL; | |
2643 | } | |
3fe0b570 | 2644 | ret = net_server(); |
055cc3e5 AB |
2645 | } else |
2646 | ret = run_tracers(); | |
d0ca268b | 2647 | |
3fe0b570 AB |
2648 | out: |
2649 | if (pfp) | |
2650 | fclose(pfp); | |
2651 | rel_devpaths(); | |
2652 | return ret; | |
2653 | } |