Print this page
OS-3820 lxbrand ptrace(2): the next generation
OS-3685 lxbrand PTRACE_O_TRACEFORK race condition
OS-3834 lxbrand 64-bit strace(1) reports 64-bit process as using x32 ABI
OS-3794 lxbrand panic on init signal death
Reviewed by: Robert Mustacchi <rm@joyent.com>
Reviewed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Reviewed by: Bryan Cantrill <bryan@joyent.com>
OS-3463 expose process argv through procfs
OS-3207 in lx zone, 'ps auxww' does not show full cmdline for processes
Reviewed by: Robert Mustacchi <rm@joyent.com>
Reviewed by: Patrick Mooney <patrick.mooney@joyent.com>
OS-3383 lx brand: node.js test test-setproctitle.js fails
OS-15 add procfs equivalent to prctl(PR_SET_NAME)
| Split |
Close |
| Expand all |
| Collapse all |
--- old/usr/src/man/man4/proc.4.man.txt
+++ new/usr/src/man/man4/proc.4.man.txt
1 1 PROC(4) File Formats and Configurations PROC(4)
2 2
3 3
4 4
5 5 NAME
6 6 proc - /proc, the process file system
7 7
8 8 DESCRIPTION
9 9 /proc is a file system that provides access to the state of each
10 10 process and light-weight process (lwp) in the system. The name of each
11 11 entry in the /proc directory is a decimal number corresponding to a
12 12 process-ID. These entries are themselves subdirectories. Access to
13 13 process state is provided by additional files contained within each
14 14 subdirectory; the hierarchy is described more completely below. In this
15 15 document, ``/proc file'' refers to a non-directory file within the
16 16 hierarchy rooted at /proc. The owner of each /proc file and
17 17 subdirectory is determined by the user-ID of the process.
18 18
19 19
20 20 /proc can be mounted on any mount point, in addition to the standard
21 21 /proc mount point, and can be mounted several places at once. Such
22 22 additional mounts are allowed in order to facilitate the confinement of
23 23 processes to subtrees of the file system via chroot(1M) and yet allow
24 24 such processes access to commands like ps(1).
25 25
26 26
27 27 Standard system calls are used to access /proc files: open(2),
28 28 close(2), read(2), and write(2) (including readv(2), writev(2),
29 29 pread(2), and pwrite(2)). Most files describe process state and can
30 30 only be opened for reading. ctl and lwpctl (control) files permit
31 31 manipulation of process state and can only be opened for writing. as
32 32 (address space) files contain the image of the running process and can
33 33 be opened for both reading and writing. An open for writing allows
34 34 process control; a read-only open allows inspection but not control. In
35 35 this document, we refer to the process as open for reading or writing
36 36 if any of its associated /proc files is open for reading or writing.
37 37
38 38
39 39 In general, more than one process can open the same /proc file at the
40 40 same time. Exclusive open is an advisory mechanism provided to allow
41 41 controlling processes to avoid collisions with each other. A process
42 42 can obtain exclusive control of a target process, with respect to other
43 43 cooperating processes, if it successfully opens any /proc file in the
44 44 target process for writing (the as or ctl files, or the lwpctl file of
45 45 any lwp) while specifying O_EXCL in the open(2). Such an open will fail
46 46 if the target process is already open for writing (that is, if an as,
47 47 ctl, or lwpctl file is already open for writing). There can be any
48 48 number of concurrent read-only opens; O_EXCL is ignored on opens for
49 49 reading. It is recommended that the first open for writing by a
50 50 controlling process use the O_EXCL flag; multiple controlling processes
51 51 usually result in chaos.
52 52
53 53
54 54 If a process opens one of its own /proc files for writing, the open
55 55 succeeds regardless of O_EXCL and regardless of whether some other
56 56 process has the process open for writing. Self-opens do not count when
57 57 another process attempts an exclusive open. (A process cannot exclude a
58 58 debugger by opening itself for writing and the application of a
59 59 debugger cannot prevent a process from opening itself.) All self-opens
60 60 for writing are forced to be close-on-exec (see the F_SETFD operation
61 61 of fcntl(2)).
62 62
63 63
64 64 Data may be transferred from or to any locations in the address space
65 65 of the traced process by applying lseek(2) to position the as file at
66 66 the virtual address of interest followed by read(2) or write(2) (or by
67 67 using pread(2) or pwrite(2) for the combined operation). The address-
68 68 map files /proc/pid/map and /proc/pid/xmap can be read to determine the
69 69 accessible areas (mappings) of the address space. I/O transfers may
70 70 span contiguous mappings. An I/O request extending into an unmapped
71 71 area is truncated at the boundary. A write request beginning at an
72 72 unmapped virtual address fails with EIO; a read request beginning at an
73 73 unmapped virtual address returns zero (an end-of-file indication).
74 74
75 75
76 76 Information and control operations are provided through additional
77 77 files. <procfs.h> contains definitions of data structures and message
78 78 formats used with these files. Some of these definitions involve the
79 79 use of sets of flags. The set types sigset_t, fltset_t, and sysset_t
80 80 correspond, respectively, to signal, fault, and system call
81 81 enumerations defined in <sys/signal.h>, <sys/fault.h>, and
82 82 <sys/syscall.h>. Each set type is large enough to hold flags for its
83 83 own enumeration. Although they are of different sizes, they have a
84 84 common structure and can be manipulated by these macros:
85 85
86 86 prfillset(&set); /* turn on all flags in set */
87 87 premptyset(&set); /* turn off all flags in set */
88 88 praddset(&set, flag); /* turn on the specified flag */
89 89 prdelset(&set, flag); /* turn off the specified flag */
90 90 r = prismember(&set, flag); /* != 0 iff flag is turned on */
91 91
92 92
93 93
94 94 One of prfillset() or premptyset() must be used to initialize set
95 95 before it is used in any other operation. flag must be a member of the
96 96 enumeration corresponding to set.
97 97
98 98
99 99 Every process contains at least one light-weight process, or lwp. Each
100 100 lwp represents a flow of execution that is independently scheduled by
101 101 the operating system. All lwps in a process share its address space as
102 102 well as many other attributes. Through the use of lwpctl and ctl files
103 103 as described below, it is possible to affect individual lwps in a
104 104 process or to affect all of them at once, depending on the operation.
105 105
106 106
107 107 When the process has more than one lwp, a representative lwp is chosen
108 108 by the system for certain process status files and control operations.
109 109 The representative lwp is a stopped lwp only if all of the process's
110 110 lwps are stopped; is stopped on an event of interest only if all of the
111 111 lwps are so stopped (excluding PR_SUSPENDED lwps); is in a PR_REQUESTED
112 112 stop only if there are no other events of interest to be found; or,
113 113 failing everything else, is in a PR_SUSPENDED stop (implying that the
114 114 process is deadlocked). See the description of the status file for
115 115 definitions of stopped states. See the PCSTOP control operation for the
116 116 definition of ``event of interest''.
117 117
118 118
119 119 The representative lwp remains fixed (it will be chosen again on the
120 120 next operation) as long as all of the lwps are stopped on events of
121 121 interest or are in a PR_SUSPENDED stop and the PCRUN control operation
122 122 is not applied to any of them.
123 123
124 124
125 125 When applied to the process control file, every /proc control operation
126 126 that must act on an lwp uses the same algorithm to choose which lwp to
127 127 act upon. Together with synchronous stopping (see PCSET), this enables
128 128 a debugger to control a multiple-lwp process using only the process-
129 129 level status and control files if it so chooses. More fine-grained
130 130 control can be achieved using the lwp-specific files.
131 131
132 132
133 133 The system supports two process data models, the traditional 32-bit
134 134 data model in which ints, longs and pointers are all 32 bits wide (the
135 135 ILP32 data model), and on some platforms the 64-bit data model in which
136 136 longs and pointers, but not ints, are 64 bits in width (the LP64 data
137 137 model). In the LP64 data model some system data types, notably size_t,
138 138 off_t, time_t and dev_t, grow from 32 bits to 64 bits as well.
139 139
140 140
141 141 The /proc interfaces described here are available to both 32-bit and
142 142 64-bit controlling processes. However, many operations attempted by a
143 143 32-bit controlling process on a 64-bit target process will fail with
144 144 EOVERFLOW because the address space range of a 32-bit process cannot
145 145 encompass a 64-bit process or because the data in some 64-bit system
146 146 data type cannot be compressed to fit into the corresponding 32-bit
147 147 type without loss of information. Operations that fail in this
148 148 circumstance include reading and writing the address space, reading the
149 149 address-map files, and setting the target process's registers. There is
150 150 no restriction on operations applied by a 64-bit process to either a
151 151 32-bit or a 64-bit target processes.
152 152
153 153
154 154 The format of the contents of any /proc file depends on the data model
155 155 of the observer (the controlling process), not on the data model of the
156 156 target process. A 64-bit debugger does not have to translate the
157 157 information it reads from a /proc file for a 32-bit process from 32-bit
158 158 format to 64-bit format. However, it usually has to be aware of the
159 159 data model of the target process. The pr_dmodel field of the status
160 160 files indicates the target process's data model.
161 161
162 162
163 163 To help deal with system data structures that are read from 32-bit
164 164 processes, a 64-bit controlling program can be compiled with the C
165 165 preprocessor symbol _SYSCALL32 defined before system header files are
166 166 included. This makes explicit 32-bit fixed-width data structures (like
167 167 cstruct stat32) visible to the 64-bit program. See types32.h(3HEAD).
168 168
169 169 DIRECTORY STRUCTURE
170 170 At the top level, the directory /proc contains entries each of which
171 171 names an existing process in the system. These entries are themselves
172 172 directories. Except where otherwise noted, the files described below
173 173 can be opened for reading only. In addition, if a process becomes a
174 174 zombie (one that has exited but whose parent has not yet performed a
175 175 wait(3C) upon it), most of its associated /proc files disappear from
176 176 the hierarchy; subsequent attempts to open them, or to read or write
177 177 files opened before the process exited, will elicit the error ENOENT.
178 178
179 179
180 180 Although process state and consequently the contents of /proc files can
181 181 change from instant to instant, a single read(2) of a /proc file is
182 182 guaranteed to return a sane representation of state; that is, the read
183 183 will be atomic with respect to the state of the process. No such
184 184 guarantee applies to successive reads applied to a /proc file for a
185 185 running process. In addition, atomicity is not guaranteed for I/O
186 186 applied to the as (address-space) file for a running process or for a
187 187 process whose address space contains memory shared by another running
188 188 process.
189 189
190 190
191 191 A number of structure definitions are used to describe the files. These
192 192 structures may grow by the addition of elements at the end in future
193 193 releases of the system and it is not legitimate for a program to assume
194 194 that they will not.
195 195
196 196 STRUCTURE OF /proc/pid
197 197 A given directory /proc/pid contains the following entries. A process
198 198 can use the invisible alias /proc/self if it wishes to open one of its
199 199 own /proc files (invisible in the sense that the name ``self'' does not
200 200 appear in a directory listing of /proc obtained from ls(1),
201 201 getdents(2), or readdir(3C)).
202 202
203 203 contracts
204 204 A directory containing references to the contracts held by the process.
205 205 Each entry is a symlink to the contract's directory under
206 206 /system/contract. See contract(4).
207 207
208 208 as
209 209 Contains the address-space image of the process; it can be opened for
210 210 both reading and writing. lseek(2) is used to position the file at the
211 211 virtual address of interest and then the address space can be examined
212 212 or changed through read(2) or write(2) (or by using pread(2) or
213 213 pwrite(2) for the combined operation).
214 214
215 215 ctl
216 216 A write-only file to which structured messages are written directing
217 217 the system to change some aspect of the process's state or control its
218 218 behavior in some way. The seek offset is not relevant when writing to
219 219 this file. Individual lwps also have associated lwpctl files in the lwp
220 220 subdirectories. A control message may be written either to the
221 221 process's ctl file or to a specific lwpctl file with operation-specific
222 222 effects. The effect of a control message is immediately reflected in
223 223 the state of the process visible through appropriate status and
224 224 information files. The types of control messages are described in
225 225 detail later. See CONTROL MESSAGES.
226 226
227 227 status
228 228 Contains state information about the process and the representative
229 229 lwp. The file contains a pstatus structure which contains an embedded
230 230 lwpstatus structure for the representative lwp, as follows:
231 231
232 232 typedef struct pstatus {
233 233 int pr_flags; /* flags (see below) */
234 234 int pr_nlwp; /* number of active lwps in the process */
235 235 int pr_nzomb; /* number of zombie lwps in the process */
236 236 pid_tpr_pid; /* process id */
237 237 pid_tpr_ppid; /* parent process id */
238 238 pid_tpr_pgid; /* process group id */
239 239 pid_tpr_sid; /* session id */
240 240 id_t pr_aslwpid; /* obsolete */
241 241 id_t pr_agentid; /* lwp-id of the agent lwp, if any */
242 242 sigset_t pr_sigpend; /* set of process pending signals */
243 243 uintptr_t pr_brkbase; /* virtual address of the process heap */
244 244 size_t pr_brksize; /* size of the process heap, in bytes */
245 245 uintptr_t pr_stkbase; /* virtual address of the process stack */
246 246 size_tpr_stksize; /* size of the process stack, in bytes */
247 247 timestruc_t pr_utime; /* process user cpu time */
248 248 timestruc_t pr_stime; /* process system cpu time */
249 249 timestruc_t pr_cutime; /* sum of children's user times */
250 250 timestruc_t pr_cstime; /* sum of children's system times */
251 251 sigset_t pr_sigtrace; /* set of traced signals */
252 252 fltset_t pr_flttrace; /* set of traced faults */
253 253 sysset_t pr_sysentry; /* set of system calls traced on entry */
254 254 sysset_t pr_sysexit; /* set of system calls traced on exit */
255 255 char pr_dmodel; /* data model of the process */
256 256 taskid_t pr_taskid; /* task id */
257 257 projid_t pr_projid; /* project id */
258 258 zoneid_t pr_zoneid; /* zone id */
259 259 lwpstatus_t pr_lwp; /* status of the representative lwp */
260 260 } pstatus_t;
261 261
262 262
263 263
264 264 pr_flags is a bit-mask holding the following process flags. For
265 265 convenience, it also contains the lwp flags for the representative lwp,
266 266 described later.
267 267
268 268 PR_ISSYS
269 269 process is a system process (see PCSTOP).
270 270
271 271
272 272 PR_VFORKP
273 273 process is the parent of a vforked child (see PCWATCH).
274 274
275 275
276 276 PR_FORK
277 277 process has its inherit-on-fork mode set (see PCSET).
278 278
279 279
280 280 PR_RLC
281 281 process has its run-on-last-close mode set (see PCSET).
282 282
283 283
284 284 PR_KLC
285 285 process has its kill-on-last-close mode set (see PCSET).
286 286
287 287
288 288 PR_ASYNC
289 289 process has its asynchronous-stop mode set (see PCSET).
290 290
291 291
292 292 PR_MSACCT
293 293 Set by default in all processes to indicate that
294 294 microstate accounting is enabled. However, this flag has
295 295 been deprecated and no longer has any effect. Microstate
296 296 accounting may not be disabled; however, it is still
297 297 possible to toggle the flag.
298 298
299 299
300 300 PR_MSFORK
301 301 Set by default in all processes to indicate that
302 302 microstate accounting will be enabled for processes that
303 303 this parent forks(). However, this flag has been
304 304 deprecated and no longer has any effect. It is possible to
305 305 toggle this flag; however, it is not possible to disable
306 306 microstate accounting.
307 307
308 308
309 309 PR_BPTADJ
310 310 process has its breakpoint adjustment mode set (see
311 311 PCSET).
312 312
313 313
314 314 PR_PTRACE
315 315 process has its ptrace-compatibility mode set (see PCSET).
316 316
317 317
318 318
319 319 pr_nlwp is the total number of active lwps in the process. pr_nzomb is
320 320 the total number of zombie lwps in the process. A zombie lwp is a non-
321 321 detached lwp that has terminated but has not been reaped with
322 322 thr_join(3C) or pthread_join(3C).
323 323
324 324
325 325 pr_pid, pr_ppid, pr_pgid, and pr_sid are, respectively, the process ID,
326 326 the ID of the process's parent, the process's process group ID, and the
327 327 process's session ID.
328 328
329 329
330 330 pr_aslwpid is obsolete and is always zero.
331 331
332 332
333 333 pr_agentid is the lwp-ID for the /proc agent lwp (see the PCAGENT
334 334 control operation). It is zero if there is no agent lwp in the process.
335 335
336 336
337 337 pr_sigpend identifies asynchronous signals pending for the process.
338 338
339 339
340 340 pr_brkbase is the virtual address of the process heap and pr_brksize is
341 341 its size in bytes. The address formed by the sum of these values is the
342 342 process break (see brk(2)). pr_stkbase and pr_stksize are,
343 343 respectively, the virtual address of the process stack and its size in
344 344 bytes. (Each lwp runs on a separate stack; the distinguishing
345 345 characteristic of the process stack is that the operating system will
346 346 grow it when necessary.)
347 347
348 348
349 349 pr_utime, pr_stime, pr_cutime, and pr_cstime are, respectively, the
350 350 user CPU and system CPU time consumed by the process, and the
351 351 cumulative user CPU and system CPU time consumed by the process's
352 352 children, in seconds and nanoseconds.
353 353
354 354
355 355 pr_sigtrace and pr_flttrace contain, respectively, the set of signals
356 356 and the set of hardware faults that are being traced (see PCSTRACE and
357 357 PCSFAULT).
358 358
359 359
360 360 pr_sysentry and pr_sysexit contain, respectively, the sets of system
361 361 calls being traced on entry and exit (see PCSENTRY and PCSEXIT).
362 362
363 363
364 364 pr_dmodel indicates the data model of the process. Possible values are:
365 365
366 366 PR_MODEL_ILP32
367 367 process data model is ILP32.
368 368
369 369
370 370 PR_MODEL_LP64
371 371 process data model is LP64.
372 372
373 373
374 374 PR_MODEL_NATIVE
375 375 process data model is native.
376 376
377 377
378 378
379 379 The pr_taskid, pr_projid, and pr_zoneid fields contain respectively,
380 380 the numeric IDs of the task, project, and zone in which the process was
381 381 running.
382 382
383 383
384 384 The constant PR_MODEL_NATIVE reflects the data model of the controlling
385 385 process, that is, its value is PR_MODEL_ILP32 or PR_MODEL_LP64
386 386 according to whether the controlling process has been compiled as a
387 387 32-bit program or a 64-bit program, respectively.
388 388
389 389
390 390 pr_lwp contains the status information for the representative lwp:
391 391
392 392 typedef struct lwpstatus {
393 393 int pr_flags; /* flags (see below) */
394 394 id_t pr_lwpid; /* specific lwp identifier */
395 395 short pr_why; /* reason for lwp stop, if stopped */
396 396 short pr_what; /* more detailed reason */
397 397 short pr_cursig; /* current signal, if any */
398 398 siginfo_t pr_info; /* info associated with signal or fault */
399 399 sigset_t pr_lwppend; /* set of signals pending to the lwp */
400 400 sigset_t pr_lwphold; /* set of signals blocked by the lwp */
401 401 struct sigaction pr_action;/* signal action for current signal */
402 402 stack_t pr_altstack; /* alternate signal stack info */
403 403 uintptr_t pr_oldcontext; /* address of previous ucontext */
404 404 short pr_syscall; /* system call number (if in syscall) */
405 405 short pr_nsysarg; /* number of arguments to this syscall */
406 406 int pr_errno; /* errno for failed syscall */
407 407 long pr_sysarg[PRSYSARGS]; /* arguments to this syscall */
408 408 long pr_rval1; /* primary syscall return value */
409 409 long pr_rval2; /* second syscall return value, if any */
410 410 char pr_clname[PRCLSZ]; /* scheduling class name */
411 411 timestruc_t pr_tstamp; /* real-time time stamp of stop */
412 412 timestruc_t pr_utime; /* lwp user cpu time */
413 413 timestruc_t pr_stime; /* lwp system cpu time */
414 414 uintptr_t pr_ustack; /* stack boundary data (stack_t) address */
415 415 ulong_t pr_instr; /* current instruction */
416 416 prgregset_t pr_reg; /* general registers */
417 417 prfpregset_t pr_fpreg; /* floating-point registers */
418 418 } lwpstatus_t;
419 419
420 420
421 421
422 422 pr_flags is a bit-mask holding the following lwp flags. For
423 423 convenience, it also contains the process flags, described previously.
424 424
425 425 PR_STOPPED
426 426 The lwp is stopped.
427 427
428 428
429 429 PR_ISTOP
430 430 The lwp is stopped on an event of interest (see PCSTOP).
431 431
432 432
433 433 PR_DSTOP
434 434 The lwp has a stop directive in effect (see PCSTOP).
435 435
436 436
437 437 PR_STEP
438 438 The lwp has a single-step directive in effect (see
439 439 PCRUN).
440 440
441 441
442 442 PR_ASLEEP
443 443 The lwp is in an interruptible sleep within a system
444 444 call.
445 445
446 446
447 447 PR_PCINVAL
448 448 The lwp's current instruction (pr_instr) is undefined.
449 449
450 450
451 451 PR_DETACH
452 452 This is a detached lwp (see pthread_create(3C) and
453 453 pthread_join(3C)).
454 454
455 455
456 456 PR_DAEMON
457 457 This is a daemon lwp (see pthread_create(3C)).
458 458
459 459
460 460 PR_ASLWP
461 461 This flag is obsolete and is never set.
462 462
463 463
464 464 PR_AGENT
465 465 This is the /proc agent lwp for the process.
466 466
467 467
468 468
469 469 pr_lwpid names the specific lwp.
470 470
471 471
472 472 pr_why and pr_what together describe, for a stopped lwp, the reason for
473 473 the stop. Possible values of pr_why and the associated pr_what are:
474 474
475 475 PR_REQUESTED
476 476 indicates that the stop occurred in response to a stop
477 477 directive, normally because PCSTOP was applied or
478 478 because another lwp stopped on an event of interest
479 479 and the asynchronous-stop flag (see PCSET) was not set
480 480 for the process. pr_what is unused in this case.
481 481
482 482
483 483 PR_SIGNALLED
484 484 indicates that the lwp stopped on receipt of a signal
485 485 (see PCSTRACE); pr_what holds the signal number that
486 486 caused the stop (for a newly-stopped lwp, the same
487 487 value is in pr_cursig).
488 488
489 489
490 490 PR_FAULTED
491 491 indicates that the lwp stopped on incurring a hardware
492 492 fault (see PCSFAULT); pr_what holds the fault number
493 493 that caused the stop.
494 494
495 495
496 496 PR_SYSENTRY
497 497 PR_SYSEXIT
498 498 indicate a stop on entry to or exit from a system call
499 499 (see PCSENTRY and PCSEXIT); pr_what holds the system
500 500 call number.
501 501
502 502
503 503 PR_JOBCONTROL
504 504 indicates that the lwp stopped due to the default
505 505 action of a job control stop signal (see
|
↓ open down ↓ |
505 lines elided |
↑ open up ↑ |
506 506 sigaction(2)); pr_what holds the stopping signal
507 507 number.
508 508
509 509
510 510 PR_SUSPENDED
511 511 indicates that the lwp stopped due to internal
512 512 synchronization of lwps within the process. pr_what is
513 513 unused in this case.
514 514
515 515
516 + PR_BRAND
517 + indicates that the lwp stopped for a brand-specific
518 + reason. Interpretation of the value of pr_what
519 + depends on which zone brand is in use. It is not
520 + generally expected that an lwp stopped in this state
521 + will be restarted by native proc(4) consumers.
516 522
523 +
524 +
517 525 pr_cursig names the current signal, that is, the next signal to be
518 526 delivered to the lwp, if any. pr_info, when the lwp is in a
519 527 PR_SIGNALLED or PR_FAULTED stop, contains additional information
520 528 pertinent to the particular signal or fault (see <sys/siginfo.h>).
521 529
522 530
523 531 pr_lwppend identifies any synchronous or directed signals pending for
524 532 the lwp. pr_lwphold identifies those signals whose delivery is being
525 533 blocked by the lwp (the signal mask).
526 534
527 535
528 536 pr_action contains the signal action information pertaining to the
529 537 current signal (see sigaction(2)); it is undefined if pr_cursig is
530 538 zero. pr_altstack contains the alternate signal stack information for
531 539 the lwp (see sigaltstack(2)).
532 540
533 541
534 542 pr_oldcontext, if not zero, contains the address on the lwp stack of a
535 543 ucontext structure describing the previous user-level context (see
536 544 ucontext.h(3HEAD)). It is non-zero only if the lwp is executing in the
537 545 context of a signal handler.
538 546
539 547
540 548 pr_syscall is the number of the system call, if any, being executed by
541 549 the lwp; it is non-zero if and only if the lwp is stopped on
542 550 PR_SYSENTRY or PR_SYSEXIT, or is asleep within a system call (
543 551 PR_ASLEEP is set). If pr_syscall is non-zero, pr_nsysarg is the number
544 552 of arguments to the system call and pr_sysarg contains the actual
545 553 arguments.
546 554
547 555
548 556 pr_rval1, pr_rval2, and pr_errno are defined only if the lwp is stopped
549 557 on PR_SYSEXIT or if the PR_VFORKP flag is set. If pr_errno is zero,
550 558 pr_rval1 and pr_rval2 contain the return values from the system call.
551 559 Otherwise, pr_errno contains the error number for the failing system
552 560 call (see <sys/errno.h>).
553 561
554 562
555 563 pr_clname contains the name of the lwp's scheduling class.
556 564
557 565
558 566 pr_tstamp, if the lwp is stopped, contains a time stamp marking when
559 567 the lwp stopped, in real time seconds and nanoseconds since an
560 568 arbitrary time in the past.
561 569
562 570
563 571 pr_utime is the amount of user level CPU time used by this LWP.
564 572
565 573
566 574 pr_stime is the amount of system level CPU time used by this LWP.
567 575
568 576
569 577 pr_ustack is the virtual address of the stack_t that contains the stack
570 578 boundaries for this LWP. See getustack(2) and _stack_grow(3C).
571 579
572 580
573 581 pr_instr contains the machine instruction to which the lwp's program
574 582 counter refers. The amount of data retrieved from the process is
575 583 machine-dependent. On SPARC based machines, it is a 32-bit word. On
576 584 x86-based machines, it is a single byte. In general, the size is that
577 585 of the machine's smallest instruction. If PR_PCINVAL is set, pr_instr
578 586 is undefined; this occurs whenever the lwp is not stopped or when the
579 587 program counter refers to an invalid virtual address.
580 588
581 589
582 590 pr_reg is an array holding the contents of a stopped lwp's general
583 591 registers.
584 592
585 593 SPARC
586 594 On SPARC-based machines, the predefined constants
587 595 R_G0 ... R_G7, R_O0 ... R_O7, R_L0 ... R_L7, R_I0
588 596 ... R_I7, R_PC, R_nPC, and R_Y can be used as
589 597 indices to refer to the corresponding registers;
590 598 previous register windows can be read from their
591 599 overflow locations on the stack (however, see the
592 600 gwindows file in the /proc/pid/lwp/lwpid
593 601 subdirectory).
594 602
595 603
596 604 SPARC V8 (32-bit)
597 605 For SPARC V8 (32-bit) controlling processes, the
598 606 predefined constants R_PSR, R_WIM, and R_TBR can
599 607 be used as indices to refer to the corresponding
600 608 special registers. For SPARC V9 (64-bit)
601 609 controlling processes, the predefined constants
602 610 R_CCR, R_ASI, and R_FPRS can be used as indices to
603 611 refer to the corresponding special registers.
604 612
605 613
606 614 x86 (32-bit)
607 615 For 32-bit x86 processes, the predefined constants
608 616 listed belowcan be used as indices to refer to the
609 617 corresponding registers.
610 618
611 619 SS
612 620 UESP
613 621 EFL
614 622 CS
615 623 EIP
616 624 ERR
617 625 TRAPNO
618 626 EAX
619 627 ECX
620 628 EDX
621 629 EBX
622 630 ESP
623 631 EBP
624 632 ESI
625 633 EDI
626 634 DS
627 635 ES
628 636 GS
629 637
630 638 The preceding constants are listed in
631 639 <sys/regset.h>.
632 640
633 641 Note that a 32-bit process can run on an x86
634 642 64-bit system, using the constants listed above.
635 643
636 644
637 645 x86 (64-bit)
638 646 To read the registers of a 32- or a 64-bit
639 647 process, a 64-bit x86 process should use the
640 648 predefined constants listed below.
641 649
642 650 REG_GSBASE
643 651 REG_FSBASE
644 652 REG_DS
645 653 REG_ES
646 654 REG_GS
647 655 REG_FS
648 656 REG_SS
649 657 REG_RSP
650 658 REG_RFL
651 659 REG_CS
652 660 REG_RIP
653 661 REG_ERR
654 662 REG_TRAPNO
655 663 REG_RAX
656 664 REG_RCX
657 665 REG_RDX
658 666 REG_RBX
659 667 REG_RBP
660 668 REG_RSI
661 669 REG_RDI
662 670 REG_R8
663 671 REG_R9
664 672 REG_R10
665 673 REG_R11
666 674 REG_R12
667 675 REG_R13
668 676 REG_R14
669 677 REG_R15
670 678
671 679 The preceding constants are listed in
672 680 <sys/regset.h>.
673 681
674 682
675 683
676 684 pr_fpreg is a structure holding the contents of the floating-point
677 685 registers.
678 686
679 687
680 688 SPARC registers, both general and floating-point, as seen by a 64-bit
681 689 controlling process are the V9 versions of the registers, even if the
682 690 target process is a 32-bit (V8) process. V8 registers are a subset of
683 691 the V9 registers.
684 692
685 693
686 694 If the lwp is not stopped, all register values are undefined.
687 695
688 696 psinfo
689 697 Contains miscellaneous information about the process and the
690 698 representative lwp needed by the ps(1) command. psinfo remains
691 699 accessible after a process becomes a zombie. The file contains a psinfo
692 700 structure which contains an embedded lwpsinfo structure for the
693 701 representative lwp, as follows:
694 702
695 703 typedef struct psinfo {
696 704 int pr_flag; /* process flags (DEPRECATED: see below) */
697 705 int pr_nlwp; /* number of active lwps in the process */
698 706 int pr_nzomb; /* number of zombie lwps in the process */
699 707 pid_t pr_pid; /* process id */
700 708 pid_t pr_ppid; /* process id of parent */
701 709 pid_t pr_pgid; /* process id of process group leader */
702 710 pid_t pr_sid; /* session id */
703 711 uid_t pr_uid; /* real user id */
704 712 uid_t pr_euid; /* effective user id */
705 713 gid_t pr_gid; /* real group id */
706 714 gid_t pr_egid; /* effective group id */
707 715 uintptr_t pr_addr; /* address of process */
708 716 size_t pr_size; /* size of process image in Kbytes */
709 717 size_t pr_rssize; /* resident set size in Kbytes */
710 718 dev_t pr_ttydev; /* controlling tty device (or PRNODEV) */
711 719 ushort_t pr_pctcpu; /* % of recent cpu time used by all lwps */
712 720 ushort_t pr_pctmem; /* % of system memory used by process */
713 721 timestruc_t pr_start; /* process start time, from the epoch */
714 722 timestruc_t pr_time; /* cpu time for this process */
715 723 timestruc_t pr_ctime; /* cpu time for reaped children */
716 724 char pr_fname[PRFNSZ]; /* name of exec'ed file */
717 725 char pr_psargs[PRARGSZ]; /* initial characters of arg list */
718 726 int pr_wstat; /* if zombie, the wait() status */
719 727 int pr_argc; /* initial argument count */
720 728 uintptr_t pr_argv; /* address of initial argument vector */
721 729 uintptr_t pr_envp; /* address of initial environment vector */
722 730 char pr_dmodel; /* data model of the process */
723 731 lwpsinfo_t pr_lwp; /* information for representative lwp */
724 732 taskid_t pr_taskid; /* task id */
725 733 projid_t pr_projid; /* project id */
726 734 poolid_t pr_poolid; /* pool id */
727 735 zoneid_t pr_zoneid; /* zone id */
728 736 ctid_t pr_contract; /* process contract id */
729 737 } psinfo_t;
730 738
731 739
732 740
733 741 Some of the entries in psinfo, such as pr_addr, refer to internal
734 742 kernel data structures and should not be expected to retain their
735 743 meanings across different versions of the operating system.
736 744
737 745
|
↓ open down ↓ |
211 lines elided |
↑ open up ↑ |
738 746 psinfo_t.pr_flag is a deprecated interface that should no longer be
739 747 used. Applications currently relying on the SSYS bit in pr_flag should
740 748 migrate to checking PR_ISSYS in the pstatus structure's pr_flags field.
741 749
742 750
743 751 pr_pctcpu and pr_pctmem are 16-bit binary fractions in the range 0.0 to
744 752 1.0 with the binary point to the right of the high-order bit (1.0 ==
745 753 0x8000). pr_pctcpu is the summation over all lwps in the process.
746 754
747 755
756 + The pr_fname and pr_psargs are writable by the owner of the process. To
757 + write to them, the psinfo file should be open for writing and the
758 + desired value for the field should be written at the file offset that
759 + corresponds to the member of structure. No other entry may be written
760 + to; if a write is attempted to an offset that does not represent one of
761 + these two memers, or if the size of the write is not exactly the size
762 + of the member being written, no bytes will be written and zero will be
763 + returned.
764 +
765 +
748 766 pr_lwp contains the ps(1) information for the representative lwp. If
749 767 the process is a zombie, pr_nlwp, pr_nzomb, and pr_lwp.pr_lwpid are
750 768 zero and the other fields of pr_lwp are undefined:
751 769
752 770 typedef struct lwpsinfo {
753 771 int pr_flag; /* lwp flags (DEPRECATED: see below) */
754 772 id_t pr_lwpid; /* lwp id */
755 773 uintptr_t pr_addr; /* internal address of lwp */
756 774 uintptr_t pr_wchan; /* wait addr for sleeping lwp */
757 775 char pr_stype; /* synchronization event type */
758 776 char pr_state; /* numeric lwp state */
759 777 char pr_sname; /* printable character for pr_state */
760 778 char pr_nice; /* nice for cpu usage */
761 779 short pr_syscall; /* system call number (if in syscall) */
762 780 char pr_oldpri; /* pre-SVR4, low value is high priority */
763 781 char pr_cpu; /* pre-SVR4, cpu usage for scheduling */
764 782 int pr_pri; /* priority, high value = high priority */
765 783 ushort_t pr_pctcpu; /* % of recent cpu time used by this lwp */
766 784 timestruc_t pr_start; /* lwp start time, from the epoch */
767 785 timestruc_t pr_time; /* cpu time for this lwp */
768 786 char pr_clname[PRCLSZ]; /* scheduling class name */
769 787 char pr_name[PRFNSZ]; /* name of system lwp */
770 788 processorid_t pr_onpro; /* processor which last ran this lwp */
771 789 processorid_t pr_bindpro;/* processor to which lwp is bound */
772 790 psetid_t pr_bindpset; /* processor set to which lwp is bound */
773 791 lgrp_id_t pr_lgrp /* home lgroup */
774 792 } lwpsinfo_t;
775 793
776 794
777 795
778 796 Some of the entries in lwpsinfo, such as pr_addr, pr_wchan, pr_stype,
779 797 pr_state, and pr_name, refer to internal kernel data structures and
780 798 should not be expected to retain their meanings across different
781 799 versions of the operating system.
782 800
783 801
784 802 lwpsinfo_t.pr_flag is a deprecated interface that should no longer be
785 803 used.
786 804
787 805
788 806 pr_pctcpu is a 16-bit binary fraction, as described above. It
789 807 represents the CPU time used by the specific lwp. On a multi-processor
790 808 machine, the maximum value is 1/N, where N is the number of CPUs.
791 809
792 810
793 811 pr_contract is the id of the process contract of which the process is a
794 812 member. See contract(4) and process(4).
795 813
796 814 cred
797 815 Contains a description of the credentials associated with the process:
798 816
799 817 typedef struct prcred {
800 818 uid_t pr_euid; /* effective user id */
801 819 uid_t pr_ruid; /* real user id */
802 820 uid_t pr_suid; /* saved user id (from exec) */
803 821 gid_t pr_egid; /* effective group id */
804 822 gid_t pr_rgid; /* real group id */
805 823 gid_t pr_sgid; /* saved group id (from exec) */
806 824 int pr_ngroups; /* number of supplementary groups */
807 825 gid_t pr_groups[1]; /* array of supplementary groups */
808 826 } prcred_t;
809 827
810 828
811 829
812 830
813 831 The array of associated supplementary groups in pr_groups is of
814 832 variable length; the cred file contains all of the supplementary
815 833 groups. pr_ngroups indicates the number of supplementary groups. (See
816 834 also the PCSCRED and PCSCREDX control operations.)
817 835
818 836 priv
819 837 Contains a description of the privileges associated with the process:
820 838
821 839 typedef struct prpriv {
822 840 uint32_t pr_nsets; /* number of privilege set */
823 841 uint32_t pr_setsize; /* size of privilege set */
824 842 uint32_t pr_infosize; /* size of supplementary data */
825 843 priv_chunk_t pr_sets[1]; /* array of sets */
826 844 } prpriv_t;
827 845
828 846
829 847
830 848 The actual dimension of the pr_sets[] field is
831 849
832 850 pr_sets[pr_nsets][pr_setsize]
833 851
834 852
835 853
836 854 which is followed by additional information about the process state
837 855 pr_infosize bytes in size.
838 856
839 857
840 858 The full size of the structure can be computed using
841 859 PRIV_PRPRIV_SIZE(prpriv_t *).
842 860
843 861 sigact
844 862 Contains an array of sigaction structures describing the current
845 863 dispositions of all signals associated with the traced process (see
|
↓ open down ↓ |
88 lines elided |
↑ open up ↑ |
846 864 sigaction(2)). Signal numbers are displaced by 1 from array indices, so
847 865 that the action for signal number n appears in position n-1 of the
848 866 array.
849 867
850 868 auxv
851 869 Contains the initial values of the process's aux vector in an array of
852 870 auxv_t structures (see <sys/auxv.h>). The values are those that were
853 871 passed by the operating system as startup information to the dynamic
854 872 linker.
855 873
874 + argv
875 + Contains the concatenation of each of the argument strings, including
876 + their NUL terminators, in the argument vector (argv) for the process.
877 + If the process has modified either its argument vector, or the contents
878 + of any of the strings referenced by that vector, those changes will be
879 + visible here.
880 +
856 881 ldt
857 882 This file exists only on x86-based machines. It is non-empty only if
858 883 the process has established a local descriptor table (LDT). If non-
859 884 empty, the file contains the array of currently active LDT entries in
860 885 an array of elements of type struct ssd, defined in <sys/sysi86.h>, one
861 886 element for each active LDT entry.
862 887
863 888 map, xmap
864 889 Contain information about the virtual address map of the process. The
865 890 map file contains an array of prmap structures while the xmap file
866 891 contains an array of prxmap structures. Each structure describes a
867 892 contiguous virtual address region in the address space of the traced
868 893 process:
869 894
870 895 typedef struct prmap {
871 896 uintptr_tpr_vaddr; /* virtual address of mapping */
872 897 size_t pr_size; /* size of mapping in bytes */
873 898 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
874 899 offset_t pr_offset; /* offset into mapped object, if any */
875 900 int pr_mflags; /* protection and attribute flags */
876 901 int pr_pagesize; /* pagesize for this mapping in bytes */
877 902 int pr_shmid; /* SysV shared memory identifier */
878 903 } prmap_t;
879 904
880 905
881 906
882 907 typedef struct prxmap {
883 908 uintptr_t pr_vaddr; /* virtual address of mapping */
884 909 size_t pr_size; /* size of mapping in bytes */
885 910 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
886 911 offset_t pr_offset; /* offset into mapped object, if any */
887 912 int pr_mflags; /* protection and attribute flags */
888 913 int pr_pagesize; /* pagesize for this mapping in bytes */
889 914 int pr_shmid; /* SysV shared memory identifier */
890 915 dev_t pr_dev; /* device of mapped object, if any */
891 916 uint64_t pr_ino; /* inode of mapped object, if any */
892 917 size_t pr_rss; /* pages of resident memory */
893 918 size_t pr_anon; /* pages of resident anonymous memory */
894 919 size_t pr_locked; /* pages of locked memory */
895 920 uint64_t pr_hatpagesize; /* pagesize of mapping */
896 921 } prxmap_t;
897 922
898 923
899 924
900 925
901 926 pr_vaddr is the virtual address of the mapping within the traced
902 927 process and pr_size is its size in bytes. pr_mapname, if it does not
903 928 contain a null string, contains the name of a file in the object
904 929 directory (see below) that can be opened read-only to obtain a file
905 930 descriptor for the mapped file associated with the mapping. This
906 931 enables a debugger to find object file symbol tables without having to
907 932 know the real path names of the executable file and shared libraries of
908 933 the process. pr_offset is the 64-bit offset within the mapped file (if
909 934 any) to which the virtual address is mapped.
910 935
911 936
912 937 pr_mflags is a bit-mask of protection and attribute flags:
913 938
914 939 MA_READ
915 940 mapping is readable by the traced process.
916 941
917 942
918 943 MA_WRITE
919 944 mapping is writable by the traced process.
920 945
921 946
922 947 MA_EXEC
923 948 mapping is executable by the traced process.
924 949
925 950
926 951 MA_SHARED
927 952 mapping changes are shared by the mapped object.
928 953
929 954
930 955 MA_ISM
931 956 mapping is intimate shared memory (shared MMU
932 957 resources)
933 958
934 959
935 960 MAP_NORESERVE
936 961 mapping does not have swap space reserved (mapped with
937 962 MAP_NORESERVE)
938 963
939 964
940 965 MA_SHM
941 966 mapping System V shared memory
942 967
943 968
944 969
945 970 A contiguous area of the address space having the same underlying
946 971 mapped object may appear as multiple mappings due to varying read,
947 972 write, and execute attributes. The underlying mapped object does not
948 973 change over the range of a single mapping. An I/O operation to a
949 974 mapping marked MA_SHARED fails if applied at a virtual address not
950 975 corresponding to a valid page in the underlying mapped object. A write
951 976 to a MA_SHARED mapping that is not marked MA_WRITE fails. Reads and
952 977 writes to private mappings always succeed. Reads and writes to unmapped
953 978 addresses fail.
954 979
955 980
956 981 pr_pagesize is the page size for the mapping, currently always the
957 982 system pagesize.
958 983
959 984
960 985 pr_shmid is the shared memory identifier, if any, for the mapping. Its
961 986 value is -1 if the mapping is not System V shared memory. See
962 987 shmget(2).
963 988
964 989
965 990 pr_dev is the device of the mapped object, if any, for the mapping. Its
966 991 value is PRNODEV (-1) if the mapping does not have a device.
967 992
968 993
969 994 pr_ino is the inode of the mapped object, if any, for the mapping. Its
970 995 contents are only valid if pr_dev is not PRNODEV.
971 996
972 997
973 998 pr_rss is the number of resident pages of memory for the mapping. The
974 999 number of resident bytes for the mapping may be determined by
975 1000 multiplying pr_rss by the page size given by pr_pagesize.
976 1001
977 1002
978 1003 pr_anon is the number of resident anonymous memory pages (pages which
979 1004 are private to this process) for the mapping.
980 1005
981 1006
982 1007 pr_locked is the number of locked pages for the mapping. Pages which
983 1008 are locked are always resident in memory.
984 1009
985 1010
986 1011 pr_hatpagesize is the size, in bytes, of the HAT (MMU) translation for
987 1012 the mapping. pr_hatpagesize may be different than pr_pagesize. The
988 1013 possible values are hardware architecture specific, and may change over
989 1014 a mapping's lifetime.
990 1015
991 1016 rmap
992 1017 Contains information about the reserved address ranges of the process.
993 1018 The file contains an array of prmap structures, as defined above for
994 1019 the map file. Each structure describes a contiguous virtual address
995 1020 region in the address space of the traced process that is reserved by
996 1021 the system in the sense that an mmap(2) system call that does not
997 1022 specify MAP_FIXED will not use any part of it for the new mapping.
998 1023 Examples of such reservations include the address ranges reserved for
999 1024 the process stack and the individual thread stacks of a multi-threaded
1000 1025 process.
1001 1026
1002 1027 cwd
1003 1028 A symbolic link to the process's current working directory. See
1004 1029 chdir(2). A readlink(2) of /proc/pid/cwd yields a null string.
1005 1030 However, it can be opened, listed, and searched as a directory, and can
1006 1031 be the target of chdir(2).
1007 1032
1008 1033 root
1009 1034 A symbolic link to the process's root directory. /proc/pid/root can
1010 1035 differ from the system root directory if the process or one of its
1011 1036 ancestors executed chroot(2) as super user. It has the same semantics
1012 1037 as /proc/pid/cwd.
1013 1038
1014 1039 fd
1015 1040 A directory containing references to the open files of the process.
1016 1041 Each entry is a decimal number corresponding to an open file descriptor
1017 1042 in the process.
1018 1043
1019 1044
1020 1045 If an entry refers to a regular file, it can be opened with normal file
1021 1046 system semantics but, to ensure that the controlling process cannot
1022 1047 gain greater access than the controlled process, with no file access
1023 1048 modes other than its read/write open modes in the controlled process.
1024 1049 If an entry refers to a directory, it can be accessed with the same
1025 1050 semantics as /proc/pid/cwd. An attempt to open any other type of entry
1026 1051 fails with EACCES.
1027 1052
1028 1053 object
1029 1054 A directory containing read-only files with names corresponding to the
1030 1055 pr_mapname entries in the map and pagedata files. Opening such a file
1031 1056 yields a file descriptor for the underlying mapped file associated with
1032 1057 an address-space mapping in the process. The file name a.out appears in
1033 1058 the directory as an alias for the process's executable file.
1034 1059
1035 1060
1036 1061 The object directory makes it possible for a controlling process to
1037 1062 gain access to the object file and any shared libraries (and
1038 1063 consequently the symbol tables) without having to know the actual path
1039 1064 names of the executable files.
1040 1065
1041 1066 path
1042 1067 A directory containing symbolic links to files opened by the process.
1043 1068 The directory includes one entry for cwd and root. The directory also
1044 1069 contains a numerical entry for each file descriptor in the fd
1045 1070 directory, and entries matching those in the object directory. If this
1046 1071 information is not available, any attempt to read the contents of the
1047 1072 symbolic link will fail. This is most common for files that do not
1048 1073 exist in the filesystem namespace (such as FIFOs and sockets), but can
1049 1074 also happen for regular files. For the file descriptor entries, the
1050 1075 path may be different from the one used by the process to open the
1051 1076 file.
1052 1077
1053 1078 pagedata
1054 1079 Opening the page data file enables tracking of address space references
1055 1080 and modifications on a per-page basis.
1056 1081
1057 1082
1058 1083 A read(2) of the page data file descriptor returns structured page data
1059 1084 and atomically clears the page data maintained for the file by the
1060 1085 system. That is to say, each read returns data collected since the last
1061 1086 read; the first read returns data collected since the file was opened.
1062 1087 When the call completes, the read buffer contains the following
1063 1088 structure as its header and thereafter contains a number of section
1064 1089 header structures and associated byte arrays that must be accessed by
1065 1090 walking linearly through the buffer.
1066 1091
1067 1092 typedef struct prpageheader {
1068 1093 timestruc_t pr_tstamp; /* real time stamp, time of read() */
1069 1094 ulong_t pr_nmap; /* number of address space mappings */
1070 1095 ulong_t pr_npage; /* total number of pages */
1071 1096 } prpageheader_t;
1072 1097
1073 1098
1074 1099
1075 1100 The header is followed by pr_nmap prasmap structures and associated
1076 1101 data arrays. The prasmap structure contains the following elements:
1077 1102
1078 1103 typedef struct prasmap {
1079 1104 uintptr_t pr_vaddr; /* virtual address of mapping */
1080 1105 ulong_t pr_npage; /* number of pages in mapping */
1081 1106 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
1082 1107 offset_t pr_offset; /* offset into mapped object, if any */
1083 1108 int pr_mflags; /* protection and attribute flags */
1084 1109 int pr_pagesize; /* pagesize for this mapping in bytes */
1085 1110 int pr_shmid; /* SysV shared memory identifier */
1086 1111 } prasmap_t;
1087 1112
1088 1113
1089 1114
1090 1115 Each section header is followed by pr_npage bytes, one byte for each
1091 1116 page in the mapping, plus 0-7 null bytes at the end so that the next
1092 1117 prasmap structure begins on an eight-byte aligned boundary. Each data
1093 1118 byte may contain these flags:
1094 1119
1095 1120 PG_REFERENCED
1096 1121 page has been referenced.
1097 1122
1098 1123
1099 1124 PG_MODIFIED
1100 1125 page has been modified.
1101 1126
1102 1127
1103 1128
1104 1129 If the read buffer is not large enough to contain all of the page data,
1105 1130 the read fails with E2BIG and the page data is not cleared. The
1106 1131 required size of the read buffer can be determined through fstat(2).
1107 1132 Application of lseek(2) to the page data file descriptor is
1108 1133 ineffective; every read starts from the beginning of the file. Closing
1109 1134 the page data file descriptor terminates the system overhead associated
1110 1135 with collecting the data.
1111 1136
1112 1137
1113 1138 More than one page data file descriptor for the same process can be
1114 1139 opened, up to a system-imposed limit per traced process. A read of one
1115 1140 does not affect the data being collected by the system for the others.
1116 1141 An open of the page data file will fail with ENOMEM if the system-
1117 1142 imposed limit would be exceeded.
1118 1143
1119 1144 watch
1120 1145 Contains an array of prwatch structures, one for each watched area
1121 1146 established by the PCWATCH control operation. See PCWATCH for details.
1122 1147
1123 1148 usage
1124 1149 Contains process usage information described by a prusage structure
1125 1150 which contains at least the following fields:
1126 1151
1127 1152 typedef struct prusage {
1128 1153 id_t pr_lwpid; /* lwp id. 0: process or defunct */
1129 1154 int pr_count; /* number of contributing lwps */
1130 1155 timestruc_t pr_tstamp; /* real time stamp, time of read() */
1131 1156 timestruc_t pr_create; /* process/lwp creation time stamp */
1132 1157 timestruc_t pr_term; /* process/lwp termination time stamp */
1133 1158 timestruc_t pr_rtime; /* total lwp real (elapsed) time */
1134 1159 timestruc_t pr_utime; /* user level CPU time */
1135 1160 timestruc_t pr_stime; /* system call CPU time */
1136 1161 timestruc_t pr_ttime; /* other system trap CPU time */
1137 1162 timestruc_t pr_tftime; /* text page fault sleep time */
1138 1163 timestruc_t pr_dftime; /* data page fault sleep time */
1139 1164 timestruc_t pr_kftime; /* kernel page fault sleep time */
1140 1165 timestruc_t pr_ltime; /* user lock wait sleep time */
1141 1166 timestruc_t pr_slptime; /* all other sleep time */
1142 1167 timestruc_t pr_wtime; /* wait-cpu (latency) time */
1143 1168 timestruc_t pr_stoptime; /* stopped time */
1144 1169 ulong_t pr_minf; /* minor page faults */
1145 1170 ulong_t pr_majf; /* major page faults */
1146 1171 ulong_t pr_nswap; /* swaps */
1147 1172 ulong_t pr_inblk; /* input blocks */
1148 1173 ulong_t pr_oublk; /* output blocks */
1149 1174 ulong_t pr_msnd; /* messages sent */
1150 1175 ulong_t pr_mrcv; /* messages received */
1151 1176 ulong_t pr_sigs; /* signals received */
1152 1177 ulong_t pr_vctx; /* voluntary context switches */
1153 1178 ulong_t pr_ictx; /* involuntary context switches */
1154 1179 ulong_t pr_sysc; /* system calls */
1155 1180 ulong_t pr_ioch; /* chars read and written */
1156 1181 } prusage_t;
1157 1182
1158 1183
1159 1184
1160 1185 Microstate accounting is now continuously enabled. While this
1161 1186 information was previously an estimate, if microstate accounting were
1162 1187 not enabled, the current information is now never an estimate
1163 1188 represents time the process has spent in various states.
1164 1189
1165 1190 lstatus
1166 1191 Contains a prheader structure followed by an array of lwpstatus
1167 1192 structures, one for each active lwp in the process (see also
1168 1193 /proc/pid/lwp/lwpid/lwpstatus, below). The prheader structure describes
1169 1194 the number and size of the array entries that follow.
1170 1195
1171 1196 typedef struct prheader {
1172 1197 long pr_nent; /* number of entries */
1173 1198 size_t pr_entsize; /* size of each entry, in bytes */
1174 1199 } prheader_t;
1175 1200
1176 1201
1177 1202
1178 1203 The lwpstatus structure may grow by the addition of elements at the end
1179 1204 in future releases of the system. Programs must use pr_entsize in the
1180 1205 file header to index through the array. These comments apply to all
1181 1206 /proc files that include a prheader structure (lpsinfo and lusage,
1182 1207 below).
1183 1208
1184 1209 lpsinfo
1185 1210 Contains a prheader structure followed by an array of lwpsinfo
1186 1211 structures, one for eachactive and zombie lwp in the process. See also
1187 1212 /proc/pid/lwp/lwpid/lwpsinfo, below.
1188 1213
1189 1214 lusage
1190 1215 Contains a prheader structure followed by an array of prusage
1191 1216 structures, one for each active lwp in the process, plus an additional
1192 1217 element at the beginning that contains the summation over all defunct
1193 1218 lwps (lwps that once existed but no longer exist in the process).
1194 1219 Excluding the pr_lwpid, pr_tstamp, pr_create, and pr_term entries, the
1195 1220 entry-by-entry summation over all these structures is the definition of
1196 1221 the process usage information obtained from the usage file. (See also
1197 1222 /proc/pid/lwp/lwpid/lwpusage, below.)
1198 1223
1199 1224 lwp
1200 1225 A directory containing entries each of which names an active or zombie
1201 1226 lwp within the process. These entries are themselves directories
1202 1227 containing additional files as described below. Only the lwpsinfo file
1203 1228 exists in the directory of a zombie lwp.
1204 1229
1205 1230 STRUCTURE OF /proc/pid/lwp/lwpid
1206 1231 A given directory /proc/pid/lwp/lwpid contains the following entries:
1207 1232
1208 1233 lwpctl
1209 1234 Write-only control file. The messages written to this file affect the
1210 1235 specific lwp rather than the representative lwp, as is the case for the
1211 1236 process's ctl file.
1212 1237
1213 1238 lwpstatus
1214 1239 lwp-specific state information. This file contains the lwpstatus
1215 1240 structure for the specific lwp as described above for the
1216 1241 representative lwp in the process's status file.
1217 1242
1218 1243 lwpsinfo
1219 1244 lwp-specific ps(1) information. This file contains the lwpsinfo
1220 1245 structure for the specific lwp as described above for the
1221 1246 representative lwp in the process's psinfo file. The lwpsinfo file
1222 1247 remains accessible after an lwp becomes a zombie.
1223 1248
1224 1249 lwpusage
1225 1250 This file contains the prusage structure for the specific lwp as
1226 1251 described above for the process's usage file.
1227 1252
1228 1253 gwindows
1229 1254 This file exists only on SPARC based machines. If it is non-empty, it
1230 1255 contains a gwindows_t structure, defined in <sys/regset.h>, with the
1231 1256 values of those SPARC register windows that could not be stored on the
1232 1257 stack when the lwp stopped. Conditions under which register windows are
1233 1258 not stored on the stack are: the stack pointer refers to nonexistent
1234 1259 process memory or the stack pointer is improperly aligned. If the lwp
1235 1260 is not stopped or if there are no register windows that could not be
1236 1261 stored on the stack, the file is empty (the usual case).
1237 1262
1238 1263 xregs
1239 1264 Extra state registers. The extra state register set is architecture
1240 1265 dependent; this file is empty if the system does not support extra
1241 1266 state registers. If the file is non-empty, it contains an architecture
1242 1267 dependent structure of type prxregset_t, defined in <procfs.h>, with
1243 1268 the values of the lwp's extra state registers. If the lwp is not
1244 1269 stopped, all register values are undefined. See also the PCSXREG
1245 1270 control operation, below.
1246 1271
1247 1272 asrs
1248 1273 This file exists only for 64-bit SPARC V9 processes. It contains an
1249 1274 asrset_t structure, defined in <sys/regset.h>, containing the values of
1250 1275 the lwp's platform-dependent ancillary state registers. If the lwp is
1251 1276 not stopped, all register values are undefined. See also the PCSASRS
1252 1277 control operation, below.
1253 1278
1254 1279 spymaster
1255 1280 For an agent lwp (see PCAGENT), this file contains a psinfo_t structure
1256 1281 that corresponds to the process that created the agent lwp at the time
1257 1282 the agent was created. This structure is identical to that retrieved
1258 1283 via the psinfo file, with one modification: the pr_time field does not
1259 1284 correspond to the CPU time for the process, but rather to the creation
1260 1285 time of the agent lwp.
1261 1286
1262 1287 templates
1263 1288 A directory which contains references to the active templates for the
1264 1289 lwp, named by the contract type. Changes made to an active template
1265 1290 descriptor do not affect the original template which was activated,
1266 1291 though they do affect the active template. It is not possible to
1267 1292 activate an active template descriptor. See contract(4).
1268 1293
1269 1294 CONTROL MESSAGES
1270 1295 Process state changes are effected through messages written to a
1271 1296 process's ctl file or to an individual lwp's lwpctl file. All control
1272 1297 messages consist of a long that names the specific operation followed
1273 1298 by additional data containing the operand, if any.
1274 1299
1275 1300
1276 1301 Multiple control messages may be combined in a single write(2) (or
1277 1302 writev(2)) to a control file, but no partial writes are permitted. That
1278 1303 is, each control message, operation code plus operand, if any, must be
1279 1304 presented in its entirety to the write(2) and not in pieces over
1280 1305 several system calls. If a control operation fails, no subsequent
1281 1306 operations contained in the same write(2) are attempted.
1282 1307
1283 1308
1284 1309 Descriptions of the allowable control messages follow. In all cases,
1285 1310 writing a message to a control file for a process or lwp that has
1286 1311 terminated elicits the error ENOENT.
1287 1312
1288 1313 PCSTOP PCDSTOP PCWSTOP PCTWSTOP
1289 1314 When applied to the process control file, PCSTOP directs all lwps to
1290 1315 stop and waits for them to stop, PCDSTOP directs all lwps to stop
1291 1316 without waiting for them to stop, and PCWSTOP simply waits for all lwps
1292 1317 to stop. When applied to an lwp control file, PCSTOP directs the
1293 1318 specific lwp to stop and waits until it has stopped, PCDSTOP directs
1294 1319 the specific lwp to stop without waiting for it to stop, and PCWSTOP
1295 1320 simply waits for the specific lwp to stop. When applied to an lwp
1296 1321 control file, PCSTOP and PCWSTOP complete when the lwp stops on an
1297 1322 event of interest, immediately if already so stopped; when applied to
1298 1323 the process control file, they complete when every lwp has stopped
1299 1324 either on an event of interest or on a PR_SUSPENDED stop.
1300 1325
1301 1326
1302 1327 PCTWSTOP is identical to PCWSTOP except that it enables the operation
1303 1328 to time out, to avoid waiting forever for a process or lwp that may
1304 1329 never stop on an event of interest. PCTWSTOP takes a long operand
1305 1330 specifying a number of milliseconds; the wait will terminate
1306 1331 successfully after the specified number of milliseconds even if the
1307 1332 process or lwp has not stopped; a timeout value of zero makes the
1308 1333 operation identical to PCWSTOP.
1309 1334
1310 1335
1311 1336 An ``event of interest'' is either a PR_REQUESTED stop or a stop that
1312 1337 has been specified in the process's tracing flags (set by PCSTRACE,
1313 1338 PCSFAULT, PCSENTRY, and PCSEXIT). PR_JOBCONTROL and PR_SUSPENDED stops
1314 1339 are specifically not events of interest. (An lwp may stop twice due to
1315 1340 a stop signal, first showing PR_SIGNALLED if the signal is traced and
1316 1341 again showing PR_JOBCONTROL if the lwp is set running without clearing
1317 1342 the signal.) If PCSTOP or PCDSTOP is applied to an lwp that is stopped,
1318 1343 but not on an event of interest, the stop directive takes effect when
1319 1344 the lwp is restarted by the competing mechanism. At that time, the lwp
1320 1345 enters a PR_REQUESTED stop before executing any user-level code.
1321 1346
1322 1347
1323 1348 A write of a control message that blocks is interruptible by a signal
1324 1349 so that, for example, an alarm(2) can be set to avoid waiting forever
1325 1350 for a process or lwp that may never stop on an event of interest. If
1326 1351 PCSTOP is interrupted, the lwp stop directives remain in effect even
1327 1352 though the write(2) returns an error. (Use of PCTWSTOP with a non-zero
1328 1353 timeout is recommended over PCWSTOP with an alarm(2).)
1329 1354
1330 1355
1331 1356 A system process (indicated by the PR_ISSYS flag) never executes at
1332 1357 user level, has no user-level address space visible through /proc, and
1333 1358 cannot be stopped. Applying one of these operations to a system process
1334 1359 or any of its lwps elicits the error EBUSY.
1335 1360
1336 1361 PCRUN
1337 1362 Make an lwp runnable again after a stop. This operation takes a long
1338 1363 operand containing zero or more of the following flags:
1339 1364
1340 1365 PRCSIG
1341 1366 clears the current signal, if any (see PCCSIG).
1342 1367
1343 1368
1344 1369 PRCFAULT
1345 1370 clears the current fault, if any (see PCCFAULT).
1346 1371
1347 1372
1348 1373 PRSTEP
1349 1374 directs the lwp to execute a single machine instruction. On
1350 1375 completion of the instruction, a trace trap occurs. If
1351 1376 FLTTRACE is being traced, the lwp stops; otherwise, it is
1352 1377 sent SIGTRAP. If SIGTRAP is being traced and is not
1353 1378 blocked, the lwp stops. When the lwp stops on an event of
1354 1379 interest, the single-step directive is cancelled, even if
1355 1380 the stop occurs before the instruction is executed. This
1356 1381 operation requires hardware and operating system support
1357 1382 and may not be implemented on all processors. It is
1358 1383 implemented on SPARC and x86-based machines.
1359 1384
1360 1385
1361 1386 PRSABORT
1362 1387 is meaningful only if the lwp is in a PR_SYSENTRY stop or
1363 1388 is marked PR_ASLEEP; it instructs the lwp to abort
1364 1389 execution of the system call (see PCSENTRY and PCSEXIT).
1365 1390
1366 1391
1367 1392 PRSTOP
1368 1393 directs the lwp to stop again as soon as possible after
1369 1394 resuming execution (see PCDSTOP). In particular, if the lwp
1370 1395 is stopped on PR_SIGNALLED or PR_FAULTED, the next stop
1371 1396 will show PR_REQUESTED, no other stop will have intervened,
1372 1397 and the lwp will not have executed any user-level code.
1373 1398
1374 1399
1375 1400
1376 1401 When applied to an lwp control file, PCRUN clears any outstanding
1377 1402 directed-stop request and makes the specific lwp runnable. The
1378 1403 operation fails with EBUSY if the specific lwp is not stopped on an
1379 1404 event of interest or has not been directed to stop or if the agent lwp
1380 1405 exists and this is not the agent lwp (see PCAGENT).
1381 1406
1382 1407
1383 1408 When applied to the process control file, a representative lwp is
1384 1409 chosen for the operation as described for /proc/pid/status. The
1385 1410 operation fails with EBUSY if the representative lwp is not stopped on
1386 1411 an event of interest or has not been directed to stop or if the agent
1387 1412 lwp exists. If PRSTEP or PRSTOP was requested, the representative lwp
1388 1413 is made runnable and its outstanding directed-stop request is cleared;
1389 1414 otherwise all outstanding directed-stop requests are cleared and, if it
1390 1415 was stopped on an event of interest, the representative lwp is marked
1391 1416 PR_REQUESTED. If, as a consequence, all lwps are in the PR_REQUESTED or
1392 1417 PR_SUSPENDED stop state, all lwps showing PR_REQUESTED are made
1393 1418 runnable.
1394 1419
1395 1420 PCSTRACE
1396 1421 Define a set of signals to be traced in the process. The receipt of one
1397 1422 of these signals by an lwp causes the lwp to stop. The set of signals
1398 1423 is defined using an operand sigset_t contained in the control message.
1399 1424 Receipt of SIGKILL cannot be traced; if specified, it is silently
1400 1425 ignored.
1401 1426
1402 1427
1403 1428 If a signal that is included in an lwp's held signal set (the signal
1404 1429 mask) is sent to the lwp, the signal is not received and does not cause
1405 1430 a stop until it is removed from the held signal set, either by the lwp
1406 1431 itself or by setting the held signal set with PCSHOLD.
1407 1432
1408 1433 PCCSIG
1409 1434 The current signal, if any, is cleared from the specific or
1410 1435 representative lwp.
1411 1436
1412 1437 PCSSIG
1413 1438 The current signal and its associated signal information for the
1414 1439 specific or representative lwp are set according to the contents of the
1415 1440 operand siginfo structure (see <sys/siginfo.h>). If the specified
1416 1441 signal number is zero, the current signal is cleared. The semantics of
1417 1442 this operation are different from those of kill(2) in that the signal
1418 1443 is delivered to the lwp immediately after execution is resumed (even if
1419 1444 it is being blocked) and an additional PR_SIGNALLED stop does not
1420 1445 intervene even if the signal is traced. Setting the current signal to
1421 1446 SIGKILL terminates the process immediately.
1422 1447
1423 1448 PCKILL
1424 1449 If applied to the process control file, a signal is sent to the process
1425 1450 with semantics identical to those of kill(2). If applied to an lwp
1426 1451 control file, a directed signal is sent to the specific lwp. The signal
1427 1452 is named in a long operand contained in the message. Sending SIGKILL
1428 1453 terminates the process immediately.
1429 1454
1430 1455 PCUNKILL
1431 1456 A signal is deleted, that is, it is removed from the set of pending
1432 1457 signals. If applied to the process control file, the signal is deleted
1433 1458 from the process's pending signals. If applied to an lwp control file,
1434 1459 the signal is deleted from the lwp's pending signals. The current
1435 1460 signal (if any) is unaffected. The signal is named in a long operand in
1436 1461 the control message. It is an error (EINVAL) to attempt to delete
1437 1462 SIGKILL.
1438 1463
1439 1464 PCSHOLD
1440 1465 Set the set of held signals for the specific or representative lwp
1441 1466 (signals whose delivery will be blocked if sent to the lwp). The set of
1442 1467 signals is specified with a sigset_t operand. SIGKILL and SIGSTOP
1443 1468 cannot be held; if specified, they are silently ignored.
1444 1469
1445 1470 PCSFAULT
1446 1471 Define a set of hardware faults to be traced in the process. On
1447 1472 incurring one of these faults, an lwp stops. The set is defined via the
1448 1473 operand fltset_t structure. Fault names are defined in <sys/fault.h>
1449 1474 and include the following. Some of these may not occur on all
1450 1475 processors; there may be processor-specific faults in addition to
1451 1476 these.
1452 1477
1453 1478 FLTILL
1454 1479 illegal instruction
1455 1480
1456 1481
1457 1482 FLTPRIV
1458 1483 privileged instruction
1459 1484
1460 1485
1461 1486 FLTBPT
1462 1487 breakpoint trap
1463 1488
1464 1489
1465 1490 FLTTRACE
1466 1491 trace trap (single-step)
1467 1492
1468 1493
1469 1494 FLTWATCH
1470 1495 watchpoint trap
1471 1496
1472 1497
1473 1498 FLTACCESS
1474 1499 memory access fault (bus error)
1475 1500
1476 1501
1477 1502 FLTBOUNDS
1478 1503 memory bounds violation
1479 1504
1480 1505
1481 1506 FLTIOVF
1482 1507 integer overflow
1483 1508
1484 1509
1485 1510 FLTIZDIV
1486 1511 integer zero divide
1487 1512
1488 1513
1489 1514 FLTFPE
1490 1515 floating-point exception
1491 1516
1492 1517
1493 1518 FLTSTACK
1494 1519 unrecoverable stack fault
1495 1520
1496 1521
1497 1522 FLTPAGE
1498 1523 recoverable page fault
1499 1524
1500 1525
1501 1526
1502 1527 When not traced, a fault normally results in the posting of a signal to
1503 1528 the lwp that incurred the fault. If an lwp stops on a fault, the signal
1504 1529 is posted to the lwp when execution is resumed unless the fault is
1505 1530 cleared by PCCFAULT or by the PRCFAULT option of PCRUN. FLTPAGE is an
1506 1531 exception; no signal is posted. The pr_info field in the lwpstatus
1507 1532 structure identifies the signal to be sent and contains machine-
1508 1533 specific information about the fault.
1509 1534
1510 1535 PCCFAULT
1511 1536 The current fault, if any, is cleared; the associated signal will not
1512 1537 be sent to the specific or representative lwp.
1513 1538
1514 1539 PCSENTRY PCSEXIT
1515 1540 These control operations instruct the process's lwps to stop on entry
1516 1541 to or exit from specified system calls. The set of system calls to be
1517 1542 traced is defined via an operand sysset_t structure.
1518 1543
1519 1544
1520 1545 When entry to a system call is being traced, an lwp stops after having
1521 1546 begun the call to the system but before the system call arguments have
1522 1547 been fetched from the lwp. When exit from a system call is being
1523 1548 traced, an lwp stops on completion of the system call just prior to
1524 1549 checking for signals and returning to user level. At this point, all
1525 1550 return values have been stored into the lwp's registers.
1526 1551
1527 1552
1528 1553 If an lwp is stopped on entry to a system call (PR_SYSENTRY) or when
1529 1554 sleeping in an interruptible system call (PR_ASLEEP is set), it may be
1530 1555 instructed to go directly to system call exit by specifying the
1531 1556 PRSABORT flag in a PCRUN control message. Unless exit from the system
1532 1557 call is being traced, the lwp returns to user level showing EINTR.
1533 1558
1534 1559 PCWATCH
1535 1560 Set or clear a watched area in the controlled process from a prwatch
1536 1561 structure operand:
1537 1562
1538 1563 typedef struct prwatch {
1539 1564 uintptr_t pr_vaddr; /* virtual address of watched area */
1540 1565 size_t pr_size; /* size of watched area in bytes */
1541 1566 int pr_wflags; /* watch type flags */
1542 1567 } prwatch_t;
1543 1568
1544 1569
1545 1570
1546 1571 pr_vaddr specifies the virtual address of an area of memory to be
1547 1572 watched in the controlled process. pr_size specifies the size of the
1548 1573 area, in bytes. pr_wflags specifies the type of memory access to be
1549 1574 monitored as a bit-mask of the following flags:
1550 1575
1551 1576 WA_READ
1552 1577 read access
1553 1578
1554 1579
1555 1580 WA_WRITE
1556 1581 write access
1557 1582
1558 1583
1559 1584 WA_EXEC
1560 1585 execution access
1561 1586
1562 1587
1563 1588 WA_TRAPAFTER
1564 1589 trap after the instruction completes
1565 1590
1566 1591
1567 1592
1568 1593 If pr_wflags is non-empty, a watched area is established for the
1569 1594 virtual address range specified by pr_vaddr and pr_size. If pr_wflags
1570 1595 is empty, any previously-established watched area starting at the
1571 1596 specified virtual address is cleared; pr_size is ignored.
1572 1597
1573 1598
1574 1599 A watchpoint is triggered when an lwp in the traced process makes a
1575 1600 memory reference that covers at least one byte of a watched area and
1576 1601 the memory reference is as specified in pr_wflags. When an lwp triggers
1577 1602 a watchpoint, it incurs a watchpoint trap. If FLTWATCH is being traced,
1578 1603 the lwp stops; otherwise, it is sent a SIGTRAP signal; if SIGTRAP is
1579 1604 being traced and is not blocked, the lwp stops.
1580 1605
1581 1606
1582 1607 The watchpoint trap occurs before the instruction completes unless
1583 1608 WA_TRAPAFTER was specified, in which case it occurs after the
1584 1609 instruction completes. If it occurs before completion, the memory is
1585 1610 not modified. If it occurs after completion, the memory is modified (if
1586 1611 the access is a write access).
1587 1612
1588 1613
1589 1614 Physical i/o is an exception for watchpoint traps. In this instance,
1590 1615 there is no guarantee that memory before the watched area has already
1591 1616 been modified (or in the case of WA_TRAPAFTER, that the memory
1592 1617 following the watched area has not been modified) when the watchpoint
1593 1618 trap occurs and the lwp stops.
1594 1619
1595 1620
1596 1621 pr_info in the lwpstatus structure contains information pertinent to
1597 1622 the watchpoint trap. In particular, the si_addr field contains the
1598 1623 virtual address of the memory reference that triggered the watchpoint,
1599 1624 and the si_code field contains one of TRAP_RWATCH, TRAP_WWATCH, or
1600 1625 TRAP_XWATCH, indicating read, write, or execute access, respectively.
1601 1626 The si_trapafter field is zero unless WA_TRAPAFTER is in effect for
1602 1627 this watched area; non-zero indicates that the current instruction is
1603 1628 not the instruction that incurred the watchpoint trap. The si_pc field
1604 1629 contains the virtual address of the instruction that incurred the trap.
1605 1630
1606 1631
1607 1632 A watchpoint trap may be triggered while executing a system call that
1608 1633 makes reference to the traced process's memory. The lwp that is
1609 1634 executing the system call incurs the watchpoint trap while still in the
1610 1635 system call. If it stops as a result, the lwpstatus structure contains
1611 1636 the system call number and its arguments. If the lwp does not stop, or
1612 1637 if it is set running again without clearing the signal or fault, the
1613 1638 system call fails with EFAULT. If WA_TRAPAFTER was specified, the
1614 1639 memory reference will have completed and the memory will have been
1615 1640 modified (if the access was a write access) when the watchpoint trap
1616 1641 occurs.
1617 1642
1618 1643
1619 1644 If more than one of WA_READ, WA_WRITE, and WA_EXEC is specified for a
1620 1645 watched area, and a single instruction incurs more than one of the
1621 1646 specified types, only one is reported when the watchpoint trap occurs.
1622 1647 The precedence is WA_EXEC, WA_READ, WA_WRITE (WA_EXEC and WA_READ take
1623 1648 precedence over WA_WRITE), unless WA_TRAPAFTER was specified, in which
1624 1649 case it is WA_WRITE, WA_READ, WA_EXEC (WA_WRITE takes precedence).
1625 1650
1626 1651
1627 1652 PCWATCH fails with EINVAL if an attempt is made to specify overlapping
1628 1653 watched areas or if pr_wflags contains flags other than those specified
1629 1654 above. It fails with ENOMEM if an attempt is made to establish more
1630 1655 watched areas than the system can support (the system can support
1631 1656 thousands).
1632 1657
1633 1658
1634 1659 The child of a vfork(2) borrows the parent's address space. When a
1635 1660 vfork(2) is executed by a traced process, all watched areas established
1636 1661 for the parent are suspended until the child terminates or performs an
1637 1662 exec(2). Any watched areas established independently in the child are
1638 1663 cancelled when the parent resumes after the child's termination or
1639 1664 exec(2). PCWATCH fails with EBUSY if applied to the parent of a
1640 1665 vfork(2) before the child has terminated or performed an exec(2). The
1641 1666 PR_VFORKP flag is set in the pstatus structure for such a parent
1642 1667 process.
1643 1668
1644 1669
1645 1670 Certain accesses of the traced process's address space by the operating
1646 1671 system are immune to watchpoints. The initial construction of a signal
1647 1672 stack frame when a signal is delivered to an lwp will not trigger a
1648 1673 watchpoint trap even if the new frame covers watched areas of the
1649 1674 stack. Once the signal handler is entered, watchpoint traps occur
1650 1675 normally. On SPARC based machines, register window overflow and
1651 1676 underflow will not trigger watchpoint traps, even if the register
1652 1677 window save areas cover watched areas of the stack.
1653 1678
1654 1679
1655 1680 Watched areas are not inherited by child processes, even if the traced
1656 1681 process's inherit-on-fork mode, PR_FORK, is set (see PCSET, below).
1657 1682 All watched areas are cancelled when the traced process performs a
1658 1683 successful exec(2).
1659 1684
1660 1685 PCSET PCUNSET
1661 1686 PCSET sets one or more modes of operation for the traced process.
1662 1687 PCUNSET unsets these modes. The modes to be set or unset are specified
1663 1688 by flags in an operand long in the control message:
1664 1689
1665 1690 PR_FORK
1666 1691 (inherit-on-fork): When set, the process's tracing flags
1667 1692 and its inherit-on-fork mode are inherited by the child of
1668 1693 a fork(2), fork1(2), or vfork(2). When unset, child
1669 1694 processes start with all tracing flags cleared.
1670 1695
1671 1696
1672 1697 PR_RLC
1673 1698 (run-on-last-close): When set and the last writable /proc
1674 1699 file descriptor referring to the traced process or any of
1675 1700 its lwps is closed, all of the process's tracing flags and
1676 1701 watched areas are cleared, any outstanding stop directives
1677 1702 are canceled, and if any lwps are stopped on events of
1678 1703 interest, they are set running as though PCRUN had been
1679 1704 applied to them. When unset, the process's tracing flags
1680 1705 and watched areas are retained and lwps are not set
1681 1706 running on last close.
1682 1707
1683 1708
1684 1709 PR_KLC
1685 1710 (kill-on-last-close): When set and the last writable /proc
1686 1711 file descriptor referring to the traced process or any of
1687 1712 its lwps is closed, the process is terminated with
1688 1713 SIGKILL.
1689 1714
1690 1715
1691 1716 PR_ASYNC
1692 1717 (asynchronous-stop): When set, a stop on an event of
1693 1718 interest by one lwp does not directly affect any other lwp
1694 1719 in the process. When unset and an lwp stops on an event of
1695 1720 interest other than PR_REQUESTED, all other lwps in the
1696 1721 process are directed to stop.
1697 1722
1698 1723
1699 1724 PR_MSACCT
1700 1725 (microstate accounting): Microstate accounting is now
1701 1726 continuously enabled. This flag is deprecated and no
1702 1727 longer has any effect upon microstate accounting.
1703 1728 Applications may toggle this flag; however, microstate
1704 1729 accounting will remain enabled regardless.
1705 1730
1706 1731
1707 1732 PR_MSFORK
1708 1733 (inherit microstate accounting): All processes now inherit
1709 1734 microstate accounting, as it is continuously enabled. This
1710 1735 flag has been deprecated and its use no longer has any
1711 1736 effect upon the behavior of microstate accounting.
1712 1737
1713 1738
1714 1739 PR_BPTADJ
1715 1740 (breakpoint trap pc adjustment): On x86-based machines, a
1716 1741 breakpoint trap leaves the program counter (the EIP)
1717 1742 referring to the breakpointed instruction plus one byte.
1718 1743 When PR_BPTADJ is set, the system will adjust the program
1719 1744 counter back to the location of the breakpointed
1720 1745 instruction when the lwp stops on a breakpoint. This flag
1721 1746 has no effect on SPARC based machines, where breakpoint
1722 1747 traps leave the program counter referring to the
1723 1748 breakpointed instruction.
1724 1749
1725 1750
1726 1751 PR_PTRACE
1727 1752 (ptrace-compatibility): When set, a stop on an event of
1728 1753 interest by the traced process is reported to the parent
1729 1754 of the traced process by wait(3C), SIGTRAP is sent to the
1730 1755 traced process when it executes a successful exec(2),
1731 1756 setuid/setgid flags are not honored for execs performed by
1732 1757 the traced process, any exec of an object file that the
1733 1758 traced process cannot read fails, and the process dies
1734 1759 when its parent dies. This mode is deprecated; it is
1735 1760 provided only to allow ptrace(3C) to be implemented as a
1736 1761 library function using /proc.
1737 1762
1738 1763
1739 1764
1740 1765 It is an error (EINVAL) to specify flags other than those described
1741 1766 above or to apply these operations to a system process. The current
1742 1767 modes are reported in the pr_flags field of /proc/pid/status and
1743 1768 /proc/pid/lwp/lwp/lwpstatus.
1744 1769
1745 1770 PCSREG
1746 1771 Set the general registers for the specific or representative lwp
1747 1772 according to the operand prgregset_t structure.
1748 1773
1749 1774
1750 1775 On SPARC based systems, only the condition-code bits of the processor-
1751 1776 status register (R_PSR) of SPARC V8 (32-bit) processes can be modified
1752 1777 by PCSREG. Other privileged registers cannot be modified at all.
1753 1778
1754 1779
1755 1780 On x86-based systems, only certain bits of the flags register (EFL) can
1756 1781 be modified by PCSREG: these include the condition codes, direction-
1757 1782 bit, and overflow-bit.
1758 1783
1759 1784
1760 1785 PCSREG fails with EBUSY if the lwp is not stopped on an event of
1761 1786 interest.
1762 1787
1763 1788 PCSVADDR
1764 1789 Set the address at which execution will resume for the specific or
1765 1790 representative lwp from the operand long. On SPARC based systems, both
1766 1791 %pc and %npc are set, with %npc set to the instruction following the
1767 1792 virtual address. On x86-based systems, only %eip is set. PCSVADDR fails
1768 1793 with EBUSY if the lwp is not stopped on an event of interest.
1769 1794
1770 1795 PCSFPREG
1771 1796 Set the floating-point registers for the specific or representative lwp
1772 1797 according to the operand prfpregset_t structure. An error (EINVAL) is
1773 1798 returned if the system does not support floating-point operations (no
1774 1799 floating-point hardware and the system does not emulate floating-point
1775 1800 machine instructions). PCSFPREG fails with EBUSY if the lwp is not
1776 1801 stopped on an event of interest.
1777 1802
1778 1803 PCSXREG
1779 1804 Set the extra state registers for the specific or representative lwp
1780 1805 according to the architecture-dependent operand prxregset_t structure.
1781 1806 An error (EINVAL) is returned if the system does not support extra
1782 1807 state registers. PCSXREG fails with EBUSY if the lwp is not stopped on
1783 1808 an event of interest.
1784 1809
1785 1810 PCSASRS
1786 1811 Set the ancillary state registers for the specific or representative
1787 1812 lwp according to the SPARC V9 platform-dependent operand asrset_t
1788 1813 structure. An error (EINVAL) is returned if either the target process
1789 1814 or the controlling process is not a 64-bit SPARC V9 process. Most of
1790 1815 the ancillary state registers are privileged registers that cannot be
1791 1816 modified. Only those that can be modified are set; all others are
1792 1817 silently ignored. PCSASRS fails with EBUSY if the lwp is not stopped on
1793 1818 an event of interest.
1794 1819
1795 1820 PCAGENT
1796 1821 Create an agent lwp in the controlled process with register values from
1797 1822 the operand prgregset_t structure (see PCSREG, above). The agent lwp is
1798 1823 created in the stopped state showing PR_REQUESTED and with its held
1799 1824 signal set (the signal mask) having all signals except SIGKILL and
1800 1825 SIGSTOP blocked.
1801 1826
1802 1827
1803 1828 The PCAGENT operation fails with EBUSY unless the process is fully
1804 1829 stopped via /proc, that is, unless all of the lwps in the process are
1805 1830 stopped either on events of interest or on PR_SUSPENDED, or are stopped
1806 1831 on PR_JOBCONTROL and have been directed to stop via PCDSTOP. It fails
1807 1832 with EBUSY if an agent lwp already exists. It fails with ENOMEM if
1808 1833 system resources for creating new lwps have been exhausted.
1809 1834
1810 1835
1811 1836 Any PCRUN operation applied to the process control file or to the
1812 1837 control file of an lwp other than the agent lwp fails with EBUSY as
1813 1838 long as the agent lwp exists. The agent lwp must be caused to terminate
1814 1839 by executing the SYS_lwp_exit system call trap before the process can
1815 1840 be restarted.
1816 1841
1817 1842
1818 1843 Once the agent lwp is created, its lwp-ID can be found by reading the
1819 1844 process status file. To facilitate opening the agent lwp's control and
1820 1845 status files, the directory name /propc/pid/lwp/agent is accepted for
1821 1846 lookup operations as an invisible alias for /proc/pid/lwp/lwpid, lwpid
1822 1847 being the lwp-ID of the agent lwp (invisible in the sense that the name
1823 1848 ``agent'' does not appear in a directory listing of /proc/pid/lwp
1824 1849 obtained from ls(1), getdents(2), or readdir(3C)).
1825 1850
1826 1851
1827 1852 The purpose of the agent lwp is to perform operations in the controlled
1828 1853 process on behalf of the controlling process: to gather information not
1829 1854 directly available via /proc files, or in general to make the process
1830 1855 change state in ways not directly available via /proc control
1831 1856 operations. To make use of an agent lwp, the controlling process must
1832 1857 be capable of making it execute system calls (specifically, the
1833 1858 SYS_lwp_exit system call trap). The register values given to the agent
1834 1859 lwp on creation are typically the registers of the representative lwp,
1835 1860 so that the agent lwp can use its stack.
1836 1861
1837 1862
1838 1863 If the controlling process neglects to force the agent lwp to execute
1839 1864 the SYS_lwp_exit system call (due to either logic error or fatal
1840 1865 failure on the part of the controlling process), the agent lwp will
1841 1866 remain in the target process. For purposes of being able to debug
1842 1867 these otherwise rogue agents, information as to the creator of the
1843 1868 agent lwp is reflected in that lwp's spymaster file in /proc. Should
1844 1869 the target process generate a core dump with the agent lwp in place,
1845 1870 this information will be available via the NT_SPYMASTER note in the
1846 1871 core file (see core(4)).
1847 1872
1848 1873
1849 1874 The agent lwp is not allowed to execute any variation of the SYS_fork
1850 1875 or SYS_exec system call traps. Attempts to do so yield ENOTSUP to the
1851 1876 agent lwp.
1852 1877
1853 1878
1854 1879 Symbolic constants for system call trap numbers like SYS_lwp_exit and
1855 1880 SYS_lwp_create can be found in the header file <sys/syscall.h>.
1856 1881
1857 1882 PCREAD PCWRITE
1858 1883 Read or write the target process's address space via a priovec
1859 1884 structure operand:
1860 1885
1861 1886 typedef struct priovec {
1862 1887 void *pio_base; /* buffer in controlling process */
1863 1888 size_t pio_len; /* size of read/write request in bytes */
1864 1889 off_t pio_offset; /* virtual address in target process */
1865 1890 } priovec_t;
1866 1891
1867 1892
1868 1893
1869 1894 These operations have the same effect as pread(2) and pwrite(2),
1870 1895 respectively, of the target process's address space file. The
1871 1896 difference is that more than one PCREAD or PCWRITE control operation
1872 1897 can be written to the control file at once, and they can be
1873 1898 interspersed with other control operations in a single write to the
1874 1899 control file. This is useful, for example, when planting many
1875 1900 breakpoint instructions in the process's address space, or when
1876 1901 stepping over a breakpointed instruction. Unlike pread(2) and
1877 1902 pwrite(2), no provision is made for partial reads or writes; if the
1878 1903 operation cannot be performed completely, it fails with EIO.
1879 1904
1880 1905 PCNICE
1881 1906 The traced process's nice(2) value is incremented by the amount in the
1882 1907 operand long. Only a process with the {PRIV_PROC_PRIOCNTL} privilege
1883 1908 asserted in its effective set can better a process's priority in this
1884 1909 way, but any user may lower the priority. This operation is not
1885 1910 meaningful for all scheduling classes.
1886 1911
1887 1912 PCSCRED
1888 1913 Set the target process credentials to the values contained in the
1889 1914 prcred_t structure operand (see /proc/pid/cred). The effective, real,
1890 1915 and saved user-IDs and group-IDs of the target process are set. The
1891 1916 target process's supplementary groups are not changed; the pr_ngroups
1892 1917 and pr_groups members of the structure operand are ignored. Only the
1893 1918 privileged processes can perform this operation; for all others it
1894 1919 fails with EPERM.
1895 1920
1896 1921 PCSCREDX
1897 1922 Operates like PCSCRED but also sets the supplementary groups; the
1898 1923 length of the data written with this control operation should be
1899 1924 "sizeof (prcred_t) + sizeof (gid_t) * (#groups - 1)".
1900 1925
1901 1926 PCSPRIV
1902 1927 Set the target process privilege to the values contained in the
1903 1928 prpriv_t operand (see /proc/pid/priv). The effective, permitted,
1904 1929 inheritable, and limit sets are all changed. Privilege flags can also
1905 1930 be set. The process is made privilege aware unless it can relinquish
1906 1931 privilege awareness. See privileges(5).
1907 1932
1908 1933
1909 1934 The limit set of the target process cannot be grown. The other
1910 1935 privilege sets must be subsets of the intersection of the effective set
1911 1936 of the calling process with the new limit set of the target process or
1912 1937 subsets of the original values of the sets in the target process.
1913 1938
1914 1939
1915 1940 If any of the above restrictions are not met, EPERM is returned. If the
1916 1941 structure written is improperly formatted, EINVAL is returned.
1917 1942
1918 1943 PROGRAMMING NOTES
1919 1944 For security reasons, except for the psinfo, usage, lpsinfo, lusage,
1920 1945 lwpsinfo, and lwpusage files, which are world-readable, and except for
1921 1946 privileged processes, an open of a /proc file fails unless both the
1922 1947 user-ID and group-ID of the caller match those of the traced process
1923 1948 and the process's object file is readable by the caller. The effective
1924 1949 set of the caller is a superset of both the inheritable and the
1925 1950 permitted set of the target process. The limit set of the caller is a
1926 1951 superset of the limit set of the target process. Except for the world-
1927 1952 readable files just mentioned, files corresponding to setuid and setgid
1928 1953 processes can be opened only by the appropriately privileged process.
1929 1954
1930 1955
1931 1956 A process that is missing the basic privilege {PRIV_PROC_INFO} cannot
1932 1957 see any processes under /proc that it cannot send a signal to.
1933 1958
1934 1959
1935 1960 A process that has {PRIV_PROC_OWNER} asserted in its effective set can
1936 1961 open any file for reading. To manipulate or control a process, the
1937 1962 controlling process must have at least as many privileges in its
1938 1963 effective set as the target process has in its effective, inheritable,
1939 1964 and permitted sets. The limit set of the controlling process must be a
1940 1965 superset of the limit set of the target process. Additional
1941 1966 restrictions apply if any of the uids of the target process are 0. See
1942 1967 privileges(5).
1943 1968
1944 1969
1945 1970 Even if held by a privileged process, an open process or lwp file
1946 1971 descriptor (other than file descriptors for the world-readable files)
1947 1972 becomes invalid if the traced process performs an exec(2) of a
1948 1973 setuid/setgid object file or an object file that the traced process
1949 1974 cannot read. Any operation performed on an invalid file descriptor,
1950 1975 except close(2), fails with EAGAIN. In this situation, if any tracing
1951 1976 flags are set and the process or any lwp file descriptor is open for
1952 1977 writing, the process will have been directed to stop and its run-on-
1953 1978 last-close flag will have been set (see PCSET). This enables a
1954 1979 controlling process (if it has permission) to reopen the /proc files to
1955 1980 get new valid file descriptors, close the invalid file descriptors,
1956 1981 unset the run-on-last-close flag (if desired), and proceed. Just
1957 1982 closing the invalid file descriptors causes the traced process to
1958 1983 resume execution with all tracing flags cleared. Any process not
1959 1984 currently open for writing via /proc, but that has left-over tracing
1960 1985 flags from a previous open, and that executes a setuid/setgid or
1961 1986 unreadable object file, will not be stopped but will have all its
1962 1987 tracing flags cleared.
1963 1988
1964 1989
1965 1990 To wait for one or more of a set of processes or lwps to stop or
1966 1991 terminate, /proc file descriptors (other than those obtained by opening
1967 1992 the cwd or root directories or by opening files in the fd or object
1968 1993 directories) can be used in a poll(2) system call. When requested and
1969 1994 returned, either of the polling events POLLPRI or POLLWRNORM indicates
1970 1995 that the process or lwp stopped on an event of interest. Although they
1971 1996 cannot be requested, the polling events POLLHUP, POLLERR, and POLLNVAL
1972 1997 may be returned. POLLHUP indicates that the process or lwp has
1973 1998 terminated. POLLERR indicates that the file descriptor has become
1974 1999 invalid. POLLNVAL is returned immediately if POLLPRI or POLLWRNORM is
1975 2000 requested on a file descriptor referring to a system process (see
1976 2001 PCSTOP). The requested events may be empty to wait simply for
1977 2002 termination.
1978 2003
1979 2004 FILES
1980 2005 /proc
1981 2006
1982 2007 directory (list of processes)
1983 2008
1984 2009
1985 2010 /proc/pid
1986 2011
1987 2012 specific process directory
1988 2013
1989 2014
1990 2015 /proc/self
1991 2016
1992 2017 alias for a process's own directory
1993 2018
1994 2019
1995 2020 /proc/pid/as
1996 2021
1997 2022 address space file
1998 2023
1999 2024
2000 2025 /proc/pid/ctl
2001 2026
2002 2027 process control file
2003 2028
2004 2029
2005 2030 /proc/pid/status
2006 2031
2007 2032 process status
2008 2033
2009 2034
2010 2035 /proc/pid/lstatus
2011 2036
2012 2037 array of lwp status structs
2013 2038
2014 2039
2015 2040 /proc/pid/psinfo
2016 2041
2017 2042 process ps(1) info
2018 2043
2019 2044
2020 2045 /proc/pid/lpsinfo
2021 2046
2022 2047 array of lwp ps(1) info structs
2023 2048
2024 2049
2025 2050 /proc/pid/map
2026 2051
2027 2052 address space map
2028 2053
2029 2054
2030 2055 /proc/pid/xmap
2031 2056
2032 2057 extended address space map
2033 2058
2034 2059
2035 2060 /proc/pid/rmap
2036 2061
2037 2062 reserved address map
2038 2063
2039 2064
2040 2065 /proc/pid/cred
2041 2066
2042 2067 process credentials
2043 2068
2044 2069
2045 2070 /proc/pid/priv
2046 2071
2047 2072 process privileges
2048 2073
2049 2074
|
↓ open down ↓ |
1184 lines elided |
↑ open up ↑ |
2050 2075 /proc/pid/sigact
2051 2076
2052 2077 process signal actions
2053 2078
2054 2079
2055 2080 /proc/pid/auxv
2056 2081
2057 2082 process aux vector
2058 2083
2059 2084
2085 + /proc/pid/argv
2086 +
2087 + process argument vector
2088 +
2089 +
2060 2090 /proc/pid/ldt
2061 2091
2062 2092 process LDT (x86 only)
2063 2093
2064 2094
2065 2095 /proc/pid/usage
2066 2096
2067 2097 process usage
2068 2098
2069 2099
2070 2100 /proc/pid/lusage
2071 2101
2072 2102 array of lwp usage structs
2073 2103
2074 2104
2075 2105 /proc/pid/path
2076 2106
2077 2107 symbolic links to process open files
2078 2108
2079 2109
2080 2110 /proc/pid/pagedata
2081 2111
2082 2112 process page data
2083 2113
2084 2114
2085 2115 /proc/pid/watch
2086 2116
2087 2117 active watchpoints
2088 2118
2089 2119
2090 2120 /proc/pid/cwd
2091 2121
2092 2122 alias for the current working directory
2093 2123
2094 2124
2095 2125 /proc/pid/root
2096 2126
2097 2127 alias for the root directory
2098 2128
2099 2129
2100 2130 /proc/pid/fd
2101 2131
2102 2132 directory (list of open files)
2103 2133
2104 2134
2105 2135 /proc/pid/fd/*
2106 2136
2107 2137 aliases for process's open files
2108 2138
2109 2139
2110 2140 /proc/pid/object
2111 2141
2112 2142 directory (list of mapped files)
2113 2143
2114 2144
2115 2145 /proc/pid/object/a.out
2116 2146
2117 2147 alias for process's executable file
2118 2148
2119 2149
2120 2150 /proc/pid/object/*
2121 2151
2122 2152 aliases for other mapped files
2123 2153
2124 2154
2125 2155 /proc/pid/lwp
2126 2156
2127 2157 directory (list of lwps)
2128 2158
2129 2159
2130 2160 /proc/pid/lwp/lwpid
2131 2161
2132 2162 specific lwp directory
2133 2163
2134 2164
2135 2165 /proc/pid/lwp/agent
2136 2166
2137 2167 alias for the agent lwp directory
2138 2168
2139 2169
2140 2170 /proc/pid/lwp/lwpid/lwpctl
2141 2171
2142 2172 lwp control file
2143 2173
2144 2174
2145 2175 /proc/pid/lwp/lwpid/lwpstatus
2146 2176
2147 2177 lwp status
2148 2178
2149 2179
2150 2180 /proc/pid/lwp/lwpid/lwpsinfo
2151 2181
2152 2182 lwp ps(1) info
2153 2183
2154 2184
2155 2185 /proc/pid/lwp/lwpid/lwpusage
2156 2186
2157 2187 lwp usage
2158 2188
2159 2189
2160 2190 /proc/pid/lwp/lwpid/gwindows
2161 2191
2162 2192 register windows (SPARC only)
2163 2193
2164 2194
2165 2195 /proc/pid/lwp/lwpid/xregs
2166 2196
2167 2197 extra state registers
2168 2198
2169 2199
2170 2200 /proc/pid/lwp/lwpid/asrs
2171 2201
2172 2202 ancillary state registers (SPARC V9 only)
2173 2203
2174 2204
2175 2205 /proc/pid/lwp/lwpid/spymaster
2176 2206
2177 2207 For an agent LWP, the controlling process
2178 2208
2179 2209
2180 2210 SEE ALSO
2181 2211 ls(1), ps(1), chroot(1M), alarm(2), brk(2), chdir(2), chroot(2),
2182 2212 close(2), creat(2), dup(2), exec(2), fcntl(2), fork(2), fork1(2),
2183 2213 fstat(2), getdents(2), getustack(2), kill(2), lseek(2), mmap(2),
2184 2214 nice(2), open(2), poll(2), pread(2), ptrace(3C), pwrite(2), read(2),
2185 2215 readlink(2), readv(2), shmget(2), sigaction(2), sigaltstack(2),
2186 2216 vfork(2), write(2), writev(2), _stack_grow(3C), readdir(3C),
2187 2217 pthread_create(3C), pthread_join(3C), siginfo.h(3HEAD),
2188 2218 signal.h(3HEAD), thr_create(3C), thr_join(3C), types32.h(3HEAD),
2189 2219 ucontext.h(3HEAD), wait(3C), contract(4), core(4), process(4),
2190 2220 lfcompile(5), privileges(5)
2191 2221
2192 2222 DIAGNOSTICS
2193 2223 Errors that can occur in addition to the errors normally associated
2194 2224 with file system access:
2195 2225
2196 2226 E2BIG
2197 2227 Data to be returned in a read(2) of the page data file
2198 2228 exceeds the size of the read buffer provided by the
2199 2229 caller.
2200 2230
2201 2231
2202 2232 EACCES
2203 2233 An attempt was made to examine a process that ran under a
2204 2234 different uid than the controlling process and
2205 2235 {PRIV_PROC_OWNER} was not asserted in the effective set.
2206 2236
2207 2237
2208 2238 EAGAIN
2209 2239 The traced process has performed an exec(2) of a
2210 2240 setuid/setgid object file or of an object file that it
2211 2241 cannot read; all further operations on the process or lwp
2212 2242 file descriptor (except close(2)) elicit this error.
2213 2243
2214 2244
2215 2245 EBUSY
2216 2246 PCSTOP, PCDSTOP, PCWSTOP, or PCTWSTOP was applied to a
2217 2247 system process; an exclusive open(2) was attempted on a
2218 2248 /proc file for a process already open for writing; PCRUN,
2219 2249 PCSREG, PCSVADDR, PCSFPREG, or PCSXREG was applied to a
2220 2250 process or lwp not stopped on an event of interest; an
2221 2251 attempt was made to mount /proc when it was already
2222 2252 mounted; PCAGENT was applied to a process that was not
2223 2253 fully stopped or that already had an agent lwp.
2224 2254
2225 2255
2226 2256 EINVAL
2227 2257 In general, this means that some invalid argument was
2228 2258 supplied to a system call. A non-exhaustive list of
2229 2259 conditions eliciting this error includes: a control
2230 2260 message operation code is undefined; an out-of-range
2231 2261 signal number was specified with PCSSIG, PCKILL, or
2232 2262 PCUNKILL; SIGKILL was specified with PCUNKILL; PCSFPREG
2233 2263 was applied on a system that does not support floating-
2234 2264 point operations; PCSXREG was applied on a system that
2235 2265 does not support extra state registers.
2236 2266
2237 2267
2238 2268 EINTR
2239 2269 A signal was received by the controlling process while
2240 2270 waiting for the traced process or lwp to stop via PCSTOP,
2241 2271 PCWSTOP, or PCTWSTOP.
2242 2272
2243 2273
2244 2274 EIO
2245 2275 A write(2) was attempted at an illegal address in the
2246 2276 traced process.
2247 2277
2248 2278
2249 2279 ENOENT
2250 2280 The traced process or lwp has terminated after being
2251 2281 opened. The basic privilege {PRIV_PROC_INFO} is not
2252 2282 asserted in the effective set of the calling process and
2253 2283 the calling process cannot send a signal to the target
2254 2284 process.
2255 2285
2256 2286
2257 2287 ENOMEM
2258 2288 The system-imposed limit on the number of page data file
2259 2289 descriptors was reached on an open of /proc/pid/pagedata;
2260 2290 an attempt was made with PCWATCH to establish more watched
2261 2291 areas than the system can support; the PCAGENT operation
2262 2292 was issued when the system was out of resources for
2263 2293 creating lwps.
2264 2294
2265 2295
2266 2296 ENOSYS
2267 2297 An attempt was made to perform an unsupported operation
2268 2298 (such as creat(2), link(2), or unlink(2)) on an entry in
2269 2299 /proc.
2270 2300
2271 2301
2272 2302 EOVERFLOW
2273 2303 A 32-bit controlling process attempted to read or write
2274 2304 the as file or attempted to read the map, rmap, or
2275 2305 pagedata file of a 64-bit target process. A 32-bit
2276 2306 controlling process attempted to apply one of the control
2277 2307 operations PCSREG, PCSXREG, PCSVADDR, PCWATCH, PCAGENT,
2278 2308 PCREAD, PCWRITE to a 64-bit target process.
2279 2309
2280 2310
2281 2311 EPERM
2282 2312 The process that issued the PCSCRED or PCSCREDX operation
2283 2313 did not have the {PRIV_PROC_SETID} privilege asserted in
2284 2314 its effective set, or the process that issued the PCNICE
2285 2315 operation did not have the {PRIV_PROC_PRIOCNTL} in its
2286 2316 effective set.
2287 2317
2288 2318 An attempt was made to control a process of which the E,
2289 2319 P, and I privilege sets were not a subset of the effective
2290 2320 set of the controlling process or the limit set of the
2291 2321 controlling process is not a superset of limit set of the
2292 2322 controlled process.
2293 2323
2294 2324 Any of the uids of the target process are 0 or an attempt
2295 2325 was made to change any of the uids to 0 using PCSCRED and
2296 2326 the security policy imposed additional restrictions. See
2297 2327 privileges(5).
2298 2328
2299 2329
2300 2330 NOTES
2301 2331 Descriptions of structures in this document include only interesting
2302 2332 structure elements, not filler and padding fields, and may show
2303 2333 elements out of order for descriptive clarity. The actual structure
2304 2334 definitions are contained in <procfs.h>.
2305 2335
2306 2336 BUGS
2307 2337 Because the old ioctl(2)-based version of /proc is currently supported
2308 2338 for binary compatibility with old applications, the top-level directory
2309 2339 for a process, /proc/pid, is not world-readable, but it is world-
2310 2340 searchable. Thus, anyone can open /proc/pid/psinfo even though ls(1)
2311 2341 applied to /proc/pid will fail for anyone but the owner or an
2312 2342 appropriately privileged process. Support for the old ioctl(2)-based
|
↓ open down ↓ |
243 lines elided |
↑ open up ↑ |
2313 2343 version of /proc will be dropped in a future release, at which time the
2314 2344 top-level directory for a process will be made world-readable.
2315 2345
2316 2346
2317 2347 On SPARC based machines, the types gregset_t and fpregset_t defined in
2318 2348 <sys/regset.h> are similar to but not the same as the types prgregset_t
2319 2349 and prfpregset_t defined in <procfs.h>.
2320 2350
2321 2351
2322 2352
2323 - March 31, 2013 PROC(4)
2353 + May 19, 2014 PROC(4)
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX