1 PROC(4) File Formats and Configurations PROC(4)
2
3 NAME
4 proc - /proc, the process file system
5
6 DESCRIPTION
7 /proc is a file system that provides access to the state of each process
8 and light-weight process (lwp) in the system. The name of each entry in
9 the /proc directory is a decimal number corresponding to a process-ID.
10 These entries are themselves subdirectories. Access to process state is
11 provided by additional files contained within each subdirectory; the
12 hierarchy is described more completely below. In this document, "/proc
13 file" refers to a non-directory file within the hierarchy rooted at
14 /proc. The owner of each /proc file and subdirectory is determined by
15 the user-ID of the process.
16
17 /proc can be mounted on any mount point, in addition to the standard
18 /proc mount point, and can be mounted several places at once. Such
19 additional mounts are allowed in order to facilitate the confinement of
20 processes to subtrees of the file system via chroot(2) and yet allow such
21 processes access to commands like ps(1).
22
23 Standard system calls are used to access /proc files: open(2), close(2),
24 read(2), and write(2) (including readv(2), writev(2), pread(2), and
25 pwrite(2)). Most files describe process state and can only be opened for
26 reading. ctl and lwpctl (control) files permit manipulation of process
27 state and can only be opened for writing. as (address space) files
28 contain the image of the running process and can be opened for both
29 reading and writing. An open for writing allows process control; a read-
30 only open allows inspection but not control. In this document, we refer
31 to the process as open for reading or writing if any of its associated
32 /proc files is open for reading or writing.
33
34 In general, more than one process can open the same /proc file at the
35 same time. Exclusive open is an advisory mechanism provided to allow
36 controlling processes to avoid collisions with each other. A process can
37 obtain exclusive control of a target process, with respect to other
38 cooperating processes, if it successfully opens any /proc file in the
39 target process for writing (the as or ctl files, or the lwpctl file of
40 any lwp) while specifying O_EXCL in the open(2). Such an open will fail
41 if the target process is already open for writing (that is, if an as,
42 ctl, or lwpctl file is already open for writing). There can be any
43 number of concurrent read-only opens; O_EXCL is ignored on opens for
44 reading. It is recommended that the first open for writing by a
45 controlling process use the O_EXCL flag; multiple controlling processes
46 usually result in chaos.
47
48 If a process opens one of its own /proc files for writing, the open
49 succeeds regardless of O_EXCL and regardless of whether some other
50 process has the process open for writing. Self-opens do not count when
51 another process attempts an exclusive open. (A process cannot exclude a
52 debugger by opening itself for writing and the application of a debugger
53 cannot prevent a process from opening itself.) All self-opens for
54 writing are forced to be close-on-exec (see the F_SETFD operation of
55 fcntl(2)).
56
57 Data may be transferred from or to any locations in the address space of
58 the traced process by applying lseek(2) to position the as file at the
59 virtual address of interest followed by read(2) or write(2) (or by using
60 pread(2) or pwrite(2) for the combined operation). The address-map files
61 /proc/pid/map and /proc/pid/xmap can be read to determine the accessible
62 areas (mappings) of the address space. I/O transfers may span contiguous
63 mappings. An I/O request extending into an unmapped area is truncated at
64 the boundary. A write request beginning at an unmapped virtual address
65 fails with EIO; a read request beginning at an unmapped virtual address
66 returns zero (an end-of-file indication).
67
68 Information and control operations are provided through additional files.
69 <procfs.h> contains definitions of data structures and message formats
70 used with these files. Some of these definitions involve the use of sets
71 of flags. The set types sigset_t, fltset_t, and sysset_t correspond,
72 respectively, to signal, fault, and system call enumerations defined in
73 <sys/signal.h>, <sys/fault.h>, and <sys/syscall.h>. Each set type is
74 large enough to hold flags for its own enumeration. Although they are of
75 different sizes, they have a common structure and can be manipulated by
76 these macros:
77
78 prfillset(&set); /* turn on all flags in set */
79 premptyset(&set); /* turn off all flags in set */
80 praddset(&set, flag); /* turn on the specified flag */
81 prdelset(&set, flag); /* turn off the specified flag */
82 r = prismember(&set, flag); /* != 0 iff flag is turned on */
83
84 One of prfillset() or premptyset() must be used to initialize set before
85 it is used in any other operation. flag must be a member of the
86 enumeration corresponding to set.
87
88 Every process contains at least one light-weight process, or lwp. Each
89 lwp represents a flow of execution that is independently scheduled by the
90 operating system. All lwps in a process share its address space as well
91 as many other attributes. Through the use of lwpctl and ctl files as
92 described below, it is possible to affect individual lwps in a process or
93 to affect all of them at once, depending on the operation.
94
95 When the process has more than one lwp, a representative lwp is chosen by
96 the system for certain process status files and control operations. The
97 representative lwp is a stopped lwp only if all of the process's lwps are
98 stopped; is stopped on an event of interest only if all of the lwps are
99 so stopped (excluding PR_SUSPENDED lwps); is in a PR_REQUESTED stop only
100 if there are no other events of interest to be found; or, failing
101 everything else, is in a PR_SUSPENDED stop (implying that the process is
102 deadlocked). See the description of the status file for definitions of
103 stopped states. See the PCSTOP control operation for the definition of
104 "event of interest".
105
106 The representative lwp remains fixed (it will be chosen again on the next
107 operation) as long as all of the lwps are stopped on events of interest
108 or are in a PR_SUSPENDED stop and the PCRUN control operation is not
109 applied to any of them.
110
111 When applied to the process control file, every /proc control operation
112 that must act on an lwp uses the same algorithm to choose which lwp to
113 act upon. Together with synchronous stopping (see PCSET), this enables a
114 debugger to control a multiple-lwp process using only the process-level
115 status and control files if it so chooses. More fine-grained control can
116 be achieved using the lwp-specific files.
117
118 The system supports two process data models, the traditional 32-bit data
119 model in which ints, longs and pointers are all 32 bits wide (the ILP32
120 data model), and on some platforms the 64-bit data model in which longs
121 and pointers, but not ints, are 64 bits in width (the LP64 data model).
122 In the LP64 data model some system data types, notably size_t, off_t,
123 time_t and dev_t, grow from 32 bits to 64 bits as well.
124
125 The /proc interfaces described here are available to both 32-bit and
126 64-bit controlling processes. However, many operations attempted by a
127 32-bit controlling process on a 64-bit target process will fail with
128 EOVERFLOW because the address space range of a 32-bit process cannot
129 encompass a 64-bit process or because the data in some 64-bit system data
130 type cannot be compressed to fit into the corresponding 32-bit type
131 without loss of information. Operations that fail in this circumstance
132 include reading and writing the address space, reading the address-map
133 files, and setting the target process's registers. There is no
134 restriction on operations applied by a 64-bit process to either a 32-bit
135 or a 64-bit target processes.
136
137 The format of the contents of any /proc file depends on the data model of
138 the observer (the controlling process), not on the data model of the
139 target process. A 64-bit debugger does not have to translate the
140 information it reads from a /proc file for a 32-bit process from 32-bit
141 format to 64-bit format. However, it usually has to be aware of the data
142 model of the target process. The pr_dmodel field of the status files
143 indicates the target process's data model.
144
145 To help deal with system data structures that are read from 32-bit
146 processes, a 64-bit controlling program can be compiled with the C
147 preprocessor symbol _SYSCALL32 defined before system header files are
148 included. This makes explicit 32-bit fixed-width data structures (like
149 struct stat32) visible to the 64-bit program. See types32.h(3HEAD).
150
151 DIRECTORY STRUCTURE
152 At the top level, the directory /proc contains entries each of which
153 names an existing process in the system. These entries are themselves
154 directories. Except where otherwise noted, the files described below can
155 be opened for reading only. In addition, if a process becomes a zombie
156 (one that has exited but whose parent has not yet performed a wait(3C)
157 upon it), most of its associated /proc files disappear from the
158 hierarchy; subsequent attempts to open them, or to read or write files
159 opened before the process exited, will elicit the error ENOENT.
160
161 Although process state and consequently the contents of /proc files can
162 change from instant to instant, a single read(2) of a /proc file is
163 guaranteed to return a sane representation of state; that is, the read
164 will be atomic with respect to the state of the process. No such
165 guarantee applies to successive reads applied to a /proc file for a
166 running process. In addition, atomicity is not guaranteed for I/O
167 applied to the as (address-space) file for a running process or for a
168 process whose address space contains memory shared by another running
169 process.
170
171 A number of structure definitions are used to describe the files. These
172 structures may grow by the addition of elements at the end in future
173 releases of the system and it is not legitimate for a program to assume
174 that they will not.
175
176 STRUCTURE OF /proc/pid
177 A given directory /proc/pid contains the following entries. A process
178 can use the invisible alias /proc/self if it wishes to open one of its
179 own /proc files (invisible in the sense that the name "self" does not
180 appear in a directory listing of /proc obtained from ls(1), getdents(2),
181 or readdir(3C)).
182
183 contracts
184 A directory containing references to the contracts held by the process.
185 Each entry is a symlink to the contract's directory under
186 /system/contract. See contract(4).
187
188 as
189 Contains the address-space image of the process; it can be opened for
190 both reading and writing. lseek(2) is used to position the file at the
191 virtual address of interest and then the address space can be examined or
192 changed through read(2) or write(2) (or by using pread(2) or pwrite(2)
193 for the combined operation).
194
195 ctl
196 A write-only file to which structured messages are written directing the
197 system to change some aspect of the process's state or control its
198 behavior in some way. The seek offset is not relevant when writing to
199 this file. Individual lwps also have associated lwpctl files in the lwp
200 subdirectories. A control message may be written either to the process's
201 ctl file or to a specific lwpctl file with operation-specific effects.
202 The effect of a control message is immediately reflected in the state of
203 the process visible through appropriate status and information files.
204 The types of control messages are described in detail later. See CONTROL
205 MESSAGES.
206
207 status
208 Contains state information about the process and the representative lwp.
209 The file contains a pstatus structure which contains an embedded
210 lwpstatus structure for the representative lwp, as follows:
211
212 typedef struct pstatus {
213 int pr_flags; /* flags (see below) */
214 int pr_nlwp; /* number of active lwps in the process */
215 int pr_nzomb; /* number of zombie lwps in the process */
216 pid_tpr_pid; /* process id */
217 pid_tpr_ppid; /* parent process id */
218 pid_tpr_pgid; /* process group id */
219 pid_tpr_sid; /* session id */
220 id_t pr_aslwpid; /* obsolete */
221 id_t pr_agentid; /* lwp-id of the agent lwp, if any */
222 sigset_t pr_sigpend; /* set of process pending signals */
223 uintptr_t pr_brkbase; /* virtual address of the process heap */
224 size_t pr_brksize; /* size of the process heap, in bytes */
225 uintptr_t pr_stkbase; /* virtual address of the process stack */
226 size_tpr_stksize; /* size of the process stack, in bytes */
227 timestruc_t pr_utime; /* process user cpu time */
228 timestruc_t pr_stime; /* process system cpu time */
229 timestruc_t pr_cutime; /* sum of children's user times */
230 timestruc_t pr_cstime; /* sum of children's system times */
231 sigset_t pr_sigtrace; /* set of traced signals */
232 fltset_t pr_flttrace; /* set of traced faults */
233 sysset_t pr_sysentry; /* set of system calls traced on entry */
234 sysset_t pr_sysexit; /* set of system calls traced on exit */
235 char pr_dmodel; /* data model of the process */
236 taskid_t pr_taskid; /* task id */
237 projid_t pr_projid; /* project id */
238 zoneid_t pr_zoneid; /* zone id */
239 lwpstatus_t pr_lwp; /* status of the representative lwp */
240 } pstatus_t;
241
242 pr_flags is a bit-mask holding the following process flags. For
243 convenience, it also contains the lwp flags for the representative lwp,
244 described later.
245
246 PR_ISSYS process is a system process (see PCSTOP).
247
248 PR_VFORKP process is the parent of a vforked child (see PCWATCH).
249
250 PR_FORK process has its inherit-on-fork mode set (see PCSET).
251
252 PR_RLC process has its run-on-last-close mode set (see PCSET).
253
254 PR_KLC process has its kill-on-last-close mode set (see PCSET).
255
256 PR_ASYNC process has its asynchronous-stop mode set (see PCSET).
257
258 PR_MSACCT Set by default in all processes to indicate that
259 microstate accounting is enabled. However, this flag
260 has been deprecated and no longer has any effect.
261 Microstate accounting may not be disabled; however, it
262 is still possible to toggle the flag.
263
264 PR_MSFORK Set by default in all processes to indicate that
265 microstate accounting will be enabled for processes that
266 this parent fork(2)s. However, this flag has been
267 deprecated and no longer has any effect. It is possible
268 to toggle this flag; however, it is not possible to
269 disable microstate accounting.
270
271 PR_BPTADJ process has its breakpoint adjustment mode set (see
272 PCSET).
273
274 PR_PTRACE process has its ptrace-compatibility mode set (see
275 PCSET).
276
277 pr_nlwp is the total number of active lwps in the process. pr_nzomb is
278 the total number of zombie lwps in the process. A zombie lwp is a non-
279 detached lwp that has terminated but has not been reaped with
280 thr_join(3C) or pthread_join(3C).
281
282 pr_pid, pr_ppi, pr_pgid, and pr_sid are, respectively, the process ID,
283 the ID of the process's parent, the process's process group ID, and the
284 process's session ID.
285
286 pr_aslwpid is obsolete and is always zero.
287
288 pr_agentid is the lwp-ID for the /proc agent lwp (see the PCAGENT control
289 operation). It is zero if there is no agent lwp in the process.
290
291 pr_sigpend identifies asynchronous signals pending for the process.
292
293 pr_brkbase is the virtual address of the process heap and pr_brksize is
294 its size in bytes. The address formed by the sum of these values is the
295 process break (see brk(2)). pr_stkbase and pr_stksize are, respectively,
296 the virtual address of the process stack and its size in bytes. (Each
297 lwp runs on a separate stack; the distinguishing characteristic of the
298 process stack is that the operating system will grow it when necessary.)
299
300 pr_utime, pr_stime, pr_cutime, and pr_cstime are, respectively, the user
301 CPU and system CPU time consumed by the process, and the cumulative user
302 CPU and system CPU time consumed by the process's children, in seconds
303 and nanoseconds.
304
305 pr_sigtrace and pr_flttrace contain, respectively, the set of signals and
306 the set of hardware faults that are being traced (see PCSTRACE and
307 PCSFAULT).
308
309 pr_sysentry and pr_sysexit contain, respectively, the sets of system
310 calls being traced on entry and exit (see PCSENTRY and PCSEXIT).
311
312 pr_dmodel indicates the data model of the process. Possible values are:
313
314 PR_MODEL_ILP32 process data model is ILP32.
315
316 PR_MODEL_LP64 process data model is LP64.
317
318 PR_MODEL_NATIVE process data model is native.
319
320 The pr_taskid, pr_projid, and pr_zoneid fields contain respectively, the
321 numeric IDs of the task, project, and zone in which the process was
322 running.
323
324 The constant PR_MODEL_NATIVE reflects the data model of the controlling
325 process, that is, its value is PR_MODEL_ILP32 or PR_MODEL_LP64 according
326 to whether the controlling process has been compiled as a 32-bit program
327 or a 64-bit program, respectively.
328
329 pr_lwp contains the status information for the representative lwp:
330
331 typedef struct lwpstatus {
332 int pr_flags; /* flags (see below) */
333 id_t pr_lwpid; /* specific lwp identifier */
334 short pr_why; /* reason for lwp stop, if stopped */
335 short pr_what; /* more detailed reason */
336 short pr_cursig; /* current signal, if any */
337 siginfo_t pr_info; /* info associated with signal or fault */
338 sigset_t pr_lwppend; /* set of signals pending to the lwp */
339 sigset_t pr_lwphold; /* set of signals blocked by the lwp */
340 struct sigaction pr_action;/* signal action for current signal */
341 stack_t pr_altstack; /* alternate signal stack info */
342 uintptr_t pr_oldcontext; /* address of previous ucontext */
343 short pr_syscall; /* system call number (if in syscall) */
344 short pr_nsysarg; /* number of arguments to this syscall */
345 int pr_errno; /* errno for failed syscall */
346 long pr_sysarg[PRSYSARGS]; /* arguments to this syscall */
347 long pr_rval1; /* primary syscall return value */
348 long pr_rval2; /* second syscall return value, if any */
349 char pr_clname[PRCLSZ]; /* scheduling class name */
350 timestruc_t pr_tstamp; /* real-time time stamp of stop */
351 timestruc_t pr_utime; /* lwp user cpu time */
352 timestruc_t pr_stime; /* lwp system cpu time */
353 uintptr_t pr_ustack; /* stack boundary data (stack_t) address */
354 ulong_t pr_instr; /* current instruction */
355 prgregset_t pr_reg; /* general registers */
356 prfpregset_t pr_fpreg; /* floating-point registers */
357 } lwpstatus_t;
358
359 pr_flags is a bit-mask holding the following lwp flags. For convenience,
360 it also contains the process flags, described previously.
361
362 PR_STOPPED The lwp is stopped.
363
364 PR_ISTOP The lwp is stopped on an event of interest (see
365 PCSTOP).
366
367 PR_DSTOP The lwp has a stop directive in effect (see PCSTOP).
368
369 PR_STEP The lwp has a single-step directive in effect (see
370 PCRUN).
371
372 PR_ASLEEP The lwp is in an interruptible sleep within a system
373 call.
374
375 PR_PCINVAL The lwp's current instruction (pr_instr) is undefined.
376
377 PR_DETACH This is a detached lwp (see pthread_create(3C) and
378 pthread_join(3C)).
379
380 PR_DAEMON This is a daemon lwp (see pthread_create(3C)).
381
382 PR_ASLWP This flag is obsolete and is never set.
383
384 PR_AGENT This is the /proc agent lwp for the process.
385
386 pr_lwpid names the specific lwp.
387
388 pr_why and pr_what together describe, for a stopped lwp, the reason for
389 the stop. Possible values of pr_why and the associated pr_what are:
390
391 PR_REQUESTED indicates that the stop occurred in response to a stop
392 directive, normally because PCSTOP was applied or
393 because another lwp stopped on an event of interest
394 and the asynchronous-stop flag (see PCSET) was not set
395 for the process. pr_what is unused in this case.
396
397 PR_SIGNALLED indicates that the lwp stopped on receipt of a signal
398 (see PCSTRACE); pr_what holds the signal number that
399 caused the stop (for a newly-stopped lwp, the same
400 value is in pr_cursig).
401
402 PR_FAULTED indicates that the lwp stopped on incurring a hardware
403 fault (see PCSFAULT); pr_what holds the fault number
404 that caused the stop.
405
406 PR_SYSENTRY
407
408 PR_SYSEXIT indicate a stop on entry to or exit from a system call
409 (see PCSENTRY and PCSEXIT); pr_what holds the system
410 call number.
411
412 PR_JOBCONTROL indicates that the lwp stopped due to the default
413 action of a job control stop signal (see
414 sigaction(2)); pr_what holds the stopping signal
415 number.
416
417 PR_SUSPENDED indicates that the lwp stopped due to internal
418 synchronization of lwps within the process. pr_what
419 is unused in this case.
420
421 pr_cursig names the current signal, that is, the next signal to be
422 delivered to the lwp, if any. pr_info, when the lwp is in a PR_SIGNALLED
423 or PR_FAULTED stop, contains additional information pertinent to the
424 particular signal or fault (see <sys/siginfo.h>).
425
426 pr_lwppend identifies any synchronous or directed signals pending for the
427 lwp. pr_lwphold identifies those signals whose delivery is being blocked
428 by the lwp (the signal mask).
429
430 pr_action contains the signal action information pertaining to the
431 current signal (see sigaction(2)); it is undefined if pr_cursig is zero.
432 pr_altstack contains the alternate signal stack information for the lwp
433 (see sigaltstack(2)).
434
435 pr_oldcontext, if not zero, contains the address on the lwp stack of a
436 ucontext structure describing the previous user-level context (see
437 ucontext.h(3HEAD)). It is non-zero only if the lwp is executing in the
438 context of a signal handler.
439
440 pr_syscall is the number of the system call, if any, being executed by
441 the lwp; it is non-zero if and only if the lwp is stopped on PR_SYSENTRY
442 or PR_SYSEXIT, or is asleep within a system call (PR_ASLEEP is set). If
443 pr_syscall is non-zero, pr_nsysarg is the number of arguments to the
444 system call and pr_sysarg contains the actual arguments.
445
446 pr_rval1, pr_rval2, and pr_errno are defined only if the lwp is stopped
447 on PR_SYSEXIT or if the PR_VFORKP flag is set. If pr_errno is zero,
448 pr_rval1 and pr_rval2 contain the return values from the system call.
449 Otherwise, pr_errno contains the error number for the failing system call
450 (see <sys/errno.h>).
451
452 pr_clname contains the name of the lwp's scheduling class.
453
454 pr_tstamp, if the lwp is stopped, contains a time stamp marking when the
455 lwp stopped, in real time seconds and nanoseconds since an arbitrary time
456 in the past.
457
458 pr_utime is the amount of user level CPU time used by this LWP.
459
460 pr_stime is the amount of system level CPU time used by this LWP.
461
462 pr_ustack is the virtual address of the stack_t that contains the stack
463 boundaries for this LWP. See getustack(2) and _stack_grow(3C).
464
465 pr_instr contains the machine instruction to which the lwp's program
466 counter refers. The amount of data retrieved from the process is
467 machine-dependent. On SPARC based machines, it is a 32-bit word. On
468 x86-based machines, it is a single byte. In general, the size is that of
469 the machine's smallest instruction. If PR_PCINVAL is set, pr_instr is
470 undefined; this occurs whenever the lwp is not stopped or when the
471 program counter refers to an invalid virtual address.
472
473 pr_reg is an array holding the contents of a stopped lwp's general
474 registers.
475
476 SPARC On SPARC-based machines, the predefined constants
477 R_G0 ... R_G7, R_O0 ... R_O7, R_L0 ... R_L7,
478 R_I0 ... R_I7, R_PC, R_nPC, and R_Y can be used
479 as indices to refer to the corresponding
480 registers; previous register windows can be read
481 from their overflow locations on the stack
482 (however, see the gwindows file in the
483 /proc/pid/lwp/lwpid subdirectory).
484
485 SPARC V8 (32-bit) For SPARC V8 (32-bit) controlling processes, the
486 predefined constants R_PSR, R_WIM, and R_TBR can
487 be used as indices to refer to the corresponding
488 special registers. For SPARC V9 (64-bit)
489 controlling processes, the predefined constants
490 R_CCR, R_ASI, and R_FPRS can be used as indices to
491 refer to the corresponding special registers.
492
493 x86 (32-bit) For 32-bit x86 processes, the predefined constants
494 listed belowcan be used as indices to refer to the
495 corresponding registers.
496 SS
497 UESP
498 EFL
499 CS
500 EIP
501 ERR
502 TRAPNO
503 EAX
504 ECX
505 EDX
506 EBX
507 ESP
508 EBP
509 ESI
510 EDI
511 DS
512 ES
513 GS
514
515 The preceding constants are listed in
516 <sys/regset.h>.
517
518 Note that a 32-bit process can run on an x86
519 64-bit system, using the constants listed above.
520
521 x86 (64-bit) To read the registers of a 32- or a 64-bit
522 process, a 64-bit x86 process should use the
523 predefined constants listed below.
524 REG_GSBASE
525 REG_FSBASE
526 REG_DS
527 REG_ES
528 REG_GS
529 REG_FS
530 REG_SS
531 REG_RSP
532 REG_RFL
533 REG_CS
534 REG_RIP
535 REG_ERR
536 REG_TRAPNO
537 REG_RAX
538 REG_RCX
539 REG_RDX
540 REG_RBX
541 REG_RBP
542 REG_RSI
543 REG_RDI
544 REG_R8
545 REG_R9
546 REG_R10
547 REG_R11
548 REG_R12
549 REG_R13
550 REG_R14
551 REG_R15
552
553 The preceding constants are listed in
554 <sys/regset.h>.
555
556 pr_fpreg is a structure holding the contents of the floating-point
557 registers.
558
559 SPARC registers, both general and floating-point, as seen by a 64-bit
560 controlling process are the V9 versions of the registers, even if the
561 target process is a 32-bit (V8) process. V8 registers are a subset of
562 the V9 registers.
563
564 If the lwp is not stopped, all register values are undefined.
565
566 psinfo
567 Contains miscellaneous information about the process and the
568 representative lwp needed by the ps(1) command. psinfo remains
569 accessible after a process becomes a zombie. The file contains a psinfo
570 structure which contains an embedded lwpsinfo structure for the
571 representative lwp, as follows:
572
573 typedef struct psinfo {
574 int pr_flag; /* process flags (DEPRECATED: see below) */
575 int pr_nlwp; /* number of active lwps in the process */
576 int pr_nzomb; /* number of zombie lwps in the process */
577 pid_t pr_pid; /* process id */
578 pid_t pr_ppid; /* process id of parent */
579 pid_t pr_pgid; /* process id of process group leader */
580 pid_t pr_sid; /* session id */
581 uid_t pr_uid; /* real user id */
582 uid_t pr_euid; /* effective user id */
583 gid_t pr_gid; /* real group id */
584 gid_t pr_egid; /* effective group id */
585 uintptr_t pr_addr; /* address of process */
586 size_t pr_size; /* size of process image in Kbytes */
587 size_t pr_rssize; /* resident set size in Kbytes */
588 dev_t pr_ttydev; /* controlling tty device (or PRNODEV) */
589 ushort_t pr_pctcpu; /* % of recent cpu time used by all lwps */
590 ushort_t pr_pctmem; /* % of system memory used by process */
591 timestruc_t pr_start; /* process start time, from the epoch */
592 timestruc_t pr_time; /* cpu time for this process */
593 timestruc_t pr_ctime; /* cpu time for reaped children */
594 char pr_fname[PRFNSZ]; /* name of exec'ed file */
595 char pr_psargs[PRARGSZ]; /* initial characters of arg list */
596 int pr_wstat; /* if zombie, the wait() status */
597 int pr_argc; /* initial argument count */
598 uintptr_t pr_argv; /* address of initial argument vector */
599 uintptr_t pr_envp; /* address of initial environment vector */
600 char pr_dmodel; /* data model of the process */
601 taskid_t pr_taskid; /* task id */
602 projid_t pr_projid; /* project id */
603 poolid_t pr_poolid; /* pool id */
604 zoneid_t pr_zoneid; /* zone id */
605 ctid_t pr_contract; /* process contract id */
606 lwpsinfo_t pr_lwp; /* information for representative lwp */
607 } psinfo_t;
608
609 Some of the entries in psinfo, such as pr_addr, refer to internal kernel
610 data structures and should not be expected to retain their meanings
611 across different versions of the operating system.
612
613 psinfo_t.pr_flag is a deprecated interface that should no longer be used.
614 Applications currently relying on the SSYS bit in pr_flag should migrate
615 to checking PR_ISSYS in the pstatus structure's pr_flags field.
616
617 pr_pctcpu and pr_pctmem are 16-bit binary fractions in the range 0.0 to
618 1.0 with the binary point to the right of the high-order bit (1.0 ==
619 0x8000). pr_pctcpu is the summation over all lwps in the process.
620
621 pr_lwp contains the ps(1) information for the representative lwp. If the
622 process is a zombie, pr_nlwp, pr_nzomb, and pr_lwp.pr_lwpid are zero and
623 the other fields of pr_lwp are undefined:
624
625 typedef struct lwpsinfo {
626 int pr_flag; /* lwp flags (DEPRECATED: see below) */
627 id_t pr_lwpid; /* lwp id */
628 uintptr_t pr_addr; /* internal address of lwp */
629 uintptr_t pr_wchan; /* wait addr for sleeping lwp */
630 char pr_stype; /* synchronization event type */
631 char pr_state; /* numeric lwp state */
632 char pr_sname; /* printable character for pr_state */
633 char pr_nice; /* nice for cpu usage */
634 short pr_syscall; /* system call number (if in syscall) */
635 char pr_oldpri; /* pre-SVR4, low value is high priority */
636 char pr_cpu; /* pre-SVR4, cpu usage for scheduling */
637 int pr_pri; /* priority, high value = high priority */
638 ushort_t pr_pctcpu; /* % of recent cpu time used by this lwp */
639 timestruc_t pr_start; /* lwp start time, from the epoch */
640 timestruc_t pr_time; /* cpu time for this lwp */
641 char pr_clname[PRCLSZ]; /* scheduling class name */
642 char pr_name[PRFNSZ]; /* name of system lwp */
643 processorid_t pr_onpro; /* processor which last ran this lwp */
644 processorid_t pr_bindpro;/* processor to which lwp is bound */
645 psetid_t pr_bindpset; /* processor set to which lwp is bound */
646 lgrp_id_t pr_lgrp; /* home lgroup */
647 } lwpsinfo_t;
648
649 Some of the entries in lwpsinfo, such as pr_addr, pr_wchan, pr_stype,
650 pr_state, and pr_name, refer to internal kernel data structures and
651 should not be expected to retain their meanings across different versions
652 of the operating system.
653
654 lwpsinfo_t.pr_flag is a deprecated interface that should no longer be
655 used.
656
657 pr_pctcpu is a 16-bit binary fraction, as described above. It represents
658 the CPU time used by the specific lwp. On a multi-processor machine, the
659 maximum value is 1/N, where N is the number of CPUs.
660
661 pr_contract is the id of the process contract of which the process is a
662 member. See contract(4) and process(4).
663
664 cred
665 Contains a description of the credentials associated with the process:
666
667 typedef struct prcred {
668 uid_t pr_euid; /* effective user id */
669 uid_t pr_ruid; /* real user id */
670 uid_t pr_suid; /* saved user id (from exec) */
671 gid_t pr_egid; /* effective group id */
672 gid_t pr_rgid; /* real group id */
673 gid_t pr_sgid; /* saved group id (from exec) */
674 int pr_ngroups; /* number of supplementary groups */
675 gid_t pr_groups[1]; /* array of supplementary groups */
676 } prcred_t;
677
678 The array of associated supplementary groups in pr_groups
679 is of variable length; the cred file contains all of the supplementary
680 groups. pr_ngroups indicates the number of supplementary groups. (See
681 also the PCSCRED and PCSCREDX control operations.)
682
683 priv
684 Contains a description of the privileges associated with the process:
685
686 typedef struct prpriv {
687 uint32_t pr_nsets; /* number of privilege set */
688 uint32_t pr_setsize; /* size of privilege set */
689 uint32_t pr_infosize; /* size of supplementary data */
690 priv_chunk_t pr_sets[1]; /* array of sets */
691 } prpriv_t;
692
693 The actual dimension of the pr_sets[] field is
694 pr_sets[pr_nsets][pr_setsize]
695
696 which is followed by additional information about the process state
697 pr_infosize bytes in size.
698
699 The full size of the structure can be computed using
700 PRIV_PRPRIV_SIZE(prpriv_t *).
701
702 secflags
703 This file contains the security-flags of the process. It contains a
704 description of the security flags associated with the process.
705
706 typedef struct prsecflags {
707 uint32_t pr_version; /* ABI Versioning of this structure */
708 secflagset_t pr_effective; /* Effective flags */
709 secflagset_t pr_inherit; /* Inheritable flags */
710 secflagset_t pr_lower; /* Lower flags */
711 secflagset_t pr_upper; /* Upper flags */
712 } prsecflags_t;
713
714 The pr_version field is a version number for the structure, currently
715 PRSECFLAGS_VERSION_1.
716
717 sigact
718 Contains an array of sigaction structures describing the current
719 dispositions of all signals associated with the traced process (see
720 sigaction(2)). Signal numbers are displaced by 1 from array indices, so
721 that the action for signal number n appears in position n-1 of the array.
722
723 auxv
724 Contains the initial values of the process's aux vector in an array of
725 auxv_t structures (see <sys/auxv.h>). The values are those that were
726 passed by the operating system as startup information to the dynamic
727 linker.
728
729 ldt
730 This file exists only on x86-based machines. It is non-empty only if the
731 process has established a local descriptor table (LDT). If non-empty,
732 the file contains the array of currently active LDT entries in an array
733 of elements of type struct ssd, defined in <sys/sysi86.h>, one element
734 for each active LDT entry.
735
736 map, xmap
737 Contain information about the virtual address map of the process. The
738 map file contains an array of prmap structures while the xmap file
739 contains an array of prxmap structures. Each structure describes a
740 contiguous virtual address region in the address space of the traced
741 process:
742
743 typedef struct prmap {
744 uintptr_tpr_vaddr; /* virtual address of mapping */
745 size_t pr_size; /* size of mapping in bytes */
746 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
747 offset_t pr_offset; /* offset into mapped object, if any */
748 int pr_mflags; /* protection and attribute flags */
749 int pr_pagesize; /* pagesize for this mapping in bytes */
750 int pr_shmid; /* SysV shared memory identifier */
751 } prmap_t;
752
753 typedef struct prxmap {
754 uintptr_t pr_vaddr; /* virtual address of mapping */
755 size_t pr_size; /* size of mapping in bytes */
756 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
757 offset_t pr_offset; /* offset into mapped object, if any */
758 int pr_mflags; /* protection and attribute flags */
759 int pr_pagesize; /* pagesize for this mapping in bytes */
760 int pr_shmid; /* SysV shared memory identifier */
761 dev_t pr_dev; /* device of mapped object, if any */
762 uint64_t pr_ino; /* inode of mapped object, if any */
763 size_t pr_rss; /* pages of resident memory */
764 size_t pr_anon; /* pages of resident anonymous memory */
765 size_t pr_locked; /* pages of locked memory */
766 uint64_t pr_hatpagesize; /* pagesize of mapping */
767 } prxmap_t;
768
769 pr_vaddr is the virtual address of the mapping within the traced process
770 and pr_size is its size in bytes. pr_mapname, if it does not contain a
771 null string, contains the name of a file in the object directory (see
772 below) that can be opened read-only to obtain a file descriptor for the
773 mapped file associated with the mapping. This enables a debugger to find
774 object file symbol tables without having to know the real path names of
775 the executable file and shared libraries of the process. pr_offset is
776 the 64-bit offset within the mapped file (if any) to which the virtual
777 address is mapped.
778
779 pr_mflags is a bit-mask of protection and attribute flags:
780
781 MA_READ mapping is readable by the traced process.
782
783 MA_WRITE mapping is writable by the traced process.
784
785 MA_EXEC mapping is executable by the traced process.
786
787 MA_SHARED mapping changes are shared by the mapped object.
788
789 MA_ISM mapping is intimate shared memory (shared MMU
790 resources)
791
792 MAP_NORESERVE
793 mapping does not have swap space reserved (mapped with
794 MAP_NORESERVE)
795
796 MA_SHM mapping System V shared memory
797
798 A contiguous area of the address space having the same underlying mapped
799 object may appear as multiple mappings due to varying read, write, and
800 execute attributes. The underlying mapped object does not change over
801 the range of a single mapping. An I/O operation to a mapping marked
802 MA_SHARED fails if applied at a virtual address not corresponding to a
803 valid page in the underlying mapped object. A write to a MA_SHARED
804 mapping that is not marked MA_WRITE fails. Reads and writes to private
805 mappings always succeed. Reads and writes to unmapped addresses fail.
806
807 pr_pagesize is the page size for the mapping, currently always the system
808 pagesize.
809
810 pr_shmid is the shared memory identifier, if any, for the mapping. Its
811 value is -1 if the mapping is not System V shared memory. See shmget(2).
812
813 pr_dev is the device of the mapped object, if any, for the mapping. Its
814 value is PRNODEV (-1) if the mapping does not have a device.
815
816 pr_ino is the inode of the mapped object, if any, for the mapping. Its
817 contents are only valid if pr_dev is not PRNODEV.
818
819 pr_rss is the number of resident pages of memory for the mapping. The
820 number of resident bytes for the mapping may be determined by multiplying
821 pr_rss by the page size given by pr_pagesize.
822
823 pr_anon is the number of resident anonymous memory pages (pages which are
824 private to this process) for the mapping.
825
826 pr_locked is the number of locked pages for the mapping. Pages which are
827 locked are always resident in memory.
828
829 pr_hatpagesize is the size, in bytes, of the HAT (MMU) translation for
830 the mapping. pr_hatpagesize may be different than pr_pagesize. The
831 possible values are hardware architecture specific, and may change over a
832 mapping's lifetime.
833
834 rmap
835 Contains information about the reserved address ranges of the process.
836 The file contains an array of prmap structures, as defined above for the
837 map file. Each structure describes a contiguous virtual address region
838 in the address space of the traced process that is reserved by the system
839 in the sense that an mmap(2) system call that does not specify MAP_FIXED
840 will not use any part of it for the new mapping. Examples of such
841 reservations include the address ranges reserved for the process stack
842 and the individual thread stacks of a multi-threaded process.
843
844 cwd
845 A symbolic link to the process's current working directory. See
846 chdir(2). A readlink(2) of /proc/pid/cwd yields a null string. However,
847 it can be opened, listed, and searched as a directory, and can be the
848 target of chdir(2).
849
850 root
851 A symbolic link to the process's root directory. /proc/pid/root can
852 differ from the system root directory if the process or one of its
853 ancestors executed chroot(2) as super user. It has the same semantics as
854 /proc/pid/cwd.
855
856 fd
857 A directory containing references to the open files of the process. Each
858 entry is a decimal number corresponding to an open file descriptor in the
859 process.
860
861 If an entry refers to a regular file, it can be opened with normal file
862 system semantics but, to ensure that the controlling process cannot gain
863 greater access than the controlled process, with no file access modes
864 other than its read/write open modes in the controlled process. If an
865 entry refers to a directory, it can be accessed with the same semantics
866 as /proc/pid/cwd. An attempt to open any other type of entry fails with
867 EACCES.
868
869 fdinfo
870 A directory containing information about each of the process's open
871 files. Each entry is a decimal number corresponding to an open file
872 descriptor in the process. Each file contains a prfdinfo_t structure
873 defined as follows:
874
875 typedef struct prfdinfo {
876 int pr_fd; /* file descriptor number */
877 mode_t pr_mode; /* (see st_mode in stat(2)) */
878 uint64_t pr_ino; /* inode number */
879 uint64_t pr_size; /* file size */
880 int64_t pr_offset; /* current offset of file descriptor */
881 uid_t pr_uid; /* owner's user id */
882 gid_t pr_gid; /* owner's group id */
883 major_t pr_major; /* major number of device containing file */
884 minor_t pr_minor; /* minor number of device containing file */
885 major_t pr_rmajor; /* major number (if special file) */
886 minor_t pr_rminor; /* minor number (if special file) */
887 int pr_fileflags; /* (see F_GETXFL in fcntl(2)) */
888 int pr_fdflags; /* (see F_GETFD in fcntl(2)) */
889 short pr_locktype; /* (see F_GETLK in fcntl(2)) */
890 pid_t pr_lockpid; /* process holding file lock (see F_GETLK) */
891 int pr_locksysid; /* sysid of locking process (see F_GETLK) */
892 pid_t pr_peerpid; /* peer process (socket, door) */
893 int pr_filler[25]; /* reserved for future use */
894 char pr_peername[PRFNSZ]; /* peer process name */
895 #if __STDC_VERSION__ >= 199901L
896 char pr_misc[]; /* self describing structures */
897 #else
898 char pr_misc[1];
899 #endif
900 } prfdinfo_t;
901
902 The pr_misc element points to a list of additional miscellaneous data
903 items, each of which has a header of type pr_misc_header_t specifying the
904 size and type, and some data which immediately follow the header.
905
906 typedef struct pr_misc_header {
907 uint_t pr_misc_size;
908 uint_t pr_misc_type;
909 } pr_misc_header_t;
910
911 The pr_misc_size field is the sum of the sizes of the header and the
912 associated data and any trailing padding bytes which will be set to zero.
913 The end of the list is indicated by a header with a zero size and a type
914 with all bits set.
915
916 The following miscellaneous data types can be present:
917
918 PR_PATHNAME The file descriptor's path in the
919 filesystem. This is a NUL-terminated
920 sequence of characters.
921
922 PR_SOCKETNAME A sockaddr structure representing the
923 local socket name for this file
924 descriptor, as would be returned by
925 calling getsockname() within the process.
926
927 PR_PEERSOCKNAME A sockaddr structure representing the peer
928 socket name for this file descriptor, as
929 would be returned by calling getpeername()
930 within the process.
931
932 PR_SOCKOPTS_BOOL_OPTS An unsigned integer which has bits set
933 corresponding to options which are set on
934 the underlying socket. The following bits
935 may be set:
936
937 PR_SO_DEBUG
938
939 PR_SO_REUSEADDR
940
941 PR_SO_REUSEPORT
942
943 PR_SO_KEEPALIVE
944
945 PR_SO_DONTROUTE
946
947 PR_SO_BROADCAST
948
949 PR_SO_OOBINLINE
950
951 PR_SO_DGRAM_ERRIND
952
953 PR_SO_ALLZONES
954
955 PR_SO_MAC_EXEMPT
956
957 PR_SO_EXCLBIND
958
959 PR_SO_PASSIVE_CONNECT
960
961 PR_SO_ACCEPTCONN
962
963 PR_UDP_NAT_T_ENDPOINT
964
965 PR_SO_VRRP
966
967 PR_SO_MAC_IMPLICIT
968
969 PR_SOCKOPT_LINGER A struct linger as would be returned by
970 calling getsockopt(SO_LINGER) within the
971 process.
972
973 PR_SOCKOPT_SNDBUF The data that would be returned by calling
974 getsockopt(SO_SNDBUF) within the process.
975
976 PR_SOCKOPT_RCVBUF The data that would be returned by calling
977 getsockopt(SO_RCVBUF) within the process.
978
979 PR_SOCKOPT_IP_NEXTHOP The data that would be returned by calling
980 getsockopt(IPPROTO_IP, IP_NEXTHOP) within
981 the process.
982
983 PR_SOCKOPT_IPV6_NEXTHOP The data that would be returned by calling
984 getsockopt(IPPROTO_IPV6, IPV6_NEXTHOP)
985 within the process.
986
987 PR_SOCKOPT_TYPE The data that would be returned by calling
988 getsockopt(SO_TYPE) within the process.
989
990 PR_SOCKOPT_TCP_CONGESTION For TCP sockets, the data that would be
991 returned by calling
992 getsockopt(IPPROTO_TCP, TCP_CONGESTION)
993 within the process. This is a NUL-
994 terminated character array containing the
995 name of the congestion algorithm in use
996 for the socket.
997
998 PR_SOCKFILTERS_PRIV Private data relating to up to the first
999 32 socket filters pushed on this
1000 descriptor.
1001
1002 object
1003 A directory containing read-only files with names corresponding to the
1004 pr_mapname entries in the map and pagedata files. Opening such a file
1005 yields a file descriptor for the underlying mapped file associated with
1006 an address-space mapping in the process. The file name a.out appears in
1007 the directory as an alias for the process's executable file.
1008
1009 The object directory makes it possible for a controlling process to gain
1010 access to the object file and any shared libraries (and consequently the
1011 symbol tables) without having to know the actual path names of the
1012 executable files.
1013
1014 path
1015 A directory containing symbolic links to files opened by the process.
1016 The directory includes one entry for cwd and root. The directory also
1017 contains a numerical entry for each file descriptor in the fd directory,
1018 and entries matching those in the object directory. If this information
1019 is not available, any attempt to read the contents of the symbolic link
1020 will fail. This is most common for files that do not exist in the
1021 filesystem namespace (such as FIFOs and sockets), but can also happen for
1022 regular files. For the file descriptor entries, the path may be
1023 different from the one used by the process to open the file.
1024
1025 pagedata
1026 Opening the page data file enables tracking of address space references
1027 and modifications on a per-page basis.
1028
1029 A read(2) of the page data file descriptor returns structured page data
1030 and atomically clears the page data maintained for the file by the
1031 system. That is to say, each read returns data collected since the last
1032 read; the first read returns data collected since the file was opened.
1033 When the call completes, the read buffer contains the following structure
1034 as its header and thereafter contains a number of section header
1035 structures and associated byte arrays that must be accessed by walking
1036 linearly through the buffer.
1037
1038 typedef struct prpageheader {
1039 timestruc_t pr_tstamp; /* real time stamp, time of read() */
1040 ulong_t pr_nmap; /* number of address space mappings */
1041 ulong_t pr_npage; /* total number of pages */
1042 } prpageheader_t;
1043
1044 The header is followed by pr_nmap prasmap structures and associated data
1045 arrays. The prasmap structure contains the following elements:
1046
1047 typedef struct prasmap {
1048 uintptr_t pr_vaddr; /* virtual address of mapping */
1049 ulong_t pr_npage; /* number of pages in mapping */
1050 char pr_mapname[PRMAPSZ]; /* name in /proc/pid/object */
1051 offset_t pr_offset; /* offset into mapped object, if any */
1052 int pr_mflags; /* protection and attribute flags */
1053 int pr_pagesize; /* pagesize for this mapping in bytes */
1054 int pr_shmid; /* SysV shared memory identifier */
1055 } prasmap_t;
1056
1057 Each section header is followed by pr_npage bytes, one byte for each page
1058 in the mapping, plus 0-7 null bytes at the end so that the next prasmap
1059 structure begins on an eight-byte aligned boundary. Each data byte may
1060 contain these flags:
1061
1062 PG_REFERENCED page has been referenced.
1063
1064 PG_MODIFIED page has been modified.
1065
1066 If the read buffer is not large enough to contain all of the page data,
1067 the read fails with E2BIG and the page data is not cleared. The required
1068 size of the read buffer can be determined through fstat(2). Application
1069 of lseek(2) to the page data file descriptor is ineffective; every read
1070 starts from the beginning of the file. Closing the page data file
1071 descriptor terminates the system overhead associated with collecting the
1072 data.
1073
1074 More than one page data file descriptor for the same process can be
1075 opened, up to a system-imposed limit per traced process. A read of one
1076 does not affect the data being collected by the system for the others.
1077 An open of the page data file will fail with ENOMEM if the system-imposed
1078 limit would be exceeded.
1079
1080 watch
1081 Contains an array of prwatch structures, one for each watched area
1082 established by the PCWATCH control operation. See PCWATCH for details.
1083
1084 usage
1085 Contains process usage information described by a prusage structure which
1086 contains at least the following fields:
1087
1088 typedef struct prusage {
1089 id_t pr_lwpid; /* lwp id. 0: process or defunct */
1090 int pr_count; /* number of contributing lwps */
1091 timestruc_t pr_tstamp; /* real time stamp, time of read() */
1092 timestruc_t pr_create; /* process/lwp creation time stamp */
1093 timestruc_t pr_term; /* process/lwp termination time stamp */
1094 timestruc_t pr_rtime; /* total lwp real (elapsed) time */
1095 timestruc_t pr_utime; /* user level CPU time */
1096 timestruc_t pr_stime; /* system call CPU time */
1097 timestruc_t pr_ttime; /* other system trap CPU time */
1098 timestruc_t pr_tftime; /* text page fault sleep time */
1099 timestruc_t pr_dftime; /* data page fault sleep time */
1100 timestruc_t pr_kftime; /* kernel page fault sleep time */
1101 timestruc_t pr_ltime; /* user lock wait sleep time */
1102 timestruc_t pr_slptime; /* all other sleep time */
1103 timestruc_t pr_wtime; /* wait-cpu (latency) time */
1104 timestruc_t pr_stoptime; /* stopped time */
1105 ulong_t pr_minf; /* minor page faults */
1106 ulong_t pr_majf; /* major page faults */
1107 ulong_t pr_nswap; /* swaps */
1108 ulong_t pr_inblk; /* input blocks */
1109 ulong_t pr_oublk; /* output blocks */
1110 ulong_t pr_msnd; /* messages sent */
1111 ulong_t pr_mrcv; /* messages received */
1112 ulong_t pr_sigs; /* signals received */
1113 ulong_t pr_vctx; /* voluntary context switches */
1114 ulong_t pr_ictx; /* involuntary context switches */
1115 ulong_t pr_sysc; /* system calls */
1116 ulong_t pr_ioch; /* chars read and written */
1117 } prusage_t;
1118
1119 Microstate accounting is now continuously enabled. While this
1120 information was previously an estimate, if microstate accounting were not
1121 enabled, the current information is now never an estimate represents time
1122 the process has spent in various states.
1123
1124 lstatus
1125 Contains a prheader structure followed by an array of lwpstatus
1126 structures, one for each active lwp in the process (see also
1127 /proc/pid/lwp/lwpid/lwpstatus, below). The prheader structure describes
1128 the number and size of the array entries that follow.
1129
1130 typedef struct prheader {
1131 long pr_nent; /* number of entries */
1132 size_t pr_entsize; /* size of each entry, in bytes */
1133 } prheader_t;
1134
1135 The lwpstatus structure may grow by the addition of elements at the end
1136 in future releases of the system. Programs must use pr_entsize in the
1137 file header to index through the array. These comments apply to all
1138 /proc files that include a prheader structure (lpsinfo and lusage,
1139 below).
1140
1141 lpsinfo
1142 Contains a prheader structure followed by an array of lwpsinfo
1143 structures, one for eachactive and zombie lwp in the process. See also
1144 /proc/pid/lwp/lwpid/lwpsinfo, below.
1145
1146 lusage
1147 Contains a prheader structure followed by an array of prusage structures,
1148 one for each active lwp in the process, plus an additional element at the
1149 beginning that contains the summation over all defunct lwps (lwps that
1150 once existed but no longer exist in the process). Excluding the
1151 pr_lwpid, pr_tstamp, pr_create, and pr_term entries, the entry-by-entry
1152 summation over all these structures is the definition of the process
1153 usage information obtained from the usage file. (See also
1154 /proc/pid/lwp/lwpid/lwpusage, below.)
1155
1156 lwp
1157 A directory containing entries each of which names an active or zombie
1158 lwp within the process. These entries are themselves directories
1159 containing additional files as described below. Only the lwpsinfo file
1160 exists in the directory of a zombie lwp.
1161
1162 STRUCTURE OF /proc/pid/lwp/lwpid
1163 A given directory /proc/pid/lwp/lwpid contains the following entries:
1164
1165 lwpctl
1166 Write-only control file. The messages written to this file affect the
1167 specific lwp rather than the representative lwp, as is the case for the
1168 process's ctl file.
1169
1170 lwpname
1171 A buffer of THREAD_NAME_MAX bytes representing the LWP name; the buffer
1172 is zero-filled if the thread name is shorter than the buffer. If no
1173 thread name is set, the buffer contains the empty string. A read with a
1174 buffer shorter than THREAD_NAME_MAX bytes is not guaranteed to be NUL-
1175 terminated. Writing to this file will set the LWP name for the specific
1176 lwp. This file may not be present in older operating system versions.
1177 THREAD_NAME_MAX may increase in the future; clients should be prepared
1178 for this.
1179
1180 lwpstatus
1181 lwp-specific state information. This file contains the lwpstatus
1182 structure for the specific lwp as described above for the representative
1183 lwp in the process's status file.
1184
1185 lwpsinfo
1186 lwp-specific ps(1) information. This file contains the lwpsinfo
1187 structure for the specific lwp as described above for the representative
1188 lwp in the process's psinfo file. The lwpsinfo file remains accessible
1189 after an lwp becomes a zombie.
1190
1191 lwpusage
1192 This file contains the prusage structure for the specific lwp as
1193 described above for the process's usage file.
1194
1195 gwindows
1196 This file exists only on SPARC based machines. If it is non-empty, it
1197 contains a gwindows_t structure, defined in <sys/regset.h>, with the
1198 values of those SPARC register windows that could not be stored on the
1199 stack when the lwp stopped. Conditions under which register windows are
1200 not stored on the stack are: the stack pointer refers to nonexistent
1201 process memory or the stack pointer is improperly aligned. If the lwp is
1202 not stopped or if there are no register windows that could not be stored
1203 on the stack, the file is empty (the usual case).
1204
1205 xregs
1206 Extra state registers. The extra state register set is architecture
1207 dependent; this file is empty if the system does not support extra state
1208 registers. If the file is non-empty, it contains an architecture
1209 dependent structure of type prxregset_t, defined in <procfs.h>, with the
1210 values of the lwp's extra state registers. If the lwp is not stopped,
1211 all register values are undefined. See also the PCSXREG control
1212 operation, below.
1213
1214 asrs
1215 This file exists only for 64-bit SPARC V9 processes. It contains an
1216 asrset_t structure, defined in <sys/regset.h>, containing the values of
1217 the lwp's platform-dependent ancillary state registers. If the lwp is
1218 not stopped, all register values are undefined. See also the PCSASRS
1219 control operation, below.
1220
1221 spymaster
1222 For an agent lwp (see PCAGENT), this file contains a psinfo_t structure
1223 that corresponds to the process that created the agent lwp at the time
1224 the agent was created. This structure is identical to that retrieved via
1225 the psinfo file, with one modification: the pr_time field does not
1226 correspond to the CPU time for the process, but rather to the creation
1227 time of the agent lwp.
1228
1229 templates
1230 A directory which contains references to the active templates for the
1231 lwp, named by the contract type. Changes made to an active template
1232 descriptor do not affect the original template which was activated,
1233 though they do affect the active template. It is not possible to
1234 activate an active template descriptor. See contract(4).
1235
1236 CONTROL MESSAGES
1237 Process state changes are effected through messages written to a
1238 process's ctl file or to an individual lwp's lwpctl file. All control
1239 messages consist of a long that names the specific operation followed by
1240 additional data containing the operand, if any.
1241
1242 Multiple control messages may be combined in a single write(2) (or
1243 writev(2)) to a control file, but no partial writes are permitted. That
1244 is, each control message, operation code plus operand, if any, must be
1245 presented in its entirety to the write(2) and not in pieces over several
1246 system calls. If a control operation fails, no subsequent operations
1247 contained in the same write(2) are attempted.
1248
1249 Descriptions of the allowable control messages follow. In all cases,
1250 writing a message to a control file for a process or lwp that has
1251 terminated elicits the error ENOENT.
1252
1253 PCSTOP PCDSTOP PCWSTOP PCTWSTOP
1254 When applied to the process control file, PCSTOP directs all lwps to stop
1255 and waits for them to stop, PCDSTOP directs all lwps to stop without
1256 waiting for them to stop, and PCWSTOP simply waits for all lwps to stop.
1257 When applied to an lwp control file, PCSTOP directs the specific lwp to
1258 stop and waits until it has stopped, PCDSTOP directs the specific lwp to
1259 stop without waiting for it to stop, and PCWSTOP
1260 simply waits for the specific lwp to stop. When applied to an lwp
1261 control file, PCSTOP and PCWSTOP complete when the lwp stops on an event
1262 of interest, immediately if already so stopped; when applied to the
1263 process control file, they complete when every lwp has stopped either on
1264 an event of interest or on a PR_SUSPENDED stop.
1265
1266 PCTWSTOP is identical to PCWSTOP except that it enables the operation to
1267 time out, to avoid waiting forever for a process or lwp that may never
1268 stop on an event of interest. PCTWSTOP takes a long operand specifying a
1269 number of milliseconds; the wait will terminate successfully after the
1270 specified number of milliseconds even if the process or lwp has not
1271 stopped; a timeout value of zero makes the operation identical to
1272 PCWSTOP.
1273
1274 An "event of interest" is either a PR_REQUESTED stop or a stop that has
1275 been specified in the process's tracing flags (set by PCSTRACE, PCSFAULT,
1276 PCSENTRY, and PCSEXIT). PR_JOBCONTROL
1277 and PR_SUSPENDED stops are specifically not events of interest. (An lwp
1278 may stop twice due to a stop signal, first showing PR_SIGNALLED if the
1279 signal is traced and again showing PR_JOBCONTROL if the lwp is set
1280 running without clearing the signal.) If PCSTOP or PCDSTOP is applied to
1281 an lwp that is stopped, but not on an event of interest, the stop
1282 directive takes effect when the lwp is restarted by the competing
1283 mechanism. At that time, the lwp enters a PR_REQUESTED stop before
1284 executing any user-level code.
1285
1286 A write of a control message that blocks is interruptible by a signal so
1287 that, for example, an alarm(2) can be set to avoid waiting forever for a
1288 process or lwp that may never stop on an event of interest. If PCSTOP is
1289 interrupted, the lwp stop directives remain in effect even though the
1290 write(2) returns an error. (Use of PCTWSTOP with a non-zero timeout is
1291 recommended over PCWSTOP with an alarm(2).)
1292
1293 A system process (indicated by the PR_ISSYS flag) never executes at user
1294 level, has no user-level address space visible through /proc, and cannot
1295 be stopped. Applying one of these operations to a system process or any
1296 of its lwps elicits the error EBUSY.
1297
1298 PCRUN
1299 Make an lwp runnable again after a stop. This operation takes a long
1300 operand containing zero or more of the following flags:
1301
1302 PRCSIG clears the current signal, if any (see PCCSIG).
1303
1304 PRCFAULT clears the current fault, if any (see PCCFAULT).
1305
1306 PRSTEP directs the lwp to execute a single machine instruction.
1307 On completion of the instruction, a trace trap occurs. If
1308 FLTTRACE is being traced, the lwp stops; otherwise, it is
1309 sent SIGTRAP. If SIGTRAP is being traced and is not
1310 blocked, the lwp stops. When the lwp stops on an event of
1311 interest, the single-step directive is cancelled, even if
1312 the stop occurs before the instruction is executed. This
1313 operation requires hardware and operating system support
1314 and may not be implemented on all processors. It is
1315 implemented on SPARC and x86-based machines.
1316
1317 PRSABORT is meaningful only if the lwp is in a PR_SYSENTRY stop or
1318 is marked PR_ASLEEP; it instructs the lwp to abort
1319 execution of the system call (see PCSENTRY and PCSEXIT).
1320
1321 PRSTOP directs the lwp to stop again as soon as possible after
1322 resuming execution (see PCDSTOP). In particular, if the
1323 lwp is stopped on PR_SIGNALLED or PR_FAULTED, the next stop
1324 will show PR_REQUESTED, no other stop will have intervened,
1325 and the lwp will not have executed any user-level code.
1326
1327 When applied to an lwp control file, PCRUN clears any outstanding
1328 directed-stop request and makes the specific lwp runnable. The operation
1329 fails with EBUSY if the specific lwp is not stopped on an event of
1330 interest or has not been directed to stop or if the agent lwp exists and
1331 this is not the agent lwp (see PCAGENT).
1332
1333 When applied to the process control file, a representative lwp is chosen
1334 for the operation as described for /proc/pid/status. The operation fails
1335 with EBUSY if the representative lwp is not stopped on an event of
1336 interest or has not been directed to stop or if the agent lwp exists. If
1337 PRSTEP or PRSTOP was requested, the representative lwp is made runnable
1338 and its outstanding directed-stop request is cleared; otherwise all
1339 outstanding directed-stop requests are cleared and, if it was stopped on
1340 an event of interest, the representative lwp is marked PR_REQUESTED. If,
1341 as a consequence, all lwps are in the PR_REQUESTED or PR_SUSPENDED stop
1342 state, all lwps showing PR_REQUESTED are made runnable.
1343
1344 PCSTRACE
1345 Define a set of signals to be traced in the process. The receipt of one
1346 of these signals by an lwp causes the lwp to stop. The set of signals is
1347 defined using an operand sigset_t contained in the control message.
1348 Receipt of SIGKILL cannot be traced; if specified, it is silently
1349 ignored.
1350
1351 If a signal that is included in an lwp's held signal set (the signal
1352 mask) is sent to the lwp, the signal is not received and does not cause a
1353 stop until it is removed from the held signal set, either by the lwp
1354 itself or by setting the held signal set with PCSHOLD.
1355
1356 PCCSIG
1357 The current signal, if any, is cleared from the specific or
1358 representative lwp.
1359
1360 PCSSIG
1361 The current signal and its associated signal information for the specific
1362 or representative lwp are set according to the contents of the operand
1363 siginfo structure (see <sys/siginfo.h>). If the specified signal number
1364 is zero, the current signal is cleared. The semantics of this operation
1365 are different from those of kill(2) in that the signal is delivered to
1366 the lwp immediately after execution is resumed (even if it is being
1367 blocked) and an additional PR_SIGNALLED stop does not intervene even if
1368 the signal is traced. Setting the current signal to SIGKILL terminates
1369 the process immediately.
1370
1371 PCKILL
1372 If applied to the process control file, a signal is sent to the process
1373 with semantics identical to those of kill(2) If applied to an lwp control
1374 file, a directed signal is sent to the specific lwp. The signal is named
1375 in a long operand contained in the message. Sending SIGKILL terminates
1376 the process immediately.
1377
1378 PCUNKILL
1379 A signal is deleted, that is, it is removed from the set of pending
1380 signals. If applied to the process control file, the signal is deleted
1381 from the process's pending signals. If applied to an lwp control file,
1382 the signal is deleted from the lwp's pending signals. The current signal
1383 (if any) is unaffected. The signal is named in a long operand in the
1384 control message. It is an error (EINVAL) to attempt to delete SIGKILL.
1385
1386 PCSHOLD
1387 Set the set of held signals for the specific or representative lwp
1388 (signals whose delivery will be blocked if sent to the lwp). The set of
1389 signals is specified with a sigset_t operand. SIGKILL and SIGSTOP cannot
1390 be held; if specified, they are silently ignored.
1391
1392 PCSFAULT
1393 Define a set of hardware faults to be traced in the process. On
1394 incurring one of these faults, an lwp stops. The set is defined via the
1395 operand fltset_t structure. Fault names are defined in <sys/fault.h> and
1396 include the following. Some of these may not occur on all processors;
1397 there may be processor-specific faults in addition to these.
1398
1399 FLTILL illegal instruction
1400
1401 FLTPRIV privileged instruction
1402
1403 FLTBPT breakpoint trap
1404
1405 FLTTRACE trace trap (single-step)
1406
1407 FLTWATCH watchpoint trap
1408
1409 FLTACCESS memory access fault (bus error)
1410
1411 FLTBOUNDS memory bounds violation
1412
1413 FLTIOVF integer overflow
1414
1415 FLTIZDIV integer zero divide
1416
1417 FLTFPE floating-point exception
1418
1419 FLTSTACK unrecoverable stack fault
1420
1421 FLTPAGE recoverable page fault
1422
1423 When not traced, a fault normally results in the posting of a signal to
1424 the lwp that incurred the fault. If an lwp stops on a fault, the signal
1425 is posted to the lwp when execution is resumed unless the fault is
1426 cleared by PCCFAULT or by the PRCFAULT option of PCRUN. FLTPAGE is an
1427 exception; no signal is posted. The pr_info field in the lwpstatus
1428 structure identifies the signal to be sent and contains machine-specific
1429 information about the fault.
1430
1431 PCCFAULT
1432 The current fault, if any, is cleared; the associated signal will not be
1433 sent to the specific or representative lwp.
1434
1435 PCSENTRY PCSEXIT
1436 These control operations instruct the process's lwps to stop on entry to
1437 or exit from specified system calls. The set of system calls to be
1438 traced is defined via an operand sysset_t structure.
1439
1440 When entry to a system call is being traced, an lwp stops after having
1441 begun the call to the system but before the system call arguments have
1442 been fetched from the lwp. When exit from a system call is being traced,
1443 an lwp stops on completion of the system call just prior to checking for
1444 signals and returning to user level. At this point, all return values
1445 have been stored into the lwp's registers.
1446
1447 If an lwp is stopped on entry to a system call (PR_SYSENTRY) or when
1448 sleeping in an interruptible system call (PR_ASLEEP is set), it may be
1449 instructed to go directly to system call exit by specifying the PRSABORT
1450 flag in a PCRUN control message. Unless exit from the system call is
1451 being traced, the lwp returns to user level showing EINTR.
1452
1453 PCWATCH
1454 Set or clear a watched area in the controlled process from a prwatch
1455 structure operand:
1456
1457 typedef struct prwatch {
1458 uintptr_t pr_vaddr; /* virtual address of watched area */
1459 size_t pr_size; /* size of watched area in bytes */
1460 int pr_wflags; /* watch type flags */
1461 } prwatch_t;
1462
1463 pr_vaddr specifies the virtual address of an area of memory to be watched
1464 in the controlled process. pr_size specifies the size of the area, in
1465 bytes. pr_wflags specifies the type of memory access to be monitored as
1466 a bit-mask of the following flags:
1467
1468 WA_READ read access
1469
1470 WA_WRITE write access
1471
1472 WA_EXEC execution access
1473
1474 WA_TRAPAFTER trap after the instruction completes
1475
1476 If pr_wflags is non-empty, a watched area is established for the virtual
1477 address range specified by pr_vaddr and pr_size. If pr_wflags is empty,
1478 any previously-established watched area starting at the specified virtual
1479 address is cleared; pr_size is ignored.
1480
1481 A watchpoint is triggered when an lwp in the traced process makes a
1482 memory reference that covers at least one byte of a watched area and the
1483 memory reference is as specified in pr_wflags. When an lwp triggers a
1484 watchpoint, it incurs a watchpoint trap. If FLTWATCH is being traced,
1485 the lwp stops; otherwise, it is sent a SIGTRAP signal; if SIGTRAP is
1486 being traced and is not blocked, the lwp stops.
1487
1488 The watchpoint trap occurs before the instruction completes unless
1489 WA_TRAPAFTER was specified, in which case it occurs after the instruction
1490 completes. If it occurs before completion, the memory is not modified.
1491 If it occurs after completion, the memory is modified (if the access is a
1492 write access).
1493
1494 Physical i/o is an exception for watchpoint traps. In this instance,
1495 there is no guarantee that memory before the watched area has already
1496 been modified (or in the case of WA_TRAPAFTER, that the memory following
1497 the watched area has not been modified) when the watchpoint trap occurs
1498 and the lwp stops.
1499
1500 pr_info in the lwpstatus structure contains information pertinent to the
1501 watchpoint trap. In particular, the si_addr field contains the virtual
1502 address of the memory reference that triggered the watchpoint, and the
1503 si_code field contains one of TRAP_RWATCH, TRAP_WWATCH, or TRAP_XWATCH,
1504 indicating read, write, or execute access, respectively. The
1505 si_trapafter field is zero unless WA_TRAPAFTER is in effect for this
1506 watched area; non-zero indicates that the current instruction is not the
1507 instruction that incurred the watchpoint trap. The si_pc field contains
1508 the virtual address of the instruction that incurred the trap.
1509
1510 A watchpoint trap may be triggered while executing a system call that
1511 makes reference to the traced process's memory. The lwp that is
1512 executing the system call incurs the watchpoint trap while still in the
1513 system call. If it stops as a result, the lwpstatus structure contains
1514 the system call number and its arguments. If the lwp does not stop, or
1515 if it is set running again without clearing the signal or fault, the
1516 system call fails with EFAULT. If WA_TRAPAFTER was specified, the memory
1517 reference will have completed and the memory will have been modified (if
1518 the access was a write access) when the watchpoint trap occurs.
1519
1520 If more than one of WA_READ, WA_WRITE, and WA_EXEC is specified for a
1521 watched area, and a single instruction incurs more than one of the
1522 specified types, only one is reported when the watchpoint trap occurs.
1523 The precedence is WA_EXEC, WA_READ, WA_WRITE (WA_EXEC and WA_READ take
1524 precedence over WA_WRITE), unless WA_TRAPAFTER was specified, in which
1525 case it is WA_WRITE, WA_READ, WA_EXEC (WA_WRITE takes precedence).
1526
1527 PCWATCH fails with EINVAL if an attempt is made to specify overlapping
1528 watched areas or if pr_wflags contains flags other than those specified
1529 above. It fails with ENOMEM if an attempt is made to establish more
1530 watched areas than the system can support (the system can support
1531 thousands).
1532
1533 The child of a vfork(2) borrows the parent's address space. When a
1534 vfork(2) is executed by a traced process, all watched areas established
1535 for the parent are suspended until the child terminates or performs an
1536 exec(2). Any watched areas established independently in the child are
1537 cancelled when the parent resumes after the child's termination or
1538 exec(2). PCWATCH fails with EBUSY if applied to the parent of a vfork(2)
1539 before the child has terminated or performed an exec(2). The PR_VFORKP
1540 flag is set in the pstatus structure for such a parent process.
1541
1542 Certain accesses of the traced process's address space by the operating
1543 system are immune to watchpoints. The initial construction of a signal
1544 stack frame when a signal is delivered to an lwp will not trigger a
1545 watchpoint trap even if the new frame covers watched areas of the stack.
1546 Once the signal handler is entered, watchpoint traps occur normally. On
1547 SPARC based machines, register window overflow and underflow will not
1548 trigger watchpoint traps, even if the register window save areas cover
1549 watched areas of the stack.
1550
1551 Watched areas are not inherited by child processes, even if the traced
1552 process's inherit-on-fork mode, PR_FORK, is set (see PCSET, below). All
1553 watched areas are cancelled when the traced process performs a successful
1554 exec(2).
1555
1556 PCSET PCUNSET
1557 PCSET sets one or more modes of operation for the traced process.
1558 PCUNSET unsets these modes. The modes to be set or unset are specified
1559 by flags in an operand long in the control message:
1560
1561 PR_FORK (inherit-on-fork): When set, the process's tracing flags
1562 and its inherit-on-fork mode are inherited by the child of
1563 a fork(2), fork1(2), or vfork(2). When unset, child
1564 processes start with all tracing flags cleared.
1565
1566 PR_RLC (run-on-last-close): When set and the last writable /proc
1567 file descriptor referring to the traced process or any of
1568 its lwps is closed, all of the process's tracing flags and
1569 watched areas are cleared, any outstanding stop directives
1570 are canceled, and if any lwps are stopped on events of
1571 interest, they are set running as though PCRUN had been
1572 applied to them. When unset, the process's tracing flags
1573 and watched areas are retained and lwps are not set
1574 running on last close.
1575
1576 PR_KLC (kill-on-last-close): When set and the last writable /proc
1577 file descriptor referring to the traced process or any of
1578 its lwps is closed, the process is terminated with
1579 SIGKILL.
1580
1581 PR_ASYNC (asynchronous-stop): When set, a stop on an event of
1582 interest by one lwp does not directly affect any other lwp
1583 in the process. When unset and an lwp stops on an event
1584 of interest other than PR_REQUESTED, all other lwps in the
1585 process are directed to stop.
1586
1587 PR_MSACCT (microstate accounting): Microstate accounting is now
1588 continuously enabled. This flag is deprecated and no
1589 longer has any effect upon microstate accounting.
1590 Applications may toggle this flag; however, microstate
1591 accounting will remain enabled regardless.
1592
1593 PR_MSFORK (inherit microstate accounting): All processes now inherit
1594 microstate accounting, as it is continuously enabled.
1595 This flag has been deprecated and its use no longer has
1596 any effect upon the behavior of microstate accounting.
1597
1598 PR_BPTADJ (breakpoint trap pc adjustment): On x86-based machines, a
1599 breakpoint trap leaves the program counter (the EIP)
1600 referring to the breakpointed instruction plus one byte.
1601 When PR_BPTADJ is set, the system will adjust the program
1602 counter back to the location of the breakpointed
1603 instruction when the lwp stops on a breakpoint. This flag
1604 has no effect on SPARC based machines, where breakpoint
1605 traps leave the program counter referring to the
1606 breakpointed instruction.
1607
1608 PR_PTRACE (ptrace-compatibility): When set, a stop on an event of
1609 interest by the traced process is reported to the parent
1610 of the traced process by wait(3C), SIGTRAP is sent to the
1611 traced process when it executes a successful exec(2),
1612 setuid/setgid flags are not honored for execs performed by
1613 the traced process, any exec of an object file that the
1614 traced process cannot read fails, and the process dies
1615 when its parent dies. This mode is deprecated; it is
1616 provided only to allow ptrace(3C) to be implemented as a
1617 library function using /proc.
1618
1619 It is an error (EINVAL) to specify flags other than those described above
1620 or to apply these operations to a system process. The current modes are
1621 reported in the pr_flags field of /proc/pid/status and
1622 /proc/pid/lwp/lwp/lwpstatus.
1623
1624 PCSREG
1625 Set the general registers for the specific or representative lwp
1626 according to the operand prgregset_t structure.
1627
1628 On SPARC based systems, only the condition-code bits of the processor-
1629 status register (R_PSR) of SPARC V8 (32-bit) processes can be modified by
1630 PCSREG. Other privileged registers cannot be modified at all.
1631
1632 On x86-based systems, only certain bits of the flags register (EFL) can
1633 be modified by PCSREG: these include the condition codes, direction-bit,
1634 and overflow-bit.
1635
1636 PCSREG fails with EBUSY if the lwp is not stopped on an event of
1637 interest.
1638
1639 PCSVADDR
1640 Set the address at which execution will resume for the specific or
1641 representative lwp from the operand long. On SPARC based systems, both
1642 %pc and %npc are set, with %npc set to the instruction following the
1643 virtual address. On x86-based systems, only %eip is set. PCSVADDR fails
1644 with EBUSY if the lwp is not stopped on an event of interest.
1645
1646 PCSFPREG
1647 Set the floating-point registers for the specific or representative lwp
1648 according to the operand prfpregset_t structure. An error (EINVAL) is
1649 returned if the system does not support floating-point operations (no
1650 floating-point hardware and the system does not emulate floating-point
1651 machine instructions). PCSFPREG fails with EBUSY if the lwp is not
1652 stopped on an event of interest.
1653
1654 PCSXREG
1655 Set the extra state registers for the specific or representative lwp
1656 according to the architecture-dependent operand prxregset_t structure.
1657 An error (EINVAL) is returned if the system does not support extra state
1658 registers. PCSXREG fails with EBUSY if the lwp is not stopped on an
1659 event of interest.
1660
1661 PCSASRS
1662 Set the ancillary state registers for the specific or representative lwp
1663 according to the SPARC V9 platform-dependent operand asrset_t structure.
1664 An error (EINVAL) is returned if either the target process or the
1665 controlling process is not a 64-bit SPARC V9 process. Most of the
1666 ancillary state registers are privileged registers that cannot be
1667 modified. Only those that can be modified are set; all others are
1668 silently ignored. PCSASRS fails with EBUSY if the lwp is not stopped on
1669 an event of interest.
1670
1671 PCAGENT
1672 Create an agent lwp in the controlled process with register values from
1673 the operand prgregset_t structure (see PCSREG, above). The agent lwp is
1674 created in the stopped state showing PR_REQUESTED and with its held
1675 signal set (the signal mask) having all signals except SIGKILL and
1676 SIGSTOP blocked.
1677
1678 The PCAGENT operation fails with EBUSY unless the process is fully
1679 stopped via /proc, that is, unless all of the lwps in the process are
1680 stopped either on events of interest or on PR_SUSPENDED, or are stopped
1681 on PR_JOBCONTROL and have been directed to stop via PCDSTOP. It fails
1682 with EBUSY if an agent lwp already exists. It fails with ENOMEM if
1683 system resources for creating new lwps have been exhausted.
1684
1685 Any PCRUN operation applied to the process control file or to the control
1686 file of an lwp other than the agent lwp fails with EBUSY as long as the
1687 agent lwp exists. The agent lwp must be caused to terminate by executing
1688 the SYS_lwp_exit system call trap before the process can be restarted.
1689
1690 Once the agent lwp is created, its lwp-ID can be found by reading the
1691 process status file. To facilitate opening the agent lwp's control and
1692 status files, the directory name /proc/pid/lwp/agent is accepted for
1693 lookup operations as an invisible alias for /proc/pid/lwp/lwpid, lwpid
1694 being the lwp-ID of the agent lwp (invisible in the sense that the name
1695 "agent" does not appear in a directory listing of /proc/pid/lwp obtained
1696 from ls(1), getdents(2), or readdir(3C).
1697
1698 The purpose of the agent lwp is to perform operations in the controlled
1699 process on behalf of the controlling process: to gather information not
1700 directly available via /proc files, or in general to make the process
1701 change state in ways not directly available via /proc control operations.
1702 To make use of an agent lwp, the controlling process must be capable of
1703 making it execute system calls (specifically, the SYS_lwp_exit system
1704 call trap). The register values given to the agent lwp on creation are
1705 typically the registers of the representative lwp, so that the agent lwp
1706 can use its stack.
1707
1708 If the controlling process neglects to force the agent lwp to execute the
1709 SYS_lwp_exit system call (due to either logic error or fatal failure on
1710 the part of the controlling process), the agent lwp will remain in the
1711 target process. For purposes of being able to debug these otherwise
1712 rogue agents, information as to the creator of the agent lwp is reflected
1713 in that lwp's spymaster file in /proc. Should the target process
1714 generate a core dump with the agent lwp in place, this information will
1715 be available via the NT_SPYMASTER note in the core file (see core(4)).
1716
1717 The agent lwp is not allowed to execute any variation of the SYS_fork or
1718 SYS_exec system call traps. Attempts to do so yield ENOTSUP to the agent
1719 lwp.
1720
1721 Symbolic constants for system call trap numbers like SYS_lwp_exit and
1722 SYS_lwp_create can be found in the header file <sys/syscall.h>.
1723
1724 PCREAD PCWRITE
1725 Read or write the target process's address space via a priovec structure
1726 operand:
1727
1728 typedef struct priovec {
1729 void *pio_base; /* buffer in controlling process */
1730 size_t pio_len; /* size of read/write request in bytes */
1731 off_t pio_offset; /* virtual address in target process */
1732 } priovec_t;
1733
1734 These operations have the same effect as pread(2) and pwrite(2),
1735 respectively, of the target process's address space file. The difference
1736 is that more than one PCREAD or PCWRITE control operation can be written
1737 to the control file at once, and they can be interspersed with other
1738 control operations in a single write to the control file. This is
1739 useful, for example, when planting many breakpoint instructions in the
1740 process's address space, or when stepping over a breakpointed
1741 instruction. Unlike pread(2) and pwrite(2), no provision is made for
1742 partial reads or writes; if the operation cannot be performed completely,
1743 it fails with EIO.
1744
1745 PCNICE
1746 The traced process's nice(2) value is incremented by the amount in the
1747 operand long. Only a process with the {PRIV_PROC_PRIOCNTL} privilege
1748 asserted in its effective set can better a process's priority in this
1749 way, but any user may lower the priority. This operation is not
1750 meaningful for all scheduling classes.
1751
1752 PCSCRED
1753 Set the target process credentials to the values contained in the
1754 prcred_t structure operand (see /proc/pid/cred). The effective, real,
1755 and saved user-IDs and group-IDs of the target process are set. The
1756 target process's supplementary groups are not changed; the pr_ngroups and
1757 pr_groups members of the structure operand are ignored. Only the
1758 privileged processes can perform this operation; for all others it fails
1759 with EPERM.
1760
1761 PCSCREDX
1762 Operates like PCSCRED but also sets the supplementary groups; the length
1763 of the data written with this control operation should be "sizeof
1764 (prcred_t) + sizeof (gid_t) * (#groups - 1)".
1765
1766 PCSPRIV
1767 Set the target process privilege to the values contained in the prpriv_t
1768 operand (see /proc/pid/priv). The effective, permitted, inheritable, and
1769 limit sets are all changed. Privilege flags can also be set. The
1770 process is made privilege aware unless it can relinquish privilege
1771 awareness. See privileges(5).
1772
1773 The limit set of the target process cannot be grown. The other privilege
1774 sets must be subsets of the intersection of the effective set of the
1775 calling process with the new limit set of the target process or subsets
1776 of the original values of the sets in the target process.
1777
1778 If any of the above restrictions are not met, EPERM is returned. If the
1779 structure written is improperly formatted, EINVAL is returned.
1780
1781 PROGRAMMING NOTES
1782 For security reasons, except for the psinfo, usage, lpsinfo, lusage,
1783 lwpsinfo, and lwpusage files, which are world-readable, and except for
1784 privileged processes, an open of a /proc file fails unless both the user-
1785 ID and group-ID of the caller match those of the traced process and the
1786 process's object file is readable by the caller. The effective set of
1787 the caller is a superset of both the inheritable and the permitted set of
1788 the target process. The limit set of the caller is a superset of the
1789 limit set of the target process. Except for the world-readable files
1790 just mentioned, files corresponding to setuid and setgid processes can be
1791 opened only by the appropriately privileged process.
1792
1793 A process that is missing the basic privilege {PRIV_PROC_INFO} cannot see
1794 any processes under /proc that it cannot send a signal to.
1795
1796 A process that has {PRIV_PROC_OWNER} asserted in its effective set can
1797 open any file for reading. To manipulate or control a process, the
1798 controlling process must have at least as many privileges in its
1799 effective set as the target process has in its effective, inheritable,
1800 and permitted sets. The limit set of the controlling process must be a
1801 superset of the limit set of the target process. Additional restrictions
1802 apply if any of the uids of the target process are 0. See privileges(5).
1803
1804 Even if held by a privileged process, an open process or lwp file
1805 descriptor (other than file descriptors for the world-readable files)
1806 becomes invalid if the traced process performs an exec(2) of a
1807 setuid/setgid object file or an object file that the traced process
1808 cannot read. Any operation performed on an invalid file descriptor,
1809 except close(2), fails with EAGAIN. In this situation, if any tracing
1810 flags are set and the process or any lwp file descriptor is open for
1811 writing, the process will have been directed to stop and its run-on-last-
1812 close flag will have been set (see PCSET). This enables a controlling
1813 process (if it has permission) to reopen the /proc files to get new valid
1814 file descriptors, close the invalid file descriptors, unset the run-on-
1815 last-close flag (if desired), and proceed. Just closing the invalid file
1816 descriptors causes the traced process to resume execution with all
1817 tracing flags cleared. Any process not currently open for writing via
1818 /proc, but that has left-over tracing flags from a previous open, and
1819 that executes a setuid/setgid or unreadable object file, will not be
1820 stopped but will have all its tracing flags cleared.
1821
1822 To wait for one or more of a set of processes or lwps to stop or
1823 terminate, /proc file descriptors (other than those obtained by opening
1824 the cwd or root directories or by opening files in the fd or object
1825 directories) can be used in a poll(2) system call. When requested and
1826 returned, either of the polling events POLLPRI or POLLWRNORM indicates
1827 that the process or lwp stopped on an event of interest. Although they
1828 cannot be requested, the polling events POLLHUP, POLLERR, and POLLNVAL
1829 may be returned. POLLHUP indicates that the process or lwp has
1830 terminated. POLLERR indicates that the file descriptor has become
1831 invalid. POLLNVAL is returned immediately if POLLPRI or POLLWRNORM is
1832 requested on a file descriptor referring to a system process (see
1833 PCSTOP). The requested events may be empty to wait simply for
1834 termination.
1835
1836 FILES
1837 /proc directory (list of processes)
1838 /proc/pid
1839 specific process directory
1840 /proc/self
1841 alias for a process's own directory
1842 /proc/pid/as
1843 address space file
1844 /proc/pid/ctl
1845 process control file
1846 /proc/pid/status
1847 process status
1848 /proc/pid/lstatus
1849 array of lwp status structs
1850 /proc/pid/psinfo
1851 process ps(1) info
1852 /proc/pid/lpsinfo
1853 array of lwp ps(1) info structs
1854 /proc/pid/map
1855 address space map
1856 /proc/pid/xmap
1857 extended address space map
1858 /proc/pid/rmap
1859 reserved address map
1860 /proc/pid/cred
1861 process credentials
1862 /proc/pid/priv
1863 process privileges
1864 /proc/pid/sigact
1865 process signal actions
1866 /proc/pid/auxv
1867 process aux vector
1868 /proc/pid/ldt
1869 process LDT (x86 only)
1870 /proc/pid/usage
1871 process usage
1872 /proc/pid/lusage
1873 array of lwp usage structs
1874 /proc/pid/path
1875 symbolic links to process open files
1876 /proc/pid/pagedata
1877 process page data
1878 /proc/pid/watch
1879 active watchpoints
1880 /proc/pid/cwd
1881 alias for the current working directory
1882 /proc/pid/root
1883 alias for the root directory
1884 /proc/pid/fd
1885 directory (list of open files)
1886 /proc/pid/fd/*
1887 aliases for process's open files
1888 /proc/pid/object
1889 directory (list of mapped files)
1890 /proc/pid/object/a.out
1891 alias for process's executable file
1892 /proc/pid/object/*
1893 aliases for other mapped files
1894 /proc/pid/lwp
1895 directory (list of lwps)
1896 /proc/pid/lwp/lwpid
1897 specific lwp directory
1898 /proc/pid/lwp/agent
1899 alias for the agent lwp directory
1900 /proc/pid/lwp/lwpid/lwpctl
1901 lwp control file
1902 /proc/pid/lwp/lwpid/lwpstatus
1903 lwp status
1904 /proc/pid/lwp/lwpid/lwpsinfo
1905 lwp ps(1) info
1906 /proc/pid/lwp/lwpid/lwpusage
1907 lwp usage
1908 /proc/pid/lwp/lwpid/gwindows
1909 register windows (SPARC only)
1910 /proc/pid/lwp/lwpid/xregs
1911 extra state registers
1912 /proc/pid/lwp/lwpid/asrs
1913 ancillary state registers (SPARC V9 only)
1914 /proc/pid/lwp/lwpid/spymaster
1915 For an agent LWP, the controlling process
1916
1917 DIAGNOSTICS
1918 Errors that can occur in addition to the errors normally associated with
1919 file system access:
1920
1921 E2BIG Data to be returned in a read(2) of the page data file
1922 exceeds the size of the read buffer provided by the
1923 caller.
1924
1925 EACCES An attempt was made to examine a process that ran under a
1926 different uid than the controlling process and
1927 {PRIV_PROC_OWNER} was not asserted in the effective set.
1928
1929 EAGAIN The traced process has performed an exec(2) of a
1930 setuid/setgid object file or of an object file that it
1931 cannot read; all further operations on the process or lwp
1932 file descriptor (except close(2)) elicit this error.
1933
1934 EBUSY PCSTOP, PCDSTOP, PCWSTOP, or PCTWSTOP was applied to a
1935 system process; an exclusive open(2) was attempted on a
1936 /proc file for a process already open for writing; PCRUN,
1937 PCSREG, PCSVADDR, PCSFPREG, or PCSXREG was applied to a
1938 process or lwp not stopped on an event of interest; an
1939 attempt was made to mount /proc when it was already
1940 mounted; PCAGENT was applied to a process that was not
1941 fully stopped or that already had an agent lwp.
1942
1943 EINVAL In general, this means that some invalid argument was
1944 supplied to a system call. A non-exhaustive list of
1945 conditions eliciting this error includes: a control
1946 message operation code is undefined; an out-of-range
1947 signal number was specified with PCSSIG, PCKILL, or
1948 PCUNKILL; SIGKILL was specified with PCUNKILL; PCSFPREG
1949 was applied on a system that does not support floating-
1950 point operations; PCSXREG was applied on a system that
1951 does not support extra state registers.
1952
1953 EINTR A signal was received by the controlling process while
1954 waiting for the traced process or lwp to stop via PCSTOP,
1955 PCWSTOP, or PCTWSTOP.
1956
1957 EIO A write(2) was attempted at an illegal address in the
1958 traced process.
1959
1960 ENOENT The traced process or lwp has terminated after being
1961 opened. The basic privilege {PRIV_PROC_INFO} is not
1962 asserted in the effective set of the calling process and
1963 the calling process cannot send a signal to the target
1964 process.
1965
1966 ENOMEM The system-imposed limit on the number of page data file
1967 descriptors was reached on an open of /proc/pid/pagedata;
1968 an attempt was made with PCWATCH to establish more watched
1969 areas than the system can support; the PCAGENT operation
1970 was issued when the system was out of resources for
1971 creating lwps.
1972
1973 ENOSYS An attempt was made to perform an unsupported operation
1974 (such as creat(2), link(2), or unlink(2)) on an entry in
1975 /proc.
1976
1977 EOVERFLOW A 32-bit controlling process attempted to read or write
1978 the as file or attempted to read the map, rmap, or
1979 pagedata file of a 64-bit target process. A 32-bit
1980 controlling process attempted to apply one of the control
1981 operations PCSREG, PCSXREG, PCSVADDR, PCWATCH, PCAGENT,
1982 PCREAD, PCWRITE to a 64-bit target process.
1983
1984 EPERM The process that issued the PCSCRED or PCSCREDX operation
1985 did not have the {PRIV_PROC_SETID} privilege asserted in
1986 its effective set, or the process that issued the PCNICE
1987 operation did not have the {PRIV_PROC_PRIOCNTL} in its
1988 effective set.
1989
1990 An attempt was made to control a process of which the E,
1991 P, and I privilege sets were not a subset of the effective
1992 set of the controlling process or the limit set of the
1993 controlling process is not a superset of limit set of the
1994 controlled process.
1995
1996 Any of the uids of the target process are 0 or an attempt
1997 was made to change any of the uids to 0 using PCSCRED and
1998 the security policy imposed additional restrictions. See
1999 privileges(5).
2000
2001 SEE ALSO
2002 ls(1), ps(1), chroot(1M), alarm(2), brk(2), chdir(2), chroot(2),
2003 close(2), creat(2), dup(2), exec(2), fcntl(2), fork(2), fork1(2),
2004 fstat(2), getdents(2), getustack(2), kill(2), lseek(2), mmap(2), nice(2),
2005 open(2), poll(2), pread(2), pwrite(2), read(2), readlink(2), readv(2),
2006 shmget(2), sigaction(2), sigaltstack(2), vfork(2), write(2), writev(2),
2007 _stack_grow(3C), pthread_create(3C), pthread_join(3C), ptrace(3C),
2008 readdir(3C), thr_create(3C), thr_join(3C), wait(3C), siginfo.h(3HEAD),
2009 signal.h(3HEAD), types32.h(3HEAD), ucontext.h(3HEAD), contract(4),
2010 core(4), process(4), lfcompile(5), privileges(5), security-flags(5)
2011
2012 NOTES
2013 Descriptions of structures in this document include only interesting
2014 structure elements, not filler and padding fields, and may show elements
2015 out of order for descriptive clarity. The actual structure definitions
2016 are contained in <procfs.h>.
2017
2018 BUGS
2019 Because the old ioctl(2)-based version of /proc is currently supported
2020 for binary compatibility with old applications, the top-level directory
2021 for a process, /proc/pid, is not world-readable, but it is world-
2022 searchable. Thus, anyone can open /proc/pid/psinfo even though ls(1)
2023 applied to /proc/pid will fail for anyone but the owner or an
2024 appropriately privileged process. Support for the old ioctl(2)-based
2025 version of /proc will be dropped in a future release, at which time the
2026 top-level directory for a process will be made world-readable.
2027
2028 On SPARC based machines, the types gregset_t and fpregset_t defined in
2029 <sys/regset.h> are similar to but not the same as the types prgregset_t
2030 and prfpregset_t defined in <procfs.h>.
2031
2032 illumos May 17, 2020 illumos