2 * Copyright (c) 1986, 1989, 1991, 1993
3 * The Regents of the University of California. All rights reserved.
4 * (c) UNIX System Laboratories, Inc.
5 * All or some portions of this file are derived from material licensed
6 * to the University of California by American Telephone and Telegraph
7 * Co. or Unix System Laboratories, Inc. and are reproduced herein with
8 * the permission of UNIX System Laboratories, Inc.
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 4. Neither the name of the University nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * @(#)proc.h 8.15 (Berkeley) 5/19/95
41 #include <sys/callout.h> /* For struct callout. */
42 #include <sys/event.h> /* For struct klist. */
43 #include <sys/condvar.h>
45 #include <sys/filedesc.h>
47 #include <sys/queue.h>
48 #include <sys/_lock.h>
49 #include <sys/lock_profile.h>
50 #include <sys/_mutex.h>
52 #include <sys/priority.h>
53 #include <sys/rtprio.h> /* XXX. */
55 #include <sys/resource.h>
56 #include <sys/sigio.h>
57 #include <sys/signal.h>
58 #include <sys/signalvar.h>
60 #include <sys/time.h> /* For structs itimerval, timeval. */
64 #include <sys/ucontext.h>
65 #include <sys/ucred.h>
66 #include <machine/proc.h> /* Machine-dependent proc substruct. */
69 * One structure allocated per session.
72 * (m) locked by s_mtx mtx
73 * (e) locked by proctree_lock sx
74 * (c) const until freeing
77 u_int s_count; /* Ref cnt; pgrps in session - atomic. */
78 struct proc *s_leader; /* (m + e) Session leader. */
79 struct vnode *s_ttyvp; /* (m) Vnode of controlling tty. */
80 struct cdev_priv *s_ttydp; /* (m) Device of controlling tty. */
81 struct tty *s_ttyp; /* (e) Controlling tty. */
82 pid_t s_sid; /* (c) Session ID. */
83 /* (m) Setlogin() name: */
84 char s_login[roundup(MAXLOGNAME, sizeof(long))];
85 struct mtx s_mtx; /* Mutex to protect members. */
89 * One structure allocated per process group.
92 * (m) locked by pg_mtx mtx
93 * (e) locked by proctree_lock sx
94 * (c) const until freeing
97 LIST_ENTRY(pgrp) pg_hash; /* (e) Hash chain. */
98 LIST_HEAD(, proc) pg_members; /* (m + e) Pointer to pgrp members. */
99 struct session *pg_session; /* (c) Pointer to session. */
100 struct sigiolst pg_sigiolst; /* (m) List of sigio sources. */
101 pid_t pg_id; /* (c) Process group id. */
102 int pg_jobc; /* (m) Job control process count. */
103 struct mtx pg_mtx; /* Mutex to protect members */
107 * pargs, used to hold a copy of the command line, if it had a sane length.
110 u_int ar_ref; /* Reference count. */
111 u_int ar_length; /* Length. */
112 u_char ar_args[1]; /* Arguments. */
116 * Description of a process.
118 * This structure contains the information needed to manage a thread of
119 * control, known in UN*X as a process; it has references to substructures
120 * containing descriptions of things that the process uses, but may share
121 * with related processes. The process structure and the substructures
122 * are always addressable except for those marked "(CPU)" below,
123 * which might be addressable only on a processor on which the process
126 * Below is a key of locks used to protect each member of struct proc. The
127 * lock is indicated by a reference to a specific character in parens in the
128 * associated comment.
129 * * - not yet protected
130 * a - only touched by curproc or parent during fork/wait
131 * b - created at fork, never changes
132 * (exception aiods switch vmspaces, but they are also
133 * marked 'P_SYSTEM' so hopefully it will be left alone)
134 * c - locked by proc mtx
135 * d - locked by allproc_lock lock
136 * e - locked by proctree_lock lock
138 * g - process group mtx
139 * h - callout_lock mtx
140 * i - by curproc or the master session mtx
141 * j - locked by proc slock
142 * k - only accessed by curthread
143 * k*- only accessed by curthread and from an interrupt
144 * l - the attaching proc or attaching proc parent
146 * n - not locked, lazy
148 * q - td_contested lock
151 * x - created at fork, only changes during single threading in exec
152 * y - created at first aio, doesn't change until exit or exec at which
153 * point we are single-threaded and only curthread changes it
154 * z - zombie threads lock
156 * If the locking key specifies two identifiers (for example, p_pptr) then
157 * either lock is sufficient for read access, but both locks must be held
162 struct kaudit_record;
164 struct kdtrace_thread;
165 struct mqueue_notifier;
179 * XXX: Does this belong in resource.h or resourcevar.h instead?
180 * Resource usage extension. The times in rusage structs in the kernel are
181 * never up to date. The actual times are kept as runtimes and tick counts
182 * (with control info in the "previous" times), and are converted when
183 * userland asks for rusage info. Backwards compatibility prevents putting
184 * this directly in the user-visible rusage struct.
186 * Locking for p_rux: (cj) means (j) for p_rux and (c) for p_crux.
187 * Locking for td_rux: (t) for all fields.
190 uint64_t rux_runtime; /* (cj) Real time. */
191 uint64_t rux_uticks; /* (cj) Statclock hits in user mode. */
192 uint64_t rux_sticks; /* (cj) Statclock hits in sys mode. */
193 uint64_t rux_iticks; /* (cj) Statclock hits in intr mode. */
194 uint64_t rux_uu; /* (c) Previous user time in usec. */
195 uint64_t rux_su; /* (c) Previous sys time in usec. */
196 uint64_t rux_tu; /* (c) Previous total time in usec. */
200 * Kernel runnable context (thread).
201 * This is what is put to sleep and reactivated.
202 * Thread context. Processes may have multiple threads.
205 struct mtx *volatile td_lock; /* replaces sched lock */
206 struct proc *td_proc; /* (*) Associated process. */
207 TAILQ_ENTRY(thread) td_plist; /* (*) All threads in this proc. */
208 TAILQ_ENTRY(thread) td_runq; /* (t) Run queue. */
209 TAILQ_ENTRY(thread) td_slpq; /* (t) Sleep queue. */
210 TAILQ_ENTRY(thread) td_lockq; /* (t) Lock queue. */
211 LIST_ENTRY(thread) td_hash; /* (d) Hash chain. */
212 struct cpuset *td_cpuset; /* (t) CPU affinity mask. */
213 struct seltd *td_sel; /* Select queue/channel. */
214 struct sleepqueue *td_sleepqueue; /* (k) Associated sleep queue. */
215 struct turnstile *td_turnstile; /* (k) Associated turnstile. */
216 struct umtx_q *td_umtxq; /* (c?) Link for when we're blocked. */
217 lwpid_t td_tid; /* (b) Thread ID. */
218 sigqueue_t td_sigqueue; /* (c) Sigs arrived, not delivered. */
219 #define td_siglist td_sigqueue.sq_signals
220 u_char td_lend_user_pri; /* (t) Lend user pri. */
222 /* Cleared during fork1() */
223 #define td_startzero td_flags
224 int td_flags; /* (t) TDF_* flags. */
225 int td_inhibitors; /* (t) Why can not run. */
226 int td_pflags; /* (k) Private thread (TDP_*) flags. */
227 int td_dupfd; /* (k) Ret value from fdopen. XXX */
228 int td_sqqueue; /* (t) Sleepqueue queue blocked on. */
229 void *td_wchan; /* (t) Sleep address. */
230 const char *td_wmesg; /* (t) Reason for sleep. */
231 u_char td_lastcpu; /* (t) Last cpu we were on. */
232 u_char td_oncpu; /* (t) Which cpu we are on. */
233 volatile u_char td_owepreempt; /* (k*) Preempt on last critical_exit */
234 u_char td_tsqueue; /* (t) Turnstile queue blocked on. */
235 short td_locks; /* (k) Count of non-spin locks. */
236 short td_rw_rlocks; /* (k) Count of rwlock read locks. */
237 short td_lk_slocks; /* (k) Count of lockmgr shared locks. */
238 short td_stopsched; /* (k) Scheduler stopped. */
239 struct turnstile *td_blocked; /* (t) Lock thread is blocked on. */
240 const char *td_lockname; /* (t) Name of lock blocked on. */
241 LIST_HEAD(, turnstile) td_contested; /* (q) Contested locks. */
242 struct lock_list_entry *td_sleeplocks; /* (k) Held sleep locks. */
243 int td_intr_nesting_level; /* (k) Interrupt recursion. */
244 int td_pinned; /* (k) Temporary cpu pin count. */
245 struct ucred *td_ucred; /* (k) Reference to credentials. */
246 u_int td_estcpu; /* (t) estimated cpu utilization */
247 int td_slptick; /* (t) Time at sleep. */
248 int td_blktick; /* (t) Time spent blocked. */
249 int td_swvoltick; /* (t) Time at last SW_VOL switch. */
250 u_int td_cow; /* (*) Number of copy-on-write faults */
251 struct rusage td_ru; /* (t) rusage information. */
252 struct rusage_ext td_rux; /* (t) Internal rusage information. */
253 uint64_t td_incruntime; /* (t) Cpu ticks to transfer to proc. */
254 uint64_t td_runtime; /* (t) How many cpu ticks we've run. */
255 u_int td_pticks; /* (t) Statclock hits for profiling */
256 u_int td_sticks; /* (t) Statclock hits in system mode. */
257 u_int td_iticks; /* (t) Statclock hits in intr mode. */
258 u_int td_uticks; /* (t) Statclock hits in user mode. */
259 int td_intrval; /* (t) Return value for sleepq. */
260 sigset_t td_oldsigmask; /* (k) Saved mask from pre sigpause. */
261 sigset_t td_sigmask; /* (c) Current signal mask. */
262 volatile u_int td_generation; /* (k) For detection of preemption */
263 stack_t td_sigstk; /* (k) Stack ptr and on-stack flag. */
264 int td_xsig; /* (c) Signal for ptrace */
265 u_long td_profil_addr; /* (k) Temporary addr until AST. */
266 u_int td_profil_ticks; /* (k) Temporary ticks until AST. */
267 char td_name[MAXCOMLEN + 1]; /* (*) Thread name. */
268 struct file *td_fpop; /* (k) file referencing cdev under op */
269 int td_dbgflags; /* (c) Userland debugger flags */
270 struct ksiginfo td_dbgksi; /* (c) ksi reflected to debugger. */
271 int td_ng_outbound; /* (k) Thread entered ng from above. */
272 struct osd td_osd; /* (k) Object specific data. */
273 struct vm_map_entry *td_map_def_user; /* (k) Deferred entries. */
274 pid_t td_dbg_forked; /* (c) Child pid for debugger. */
275 #define td_endzero td_rqindex
277 /* Copied during fork1() or thread_sched_upcall(). */
278 #define td_startcopy td_endzero
279 u_char td_rqindex; /* (t) Run queue index. */
280 u_char td_base_pri; /* (t) Thread base kernel priority. */
281 u_char td_priority; /* (t) Thread active priority. */
282 u_char td_pri_class; /* (t) Scheduling class. */
283 u_char td_user_pri; /* (t) User pri from estcpu and nice. */
284 u_char td_base_user_pri; /* (t) Base user pri */
285 #define td_endcopy td_pcb
288 * Fields that must be manually set in fork1() or thread_sched_upcall()
289 * or already have been set in the allocator, constructor, etc.
291 struct pcb *td_pcb; /* (k) Kernel VA of pcb and kstack. */
298 } td_state; /* (t) thread state */
299 register_t td_retval[2]; /* (k) Syscall aux returns. */
300 struct callout td_slpcallout; /* (h) Callout for sleep. */
301 struct trapframe *td_frame; /* (k) */
302 struct vm_object *td_kstack_obj;/* (a) Kstack object. */
303 vm_offset_t td_kstack; /* (a) Kernel VA of kstack. */
304 int td_kstack_pages; /* (a) Size of the kstack. */
305 volatile u_int td_critnest; /* (k*) Critical section nest level. */
306 struct mdthread td_md; /* (k) Any machine-dependent fields. */
307 struct td_sched *td_sched; /* (*) Scheduler-specific data. */
308 struct kaudit_record *td_ar; /* (k) Active audit record, if any. */
309 struct lpohead td_lprof[2]; /* (a) lock profiling objects. */
310 struct kdtrace_thread *td_dtrace; /* (*) DTrace-specific data. */
311 int td_errno; /* Error returned by last syscall. */
312 struct vnet *td_vnet; /* (k) Effective vnet. */
313 const char *td_vnet_lpush; /* (k) Debugging vnet push / pop. */
314 struct trapframe *td_intr_frame;/* (k) Frame of the current irq */
315 struct proc *td_rfppwait_p; /* (k) The vforked child */
318 struct mtx *thread_lock_block(struct thread *);
319 void thread_lock_unblock(struct thread *, struct mtx *);
320 void thread_lock_set(struct thread *, struct mtx *);
321 #define THREAD_LOCK_ASSERT(td, type) \
323 struct mtx *__m = (td)->td_lock; \
324 if (__m != &blocked_lock) \
325 mtx_assert(__m, (type)); \
329 #define THREAD_LOCKPTR_ASSERT(td, lock) \
331 struct mtx *__m = (td)->td_lock; \
332 KASSERT((__m == &blocked_lock || __m == (lock)), \
333 ("Thread %p lock %p does not match %p", td, __m, (lock))); \
336 #define THREAD_LOCKPTR_ASSERT(td, lock)
339 #define CRITICAL_ASSERT(td) \
340 KASSERT((td)->td_critnest >= 1, ("Not in critical section"));
343 * Flags kept in td_flags:
344 * To change these you MUST have the scheduler lock.
346 #define TDF_BORROWING 0x00000001 /* Thread is borrowing pri from another. */
347 #define TDF_INPANIC 0x00000002 /* Caused a panic, let it drive crashdump. */
348 #define TDF_INMEM 0x00000004 /* Thread's stack is in memory. */
349 #define TDF_SINTR 0x00000008 /* Sleep is interruptible. */
350 #define TDF_TIMEOUT 0x00000010 /* Timing out during sleep. */
351 #define TDF_IDLETD 0x00000020 /* This is a per-CPU idle thread. */
352 #define TDF_CANSWAP 0x00000040 /* Thread can be swapped. */
353 #define TDF_SLEEPABORT 0x00000080 /* sleepq_abort was called. */
354 #define TDF_KTH_SUSP 0x00000100 /* kthread is suspended */
355 #define TDF_UNUSED09 0x00000200 /* --available-- */
356 #define TDF_BOUNDARY 0x00000400 /* Thread suspended at user boundary */
357 #define TDF_ASTPENDING 0x00000800 /* Thread has some asynchronous events. */
358 #define TDF_TIMOFAIL 0x00001000 /* Timeout from sleep after we were awake. */
359 #define TDF_SBDRY 0x00002000 /* Stop only on usermode boundary. */
360 #define TDF_UPIBLOCKED 0x00004000 /* Thread blocked on user PI mutex. */
361 #define TDF_NEEDSUSPCHK 0x00008000 /* Thread may need to suspend. */
362 #define TDF_NEEDRESCHED 0x00010000 /* Thread needs to yield. */
363 #define TDF_NEEDSIGCHK 0x00020000 /* Thread may need signal delivery. */
364 #define TDF_NOLOAD 0x00040000 /* Ignore during load avg calculations. */
365 #define TDF_UNUSED19 0x00080000 /* --available-- */
366 #define TDF_THRWAKEUP 0x00100000 /* Libthr thread must not suspend itself. */
367 #define TDF_UNUSED21 0x00200000 /* --available-- */
368 #define TDF_SWAPINREQ 0x00400000 /* Swapin request due to wakeup. */
369 #define TDF_UNUSED23 0x00800000 /* --available-- */
370 #define TDF_SCHED0 0x01000000 /* Reserved for scheduler private use */
371 #define TDF_SCHED1 0x02000000 /* Reserved for scheduler private use */
372 #define TDF_SCHED2 0x04000000 /* Reserved for scheduler private use */
373 #define TDF_SCHED3 0x08000000 /* Reserved for scheduler private use */
374 #define TDF_ALRMPEND 0x10000000 /* Pending SIGVTALRM needs to be posted. */
375 #define TDF_PROFPEND 0x20000000 /* Pending SIGPROF needs to be posted. */
376 #define TDF_MACPEND 0x40000000 /* AST-based MAC event pending. */
378 /* Userland debug flags */
379 #define TDB_SUSPEND 0x00000001 /* Thread is suspended by debugger */
380 #define TDB_XSIG 0x00000002 /* Thread is exchanging signal under trace */
381 #define TDB_USERWR 0x00000004 /* Debugger modified memory or registers */
382 #define TDB_SCE 0x00000008 /* Thread performs syscall enter */
383 #define TDB_SCX 0x00000010 /* Thread performs syscall exit */
384 #define TDB_EXEC 0x00000020 /* TDB_SCX from exec(2) family */
385 #define TDB_FORK 0x00000040 /* TDB_SCX from fork(2) that created new
387 #define TDB_STOPATFORK 0x00000080 /* Stop at the return from fork (child
389 #define TDB_CHILD 0x00000100 /* New child indicator for ptrace() */
392 * "Private" flags kept in td_pflags:
393 * These are only written by curthread and thus need no locking.
395 #define TDP_OLDMASK 0x00000001 /* Need to restore mask after suspend. */
396 #define TDP_INKTR 0x00000002 /* Thread is currently in KTR code. */
397 #define TDP_INKTRACE 0x00000004 /* Thread is currently in KTRACE code. */
398 #define TDP_BUFNEED 0x00000008 /* Do not recurse into the buf flush */
399 #define TDP_COWINPROGRESS 0x00000010 /* Snapshot copy-on-write in progress. */
400 #define TDP_ALTSTACK 0x00000020 /* Have alternate signal stack. */
401 #define TDP_DEADLKTREAT 0x00000040 /* Lock aquisition - deadlock treatment. */
402 #define TDP_NOFAULTING 0x00000080 /* Do not handle page faults. */
403 #define TDP_NOSLEEPING 0x00000100 /* Thread is not allowed to sleep on a sq. */
404 #define TDP_OWEUPC 0x00000200 /* Call addupc() at next AST. */
405 #define TDP_ITHREAD 0x00000400 /* Thread is an interrupt thread. */
406 #define TDP_SYNCIO 0x00000800 /* Local override, disable async i/o. */
407 #define TDP_SCHED1 0x00001000 /* Reserved for scheduler private use */
408 #define TDP_SCHED2 0x00002000 /* Reserved for scheduler private use */
409 #define TDP_SCHED3 0x00004000 /* Reserved for scheduler private use */
410 #define TDP_SCHED4 0x00008000 /* Reserved for scheduler private use */
411 #define TDP_GEOM 0x00010000 /* Settle GEOM before finishing syscall */
412 #define TDP_SOFTDEP 0x00020000 /* Stuck processing softdep worklist */
413 #define TDP_NORUNNINGBUF 0x00040000 /* Ignore runningbufspace check */
414 #define TDP_WAKEUP 0x00080000 /* Don't sleep in umtx cond_wait */
415 #define TDP_INBDFLUSH 0x00100000 /* Already in BO_BDFLUSH, do not recurse */
416 #define TDP_KTHREAD 0x00200000 /* This is an official kernel thread */
417 #define TDP_CALLCHAIN 0x00400000 /* Capture thread's callchain */
418 #define TDP_IGNSUSP 0x00800000 /* Permission to ignore the MNTK_SUSPEND* */
419 #define TDP_AUDITREC 0x01000000 /* Audit record pending on thread */
420 #define TDP_RFPPWAIT 0x02000000 /* Handle RFPPWAIT on syscall exit */
421 #define TDP_RESETSPUR 0x04000000 /* Reset spurious page fault history. */
422 #define TDP_NERRNO 0x08000000 /* Last errno is already in td_errno */
425 * Reasons that the current thread can not be run yet.
426 * More than one may apply.
428 #define TDI_SUSPENDED 0x0001 /* On suspension queue. */
429 #define TDI_SLEEPING 0x0002 /* Actually asleep! (tricky). */
430 #define TDI_SWAPPED 0x0004 /* Stack not in mem. Bad juju if run. */
431 #define TDI_LOCK 0x0008 /* Stopped on a lock. */
432 #define TDI_IWAIT 0x0010 /* Awaiting interrupt. */
434 #define TD_IS_SLEEPING(td) ((td)->td_inhibitors & TDI_SLEEPING)
435 #define TD_ON_SLEEPQ(td) ((td)->td_wchan != NULL)
436 #define TD_IS_SUSPENDED(td) ((td)->td_inhibitors & TDI_SUSPENDED)
437 #define TD_IS_SWAPPED(td) ((td)->td_inhibitors & TDI_SWAPPED)
438 #define TD_ON_LOCK(td) ((td)->td_inhibitors & TDI_LOCK)
439 #define TD_AWAITING_INTR(td) ((td)->td_inhibitors & TDI_IWAIT)
440 #define TD_IS_RUNNING(td) ((td)->td_state == TDS_RUNNING)
441 #define TD_ON_RUNQ(td) ((td)->td_state == TDS_RUNQ)
442 #define TD_CAN_RUN(td) ((td)->td_state == TDS_CAN_RUN)
443 #define TD_IS_INHIBITED(td) ((td)->td_state == TDS_INHIBITED)
444 #define TD_ON_UPILOCK(td) ((td)->td_flags & TDF_UPIBLOCKED)
445 #define TD_IS_IDLETHREAD(td) ((td)->td_flags & TDF_IDLETD)
448 #define TD_SET_INHIB(td, inhib) do { \
449 (td)->td_state = TDS_INHIBITED; \
450 (td)->td_inhibitors |= (inhib); \
453 #define TD_CLR_INHIB(td, inhib) do { \
454 if (((td)->td_inhibitors & (inhib)) && \
455 (((td)->td_inhibitors &= ~(inhib)) == 0)) \
456 (td)->td_state = TDS_CAN_RUN; \
459 #define TD_SET_SLEEPING(td) TD_SET_INHIB((td), TDI_SLEEPING)
460 #define TD_SET_SWAPPED(td) TD_SET_INHIB((td), TDI_SWAPPED)
461 #define TD_SET_LOCK(td) TD_SET_INHIB((td), TDI_LOCK)
462 #define TD_SET_SUSPENDED(td) TD_SET_INHIB((td), TDI_SUSPENDED)
463 #define TD_SET_IWAIT(td) TD_SET_INHIB((td), TDI_IWAIT)
464 #define TD_SET_EXITING(td) TD_SET_INHIB((td), TDI_EXITING)
466 #define TD_CLR_SLEEPING(td) TD_CLR_INHIB((td), TDI_SLEEPING)
467 #define TD_CLR_SWAPPED(td) TD_CLR_INHIB((td), TDI_SWAPPED)
468 #define TD_CLR_LOCK(td) TD_CLR_INHIB((td), TDI_LOCK)
469 #define TD_CLR_SUSPENDED(td) TD_CLR_INHIB((td), TDI_SUSPENDED)
470 #define TD_CLR_IWAIT(td) TD_CLR_INHIB((td), TDI_IWAIT)
472 #define TD_SET_RUNNING(td) (td)->td_state = TDS_RUNNING
473 #define TD_SET_RUNQ(td) (td)->td_state = TDS_RUNQ
474 #define TD_SET_CAN_RUN(td) (td)->td_state = TDS_CAN_RUN
480 LIST_ENTRY(proc) p_list; /* (d) List of all processes. */
481 TAILQ_HEAD(, thread) p_threads; /* (c) all threads. */
482 struct mtx p_slock; /* process spin lock */
483 struct ucred *p_ucred; /* (c) Process owner's identity. */
484 struct filedesc *p_fd; /* (b) Open files. */
485 struct filedesc_to_leader *p_fdtol; /* (b) Tracking node */
486 struct pstats *p_stats; /* (b) Accounting/statistics (CPU). */
487 struct plimit *p_limit; /* (c) Process limits. */
488 struct callout p_limco; /* (c) Limit callout handle */
489 struct sigacts *p_sigacts; /* (x) Signal actions, state (CPU). */
492 * The following don't make too much sense.
493 * See the td_ or ke_ versions of the same flags.
495 int p_flag; /* (c) P_* flags. */
497 PRS_NEW = 0, /* In creation */
498 PRS_NORMAL, /* threads can be run. */
500 } p_state; /* (j/c) Process status. */
501 pid_t p_pid; /* (b) Process identifier. */
502 LIST_ENTRY(proc) p_hash; /* (d) Hash chain. */
503 LIST_ENTRY(proc) p_pglist; /* (g + e) List of processes in pgrp. */
504 struct proc *p_pptr; /* (c + e) Pointer to parent process. */
505 LIST_ENTRY(proc) p_sibling; /* (e) List of sibling processes. */
506 LIST_HEAD(, proc) p_children; /* (e) Pointer to list of children. */
507 struct mtx p_mtx; /* (n) Lock for this struct. */
508 struct ksiginfo *p_ksi; /* Locked by parent proc lock */
509 sigqueue_t p_sigqueue; /* (c) Sigs not delivered to a td. */
510 #define p_siglist p_sigqueue.sq_signals
512 /* The following fields are all zeroed upon creation in fork. */
513 #define p_startzero p_oppid
514 pid_t p_oppid; /* (c + e) Save ppid in ptrace. XXX */
516 struct vmspace *p_vmspace; /* (b) Address space. */
517 u_int p_swtick; /* (c) Tick when swapped in or out. */
518 struct itimerval p_realtimer; /* (c) Alarm timer. */
519 struct rusage p_ru; /* (a) Exit information. */
520 struct rusage_ext p_rux; /* (cj) Internal resource usage. */
521 struct rusage_ext p_crux; /* (c) Internal child resource usage. */
522 int p_profthreads; /* (c) Num threads in addupc_task. */
523 volatile int p_exitthreads; /* (j) Number of threads exiting */
524 int p_traceflag; /* (o) Kernel trace points. */
525 struct vnode *p_tracevp; /* (c + o) Trace to vnode. */
526 struct ucred *p_tracecred; /* (o) Credentials to trace with. */
527 struct vnode *p_textvp; /* (b) Vnode of executable. */
528 u_int p_lock; /* (c) Proclock (prevent swap) count. */
529 struct sigiolst p_sigiolst; /* (c) List of sigio sources. */
530 int p_sigparent; /* (c) Signal to parent on exit. */
531 int p_sig; /* (n) For core dump/debugger XXX. */
532 u_long p_code; /* (n) For core dump/debugger XXX. */
533 u_int p_stops; /* (c) Stop event bitmask. */
534 u_int p_stype; /* (c) Stop event type. */
535 char p_step; /* (c) Process is stopped. */
536 u_char p_pfsflags; /* (c) Procfs flags. */
537 struct nlminfo *p_nlminfo; /* (?) Only used by/for lockd. */
538 struct kaioinfo *p_aioinfo; /* (y) ASYNC I/O info. */
539 struct thread *p_singlethread;/* (c + j) If single threading this is it */
540 int p_suspcount; /* (j) Num threads in suspended mode. */
541 struct thread *p_xthread; /* (c) Trap thread */
542 int p_boundary_count;/* (j) Num threads at user boundary */
543 int p_pendingcnt; /* how many signals are pending */
544 struct itimers *p_itimers; /* (c) POSIX interval timers. */
545 struct procdesc *p_procdesc; /* (e) Process descriptor, if any. */
546 /* End area that is zeroed on creation. */
547 #define p_endzero p_magic
549 /* The following fields are all copied upon creation in fork. */
550 #define p_startcopy p_endzero
551 u_int p_magic; /* (b) Magic number. */
552 int p_osrel; /* (x) osreldate for the
553 binary (from ELF note, if any) */
554 char p_comm[MAXCOMLEN + 1]; /* (b) Process name. */
555 struct pgrp *p_pgrp; /* (c + e) Pointer to process group. */
556 struct sysentvec *p_sysent; /* (b) Syscall dispatch info. */
557 struct pargs *p_args; /* (c) Process arguments. */
558 rlim_t p_cpulimit; /* (c) Current CPU limit in seconds. */
559 signed char p_nice; /* (c) Process "nice" value. */
560 int p_fibnum; /* in this routing domain XXX MRT */
561 /* End area that is copied on creation. */
562 #define p_endcopy p_xstat
564 u_short p_xstat; /* (c) Exit status; also stop sig. */
565 struct knlist p_klist; /* (c) Knotes attached to this proc. */
566 int p_numthreads; /* (c) Number of threads. */
567 struct mdproc p_md; /* Any machine-dependent fields. */
568 struct callout p_itcallout; /* (h + c) Interval timer callout. */
569 u_short p_acflag; /* (c) Accounting flags. */
570 struct proc *p_peers; /* (r) */
571 struct proc *p_leader; /* (b) */
572 void *p_emuldata; /* (c) Emulator state data. */
573 struct label *p_label; /* (*) Proc (not subject) MAC label. */
574 struct p_sched *p_sched; /* (*) Scheduler-specific data. */
575 STAILQ_HEAD(, ktr_request) p_ktr; /* (o) KTR event queue. */
576 LIST_HEAD(, mqueue_notifier) p_mqnotifier; /* (c) mqueue notifiers.*/
577 struct kdtrace_proc *p_dtrace; /* (*) DTrace-specific data. */
578 struct cv p_pwait; /* (*) wait cv for exit/exec. */
579 struct cv p_dbgwait; /* (*) wait cv for debugger attach
581 uint64_t p_prev_runtime; /* (c) Resource usage accounting. */
582 struct racct *p_racct; /* (b) Resource accounting. */
584 * An orphan is the child that has beed re-parented to the
585 * debugger as a result of attaching to it. Need to keep
586 * track of them for parent to be able to collect the exit
587 * status of what used to be children.
589 LIST_ENTRY(proc) p_orphan; /* (e) List of orphan processes. */
590 LIST_HEAD(, proc) p_orphans; /* (e) Pointer to list of orphans. */
593 #define p_session p_pgrp->pg_session
594 #define p_pgid p_pgrp->pg_id
596 #define NOCPU 0xff /* For when we aren't on a CPU. */
598 #define PROC_SLOCK(p) mtx_lock_spin(&(p)->p_slock)
599 #define PROC_SUNLOCK(p) mtx_unlock_spin(&(p)->p_slock)
600 #define PROC_SLOCK_ASSERT(p, type) mtx_assert(&(p)->p_slock, (type))
602 /* These flags are kept in p_flag. */
603 #define P_ADVLOCK 0x00001 /* Process may hold a POSIX advisory lock. */
604 #define P_CONTROLT 0x00002 /* Has a controlling terminal. */
605 #define P_KTHREAD 0x00004 /* Kernel thread (*). */
606 #define P_FOLLOWFORK 0x00008 /* Attach parent debugger to children. */
607 #define P_PPWAIT 0x00010 /* Parent is waiting for child to exec/exit. */
608 #define P_PROFIL 0x00020 /* Has started profiling. */
609 #define P_STOPPROF 0x00040 /* Has thread requesting to stop profiling. */
610 #define P_HADTHREADS 0x00080 /* Has had threads (no cleanup shortcuts) */
611 #define P_SUGID 0x00100 /* Had set id privileges since last exec. */
612 #define P_SYSTEM 0x00200 /* System proc: no sigs, stats or swapping. */
613 #define P_SINGLE_EXIT 0x00400 /* Threads suspending should exit, not wait. */
614 #define P_TRACED 0x00800 /* Debugged process being traced. */
615 #define P_WAITED 0x01000 /* Someone is waiting for us. */
616 #define P_WEXIT 0x02000 /* Working on exiting. */
617 #define P_EXEC 0x04000 /* Process called exec. */
618 #define P_WKILLED 0x08000 /* Killed, go to kernel/user boundary ASAP. */
619 #define P_CONTINUED 0x10000 /* Proc has continued from a stopped state. */
620 #define P_STOPPED_SIG 0x20000 /* Stopped due to SIGSTOP/SIGTSTP. */
621 #define P_STOPPED_TRACE 0x40000 /* Stopped because of tracing. */
622 #define P_STOPPED_SINGLE 0x80000 /* Only 1 thread can continue (not to user). */
623 #define P_PROTECTED 0x100000 /* Do not kill on memory overcommit. */
624 #define P_SIGEVENT 0x200000 /* Process pending signals changed. */
625 #define P_SINGLE_BOUNDARY 0x400000 /* Threads should suspend at user boundary. */
626 #define P_HWPMC 0x800000 /* Process is using HWPMCs */
627 #define P_JAILED 0x1000000 /* Process is in jail. */
628 #define P_ORPHAN 0x2000000 /* Orphaned. */
629 #define P_INEXEC 0x4000000 /* Process is in execve(). */
630 #define P_STATCHILD 0x8000000 /* Child process stopped or exited. */
631 #define P_INMEM 0x10000000 /* Loaded into memory. */
632 #define P_SWAPPINGOUT 0x20000000 /* Process is being swapped out. */
633 #define P_SWAPPINGIN 0x40000000 /* Process is being swapped in. */
635 #define P_STOPPED (P_STOPPED_SIG|P_STOPPED_SINGLE|P_STOPPED_TRACE)
636 #define P_SHOULDSTOP(p) ((p)->p_flag & P_STOPPED)
637 #define P_KILLED(p) ((p)->p_flag & P_WKILLED)
640 * These were process status values (p_stat), now they are only used in
641 * legacy conversion code.
643 #define SIDL 1 /* Process being created by fork. */
644 #define SRUN 2 /* Currently runnable. */
645 #define SSLEEP 3 /* Sleeping on an address. */
646 #define SSTOP 4 /* Process debugging or suspension. */
647 #define SZOMB 5 /* Awaiting collection by parent. */
648 #define SWAIT 6 /* Waiting for interrupt. */
649 #define SLOCK 7 /* Blocked on a lock. */
651 #define P_MAGIC 0xbeefface
655 /* Types and flags for mi_switch(). */
656 #define SW_TYPE_MASK 0xff /* First 8 bits are switch type */
657 #define SWT_NONE 0 /* Unspecified switch. */
658 #define SWT_PREEMPT 1 /* Switching due to preemption. */
659 #define SWT_OWEPREEMPT 2 /* Switching due to opepreempt. */
660 #define SWT_TURNSTILE 3 /* Turnstile contention. */
661 #define SWT_SLEEPQ 4 /* Sleepq wait. */
662 #define SWT_SLEEPQTIMO 5 /* Sleepq timeout wait. */
663 #define SWT_RELINQUISH 6 /* yield call. */
664 #define SWT_NEEDRESCHED 7 /* NEEDRESCHED was set. */
665 #define SWT_IDLE 8 /* Switching from the idle thread. */
666 #define SWT_IWAIT 9 /* Waiting for interrupts. */
667 #define SWT_SUSPEND 10 /* Thread suspended. */
668 #define SWT_REMOTEPREEMPT 11 /* Remote processor preempted. */
669 #define SWT_REMOTEWAKEIDLE 12 /* Remote processor preempted idle. */
670 #define SWT_COUNT 13 /* Number of switch types. */
672 #define SW_VOL 0x0100 /* Voluntary switch. */
673 #define SW_INVOL 0x0200 /* Involuntary switch. */
674 #define SW_PREEMPT 0x0400 /* The invol switch is a preemption */
676 /* How values for thread_single(). */
677 #define SINGLE_NO_EXIT 0
678 #define SINGLE_EXIT 1
679 #define SINGLE_BOUNDARY 2
681 #ifdef MALLOC_DECLARE
682 MALLOC_DECLARE(M_PARGS);
683 MALLOC_DECLARE(M_PGRP);
684 MALLOC_DECLARE(M_SESSION);
685 MALLOC_DECLARE(M_SUBPROC);
688 #define FOREACH_PROC_IN_SYSTEM(p) \
689 LIST_FOREACH((p), &allproc, p_list)
690 #define FOREACH_THREAD_IN_PROC(p, td) \
691 TAILQ_FOREACH((td), &(p)->p_threads, td_plist)
693 #define FIRST_THREAD_IN_PROC(p) TAILQ_FIRST(&(p)->p_threads)
696 * We use process IDs <= PID_MAX; PID_MAX + 1 must also fit in a pid_t,
697 * as it is used to represent "no process group".
699 #define PID_MAX 99999
700 #define NO_PID 100000
702 #define SESS_LEADER(p) ((p)->p_session->s_leader == (p))
705 #define STOPEVENT(p, e, v) do { \
706 if ((p)->p_stops & (e)) { \
708 stopevent((p), (e), (v)); \
712 #define _STOPEVENT(p, e, v) do { \
713 PROC_LOCK_ASSERT(p, MA_OWNED); \
714 WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, &p->p_mtx.lock_object, \
715 "checking stopevent %d", (e)); \
716 if ((p)->p_stops & (e)) \
717 stopevent((p), (e), (v)); \
720 /* Lock and unlock a process. */
721 #define PROC_LOCK(p) mtx_lock(&(p)->p_mtx)
722 #define PROC_TRYLOCK(p) mtx_trylock(&(p)->p_mtx)
723 #define PROC_UNLOCK(p) mtx_unlock(&(p)->p_mtx)
724 #define PROC_LOCKED(p) mtx_owned(&(p)->p_mtx)
725 #define PROC_LOCK_ASSERT(p, type) mtx_assert(&(p)->p_mtx, (type))
727 /* Lock and unlock a process group. */
728 #define PGRP_LOCK(pg) mtx_lock(&(pg)->pg_mtx)
729 #define PGRP_UNLOCK(pg) mtx_unlock(&(pg)->pg_mtx)
730 #define PGRP_LOCKED(pg) mtx_owned(&(pg)->pg_mtx)
731 #define PGRP_LOCK_ASSERT(pg, type) mtx_assert(&(pg)->pg_mtx, (type))
733 #define PGRP_LOCK_PGSIGNAL(pg) do { \
737 #define PGRP_UNLOCK_PGSIGNAL(pg) do { \
742 /* Lock and unlock a session. */
743 #define SESS_LOCK(s) mtx_lock(&(s)->s_mtx)
744 #define SESS_UNLOCK(s) mtx_unlock(&(s)->s_mtx)
745 #define SESS_LOCKED(s) mtx_owned(&(s)->s_mtx)
746 #define SESS_LOCK_ASSERT(s, type) mtx_assert(&(s)->s_mtx, (type))
748 /* Hold process U-area in memory, normally for ptrace/procfs work. */
749 #define PHOLD(p) do { \
754 #define _PHOLD(p) do { \
755 PROC_LOCK_ASSERT((p), MA_OWNED); \
756 KASSERT(!((p)->p_flag & P_WEXIT) || (p) == curproc, \
757 ("PHOLD of exiting process")); \
759 if (((p)->p_flag & P_INMEM) == 0) \
762 #define PROC_ASSERT_HELD(p) do { \
763 KASSERT((p)->p_lock > 0, ("process not held")); \
766 #define PRELE(p) do { \
771 #define _PRELE(p) do { \
772 PROC_LOCK_ASSERT((p), MA_OWNED); \
773 PROC_ASSERT_HELD(p); \
775 if (((p)->p_flag & P_WEXIT) && (p)->p_lock == 0) \
776 wakeup(&(p)->p_lock); \
778 #define PROC_ASSERT_NOT_HELD(p) do { \
779 KASSERT((p)->p_lock == 0, ("process held")); \
782 /* Check whether a thread is safe to be swapped out. */
783 #define thread_safetoswapout(td) ((td)->td_flags & TDF_CANSWAP)
785 /* Control whether or not it is safe for curthread to sleep. */
786 #define THREAD_NO_SLEEPING() do { \
787 KASSERT(!(curthread->td_pflags & TDP_NOSLEEPING), \
788 ("nested no sleeping")); \
789 curthread->td_pflags |= TDP_NOSLEEPING; \
792 #define THREAD_SLEEPING_OK() do { \
793 KASSERT((curthread->td_pflags & TDP_NOSLEEPING), \
794 ("nested sleeping ok")); \
795 curthread->td_pflags &= ~TDP_NOSLEEPING; \
798 #define PIDHASH(pid) (&pidhashtbl[(pid) & pidhash])
799 extern LIST_HEAD(pidhashhead, proc) *pidhashtbl;
800 extern u_long pidhash;
801 #define TIDHASH(tid) (&tidhashtbl[(tid) & tidhash])
802 extern LIST_HEAD(tidhashhead, thread) *tidhashtbl;
803 extern u_long tidhash;
804 extern struct rwlock tidhash_lock;
806 #define PGRPHASH(pgid) (&pgrphashtbl[(pgid) & pgrphash])
807 extern LIST_HEAD(pgrphashhead, pgrp) *pgrphashtbl;
808 extern u_long pgrphash;
810 extern struct sx allproc_lock;
811 extern struct sx proctree_lock;
812 extern struct mtx ppeers_lock;
813 extern struct proc proc0; /* Process slot for swapper. */
814 extern struct thread thread0; /* Primary thread in proc0. */
815 extern struct vmspace vmspace0; /* VM space for proc0. */
816 extern int hogticks; /* Limit on kernel cpu hogs. */
818 extern int nprocs, maxproc; /* Current and max number of procs. */
819 extern int maxprocperuid; /* Max procs per uid. */
820 extern u_long ps_arg_cache_limit;
822 LIST_HEAD(proclist, proc);
823 TAILQ_HEAD(procqueue, proc);
824 TAILQ_HEAD(threadqueue, thread);
825 extern struct proclist allproc; /* List of all processes. */
826 extern struct proclist zombproc; /* List of zombie processes. */
827 extern struct proc *initproc, *pageproc; /* Process slots for init, pager. */
829 extern struct uma_zone *proc_zone;
831 struct proc *pfind(pid_t); /* Find process by id. */
832 struct pgrp *pgfind(pid_t); /* Find process group by id. */
833 struct proc *zpfind(pid_t); /* Find zombie process by id. */
838 #define PGET_HOLD 0x00001 /* Hold the process. */
839 #define PGET_CANSEE 0x00002 /* Check against p_cansee(). */
840 #define PGET_CANDEBUG 0x00004 /* Check against p_candebug(). */
841 #define PGET_ISCURRENT 0x00008 /* Check that the found process is current. */
842 #define PGET_NOTWEXIT 0x00010 /* Check that the process is not in P_WEXIT. */
843 #define PGET_NOTINEXEC 0x00020 /* Check that the process is not in P_INEXEC. */
844 #define PGET_NOTID 0x00040 /* Do not assume tid if pid > PID_MAX. */
846 #define PGET_WANTREAD (PGET_HOLD | PGET_CANDEBUG | PGET_NOTWEXIT)
848 int pget(pid_t pid, int flags, struct proc **pp);
850 void ast(struct trapframe *framep);
851 struct thread *choosethread(void);
852 int cr_cansignal(struct ucred *cred, struct proc *proc, int signum);
853 int enterpgrp(struct proc *p, pid_t pgid, struct pgrp *pgrp,
854 struct session *sess);
855 int enterthispgrp(struct proc *p, struct pgrp *pgrp);
856 void faultin(struct proc *p);
857 void fixjobc(struct proc *p, struct pgrp *pgrp, int entering);
858 int fork1(struct thread *, int, int, struct proc **, int *, int);
859 void fork_exit(void (*)(void *, struct trapframe *), void *,
861 void fork_return(struct thread *, struct trapframe *);
862 int inferior(struct proc *p);
863 void kern_yield(int);
864 void kick_proc0(void);
865 int leavepgrp(struct proc *p);
866 int maybe_preempt(struct thread *td);
867 void maybe_yield(void);
868 void mi_switch(int flags, struct thread *newtd);
869 int p_candebug(struct thread *td, struct proc *p);
870 int p_cansee(struct thread *td, struct proc *p);
871 int p_cansched(struct thread *td, struct proc *p);
872 int p_cansignal(struct thread *td, struct proc *p, int signum);
873 int p_canwait(struct thread *td, struct proc *p);
874 struct pargs *pargs_alloc(int len);
875 void pargs_drop(struct pargs *pa);
876 void pargs_hold(struct pargs *pa);
877 int proc_getargv(struct thread *td, struct proc *p, struct sbuf *sb);
878 int proc_getenvv(struct thread *td, struct proc *p, struct sbuf *sb);
880 void proc_linkup0(struct proc *p, struct thread *td);
881 void proc_linkup(struct proc *p, struct thread *td);
882 void proc_reap(struct thread *td, struct proc *p, int *status, int options,
883 struct rusage *rusage);
884 void proc_reparent(struct proc *child, struct proc *newparent);
885 struct pstats *pstats_alloc(void);
886 void pstats_fork(struct pstats *src, struct pstats *dst);
887 void pstats_free(struct pstats *ps);
888 int securelevel_ge(struct ucred *cr, int level);
889 int securelevel_gt(struct ucred *cr, int level);
890 void sess_hold(struct session *);
891 void sess_release(struct session *);
892 int setrunnable(struct thread *);
893 void setsugid(struct proc *p);
894 int should_yield(void);
895 int sigonstack(size_t sp);
896 void sleepinit(void);
897 void stopevent(struct proc *, u_int, u_int);
898 struct thread *tdfind(lwpid_t, pid_t);
899 void threadinit(void);
900 void tidhash_add(struct thread *);
901 void tidhash_remove(struct thread *);
903 int cpu_idle_wakeup(int);
904 extern void (*cpu_idle_hook)(void); /* Hook to machdep CPU idler. */
905 void cpu_switch(struct thread *, struct thread *, struct mtx *);
906 void cpu_throw(struct thread *, struct thread *) __dead2;
907 void unsleep(struct thread *);
908 void userret(struct thread *, struct trapframe *);
910 void cpu_exit(struct thread *);
911 void exit1(struct thread *, int) __dead2;
913 int cpu_fetch_syscall_args(struct thread *td, struct syscall_args *sa);
914 void cpu_fork(struct thread *, struct proc *, struct thread *, int);
915 void cpu_set_fork_handler(struct thread *, void (*)(void *), void *);
916 void cpu_set_syscall_retval(struct thread *, int);
917 void cpu_set_upcall(struct thread *td, struct thread *td0);
918 void cpu_set_upcall_kse(struct thread *, void (*)(void *), void *,
920 int cpu_set_user_tls(struct thread *, void *tls_base);
921 void cpu_thread_alloc(struct thread *);
922 void cpu_thread_clean(struct thread *);
923 void cpu_thread_exit(struct thread *);
924 void cpu_thread_free(struct thread *);
925 void cpu_thread_swapin(struct thread *);
926 void cpu_thread_swapout(struct thread *);
927 struct thread *thread_alloc(int pages);
928 int thread_alloc_stack(struct thread *, int pages);
929 void thread_exit(void) __dead2;
930 void thread_free(struct thread *td);
931 void thread_link(struct thread *td, struct proc *p);
932 void thread_reap(void);
933 int thread_single(int how);
934 void thread_single_end(void);
935 void thread_stash(struct thread *td);
936 void thread_stopped(struct proc *p);
937 void childproc_stopped(struct proc *child, int reason);
938 void childproc_continued(struct proc *child);
939 void childproc_exited(struct proc *child);
940 int thread_suspend_check(int how);
941 void thread_suspend_switch(struct thread *);
942 void thread_suspend_one(struct thread *td);
943 void thread_unlink(struct thread *td);
944 void thread_unsuspend(struct proc *p);
945 int thread_unsuspend_one(struct thread *td);
946 void thread_unthread(struct thread *td);
947 void thread_wait(struct proc *p);
948 struct thread *thread_find(struct proc *p, lwpid_t tid);
951 curthread_pflags_set(int flags)
957 save = ~flags | (td->td_pflags & flags);
958 td->td_pflags |= flags;
963 curthread_pflags_restore(int save)
966 curthread->td_pflags &= save;
971 #endif /* !_SYS_PROC_H_ */