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 * 3. All advertising materials mentioning features or use of this software
19 * must display the following acknowledgement:
20 * This product includes software developed by the University of
21 * California, Berkeley and its contributors.
22 * 4. Neither the name of the University nor the names of its contributors
23 * may be used to endorse or promote products derived from this software
24 * without specific prior written permission.
26 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
38 * @(#)proc.h 8.15 (Berkeley) 5/19/95
39 * $FreeBSD: src/sys/sys/proc.h,v 1.99.2.9 2003/06/06 20:21:32 tegge Exp $
40 * $DragonFly: src/sys/sys/proc.h,v 1.121 2008/07/23 17:22:33 dillon Exp $
46 #if !defined(_KERNEL) && !defined(_KERNEL_STRUCTURES)
48 #error "Userland must include sys/user.h instead of sys/proc.h"
52 #include <sys/callout.h> /* For struct callout_handle. */
53 #include <sys/filedesc.h>
54 #include <sys/queue.h>
56 #include <sys/rtprio.h> /* For struct rtprio. */
57 #include <sys/signal.h>
60 #include <sys/time.h> /* For structs itimerval, timeval. */
62 #include <sys/ucred.h>
63 #include <sys/event.h> /* For struct klist */
64 #include <sys/eventvar.h>
65 #include <sys/sysent.h> /* For struct sysentvec */
66 #include <sys/thread.h>
67 #include <sys/varsym.h>
68 #include <sys/upcall.h>
69 #include <sys/resourcevar.h>
71 #include <sys/globaldata.h>
73 #include <sys/systimer.h>
74 #include <sys/iosched.h>
75 #include <sys/usched.h>
76 #include <machine/proc.h> /* Machine-dependent proc substruct. */
77 #include <machine/atomic.h> /* Machine-dependent proc substruct. */
78 #include <sys/signalvar.h>
80 LIST_HEAD(proclist, proc);
81 LIST_HEAD(lwplist, lwp);
84 RB_HEAD(lwp_rb_tree, lwp);
85 RB_PROTOTYPE2(lwp_rb_tree, lwp, u.lwp_rbnode, rb_lwp_compare, lwpid_t);
88 * One structure allocated per session.
91 int s_count; /* Ref cnt; pgrps in session. */
92 struct proc *s_leader; /* Session leader. */
93 struct vnode *s_ttyvp; /* Vnode of controlling terminal. */
94 struct tty *s_ttyp; /* Controlling terminal. */
95 pid_t s_sid; /* Session ID */
96 char s_login[roundup(MAXLOGNAME, sizeof(long))]; /* Setlogin() name. */
100 * One structure allocated per process group.
103 LIST_ENTRY(pgrp) pg_hash; /* Hash chain. */
104 struct proclist pg_members; /* Pointer to pgrp members. */
105 struct session *pg_session; /* Pointer to session. */
106 struct sigiolst pg_sigiolst; /* List of sigio sources. */
107 pid_t pg_id; /* Pgrp id. */
108 int pg_jobc; /* # procs qualifying pgrp for job control */
110 struct lwkt_token pg_token;
114 #define PS_NOCLDWAIT 0x0001 /* No zombies if child dies */
115 #define PS_NOCLDSTOP 0x0002 /* No SIGCHLD when children stop. */
118 * pargs, used to hold a copy of the command line, if it had a sane
122 u_int ar_ref; /* Reference count */
123 u_int ar_length; /* Length */
124 u_char ar_args[0]; /* Arguments */
128 * Description of a process.
130 * This structure contains the information needed to manage a thread of
131 * control, known in UN*X as a process; it has references to substructures
132 * containing descriptions of things that the process uses, but may share
133 * with related processes. The process structure and the substructures
134 * are always addressable except for those marked "(PROC ONLY)" below,
135 * which might be addressable only on a processor on which the process
138 * NOTE! The process start time is stored in the thread structure associated
139 * with the process. If the process is a Zombie, then this field will be
140 * inaccessible due to the thread structure being free'd in kern_wait1().
146 struct vmspace_entry;
163 TAILQ_ENTRY(lwp) lwp_procq; /* run/sleep queue. */
165 RB_ENTRY(lwp) lwp_rbnode; /* RB tree node - lwp in proc */
166 LIST_ENTRY(lwp) lwp_reap_entry; /* reaper list */
169 struct proc *lwp_proc; /* Link to our proc. */
170 struct vmspace *lwp_vmspace; /* Inherited from p_vmspace */
171 struct vkernel_lwp *lwp_vkernel;/* VKernel support, lwp part */
173 lwpid_t lwp_tid; /* Our thread id . */
175 int lwp_flag; /* LWP_* flags. */
176 enum lwpstat lwp_stat; /* LS* lwp status. */
177 int lwp_lock; /* lwp lock (prevent destruct) count */
179 int lwp_dupfd; /* Sideways return value from fdopen. XXX */
182 * The following two fields are marked XXX since (at least) the
183 * 4.4BSD-Lite2 import. I can only guess the reason: It is ugly.
184 * These fields are used to pass the trap code from trapsignal() to
185 * postsig(), which gets called later from userret().
187 * The correct "fix" for these XXX is to convert our signal system
188 * to use signal queues, where each signal can carry its own meta
191 int lwp_sig; /* for core dump/debugger XXX */
192 u_long lwp_code; /* for core dump/debugger XXX */
197 sysclock_t lwp_cpticks; /* cpu used in sched clock ticks */
198 sysclock_t lwp_cpbase; /* Measurement base */
199 fixpt_t lwp_pctcpu; /* %cpu for this process */
200 u_int lwp_slptime; /* Time since last blocked. */
202 int lwp_traceflag; /* Kernel trace points. */
204 struct rusage lwp_ru; /* stats for this lwp */
206 union usched_data lwp_usdata; /* User scheduler specific */
208 #define lwp_startcopy lwp_cpumask
209 cpumask_t lwp_cpumask;
210 sigset_t lwp_siglist; /* Signals arrived but not delivered. */
211 sigset_t lwp_oldsigmask; /* saved mask from before sigpause */
212 sigset_t lwp_sigmask; /* Current signal mask. */
213 stack_t lwp_sigstk; /* sp & on stack state variable */
215 struct rtprio lwp_rtprio; /* Realtime priority. */
216 #define lwp_endcopy lwp_md
218 struct mdproc lwp_md; /* Any machine-dependent fields. */
220 struct thread *lwp_thread; /* backpointer to proc's thread */
221 struct upcall *lwp_upcall; /* REGISTERED USERLAND POINTER! */
222 struct kqueue lwp_kqueue; /* for select/poll */
223 u_int lwp_kqueue_serial;
227 LIST_ENTRY(proc) p_list; /* List of all processes. */
230 struct ucred *p_ucred; /* Process owner's identity. */
231 struct filedesc *p_fd; /* Ptr to open files structure. */
232 struct filedesc_to_leader *p_fdtol; /* Ptr to tracking node XXX lwp */
233 struct plimit *p_limit; /* Process limits. */
234 struct pstats *p_stats;
235 u_int p_mqueue_cnt; /* Count of open mqueues. */
237 struct sigacts *p_sigacts;
238 #define p_sigignore p_sigacts->ps_sigignore
239 #define p_sigcatch p_sigacts->ps_sigcatch
240 #define p_rlimit p_limit->pl_rlimit
242 int p_flag; /* P_* flags. */
243 enum procstat p_stat; /* S* process status. */
246 pid_t p_pid; /* Process identifier. */
247 LIST_ENTRY(proc) p_hash; /* Hash chain. */
248 LIST_ENTRY(proc) p_pglist; /* List of processes in pgrp. */
249 struct proc *p_pptr; /* Pointer to parent process. */
250 LIST_ENTRY(proc) p_sibling; /* List of sibling processes. */
251 struct proclist p_children; /* Pointer to list of children. */
252 struct callout p_ithandle; /* for scheduling p_realtimer */
253 struct varsymset p_varsymset;
254 struct iosched_data p_iosdata; /* Dynamic I/O scheduling data */
256 pid_t p_oppid; /* Save parent pid during ptrace. XXX */
258 struct vmspace *p_vmspace; /* Current address space. */
260 unsigned int p_swtime; /* Time swapped in or out */
262 struct itimerval p_realtimer; /* Alarm timer. */
263 struct itimerval p_timer[3]; /* Virtual-time timers. */
265 int p_traceflag; /* Kernel trace points. */
266 struct ktrace_node *p_tracenode; /* Trace to vnode. */
268 sigset_t p_siglist; /* Signals arrived but not delivered. */
270 struct vnode *p_textvp; /* Vnode of executable. */
271 struct nchandle p_textnch; /* namecache handle of executable. */
273 unsigned int p_stops; /* procfs event bitmask */
274 unsigned int p_stype; /* procfs stop event type */
275 char p_step; /* procfs stop *once* flag */
276 unsigned char p_pfsflags; /* procfs flags */
277 char p_pad2[2]; /* padding for alignment */
278 struct sigiolst p_sigiolst; /* list of sigio sources */
279 int p_sigparent; /* signal to parent on exit */
280 struct klist p_klist; /* knotes attached to this process */
282 struct timeval p_start; /* start time for a process */
284 struct rusage p_ru; /* stats for this proc */
285 struct rusage p_cru; /* sum of stats for reaped children */
286 void *p_dsched_priv1;
288 /* The following fields are all copied upon creation in fork. */
289 #define p_startcopy p_comm
291 char p_comm[MAXCOMLEN+1]; /* typ 16+1 bytes */
292 char p_pad3; /* Process lock (prevent destruct) count. */
293 char p_nice; /* Process "nice" value. */
296 struct pgrp *p_pgrp; /* Pointer to process group. */
298 struct sysentvec *p_sysent; /* System call dispatch information. */
300 struct uprof p_prof; /* Profiling arguments. */
301 struct rtprio p_rtprio; /* Realtime priority. */
302 struct pargs *p_args;
303 u_short p_xstat; /* Exit status or last stop signal */
306 void *p_dsched_priv2;
307 /* End area that is copied on creation. */
308 #define p_endcopy p_dsched_priv2
309 u_short p_acflag; /* Accounting flags. */
311 int p_lock; /* Prevent proc destruction */
312 int p_nthreads; /* Number of threads in this process. */
313 int p_nstopped; /* Number of stopped threads. */
314 int p_lasttid; /* Last tid used. */
315 struct lwp_rb_tree p_lwp_tree; /* RB tree of LWPs for this process */
316 void *p_aioinfo; /* ASYNC I/O info */
317 int p_wakeup; /* thread id XXX lwp */
318 struct proc *p_peers; /* XXX lwp */
319 struct proc *p_leader; /* XXX lwp */
320 void *p_emuldata; /* process-specific emulator state */
321 struct usched *p_usched; /* Userland scheduling control */
322 struct vkernel_proc *p_vkernel; /* VKernel support, proc part */
323 int p_numposixlocks; /* number of POSIX locks */
324 void (*p_userret)(void);/* p: return-to-user hook */
326 struct spinlock p_spin; /* Spinlock for LWP access to proc */
327 struct lwkt_token p_token; /* Token for LWP access to proc */
330 #define lwp_wchan lwp_thread->td_wchan
331 #define lwp_wmesg lwp_thread->td_wmesg
332 #define p_session p_pgrp->pg_session
333 #define p_pgid p_pgrp->pg_id
335 /* These flags are kept in p_flags. */
336 #define P_ADVLOCK 0x00001 /* Process may hold a POSIX advisory lock. */
337 #define P_CONTROLT 0x00002 /* Has a controlling terminal. */
338 #define P_SWAPPEDOUT 0x00004 /* Swapped out of memory */
339 #define P_UNUSED3 0x00008 /* was: Event pending, break tsleep on sigcont */
340 #define P_PPWAIT 0x00010 /* Parent is waiting for child to exec/exit. */
341 #define P_PROFIL 0x00020 /* Has started profiling. */
342 #define P_UNUSED5 0x00040 /* was: Selecting; wakeup/waiting danger. */
343 #define P_UNUSED4 0x00080 /* was: Sleep is interruptible. */
344 #define P_SUGID 0x00100 /* Had set id privileges since last exec. */
345 #define P_SYSTEM 0x00200 /* System proc: no sigs, stats or swapping. */
346 #define P_UNUSED2 0x00400 /* was: SIGSTOP status */
347 #define P_TRACED 0x00800 /* Debugged process being traced. */
348 #define P_WAITED 0x01000 /* SIGSTOP status was returned by wait3/4 */
349 #define P_WEXIT 0x02000 /* Working on exiting (master exit) */
350 #define P_EXEC 0x04000 /* Process called exec. */
351 #define P_CONTINUED 0x08000 /* Proc has continued from a stopped state. */
353 /* Should probably be changed into a hold count. */
354 /* was P_NOSWAP 0x08000 was: Do not swap upages; p->p_hold */
355 #define P_MAILBOX 0x10000 /* Possible mailbox signal pending */
357 #define P_UPCALLPEND 0x20000 /* an upcall is pending */
359 #define P_SWAPWAIT 0x40000 /* Waiting for a swapin */
360 #define P_UNUSED6 0x80000 /* was: Now in a zombied state */
362 /* Marked a kernel thread */
363 #define P_UNUSED07 0x100000 /* was: on a user scheduling run queue */
364 #define P_KTHREADP 0x200000 /* Process is really a kernel thread */
365 #define P_IDLESWAP 0x400000 /* Swapout was due to idleswap, not load */
366 #define P_DEADLKTREAT 0x800000 /* lock aquisition - deadlock treatment */
368 #define P_JAILED 0x1000000 /* Process is in jail */
369 #define P_UNUSED0 0x2000000 /* need to restore mask before pause */
370 #define P_UNUSED1 0x4000000 /* have alternate signal stack */
371 #define P_INEXEC 0x8000000 /* Process is in execve(). */
372 #define P_PASSIVE_ACQ 0x10000000 /* Passive acquire cpu (see kern_switch) */
373 #define P_UPCALLWAIT 0x20000000 /* Wait for upcall or signal */
374 #define P_XCPU 0x40000000 /* SIGXCPU */
377 * LWP_WSTOP: When set the thread will stop prior to return to userland
378 * and has been counted in the process stop-threads-count, but
379 * may still be running in kernel-land.
381 * LWP_WEXIT: When set the thread has been asked to exit and will not return
382 * to userland. p_nthreads will not be decremented until the
383 * thread has actually exited.
385 #define LWP_ALTSTACK 0x0000001 /* have alternate signal stack */
386 #define LWP_OLDMASK 0x0000002 /* need to restore mask before pause */
387 #define LWP_BREAKTSLEEP 0x0000004 /* Event pending, break tsleep on sigcont */
388 #define LWP_SINTR 0x0000008 /* Sleep is interruptible. */
389 #define LWP_SELECT 0x0000010 /* Selecting; wakeup/waiting danger. */
390 #define LWP_ONRUNQ 0x0000020 /* on a user scheduling run queue */
391 #define LWP_WEXIT 0x0000040 /* working on exiting */
392 #define LWP_WSTOP 0x0000080 /* working on stopping */
394 #define FIRST_LWP_IN_PROC(p) RB_FIRST(lwp_rb_tree, &(p)->p_lwp_tree)
395 #define FOREACH_LWP_IN_PROC(lp, p) \
396 RB_FOREACH(lp, lwp_rb_tree, &(p)->p_lwp_tree)
397 #define ONLY_LWP_IN_PROC(p) \
398 (p->p_nthreads != 1 && \
399 (panic("%s: proc %p (pid %d cmd %s) has more than one thread", \
400 __func__, p, p->p_pid, p->p_comm), 1), \
401 RB_ROOT(&p->p_lwp_tree))
404 * We use process IDs <= PID_MAX; PID_MAX + 1 must also fit in a pid_t,
405 * as it is used to represent "no process group".
407 #define PID_MAX 99999
408 #define NO_PID 100000
410 #define SESS_LEADER(p) ((p)->p_session->s_leader == (p))
414 #ifdef MALLOC_DECLARE
415 MALLOC_DECLARE(M_SESSION);
416 MALLOC_DECLARE(M_PROC);
417 MALLOC_DECLARE(M_LWP);
418 MALLOC_DECLARE(M_SUBPROC);
419 MALLOC_DECLARE(M_PARGS);
422 /* for priv_check_cred() */
423 #define NULL_CRED_OKAY 0x2
425 /* Handy macro to determine if p1 can mangle p2 */
427 #define PRISON_CHECK(cr1, cr2) \
428 ((!(cr1)->cr_prison) || (cr1)->cr_prison == (cr2)->cr_prison)
433 extern void stopevent(struct proc*, unsigned int, unsigned int);
434 #define STOPEVENT(p,e,v) \
436 if ((p)->p_stops & (e)) { \
442 * Hold process in memory, don't destruct, used by ktrace, procfs, sigio,
443 * and signaling code (e.g. ksignal()).
447 #define PHOLD(p) atomic_add_int(&(p)->p_lock, 1)
448 #define PRELE(p) atomic_add_int(&(p)->p_lock, -1)
451 * Hold lwp in memory, don't destruct, normally for ptrace/procfs work
452 * atomic ops because they can occur from an IPI.
455 #define LWPHOLD(lp) atomic_add_int(&(lp)->lwp_lock, 1)
456 #define LWPRELE(lp) atomic_add_int(&(lp)->lwp_lock, -1)
458 #define PIDHASH(pid) (&pidhashtbl[(pid) & pidhash])
459 extern LIST_HEAD(pidhashhead, proc) *pidhashtbl;
460 extern u_long pidhash;
462 #define PGRPHASH(pgid) (&pgrphashtbl[(pgid) & pgrphash])
463 extern LIST_HEAD(pgrphashhead, pgrp) *pgrphashtbl;
464 extern u_long pgrphash;
468 #define SET_CURPROC(p) (curproc = (p))
472 extern struct proc proc0; /* Process slot for swapper. */
473 extern struct lwp lwp0; /* LWP slot for swapper. */
474 extern struct thread thread0; /* Thread slot for swapper. */
475 extern int hogticks; /* Limit on kernel cpu hogs. */
476 extern int nprocs, maxproc; /* Current and max number of procs. */
477 extern int maxprocperuid; /* Max procs per uid. */
478 extern int sched_quantum; /* Scheduling quantum in ticks */
480 extern struct proclist allproc; /* List of all processes. */
481 extern struct proclist zombproc; /* List of zombie processes. */
482 extern struct proc *initproc; /* Process slot for init */
483 extern struct thread *pagethread, *updatethread;
486 * Scheduler independant variables. The primary scheduler polling frequency,
487 * the maximum ESTCPU value, and the weighting factor for nice values. A
488 * cpu bound program's estcpu will increase to ESTCPUMAX - 1.
490 #define ESTCPUFREQ 50
492 extern u_long ps_arg_cache_limit;
493 extern int ps_argsopen;
494 extern int ps_showallprocs;
496 struct proc *pfind (pid_t); /* Find process by id w/ref */
497 struct proc *pfindn (pid_t); /* Find process by id wo/ref */
498 struct pgrp *pgfind (pid_t); /* Find process group by id w/ref */
499 struct proc *zpfind (pid_t); /* Find zombie process by id w/ref */
500 void pgref (struct pgrp *); /* Ref pgrp preventing disposal */
501 void pgrel (struct pgrp *); /* Deref pgrp & dispose on 1->0 trans */
506 int enterpgrp (struct proc *p, pid_t pgid, int mksess);
507 void proc_add_allproc(struct proc *p);
508 void proc_move_allproc_zombie(struct proc *);
509 void proc_remove_zombie(struct proc *);
510 void allproc_scan(int (*callback)(struct proc *, void *), void *data);
511 void alllwp_scan(int (*callback)(struct lwp *, void *), void *data);
512 void zombproc_scan(int (*callback)(struct proc *, void *), void *data);
513 void fixjobc (struct proc *p, struct pgrp *pgrp, int entering);
514 void updatepcpu(struct lwp *, int, int);
515 int inferior (struct proc *p);
516 int leavepgrp (struct proc *p);
517 void sess_hold(struct session *sp);
518 void sess_rele(struct session *sp);
519 void procinit (void);
520 void relscurproc(struct proc *curp);
521 int p_trespass (struct ucred *cr1, struct ucred *cr2);
522 void setrunnable (struct lwp *);
523 void proc_stop (struct proc *);
524 void proc_unstop (struct proc *);
525 void sleep_gdinit (struct globaldata *);
526 int suser (struct thread *td);
527 int suser_cred (struct ucred *cred, int flag);
528 thread_t cpu_heavy_switch (struct thread *);
529 thread_t cpu_lwkt_switch (struct thread *);
531 void cpu_lwp_exit (void) __dead2;
532 void cpu_thread_exit (void) __dead2;
533 void lwp_exit (int masterexit) __dead2;
534 void lwp_dispose (struct lwp *);
535 int killalllwps (int);
536 void exit1 (int) __dead2;
537 void cpu_fork (struct lwp *, struct lwp *, int);
538 int cpu_prepare_lwp(struct lwp *, struct lwp_params *);
539 void cpu_set_fork_handler (struct lwp *, void (*)(void *, struct trapframe *), void *);
540 void cpu_set_thread_handler(struct thread *td, void (*retfunc)(void), void *func, void *arg);
541 int fork1 (struct lwp *, int, struct proc **);
542 void start_forked_proc (struct lwp *, struct proc *);
543 int trace_req (struct proc *);
544 void cpu_proc_wait (struct proc *);
545 void cpu_thread_wait (struct thread *);
546 void setsugid (void);
547 void faultin (struct proc *p);
548 void swapin_request (void);
550 u_int32_t procrunnable (void);
554 #endif /* _KERNEL || _KERNEL_STRUCTURES */
555 #endif /* !_SYS_PROC_H_ */