proc.h revision 267017
1/*-
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.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
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.
21 *
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
32 * SUCH DAMAGE.
33 *
34 *	@(#)proc.h	8.15 (Berkeley) 5/19/95
35 * $FreeBSD: releng/10.0/sys/sys/proc.h 267017 2014-06-03 19:02:52Z delphij $
36 */
37
38#ifndef _SYS_PROC_H_
39#define	_SYS_PROC_H_
40
41#include <sys/callout.h>		/* For struct callout. */
42#include <sys/event.h>			/* For struct klist. */
43#include <sys/condvar.h>
44#ifndef _KERNEL
45#include <sys/filedesc.h>
46#endif
47#include <sys/queue.h>
48#include <sys/_lock.h>
49#include <sys/lock_profile.h>
50#include <sys/_mutex.h>
51#include <sys/osd.h>
52#include <sys/priority.h>
53#include <sys/rtprio.h>			/* XXX. */
54#include <sys/runq.h>
55#include <sys/resource.h>
56#include <sys/sigio.h>
57#include <sys/signal.h>
58#include <sys/signalvar.h>
59#ifndef _KERNEL
60#include <sys/time.h>			/* For structs itimerval, timeval. */
61#else
62#include <sys/pcpu.h>
63#endif
64#include <sys/ucontext.h>
65#include <sys/ucred.h>
66#include <machine/proc.h>		/* Machine-dependent proc substruct. */
67
68/*
69 * One structure allocated per session.
70 *
71 * List of locks
72 * (m)		locked by s_mtx mtx
73 * (e)		locked by proctree_lock sx
74 * (c)		const until freeing
75 */
76struct session {
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. */
86};
87
88/*
89 * One structure allocated per process group.
90 *
91 * List of locks
92 * (m)		locked by pg_mtx mtx
93 * (e)		locked by proctree_lock sx
94 * (c)		const until freeing
95 */
96struct pgrp {
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 */
104};
105
106/*
107 * pargs, used to hold a copy of the command line, if it had a sane length.
108 */
109struct pargs {
110	u_int	ar_ref;		/* Reference count. */
111	u_int	ar_length;	/* Length. */
112	u_char	ar_args[1];	/* Arguments. */
113};
114
115/*-
116 * Description of a process.
117 *
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
124 * is running.
125 *
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
137 *      f - session mtx
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
145 *      m - Giant
146 *      n - not locked, lazy
147 *      o - ktrace lock
148 *      q - td_contested lock
149 *      r - p_peers lock
150 *      t - thread 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
155 *
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
158 * for write access.
159 */
160struct cpuset;
161struct kaioinfo;
162struct kaudit_record;
163struct kdtrace_proc;
164struct kdtrace_thread;
165struct mqueue_notifier;
166struct nlminfo;
167struct p_sched;
168struct proc;
169struct procdesc;
170struct racct;
171struct sbuf;
172struct sleepqueue;
173struct td_sched;
174struct thread;
175struct trapframe;
176struct turnstile;
177
178/*
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.
185 *
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.
188 */
189struct rusage_ext {
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. */
197};
198
199/*
200 * Kernel runnable context (thread).
201 * This is what is put to sleep and reactivated.
202 * Thread context.  Processes may have multiple threads.
203 */
204struct thread {
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 rl_q_entry *td_rlqe;	/* (k) Associated range lock entry. */
217	struct umtx_q   *td_umtxq;	/* (c?) Link for when we're blocked. */
218	lwpid_t		td_tid;		/* (b) Thread ID. */
219	sigqueue_t	td_sigqueue;	/* (c) Sigs arrived, not delivered. */
220#define	td_siglist	td_sigqueue.sq_signals
221	u_char		td_lend_user_pri; /* (t) Lend user pri. */
222
223/* Cleared during fork1() */
224#define	td_startzero td_flags
225	int		td_flags;	/* (t) TDF_* flags. */
226	int		td_inhibitors;	/* (t) Why can not run. */
227	int		td_pflags;	/* (k) Private thread (TDP_*) flags. */
228	int		td_dupfd;	/* (k) Ret value from fdopen. XXX */
229	int		td_sqqueue;	/* (t) Sleepqueue queue blocked on. */
230	void		*td_wchan;	/* (t) Sleep address. */
231	const char	*td_wmesg;	/* (t) Reason for sleep. */
232	u_char		td_lastcpu;	/* (t) Last cpu we were on. */
233	u_char		td_oncpu;	/* (t) Which cpu we are on. */
234	volatile u_char td_owepreempt;  /* (k*) Preempt on last critical_exit */
235	u_char		td_tsqueue;	/* (t) Turnstile queue blocked on. */
236	short		td_locks;	/* (k) Count of non-spin locks. */
237	short		td_rw_rlocks;	/* (k) Count of rwlock read locks. */
238	short		td_lk_slocks;	/* (k) Count of lockmgr shared locks. */
239	short		td_stopsched;	/* (k) Scheduler stopped. */
240	struct turnstile *td_blocked;	/* (t) Lock thread is blocked on. */
241	const char	*td_lockname;	/* (t) Name of lock blocked on. */
242	LIST_HEAD(, turnstile) td_contested;	/* (q) Contested locks. */
243	struct lock_list_entry *td_sleeplocks; /* (k) Held sleep locks. */
244	int		td_intr_nesting_level; /* (k) Interrupt recursion. */
245	int		td_pinned;	/* (k) Temporary cpu pin count. */
246	struct ucred	*td_ucred;	/* (k) Reference to credentials. */
247	u_int		td_estcpu;	/* (t) estimated cpu utilization */
248	int		td_slptick;	/* (t) Time at sleep. */
249	int		td_blktick;	/* (t) Time spent blocked. */
250	int		td_swvoltick;	/* (t) Time at last SW_VOL switch. */
251	u_int		td_cow;		/* (*) Number of copy-on-write faults */
252	struct rusage	td_ru;		/* (t) rusage information. */
253	struct rusage_ext td_rux;	/* (t) Internal rusage information. */
254	uint64_t	td_incruntime;	/* (t) Cpu ticks to transfer to proc. */
255	uint64_t	td_runtime;	/* (t) How many cpu ticks we've run. */
256	u_int 		td_pticks;	/* (t) Statclock hits for profiling */
257	u_int		td_sticks;	/* (t) Statclock hits in system mode. */
258	u_int		td_iticks;	/* (t) Statclock hits in intr mode. */
259	u_int		td_uticks;	/* (t) Statclock hits in user mode. */
260	int		td_intrval;	/* (t) Return value for sleepq. */
261	sigset_t	td_oldsigmask;	/* (k) Saved mask from pre sigpause. */
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	u_int		td_vp_reserv;	/* (k) Count of reserved vnodes. */
276	int		td_no_sleeping;	/* (k) Sleeping disabled count. */
277	int		td_dom_rr_idx;	/* (k) RR Numa domain selection. */
278#define	td_endzero td_sigmask
279
280/* Copied during fork1() or create_thread(). */
281#define	td_startcopy td_endzero
282	sigset_t	td_sigmask;	/* (c) Current signal mask. */
283	u_char		td_rqindex;	/* (t) Run queue index. */
284	u_char		td_base_pri;	/* (t) Thread base kernel priority. */
285	u_char		td_priority;	/* (t) Thread active priority. */
286	u_char		td_pri_class;	/* (t) Scheduling class. */
287	u_char		td_user_pri;	/* (t) User pri from estcpu and nice. */
288	u_char		td_base_user_pri; /* (t) Base user pri */
289#define	td_endcopy td_pcb
290
291/*
292 * Fields that must be manually set in fork1() or create_thread()
293 * or already have been set in the allocator, constructor, etc.
294 */
295	struct pcb	*td_pcb;	/* (k) Kernel VA of pcb and kstack. */
296	enum {
297		TDS_INACTIVE = 0x0,
298		TDS_INHIBITED,
299		TDS_CAN_RUN,
300		TDS_RUNQ,
301		TDS_RUNNING
302	} td_state;			/* (t) thread state */
303	register_t	td_retval[2];	/* (k) Syscall aux returns. */
304	struct callout	td_slpcallout;	/* (h) Callout for sleep. */
305	struct trapframe *td_frame;	/* (k) */
306	struct vm_object *td_kstack_obj;/* (a) Kstack object. */
307	vm_offset_t	td_kstack;	/* (a) Kernel VA of kstack. */
308	int		td_kstack_pages; /* (a) Size of the kstack. */
309	volatile u_int	td_critnest;	/* (k*) Critical section nest level. */
310	struct mdthread td_md;		/* (k) Any machine-dependent fields. */
311	struct td_sched	*td_sched;	/* (*) Scheduler-specific data. */
312	struct kaudit_record	*td_ar;	/* (k) Active audit record, if any. */
313	struct lpohead	td_lprof[2];	/* (a) lock profiling objects. */
314	struct kdtrace_thread	*td_dtrace; /* (*) DTrace-specific data. */
315	int		td_errno;	/* Error returned by last syscall. */
316	struct vnet	*td_vnet;	/* (k) Effective vnet. */
317	const char	*td_vnet_lpush;	/* (k) Debugging vnet push / pop. */
318	struct trapframe *td_intr_frame;/* (k) Frame of the current irq */
319	struct proc	*td_rfppwait_p;	/* (k) The vforked child */
320	struct vm_page	**td_ma;	/* (k) uio pages held */
321	int		td_ma_cnt;	/* (k) size of *td_ma */
322};
323
324struct mtx *thread_lock_block(struct thread *);
325void thread_lock_unblock(struct thread *, struct mtx *);
326void thread_lock_set(struct thread *, struct mtx *);
327#define	THREAD_LOCK_ASSERT(td, type)					\
328do {									\
329	struct mtx *__m = (td)->td_lock;				\
330	if (__m != &blocked_lock)					\
331		mtx_assert(__m, (type));				\
332} while (0)
333
334#ifdef INVARIANTS
335#define	THREAD_LOCKPTR_ASSERT(td, lock)					\
336do {									\
337	struct mtx *__m = (td)->td_lock;				\
338	KASSERT((__m == &blocked_lock || __m == (lock)),		\
339	    ("Thread %p lock %p does not match %p", td, __m, (lock)));	\
340} while (0)
341#else
342#define	THREAD_LOCKPTR_ASSERT(td, lock)
343#endif
344
345/*
346 * Flags kept in td_flags:
347 * To change these you MUST have the scheduler lock.
348 */
349#define	TDF_BORROWING	0x00000001 /* Thread is borrowing pri from another. */
350#define	TDF_INPANIC	0x00000002 /* Caused a panic, let it drive crashdump. */
351#define	TDF_INMEM	0x00000004 /* Thread's stack is in memory. */
352#define	TDF_SINTR	0x00000008 /* Sleep is interruptible. */
353#define	TDF_TIMEOUT	0x00000010 /* Timing out during sleep. */
354#define	TDF_IDLETD	0x00000020 /* This is a per-CPU idle thread. */
355#define	TDF_CANSWAP	0x00000040 /* Thread can be swapped. */
356#define	TDF_SLEEPABORT	0x00000080 /* sleepq_abort was called. */
357#define	TDF_KTH_SUSP	0x00000100 /* kthread is suspended */
358#define	TDF_UNUSED09	0x00000200 /* --available-- */
359#define	TDF_BOUNDARY	0x00000400 /* Thread suspended at user boundary */
360#define	TDF_ASTPENDING	0x00000800 /* Thread has some asynchronous events. */
361#define	TDF_TIMOFAIL	0x00001000 /* Timeout from sleep after we were awake. */
362#define	TDF_SBDRY	0x00002000 /* Stop only on usermode boundary. */
363#define	TDF_UPIBLOCKED	0x00004000 /* Thread blocked on user PI mutex. */
364#define	TDF_NEEDSUSPCHK	0x00008000 /* Thread may need to suspend. */
365#define	TDF_NEEDRESCHED	0x00010000 /* Thread needs to yield. */
366#define	TDF_NEEDSIGCHK	0x00020000 /* Thread may need signal delivery. */
367#define	TDF_NOLOAD	0x00040000 /* Ignore during load avg calculations. */
368#define	TDF_UNUSED19	0x00080000 /* --available-- */
369#define	TDF_THRWAKEUP	0x00100000 /* Libthr thread must not suspend itself. */
370#define	TDF_UNUSED21	0x00200000 /* --available-- */
371#define	TDF_SWAPINREQ	0x00400000 /* Swapin request due to wakeup. */
372#define	TDF_UNUSED23	0x00800000 /* --available-- */
373#define	TDF_SCHED0	0x01000000 /* Reserved for scheduler private use */
374#define	TDF_SCHED1	0x02000000 /* Reserved for scheduler private use */
375#define	TDF_SCHED2	0x04000000 /* Reserved for scheduler private use */
376#define	TDF_SCHED3	0x08000000 /* Reserved for scheduler private use */
377#define	TDF_ALRMPEND	0x10000000 /* Pending SIGVTALRM needs to be posted. */
378#define	TDF_PROFPEND	0x20000000 /* Pending SIGPROF needs to be posted. */
379#define	TDF_MACPEND	0x40000000 /* AST-based MAC event pending. */
380
381/* Userland debug flags */
382#define	TDB_SUSPEND	0x00000001 /* Thread is suspended by debugger */
383#define	TDB_XSIG	0x00000002 /* Thread is exchanging signal under trace */
384#define	TDB_USERWR	0x00000004 /* Debugger modified memory or registers */
385#define	TDB_SCE		0x00000008 /* Thread performs syscall enter */
386#define	TDB_SCX		0x00000010 /* Thread performs syscall exit */
387#define	TDB_EXEC	0x00000020 /* TDB_SCX from exec(2) family */
388#define	TDB_FORK	0x00000040 /* TDB_SCX from fork(2) that created new
389				      process */
390#define	TDB_STOPATFORK	0x00000080 /* Stop at the return from fork (child
391				      only) */
392#define	TDB_CHILD	0x00000100 /* New child indicator for ptrace() */
393
394/*
395 * "Private" flags kept in td_pflags:
396 * These are only written by curthread and thus need no locking.
397 */
398#define	TDP_OLDMASK	0x00000001 /* Need to restore mask after suspend. */
399#define	TDP_INKTR	0x00000002 /* Thread is currently in KTR code. */
400#define	TDP_INKTRACE	0x00000004 /* Thread is currently in KTRACE code. */
401#define	TDP_BUFNEED	0x00000008 /* Do not recurse into the buf flush */
402#define	TDP_COWINPROGRESS 0x00000010 /* Snapshot copy-on-write in progress. */
403#define	TDP_ALTSTACK	0x00000020 /* Have alternate signal stack. */
404#define	TDP_DEADLKTREAT	0x00000040 /* Lock aquisition - deadlock treatment. */
405#define	TDP_NOFAULTING	0x00000080 /* Do not handle page faults. */
406#define	TDP_UNUSED9	0x00000100 /* --available-- */
407#define	TDP_OWEUPC	0x00000200 /* Call addupc() at next AST. */
408#define	TDP_ITHREAD	0x00000400 /* Thread is an interrupt thread. */
409#define	TDP_SYNCIO	0x00000800 /* Local override, disable async i/o. */
410#define	TDP_SCHED1	0x00001000 /* Reserved for scheduler private use */
411#define	TDP_SCHED2	0x00002000 /* Reserved for scheduler private use */
412#define	TDP_SCHED3	0x00004000 /* Reserved for scheduler private use */
413#define	TDP_SCHED4	0x00008000 /* Reserved for scheduler private use */
414#define	TDP_GEOM	0x00010000 /* Settle GEOM before finishing syscall */
415#define	TDP_SOFTDEP	0x00020000 /* Stuck processing softdep worklist */
416#define	TDP_NORUNNINGBUF 0x00040000 /* Ignore runningbufspace check */
417#define	TDP_WAKEUP	0x00080000 /* Don't sleep in umtx cond_wait */
418#define	TDP_INBDFLUSH	0x00100000 /* Already in BO_BDFLUSH, do not recurse */
419#define	TDP_KTHREAD	0x00200000 /* This is an official kernel thread */
420#define	TDP_CALLCHAIN	0x00400000 /* Capture thread's callchain */
421#define	TDP_IGNSUSP	0x00800000 /* Permission to ignore the MNTK_SUSPEND* */
422#define	TDP_AUDITREC	0x01000000 /* Audit record pending on thread */
423#define	TDP_RFPPWAIT	0x02000000 /* Handle RFPPWAIT on syscall exit */
424#define	TDP_RESETSPUR	0x04000000 /* Reset spurious page fault history. */
425#define	TDP_NERRNO	0x08000000 /* Last errno is already in td_errno */
426#define	TDP_UIOHELD	0x10000000 /* Current uio has pages held in td_ma */
427
428/*
429 * Reasons that the current thread can not be run yet.
430 * More than one may apply.
431 */
432#define	TDI_SUSPENDED	0x0001	/* On suspension queue. */
433#define	TDI_SLEEPING	0x0002	/* Actually asleep! (tricky). */
434#define	TDI_SWAPPED	0x0004	/* Stack not in mem.  Bad juju if run. */
435#define	TDI_LOCK	0x0008	/* Stopped on a lock. */
436#define	TDI_IWAIT	0x0010	/* Awaiting interrupt. */
437
438#define	TD_IS_SLEEPING(td)	((td)->td_inhibitors & TDI_SLEEPING)
439#define	TD_ON_SLEEPQ(td)	((td)->td_wchan != NULL)
440#define	TD_IS_SUSPENDED(td)	((td)->td_inhibitors & TDI_SUSPENDED)
441#define	TD_IS_SWAPPED(td)	((td)->td_inhibitors & TDI_SWAPPED)
442#define	TD_ON_LOCK(td)		((td)->td_inhibitors & TDI_LOCK)
443#define	TD_AWAITING_INTR(td)	((td)->td_inhibitors & TDI_IWAIT)
444#define	TD_IS_RUNNING(td)	((td)->td_state == TDS_RUNNING)
445#define	TD_ON_RUNQ(td)		((td)->td_state == TDS_RUNQ)
446#define	TD_CAN_RUN(td)		((td)->td_state == TDS_CAN_RUN)
447#define	TD_IS_INHIBITED(td)	((td)->td_state == TDS_INHIBITED)
448#define	TD_ON_UPILOCK(td)	((td)->td_flags & TDF_UPIBLOCKED)
449#define TD_IS_IDLETHREAD(td)	((td)->td_flags & TDF_IDLETD)
450
451
452#define	TD_SET_INHIB(td, inhib) do {			\
453	(td)->td_state = TDS_INHIBITED;			\
454	(td)->td_inhibitors |= (inhib);			\
455} while (0)
456
457#define	TD_CLR_INHIB(td, inhib) do {			\
458	if (((td)->td_inhibitors & (inhib)) &&		\
459	    (((td)->td_inhibitors &= ~(inhib)) == 0))	\
460		(td)->td_state = TDS_CAN_RUN;		\
461} while (0)
462
463#define	TD_SET_SLEEPING(td)	TD_SET_INHIB((td), TDI_SLEEPING)
464#define	TD_SET_SWAPPED(td)	TD_SET_INHIB((td), TDI_SWAPPED)
465#define	TD_SET_LOCK(td)		TD_SET_INHIB((td), TDI_LOCK)
466#define	TD_SET_SUSPENDED(td)	TD_SET_INHIB((td), TDI_SUSPENDED)
467#define	TD_SET_IWAIT(td)	TD_SET_INHIB((td), TDI_IWAIT)
468#define	TD_SET_EXITING(td)	TD_SET_INHIB((td), TDI_EXITING)
469
470#define	TD_CLR_SLEEPING(td)	TD_CLR_INHIB((td), TDI_SLEEPING)
471#define	TD_CLR_SWAPPED(td)	TD_CLR_INHIB((td), TDI_SWAPPED)
472#define	TD_CLR_LOCK(td)		TD_CLR_INHIB((td), TDI_LOCK)
473#define	TD_CLR_SUSPENDED(td)	TD_CLR_INHIB((td), TDI_SUSPENDED)
474#define	TD_CLR_IWAIT(td)	TD_CLR_INHIB((td), TDI_IWAIT)
475
476#define	TD_SET_RUNNING(td)	(td)->td_state = TDS_RUNNING
477#define	TD_SET_RUNQ(td)		(td)->td_state = TDS_RUNQ
478#define	TD_SET_CAN_RUN(td)	(td)->td_state = TDS_CAN_RUN
479
480/*
481 * Process structure.
482 */
483struct proc {
484	LIST_ENTRY(proc) p_list;	/* (d) List of all processes. */
485	TAILQ_HEAD(, thread) p_threads;	/* (c) all threads. */
486	struct mtx	p_slock;	/* process spin lock */
487	struct ucred	*p_ucred;	/* (c) Process owner's identity. */
488	struct filedesc	*p_fd;		/* (b) Open files. */
489	struct filedesc_to_leader *p_fdtol; /* (b) Tracking node */
490	struct pstats	*p_stats;	/* (b) Accounting/statistics (CPU). */
491	struct plimit	*p_limit;	/* (c) Process limits. */
492	struct callout	p_limco;	/* (c) Limit callout handle */
493	struct sigacts	*p_sigacts;	/* (x) Signal actions, state (CPU). */
494
495	int		p_flag;		/* (c) P_* flags. */
496	int		p_flag2;	/* (c) P2_* flags. */
497	enum {
498		PRS_NEW = 0,		/* In creation */
499		PRS_NORMAL,		/* threads can be run. */
500		PRS_ZOMBIE
501	} p_state;			/* (j/c) Process status. */
502	pid_t		p_pid;		/* (b) Process identifier. */
503	LIST_ENTRY(proc) p_hash;	/* (d) Hash chain. */
504	LIST_ENTRY(proc) p_pglist;	/* (g + e) List of processes in pgrp. */
505	struct proc	*p_pptr;	/* (c + e) Pointer to parent process. */
506	LIST_ENTRY(proc) p_sibling;	/* (e) List of sibling processes. */
507	LIST_HEAD(, proc) p_children;	/* (e) Pointer to list of children. */
508	struct mtx	p_mtx;		/* (n) Lock for this struct. */
509	struct ksiginfo *p_ksi;	/* Locked by parent proc lock */
510	sigqueue_t	p_sigqueue;	/* (c) Sigs not delivered to a td. */
511#define p_siglist	p_sigqueue.sq_signals
512
513/* The following fields are all zeroed upon creation in fork. */
514#define	p_startzero	p_oppid
515	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
548
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
563
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
580					   after fork. */
581	uint64_t	p_prev_runtime;	/* (c) Resource usage accounting. */
582	struct racct	*p_racct;	/* (b) Resource accounting. */
583	u_char		p_throttled;	/* (c) Flag for racct pcpu throttling */
584	/*
585	 * An orphan is the child that has beed re-parented to the
586	 * debugger as a result of attaching to it.  Need to keep
587	 * track of them for parent to be able to collect the exit
588	 * status of what used to be children.
589	 */
590	LIST_ENTRY(proc) p_orphan;	/* (e) List of orphan processes. */
591	LIST_HEAD(, proc) p_orphans;	/* (e) Pointer to list of orphans. */
592};
593
594#define	p_session	p_pgrp->pg_session
595#define	p_pgid		p_pgrp->pg_id
596
597#define	NOCPU	0xff		/* For when we aren't on a CPU. */
598
599#define	PROC_SLOCK(p)	mtx_lock_spin(&(p)->p_slock)
600#define	PROC_SUNLOCK(p)	mtx_unlock_spin(&(p)->p_slock)
601#define	PROC_SLOCK_ASSERT(p, type)	mtx_assert(&(p)->p_slock, (type))
602
603/* These flags are kept in p_flag. */
604#define	P_ADVLOCK	0x00001	/* Process may hold a POSIX advisory lock. */
605#define	P_CONTROLT	0x00002	/* Has a controlling terminal. */
606#define	P_KTHREAD	0x00004	/* Kernel thread (*). */
607#define	P_FOLLOWFORK	0x00008	/* Attach parent debugger to children. */
608#define	P_PPWAIT	0x00010	/* Parent is waiting for child to exec/exit. */
609#define	P_PROFIL	0x00020	/* Has started profiling. */
610#define	P_STOPPROF	0x00040	/* Has thread requesting to stop profiling. */
611#define	P_HADTHREADS	0x00080	/* Has had threads (no cleanup shortcuts) */
612#define	P_SUGID		0x00100	/* Had set id privileges since last exec. */
613#define	P_SYSTEM	0x00200	/* System proc: no sigs, stats or swapping. */
614#define	P_SINGLE_EXIT	0x00400	/* Threads suspending should exit, not wait. */
615#define	P_TRACED	0x00800	/* Debugged process being traced. */
616#define	P_WAITED	0x01000	/* Someone is waiting for us. */
617#define	P_WEXIT		0x02000	/* Working on exiting. */
618#define	P_EXEC		0x04000	/* Process called exec. */
619#define	P_WKILLED	0x08000	/* Killed, go to kernel/user boundary ASAP. */
620#define	P_CONTINUED	0x10000	/* Proc has continued from a stopped state. */
621#define	P_STOPPED_SIG	0x20000	/* Stopped due to SIGSTOP/SIGTSTP. */
622#define	P_STOPPED_TRACE	0x40000	/* Stopped because of tracing. */
623#define	P_STOPPED_SINGLE 0x80000 /* Only 1 thread can continue (not to user). */
624#define	P_PROTECTED	0x100000 /* Do not kill on memory overcommit. */
625#define	P_SIGEVENT	0x200000 /* Process pending signals changed. */
626#define	P_SINGLE_BOUNDARY 0x400000 /* Threads should suspend at user boundary. */
627#define	P_HWPMC		0x800000 /* Process is using HWPMCs */
628#define	P_JAILED	0x1000000 /* Process is in jail. */
629#define	P_ORPHAN	0x2000000 /* Orphaned. */
630#define	P_INEXEC	0x4000000 /* Process is in execve(). */
631#define	P_STATCHILD	0x8000000 /* Child process stopped or exited. */
632#define	P_INMEM		0x10000000 /* Loaded into memory. */
633#define	P_SWAPPINGOUT	0x20000000 /* Process is being swapped out. */
634#define	P_SWAPPINGIN	0x40000000 /* Process is being swapped in. */
635#define	P_PPTRACE	0x80000000 /* PT_TRACEME by vforked child. */
636
637#define	P_STOPPED	(P_STOPPED_SIG|P_STOPPED_SINGLE|P_STOPPED_TRACE)
638#define	P_SHOULDSTOP(p)	((p)->p_flag & P_STOPPED)
639#define	P_KILLED(p)	((p)->p_flag & P_WKILLED)
640
641/* These flags are kept in p_flag2. */
642#define	P2_INHERIT_PROTECTED 0x00000001 /* New children get P_PROTECTED. */
643
644/*
645 * These were process status values (p_stat), now they are only used in
646 * legacy conversion code.
647 */
648#define	SIDL	1		/* Process being created by fork. */
649#define	SRUN	2		/* Currently runnable. */
650#define	SSLEEP	3		/* Sleeping on an address. */
651#define	SSTOP	4		/* Process debugging or suspension. */
652#define	SZOMB	5		/* Awaiting collection by parent. */
653#define	SWAIT	6		/* Waiting for interrupt. */
654#define	SLOCK	7		/* Blocked on a lock. */
655
656#define	P_MAGIC		0xbeefface
657
658#ifdef _KERNEL
659
660/* Types and flags for mi_switch(). */
661#define	SW_TYPE_MASK		0xff	/* First 8 bits are switch type */
662#define	SWT_NONE		0	/* Unspecified switch. */
663#define	SWT_PREEMPT		1	/* Switching due to preemption. */
664#define	SWT_OWEPREEMPT		2	/* Switching due to opepreempt. */
665#define	SWT_TURNSTILE		3	/* Turnstile contention. */
666#define	SWT_SLEEPQ		4	/* Sleepq wait. */
667#define	SWT_SLEEPQTIMO		5	/* Sleepq timeout wait. */
668#define	SWT_RELINQUISH		6	/* yield call. */
669#define	SWT_NEEDRESCHED		7	/* NEEDRESCHED was set. */
670#define	SWT_IDLE		8	/* Switching from the idle thread. */
671#define	SWT_IWAIT		9	/* Waiting for interrupts. */
672#define	SWT_SUSPEND		10	/* Thread suspended. */
673#define	SWT_REMOTEPREEMPT	11	/* Remote processor preempted. */
674#define	SWT_REMOTEWAKEIDLE	12	/* Remote processor preempted idle. */
675#define	SWT_COUNT		13	/* Number of switch types. */
676/* Flags */
677#define	SW_VOL		0x0100		/* Voluntary switch. */
678#define	SW_INVOL	0x0200		/* Involuntary switch. */
679#define SW_PREEMPT	0x0400		/* The invol switch is a preemption */
680
681/* How values for thread_single(). */
682#define	SINGLE_NO_EXIT	0
683#define	SINGLE_EXIT	1
684#define	SINGLE_BOUNDARY	2
685
686#ifdef MALLOC_DECLARE
687MALLOC_DECLARE(M_PARGS);
688MALLOC_DECLARE(M_PGRP);
689MALLOC_DECLARE(M_SESSION);
690MALLOC_DECLARE(M_SUBPROC);
691#endif
692
693#define	FOREACH_PROC_IN_SYSTEM(p)					\
694	LIST_FOREACH((p), &allproc, p_list)
695#define	FOREACH_THREAD_IN_PROC(p, td)					\
696	TAILQ_FOREACH((td), &(p)->p_threads, td_plist)
697
698#define	FIRST_THREAD_IN_PROC(p)	TAILQ_FIRST(&(p)->p_threads)
699
700/*
701 * We use process IDs <= pid_max <= PID_MAX; PID_MAX + 1 must also fit
702 * in a pid_t, as it is used to represent "no process group".
703 */
704#define	PID_MAX		99999
705#define	NO_PID		100000
706extern pid_t pid_max;
707
708#define	SESS_LEADER(p)	((p)->p_session->s_leader == (p))
709
710
711#define	STOPEVENT(p, e, v) do {						\
712	if ((p)->p_stops & (e))	{					\
713		PROC_LOCK(p);						\
714		stopevent((p), (e), (v));				\
715		PROC_UNLOCK(p);						\
716	}								\
717} while (0)
718#define	_STOPEVENT(p, e, v) do {					\
719	PROC_LOCK_ASSERT(p, MA_OWNED);					\
720	WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, &p->p_mtx.lock_object, \
721 	    "checking stopevent %d", (e));				\
722	if ((p)->p_stops & (e))						\
723		stopevent((p), (e), (v));				\
724} while (0)
725
726/* Lock and unlock a process. */
727#define	PROC_LOCK(p)	mtx_lock(&(p)->p_mtx)
728#define	PROC_TRYLOCK(p)	mtx_trylock(&(p)->p_mtx)
729#define	PROC_UNLOCK(p)	mtx_unlock(&(p)->p_mtx)
730#define	PROC_LOCKED(p)	mtx_owned(&(p)->p_mtx)
731#define	PROC_LOCK_ASSERT(p, type)	mtx_assert(&(p)->p_mtx, (type))
732
733/* Lock and unlock a process group. */
734#define	PGRP_LOCK(pg)	mtx_lock(&(pg)->pg_mtx)
735#define	PGRP_UNLOCK(pg)	mtx_unlock(&(pg)->pg_mtx)
736#define	PGRP_LOCKED(pg)	mtx_owned(&(pg)->pg_mtx)
737#define	PGRP_LOCK_ASSERT(pg, type)	mtx_assert(&(pg)->pg_mtx, (type))
738
739#define	PGRP_LOCK_PGSIGNAL(pg) do {					\
740	if ((pg) != NULL)						\
741		PGRP_LOCK(pg);						\
742} while (0)
743#define	PGRP_UNLOCK_PGSIGNAL(pg) do {					\
744	if ((pg) != NULL)						\
745		PGRP_UNLOCK(pg);					\
746} while (0)
747
748/* Lock and unlock a session. */
749#define	SESS_LOCK(s)	mtx_lock(&(s)->s_mtx)
750#define	SESS_UNLOCK(s)	mtx_unlock(&(s)->s_mtx)
751#define	SESS_LOCKED(s)	mtx_owned(&(s)->s_mtx)
752#define	SESS_LOCK_ASSERT(s, type)	mtx_assert(&(s)->s_mtx, (type))
753
754/* Hold process U-area in memory, normally for ptrace/procfs work. */
755#define	PHOLD(p) do {							\
756	PROC_LOCK(p);							\
757	_PHOLD(p);							\
758	PROC_UNLOCK(p);							\
759} while (0)
760#define	_PHOLD(p) do {							\
761	PROC_LOCK_ASSERT((p), MA_OWNED);				\
762	KASSERT(!((p)->p_flag & P_WEXIT) || (p) == curproc,		\
763	    ("PHOLD of exiting process"));				\
764	(p)->p_lock++;							\
765	if (((p)->p_flag & P_INMEM) == 0)				\
766		faultin((p));						\
767} while (0)
768#define PROC_ASSERT_HELD(p) do {					\
769	KASSERT((p)->p_lock > 0, ("process not held"));			\
770} while (0)
771
772#define	PRELE(p) do {							\
773	PROC_LOCK((p));							\
774	_PRELE((p));							\
775	PROC_UNLOCK((p));						\
776} while (0)
777#define	_PRELE(p) do {							\
778	PROC_LOCK_ASSERT((p), MA_OWNED);				\
779	PROC_ASSERT_HELD(p);						\
780	(--(p)->p_lock);						\
781	if (((p)->p_flag & P_WEXIT) && (p)->p_lock == 0)		\
782		wakeup(&(p)->p_lock);					\
783} while (0)
784#define PROC_ASSERT_NOT_HELD(p) do {					\
785	KASSERT((p)->p_lock == 0, ("process held"));			\
786} while (0)
787
788/* Check whether a thread is safe to be swapped out. */
789#define	thread_safetoswapout(td)	((td)->td_flags & TDF_CANSWAP)
790
791/* Control whether or not it is safe for curthread to sleep. */
792#define	THREAD_NO_SLEEPING()		((curthread)->td_no_sleeping++)
793
794#define	THREAD_SLEEPING_OK()		((curthread)->td_no_sleeping--)
795
796#define	PIDHASH(pid)	(&pidhashtbl[(pid) & pidhash])
797extern LIST_HEAD(pidhashhead, proc) *pidhashtbl;
798extern u_long pidhash;
799#define	TIDHASH(tid)	(&tidhashtbl[(tid) & tidhash])
800extern LIST_HEAD(tidhashhead, thread) *tidhashtbl;
801extern u_long tidhash;
802extern struct rwlock tidhash_lock;
803
804#define	PGRPHASH(pgid)	(&pgrphashtbl[(pgid) & pgrphash])
805extern LIST_HEAD(pgrphashhead, pgrp) *pgrphashtbl;
806extern u_long pgrphash;
807
808extern struct sx allproc_lock;
809extern struct sx proctree_lock;
810extern struct mtx ppeers_lock;
811extern struct proc proc0;		/* Process slot for swapper. */
812extern struct thread thread0;		/* Primary thread in proc0. */
813extern struct vmspace vmspace0;		/* VM space for proc0. */
814extern int hogticks;			/* Limit on kernel cpu hogs. */
815extern int lastpid;
816extern int nprocs, maxproc;		/* Current and max number of procs. */
817extern int maxprocperuid;		/* Max procs per uid. */
818extern u_long ps_arg_cache_limit;
819
820LIST_HEAD(proclist, proc);
821TAILQ_HEAD(procqueue, proc);
822TAILQ_HEAD(threadqueue, thread);
823extern struct proclist allproc;		/* List of all processes. */
824extern struct proclist zombproc;	/* List of zombie processes. */
825extern struct proc *initproc, *pageproc; /* Process slots for init, pager. */
826
827extern struct uma_zone *proc_zone;
828
829struct	proc *pfind(pid_t);		/* Find process by id. */
830struct	proc *pfind_locked(pid_t pid);
831struct	pgrp *pgfind(pid_t);		/* Find process group by id. */
832struct	proc *zpfind(pid_t);		/* Find zombie process by id. */
833
834/*
835 * pget() flags.
836 */
837#define	PGET_HOLD	0x00001	/* Hold the process. */
838#define	PGET_CANSEE	0x00002	/* Check against p_cansee(). */
839#define	PGET_CANDEBUG	0x00004	/* Check against p_candebug(). */
840#define	PGET_ISCURRENT	0x00008	/* Check that the found process is current. */
841#define	PGET_NOTWEXIT	0x00010	/* Check that the process is not in P_WEXIT. */
842#define	PGET_NOTINEXEC	0x00020	/* Check that the process is not in P_INEXEC. */
843#define	PGET_NOTID	0x00040	/* Do not assume tid if pid > PID_MAX. */
844
845#define	PGET_WANTREAD	(PGET_HOLD | PGET_CANDEBUG | PGET_NOTWEXIT)
846
847int	pget(pid_t pid, int flags, struct proc **pp);
848
849void	ast(struct trapframe *framep);
850struct	thread *choosethread(void);
851int	cr_cansignal(struct ucred *cred, struct proc *proc, int signum);
852int	enterpgrp(struct proc *p, pid_t pgid, struct pgrp *pgrp,
853	    struct session *sess);
854int	enterthispgrp(struct proc *p, struct pgrp *pgrp);
855void	faultin(struct proc *p);
856void	fixjobc(struct proc *p, struct pgrp *pgrp, int entering);
857int	fork1(struct thread *, int, int, struct proc **, int *, int);
858void	fork_exit(void (*)(void *, struct trapframe *), void *,
859	    struct trapframe *);
860void	fork_return(struct thread *, struct trapframe *);
861int	inferior(struct proc *p);
862void	kern_yield(int);
863void 	kick_proc0(void);
864int	leavepgrp(struct proc *p);
865int	maybe_preempt(struct thread *td);
866void	maybe_yield(void);
867void	mi_switch(int flags, struct thread *newtd);
868int	p_candebug(struct thread *td, struct proc *p);
869int	p_cansee(struct thread *td, struct proc *p);
870int	p_cansched(struct thread *td, struct proc *p);
871int	p_cansignal(struct thread *td, struct proc *p, int signum);
872int	p_canwait(struct thread *td, struct proc *p);
873struct	pargs *pargs_alloc(int len);
874void	pargs_drop(struct pargs *pa);
875void	pargs_hold(struct pargs *pa);
876int	proc_getargv(struct thread *td, struct proc *p, struct sbuf *sb);
877int	proc_getauxv(struct thread *td, struct proc *p, struct sbuf *sb);
878int	proc_getenvv(struct thread *td, struct proc *p, struct sbuf *sb);
879void	procinit(void);
880void	proc_linkup0(struct proc *p, struct thread *td);
881void	proc_linkup(struct proc *p, struct thread *td);
882void	proc_reap(struct thread *td, struct proc *p, int *status, int options);
883void	proc_reparent(struct proc *child, struct proc *newparent);
884struct	pstats *pstats_alloc(void);
885void	pstats_fork(struct pstats *src, struct pstats *dst);
886void	pstats_free(struct pstats *ps);
887int	securelevel_ge(struct ucred *cr, int level);
888int	securelevel_gt(struct ucred *cr, int level);
889void	sess_hold(struct session *);
890void	sess_release(struct session *);
891int	setrunnable(struct thread *);
892void	setsugid(struct proc *p);
893int	should_yield(void);
894int	sigonstack(size_t sp);
895void	stopevent(struct proc *, u_int, u_int);
896struct	thread *tdfind(lwpid_t, pid_t);
897void	threadinit(void);
898void	tidhash_add(struct thread *);
899void	tidhash_remove(struct thread *);
900void	cpu_idle(int);
901int	cpu_idle_wakeup(int);
902extern	void (*cpu_idle_hook)(sbintime_t);	/* Hook to machdep CPU idler. */
903void	cpu_switch(struct thread *, struct thread *, struct mtx *);
904void	cpu_throw(struct thread *, struct thread *) __dead2;
905void	unsleep(struct thread *);
906void	userret(struct thread *, struct trapframe *);
907
908void	cpu_exit(struct thread *);
909void	exit1(struct thread *, int) __dead2;
910struct syscall_args;
911int	cpu_fetch_syscall_args(struct thread *td, struct syscall_args *sa);
912void	cpu_fork(struct thread *, struct proc *, struct thread *, int);
913void	cpu_set_fork_handler(struct thread *, void (*)(void *), void *);
914void	cpu_set_syscall_retval(struct thread *, int);
915void	cpu_set_upcall(struct thread *td, struct thread *td0);
916void	cpu_set_upcall_kse(struct thread *, void (*)(void *), void *,
917	    stack_t *);
918int	cpu_set_user_tls(struct thread *, void *tls_base);
919void	cpu_thread_alloc(struct thread *);
920void	cpu_thread_clean(struct thread *);
921void	cpu_thread_exit(struct thread *);
922void	cpu_thread_free(struct thread *);
923void	cpu_thread_swapin(struct thread *);
924void	cpu_thread_swapout(struct thread *);
925struct	thread *thread_alloc(int pages);
926int	thread_alloc_stack(struct thread *, int pages);
927void	thread_exit(void) __dead2;
928void	thread_free(struct thread *td);
929void	thread_link(struct thread *td, struct proc *p);
930void	thread_reap(void);
931int	thread_single(int how);
932void	thread_single_end(void);
933void	thread_stash(struct thread *td);
934void	thread_stopped(struct proc *p);
935void	childproc_stopped(struct proc *child, int reason);
936void	childproc_continued(struct proc *child);
937void	childproc_exited(struct proc *child);
938int	thread_suspend_check(int how);
939void	thread_suspend_switch(struct thread *);
940void	thread_suspend_one(struct thread *td);
941void	thread_unlink(struct thread *td);
942void	thread_unsuspend(struct proc *p);
943int	thread_unsuspend_one(struct thread *td);
944void	thread_unthread(struct thread *td);
945void	thread_wait(struct proc *p);
946struct thread	*thread_find(struct proc *p, lwpid_t tid);
947
948static __inline int
949curthread_pflags_set(int flags)
950{
951	struct thread *td;
952	int save;
953
954	td = curthread;
955	save = ~flags | (td->td_pflags & flags);
956	td->td_pflags |= flags;
957	return (save);
958}
959
960static __inline void
961curthread_pflags_restore(int save)
962{
963
964	curthread->td_pflags &= save;
965}
966
967#endif	/* _KERNEL */
968
969#define	TDP_EXECVMSPC	0x40000000 /* Execve destroyed old vmspace */
970#endif	/* !_SYS_PROC_H_ */
971