diff --git a/sys/kern/kern_proc.c b/sys/kern/kern_proc.c
index 77ae6cbddea5..463ecb025317 100644
--- a/sys/kern/kern_proc.c
+++ b/sys/kern/kern_proc.c
@@ -1,3636 +1,3636 @@
/*-
* SPDX-License-Identifier: BSD-3-Clause
*
* Copyright (c) 1982, 1986, 1989, 1991, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
#include "opt_ddb.h"
#include "opt_ktrace.h"
#include "opt_kstack_pages.h"
#include "opt_stack.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/bitstring.h>
#include <sys/conf.h>
#include <sys/elf.h>
#include <sys/eventhandler.h>
#include <sys/exec.h>
#include <sys/fcntl.h>
#include <sys/ipc.h>
#include <sys/jail.h>
#include <sys/kernel.h>
#include <sys/limits.h>
#include <sys/lock.h>
#include <sys/loginclass.h>
#include <sys/malloc.h>
#include <sys/mman.h>
#include <sys/mount.h>
#include <sys/mutex.h>
#include <sys/namei.h>
#include <sys/proc.h>
#include <sys/ptrace.h>
#include <sys/refcount.h>
#include <sys/resourcevar.h>
#include <sys/rwlock.h>
#include <sys/sbuf.h>
#include <sys/sysent.h>
#include <sys/sched.h>
#include <sys/shm.h>
#include <sys/smp.h>
#include <sys/stack.h>
#include <sys/stat.h>
#include <sys/dtrace_bsd.h>
#include <sys/sysctl.h>
#include <sys/filedesc.h>
#include <sys/tty.h>
#include <sys/signalvar.h>
#include <sys/sdt.h>
#include <sys/sx.h>
#include <sys/user.h>
#include <sys/vnode.h>
#include <sys/wait.h>
#ifdef KTRACE
#include <sys/ktrace.h>
#endif
#ifdef DDB
#include <ddb/ddb.h>
#endif
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/vm_extern.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/uma.h>
#include <fs/devfs/devfs.h>
#ifdef COMPAT_FREEBSD32
#include <compat/freebsd32/freebsd32.h>
#include <compat/freebsd32/freebsd32_util.h>
#endif
SDT_PROVIDER_DEFINE(proc);
MALLOC_DEFINE(M_SESSION, "session", "session header");
static MALLOC_DEFINE(M_PROC, "proc", "Proc structures");
MALLOC_DEFINE(M_SUBPROC, "subproc", "Proc sub-structures");
static void doenterpgrp(struct proc *, struct pgrp *);
static void orphanpg(struct pgrp *pg);
static void fill_kinfo_aggregate(struct proc *p, struct kinfo_proc *kp);
static void fill_kinfo_proc_only(struct proc *p, struct kinfo_proc *kp);
static void fill_kinfo_thread(struct thread *td, struct kinfo_proc *kp,
int preferthread);
static void pgdelete(struct pgrp *);
static int pgrp_init(void *mem, int size, int flags);
static int proc_ctor(void *mem, int size, void *arg, int flags);
static void proc_dtor(void *mem, int size, void *arg);
static int proc_init(void *mem, int size, int flags);
static void proc_fini(void *mem, int size);
static void pargs_free(struct pargs *pa);
/*
* Other process lists
*/
struct pidhashhead *pidhashtbl = NULL;
struct sx *pidhashtbl_lock;
u_long pidhash;
u_long pidhashlock;
struct pgrphashhead *pgrphashtbl;
u_long pgrphash;
struct proclist allproc = LIST_HEAD_INITIALIZER(allproc);
struct sx __exclusive_cache_line allproc_lock;
struct sx __exclusive_cache_line proctree_lock;
struct mtx __exclusive_cache_line ppeers_lock;
struct mtx __exclusive_cache_line procid_lock;
uma_zone_t proc_zone;
uma_zone_t pgrp_zone;
/*
* The offset of various fields in struct proc and struct thread.
* These are used by kernel debuggers to enumerate kernel threads and
* processes.
*/
const int proc_off_p_pid = offsetof(struct proc, p_pid);
const int proc_off_p_comm = offsetof(struct proc, p_comm);
const int proc_off_p_list = offsetof(struct proc, p_list);
const int proc_off_p_hash = offsetof(struct proc, p_hash);
const int proc_off_p_threads = offsetof(struct proc, p_threads);
const int thread_off_td_tid = offsetof(struct thread, td_tid);
const int thread_off_td_name = offsetof(struct thread, td_name);
const int thread_off_td_oncpu = offsetof(struct thread, td_oncpu);
const int thread_off_td_pcb = offsetof(struct thread, td_pcb);
const int thread_off_td_plist = offsetof(struct thread, td_plist);
EVENTHANDLER_LIST_DEFINE(process_ctor);
EVENTHANDLER_LIST_DEFINE(process_dtor);
EVENTHANDLER_LIST_DEFINE(process_init);
EVENTHANDLER_LIST_DEFINE(process_fini);
EVENTHANDLER_LIST_DEFINE(process_exit);
EVENTHANDLER_LIST_DEFINE(process_fork);
EVENTHANDLER_LIST_DEFINE(process_exec);
int kstack_pages = KSTACK_PAGES;
SYSCTL_INT(_kern, OID_AUTO, kstack_pages, CTLFLAG_RDTUN | CTLFLAG_NOFETCH,
&kstack_pages, 0,
"Kernel stack size in pages");
static int vmmap_skip_res_cnt = 0;
SYSCTL_INT(_kern, OID_AUTO, proc_vmmap_skip_resident_count, CTLFLAG_RW,
&vmmap_skip_res_cnt, 0,
"Skip calculation of the pages resident count in kern.proc.vmmap");
CTASSERT(sizeof(struct kinfo_proc) == KINFO_PROC_SIZE);
#ifdef COMPAT_FREEBSD32
CTASSERT(sizeof(struct kinfo_proc32) == KINFO_PROC32_SIZE);
#endif
/*
* Initialize global process hashing structures.
*/
void
procinit(void)
{
u_long i;
sx_init(&allproc_lock, "allproc");
sx_init(&proctree_lock, "proctree");
mtx_init(&ppeers_lock, "p_peers", NULL, MTX_DEF);
mtx_init(&procid_lock, "procid", NULL, MTX_DEF);
pidhashtbl = hashinit(maxproc / 4, M_PROC, &pidhash);
pidhashlock = (pidhash + 1) / 64;
if (pidhashlock > 0)
pidhashlock--;
pidhashtbl_lock = malloc(sizeof(*pidhashtbl_lock) * (pidhashlock + 1),
M_PROC, M_WAITOK | M_ZERO);
for (i = 0; i < pidhashlock + 1; i++)
sx_init_flags(&pidhashtbl_lock[i], "pidhash", SX_DUPOK);
pgrphashtbl = hashinit(maxproc / 4, M_PROC, &pgrphash);
proc_zone = uma_zcreate("PROC", sched_sizeof_proc(),
proc_ctor, proc_dtor, proc_init, proc_fini,
UMA_ALIGN_PTR, UMA_ZONE_NOFREE);
pgrp_zone = uma_zcreate("PGRP", sizeof(struct pgrp), NULL, NULL,
pgrp_init, NULL, UMA_ALIGN_PTR, UMA_ZONE_NOFREE);
uihashinit();
}
/*
* Prepare a proc for use.
*/
static int
proc_ctor(void *mem, int size, void *arg, int flags)
{
struct proc *p;
struct thread *td;
p = (struct proc *)mem;
#ifdef KDTRACE_HOOKS
kdtrace_proc_ctor(p);
#endif
EVENTHANDLER_DIRECT_INVOKE(process_ctor, p);
td = FIRST_THREAD_IN_PROC(p);
if (td != NULL) {
/* Make sure all thread constructors are executed */
EVENTHANDLER_DIRECT_INVOKE(thread_ctor, td);
}
return (0);
}
/*
* Reclaim a proc after use.
*/
static void
proc_dtor(void *mem, int size, void *arg)
{
struct proc *p;
struct thread *td;
/* INVARIANTS checks go here */
p = (struct proc *)mem;
td = FIRST_THREAD_IN_PROC(p);
if (td != NULL) {
#ifdef INVARIANTS
KASSERT((p->p_numthreads == 1),
("bad number of threads in exiting process"));
KASSERT(STAILQ_EMPTY(&p->p_ktr), ("proc_dtor: non-empty p_ktr"));
#endif
/* Free all OSD associated to this thread. */
osd_thread_exit(td);
ast_kclear(td);
/* Make sure all thread destructors are executed */
EVENTHANDLER_DIRECT_INVOKE(thread_dtor, td);
}
EVENTHANDLER_DIRECT_INVOKE(process_dtor, p);
#ifdef KDTRACE_HOOKS
kdtrace_proc_dtor(p);
#endif
if (p->p_ksi != NULL)
KASSERT(! KSI_ONQ(p->p_ksi), ("SIGCHLD queue"));
}
/*
* Initialize type-stable parts of a proc (when newly created).
*/
static int
proc_init(void *mem, int size, int flags)
{
struct proc *p;
p = (struct proc *)mem;
mtx_init(&p->p_mtx, "process lock", NULL, MTX_DEF | MTX_DUPOK | MTX_NEW);
mtx_init(&p->p_slock, "process slock", NULL, MTX_SPIN | MTX_NEW);
mtx_init(&p->p_statmtx, "pstatl", NULL, MTX_SPIN | MTX_NEW);
mtx_init(&p->p_itimmtx, "pitiml", NULL, MTX_SPIN | MTX_NEW);
mtx_init(&p->p_profmtx, "pprofl", NULL, MTX_SPIN | MTX_NEW);
cv_init(&p->p_pwait, "ppwait");
TAILQ_INIT(&p->p_threads); /* all threads in proc */
EVENTHANDLER_DIRECT_INVOKE(process_init, p);
p->p_stats = pstats_alloc();
p->p_pgrp = NULL;
TAILQ_INIT(&p->p_kqtim_stop);
return (0);
}
/*
* UMA should ensure that this function is never called.
* Freeing a proc structure would violate type stability.
*/
static void
proc_fini(void *mem, int size)
{
#ifdef notnow
struct proc *p;
p = (struct proc *)mem;
EVENTHANDLER_DIRECT_INVOKE(process_fini, p);
pstats_free(p->p_stats);
thread_free(FIRST_THREAD_IN_PROC(p));
mtx_destroy(&p->p_mtx);
if (p->p_ksi != NULL)
ksiginfo_free(p->p_ksi);
#else
panic("proc reclaimed");
#endif
}
static int
pgrp_init(void *mem, int size, int flags)
{
struct pgrp *pg;
pg = mem;
mtx_init(&pg->pg_mtx, "process group", NULL, MTX_DEF | MTX_DUPOK);
sx_init(&pg->pg_killsx, "killpg racer");
return (0);
}
/*
* PID space management.
*
* These bitmaps are used by fork_findpid.
*/
bitstr_t bit_decl(proc_id_pidmap, PID_MAX);
bitstr_t bit_decl(proc_id_grpidmap, PID_MAX);
bitstr_t bit_decl(proc_id_sessidmap, PID_MAX);
bitstr_t bit_decl(proc_id_reapmap, PID_MAX);
static bitstr_t *proc_id_array[] = {
proc_id_pidmap,
proc_id_grpidmap,
proc_id_sessidmap,
proc_id_reapmap,
};
void
proc_id_set(int type, pid_t id)
{
KASSERT(type >= 0 && type < nitems(proc_id_array),
("invalid type %d\n", type));
mtx_lock(&procid_lock);
KASSERT(bit_test(proc_id_array[type], id) == 0,
("bit %d already set in %d\n", id, type));
bit_set(proc_id_array[type], id);
mtx_unlock(&procid_lock);
}
void
proc_id_set_cond(int type, pid_t id)
{
KASSERT(type >= 0 && type < nitems(proc_id_array),
("invalid type %d\n", type));
if (bit_test(proc_id_array[type], id))
return;
mtx_lock(&procid_lock);
bit_set(proc_id_array[type], id);
mtx_unlock(&procid_lock);
}
void
proc_id_clear(int type, pid_t id)
{
KASSERT(type >= 0 && type < nitems(proc_id_array),
("invalid type %d\n", type));
mtx_lock(&procid_lock);
KASSERT(bit_test(proc_id_array[type], id) != 0,
("bit %d not set in %d\n", id, type));
bit_clear(proc_id_array[type], id);
mtx_unlock(&procid_lock);
}
/*
* Is p an inferior of the current process?
*/
int
inferior(struct proc *p)
{
sx_assert(&proctree_lock, SX_LOCKED);
PROC_LOCK_ASSERT(p, MA_OWNED);
for (; p != curproc; p = proc_realparent(p)) {
if (p->p_pid == 0)
return (0);
}
return (1);
}
/*
* Shared lock all the pid hash lists.
*/
void
pidhash_slockall(void)
{
u_long i;
for (i = 0; i < pidhashlock + 1; i++)
sx_slock(&pidhashtbl_lock[i]);
}
/*
* Shared unlock all the pid hash lists.
*/
void
pidhash_sunlockall(void)
{
u_long i;
for (i = 0; i < pidhashlock + 1; i++)
sx_sunlock(&pidhashtbl_lock[i]);
}
/*
* Similar to pfind(), this function locate a process by number.
*/
struct proc *
pfind_any_locked(pid_t pid)
{
struct proc *p;
sx_assert(PIDHASHLOCK(pid), SX_LOCKED);
LIST_FOREACH(p, PIDHASH(pid), p_hash) {
if (p->p_pid == pid) {
PROC_LOCK(p);
if (p->p_state == PRS_NEW) {
PROC_UNLOCK(p);
p = NULL;
}
break;
}
}
return (p);
}
/*
* Locate a process by number.
*
* By not returning processes in the PRS_NEW state, we allow callers to avoid
* testing for that condition to avoid dereferencing p_ucred, et al.
*/
static __always_inline struct proc *
_pfind(pid_t pid, bool zombie)
{
struct proc *p;
p = curproc;
if (p->p_pid == pid) {
PROC_LOCK(p);
return (p);
}
sx_slock(PIDHASHLOCK(pid));
LIST_FOREACH(p, PIDHASH(pid), p_hash) {
if (p->p_pid == pid) {
PROC_LOCK(p);
if (p->p_state == PRS_NEW ||
(!zombie && p->p_state == PRS_ZOMBIE)) {
PROC_UNLOCK(p);
p = NULL;
}
break;
}
}
sx_sunlock(PIDHASHLOCK(pid));
return (p);
}
struct proc *
pfind(pid_t pid)
{
return (_pfind(pid, false));
}
/*
* Same as pfind but allow zombies.
*/
struct proc *
pfind_any(pid_t pid)
{
return (_pfind(pid, true));
}
/*
* Locate a process group by number.
* The caller must hold proctree_lock.
*/
struct pgrp *
pgfind(pid_t pgid)
{
struct pgrp *pgrp;
sx_assert(&proctree_lock, SX_LOCKED);
LIST_FOREACH(pgrp, PGRPHASH(pgid), pg_hash) {
if (pgrp->pg_id == pgid) {
PGRP_LOCK(pgrp);
return (pgrp);
}
}
return (NULL);
}
/*
* Locate process and do additional manipulations, depending on flags.
*/
int
pget(pid_t pid, int flags, struct proc **pp)
{
struct proc *p;
struct thread *td1;
int error;
p = curproc;
if (p->p_pid == pid) {
PROC_LOCK(p);
} else {
p = NULL;
if (pid <= PID_MAX) {
if ((flags & PGET_NOTWEXIT) == 0)
p = pfind_any(pid);
else
p = pfind(pid);
} else if ((flags & PGET_NOTID) == 0) {
td1 = tdfind(pid, -1);
if (td1 != NULL)
p = td1->td_proc;
}
if (p == NULL)
return (ESRCH);
if ((flags & PGET_CANSEE) != 0) {
error = p_cansee(curthread, p);
if (error != 0)
goto errout;
}
}
if ((flags & PGET_CANDEBUG) != 0) {
error = p_candebug(curthread, p);
if (error != 0)
goto errout;
}
if ((flags & PGET_ISCURRENT) != 0 && curproc != p) {
error = EPERM;
goto errout;
}
if ((flags & PGET_NOTWEXIT) != 0 && (p->p_flag & P_WEXIT) != 0) {
error = ESRCH;
goto errout;
}
if ((flags & PGET_NOTINEXEC) != 0 && (p->p_flag & P_INEXEC) != 0) {
/*
* XXXRW: Not clear ESRCH is the right error during proc
* execve().
*/
error = ESRCH;
goto errout;
}
if ((flags & PGET_HOLD) != 0) {
_PHOLD(p);
PROC_UNLOCK(p);
}
*pp = p;
return (0);
errout:
PROC_UNLOCK(p);
return (error);
}
/*
* Create a new process group.
* pgid must be equal to the pid of p.
* Begin a new session if required.
*/
int
enterpgrp(struct proc *p, pid_t pgid, struct pgrp *pgrp, struct session *sess)
{
struct pgrp *old_pgrp;
sx_assert(&proctree_lock, SX_XLOCKED);
KASSERT(pgrp != NULL, ("enterpgrp: pgrp == NULL"));
KASSERT(p->p_pid == pgid,
("enterpgrp: new pgrp and pid != pgid"));
KASSERT(pgfind(pgid) == NULL,
("enterpgrp: pgrp with pgid exists"));
KASSERT(!SESS_LEADER(p),
("enterpgrp: session leader attempted setpgrp"));
old_pgrp = p->p_pgrp;
if (!sx_try_xlock(&old_pgrp->pg_killsx)) {
sx_xunlock(&proctree_lock);
sx_xlock(&old_pgrp->pg_killsx);
sx_xunlock(&old_pgrp->pg_killsx);
return (ERESTART);
}
MPASS(old_pgrp == p->p_pgrp);
if (sess != NULL) {
/*
* new session
*/
mtx_init(&sess->s_mtx, "session", NULL, MTX_DEF);
PROC_LOCK(p);
p->p_flag &= ~P_CONTROLT;
PROC_UNLOCK(p);
PGRP_LOCK(pgrp);
sess->s_leader = p;
sess->s_sid = p->p_pid;
proc_id_set(PROC_ID_SESSION, p->p_pid);
refcount_init(&sess->s_count, 1);
sess->s_ttyvp = NULL;
sess->s_ttydp = NULL;
sess->s_ttyp = NULL;
bcopy(p->p_session->s_login, sess->s_login,
sizeof(sess->s_login));
pgrp->pg_session = sess;
KASSERT(p == curproc,
("enterpgrp: mksession and p != curproc"));
} else {
pgrp->pg_session = p->p_session;
sess_hold(pgrp->pg_session);
PGRP_LOCK(pgrp);
}
pgrp->pg_id = pgid;
proc_id_set(PROC_ID_GROUP, p->p_pid);
LIST_INIT(&pgrp->pg_members);
pgrp->pg_flags = 0;
/*
* As we have an exclusive lock of proctree_lock,
* this should not deadlock.
*/
LIST_INSERT_HEAD(PGRPHASH(pgid), pgrp, pg_hash);
SLIST_INIT(&pgrp->pg_sigiolst);
PGRP_UNLOCK(pgrp);
doenterpgrp(p, pgrp);
sx_xunlock(&old_pgrp->pg_killsx);
return (0);
}
/*
* Move p to an existing process group
*/
int
enterthispgrp(struct proc *p, struct pgrp *pgrp)
{
struct pgrp *old_pgrp;
sx_assert(&proctree_lock, SX_XLOCKED);
PROC_LOCK_ASSERT(p, MA_NOTOWNED);
PGRP_LOCK_ASSERT(pgrp, MA_NOTOWNED);
PGRP_LOCK_ASSERT(p->p_pgrp, MA_NOTOWNED);
SESS_LOCK_ASSERT(p->p_session, MA_NOTOWNED);
KASSERT(pgrp->pg_session == p->p_session,
("%s: pgrp's session %p, p->p_session %p proc %p\n",
__func__, pgrp->pg_session, p->p_session, p));
KASSERT(pgrp != p->p_pgrp,
("%s: p %p belongs to pgrp %p", __func__, p, pgrp));
old_pgrp = p->p_pgrp;
if (!sx_try_xlock(&old_pgrp->pg_killsx)) {
sx_xunlock(&proctree_lock);
sx_xlock(&old_pgrp->pg_killsx);
sx_xunlock(&old_pgrp->pg_killsx);
return (ERESTART);
}
MPASS(old_pgrp == p->p_pgrp);
if (!sx_try_xlock(&pgrp->pg_killsx)) {
sx_xunlock(&old_pgrp->pg_killsx);
sx_xunlock(&proctree_lock);
sx_xlock(&pgrp->pg_killsx);
sx_xunlock(&pgrp->pg_killsx);
return (ERESTART);
}
doenterpgrp(p, pgrp);
sx_xunlock(&pgrp->pg_killsx);
sx_xunlock(&old_pgrp->pg_killsx);
return (0);
}
/*
* If true, any child of q which belongs to group pgrp, qualifies the
* process group pgrp as not orphaned.
*/
static bool
isjobproc(struct proc *q, struct pgrp *pgrp)
{
sx_assert(&proctree_lock, SX_LOCKED);
return (q->p_pgrp != pgrp &&
q->p_pgrp->pg_session == pgrp->pg_session);
}
static struct proc *
jobc_reaper(struct proc *p)
{
struct proc *pp;
sx_assert(&proctree_lock, SA_LOCKED);
for (pp = p;;) {
pp = pp->p_reaper;
if (pp->p_reaper == pp ||
(pp->p_treeflag & P_TREE_GRPEXITED) == 0)
return (pp);
}
}
static struct proc *
jobc_parent(struct proc *p, struct proc *p_exiting)
{
struct proc *pp;
sx_assert(&proctree_lock, SA_LOCKED);
pp = proc_realparent(p);
if (pp->p_pptr == NULL || pp == p_exiting ||
(pp->p_treeflag & P_TREE_GRPEXITED) == 0)
return (pp);
return (jobc_reaper(pp));
}
-static int
+int
pgrp_calc_jobc(struct pgrp *pgrp)
{
struct proc *q;
int cnt;
#ifdef INVARIANTS
if (!mtx_owned(&pgrp->pg_mtx))
sx_assert(&proctree_lock, SA_LOCKED);
#endif
cnt = 0;
LIST_FOREACH(q, &pgrp->pg_members, p_pglist) {
if ((q->p_treeflag & P_TREE_GRPEXITED) != 0 ||
q->p_pptr == NULL)
continue;
if (isjobproc(jobc_parent(q, NULL), pgrp))
cnt++;
}
return (cnt);
}
/*
* Move p to a process group
*/
static void
doenterpgrp(struct proc *p, struct pgrp *pgrp)
{
struct pgrp *savepgrp;
struct proc *pp;
sx_assert(&proctree_lock, SX_XLOCKED);
PROC_LOCK_ASSERT(p, MA_NOTOWNED);
PGRP_LOCK_ASSERT(pgrp, MA_NOTOWNED);
PGRP_LOCK_ASSERT(p->p_pgrp, MA_NOTOWNED);
SESS_LOCK_ASSERT(p->p_session, MA_NOTOWNED);
savepgrp = p->p_pgrp;
pp = jobc_parent(p, NULL);
PGRP_LOCK(pgrp);
PGRP_LOCK(savepgrp);
if (isjobproc(pp, savepgrp) && pgrp_calc_jobc(savepgrp) == 1)
orphanpg(savepgrp);
PROC_LOCK(p);
LIST_REMOVE(p, p_pglist);
p->p_pgrp = pgrp;
PROC_UNLOCK(p);
LIST_INSERT_HEAD(&pgrp->pg_members, p, p_pglist);
if (isjobproc(pp, pgrp))
pgrp->pg_flags &= ~PGRP_ORPHANED;
PGRP_UNLOCK(savepgrp);
PGRP_UNLOCK(pgrp);
if (LIST_EMPTY(&savepgrp->pg_members))
pgdelete(savepgrp);
}
/*
* remove process from process group
*/
int
leavepgrp(struct proc *p)
{
struct pgrp *savepgrp;
sx_assert(&proctree_lock, SX_XLOCKED);
savepgrp = p->p_pgrp;
PGRP_LOCK(savepgrp);
PROC_LOCK(p);
LIST_REMOVE(p, p_pglist);
p->p_pgrp = NULL;
PROC_UNLOCK(p);
PGRP_UNLOCK(savepgrp);
if (LIST_EMPTY(&savepgrp->pg_members))
pgdelete(savepgrp);
return (0);
}
/*
* delete a process group
*/
static void
pgdelete(struct pgrp *pgrp)
{
struct session *savesess;
struct tty *tp;
sx_assert(&proctree_lock, SX_XLOCKED);
PGRP_LOCK_ASSERT(pgrp, MA_NOTOWNED);
SESS_LOCK_ASSERT(pgrp->pg_session, MA_NOTOWNED);
/*
* Reset any sigio structures pointing to us as a result of
* F_SETOWN with our pgid. The proctree lock ensures that
* new sigio structures will not be added after this point.
*/
funsetownlst(&pgrp->pg_sigiolst);
PGRP_LOCK(pgrp);
tp = pgrp->pg_session->s_ttyp;
LIST_REMOVE(pgrp, pg_hash);
savesess = pgrp->pg_session;
PGRP_UNLOCK(pgrp);
/* Remove the reference to the pgrp before deallocating it. */
if (tp != NULL) {
tty_lock(tp);
tty_rel_pgrp(tp, pgrp);
}
proc_id_clear(PROC_ID_GROUP, pgrp->pg_id);
uma_zfree(pgrp_zone, pgrp);
sess_release(savesess);
}
static void
fixjobc_kill(struct proc *p)
{
struct proc *q;
struct pgrp *pgrp;
sx_assert(&proctree_lock, SX_LOCKED);
PROC_LOCK_ASSERT(p, MA_NOTOWNED);
pgrp = p->p_pgrp;
PGRP_LOCK_ASSERT(pgrp, MA_NOTOWNED);
SESS_LOCK_ASSERT(pgrp->pg_session, MA_NOTOWNED);
/*
* p no longer affects process group orphanage for children.
* It is marked by the flag because p is only physically
* removed from its process group on wait(2).
*/
MPASS((p->p_treeflag & P_TREE_GRPEXITED) == 0);
p->p_treeflag |= P_TREE_GRPEXITED;
/*
* Check if exiting p orphans its own group.
*/
pgrp = p->p_pgrp;
if (isjobproc(jobc_parent(p, NULL), pgrp)) {
PGRP_LOCK(pgrp);
if (pgrp_calc_jobc(pgrp) == 0)
orphanpg(pgrp);
PGRP_UNLOCK(pgrp);
}
/*
* Check this process' children to see whether they qualify
* their process groups after reparenting to reaper.
*/
LIST_FOREACH(q, &p->p_children, p_sibling) {
pgrp = q->p_pgrp;
PGRP_LOCK(pgrp);
if (pgrp_calc_jobc(pgrp) == 0) {
/*
* We want to handle exactly the children that
* has p as realparent. Then, when calculating
* jobc_parent for children, we should ignore
* P_TREE_GRPEXITED flag already set on p.
*/
if (jobc_parent(q, p) == p && isjobproc(p, pgrp))
orphanpg(pgrp);
} else
pgrp->pg_flags &= ~PGRP_ORPHANED;
PGRP_UNLOCK(pgrp);
}
LIST_FOREACH(q, &p->p_orphans, p_orphan) {
pgrp = q->p_pgrp;
PGRP_LOCK(pgrp);
if (pgrp_calc_jobc(pgrp) == 0) {
if (isjobproc(p, pgrp))
orphanpg(pgrp);
} else
pgrp->pg_flags &= ~PGRP_ORPHANED;
PGRP_UNLOCK(pgrp);
}
}
void
killjobc(void)
{
struct session *sp;
struct tty *tp;
struct proc *p;
struct vnode *ttyvp;
p = curproc;
MPASS(p->p_flag & P_WEXIT);
sx_assert(&proctree_lock, SX_LOCKED);
if (SESS_LEADER(p)) {
sp = p->p_session;
/*
* s_ttyp is not zero'd; we use this to indicate that
* the session once had a controlling terminal. (for
* logging and informational purposes)
*/
SESS_LOCK(sp);
ttyvp = sp->s_ttyvp;
tp = sp->s_ttyp;
sp->s_ttyvp = NULL;
sp->s_ttydp = NULL;
sp->s_leader = NULL;
SESS_UNLOCK(sp);
/*
* Signal foreground pgrp and revoke access to
* controlling terminal if it has not been revoked
* already.
*
* Because the TTY may have been revoked in the mean
* time and could already have a new session associated
* with it, make sure we don't send a SIGHUP to a
* foreground process group that does not belong to this
* session.
*/
if (tp != NULL) {
tty_lock(tp);
if (tp->t_session == sp)
tty_signal_pgrp(tp, SIGHUP);
tty_unlock(tp);
}
if (ttyvp != NULL) {
sx_xunlock(&proctree_lock);
if (vn_lock(ttyvp, LK_EXCLUSIVE) == 0) {
VOP_REVOKE(ttyvp, REVOKEALL);
VOP_UNLOCK(ttyvp);
}
devfs_ctty_unref(ttyvp);
sx_xlock(&proctree_lock);
}
}
fixjobc_kill(p);
}
/*
* A process group has become orphaned, mark it as such for signal
* delivery code. If there are any stopped processes in the group,
* hang-up all process in that group.
*/
static void
orphanpg(struct pgrp *pg)
{
struct proc *p;
PGRP_LOCK_ASSERT(pg, MA_OWNED);
pg->pg_flags |= PGRP_ORPHANED;
LIST_FOREACH(p, &pg->pg_members, p_pglist) {
PROC_LOCK(p);
if (P_SHOULDSTOP(p) == P_STOPPED_SIG) {
PROC_UNLOCK(p);
LIST_FOREACH(p, &pg->pg_members, p_pglist) {
PROC_LOCK(p);
kern_psignal(p, SIGHUP);
kern_psignal(p, SIGCONT);
PROC_UNLOCK(p);
}
return;
}
PROC_UNLOCK(p);
}
}
void
sess_hold(struct session *s)
{
refcount_acquire(&s->s_count);
}
void
sess_release(struct session *s)
{
if (refcount_release(&s->s_count)) {
if (s->s_ttyp != NULL) {
tty_lock(s->s_ttyp);
tty_rel_sess(s->s_ttyp, s);
}
proc_id_clear(PROC_ID_SESSION, s->s_sid);
mtx_destroy(&s->s_mtx);
free(s, M_SESSION);
}
}
#ifdef DDB
static void
db_print_pgrp_one(struct pgrp *pgrp, struct proc *p)
{
db_printf(
" pid %d at %p pr %d pgrp %p e %d jc %d\n",
p->p_pid, p, p->p_pptr == NULL ? -1 : p->p_pptr->p_pid,
p->p_pgrp, (p->p_treeflag & P_TREE_GRPEXITED) != 0,
p->p_pptr == NULL ? 0 : isjobproc(p->p_pptr, pgrp));
}
DB_SHOW_COMMAND_FLAGS(pgrpdump, pgrpdump, DB_CMD_MEMSAFE)
{
struct pgrp *pgrp;
struct proc *p;
int i;
for (i = 0; i <= pgrphash; i++) {
if (!LIST_EMPTY(&pgrphashtbl[i])) {
db_printf("indx %d\n", i);
LIST_FOREACH(pgrp, &pgrphashtbl[i], pg_hash) {
db_printf(
" pgrp %p, pgid %d, sess %p, sesscnt %d, mem %p\n",
pgrp, (int)pgrp->pg_id, pgrp->pg_session,
pgrp->pg_session->s_count,
LIST_FIRST(&pgrp->pg_members));
LIST_FOREACH(p, &pgrp->pg_members, p_pglist)
db_print_pgrp_one(pgrp, p);
}
}
}
}
#endif /* DDB */
/*
* Calculate the kinfo_proc members which contain process-wide
* informations.
* Must be called with the target process locked.
*/
static void
fill_kinfo_aggregate(struct proc *p, struct kinfo_proc *kp)
{
struct thread *td;
PROC_LOCK_ASSERT(p, MA_OWNED);
kp->ki_estcpu = 0;
kp->ki_pctcpu = 0;
FOREACH_THREAD_IN_PROC(p, td) {
thread_lock(td);
kp->ki_pctcpu += sched_pctcpu(td);
kp->ki_estcpu += sched_estcpu(td);
thread_unlock(td);
}
}
/*
* Fill in any information that is common to all threads in the process.
* Must be called with the target process locked.
*/
static void
fill_kinfo_proc_only(struct proc *p, struct kinfo_proc *kp)
{
struct thread *td0;
struct ucred *cred;
struct sigacts *ps;
struct timeval boottime;
PROC_LOCK_ASSERT(p, MA_OWNED);
kp->ki_structsize = sizeof(*kp);
kp->ki_paddr = p;
kp->ki_addr =/* p->p_addr; */0; /* XXX */
kp->ki_args = p->p_args;
kp->ki_textvp = p->p_textvp;
#ifdef KTRACE
kp->ki_tracep = ktr_get_tracevp(p, false);
kp->ki_traceflag = p->p_traceflag;
#endif
kp->ki_fd = p->p_fd;
kp->ki_pd = p->p_pd;
kp->ki_vmspace = p->p_vmspace;
kp->ki_flag = p->p_flag;
kp->ki_flag2 = p->p_flag2;
cred = p->p_ucred;
if (cred) {
kp->ki_uid = cred->cr_uid;
kp->ki_ruid = cred->cr_ruid;
kp->ki_svuid = cred->cr_svuid;
kp->ki_cr_flags = 0;
if (cred->cr_flags & CRED_FLAG_CAPMODE)
kp->ki_cr_flags |= KI_CRF_CAPABILITY_MODE;
/* XXX bde doesn't like KI_NGROUPS */
if (cred->cr_ngroups > KI_NGROUPS) {
kp->ki_ngroups = KI_NGROUPS;
kp->ki_cr_flags |= KI_CRF_GRP_OVERFLOW;
} else
kp->ki_ngroups = cred->cr_ngroups;
bcopy(cred->cr_groups, kp->ki_groups,
kp->ki_ngroups * sizeof(gid_t));
kp->ki_rgid = cred->cr_rgid;
kp->ki_svgid = cred->cr_svgid;
/* If jailed(cred), emulate the old P_JAILED flag. */
if (jailed(cred)) {
kp->ki_flag |= P_JAILED;
/* If inside the jail, use 0 as a jail ID. */
if (cred->cr_prison != curthread->td_ucred->cr_prison)
kp->ki_jid = cred->cr_prison->pr_id;
}
strlcpy(kp->ki_loginclass, cred->cr_loginclass->lc_name,
sizeof(kp->ki_loginclass));
}
ps = p->p_sigacts;
if (ps) {
mtx_lock(&ps->ps_mtx);
kp->ki_sigignore = ps->ps_sigignore;
kp->ki_sigcatch = ps->ps_sigcatch;
mtx_unlock(&ps->ps_mtx);
}
if (p->p_state != PRS_NEW &&
p->p_state != PRS_ZOMBIE &&
p->p_vmspace != NULL) {
struct vmspace *vm = p->p_vmspace;
kp->ki_size = vm->vm_map.size;
kp->ki_rssize = vmspace_resident_count(vm); /*XXX*/
FOREACH_THREAD_IN_PROC(p, td0)
kp->ki_rssize += td0->td_kstack_pages;
kp->ki_swrss = vm->vm_swrss;
kp->ki_tsize = vm->vm_tsize;
kp->ki_dsize = vm->vm_dsize;
kp->ki_ssize = vm->vm_ssize;
} else if (p->p_state == PRS_ZOMBIE)
kp->ki_stat = SZOMB;
kp->ki_sflag = PS_INMEM;
/* Calculate legacy swtime as seconds since 'swtick'. */
kp->ki_swtime = (ticks - p->p_swtick) / hz;
kp->ki_pid = p->p_pid;
kp->ki_nice = p->p_nice;
kp->ki_fibnum = p->p_fibnum;
kp->ki_start = p->p_stats->p_start;
getboottime(&boottime);
timevaladd(&kp->ki_start, &boottime);
PROC_STATLOCK(p);
rufetch(p, &kp->ki_rusage);
kp->ki_runtime = cputick2usec(p->p_rux.rux_runtime);
calcru(p, &kp->ki_rusage.ru_utime, &kp->ki_rusage.ru_stime);
PROC_STATUNLOCK(p);
calccru(p, &kp->ki_childutime, &kp->ki_childstime);
/* Some callers want child times in a single value. */
kp->ki_childtime = kp->ki_childstime;
timevaladd(&kp->ki_childtime, &kp->ki_childutime);
FOREACH_THREAD_IN_PROC(p, td0)
kp->ki_cow += td0->td_cow;
if (p->p_comm[0] != '\0')
strlcpy(kp->ki_comm, p->p_comm, sizeof(kp->ki_comm));
if (p->p_sysent && p->p_sysent->sv_name != NULL &&
p->p_sysent->sv_name[0] != '\0')
strlcpy(kp->ki_emul, p->p_sysent->sv_name, sizeof(kp->ki_emul));
kp->ki_siglist = p->p_siglist;
kp->ki_xstat = KW_EXITCODE(p->p_xexit, p->p_xsig);
kp->ki_acflag = p->p_acflag;
kp->ki_lock = p->p_lock;
if (p->p_pptr) {
kp->ki_ppid = p->p_oppid;
if (p->p_flag & P_TRACED)
kp->ki_tracer = p->p_pptr->p_pid;
}
}
/*
* Fill job-related process information.
*/
static void
fill_kinfo_proc_pgrp(struct proc *p, struct kinfo_proc *kp)
{
struct tty *tp;
struct session *sp;
struct pgrp *pgrp;
sx_assert(&proctree_lock, SA_LOCKED);
PROC_LOCK_ASSERT(p, MA_OWNED);
pgrp = p->p_pgrp;
if (pgrp == NULL)
return;
kp->ki_pgid = pgrp->pg_id;
kp->ki_jobc = pgrp_calc_jobc(pgrp);
sp = pgrp->pg_session;
tp = NULL;
if (sp != NULL) {
kp->ki_sid = sp->s_sid;
SESS_LOCK(sp);
strlcpy(kp->ki_login, sp->s_login, sizeof(kp->ki_login));
if (sp->s_ttyvp)
kp->ki_kiflag |= KI_CTTY;
if (SESS_LEADER(p))
kp->ki_kiflag |= KI_SLEADER;
tp = sp->s_ttyp;
SESS_UNLOCK(sp);
}
if ((p->p_flag & P_CONTROLT) && tp != NULL) {
kp->ki_tdev = tty_udev(tp);
kp->ki_tdev_freebsd11 = kp->ki_tdev; /* truncate */
kp->ki_tpgid = tp->t_pgrp ? tp->t_pgrp->pg_id : NO_PID;
if (tp->t_session)
kp->ki_tsid = tp->t_session->s_sid;
} else {
kp->ki_tdev = NODEV;
kp->ki_tdev_freebsd11 = kp->ki_tdev; /* truncate */
}
}
/*
* Fill in information that is thread specific. Must be called with
* target process locked. If 'preferthread' is set, overwrite certain
* process-related fields that are maintained for both threads and
* processes.
*/
static void
fill_kinfo_thread(struct thread *td, struct kinfo_proc *kp, int preferthread)
{
struct proc *p;
p = td->td_proc;
kp->ki_tdaddr = td;
PROC_LOCK_ASSERT(p, MA_OWNED);
if (preferthread)
PROC_STATLOCK(p);
thread_lock(td);
if (td->td_wmesg != NULL)
strlcpy(kp->ki_wmesg, td->td_wmesg, sizeof(kp->ki_wmesg));
else
bzero(kp->ki_wmesg, sizeof(kp->ki_wmesg));
if (strlcpy(kp->ki_tdname, td->td_name, sizeof(kp->ki_tdname)) >=
sizeof(kp->ki_tdname)) {
strlcpy(kp->ki_moretdname,
td->td_name + sizeof(kp->ki_tdname) - 1,
sizeof(kp->ki_moretdname));
} else {
bzero(kp->ki_moretdname, sizeof(kp->ki_moretdname));
}
if (TD_ON_LOCK(td)) {
kp->ki_kiflag |= KI_LOCKBLOCK;
strlcpy(kp->ki_lockname, td->td_lockname,
sizeof(kp->ki_lockname));
} else {
kp->ki_kiflag &= ~KI_LOCKBLOCK;
bzero(kp->ki_lockname, sizeof(kp->ki_lockname));
}
if (p->p_state == PRS_NORMAL) { /* approximate. */
if (TD_ON_RUNQ(td) ||
TD_CAN_RUN(td) ||
TD_IS_RUNNING(td)) {
kp->ki_stat = SRUN;
} else if (P_SHOULDSTOP(p)) {
kp->ki_stat = SSTOP;
} else if (TD_IS_SLEEPING(td)) {
kp->ki_stat = SSLEEP;
} else if (TD_ON_LOCK(td)) {
kp->ki_stat = SLOCK;
} else {
kp->ki_stat = SWAIT;
}
} else if (p->p_state == PRS_ZOMBIE) {
kp->ki_stat = SZOMB;
} else {
kp->ki_stat = SIDL;
}
/* Things in the thread */
kp->ki_wchan = td->td_wchan;
kp->ki_pri.pri_level = td->td_priority;
kp->ki_pri.pri_native = td->td_base_pri;
/*
* Note: legacy fields; clamp at the old NOCPU value and/or
* the maximum u_char CPU value.
*/
if (td->td_lastcpu == NOCPU)
kp->ki_lastcpu_old = NOCPU_OLD;
else if (td->td_lastcpu > MAXCPU_OLD)
kp->ki_lastcpu_old = MAXCPU_OLD;
else
kp->ki_lastcpu_old = td->td_lastcpu;
if (td->td_oncpu == NOCPU)
kp->ki_oncpu_old = NOCPU_OLD;
else if (td->td_oncpu > MAXCPU_OLD)
kp->ki_oncpu_old = MAXCPU_OLD;
else
kp->ki_oncpu_old = td->td_oncpu;
kp->ki_lastcpu = td->td_lastcpu;
kp->ki_oncpu = td->td_oncpu;
kp->ki_tdflags = td->td_flags;
kp->ki_tid = td->td_tid;
kp->ki_numthreads = p->p_numthreads;
kp->ki_pcb = td->td_pcb;
kp->ki_kstack = (void *)td->td_kstack;
kp->ki_slptime = (ticks - td->td_slptick) / hz;
kp->ki_pri.pri_class = td->td_pri_class;
kp->ki_pri.pri_user = td->td_user_pri;
if (preferthread) {
rufetchtd(td, &kp->ki_rusage);
kp->ki_runtime = cputick2usec(td->td_rux.rux_runtime);
kp->ki_pctcpu = sched_pctcpu(td);
kp->ki_estcpu = sched_estcpu(td);
kp->ki_cow = td->td_cow;
}
/* We can't get this anymore but ps etc never used it anyway. */
kp->ki_rqindex = 0;
if (preferthread)
kp->ki_siglist = td->td_siglist;
kp->ki_sigmask = td->td_sigmask;
thread_unlock(td);
if (preferthread)
PROC_STATUNLOCK(p);
}
/*
* Fill in a kinfo_proc structure for the specified process.
* Must be called with the target process locked.
*/
void
fill_kinfo_proc(struct proc *p, struct kinfo_proc *kp)
{
MPASS(FIRST_THREAD_IN_PROC(p) != NULL);
bzero(kp, sizeof(*kp));
fill_kinfo_proc_pgrp(p,kp);
fill_kinfo_proc_only(p, kp);
fill_kinfo_thread(FIRST_THREAD_IN_PROC(p), kp, 0);
fill_kinfo_aggregate(p, kp);
}
struct pstats *
pstats_alloc(void)
{
return (malloc(sizeof(struct pstats), M_SUBPROC, M_ZERO|M_WAITOK));
}
/*
* Copy parts of p_stats; zero the rest of p_stats (statistics).
*/
void
pstats_fork(struct pstats *src, struct pstats *dst)
{
bzero(&dst->pstat_startzero,
__rangeof(struct pstats, pstat_startzero, pstat_endzero));
bcopy(&src->pstat_startcopy, &dst->pstat_startcopy,
__rangeof(struct pstats, pstat_startcopy, pstat_endcopy));
}
void
pstats_free(struct pstats *ps)
{
free(ps, M_SUBPROC);
}
#ifdef COMPAT_FREEBSD32
/*
* This function is typically used to copy out the kernel address, so
* it can be replaced by assignment of zero.
*/
static inline uint32_t
ptr32_trim(const void *ptr)
{
uintptr_t uptr;
uptr = (uintptr_t)ptr;
return ((uptr > UINT_MAX) ? 0 : uptr);
}
#define PTRTRIM_CP(src,dst,fld) \
do { (dst).fld = ptr32_trim((src).fld); } while (0)
static void
freebsd32_kinfo_proc_out(const struct kinfo_proc *ki, struct kinfo_proc32 *ki32)
{
int i;
bzero(ki32, sizeof(struct kinfo_proc32));
ki32->ki_structsize = sizeof(struct kinfo_proc32);
CP(*ki, *ki32, ki_layout);
PTRTRIM_CP(*ki, *ki32, ki_args);
PTRTRIM_CP(*ki, *ki32, ki_paddr);
PTRTRIM_CP(*ki, *ki32, ki_addr);
PTRTRIM_CP(*ki, *ki32, ki_tracep);
PTRTRIM_CP(*ki, *ki32, ki_textvp);
PTRTRIM_CP(*ki, *ki32, ki_fd);
PTRTRIM_CP(*ki, *ki32, ki_vmspace);
PTRTRIM_CP(*ki, *ki32, ki_wchan);
CP(*ki, *ki32, ki_pid);
CP(*ki, *ki32, ki_ppid);
CP(*ki, *ki32, ki_pgid);
CP(*ki, *ki32, ki_tpgid);
CP(*ki, *ki32, ki_sid);
CP(*ki, *ki32, ki_tsid);
CP(*ki, *ki32, ki_jobc);
CP(*ki, *ki32, ki_tdev);
CP(*ki, *ki32, ki_tdev_freebsd11);
CP(*ki, *ki32, ki_siglist);
CP(*ki, *ki32, ki_sigmask);
CP(*ki, *ki32, ki_sigignore);
CP(*ki, *ki32, ki_sigcatch);
CP(*ki, *ki32, ki_uid);
CP(*ki, *ki32, ki_ruid);
CP(*ki, *ki32, ki_svuid);
CP(*ki, *ki32, ki_rgid);
CP(*ki, *ki32, ki_svgid);
CP(*ki, *ki32, ki_ngroups);
for (i = 0; i < KI_NGROUPS; i++)
CP(*ki, *ki32, ki_groups[i]);
CP(*ki, *ki32, ki_size);
CP(*ki, *ki32, ki_rssize);
CP(*ki, *ki32, ki_swrss);
CP(*ki, *ki32, ki_tsize);
CP(*ki, *ki32, ki_dsize);
CP(*ki, *ki32, ki_ssize);
CP(*ki, *ki32, ki_xstat);
CP(*ki, *ki32, ki_acflag);
CP(*ki, *ki32, ki_pctcpu);
CP(*ki, *ki32, ki_estcpu);
CP(*ki, *ki32, ki_slptime);
CP(*ki, *ki32, ki_swtime);
CP(*ki, *ki32, ki_cow);
CP(*ki, *ki32, ki_runtime);
TV_CP(*ki, *ki32, ki_start);
TV_CP(*ki, *ki32, ki_childtime);
CP(*ki, *ki32, ki_flag);
CP(*ki, *ki32, ki_kiflag);
CP(*ki, *ki32, ki_traceflag);
CP(*ki, *ki32, ki_stat);
CP(*ki, *ki32, ki_nice);
CP(*ki, *ki32, ki_lock);
CP(*ki, *ki32, ki_rqindex);
CP(*ki, *ki32, ki_oncpu);
CP(*ki, *ki32, ki_lastcpu);
/* XXX TODO: wrap cpu value as appropriate */
CP(*ki, *ki32, ki_oncpu_old);
CP(*ki, *ki32, ki_lastcpu_old);
bcopy(ki->ki_tdname, ki32->ki_tdname, TDNAMLEN + 1);
bcopy(ki->ki_wmesg, ki32->ki_wmesg, WMESGLEN + 1);
bcopy(ki->ki_login, ki32->ki_login, LOGNAMELEN + 1);
bcopy(ki->ki_lockname, ki32->ki_lockname, LOCKNAMELEN + 1);
bcopy(ki->ki_comm, ki32->ki_comm, COMMLEN + 1);
bcopy(ki->ki_emul, ki32->ki_emul, KI_EMULNAMELEN + 1);
bcopy(ki->ki_loginclass, ki32->ki_loginclass, LOGINCLASSLEN + 1);
bcopy(ki->ki_moretdname, ki32->ki_moretdname, MAXCOMLEN - TDNAMLEN + 1);
CP(*ki, *ki32, ki_tracer);
CP(*ki, *ki32, ki_flag2);
CP(*ki, *ki32, ki_fibnum);
CP(*ki, *ki32, ki_cr_flags);
CP(*ki, *ki32, ki_jid);
CP(*ki, *ki32, ki_numthreads);
CP(*ki, *ki32, ki_tid);
CP(*ki, *ki32, ki_pri);
freebsd32_rusage_out(&ki->ki_rusage, &ki32->ki_rusage);
freebsd32_rusage_out(&ki->ki_rusage_ch, &ki32->ki_rusage_ch);
PTRTRIM_CP(*ki, *ki32, ki_pcb);
PTRTRIM_CP(*ki, *ki32, ki_kstack);
PTRTRIM_CP(*ki, *ki32, ki_udata);
PTRTRIM_CP(*ki, *ki32, ki_tdaddr);
CP(*ki, *ki32, ki_sflag);
CP(*ki, *ki32, ki_tdflags);
}
#endif
static ssize_t
kern_proc_out_size(struct proc *p, int flags)
{
ssize_t size = 0;
PROC_LOCK_ASSERT(p, MA_OWNED);
if ((flags & KERN_PROC_NOTHREADS) != 0) {
#ifdef COMPAT_FREEBSD32
if ((flags & KERN_PROC_MASK32) != 0) {
size += sizeof(struct kinfo_proc32);
} else
#endif
size += sizeof(struct kinfo_proc);
} else {
#ifdef COMPAT_FREEBSD32
if ((flags & KERN_PROC_MASK32) != 0)
size += sizeof(struct kinfo_proc32) * p->p_numthreads;
else
#endif
size += sizeof(struct kinfo_proc) * p->p_numthreads;
}
PROC_UNLOCK(p);
return (size);
}
int
kern_proc_out(struct proc *p, struct sbuf *sb, int flags)
{
struct thread *td;
struct kinfo_proc ki;
#ifdef COMPAT_FREEBSD32
struct kinfo_proc32 ki32;
#endif
int error;
PROC_LOCK_ASSERT(p, MA_OWNED);
MPASS(FIRST_THREAD_IN_PROC(p) != NULL);
error = 0;
fill_kinfo_proc(p, &ki);
if ((flags & KERN_PROC_NOTHREADS) != 0) {
#ifdef COMPAT_FREEBSD32
if ((flags & KERN_PROC_MASK32) != 0) {
freebsd32_kinfo_proc_out(&ki, &ki32);
if (sbuf_bcat(sb, &ki32, sizeof(ki32)) != 0)
error = ENOMEM;
} else
#endif
if (sbuf_bcat(sb, &ki, sizeof(ki)) != 0)
error = ENOMEM;
} else {
FOREACH_THREAD_IN_PROC(p, td) {
fill_kinfo_thread(td, &ki, 1);
#ifdef COMPAT_FREEBSD32
if ((flags & KERN_PROC_MASK32) != 0) {
freebsd32_kinfo_proc_out(&ki, &ki32);
if (sbuf_bcat(sb, &ki32, sizeof(ki32)) != 0)
error = ENOMEM;
} else
#endif
if (sbuf_bcat(sb, &ki, sizeof(ki)) != 0)
error = ENOMEM;
if (error != 0)
break;
}
}
PROC_UNLOCK(p);
return (error);
}
static int
sysctl_out_proc(struct proc *p, struct sysctl_req *req, int flags)
{
struct sbuf sb;
struct kinfo_proc ki;
int error, error2;
if (req->oldptr == NULL)
return (SYSCTL_OUT(req, 0, kern_proc_out_size(p, flags)));
sbuf_new_for_sysctl(&sb, (char *)&ki, sizeof(ki), req);
sbuf_clear_flags(&sb, SBUF_INCLUDENUL);
error = kern_proc_out(p, &sb, flags);
error2 = sbuf_finish(&sb);
sbuf_delete(&sb);
if (error != 0)
return (error);
else if (error2 != 0)
return (error2);
return (0);
}
int
proc_iterate(int (*cb)(struct proc *, void *), void *cbarg)
{
struct proc *p;
int error, i, j;
for (i = 0; i < pidhashlock + 1; i++) {
sx_slock(&proctree_lock);
sx_slock(&pidhashtbl_lock[i]);
for (j = i; j <= pidhash; j += pidhashlock + 1) {
LIST_FOREACH(p, &pidhashtbl[j], p_hash) {
if (p->p_state == PRS_NEW)
continue;
error = cb(p, cbarg);
PROC_LOCK_ASSERT(p, MA_NOTOWNED);
if (error != 0) {
sx_sunlock(&pidhashtbl_lock[i]);
sx_sunlock(&proctree_lock);
return (error);
}
}
}
sx_sunlock(&pidhashtbl_lock[i]);
sx_sunlock(&proctree_lock);
}
return (0);
}
struct kern_proc_out_args {
struct sysctl_req *req;
int flags;
int oid_number;
int *name;
};
static int
sysctl_kern_proc_iterate(struct proc *p, void *origarg)
{
struct kern_proc_out_args *arg = origarg;
int *name = arg->name;
int oid_number = arg->oid_number;
int flags = arg->flags;
struct sysctl_req *req = arg->req;
int error = 0;
PROC_LOCK(p);
KASSERT(p->p_ucred != NULL,
("process credential is NULL for non-NEW proc"));
/*
* Show a user only appropriate processes.
*/
if (p_cansee(curthread, p))
goto skip;
/*
* TODO - make more efficient (see notes below).
* do by session.
*/
switch (oid_number) {
case KERN_PROC_GID:
if (p->p_ucred->cr_gid != (gid_t)name[0])
goto skip;
break;
case KERN_PROC_PGRP:
/* could do this by traversing pgrp */
if (p->p_pgrp == NULL ||
p->p_pgrp->pg_id != (pid_t)name[0])
goto skip;
break;
case KERN_PROC_RGID:
if (p->p_ucred->cr_rgid != (gid_t)name[0])
goto skip;
break;
case KERN_PROC_SESSION:
if (p->p_session == NULL ||
p->p_session->s_sid != (pid_t)name[0])
goto skip;
break;
case KERN_PROC_TTY:
if ((p->p_flag & P_CONTROLT) == 0 ||
p->p_session == NULL)
goto skip;
/* XXX proctree_lock */
SESS_LOCK(p->p_session);
if (p->p_session->s_ttyp == NULL ||
tty_udev(p->p_session->s_ttyp) !=
(dev_t)name[0]) {
SESS_UNLOCK(p->p_session);
goto skip;
}
SESS_UNLOCK(p->p_session);
break;
case KERN_PROC_UID:
if (p->p_ucred->cr_uid != (uid_t)name[0])
goto skip;
break;
case KERN_PROC_RUID:
if (p->p_ucred->cr_ruid != (uid_t)name[0])
goto skip;
break;
case KERN_PROC_PROC:
break;
default:
break;
}
error = sysctl_out_proc(p, req, flags);
PROC_LOCK_ASSERT(p, MA_NOTOWNED);
return (error);
skip:
PROC_UNLOCK(p);
return (0);
}
static int
sysctl_kern_proc(SYSCTL_HANDLER_ARGS)
{
struct kern_proc_out_args iterarg;
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
int flags, oid_number;
int error = 0;
oid_number = oidp->oid_number;
if (oid_number != KERN_PROC_ALL &&
(oid_number & KERN_PROC_INC_THREAD) == 0)
flags = KERN_PROC_NOTHREADS;
else {
flags = 0;
oid_number &= ~KERN_PROC_INC_THREAD;
}
#ifdef COMPAT_FREEBSD32
if (req->flags & SCTL_MASK32)
flags |= KERN_PROC_MASK32;
#endif
if (oid_number == KERN_PROC_PID) {
if (namelen != 1)
return (EINVAL);
error = sysctl_wire_old_buffer(req, 0);
if (error)
return (error);
sx_slock(&proctree_lock);
error = pget((pid_t)name[0], PGET_CANSEE, &p);
if (error == 0)
error = sysctl_out_proc(p, req, flags);
sx_sunlock(&proctree_lock);
return (error);
}
switch (oid_number) {
case KERN_PROC_ALL:
if (namelen != 0)
return (EINVAL);
break;
case KERN_PROC_PROC:
if (namelen != 0 && namelen != 1)
return (EINVAL);
break;
default:
if (namelen != 1)
return (EINVAL);
break;
}
if (req->oldptr == NULL) {
/* overestimate by 5 procs */
error = SYSCTL_OUT(req, 0, sizeof (struct kinfo_proc) * 5);
if (error)
return (error);
} else {
error = sysctl_wire_old_buffer(req, 0);
if (error != 0)
return (error);
}
iterarg.flags = flags;
iterarg.oid_number = oid_number;
iterarg.req = req;
iterarg.name = name;
error = proc_iterate(sysctl_kern_proc_iterate, &iterarg);
return (error);
}
struct pargs *
pargs_alloc(int len)
{
struct pargs *pa;
pa = malloc(sizeof(struct pargs) + len, M_PARGS,
M_WAITOK);
refcount_init(&pa->ar_ref, 1);
pa->ar_length = len;
return (pa);
}
static void
pargs_free(struct pargs *pa)
{
free(pa, M_PARGS);
}
void
pargs_hold(struct pargs *pa)
{
if (pa == NULL)
return;
refcount_acquire(&pa->ar_ref);
}
void
pargs_drop(struct pargs *pa)
{
if (pa == NULL)
return;
if (refcount_release(&pa->ar_ref))
pargs_free(pa);
}
static int
proc_read_string(struct thread *td, struct proc *p, const char *sptr, char *buf,
size_t len)
{
ssize_t n;
/*
* This may return a short read if the string is shorter than the chunk
* and is aligned at the end of the page, and the following page is not
* mapped.
*/
n = proc_readmem(td, p, (vm_offset_t)sptr, buf, len);
if (n <= 0)
return (ENOMEM);
return (0);
}
#define PROC_AUXV_MAX 256 /* Safety limit on auxv size. */
enum proc_vector_type {
PROC_ARG,
PROC_ENV,
PROC_AUX,
};
#ifdef COMPAT_FREEBSD32
static int
get_proc_vector32(struct thread *td, struct proc *p, char ***proc_vectorp,
size_t *vsizep, enum proc_vector_type type)
{
struct freebsd32_ps_strings pss;
Elf32_Auxinfo aux;
vm_offset_t vptr, ptr;
uint32_t *proc_vector32;
char **proc_vector;
size_t vsize, size;
int i, error;
error = 0;
if (proc_readmem(td, p, PROC_PS_STRINGS(p), &pss, sizeof(pss)) !=
sizeof(pss))
return (ENOMEM);
switch (type) {
case PROC_ARG:
vptr = (vm_offset_t)PTRIN(pss.ps_argvstr);
vsize = pss.ps_nargvstr;
if (vsize > ARG_MAX)
return (ENOEXEC);
size = vsize * sizeof(int32_t);
break;
case PROC_ENV:
vptr = (vm_offset_t)PTRIN(pss.ps_envstr);
vsize = pss.ps_nenvstr;
if (vsize > ARG_MAX)
return (ENOEXEC);
size = vsize * sizeof(int32_t);
break;
case PROC_AUX:
vptr = (vm_offset_t)PTRIN(pss.ps_envstr) +
(pss.ps_nenvstr + 1) * sizeof(int32_t);
if (vptr % 4 != 0)
return (ENOEXEC);
for (ptr = vptr, i = 0; i < PROC_AUXV_MAX; i++) {
if (proc_readmem(td, p, ptr, &aux, sizeof(aux)) !=
sizeof(aux))
return (ENOMEM);
if (aux.a_type == AT_NULL)
break;
ptr += sizeof(aux);
}
if (aux.a_type != AT_NULL)
return (ENOEXEC);
vsize = i + 1;
size = vsize * sizeof(aux);
break;
default:
KASSERT(0, ("Wrong proc vector type: %d", type));
return (EINVAL);
}
proc_vector32 = malloc(size, M_TEMP, M_WAITOK);
if (proc_readmem(td, p, vptr, proc_vector32, size) != size) {
error = ENOMEM;
goto done;
}
if (type == PROC_AUX) {
*proc_vectorp = (char **)proc_vector32;
*vsizep = vsize;
return (0);
}
proc_vector = malloc(vsize * sizeof(char *), M_TEMP, M_WAITOK);
for (i = 0; i < (int)vsize; i++)
proc_vector[i] = PTRIN(proc_vector32[i]);
*proc_vectorp = proc_vector;
*vsizep = vsize;
done:
free(proc_vector32, M_TEMP);
return (error);
}
#endif
static int
get_proc_vector(struct thread *td, struct proc *p, char ***proc_vectorp,
size_t *vsizep, enum proc_vector_type type)
{
struct ps_strings pss;
Elf_Auxinfo aux;
vm_offset_t vptr, ptr;
char **proc_vector;
size_t vsize, size;
int i;
#ifdef COMPAT_FREEBSD32
if (SV_PROC_FLAG(p, SV_ILP32) != 0)
return (get_proc_vector32(td, p, proc_vectorp, vsizep, type));
#endif
if (proc_readmem(td, p, PROC_PS_STRINGS(p), &pss, sizeof(pss)) !=
sizeof(pss))
return (ENOMEM);
switch (type) {
case PROC_ARG:
vptr = (vm_offset_t)pss.ps_argvstr;
vsize = pss.ps_nargvstr;
if (vsize > ARG_MAX)
return (ENOEXEC);
size = vsize * sizeof(char *);
break;
case PROC_ENV:
vptr = (vm_offset_t)pss.ps_envstr;
vsize = pss.ps_nenvstr;
if (vsize > ARG_MAX)
return (ENOEXEC);
size = vsize * sizeof(char *);
break;
case PROC_AUX:
/*
* The aux array is just above env array on the stack. Check
* that the address is naturally aligned.
*/
vptr = (vm_offset_t)pss.ps_envstr + (pss.ps_nenvstr + 1)
* sizeof(char *);
#if __ELF_WORD_SIZE == 64
if (vptr % sizeof(uint64_t) != 0)
#else
if (vptr % sizeof(uint32_t) != 0)
#endif
return (ENOEXEC);
/*
* We count the array size reading the aux vectors from the
* stack until AT_NULL vector is returned. So (to keep the code
* simple) we read the process stack twice: the first time here
* to find the size and the second time when copying the vectors
* to the allocated proc_vector.
*/
for (ptr = vptr, i = 0; i < PROC_AUXV_MAX; i++) {
if (proc_readmem(td, p, ptr, &aux, sizeof(aux)) !=
sizeof(aux))
return (ENOMEM);
if (aux.a_type == AT_NULL)
break;
ptr += sizeof(aux);
}
/*
* If the PROC_AUXV_MAX entries are iterated over, and we have
* not reached AT_NULL, it is most likely we are reading wrong
* data: either the process doesn't have auxv array or data has
* been modified. Return the error in this case.
*/
if (aux.a_type != AT_NULL)
return (ENOEXEC);
vsize = i + 1;
size = vsize * sizeof(aux);
break;
default:
KASSERT(0, ("Wrong proc vector type: %d", type));
return (EINVAL); /* In case we are built without INVARIANTS. */
}
proc_vector = malloc(size, M_TEMP, M_WAITOK);
if (proc_readmem(td, p, vptr, proc_vector, size) != size) {
free(proc_vector, M_TEMP);
return (ENOMEM);
}
*proc_vectorp = proc_vector;
*vsizep = vsize;
return (0);
}
#define GET_PS_STRINGS_CHUNK_SZ 256 /* Chunk size (bytes) for ps_strings operations. */
static int
get_ps_strings(struct thread *td, struct proc *p, struct sbuf *sb,
enum proc_vector_type type)
{
size_t done, len, nchr, vsize;
int error, i;
char **proc_vector, *sptr;
char pss_string[GET_PS_STRINGS_CHUNK_SZ];
PROC_ASSERT_HELD(p);
/*
* We are not going to read more than 2 * (PATH_MAX + ARG_MAX) bytes.
*/
nchr = 2 * (PATH_MAX + ARG_MAX);
error = get_proc_vector(td, p, &proc_vector, &vsize, type);
if (error != 0)
return (error);
for (done = 0, i = 0; i < (int)vsize && done < nchr; i++) {
/*
* The program may have scribbled into its argv array, e.g. to
* remove some arguments. If that has happened, break out
* before trying to read from NULL.
*/
if (proc_vector[i] == NULL)
break;
for (sptr = proc_vector[i]; ; sptr += GET_PS_STRINGS_CHUNK_SZ) {
error = proc_read_string(td, p, sptr, pss_string,
sizeof(pss_string));
if (error != 0)
goto done;
len = strnlen(pss_string, GET_PS_STRINGS_CHUNK_SZ);
if (done + len >= nchr)
len = nchr - done - 1;
sbuf_bcat(sb, pss_string, len);
if (len != GET_PS_STRINGS_CHUNK_SZ)
break;
done += GET_PS_STRINGS_CHUNK_SZ;
}
sbuf_bcat(sb, "", 1);
done += len + 1;
}
done:
free(proc_vector, M_TEMP);
return (error);
}
int
proc_getargv(struct thread *td, struct proc *p, struct sbuf *sb)
{
return (get_ps_strings(curthread, p, sb, PROC_ARG));
}
int
proc_getenvv(struct thread *td, struct proc *p, struct sbuf *sb)
{
return (get_ps_strings(curthread, p, sb, PROC_ENV));
}
int
proc_getauxv(struct thread *td, struct proc *p, struct sbuf *sb)
{
size_t vsize, size;
char **auxv;
int error;
error = get_proc_vector(td, p, &auxv, &vsize, PROC_AUX);
if (error == 0) {
#ifdef COMPAT_FREEBSD32
if (SV_PROC_FLAG(p, SV_ILP32) != 0)
size = vsize * sizeof(Elf32_Auxinfo);
else
#endif
size = vsize * sizeof(Elf_Auxinfo);
if (sbuf_bcat(sb, auxv, size) != 0)
error = ENOMEM;
free(auxv, M_TEMP);
}
return (error);
}
/*
* This sysctl allows a process to retrieve the argument list or process
* title for another process without groping around in the address space
* of the other process. It also allow a process to set its own "process
* title to a string of its own choice.
*/
static int
sysctl_kern_proc_args(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct pargs *newpa, *pa;
struct proc *p;
struct sbuf sb;
int flags, error = 0, error2;
pid_t pid;
if (namelen != 1)
return (EINVAL);
p = curproc;
pid = (pid_t)name[0];
if (pid == -1) {
pid = p->p_pid;
}
/*
* If the query is for this process and it is single-threaded, there
* is nobody to modify pargs, thus we can just read.
*/
if (pid == p->p_pid && p->p_numthreads == 1 && req->newptr == NULL &&
(pa = p->p_args) != NULL)
return (SYSCTL_OUT(req, pa->ar_args, pa->ar_length));
flags = PGET_CANSEE;
if (req->newptr != NULL)
flags |= PGET_ISCURRENT;
error = pget(pid, flags, &p);
if (error)
return (error);
pa = p->p_args;
if (pa != NULL) {
pargs_hold(pa);
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, pa->ar_args, pa->ar_length);
pargs_drop(pa);
} else if ((p->p_flag & (P_WEXIT | P_SYSTEM)) == 0) {
_PHOLD(p);
PROC_UNLOCK(p);
sbuf_new_for_sysctl(&sb, NULL, GET_PS_STRINGS_CHUNK_SZ, req);
sbuf_clear_flags(&sb, SBUF_INCLUDENUL);
error = proc_getargv(curthread, p, &sb);
error2 = sbuf_finish(&sb);
PRELE(p);
sbuf_delete(&sb);
if (error == 0 && error2 != 0)
error = error2;
} else {
PROC_UNLOCK(p);
}
if (error != 0 || req->newptr == NULL)
return (error);
if (req->newlen > ps_arg_cache_limit - sizeof(struct pargs))
return (ENOMEM);
if (req->newlen == 0) {
/*
* Clear the argument pointer, so that we'll fetch arguments
* with proc_getargv() until further notice.
*/
newpa = NULL;
} else {
newpa = pargs_alloc(req->newlen);
error = SYSCTL_IN(req, newpa->ar_args, req->newlen);
if (error != 0) {
pargs_free(newpa);
return (error);
}
}
PROC_LOCK(p);
pa = p->p_args;
p->p_args = newpa;
PROC_UNLOCK(p);
pargs_drop(pa);
return (0);
}
/*
* This sysctl allows a process to retrieve environment of another process.
*/
static int
sysctl_kern_proc_env(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
struct sbuf sb;
int error, error2;
if (namelen != 1)
return (EINVAL);
error = pget((pid_t)name[0], PGET_WANTREAD, &p);
if (error != 0)
return (error);
if ((p->p_flag & P_SYSTEM) != 0) {
PRELE(p);
return (0);
}
sbuf_new_for_sysctl(&sb, NULL, GET_PS_STRINGS_CHUNK_SZ, req);
sbuf_clear_flags(&sb, SBUF_INCLUDENUL);
error = proc_getenvv(curthread, p, &sb);
error2 = sbuf_finish(&sb);
PRELE(p);
sbuf_delete(&sb);
return (error != 0 ? error : error2);
}
/*
* This sysctl allows a process to retrieve ELF auxiliary vector of
* another process.
*/
static int
sysctl_kern_proc_auxv(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
struct sbuf sb;
int error, error2;
if (namelen != 1)
return (EINVAL);
error = pget((pid_t)name[0], PGET_WANTREAD, &p);
if (error != 0)
return (error);
if ((p->p_flag & P_SYSTEM) != 0) {
PRELE(p);
return (0);
}
sbuf_new_for_sysctl(&sb, NULL, GET_PS_STRINGS_CHUNK_SZ, req);
sbuf_clear_flags(&sb, SBUF_INCLUDENUL);
error = proc_getauxv(curthread, p, &sb);
error2 = sbuf_finish(&sb);
PRELE(p);
sbuf_delete(&sb);
return (error != 0 ? error : error2);
}
/*
* Look up the canonical executable path running in the specified process.
* It tries to return the same hardlink name as was used for execve(2).
* This allows the programs that modify their behavior based on their progname,
* to operate correctly.
*
* Result is returned in retbuf, it must not be freed, similar to vn_fullpath()
* calling conventions.
* binname is a pointer to temporary string buffer of length MAXPATHLEN,
* allocated and freed by caller.
* freebuf should be freed by caller, from the M_TEMP malloc type.
*/
int
proc_get_binpath(struct proc *p, char *binname, char **retbuf,
char **freebuf)
{
struct nameidata nd;
struct vnode *vp, *dvp;
size_t freepath_size;
int error;
bool do_fullpath;
PROC_LOCK_ASSERT(p, MA_OWNED);
vp = p->p_textvp;
if (vp == NULL) {
PROC_UNLOCK(p);
*retbuf = "";
*freebuf = NULL;
return (0);
}
vref(vp);
dvp = p->p_textdvp;
if (dvp != NULL)
vref(dvp);
if (p->p_binname != NULL)
strlcpy(binname, p->p_binname, MAXPATHLEN);
PROC_UNLOCK(p);
do_fullpath = true;
*freebuf = NULL;
if (dvp != NULL && binname[0] != '\0') {
freepath_size = MAXPATHLEN;
if (vn_fullpath_hardlink(vp, dvp, binname, strlen(binname),
retbuf, freebuf, &freepath_size) == 0) {
/*
* Recheck the looked up path. The binary
* might have been renamed or replaced, in
* which case we should not report old name.
*/
NDINIT(&nd, LOOKUP, FOLLOW, UIO_SYSSPACE, *retbuf);
error = namei(&nd);
if (error == 0) {
if (nd.ni_vp == vp)
do_fullpath = false;
vrele(nd.ni_vp);
NDFREE_PNBUF(&nd);
}
}
}
if (do_fullpath) {
free(*freebuf, M_TEMP);
*freebuf = NULL;
error = vn_fullpath(vp, retbuf, freebuf);
}
vrele(vp);
if (dvp != NULL)
vrele(dvp);
return (error);
}
/*
* This sysctl allows a process to retrieve the path of the executable for
* itself or another process.
*/
static int
sysctl_kern_proc_pathname(SYSCTL_HANDLER_ARGS)
{
pid_t *pidp = (pid_t *)arg1;
unsigned int arglen = arg2;
struct proc *p;
char *retbuf, *freebuf, *binname;
int error;
if (arglen != 1)
return (EINVAL);
binname = malloc(MAXPATHLEN, M_TEMP, M_WAITOK);
binname[0] = '\0';
if (*pidp == -1) { /* -1 means this process */
error = 0;
p = req->td->td_proc;
PROC_LOCK(p);
} else {
error = pget(*pidp, PGET_CANSEE, &p);
}
if (error == 0)
error = proc_get_binpath(p, binname, &retbuf, &freebuf);
free(binname, M_TEMP);
if (error != 0)
return (error);
error = SYSCTL_OUT(req, retbuf, strlen(retbuf) + 1);
free(freebuf, M_TEMP);
return (error);
}
static int
sysctl_kern_proc_sv_name(SYSCTL_HANDLER_ARGS)
{
struct proc *p;
char *sv_name;
int *name;
int namelen;
int error;
namelen = arg2;
if (namelen != 1)
return (EINVAL);
name = (int *)arg1;
error = pget((pid_t)name[0], PGET_CANSEE, &p);
if (error != 0)
return (error);
sv_name = p->p_sysent->sv_name;
PROC_UNLOCK(p);
return (sysctl_handle_string(oidp, sv_name, 0, req));
}
#ifdef KINFO_OVMENTRY_SIZE
CTASSERT(sizeof(struct kinfo_ovmentry) == KINFO_OVMENTRY_SIZE);
#endif
#ifdef COMPAT_FREEBSD7
static int
sysctl_kern_proc_ovmmap(SYSCTL_HANDLER_ARGS)
{
vm_map_entry_t entry, tmp_entry;
unsigned int last_timestamp, namelen;
char *fullpath, *freepath;
struct kinfo_ovmentry *kve;
struct vattr va;
struct ucred *cred;
int error, *name;
struct vnode *vp;
struct proc *p;
vm_map_t map;
struct vmspace *vm;
namelen = arg2;
if (namelen != 1)
return (EINVAL);
name = (int *)arg1;
error = pget((pid_t)name[0], PGET_WANTREAD, &p);
if (error != 0)
return (error);
vm = vmspace_acquire_ref(p);
if (vm == NULL) {
PRELE(p);
return (ESRCH);
}
kve = malloc(sizeof(*kve), M_TEMP, M_WAITOK);
map = &vm->vm_map;
vm_map_lock_read(map);
VM_MAP_ENTRY_FOREACH(entry, map) {
vm_object_t obj, tobj, lobj;
vm_offset_t addr;
if (entry->eflags & MAP_ENTRY_IS_SUB_MAP)
continue;
bzero(kve, sizeof(*kve));
kve->kve_structsize = sizeof(*kve);
kve->kve_private_resident = 0;
obj = entry->object.vm_object;
if (obj != NULL) {
VM_OBJECT_RLOCK(obj);
if (obj->shadow_count == 1)
kve->kve_private_resident =
obj->resident_page_count;
}
kve->kve_resident = 0;
addr = entry->start;
while (addr < entry->end) {
if (pmap_extract(map->pmap, addr))
kve->kve_resident++;
addr += PAGE_SIZE;
}
for (lobj = tobj = obj; tobj; tobj = tobj->backing_object) {
if (tobj != obj) {
VM_OBJECT_RLOCK(tobj);
kve->kve_offset += tobj->backing_object_offset;
}
if (lobj != obj)
VM_OBJECT_RUNLOCK(lobj);
lobj = tobj;
}
kve->kve_start = (void*)entry->start;
kve->kve_end = (void*)entry->end;
kve->kve_offset += (off_t)entry->offset;
if (entry->protection & VM_PROT_READ)
kve->kve_protection |= KVME_PROT_READ;
if (entry->protection & VM_PROT_WRITE)
kve->kve_protection |= KVME_PROT_WRITE;
if (entry->protection & VM_PROT_EXECUTE)
kve->kve_protection |= KVME_PROT_EXEC;
if (entry->eflags & MAP_ENTRY_COW)
kve->kve_flags |= KVME_FLAG_COW;
if (entry->eflags & MAP_ENTRY_NEEDS_COPY)
kve->kve_flags |= KVME_FLAG_NEEDS_COPY;
if (entry->eflags & MAP_ENTRY_NOCOREDUMP)
kve->kve_flags |= KVME_FLAG_NOCOREDUMP;
last_timestamp = map->timestamp;
vm_map_unlock_read(map);
kve->kve_fileid = 0;
kve->kve_fsid = 0;
freepath = NULL;
fullpath = "";
if (lobj) {
kve->kve_type = vm_object_kvme_type(lobj, &vp);
if (kve->kve_type == KVME_TYPE_MGTDEVICE)
kve->kve_type = KVME_TYPE_UNKNOWN;
if (vp != NULL)
vref(vp);
if (lobj != obj)
VM_OBJECT_RUNLOCK(lobj);
kve->kve_ref_count = obj->ref_count;
kve->kve_shadow_count = obj->shadow_count;
VM_OBJECT_RUNLOCK(obj);
if (vp != NULL) {
vn_fullpath(vp, &fullpath, &freepath);
cred = curthread->td_ucred;
vn_lock(vp, LK_SHARED | LK_RETRY);
if (VOP_GETATTR(vp, &va, cred) == 0) {
kve->kve_fileid = va.va_fileid;
/* truncate */
kve->kve_fsid = va.va_fsid;
}
vput(vp);
}
} else {
kve->kve_type = KVME_TYPE_NONE;
kve->kve_ref_count = 0;
kve->kve_shadow_count = 0;
}
strlcpy(kve->kve_path, fullpath, sizeof(kve->kve_path));
if (freepath != NULL)
free(freepath, M_TEMP);
error = SYSCTL_OUT(req, kve, sizeof(*kve));
vm_map_lock_read(map);
if (error)
break;
if (last_timestamp != map->timestamp) {
vm_map_lookup_entry(map, addr - 1, &tmp_entry);
entry = tmp_entry;
}
}
vm_map_unlock_read(map);
vmspace_free(vm);
PRELE(p);
free(kve, M_TEMP);
return (error);
}
#endif /* COMPAT_FREEBSD7 */
#ifdef KINFO_VMENTRY_SIZE
CTASSERT(sizeof(struct kinfo_vmentry) == KINFO_VMENTRY_SIZE);
#endif
void
kern_proc_vmmap_resident(vm_map_t map, vm_map_entry_t entry,
int *resident_count, bool *super)
{
vm_object_t obj, tobj;
vm_page_t m, m_adv;
vm_offset_t addr;
vm_paddr_t pa;
vm_pindex_t pi, pi_adv, pindex;
int incore;
*super = false;
*resident_count = 0;
if (vmmap_skip_res_cnt)
return;
pa = 0;
obj = entry->object.vm_object;
addr = entry->start;
m_adv = NULL;
pi = OFF_TO_IDX(entry->offset);
for (; addr < entry->end; addr += IDX_TO_OFF(pi_adv), pi += pi_adv) {
if (m_adv != NULL) {
m = m_adv;
} else {
pi_adv = atop(entry->end - addr);
pindex = pi;
for (tobj = obj;; tobj = tobj->backing_object) {
m = vm_page_find_least(tobj, pindex);
if (m != NULL) {
if (m->pindex == pindex)
break;
if (pi_adv > m->pindex - pindex) {
pi_adv = m->pindex - pindex;
m_adv = m;
}
}
if (tobj->backing_object == NULL)
goto next;
pindex += OFF_TO_IDX(tobj->
backing_object_offset);
}
}
m_adv = NULL;
if (m->psind != 0 && addr + pagesizes[1] <= entry->end &&
(addr & (pagesizes[1] - 1)) == 0 && (incore =
pmap_mincore(map->pmap, addr, &pa) & MINCORE_SUPER) != 0) {
*super = true;
/*
* The virtual page might be smaller than the physical
* page, so we use the page size reported by the pmap
* rather than m->psind.
*/
pi_adv = atop(pagesizes[incore >> MINCORE_PSIND_SHIFT]);
} else {
/*
* We do not test the found page on validity.
* Either the page is busy and being paged in,
* or it was invalidated. The first case
* should be counted as resident, the second
* is not so clear; we do account both.
*/
pi_adv = 1;
}
*resident_count += pi_adv;
next:;
}
}
/*
* Must be called with the process locked and will return unlocked.
*/
int
kern_proc_vmmap_out(struct proc *p, struct sbuf *sb, ssize_t maxlen, int flags)
{
vm_map_entry_t entry, tmp_entry;
struct vattr va;
vm_map_t map;
vm_object_t lobj, nobj, obj, tobj;
char *fullpath, *freepath;
struct kinfo_vmentry *kve;
struct ucred *cred;
struct vnode *vp;
struct vmspace *vm;
struct cdev *cdev;
struct cdevsw *csw;
vm_offset_t addr;
unsigned int last_timestamp;
int error, ref;
key_t key;
unsigned short seq;
bool guard, super;
PROC_LOCK_ASSERT(p, MA_OWNED);
_PHOLD(p);
PROC_UNLOCK(p);
vm = vmspace_acquire_ref(p);
if (vm == NULL) {
PRELE(p);
return (ESRCH);
}
kve = malloc(sizeof(*kve), M_TEMP, M_WAITOK | M_ZERO);
error = 0;
map = &vm->vm_map;
vm_map_lock_read(map);
VM_MAP_ENTRY_FOREACH(entry, map) {
if (entry->eflags & MAP_ENTRY_IS_SUB_MAP)
continue;
addr = entry->end;
bzero(kve, sizeof(*kve));
obj = entry->object.vm_object;
if (obj != NULL) {
if ((obj->flags & OBJ_ANON) != 0)
kve->kve_obj = (uintptr_t)obj;
for (tobj = obj; tobj != NULL;
tobj = tobj->backing_object) {
VM_OBJECT_RLOCK(tobj);
kve->kve_offset += tobj->backing_object_offset;
lobj = tobj;
}
if (obj->backing_object == NULL)
kve->kve_private_resident =
obj->resident_page_count;
kern_proc_vmmap_resident(map, entry,
&kve->kve_resident, &super);
if (super)
kve->kve_flags |= KVME_FLAG_SUPER;
for (tobj = obj; tobj != NULL; tobj = nobj) {
nobj = tobj->backing_object;
if (tobj != obj && tobj != lobj)
VM_OBJECT_RUNLOCK(tobj);
}
} else {
lobj = NULL;
}
kve->kve_start = entry->start;
kve->kve_end = entry->end;
kve->kve_offset += entry->offset;
if (entry->protection & VM_PROT_READ)
kve->kve_protection |= KVME_PROT_READ;
if (entry->protection & VM_PROT_WRITE)
kve->kve_protection |= KVME_PROT_WRITE;
if (entry->protection & VM_PROT_EXECUTE)
kve->kve_protection |= KVME_PROT_EXEC;
if (entry->max_protection & VM_PROT_READ)
kve->kve_protection |= KVME_MAX_PROT_READ;
if (entry->max_protection & VM_PROT_WRITE)
kve->kve_protection |= KVME_MAX_PROT_WRITE;
if (entry->max_protection & VM_PROT_EXECUTE)
kve->kve_protection |= KVME_MAX_PROT_EXEC;
if (entry->eflags & MAP_ENTRY_COW)
kve->kve_flags |= KVME_FLAG_COW;
if (entry->eflags & MAP_ENTRY_NEEDS_COPY)
kve->kve_flags |= KVME_FLAG_NEEDS_COPY;
if (entry->eflags & MAP_ENTRY_NOCOREDUMP)
kve->kve_flags |= KVME_FLAG_NOCOREDUMP;
if (entry->eflags & MAP_ENTRY_GROWS_DOWN)
kve->kve_flags |= KVME_FLAG_GROWS_DOWN;
if (entry->eflags & MAP_ENTRY_USER_WIRED)
kve->kve_flags |= KVME_FLAG_USER_WIRED;
guard = (entry->eflags & MAP_ENTRY_GUARD) != 0;
last_timestamp = map->timestamp;
vm_map_unlock_read(map);
freepath = NULL;
fullpath = "";
if (lobj != NULL) {
kve->kve_type = vm_object_kvme_type(lobj, &vp);
if (vp != NULL)
vref(vp);
if (lobj != obj)
VM_OBJECT_RUNLOCK(lobj);
kve->kve_ref_count = obj->ref_count;
kve->kve_shadow_count = obj->shadow_count;
if ((obj->type == OBJT_DEVICE ||
obj->type == OBJT_MGTDEVICE) &&
(obj->flags & OBJ_CDEVH) != 0) {
cdev = obj->un_pager.devp.handle;
if (cdev != NULL) {
csw = dev_refthread(cdev, &ref);
if (csw != NULL) {
strlcpy(kve->kve_path,
cdev->si_name, sizeof(
kve->kve_path));
dev_relthread(cdev, ref);
}
}
}
VM_OBJECT_RUNLOCK(obj);
if ((lobj->flags & OBJ_SYSVSHM) != 0) {
kve->kve_flags |= KVME_FLAG_SYSVSHM;
shmobjinfo(lobj, &key, &seq);
kve->kve_vn_fileid = key;
kve->kve_vn_fsid_freebsd11 = seq;
}
if ((lobj->flags & OBJ_POSIXSHM) != 0) {
kve->kve_flags |= KVME_FLAG_POSIXSHM;
shm_get_path(lobj, kve->kve_path,
sizeof(kve->kve_path));
}
if (vp != NULL) {
vn_fullpath(vp, &fullpath, &freepath);
kve->kve_vn_type = vntype_to_kinfo(vp->v_type);
cred = curthread->td_ucred;
vn_lock(vp, LK_SHARED | LK_RETRY);
if (VOP_GETATTR(vp, &va, cred) == 0) {
kve->kve_vn_fileid = va.va_fileid;
kve->kve_vn_fsid = va.va_fsid;
kve->kve_vn_fsid_freebsd11 =
kve->kve_vn_fsid; /* truncate */
kve->kve_vn_mode =
MAKEIMODE(va.va_type, va.va_mode);
kve->kve_vn_size = va.va_size;
kve->kve_vn_rdev = va.va_rdev;
kve->kve_vn_rdev_freebsd11 =
kve->kve_vn_rdev; /* truncate */
kve->kve_status = KF_ATTR_VALID;
}
vput(vp);
strlcpy(kve->kve_path, fullpath, sizeof(
kve->kve_path));
free(freepath, M_TEMP);
}
} else {
kve->kve_type = guard ? KVME_TYPE_GUARD :
KVME_TYPE_NONE;
kve->kve_ref_count = 0;
kve->kve_shadow_count = 0;
}
/* Pack record size down */
if ((flags & KERN_VMMAP_PACK_KINFO) != 0)
kve->kve_structsize =
offsetof(struct kinfo_vmentry, kve_path) +
strlen(kve->kve_path) + 1;
else
kve->kve_structsize = sizeof(*kve);
kve->kve_structsize = roundup(kve->kve_structsize,
sizeof(uint64_t));
/* Halt filling and truncate rather than exceeding maxlen */
if (maxlen != -1 && maxlen < kve->kve_structsize) {
error = 0;
vm_map_lock_read(map);
break;
} else if (maxlen != -1)
maxlen -= kve->kve_structsize;
if (sbuf_bcat(sb, kve, kve->kve_structsize) != 0)
error = ENOMEM;
vm_map_lock_read(map);
if (error != 0)
break;
if (last_timestamp != map->timestamp) {
vm_map_lookup_entry(map, addr - 1, &tmp_entry);
entry = tmp_entry;
}
}
vm_map_unlock_read(map);
vmspace_free(vm);
PRELE(p);
free(kve, M_TEMP);
return (error);
}
static int
sysctl_kern_proc_vmmap(SYSCTL_HANDLER_ARGS)
{
struct proc *p;
struct sbuf sb;
u_int namelen;
int error, error2, *name;
namelen = arg2;
if (namelen != 1)
return (EINVAL);
name = (int *)arg1;
sbuf_new_for_sysctl(&sb, NULL, sizeof(struct kinfo_vmentry), req);
sbuf_clear_flags(&sb, SBUF_INCLUDENUL);
error = pget((pid_t)name[0], PGET_CANDEBUG | PGET_NOTWEXIT, &p);
if (error != 0) {
sbuf_delete(&sb);
return (error);
}
error = kern_proc_vmmap_out(p, &sb, -1, KERN_VMMAP_PACK_KINFO);
error2 = sbuf_finish(&sb);
sbuf_delete(&sb);
return (error != 0 ? error : error2);
}
#if defined(STACK) || defined(DDB)
static int
sysctl_kern_proc_kstack(SYSCTL_HANDLER_ARGS)
{
struct kinfo_kstack *kkstp;
int error, i, *name, numthreads;
lwpid_t *lwpidarray;
struct thread *td;
struct stack *st;
struct sbuf sb;
struct proc *p;
u_int namelen;
namelen = arg2;
if (namelen != 1)
return (EINVAL);
name = (int *)arg1;
error = pget((pid_t)name[0], PGET_NOTINEXEC | PGET_WANTREAD, &p);
if (error != 0)
return (error);
kkstp = malloc(sizeof(*kkstp), M_TEMP, M_WAITOK);
st = stack_create(M_WAITOK);
lwpidarray = NULL;
PROC_LOCK(p);
do {
if (lwpidarray != NULL) {
free(lwpidarray, M_TEMP);
lwpidarray = NULL;
}
numthreads = p->p_numthreads;
PROC_UNLOCK(p);
lwpidarray = malloc(sizeof(*lwpidarray) * numthreads, M_TEMP,
M_WAITOK | M_ZERO);
PROC_LOCK(p);
} while (numthreads < p->p_numthreads);
/*
* XXXRW: During the below loop, execve(2) and countless other sorts
* of changes could have taken place. Should we check to see if the
* vmspace has been replaced, or the like, in order to prevent
* giving a snapshot that spans, say, execve(2), with some threads
* before and some after? Among other things, the credentials could
* have changed, in which case the right to extract debug info might
* no longer be assured.
*/
i = 0;
FOREACH_THREAD_IN_PROC(p, td) {
KASSERT(i < numthreads,
("sysctl_kern_proc_kstack: numthreads"));
lwpidarray[i] = td->td_tid;
i++;
}
PROC_UNLOCK(p);
numthreads = i;
for (i = 0; i < numthreads; i++) {
td = tdfind(lwpidarray[i], p->p_pid);
if (td == NULL) {
continue;
}
bzero(kkstp, sizeof(*kkstp));
(void)sbuf_new(&sb, kkstp->kkst_trace,
sizeof(kkstp->kkst_trace), SBUF_FIXEDLEN);
thread_lock(td);
kkstp->kkst_tid = td->td_tid;
if (stack_save_td(st, td) == 0)
kkstp->kkst_state = KKST_STATE_STACKOK;
else
kkstp->kkst_state = KKST_STATE_RUNNING;
thread_unlock(td);
PROC_UNLOCK(p);
stack_sbuf_print(&sb, st);
sbuf_finish(&sb);
sbuf_delete(&sb);
error = SYSCTL_OUT(req, kkstp, sizeof(*kkstp));
if (error)
break;
}
PRELE(p);
if (lwpidarray != NULL)
free(lwpidarray, M_TEMP);
stack_destroy(st);
free(kkstp, M_TEMP);
return (error);
}
#endif
/*
* This sysctl allows a process to retrieve the full list of groups from
* itself or another process.
*/
static int
sysctl_kern_proc_groups(SYSCTL_HANDLER_ARGS)
{
pid_t *pidp = (pid_t *)arg1;
unsigned int arglen = arg2;
struct proc *p;
struct ucred *cred;
int error;
if (arglen != 1)
return (EINVAL);
if (*pidp == -1) { /* -1 means this process */
p = req->td->td_proc;
PROC_LOCK(p);
} else {
error = pget(*pidp, PGET_CANSEE, &p);
if (error != 0)
return (error);
}
cred = crhold(p->p_ucred);
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, cred->cr_groups,
cred->cr_ngroups * sizeof(gid_t));
crfree(cred);
return (error);
}
/*
* This sysctl allows a process to retrieve or/and set the resource limit for
* another process.
*/
static int
sysctl_kern_proc_rlimit(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct rlimit rlim;
struct proc *p;
u_int which;
int flags, error;
if (namelen != 2)
return (EINVAL);
which = (u_int)name[1];
if (which >= RLIM_NLIMITS)
return (EINVAL);
if (req->newptr != NULL && req->newlen != sizeof(rlim))
return (EINVAL);
flags = PGET_HOLD | PGET_NOTWEXIT;
if (req->newptr != NULL)
flags |= PGET_CANDEBUG;
else
flags |= PGET_CANSEE;
error = pget((pid_t)name[0], flags, &p);
if (error != 0)
return (error);
/*
* Retrieve limit.
*/
if (req->oldptr != NULL) {
PROC_LOCK(p);
lim_rlimit_proc(p, which, &rlim);
PROC_UNLOCK(p);
}
error = SYSCTL_OUT(req, &rlim, sizeof(rlim));
if (error != 0)
goto errout;
/*
* Set limit.
*/
if (req->newptr != NULL) {
error = SYSCTL_IN(req, &rlim, sizeof(rlim));
if (error == 0)
error = kern_proc_setrlimit(curthread, p, which, &rlim);
}
errout:
PRELE(p);
return (error);
}
/*
* This sysctl allows a process to retrieve ps_strings structure location of
* another process.
*/
static int
sysctl_kern_proc_ps_strings(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
vm_offset_t ps_strings;
int error;
#ifdef COMPAT_FREEBSD32
uint32_t ps_strings32;
#endif
if (namelen != 1)
return (EINVAL);
error = pget((pid_t)name[0], PGET_CANDEBUG, &p);
if (error != 0)
return (error);
#ifdef COMPAT_FREEBSD32
if ((req->flags & SCTL_MASK32) != 0) {
/*
* We return 0 if the 32 bit emulation request is for a 64 bit
* process.
*/
ps_strings32 = SV_PROC_FLAG(p, SV_ILP32) != 0 ?
PTROUT(PROC_PS_STRINGS(p)) : 0;
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, &ps_strings32, sizeof(ps_strings32));
return (error);
}
#endif
ps_strings = PROC_PS_STRINGS(p);
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, &ps_strings, sizeof(ps_strings));
return (error);
}
/*
* This sysctl allows a process to retrieve umask of another process.
*/
static int
sysctl_kern_proc_umask(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
int error;
u_short cmask;
pid_t pid;
if (namelen != 1)
return (EINVAL);
pid = (pid_t)name[0];
p = curproc;
if (pid == p->p_pid || pid == 0) {
cmask = p->p_pd->pd_cmask;
goto out;
}
error = pget(pid, PGET_WANTREAD, &p);
if (error != 0)
return (error);
cmask = p->p_pd->pd_cmask;
PRELE(p);
out:
error = SYSCTL_OUT(req, &cmask, sizeof(cmask));
return (error);
}
/*
* This sysctl allows a process to set and retrieve binary osreldate of
* another process.
*/
static int
sysctl_kern_proc_osrel(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
int flags, error, osrel;
if (namelen != 1)
return (EINVAL);
if (req->newptr != NULL && req->newlen != sizeof(osrel))
return (EINVAL);
flags = PGET_HOLD | PGET_NOTWEXIT;
if (req->newptr != NULL)
flags |= PGET_CANDEBUG;
else
flags |= PGET_CANSEE;
error = pget((pid_t)name[0], flags, &p);
if (error != 0)
return (error);
error = SYSCTL_OUT(req, &p->p_osrel, sizeof(p->p_osrel));
if (error != 0)
goto errout;
if (req->newptr != NULL) {
error = SYSCTL_IN(req, &osrel, sizeof(osrel));
if (error != 0)
goto errout;
if (osrel < 0) {
error = EINVAL;
goto errout;
}
p->p_osrel = osrel;
}
errout:
PRELE(p);
return (error);
}
static int
sysctl_kern_proc_sigtramp(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
struct proc *p;
struct kinfo_sigtramp kst;
const struct sysentvec *sv;
int error;
#ifdef COMPAT_FREEBSD32
struct kinfo_sigtramp32 kst32;
#endif
if (namelen != 1)
return (EINVAL);
error = pget((pid_t)name[0], PGET_CANDEBUG, &p);
if (error != 0)
return (error);
sv = p->p_sysent;
#ifdef COMPAT_FREEBSD32
if ((req->flags & SCTL_MASK32) != 0) {
bzero(&kst32, sizeof(kst32));
if (SV_PROC_FLAG(p, SV_ILP32)) {
if (PROC_HAS_SHP(p)) {
kst32.ksigtramp_start = PROC_SIGCODE(p);
kst32.ksigtramp_end = kst32.ksigtramp_start +
((sv->sv_flags & SV_DSO_SIG) == 0 ?
*sv->sv_szsigcode :
(uintptr_t)sv->sv_szsigcode);
} else {
kst32.ksigtramp_start = PROC_PS_STRINGS(p) -
*sv->sv_szsigcode;
kst32.ksigtramp_end = PROC_PS_STRINGS(p);
}
}
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, &kst32, sizeof(kst32));
return (error);
}
#endif
bzero(&kst, sizeof(kst));
if (PROC_HAS_SHP(p)) {
kst.ksigtramp_start = (char *)PROC_SIGCODE(p);
kst.ksigtramp_end = (char *)kst.ksigtramp_start +
((sv->sv_flags & SV_DSO_SIG) == 0 ? *sv->sv_szsigcode :
(uintptr_t)sv->sv_szsigcode);
} else {
kst.ksigtramp_start = (char *)PROC_PS_STRINGS(p) -
*sv->sv_szsigcode;
kst.ksigtramp_end = (char *)PROC_PS_STRINGS(p);
}
PROC_UNLOCK(p);
error = SYSCTL_OUT(req, &kst, sizeof(kst));
return (error);
}
static int
sysctl_kern_proc_sigfastblk(SYSCTL_HANDLER_ARGS)
{
int *name = (int *)arg1;
u_int namelen = arg2;
pid_t pid;
struct proc *p;
struct thread *td1;
uintptr_t addr;
#ifdef COMPAT_FREEBSD32
uint32_t addr32;
#endif
int error;
if (namelen != 1 || req->newptr != NULL)
return (EINVAL);
pid = (pid_t)name[0];
error = pget(pid, PGET_HOLD | PGET_NOTWEXIT | PGET_CANDEBUG, &p);
if (error != 0)
return (error);
PROC_LOCK(p);
#ifdef COMPAT_FREEBSD32
if (SV_CURPROC_FLAG(SV_ILP32)) {
if (!SV_PROC_FLAG(p, SV_ILP32)) {
error = EINVAL;
goto errlocked;
}
}
#endif
if (pid <= PID_MAX) {
td1 = FIRST_THREAD_IN_PROC(p);
} else {
FOREACH_THREAD_IN_PROC(p, td1) {
if (td1->td_tid == pid)
break;
}
}
if (td1 == NULL) {
error = ESRCH;
goto errlocked;
}
/*
* The access to the private thread flags. It is fine as far
* as no out-of-thin-air values are read from td_pflags, and
* usermode read of the td_sigblock_ptr is racy inherently,
* since target process might have already changed it
* meantime.
*/
if ((td1->td_pflags & TDP_SIGFASTBLOCK) != 0)
addr = (uintptr_t)td1->td_sigblock_ptr;
else
error = ENOTTY;
errlocked:
_PRELE(p);
PROC_UNLOCK(p);
if (error != 0)
return (error);
#ifdef COMPAT_FREEBSD32
if (SV_CURPROC_FLAG(SV_ILP32)) {
addr32 = addr;
error = SYSCTL_OUT(req, &addr32, sizeof(addr32));
} else
#endif
error = SYSCTL_OUT(req, &addr, sizeof(addr));
return (error);
}
static int
sysctl_kern_proc_vm_layout(SYSCTL_HANDLER_ARGS)
{
struct kinfo_vm_layout kvm;
struct proc *p;
struct vmspace *vmspace;
int error, *name;
name = (int *)arg1;
if ((u_int)arg2 != 1)
return (EINVAL);
error = pget((pid_t)name[0], PGET_CANDEBUG, &p);
if (error != 0)
return (error);
#ifdef COMPAT_FREEBSD32
if (SV_CURPROC_FLAG(SV_ILP32)) {
if (!SV_PROC_FLAG(p, SV_ILP32)) {
PROC_UNLOCK(p);
return (EINVAL);
}
}
#endif
vmspace = vmspace_acquire_ref(p);
PROC_UNLOCK(p);
memset(&kvm, 0, sizeof(kvm));
kvm.kvm_min_user_addr = vm_map_min(&vmspace->vm_map);
kvm.kvm_max_user_addr = vm_map_max(&vmspace->vm_map);
kvm.kvm_text_addr = (uintptr_t)vmspace->vm_taddr;
kvm.kvm_text_size = vmspace->vm_tsize;
kvm.kvm_data_addr = (uintptr_t)vmspace->vm_daddr;
kvm.kvm_data_size = vmspace->vm_dsize;
kvm.kvm_stack_addr = (uintptr_t)vmspace->vm_maxsaddr;
kvm.kvm_stack_size = vmspace->vm_ssize;
kvm.kvm_shp_addr = vmspace->vm_shp_base;
kvm.kvm_shp_size = p->p_sysent->sv_shared_page_len;
if ((vmspace->vm_map.flags & MAP_WIREFUTURE) != 0)
kvm.kvm_map_flags |= KMAP_FLAG_WIREFUTURE;
if ((vmspace->vm_map.flags & MAP_ASLR) != 0)
kvm.kvm_map_flags |= KMAP_FLAG_ASLR;
if ((vmspace->vm_map.flags & MAP_ASLR_IGNSTART) != 0)
kvm.kvm_map_flags |= KMAP_FLAG_ASLR_IGNSTART;
if ((vmspace->vm_map.flags & MAP_WXORX) != 0)
kvm.kvm_map_flags |= KMAP_FLAG_WXORX;
if ((vmspace->vm_map.flags & MAP_ASLR_STACK) != 0)
kvm.kvm_map_flags |= KMAP_FLAG_ASLR_STACK;
if (vmspace->vm_shp_base != p->p_sysent->sv_shared_page_base &&
PROC_HAS_SHP(p))
kvm.kvm_map_flags |= KMAP_FLAG_ASLR_SHARED_PAGE;
#ifdef COMPAT_FREEBSD32
if (SV_CURPROC_FLAG(SV_ILP32)) {
struct kinfo_vm_layout32 kvm32;
memset(&kvm32, 0, sizeof(kvm32));
kvm32.kvm_min_user_addr = (uint32_t)kvm.kvm_min_user_addr;
kvm32.kvm_max_user_addr = (uint32_t)kvm.kvm_max_user_addr;
kvm32.kvm_text_addr = (uint32_t)kvm.kvm_text_addr;
kvm32.kvm_text_size = (uint32_t)kvm.kvm_text_size;
kvm32.kvm_data_addr = (uint32_t)kvm.kvm_data_addr;
kvm32.kvm_data_size = (uint32_t)kvm.kvm_data_size;
kvm32.kvm_stack_addr = (uint32_t)kvm.kvm_stack_addr;
kvm32.kvm_stack_size = (uint32_t)kvm.kvm_stack_size;
kvm32.kvm_shp_addr = (uint32_t)kvm.kvm_shp_addr;
kvm32.kvm_shp_size = (uint32_t)kvm.kvm_shp_size;
kvm32.kvm_map_flags = kvm.kvm_map_flags;
error = SYSCTL_OUT(req, &kvm32, sizeof(kvm32));
goto out;
}
#endif
error = SYSCTL_OUT(req, &kvm, sizeof(kvm));
#ifdef COMPAT_FREEBSD32
out:
#endif
vmspace_free(vmspace);
return (error);
}
SYSCTL_NODE(_kern, KERN_PROC, proc, CTLFLAG_RD | CTLFLAG_MPSAFE, 0,
"Process table");
SYSCTL_PROC(_kern_proc, KERN_PROC_ALL, all, CTLFLAG_RD|CTLTYPE_STRUCT|
CTLFLAG_MPSAFE, 0, 0, sysctl_kern_proc, "S,proc",
"Return entire process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_GID, gid, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_PGRP, pgrp, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_RGID, rgid, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_SESSION, sid, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_TTY, tty, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_UID, uid, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_RUID, ruid, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_PID, pid, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, KERN_PROC_PROC, proc, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc, "Return process table, no threads");
static SYSCTL_NODE(_kern_proc, KERN_PROC_ARGS, args,
CTLFLAG_RW | CTLFLAG_CAPWR | CTLFLAG_ANYBODY | CTLFLAG_MPSAFE,
sysctl_kern_proc_args, "Process argument list");
static SYSCTL_NODE(_kern_proc, KERN_PROC_ENV, env, CTLFLAG_RD | CTLFLAG_MPSAFE,
sysctl_kern_proc_env, "Process environment");
static SYSCTL_NODE(_kern_proc, KERN_PROC_AUXV, auxv, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_auxv, "Process ELF auxiliary vector");
static SYSCTL_NODE(_kern_proc, KERN_PROC_PATHNAME, pathname, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_pathname, "Process executable path");
static SYSCTL_NODE(_kern_proc, KERN_PROC_SV_NAME, sv_name, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_sv_name,
"Process syscall vector name (ABI type)");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_GID | KERN_PROC_INC_THREAD), gid_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_PGRP | KERN_PROC_INC_THREAD), pgrp_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_RGID | KERN_PROC_INC_THREAD), rgid_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_SESSION | KERN_PROC_INC_THREAD),
sid_td, CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_TTY | KERN_PROC_INC_THREAD), tty_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_UID | KERN_PROC_INC_THREAD), uid_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_RUID | KERN_PROC_INC_THREAD), ruid_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_PID | KERN_PROC_INC_THREAD), pid_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc, "Process table");
static SYSCTL_NODE(_kern_proc, (KERN_PROC_PROC | KERN_PROC_INC_THREAD), proc_td,
CTLFLAG_RD | CTLFLAG_MPSAFE, sysctl_kern_proc,
"Return process table, including threads");
#ifdef COMPAT_FREEBSD7
static SYSCTL_NODE(_kern_proc, KERN_PROC_OVMMAP, ovmmap, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_ovmmap, "Old Process vm map entries");
#endif
static SYSCTL_NODE(_kern_proc, KERN_PROC_VMMAP, vmmap, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_vmmap, "Process vm map entries");
#if defined(STACK) || defined(DDB)
static SYSCTL_NODE(_kern_proc, KERN_PROC_KSTACK, kstack, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_kstack, "Process kernel stacks");
#endif
static SYSCTL_NODE(_kern_proc, KERN_PROC_GROUPS, groups, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_groups, "Process groups");
static SYSCTL_NODE(_kern_proc, KERN_PROC_RLIMIT, rlimit, CTLFLAG_RW |
CTLFLAG_ANYBODY | CTLFLAG_MPSAFE, sysctl_kern_proc_rlimit,
"Process resource limits");
static SYSCTL_NODE(_kern_proc, KERN_PROC_PS_STRINGS, ps_strings, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_ps_strings,
"Process ps_strings location");
static SYSCTL_NODE(_kern_proc, KERN_PROC_UMASK, umask, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_umask, "Process umask");
static SYSCTL_NODE(_kern_proc, KERN_PROC_OSREL, osrel, CTLFLAG_RW |
CTLFLAG_ANYBODY | CTLFLAG_MPSAFE, sysctl_kern_proc_osrel,
"Process binary osreldate");
static SYSCTL_NODE(_kern_proc, KERN_PROC_SIGTRAMP, sigtramp, CTLFLAG_RD |
CTLFLAG_MPSAFE, sysctl_kern_proc_sigtramp,
"Process signal trampoline location");
static SYSCTL_NODE(_kern_proc, KERN_PROC_SIGFASTBLK, sigfastblk, CTLFLAG_RD |
CTLFLAG_ANYBODY | CTLFLAG_MPSAFE, sysctl_kern_proc_sigfastblk,
"Thread sigfastblock address");
static SYSCTL_NODE(_kern_proc, KERN_PROC_VM_LAYOUT, vm_layout, CTLFLAG_RD |
CTLFLAG_ANYBODY | CTLFLAG_MPSAFE, sysctl_kern_proc_vm_layout,
"Process virtual address space layout info");
static struct sx stop_all_proc_blocker;
SX_SYSINIT(stop_all_proc_blocker, &stop_all_proc_blocker, "sapblk");
bool
stop_all_proc_block(void)
{
return (sx_xlock_sig(&stop_all_proc_blocker) == 0);
}
void
stop_all_proc_unblock(void)
{
sx_xunlock(&stop_all_proc_blocker);
}
int allproc_gen;
/*
* stop_all_proc() purpose is to stop all process which have usermode,
* except current process for obvious reasons. This makes it somewhat
* unreliable when invoked from multithreaded process. The service
* must not be user-callable anyway.
*/
void
stop_all_proc(void)
{
struct proc *cp, *p;
int r, gen;
bool restart, seen_stopped, seen_exiting, stopped_some;
if (!stop_all_proc_block())
return;
cp = curproc;
allproc_loop:
sx_xlock(&allproc_lock);
gen = allproc_gen;
seen_exiting = seen_stopped = stopped_some = restart = false;
LIST_REMOVE(cp, p_list);
LIST_INSERT_HEAD(&allproc, cp, p_list);
for (;;) {
p = LIST_NEXT(cp, p_list);
if (p == NULL)
break;
LIST_REMOVE(cp, p_list);
LIST_INSERT_AFTER(p, cp, p_list);
PROC_LOCK(p);
if ((p->p_flag & (P_KPROC | P_SYSTEM | P_TOTAL_STOP |
P_STOPPED_SIG)) != 0) {
PROC_UNLOCK(p);
continue;
}
if ((p->p_flag2 & P2_WEXIT) != 0) {
seen_exiting = true;
PROC_UNLOCK(p);
continue;
}
if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE) {
/*
* Stopped processes are tolerated when there
* are no other processes which might continue
* them. P_STOPPED_SINGLE but not
* P_TOTAL_STOP process still has at least one
* thread running.
*/
seen_stopped = true;
PROC_UNLOCK(p);
continue;
}
if ((p->p_flag & P_TRACED) != 0) {
/*
* thread_single() below cannot stop traced p,
* so skip it. OTOH, we cannot require
* restart because debugger might be either
* already stopped or traced as well.
*/
PROC_UNLOCK(p);
continue;
}
sx_xunlock(&allproc_lock);
_PHOLD(p);
r = thread_single(p, SINGLE_ALLPROC);
if (r != 0)
restart = true;
else
stopped_some = true;
_PRELE(p);
PROC_UNLOCK(p);
sx_xlock(&allproc_lock);
}
/* Catch forked children we did not see in iteration. */
if (gen != allproc_gen)
restart = true;
sx_xunlock(&allproc_lock);
if (restart || stopped_some || seen_exiting || seen_stopped) {
kern_yield(PRI_USER);
goto allproc_loop;
}
}
void
resume_all_proc(void)
{
struct proc *cp, *p;
cp = curproc;
sx_xlock(&allproc_lock);
again:
LIST_REMOVE(cp, p_list);
LIST_INSERT_HEAD(&allproc, cp, p_list);
for (;;) {
p = LIST_NEXT(cp, p_list);
if (p == NULL)
break;
LIST_REMOVE(cp, p_list);
LIST_INSERT_AFTER(p, cp, p_list);
PROC_LOCK(p);
if ((p->p_flag & P_TOTAL_STOP) != 0) {
sx_xunlock(&allproc_lock);
_PHOLD(p);
thread_single_end(p, SINGLE_ALLPROC);
_PRELE(p);
PROC_UNLOCK(p);
sx_xlock(&allproc_lock);
} else {
PROC_UNLOCK(p);
}
}
/* Did the loop above missed any stopped process ? */
FOREACH_PROC_IN_SYSTEM(p) {
/* No need for proc lock. */
if ((p->p_flag & P_TOTAL_STOP) != 0)
goto again;
}
sx_xunlock(&allproc_lock);
stop_all_proc_unblock();
}
/* #define TOTAL_STOP_DEBUG 1 */
#ifdef TOTAL_STOP_DEBUG
volatile static int ap_resume;
#include <sys/mount.h>
static int
sysctl_debug_stop_all_proc(SYSCTL_HANDLER_ARGS)
{
int error, val;
val = 0;
ap_resume = 0;
error = sysctl_handle_int(oidp, &val, 0, req);
if (error != 0 || req->newptr == NULL)
return (error);
if (val != 0) {
stop_all_proc();
syncer_suspend();
while (ap_resume == 0)
;
syncer_resume();
resume_all_proc();
}
return (0);
}
SYSCTL_PROC(_debug, OID_AUTO, stop_all_proc, CTLTYPE_INT | CTLFLAG_RW |
CTLFLAG_MPSAFE, __DEVOLATILE(int *, &ap_resume), 0,
sysctl_debug_stop_all_proc, "I",
"");
#endif
diff --git a/sys/sys/proc.h b/sys/sys/proc.h
index df7ce6de91d9..2615fa0dc275 100644
--- a/sys/sys/proc.h
+++ b/sys/sys/proc.h
@@ -1,1342 +1,1343 @@
/*-
* SPDX-License-Identifier: BSD-3-Clause
*
* Copyright (c) 1986, 1989, 1991, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#ifndef _SYS_PROC_H_
#define _SYS_PROC_H_
#include <sys/callout.h> /* For struct callout. */
#include <sys/event.h> /* For struct klist. */
#ifdef _KERNEL
#include <sys/_eventhandler.h>
#endif
#include <sys/condvar.h>
#ifndef _KERNEL
#include <sys/filedesc.h>
#endif
#include <sys/queue.h>
#include <sys/_lock.h>
#include <sys/lock_profile.h>
#include <sys/_mutex.h>
#include <sys/osd.h>
#include <sys/priority.h>
#include <sys/rtprio.h> /* XXX. */
#include <sys/runq.h>
#include <sys/resource.h>
#include <sys/sigio.h>
#include <sys/signal.h>
#include <sys/signalvar.h>
#ifndef _KERNEL
#include <sys/time.h> /* For structs itimerval, timeval. */
#else
#include <sys/pcpu.h>
#include <sys/systm.h>
#endif
#include <sys/ucontext.h>
#include <sys/ucred.h>
#include <sys/types.h>
#include <sys/_domainset.h>
#include <machine/proc.h> /* Machine-dependent proc substruct. */
#ifdef _KERNEL
#include <machine/cpu.h>
#endif
/*
* One structure allocated per session.
*
* List of locks
* (m) locked by s_mtx mtx
* (e) locked by proctree_lock sx
* (c) const until freeing
*/
struct session {
u_int s_count; /* Ref cnt; pgrps in session - atomic. */
struct proc *s_leader; /* (m + e) Session leader. */
struct vnode *s_ttyvp; /* (m) Vnode of controlling tty. */
struct cdev_priv *s_ttydp; /* (m) Device of controlling tty. */
struct tty *s_ttyp; /* (e) Controlling tty. */
pid_t s_sid; /* (c) Session ID. */
/* (m) Setlogin() name: */
char s_login[roundup(MAXLOGNAME, sizeof(long))];
struct mtx s_mtx; /* Mutex to protect members. */
};
/*
* One structure allocated per process group.
*
* List of locks
* (m) locked by pg_mtx mtx
* (e) locked by proctree_lock sx
* (c) const until freeing
*/
struct pgrp {
LIST_ENTRY(pgrp) pg_hash; /* (e) Hash chain. */
LIST_HEAD(, proc) pg_members; /* (m + e) Pointer to pgrp members. */
struct session *pg_session; /* (c) Pointer to session. */
struct sigiolst pg_sigiolst; /* (m) List of sigio sources. */
pid_t pg_id; /* (c) Process group id. */
struct mtx pg_mtx; /* Mutex to protect members */
int pg_flags; /* (m) PGRP_ flags */
struct sx pg_killsx; /* Mutual exclusion between group member
* fork() and killpg() */
};
#define PGRP_ORPHANED 0x00000001 /* Group is orphaned */
/*
* pargs, used to hold a copy of the command line, if it had a sane length.
*/
struct pargs {
u_int ar_ref; /* Reference count. */
u_int ar_length; /* Length. */
u_char ar_args[1]; /* Arguments. */
};
/*-
* Description of a process.
*
* This structure contains the information needed to manage a thread of
* control, known in UN*X as a process; it has references to substructures
* containing descriptions of things that the process uses, but may share
* with related processes. The process structure and the substructures
* are always addressable except for those marked "(CPU)" below,
* which might be addressable only on a processor on which the process
* is running.
*
* Below is a key of locks used to protect each member of struct proc. The
* lock is indicated by a reference to a specific character in parens in the
* associated comment.
* * - not yet protected
* a - only touched by curproc or parent during fork/wait
* b - created at fork, never changes
* (exception aiods switch vmspaces, but they are also
* marked 'P_SYSTEM' so hopefully it will be left alone)
* c - locked by proc mtx
* d - locked by allproc_lock lock
* e - locked by proctree_lock lock
* f - session mtx
* g - process group mtx
* h - callout_lock mtx
* i - by curproc or the master session mtx
* j - locked by proc slock
* k - only accessed by curthread
* k*- only accessed by curthread and from an interrupt
* kx- only accessed by curthread and by debugger
* l - the attaching proc or attaching proc parent
* n - not locked, lazy
* o - ktrace lock
* q - td_contested lock
* r - p_peers lock
* s - see sleepq_switch(), sleeping_on_old_rtc(), and sleep(9)
* t - thread lock
* u - process stat lock
* w - process timer lock
* x - created at fork, only changes during single threading in exec
* y - created at first aio, doesn't change until exit or exec at which
* point we are single-threaded and only curthread changes it
*
* If the locking key specifies two identifiers (for example, p_pptr) then
* either lock is sufficient for read access, but both locks must be held
* for write access.
*/
struct cpuset;
struct filecaps;
struct filemon;
struct kaioinfo;
struct kaudit_record;
struct kcov_info;
struct kdtrace_proc;
struct kdtrace_thread;
struct kmsan_td;
struct kq_timer_cb_data;
struct mqueue_notifier;
struct p_sched;
struct proc;
struct procdesc;
struct racct;
struct sbuf;
struct sleepqueue;
struct socket;
struct td_sched;
struct thread;
struct trapframe;
struct turnstile;
struct vm_map;
struct vm_map_entry;
struct epoch_tracker;
struct syscall_args {
u_int code;
u_int original_code;
struct sysent *callp;
register_t args[8];
};
/*
* XXX: Does this belong in resource.h or resourcevar.h instead?
* Resource usage extension. The times in rusage structs in the kernel are
* never up to date. The actual times are kept as runtimes and tick counts
* (with control info in the "previous" times), and are converted when
* userland asks for rusage info. Backwards compatibility prevents putting
* this directly in the user-visible rusage struct.
*
* Locking for p_rux: (cu) means (u) for p_rux and (c) for p_crux.
* Locking for td_rux: (t) for all fields.
*/
struct rusage_ext {
uint64_t rux_runtime; /* (cu) Real time. */
uint64_t rux_uticks; /* (cu) Statclock hits in user mode. */
uint64_t rux_sticks; /* (cu) Statclock hits in sys mode. */
uint64_t rux_iticks; /* (cu) Statclock hits in intr mode. */
uint64_t rux_uu; /* (c) Previous user time in usec. */
uint64_t rux_su; /* (c) Previous sys time in usec. */
uint64_t rux_tu; /* (c) Previous total time in usec. */
};
/*
* Kernel runnable context (thread).
* This is what is put to sleep and reactivated.
* Thread context. Processes may have multiple threads.
*/
struct thread {
struct mtx *volatile td_lock; /* replaces sched lock */
struct proc *td_proc; /* (*) Associated process. */
TAILQ_ENTRY(thread) td_plist; /* (*) All threads in this proc. */
TAILQ_ENTRY(thread) td_runq; /* (t) Run queue. */
union {
TAILQ_ENTRY(thread) td_slpq; /* (t) Sleep queue. */
struct thread *td_zombie; /* Zombie list linkage */
};
TAILQ_ENTRY(thread) td_lockq; /* (t) Lock queue. */
LIST_ENTRY(thread) td_hash; /* (d) Hash chain. */
struct cpuset *td_cpuset; /* (t) CPU affinity mask. */
struct domainset_ref td_domain; /* (a) NUMA policy */
struct seltd *td_sel; /* Select queue/channel. */
struct sleepqueue *td_sleepqueue; /* (k) Associated sleep queue. */
struct turnstile *td_turnstile; /* (k) Associated turnstile. */
void *td_pad1; /* Available */
struct umtx_q *td_umtxq; /* (c?) Link for when we're blocked. */
lwpid_t td_tid; /* (b) Thread ID. */
sigqueue_t td_sigqueue; /* (c) Sigs arrived, not delivered. */
#define td_siglist td_sigqueue.sq_signals
u_char td_lend_user_pri; /* (t) Lend user pri. */
u_char td_allocdomain; /* (b) NUMA domain backing this struct thread. */
u_char td_base_ithread_pri; /* (t) Base ithread pri */
struct kmsan_td *td_kmsan; /* (k) KMSAN state */
/* Cleared during fork1(), thread_create(), or kthread_add(). */
#define td_startzero td_flags
int td_flags; /* (t) TDF_* flags. */
int td_ast; /* (t) TDA_* indicators */
int td_inhibitors; /* (t) Why can not run. */
int td_pflags; /* (k) Private thread (TDP_*) flags. */
int td_pflags2; /* (k) Private thread (TDP2_*) flags. */
int td_dupfd; /* (k) Ret value from fdopen. XXX */
int td_sqqueue; /* (t) Sleepqueue queue blocked on. */
const void *td_wchan; /* (t) Sleep address. */
const char *td_wmesg; /* (t) Reason for sleep. */
volatile u_char td_owepreempt; /* (k*) Preempt on last critical_exit */
u_char td_tsqueue; /* (t) Turnstile queue blocked on. */
u_char _td_pad0[2]; /* Available. */
int td_locks; /* (k) Debug: count of non-spin locks */
int td_rw_rlocks; /* (k) Count of rwlock read locks. */
int td_sx_slocks; /* (k) Count of sx shared locks. */
int td_lk_slocks; /* (k) Count of lockmgr shared locks. */
struct lock_object *td_wantedlock; /* (k) Lock we are contending on */
struct turnstile *td_blocked; /* (t) Lock thread is blocked on. */
const char *td_lockname; /* (t) Name of lock blocked on. */
LIST_HEAD(, turnstile) td_contested; /* (q) Contested locks. */
struct lock_list_entry *td_sleeplocks; /* (k) Held sleep locks. */
int td_intr_nesting_level; /* (k) Interrupt recursion. */
int td_pinned; /* (k) Temporary cpu pin count. */
struct ucred *td_realucred; /* (k) Reference to credentials. */
struct ucred *td_ucred; /* (k) Used credentials, temporarily switchable. */
struct plimit *td_limit; /* (k) Resource limits. */
int td_slptick; /* (t) Time at sleep. */
int td_blktick; /* (t) Time spent blocked. */
int td_swvoltick; /* (t) Time at last SW_VOL switch. */
int td_swinvoltick; /* (t) Time at last SW_INVOL switch. */
u_int td_cow; /* (*) Number of copy-on-write faults */
struct rusage td_ru; /* (t) rusage information. */
struct rusage_ext td_rux; /* (t) Internal rusage information. */
uint64_t td_incruntime; /* (t) Cpu ticks to transfer to proc. */
uint64_t td_runtime; /* (t) How many cpu ticks we've run. */
u_int td_pticks; /* (t) Statclock hits for profiling */
u_int td_sticks; /* (t) Statclock hits in system mode. */
u_int td_iticks; /* (t) Statclock hits in intr mode. */
u_int td_uticks; /* (t) Statclock hits in user mode. */
int td_intrval; /* (t) Return value for sleepq. */
sigset_t td_oldsigmask; /* (k) Saved mask from pre sigpause. */
volatile u_int td_generation; /* (k) For detection of preemption */
stack_t td_sigstk; /* (k) Stack ptr and on-stack flag. */
int td_xsig; /* (c) Signal for ptrace */
u_long td_profil_addr; /* (k) Temporary addr until AST. */
u_int td_profil_ticks; /* (k) Temporary ticks until AST. */
char td_name[MAXCOMLEN + 1]; /* (*) Thread name. */
struct file *td_fpop; /* (k) file referencing cdev under op */
int td_dbgflags; /* (c) Userland debugger flags */
siginfo_t td_si; /* (c) For debugger or core file */
int td_ng_outbound; /* (k) Thread entered ng from above. */
struct osd td_osd; /* (k) Object specific data. */
struct vm_map_entry *td_map_def_user; /* (k) Deferred entries. */
pid_t td_dbg_forked; /* (c) Child pid for debugger. */
u_int td_no_sleeping; /* (k) Sleeping disabled count. */
struct vnode *td_vp_reserved;/* (k) Preallocated vnode. */
void *td_su; /* (k) FFS SU private */
sbintime_t td_sleeptimo; /* (t) Sleep timeout. */
int td_rtcgen; /* (s) rtc_generation of abs. sleep */
int td_errno; /* (k) Error from last syscall. */
size_t td_vslock_sz; /* (k) amount of vslock-ed space */
struct kcov_info *td_kcov_info; /* (*) Kernel code coverage data */
long td_ucredref; /* (k) references on td_realucred */
#define td_endzero td_sigmask
/* Copied during fork1(), thread_create(), or kthread_add(). */
#define td_startcopy td_endzero
sigset_t td_sigmask; /* (c) Current signal mask. */
u_char td_rqindex; /* (t) Run queue index. */
u_char td_base_pri; /* (t) Thread base kernel priority. */
u_char td_priority; /* (t) Thread active priority. */
u_char td_pri_class; /* (t) Scheduling class. */
u_char td_user_pri; /* (t) User pri from estcpu and nice. */
u_char td_base_user_pri; /* (t) Base user pri */
uintptr_t td_rb_list; /* (k) Robust list head. */
uintptr_t td_rbp_list; /* (k) Robust priv list head. */
uintptr_t td_rb_inact; /* (k) Current in-action mutex loc. */
struct syscall_args td_sa; /* (kx) Syscall parameters. Copied on
fork for child tracing. */
void *td_sigblock_ptr; /* (k) uptr for fast sigblock. */
uint32_t td_sigblock_val; /* (k) fast sigblock value read at
td_sigblock_ptr on kern entry */
#define td_endcopy td_pcb
/*
* Fields that must be manually set in fork1(), thread_create(), kthread_add(),
* or already have been set in the allocator, constructor, etc.
*/
struct pcb *td_pcb; /* (k) Kernel VA of pcb and kstack. */
enum td_states {
TDS_INACTIVE = 0x0,
TDS_INHIBITED,
TDS_CAN_RUN,
TDS_RUNQ,
TDS_RUNNING
} td_state; /* (t) thread state */
/* Note: td_state must be accessed using TD_{GET,SET}_STATE(). */
union {
syscallarg_t tdu_retval[2];
off_t tdu_off;
} td_uretoff; /* (k) Syscall aux returns. */
#define td_retval td_uretoff.tdu_retval
u_int td_cowgen; /* (k) Generation of COW pointers. */
/* LP64 hole */
struct callout td_slpcallout; /* (h) Callout for sleep. */
struct trapframe *td_frame; /* (k) */
vm_offset_t td_kstack; /* (a) Kernel VA of kstack. */
u_short td_kstack_pages; /* (a) Size of the kstack. */
u_short td_kstack_domain; /* (a) Domain backing kstack KVA. */
volatile u_int td_critnest; /* (k*) Critical section nest level. */
struct mdthread td_md; /* (k) Any machine-dependent fields. */
struct kaudit_record *td_ar; /* (k) Active audit record, if any. */
struct lpohead td_lprof[2]; /* (a) lock profiling objects. */
struct kdtrace_thread *td_dtrace; /* (*) DTrace-specific data. */
struct vnet *td_vnet; /* (k) Effective vnet. */
const char *td_vnet_lpush; /* (k) Debugging vnet push / pop. */
struct trapframe *td_intr_frame;/* (k) Frame of the current irq */
struct proc *td_rfppwait_p; /* (k) The vforked child */
struct vm_page **td_ma; /* (k) uio pages held */
int td_ma_cnt; /* (k) size of *td_ma */
/* LP64 hole */
void *td_emuldata; /* Emulator state data */
int td_lastcpu; /* (t) Last cpu we were on. */
int td_oncpu; /* (t) Which cpu we are on. */
void *td_lkpi_task; /* LinuxKPI task struct pointer */
int td_pmcpend;
void *td_remotereq; /* (c) dbg remote request. */
off_t td_ktr_io_lim; /* (k) limit for ktrace file size */
#ifdef EPOCH_TRACE
SLIST_HEAD(, epoch_tracker) td_epochs;
#endif
};
struct thread0_storage {
struct thread t0st_thread;
uint64_t t0st_sched[10];
};
struct mtx *thread_lock_block(struct thread *);
void thread_lock_block_wait(struct thread *);
void thread_lock_set(struct thread *, struct mtx *);
void thread_lock_unblock(struct thread *, struct mtx *);
#define THREAD_LOCK_ASSERT(td, type) \
mtx_assert((td)->td_lock, (type))
#define THREAD_LOCK_BLOCKED_ASSERT(td, type) \
do { \
struct mtx *__m = (td)->td_lock; \
if (__m != &blocked_lock) \
mtx_assert(__m, (type)); \
} while (0)
#ifdef INVARIANTS
#define THREAD_LOCKPTR_ASSERT(td, lock) \
do { \
struct mtx *__m; \
__m = (td)->td_lock; \
KASSERT(__m == (lock), \
("Thread %p lock %p does not match %p", td, __m, (lock))); \
} while (0)
#define THREAD_LOCKPTR_BLOCKED_ASSERT(td, lock) \
do { \
struct mtx *__m; \
__m = (td)->td_lock; \
KASSERT(__m == (lock) || __m == &blocked_lock, \
("Thread %p lock %p does not match %p", td, __m, (lock))); \
} while (0)
#define TD_LOCKS_INC(td) ((td)->td_locks++)
#define TD_LOCKS_DEC(td) do { \
KASSERT(SCHEDULER_STOPPED() || (td)->td_locks > 0, \
("Thread %p owns no locks", (td))); \
(td)->td_locks--; \
} while (0)
#else
#define THREAD_LOCKPTR_ASSERT(td, lock)
#define THREAD_LOCKPTR_BLOCKED_ASSERT(td, lock)
#define TD_LOCKS_INC(td)
#define TD_LOCKS_DEC(td)
#endif
/*
* Flags kept in td_flags:
* To change these you MUST have the scheduler lock.
*/
#define TDF_BORROWING 0x00000001 /* Thread is borrowing pri from another. */
#define TDF_INPANIC 0x00000002 /* Caused a panic, let it drive crashdump. */
#define TDF_INMEM 0x00000004 /* Thread's stack is in memory. */
#define TDF_SINTR 0x00000008 /* Sleep is interruptible. */
#define TDF_TIMEOUT 0x00000010 /* Timing out during sleep. */
#define TDF_IDLETD 0x00000020 /* This is a per-CPU idle thread. */
#define TDF_UNUSED11 0x00000040 /* Available */
#define TDF_SIGWAIT 0x00000080 /* Ignore ignored signals */
#define TDF_KTH_SUSP 0x00000100 /* kthread is suspended */
#define TDF_ALLPROCSUSP 0x00000200 /* suspended by SINGLE_ALLPROC */
#define TDF_BOUNDARY 0x00000400 /* Thread suspended at user boundary */
#define TDF_UNUSED1 0x00000800 /* Available */
#define TDF_UNUSED2 0x00001000 /* Available */
#define TDF_SBDRY 0x00002000 /* Stop only on usermode boundary. */
#define TDF_UPIBLOCKED 0x00004000 /* Thread blocked on user PI mutex. */
#define TDF_UNUSED3 0x00008000 /* Available */
#define TDF_UNUSED4 0x00010000 /* Available */
#define TDF_UNUSED5 0x00020000 /* Available */
#define TDF_NOLOAD 0x00040000 /* Ignore during load avg calculations. */
#define TDF_SERESTART 0x00080000 /* ERESTART on stop attempts. */
#define TDF_THRWAKEUP 0x00100000 /* Libthr thread must not suspend itself. */
#define TDF_SEINTR 0x00200000 /* EINTR on stop attempts. */
#define TDF_UNUSED12 0x00400000 /* Available */
#define TDF_UNUSED6 0x00800000 /* Available */
#define TDF_SCHED0 0x01000000 /* Reserved for scheduler private use */
#define TDF_SCHED1 0x02000000 /* Reserved for scheduler private use */
#define TDF_SCHED2 0x04000000 /* Reserved for scheduler private use */
#define TDF_SCHED3 0x08000000 /* Reserved for scheduler private use */
#define TDF_UNUSED7 0x10000000 /* Available */
#define TDF_UNUSED8 0x20000000 /* Available */
#define TDF_UNUSED9 0x40000000 /* Available */
#define TDF_UNUSED10 0x80000000 /* Available */
enum {
TDA_AST = 0, /* Special: call all non-flagged AST handlers */
TDA_OWEUPC,
TDA_HWPMC,
TDA_VFORK,
TDA_ALRM,
TDA_PROF,
TDA_MAC,
TDA_SCHED,
TDA_UFS,
TDA_GEOM,
TDA_KQUEUE,
TDA_RACCT,
TDA_MOD1, /* For third party use, before signals are */
TDA_MOD2, /* processed .. */
TDA_PSELECT, /* For discarding temporary signal mask */
TDA_SIG,
TDA_KTRACE,
TDA_SUSPEND,
TDA_SIGSUSPEND,
TDA_MOD3, /* .. and after */
TDA_MOD4,
TDA_MAX,
};
#define TDAI(tda) (1U << (tda))
#define td_ast_pending(td, tda) ((td->td_ast & TDAI(tda)) != 0)
/* Userland debug flags */
#define TDB_SUSPEND 0x00000001 /* Thread is suspended by debugger */
#define TDB_XSIG 0x00000002 /* Thread is exchanging signal under trace */
#define TDB_USERWR 0x00000004 /* Debugger modified memory or registers */
#define TDB_SCE 0x00000008 /* Thread performs syscall enter */
#define TDB_SCX 0x00000010 /* Thread performs syscall exit */
#define TDB_EXEC 0x00000020 /* TDB_SCX from exec(2) family */
#define TDB_FORK 0x00000040 /* TDB_SCX from fork(2) that created new
process */
#define TDB_STOPATFORK 0x00000080 /* Stop at the return from fork (child
only) */
#define TDB_CHILD 0x00000100 /* New child indicator for ptrace() */
#define TDB_BORN 0x00000200 /* New LWP indicator for ptrace() */
#define TDB_EXIT 0x00000400 /* Exiting LWP indicator for ptrace() */
#define TDB_VFORK 0x00000800 /* vfork indicator for ptrace() */
#define TDB_FSTP 0x00001000 /* The thread is PT_ATTACH leader */
#define TDB_STEP 0x00002000 /* (x86) PSL_T set for PT_STEP */
#define TDB_SSWITCH 0x00004000 /* Suspended in ptracestop */
#define TDB_BOUNDARY 0x00008000 /* ptracestop() at boundary */
#define TDB_COREDUMPREQ 0x00010000 /* Coredump request */
#define TDB_SCREMOTEREQ 0x00020000 /* Remote syscall request */
/*
* "Private" flags kept in td_pflags:
* These are only written by curthread and thus need no locking.
*/
#define TDP_OLDMASK 0x00000001 /* Need to restore mask after suspend. */
#define TDP_INKTR 0x00000002 /* Thread is currently in KTR code. */
#define TDP_INKTRACE 0x00000004 /* Thread is currently in KTRACE code. */
#define TDP_BUFNEED 0x00000008 /* Do not recurse into the buf flush */
#define TDP_COWINPROGRESS 0x00000010 /* Snapshot copy-on-write in progress. */
#define TDP_ALTSTACK 0x00000020 /* Have alternate signal stack. */
#define TDP_DEADLKTREAT 0x00000040 /* Lock acquisition - deadlock treatment. */
#define TDP_NOFAULTING 0x00000080 /* Do not handle page faults. */
#define TDP_SIGFASTBLOCK 0x00000100 /* Fast sigblock active */
#define TDP_OWEUPC 0x00000200 /* Call addupc() at next AST. */
#define TDP_ITHREAD 0x00000400 /* Thread is an interrupt thread. */
#define TDP_SYNCIO 0x00000800 /* Local override, disable async i/o. */
#define TDP_SCHED1 0x00001000 /* Reserved for scheduler private use */
#define TDP_SCHED2 0x00002000 /* Reserved for scheduler private use */
#define TDP_SCHED3 0x00004000 /* Reserved for scheduler private use */
#define TDP_SCHED4 0x00008000 /* Reserved for scheduler private use */
#define TDP_GEOM 0x00010000 /* Settle GEOM before finishing syscall */
#define TDP_SOFTDEP 0x00020000 /* Stuck processing softdep worklist */
#define TDP_NORUNNINGBUF 0x00040000 /* Ignore runningbufspace check */
#define TDP_WAKEUP 0x00080000 /* Don't sleep in umtx cond_wait */
#define TDP_INBDFLUSH 0x00100000 /* Already in BO_BDFLUSH, do not recurse */
#define TDP_KTHREAD 0x00200000 /* This is an official kernel thread */
#define TDP_CALLCHAIN 0x00400000 /* Capture thread's callchain */
#define TDP_IGNSUSP 0x00800000 /* Permission to ignore the MNTK_SUSPEND* */
#define TDP_AUDITREC 0x01000000 /* Audit record pending on thread */
#define TDP_RFPPWAIT 0x02000000 /* Handle RFPPWAIT on syscall exit */
#define TDP_RESETSPUR 0x04000000 /* Reset spurious page fault history. */
#define TDP_NERRNO 0x08000000 /* Last errno is already in td_errno */
#define TDP_UIOHELD 0x10000000 /* Current uio has pages held in td_ma */
#define TDP_EFIRT 0x20000000 /* In firmware (EFI RT) call */
#define TDP_EXECVMSPC 0x40000000 /* Execve destroyed old vmspace */
#define TDP_SIGFASTPENDING 0x80000000 /* Pending signal due to sigfastblock */
#define TDP2_SBPAGES 0x00000001 /* Owns sbusy on some pages */
#define TDP2_COMPAT32RB 0x00000002 /* compat32 ABI for robust lists */
#define TDP2_ACCT 0x00000004 /* Doing accounting */
#define TDP2_SAN_QUIET 0x00000008 /* Disable warnings from K(A|M)SAN */
/*
* Reasons that the current thread can not be run yet.
* More than one may apply.
*/
#define TDI_SUSPENDED 0x0001 /* On suspension queue. */
#define TDI_SLEEPING 0x0002 /* Actually asleep! (tricky). */
#define TDI_LOCK 0x0008 /* Stopped on a lock. */
#define TDI_IWAIT 0x0010 /* Awaiting interrupt. */
#define TD_IS_SLEEPING(td) ((td)->td_inhibitors & TDI_SLEEPING)
#define TD_ON_SLEEPQ(td) ((td)->td_wchan != NULL)
#define TD_IS_SUSPENDED(td) ((td)->td_inhibitors & TDI_SUSPENDED)
#define TD_ON_LOCK(td) ((td)->td_inhibitors & TDI_LOCK)
#define TD_AWAITING_INTR(td) ((td)->td_inhibitors & TDI_IWAIT)
#ifdef _KERNEL
#define TD_GET_STATE(td) atomic_load_int(&(td)->td_state)
#else
#define TD_GET_STATE(td) ((td)->td_state)
#endif
#define TD_IS_RUNNING(td) (TD_GET_STATE(td) == TDS_RUNNING)
#define TD_ON_RUNQ(td) (TD_GET_STATE(td) == TDS_RUNQ)
#define TD_CAN_RUN(td) (TD_GET_STATE(td) == TDS_CAN_RUN)
#define TD_IS_INHIBITED(td) (TD_GET_STATE(td) == TDS_INHIBITED)
#define TD_ON_UPILOCK(td) ((td)->td_flags & TDF_UPIBLOCKED)
#define TD_IS_IDLETHREAD(td) ((td)->td_flags & TDF_IDLETD)
#define TD_CAN_ABORT(td) (TD_ON_SLEEPQ((td)) && \
((td)->td_flags & TDF_SINTR) != 0)
#define KTDSTATE(td) \
(((td)->td_inhibitors & TDI_SLEEPING) != 0 ? "sleep" : \
((td)->td_inhibitors & TDI_SUSPENDED) != 0 ? "suspended" : \
((td)->td_inhibitors & TDI_LOCK) != 0 ? "blocked" : \
((td)->td_inhibitors & TDI_IWAIT) != 0 ? "iwait" : "yielding")
#define TD_SET_INHIB(td, inhib) do { \
TD_SET_STATE(td, TDS_INHIBITED); \
(td)->td_inhibitors |= (inhib); \
} while (0)
#define TD_CLR_INHIB(td, inhib) do { \
if (((td)->td_inhibitors & (inhib)) && \
(((td)->td_inhibitors &= ~(inhib)) == 0)) \
TD_SET_STATE(td, TDS_CAN_RUN); \
} while (0)
#define TD_SET_SLEEPING(td) TD_SET_INHIB((td), TDI_SLEEPING)
#define TD_SET_LOCK(td) TD_SET_INHIB((td), TDI_LOCK)
#define TD_SET_SUSPENDED(td) TD_SET_INHIB((td), TDI_SUSPENDED)
#define TD_SET_IWAIT(td) TD_SET_INHIB((td), TDI_IWAIT)
#define TD_SET_EXITING(td) TD_SET_INHIB((td), TDI_EXITING)
#define TD_CLR_SLEEPING(td) TD_CLR_INHIB((td), TDI_SLEEPING)
#define TD_CLR_LOCK(td) TD_CLR_INHIB((td), TDI_LOCK)
#define TD_CLR_SUSPENDED(td) TD_CLR_INHIB((td), TDI_SUSPENDED)
#define TD_CLR_IWAIT(td) TD_CLR_INHIB((td), TDI_IWAIT)
#ifdef _KERNEL
#define TD_SET_STATE(td, state) atomic_store_int(&(td)->td_state, state)
#else
#define TD_SET_STATE(td, state) (td)->td_state = state
#endif
#define TD_SET_RUNNING(td) TD_SET_STATE(td, TDS_RUNNING)
#define TD_SET_RUNQ(td) TD_SET_STATE(td, TDS_RUNQ)
#define TD_SET_CAN_RUN(td) TD_SET_STATE(td, TDS_CAN_RUN)
#define TD_SBDRY_INTR(td) \
(((td)->td_flags & (TDF_SEINTR | TDF_SERESTART)) != 0)
#define TD_SBDRY_ERRNO(td) \
(((td)->td_flags & TDF_SEINTR) != 0 ? EINTR : ERESTART)
/*
* Process structure.
*/
struct proc {
LIST_ENTRY(proc) p_list; /* (d) List of all processes. */
TAILQ_HEAD(, thread) p_threads; /* (c) all threads. */
struct mtx p_slock; /* process spin lock */
struct ucred *p_ucred; /* (c) Process owner's identity. */
struct filedesc *p_fd; /* (b) Open files. */
struct filedesc_to_leader *p_fdtol; /* (b) Tracking node */
struct pwddesc *p_pd; /* (b) Cwd, chroot, jail, umask */
struct pstats *p_stats; /* (b) Accounting/statistics (CPU). */
struct plimit *p_limit; /* (c) Resource limits. */
struct callout p_limco; /* (c) Limit callout handle */
struct sigacts *p_sigacts; /* (x) Signal actions, state (CPU). */
int p_flag; /* (c) P_* flags. */
int p_flag2; /* (c) P2_* flags. */
enum p_states {
PRS_NEW = 0, /* In creation */
PRS_NORMAL, /* threads can be run. */
PRS_ZOMBIE
} p_state; /* (j/c) Process status. */
pid_t p_pid; /* (b) Process identifier. */
LIST_ENTRY(proc) p_hash; /* (d) Hash chain. */
LIST_ENTRY(proc) p_pglist; /* (g + e) List of processes in pgrp. */
struct proc *p_pptr; /* (c + e) Pointer to parent process. */
LIST_ENTRY(proc) p_sibling; /* (e) List of sibling processes. */
LIST_HEAD(, proc) p_children; /* (e) Pointer to list of children. */
struct proc *p_reaper; /* (e) My reaper. */
LIST_HEAD(, proc) p_reaplist; /* (e) List of my descendants
(if I am reaper). */
LIST_ENTRY(proc) p_reapsibling; /* (e) List of siblings - descendants of
the same reaper. */
struct mtx p_mtx; /* (n) Lock for this struct. */
struct mtx p_statmtx; /* Lock for the stats */
struct mtx p_itimmtx; /* Lock for the virt/prof timers */
struct mtx p_profmtx; /* Lock for the profiling */
struct ksiginfo *p_ksi; /* Locked by parent proc lock */
sigqueue_t p_sigqueue; /* (c) Sigs not delivered to a td. */
#define p_siglist p_sigqueue.sq_signals
pid_t p_oppid; /* (c + e) Real parent pid. */
/* The following fields are all zeroed upon creation in fork. */
#define p_startzero p_vmspace
struct vmspace *p_vmspace; /* (b) Address space. */
u_int p_swtick; /* (c) Tick when swapped in or out. */
u_int p_cowgen; /* (c) Generation of COW pointers. */
struct itimerval p_realtimer; /* (c) Alarm timer. */
struct rusage p_ru; /* (a) Exit information. */
struct rusage_ext p_rux; /* (cu) Internal resource usage. */
struct rusage_ext p_crux; /* (c) Internal child resource usage. */
int p_profthreads; /* (c) Num threads in addupc_task. */
volatile int p_exitthreads; /* (j) Number of threads exiting */
int p_traceflag; /* (o) Kernel trace points. */
struct ktr_io_params *p_ktrioparms; /* (c + o) Params for ktrace. */
struct vnode *p_textvp; /* (b) Vnode of executable. */
struct vnode *p_textdvp; /* (b) Dir containing textvp. */
char *p_binname; /* (b) Binary hardlink name. */
u_int p_lock; /* (c) Prevent exit. */
struct sigiolst p_sigiolst; /* (c) List of sigio sources. */
int p_sigparent; /* (c) Signal to parent on exit. */
int p_sig; /* (n) For core dump/debugger XXX. */
u_int p_ptevents; /* (c + e) ptrace() event mask. */
struct kaioinfo *p_aioinfo; /* (y) ASYNC I/O info. */
struct thread *p_singlethread;/* (c + j) If single threading this is it */
int p_suspcount; /* (j) Num threads in suspended mode. */
struct thread *p_xthread; /* (c) Trap thread */
int p_boundary_count;/* (j) Num threads at user boundary */
int p_pendingcnt; /* (c) how many signals are pending */
struct itimers *p_itimers; /* (c) POSIX interval timers. */
struct procdesc *p_procdesc; /* (e) Process descriptor, if any. */
u_int p_treeflag; /* (e) P_TREE flags */
int p_pendingexits; /* (c) Count of pending thread exits. */
struct filemon *p_filemon; /* (c) filemon-specific data. */
int p_pdeathsig; /* (c) Signal from parent on exit. */
/* End area that is zeroed on creation. */
#define p_endzero p_magic
/* The following fields are all copied upon creation in fork. */
#define p_startcopy p_endzero
u_int p_magic; /* (b) Magic number. */
int p_osrel; /* (x) osreldate for the
binary (from ELF note, if any) */
uint32_t p_fctl0; /* (x) ABI feature control, ELF note */
char p_comm[MAXCOMLEN + 1]; /* (x) Process name. */
struct sysentvec *p_sysent; /* (b) Syscall dispatch info. */
struct pargs *p_args; /* (c) Process arguments. */
rlim_t p_cpulimit; /* (c) Current CPU limit in seconds. */
signed char p_nice; /* (c) Process "nice" value. */
int p_fibnum; /* in this routing domain XXX MRT */
pid_t p_reapsubtree; /* (e) Pid of the direct child of the
reaper which spawned
our subtree. */
uint64_t p_elf_flags; /* (x) ELF flags */
void *p_elf_brandinfo; /* (x) Elf_Brandinfo, NULL for
non ELF binaries. */
sbintime_t p_umtx_min_timeout;
/* End area that is copied on creation. */
#define p_endcopy p_xexit
u_int p_xexit; /* (c) Exit code. */
u_int p_xsig; /* (c) Stop/kill sig. */
struct pgrp *p_pgrp; /* (c + e) Pointer to process group. */
struct knlist *p_klist; /* (c) Knotes attached to this proc. */
int p_numthreads; /* (c) Number of threads. */
struct mdproc p_md; /* Any machine-dependent fields. */
struct callout p_itcallout; /* (h + c) Interval timer callout. */
u_short p_acflag; /* (c) Accounting flags. */
struct proc *p_peers; /* (r) */
struct proc *p_leader; /* (b) */
void *p_emuldata; /* (c) Emulator state data. */
struct label *p_label; /* (*) Proc (not subject) MAC label. */
STAILQ_HEAD(, ktr_request) p_ktr; /* (o) KTR event queue. */
LIST_HEAD(, mqueue_notifier) p_mqnotifier; /* (c) mqueue notifiers.*/
struct kdtrace_proc *p_dtrace; /* (*) DTrace-specific data. */
struct cv p_pwait; /* (*) wait cv for exit/exec. */
uint64_t p_prev_runtime; /* (c) Resource usage accounting. */
struct racct *p_racct; /* (b) Resource accounting. */
int p_throttled; /* (c) Flag for racct pcpu throttling */
/*
* An orphan is the child that has been re-parented to the
* debugger as a result of attaching to it. Need to keep
* track of them for parent to be able to collect the exit
* status of what used to be children.
*/
LIST_ENTRY(proc) p_orphan; /* (e) List of orphan processes. */
LIST_HEAD(, proc) p_orphans; /* (e) Pointer to list of orphans. */
TAILQ_HEAD(, kq_timer_cb_data) p_kqtim_stop; /* (c) */
LIST_ENTRY(proc) p_jaillist; /* (d) Jail process linkage. */
};
#define p_session p_pgrp->pg_session
#define p_pgid p_pgrp->pg_id
#define NOCPU (-1) /* For when we aren't on a CPU. */
#define NOCPU_OLD (255)
#define MAXCPU_OLD (254)
#define PROC_SLOCK(p) mtx_lock_spin(&(p)->p_slock)
#define PROC_SUNLOCK(p) mtx_unlock_spin(&(p)->p_slock)
#define PROC_SLOCK_ASSERT(p, type) mtx_assert(&(p)->p_slock, (type))
#define PROC_STATLOCK(p) mtx_lock_spin(&(p)->p_statmtx)
#define PROC_STATUNLOCK(p) mtx_unlock_spin(&(p)->p_statmtx)
#define PROC_STATLOCK_ASSERT(p, type) mtx_assert(&(p)->p_statmtx, (type))
#define PROC_ITIMLOCK(p) mtx_lock_spin(&(p)->p_itimmtx)
#define PROC_ITIMUNLOCK(p) mtx_unlock_spin(&(p)->p_itimmtx)
#define PROC_ITIMLOCK_ASSERT(p, type) mtx_assert(&(p)->p_itimmtx, (type))
#define PROC_PROFLOCK(p) mtx_lock_spin(&(p)->p_profmtx)
#define PROC_PROFUNLOCK(p) mtx_unlock_spin(&(p)->p_profmtx)
#define PROC_PROFLOCK_ASSERT(p, type) mtx_assert(&(p)->p_profmtx, (type))
/* These flags are kept in p_flag. */
#define P_ADVLOCK 0x00000001 /* Process may hold a POSIX advisory
lock. */
#define P_CONTROLT 0x00000002 /* Has a controlling terminal. */
#define P_KPROC 0x00000004 /* Kernel process. */
#define P_UNUSED3 0x00000008 /* --available-- */
#define P_PPWAIT 0x00000010 /* Parent is waiting for child to
exec/exit. */
#define P_PROFIL 0x00000020 /* Has started profiling. */
#define P_STOPPROF 0x00000040 /* Has thread requesting to stop
profiling. */
#define P_HADTHREADS 0x00000080 /* Has had threads (no cleanup
shortcuts) */
#define P_SUGID 0x00000100 /* Had set id privileges since last
exec. */
#define P_SYSTEM 0x00000200 /* System proc: no sigs or stats. */
#define P_SINGLE_EXIT 0x00000400 /* Threads suspending should exit,
not wait. */
#define P_TRACED 0x00000800 /* Debugged process being traced. */
#define P_WAITED 0x00001000 /* Someone is waiting for us. */
#define P_WEXIT 0x00002000 /* Working on exiting. */
#define P_EXEC 0x00004000 /* Process called exec. */
#define P_WKILLED 0x00008000 /* Killed, go to kernel/user boundary
ASAP. */
#define P_CONTINUED 0x00010000 /* Proc has continued from a stopped
state. */
#define P_STOPPED_SIG 0x00020000 /* Stopped due to SIGSTOP/SIGTSTP. */
#define P_STOPPED_TRACE 0x00040000 /* Stopped because of tracing. */
#define P_STOPPED_SINGLE 0x00080000 /* Only 1 thread can continue (not to
user). */
#define P_PROTECTED 0x00100000 /* Do not kill on memory overcommit. */
#define P_SIGEVENT 0x00200000 /* Process pending signals changed. */
#define P_SINGLE_BOUNDARY 0x00400000 /* Threads should suspend at user
boundary. */
#define P_HWPMC 0x00800000 /* Process is using HWPMCs */
#define P_JAILED 0x01000000 /* Process is in jail. */
#define P_TOTAL_STOP 0x02000000 /* Stopped in stop_all_proc. */
#define P_INEXEC 0x04000000 /* Process is in execve(). */
#define P_STATCHILD 0x08000000 /* Child process stopped or exited. */
#define P_INMEM 0x10000000 /* Loaded into memory, always set. */
#define P_UNUSED1 0x20000000 /* --available-- */
#define P_UNUSED2 0x40000000 /* --available-- */
#define P_PPTRACE 0x80000000 /* PT_TRACEME by vforked child. */
#define P_STOPPED (P_STOPPED_SIG|P_STOPPED_SINGLE|P_STOPPED_TRACE)
#define P_SHOULDSTOP(p) ((p)->p_flag & P_STOPPED)
#define P_KILLED(p) ((p)->p_flag & P_WKILLED)
/* These flags are kept in p_flag2. */
#define P2_INHERIT_PROTECTED 0x00000001 /* New children get
P_PROTECTED. */
#define P2_NOTRACE 0x00000002 /* No ptrace(2) attach or
coredumps. */
#define P2_NOTRACE_EXEC 0x00000004 /* Keep P2_NOPTRACE on
exec(2). */
#define P2_AST_SU 0x00000008 /* Handles SU ast for
kthreads. */
#define P2_PTRACE_FSTP 0x00000010 /* SIGSTOP from PT_ATTACH not
yet handled. */
#define P2_TRAPCAP 0x00000020 /* SIGTRAP on ENOTCAPABLE */
#define P2_ASLR_ENABLE 0x00000040 /* Force enable ASLR. */
#define P2_ASLR_DISABLE 0x00000080 /* Force disable ASLR. */
#define P2_ASLR_IGNSTART 0x00000100 /* Enable ASLR to consume sbrk
area. */
#define P2_PROTMAX_ENABLE 0x00000200 /* Force enable implied
PROT_MAX. */
#define P2_PROTMAX_DISABLE 0x00000400 /* Force disable implied
PROT_MAX. */
#define P2_STKGAP_DISABLE 0x00000800 /* Disable stack gap for
MAP_STACK */
#define P2_STKGAP_DISABLE_EXEC 0x00001000 /* Stack gap disabled
after exec */
#define P2_ITSTOPPED 0x00002000 /* itimers stopped */
#define P2_PTRACEREQ 0x00004000 /* Active ptrace req */
#define P2_NO_NEW_PRIVS 0x00008000 /* Ignore setuid */
#define P2_WXORX_DISABLE 0x00010000 /* WX mappings enabled */
#define P2_WXORX_ENABLE_EXEC 0x00020000 /* WXORX enabled after exec */
#define P2_WEXIT 0x00040000 /* exit just started, no
external thread_single() is
permitted */
#define P2_REAPKILLED 0x00080000 /* REAP_KILL pass touched me */
#define P2_MEMBAR_PRIVE 0x00100000 /* membar private expedited
registered */
#define P2_MEMBAR_PRIVE_SYNCORE 0x00200000 /* membar private expedited
sync core registered */
#define P2_MEMBAR_GLOBE 0x00400000 /* membar global expedited
registered */
/* Flags protected by proctree_lock, kept in p_treeflags. */
#define P_TREE_ORPHANED 0x00000001 /* Reparented, on orphan list */
#define P_TREE_FIRST_ORPHAN 0x00000002 /* First element of orphan
list */
#define P_TREE_REAPER 0x00000004 /* Reaper of subtree */
#define P_TREE_GRPEXITED 0x00000008 /* exit1() done with job ctl */
/*
* These were process status values (p_stat), now they are only used in
* legacy conversion code.
*/
#define SIDL 1 /* Process being created by fork. */
#define SRUN 2 /* Currently runnable. */
#define SSLEEP 3 /* Sleeping on an address. */
#define SSTOP 4 /* Process debugging or suspension. */
#define SZOMB 5 /* Awaiting collection by parent. */
#define SWAIT 6 /* Waiting for interrupt. */
#define SLOCK 7 /* Blocked on a lock. */
#define P_MAGIC 0xbeefface
#ifdef _KERNEL
/* Types and flags for mi_switch(9). */
#define SW_TYPE_MASK 0xff /* First 8 bits are switch type */
#define SWT_OWEPREEMPT 1 /* Switching due to owepreempt. */
#define SWT_TURNSTILE 2 /* Turnstile contention. */
#define SWT_SLEEPQ 3 /* Sleepq wait. */
#define SWT_RELINQUISH 4 /* yield call. */
#define SWT_NEEDRESCHED 5 /* NEEDRESCHED was set. */
#define SWT_IDLE 6 /* Switching from the idle thread. */
#define SWT_IWAIT 7 /* Waiting for interrupts. */
#define SWT_SUSPEND 8 /* Thread suspended. */
#define SWT_REMOTEPREEMPT 9 /* Remote processor preempted. */
#define SWT_REMOTEWAKEIDLE 10 /* Remote processor preempted idle. */
#define SWT_BIND 11 /* Thread bound to a new CPU. */
#define SWT_COUNT 12 /* Number of switch types. */
/* Flags */
#define SW_VOL 0x0100 /* Voluntary switch. */
#define SW_INVOL 0x0200 /* Involuntary switch. */
#define SW_PREEMPT 0x0400 /* The invol switch is a preemption */
/* How values for thread_single(). */
#define SINGLE_NO_EXIT 0
#define SINGLE_EXIT 1
#define SINGLE_BOUNDARY 2
#define SINGLE_ALLPROC 3
#define FOREACH_PROC_IN_SYSTEM(p) \
LIST_FOREACH((p), &allproc, p_list)
#define FOREACH_THREAD_IN_PROC(p, td) \
TAILQ_FOREACH((td), &(p)->p_threads, td_plist)
#define FIRST_THREAD_IN_PROC(p) TAILQ_FIRST(&(p)->p_threads)
/*
* We use process IDs <= pid_max <= PID_MAX; PID_MAX + 1 must also fit
* in a pid_t, as it is used to represent "no process group".
*/
#define PID_MAX 99999
#define NO_PID (PID_MAX + 1)
#define THREAD0_TID NO_PID
extern pid_t pid_max;
#define SESS_LEADER(p) ((p)->p_session->s_leader == (p))
/* Lock and unlock a process. */
#define PROC_LOCK(p) mtx_lock(&(p)->p_mtx)
#define PROC_TRYLOCK(p) mtx_trylock(&(p)->p_mtx)
#define PROC_UNLOCK(p) mtx_unlock(&(p)->p_mtx)
#define PROC_LOCKED(p) mtx_owned(&(p)->p_mtx)
#define PROC_WAIT_UNLOCKED(p) mtx_wait_unlocked(&(p)->p_mtx)
#define PROC_LOCK_ASSERT(p, type) mtx_assert(&(p)->p_mtx, (type))
/* Lock and unlock a process group. */
#define PGRP_LOCK(pg) mtx_lock(&(pg)->pg_mtx)
#define PGRP_UNLOCK(pg) mtx_unlock(&(pg)->pg_mtx)
#define PGRP_LOCKED(pg) mtx_owned(&(pg)->pg_mtx)
#define PGRP_LOCK_ASSERT(pg, type) mtx_assert(&(pg)->pg_mtx, (type))
#define PGRP_LOCK_PGSIGNAL(pg) do { \
if ((pg) != NULL) \
PGRP_LOCK(pg); \
} while (0)
#define PGRP_UNLOCK_PGSIGNAL(pg) do { \
if ((pg) != NULL) \
PGRP_UNLOCK(pg); \
} while (0)
/* Lock and unlock a session. */
#define SESS_LOCK(s) mtx_lock(&(s)->s_mtx)
#define SESS_UNLOCK(s) mtx_unlock(&(s)->s_mtx)
#define SESS_LOCKED(s) mtx_owned(&(s)->s_mtx)
#define SESS_LOCK_ASSERT(s, type) mtx_assert(&(s)->s_mtx, (type))
/*
* A non-zero p_lock prevents the process from exiting; it will sleep in exit1()
* until the count reaches zero.
*
* PHOLD() asserts that the process (except the current process) is
* not exiting and increments p_lock.
* _PHOLD() is same as PHOLD(), it takes the process locked.
*/
#define PHOLD(p) do { \
PROC_LOCK(p); \
_PHOLD(p); \
PROC_UNLOCK(p); \
} while (0)
#define _PHOLD(p) do { \
PROC_LOCK_ASSERT((p), MA_OWNED); \
KASSERT(!((p)->p_flag & P_WEXIT) || (p) == curproc, \
("PHOLD of exiting process %p", p)); \
(p)->p_lock++; \
} while (0)
#define PROC_ASSERT_HELD(p) do { \
KASSERT((p)->p_lock > 0, ("process %p not held", p)); \
} while (0)
#define PRELE(p) do { \
PROC_LOCK((p)); \
_PRELE((p)); \
PROC_UNLOCK((p)); \
} while (0)
#define _PRELE(p) do { \
PROC_LOCK_ASSERT((p), MA_OWNED); \
PROC_ASSERT_HELD(p); \
(--(p)->p_lock); \
if (((p)->p_flag & P_WEXIT) && (p)->p_lock == 0) \
wakeup(&(p)->p_lock); \
} while (0)
#define PROC_ASSERT_NOT_HELD(p) do { \
KASSERT((p)->p_lock == 0, ("process %p held", p)); \
} while (0)
#define PROC_UPDATE_COW(p) do { \
struct proc *_p = (p); \
PROC_LOCK_ASSERT((_p), MA_OWNED); \
atomic_store_int(&_p->p_cowgen, _p->p_cowgen + 1); \
} while (0)
#define PROC_COW_CHANGECOUNT(td, p) ({ \
struct thread *_td = (td); \
struct proc *_p = (p); \
MPASS(_td == curthread); \
PROC_LOCK_ASSERT(_p, MA_OWNED); \
_p->p_cowgen - _td->td_cowgen; \
})
/* Control whether or not it is safe for curthread to sleep. */
#define THREAD_NO_SLEEPING() do { \
curthread->td_no_sleeping++; \
MPASS(curthread->td_no_sleeping > 0); \
} while (0)
#define THREAD_SLEEPING_OK() do { \
MPASS(curthread->td_no_sleeping > 0); \
curthread->td_no_sleeping--; \
} while (0)
#define THREAD_CAN_SLEEP() ((curthread)->td_no_sleeping == 0)
#define THREAD_CONTENDS_ON_LOCK(lo) do { \
MPASS(curthread->td_wantedlock == NULL); \
curthread->td_wantedlock = lo; \
} while (0)
#define THREAD_CONTENTION_DONE(lo) do { \
MPASS(curthread->td_wantedlock == lo); \
curthread->td_wantedlock = NULL; \
} while (0)
#define PIDHASH(pid) (&pidhashtbl[(pid) & pidhash])
#define PIDHASHLOCK(pid) (&pidhashtbl_lock[((pid) & pidhashlock)])
extern LIST_HEAD(pidhashhead, proc) *pidhashtbl;
extern struct sx *pidhashtbl_lock;
extern u_long pidhash;
extern u_long pidhashlock;
#define PGRPHASH(pgid) (&pgrphashtbl[(pgid) & pgrphash])
extern LIST_HEAD(pgrphashhead, pgrp) *pgrphashtbl;
extern u_long pgrphash;
extern struct sx allproc_lock;
extern int allproc_gen;
extern struct sx proctree_lock;
extern struct mtx ppeers_lock;
extern struct mtx procid_lock;
extern struct proc proc0; /* Process slot for swapper. */
extern struct thread0_storage thread0_st; /* Primary thread in proc0. */
#define thread0 (thread0_st.t0st_thread)
extern struct vmspace vmspace0; /* VM space for proc0. */
extern int hogticks; /* Limit on kernel cpu hogs. */
extern int lastpid;
extern int nprocs, maxproc; /* Current and max number of procs. */
extern int maxprocperuid; /* Max procs per uid. */
extern u_long ps_arg_cache_limit;
LIST_HEAD(proclist, proc);
TAILQ_HEAD(procqueue, proc);
TAILQ_HEAD(threadqueue, thread);
extern struct proclist allproc; /* List of all processes. */
extern struct proc *initproc, *pageproc; /* Process slots for init, pager. */
extern struct uma_zone *proc_zone;
extern struct uma_zone *pgrp_zone;
struct proc *pfind(pid_t); /* Find process by id. */
struct proc *pfind_any(pid_t); /* Find (zombie) process by id. */
struct proc *pfind_any_locked(pid_t pid); /* Find process by id, locked. */
struct pgrp *pgfind(pid_t); /* Find process group by id. */
void pidhash_slockall(void); /* Shared lock all pid hash lists. */
void pidhash_sunlockall(void); /* Shared unlock all pid hash lists. */
struct fork_req {
int fr_flags;
int fr_pages;
int *fr_pidp;
struct proc **fr_procp;
int *fr_pd_fd;
int fr_pd_flags;
struct filecaps *fr_pd_fcaps;
int fr_flags2;
#define FR2_DROPSIG_CAUGHT 0x00000001 /* Drop caught non-DFL signals */
#define FR2_SHARE_PATHS 0x00000002 /* Invert sense of RFFDG for paths */
#define FR2_KPROC 0x00000004 /* Create a kernel process */
};
/*
* pget() flags.
*/
#define PGET_HOLD 0x00001 /* Hold the process. */
#define PGET_CANSEE 0x00002 /* Check against p_cansee(). */
#define PGET_CANDEBUG 0x00004 /* Check against p_candebug(). */
#define PGET_ISCURRENT 0x00008 /* Check that the found process is current. */
#define PGET_NOTWEXIT 0x00010 /* Check that the process is not in P_WEXIT. */
#define PGET_NOTINEXEC 0x00020 /* Check that the process is not in P_INEXEC. */
#define PGET_NOTID 0x00040 /* Do not assume tid if pid > PID_MAX. */
#define PGET_WANTREAD (PGET_HOLD | PGET_CANDEBUG | PGET_NOTWEXIT)
int pget(pid_t pid, int flags, struct proc **pp);
/* ast_register() flags */
#define ASTR_ASTF_REQUIRED 0x0001 /* td_ast TDAI(TDA_X) flag set is
required for call */
#define ASTR_TDP 0x0002 /* td_pflags flag set is required */
#define ASTR_KCLEAR 0x0004 /* call me on ast_kclear() */
#define ASTR_UNCOND 0x0008 /* call me always */
void ast(struct trapframe *framep);
void ast_kclear(struct thread *td);
void ast_register(int ast, int ast_flags, int tdp,
void (*f)(struct thread *td, int asts));
void ast_deregister(int tda);
void ast_sched_locked(struct thread *td, int tda);
void ast_sched_mask(struct thread *td, int ast);
void ast_sched(struct thread *td, int tda);
void ast_unsched_locked(struct thread *td, int tda);
struct thread *choosethread(void);
int cr_bsd_visible(struct ucred *u1, struct ucred *u2);
int cr_cansee(struct ucred *u1, struct ucred *u2);
int cr_canseesocket(struct ucred *cred, struct socket *so);
int cr_cansignal(struct ucred *cred, struct proc *proc, int signum);
int enterpgrp(struct proc *p, pid_t pgid, struct pgrp *pgrp,
struct session *sess);
int enterthispgrp(struct proc *p, struct pgrp *pgrp);
int fork1(struct thread *, struct fork_req *);
void fork_exit(void (*)(void *, struct trapframe *), void *,
struct trapframe *);
void fork_return(struct thread *, struct trapframe *);
int inferior(struct proc *p);
void itimer_proc_continue(struct proc *p);
void kqtimer_proc_continue(struct proc *p);
void kern_proc_vmmap_resident(struct vm_map *map, struct vm_map_entry *entry,
int *resident_count, bool *super);
void kern_yield(int);
void killjobc(void);
int leavepgrp(struct proc *p);
int maybe_preempt(struct thread *td);
void maybe_yield(void);
void mi_switch(int flags);
int p_candebug(struct thread *td, struct proc *p);
int p_cansee(struct thread *td, struct proc *p);
int p_cansched(struct thread *td, struct proc *p);
int p_cansignal(struct thread *td, struct proc *p, int signum);
int p_canwait(struct thread *td, struct proc *p);
struct pargs *pargs_alloc(int len);
void pargs_drop(struct pargs *pa);
void pargs_hold(struct pargs *pa);
+int pgrp_calc_jobc(struct pgrp *pgrp);
void proc_add_orphan(struct proc *child, struct proc *parent);
int proc_get_binpath(struct proc *p, char *binname, char **fullpath,
char **freepath);
int proc_getargv(struct thread *td, struct proc *p, struct sbuf *sb);
int proc_getauxv(struct thread *td, struct proc *p, struct sbuf *sb);
int proc_getenvv(struct thread *td, struct proc *p, struct sbuf *sb);
void procinit(void);
int proc_iterate(int (*cb)(struct proc *, void *), void *cbarg);
void proc_linkup0(struct proc *p, struct thread *td);
void proc_linkup(struct proc *p, struct thread *td);
struct proc *proc_realparent(struct proc *child);
void proc_reap(struct thread *td, struct proc *p, int *status, int options);
void proc_reparent(struct proc *child, struct proc *newparent, bool set_oppid);
void proc_set_p2_wexit(struct proc *p);
void proc_set_traced(struct proc *p, bool stop);
void proc_wkilled(struct proc *p);
struct pstats *pstats_alloc(void);
void pstats_fork(struct pstats *src, struct pstats *dst);
void pstats_free(struct pstats *ps);
void proc_clear_orphan(struct proc *p);
void reaper_abandon_children(struct proc *p, bool exiting);
int securelevel_ge(struct ucred *cr, int level);
int securelevel_gt(struct ucred *cr, int level);
void sess_hold(struct session *);
void sess_release(struct session *);
void setrunnable(struct thread *, int);
void setsugid(struct proc *p);
bool should_yield(void);
int sigonstack(size_t sp);
void stopevent(struct proc *, u_int, u_int);
struct thread *tdfind(lwpid_t, pid_t);
void threadinit(void);
void tidhash_add(struct thread *);
void tidhash_remove(struct thread *);
void cpu_idle(int);
int cpu_idle_wakeup(int);
extern void (*cpu_idle_hook)(sbintime_t); /* Hook to machdep CPU idler. */
void cpu_switch(struct thread *, struct thread *, struct mtx *);
void cpu_sync_core(void);
void cpu_throw(struct thread *, struct thread *) __dead2;
bool curproc_sigkilled(void);
void userret(struct thread *, struct trapframe *);
void cpu_exit(struct thread *);
void exit1(struct thread *, int, int) __dead2;
void cpu_copy_thread(struct thread *td, struct thread *td0);
bool cpu_exec_vmspace_reuse(struct proc *p, struct vm_map *map);
int cpu_fetch_syscall_args(struct thread *td);
void cpu_fork(struct thread *, struct proc *, struct thread *, int);
void cpu_fork_kthread_handler(struct thread *, void (*)(void *), void *);
int cpu_procctl(struct thread *td, int idtype, id_t id, int com,
void *data);
void cpu_set_syscall_retval(struct thread *, int);
int cpu_set_upcall(struct thread *, void (*)(void *), void *,
stack_t *);
int cpu_set_user_tls(struct thread *, void *tls_base);
void cpu_thread_alloc(struct thread *);
void cpu_thread_clean(struct thread *);
void cpu_thread_exit(struct thread *);
void cpu_thread_free(struct thread *);
struct thread *thread_alloc(int pages);
int thread_check_susp(struct thread *td, bool sleep);
void thread_cow_get_proc(struct thread *newtd, struct proc *p);
void thread_cow_get(struct thread *newtd, struct thread *td);
void thread_cow_free(struct thread *td);
void thread_cow_update(struct thread *td);
void thread_cow_synced(struct thread *td);
int thread_create(struct thread *td, struct rtprio *rtp,
int (*initialize_thread)(struct thread *, void *), void *thunk);
void thread_exit(void) __dead2;
void thread_free(struct thread *td);
void thread_link(struct thread *td, struct proc *p);
void thread_reap_barrier(void);
int thread_recycle(struct thread *, int pages);
int thread_single(struct proc *p, int how);
void thread_single_end(struct proc *p, int how);
void thread_stash(struct thread *td);
void thread_stopped(struct proc *p);
void childproc_stopped(struct proc *child, int reason);
void childproc_continued(struct proc *child);
void childproc_exited(struct proc *child);
void thread_run_flash(struct thread *td);
int thread_suspend_check(int how);
bool thread_suspend_check_needed(void);
void thread_suspend_switch(struct thread *, struct proc *p);
void thread_suspend_one(struct thread *td);
void thread_unlink(struct thread *td);
void thread_unsuspend(struct proc *p);
void thread_wait(struct proc *p);
bool stop_all_proc_block(void);
void stop_all_proc_unblock(void);
void stop_all_proc(void);
void resume_all_proc(void);
static __inline int
curthread_pflags_set(int flags)
{
struct thread *td;
int save;
td = curthread;
save = ~flags | (td->td_pflags & flags);
td->td_pflags |= flags;
return (save);
}
static __inline void
curthread_pflags_restore(int save)
{
curthread->td_pflags &= save;
}
static __inline int
curthread_pflags2_set(int flags)
{
struct thread *td;
int save;
td = curthread;
save = ~flags | (td->td_pflags2 & flags);
td->td_pflags2 |= flags;
return (save);
}
static __inline void
curthread_pflags2_restore(int save)
{
curthread->td_pflags2 &= save;
}
static __inline __pure2 struct td_sched *
td_get_sched(struct thread *td)
{
return ((struct td_sched *)&td[1]);
}
#define PROC_ID_PID 0
#define PROC_ID_GROUP 1
#define PROC_ID_SESSION 2
#define PROC_ID_REAP 3
void proc_id_set(int type, pid_t id);
void proc_id_set_cond(int type, pid_t id);
void proc_id_clear(int type, pid_t id);
EVENTHANDLER_LIST_DECLARE(process_ctor);
EVENTHANDLER_LIST_DECLARE(process_dtor);
EVENTHANDLER_LIST_DECLARE(process_init);
EVENTHANDLER_LIST_DECLARE(process_fini);
EVENTHANDLER_LIST_DECLARE(process_exit);
EVENTHANDLER_LIST_DECLARE(process_fork);
EVENTHANDLER_LIST_DECLARE(process_exec);
EVENTHANDLER_LIST_DECLARE(thread_ctor);
EVENTHANDLER_LIST_DECLARE(thread_dtor);
EVENTHANDLER_LIST_DECLARE(thread_init);
#endif /* _KERNEL */
#endif /* !_SYS_PROC_H_ */