kern_proc.c revision 1.171
1 /* $NetBSD: kern_proc.c,v 1.171 2011/01/28 20:31:10 pooka Exp $ */ 2 3 /*- 4 * Copyright (c) 1999, 2006, 2007, 2008 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility, 9 * NASA Ames Research Center, and by Andrew Doran. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 21 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 22 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 23 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 30 * POSSIBILITY OF SUCH DAMAGE. 31 */ 32 33 /* 34 * Copyright (c) 1982, 1986, 1989, 1991, 1993 35 * The Regents of the University of California. All rights reserved. 36 * 37 * Redistribution and use in source and binary forms, with or without 38 * modification, are permitted provided that the following conditions 39 * are met: 40 * 1. Redistributions of source code must retain the above copyright 41 * notice, this list of conditions and the following disclaimer. 42 * 2. Redistributions in binary form must reproduce the above copyright 43 * notice, this list of conditions and the following disclaimer in the 44 * documentation and/or other materials provided with the distribution. 45 * 3. Neither the name of the University nor the names of its contributors 46 * may be used to endorse or promote products derived from this software 47 * without specific prior written permission. 48 * 49 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 50 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 51 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 52 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 53 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 54 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 55 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 56 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 57 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 58 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 59 * SUCH DAMAGE. 60 * 61 * @(#)kern_proc.c 8.7 (Berkeley) 2/14/95 62 */ 63 64 #include <sys/cdefs.h> 65 __KERNEL_RCSID(0, "$NetBSD: kern_proc.c,v 1.171 2011/01/28 20:31:10 pooka Exp $"); 66 67 #ifdef _KERNEL_OPT 68 #include "opt_kstack.h" 69 #include "opt_maxuprc.h" 70 #include "opt_dtrace.h" 71 #include "opt_compat_netbsd32.h" 72 #endif 73 74 #include <sys/param.h> 75 #include <sys/systm.h> 76 #include <sys/kernel.h> 77 #include <sys/proc.h> 78 #include <sys/resourcevar.h> 79 #include <sys/buf.h> 80 #include <sys/acct.h> 81 #include <sys/wait.h> 82 #include <sys/file.h> 83 #include <ufs/ufs/quota.h> 84 #include <sys/uio.h> 85 #include <sys/pool.h> 86 #include <sys/pset.h> 87 #include <sys/mbuf.h> 88 #include <sys/ioctl.h> 89 #include <sys/tty.h> 90 #include <sys/signalvar.h> 91 #include <sys/ras.h> 92 #include <sys/sa.h> 93 #include <sys/savar.h> 94 #include <sys/filedesc.h> 95 #include "sys/syscall_stats.h" 96 #include <sys/kauth.h> 97 #include <sys/sleepq.h> 98 #include <sys/atomic.h> 99 #include <sys/kmem.h> 100 #include <sys/dtrace_bsd.h> 101 #include <sys/sysctl.h> 102 #include <sys/exec.h> 103 #include <sys/cpu.h> 104 105 #include <uvm/uvm_extern.h> 106 #include <uvm/uvm_extern.h> 107 108 #ifdef COMPAT_NETBSD32 109 #include <compat/netbsd32/netbsd32.h> 110 #endif 111 112 /* 113 * Other process lists 114 */ 115 116 struct proclist allproc; 117 struct proclist zombproc; /* resources have been freed */ 118 119 kmutex_t *proc_lock; 120 121 /* 122 * pid to proc lookup is done by indexing the pid_table array. 123 * Since pid numbers are only allocated when an empty slot 124 * has been found, there is no need to search any lists ever. 125 * (an orphaned pgrp will lock the slot, a session will lock 126 * the pgrp with the same number.) 127 * If the table is too small it is reallocated with twice the 128 * previous size and the entries 'unzipped' into the two halves. 129 * A linked list of free entries is passed through the pt_proc 130 * field of 'free' items - set odd to be an invalid ptr. 131 */ 132 133 struct pid_table { 134 struct proc *pt_proc; 135 struct pgrp *pt_pgrp; 136 pid_t pt_pid; 137 }; 138 #if 1 /* strongly typed cast - should be a noop */ 139 static inline uint p2u(struct proc *p) { return (uint)(uintptr_t)p; } 140 #else 141 #define p2u(p) ((uint)p) 142 #endif 143 #define P_VALID(p) (!(p2u(p) & 1)) 144 #define P_NEXT(p) (p2u(p) >> 1) 145 #define P_FREE(pid) ((struct proc *)(uintptr_t)((pid) << 1 | 1)) 146 147 #define INITIAL_PID_TABLE_SIZE (1 << 5) 148 static struct pid_table *pid_table; 149 static uint pid_tbl_mask = INITIAL_PID_TABLE_SIZE - 1; 150 static uint pid_alloc_lim; /* max we allocate before growing table */ 151 static uint pid_alloc_cnt; /* number of allocated pids */ 152 153 /* links through free slots - never empty! */ 154 static uint next_free_pt, last_free_pt; 155 static pid_t pid_max = PID_MAX; /* largest value we allocate */ 156 157 /* Components of the first process -- never freed. */ 158 159 extern struct emul emul_netbsd; /* defined in kern_exec.c */ 160 161 struct session session0 = { 162 .s_count = 1, 163 .s_sid = 0, 164 }; 165 struct pgrp pgrp0 = { 166 .pg_members = LIST_HEAD_INITIALIZER(&pgrp0.pg_members), 167 .pg_session = &session0, 168 }; 169 filedesc_t filedesc0; 170 struct cwdinfo cwdi0 = { 171 .cwdi_cmask = CMASK, /* see cmask below */ 172 .cwdi_refcnt = 1, 173 }; 174 struct plimit limit0; 175 struct pstats pstat0; 176 struct vmspace vmspace0; 177 struct sigacts sigacts0; 178 struct proc proc0 = { 179 .p_lwps = LIST_HEAD_INITIALIZER(&proc0.p_lwps), 180 .p_sigwaiters = LIST_HEAD_INITIALIZER(&proc0.p_sigwaiters), 181 .p_nlwps = 1, 182 .p_nrlwps = 1, 183 .p_nlwpid = 1, /* must match lwp0.l_lid */ 184 .p_pgrp = &pgrp0, 185 .p_comm = "system", 186 /* 187 * Set P_NOCLDWAIT so that kernel threads are reparented to init(8) 188 * when they exit. init(8) can easily wait them out for us. 189 */ 190 .p_flag = PK_SYSTEM | PK_NOCLDWAIT, 191 .p_stat = SACTIVE, 192 .p_nice = NZERO, 193 .p_emul = &emul_netbsd, 194 .p_cwdi = &cwdi0, 195 .p_limit = &limit0, 196 .p_fd = &filedesc0, 197 .p_vmspace = &vmspace0, 198 .p_stats = &pstat0, 199 .p_sigacts = &sigacts0, 200 }; 201 kauth_cred_t cred0; 202 203 int nofile = NOFILE; 204 int maxuprc = MAXUPRC; 205 int cmask = CMASK; 206 207 MALLOC_DEFINE(M_EMULDATA, "emuldata", "Per-process emulation data"); 208 MALLOC_DEFINE(M_SUBPROC, "subproc", "Proc sub-structures"); 209 210 static int sysctl_doeproc(SYSCTLFN_PROTO); 211 static int sysctl_kern_proc_args(SYSCTLFN_PROTO); 212 static void fill_kproc2(struct proc *, struct kinfo_proc2 *, bool); 213 214 /* 215 * The process list descriptors, used during pid allocation and 216 * by sysctl. No locking on this data structure is needed since 217 * it is completely static. 218 */ 219 const struct proclist_desc proclists[] = { 220 { &allproc }, 221 { &zombproc }, 222 { NULL }, 223 }; 224 225 static struct pgrp * pg_remove(pid_t); 226 static void pg_delete(pid_t); 227 static void orphanpg(struct pgrp *); 228 229 static specificdata_domain_t proc_specificdata_domain; 230 231 static pool_cache_t proc_cache; 232 233 static kauth_listener_t proc_listener; 234 235 static int 236 proc_listener_cb(kauth_cred_t cred, kauth_action_t action, void *cookie, 237 void *arg0, void *arg1, void *arg2, void *arg3) 238 { 239 struct proc *p; 240 int result; 241 242 result = KAUTH_RESULT_DEFER; 243 p = arg0; 244 245 switch (action) { 246 case KAUTH_PROCESS_CANSEE: { 247 enum kauth_process_req req; 248 249 req = (enum kauth_process_req)arg1; 250 251 switch (req) { 252 case KAUTH_REQ_PROCESS_CANSEE_ARGS: 253 case KAUTH_REQ_PROCESS_CANSEE_ENTRY: 254 case KAUTH_REQ_PROCESS_CANSEE_OPENFILES: 255 result = KAUTH_RESULT_ALLOW; 256 257 break; 258 259 case KAUTH_REQ_PROCESS_CANSEE_ENV: 260 if (kauth_cred_getuid(cred) != 261 kauth_cred_getuid(p->p_cred) || 262 kauth_cred_getuid(cred) != 263 kauth_cred_getsvuid(p->p_cred)) 264 break; 265 266 result = KAUTH_RESULT_ALLOW; 267 268 break; 269 270 default: 271 break; 272 } 273 274 break; 275 } 276 277 case KAUTH_PROCESS_FORK: { 278 int lnprocs = (int)(unsigned long)arg2; 279 280 /* 281 * Don't allow a nonprivileged user to use the last few 282 * processes. The variable lnprocs is the current number of 283 * processes, maxproc is the limit. 284 */ 285 if (__predict_false((lnprocs >= maxproc - 5))) 286 break; 287 288 result = KAUTH_RESULT_ALLOW; 289 290 break; 291 } 292 293 case KAUTH_PROCESS_CORENAME: 294 case KAUTH_PROCESS_STOPFLAG: 295 if (proc_uidmatch(cred, p->p_cred) == 0) 296 result = KAUTH_RESULT_ALLOW; 297 298 break; 299 300 default: 301 break; 302 } 303 304 return result; 305 } 306 307 /* 308 * Initialize global process hashing structures. 309 */ 310 void 311 procinit(void) 312 { 313 const struct proclist_desc *pd; 314 u_int i; 315 #define LINK_EMPTY ((PID_MAX + INITIAL_PID_TABLE_SIZE) & ~(INITIAL_PID_TABLE_SIZE - 1)) 316 317 for (pd = proclists; pd->pd_list != NULL; pd++) 318 LIST_INIT(pd->pd_list); 319 320 proc_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE); 321 pid_table = kmem_alloc(INITIAL_PID_TABLE_SIZE 322 * sizeof(struct pid_table), KM_SLEEP); 323 324 /* Set free list running through table... 325 Preset 'use count' above PID_MAX so we allocate pid 1 next. */ 326 for (i = 0; i <= pid_tbl_mask; i++) { 327 pid_table[i].pt_proc = P_FREE(LINK_EMPTY + i + 1); 328 pid_table[i].pt_pgrp = 0; 329 pid_table[i].pt_pid = 0; 330 } 331 /* slot 0 is just grabbed */ 332 next_free_pt = 1; 333 /* Need to fix last entry. */ 334 last_free_pt = pid_tbl_mask; 335 pid_table[last_free_pt].pt_proc = P_FREE(LINK_EMPTY); 336 /* point at which we grow table - to avoid reusing pids too often */ 337 pid_alloc_lim = pid_tbl_mask - 1; 338 #undef LINK_EMPTY 339 340 proc_specificdata_domain = specificdata_domain_create(); 341 KASSERT(proc_specificdata_domain != NULL); 342 343 proc_cache = pool_cache_init(sizeof(struct proc), 0, 0, 0, 344 "procpl", NULL, IPL_NONE, NULL, NULL, NULL); 345 346 proc_listener = kauth_listen_scope(KAUTH_SCOPE_PROCESS, 347 proc_listener_cb, NULL); 348 } 349 350 void 351 procinit_sysctl(void) 352 { 353 static struct sysctllog *clog; 354 355 sysctl_createv(&clog, 0, NULL, NULL, 356 CTLFLAG_PERMANENT, 357 CTLTYPE_NODE, "kern", NULL, 358 NULL, 0, NULL, 0, 359 CTL_KERN, CTL_EOL); 360 361 sysctl_createv(&clog, 0, NULL, NULL, 362 CTLFLAG_PERMANENT, 363 CTLTYPE_NODE, "proc", 364 SYSCTL_DESCR("System-wide process information"), 365 sysctl_doeproc, 0, NULL, 0, 366 CTL_KERN, KERN_PROC, CTL_EOL); 367 sysctl_createv(&clog, 0, NULL, NULL, 368 CTLFLAG_PERMANENT, 369 CTLTYPE_NODE, "proc2", 370 SYSCTL_DESCR("Machine-independent process information"), 371 sysctl_doeproc, 0, NULL, 0, 372 CTL_KERN, KERN_PROC2, CTL_EOL); 373 sysctl_createv(&clog, 0, NULL, NULL, 374 CTLFLAG_PERMANENT, 375 CTLTYPE_NODE, "proc_args", 376 SYSCTL_DESCR("Process argument information"), 377 sysctl_kern_proc_args, 0, NULL, 0, 378 CTL_KERN, KERN_PROC_ARGS, CTL_EOL); 379 380 /* 381 "nodes" under these: 382 383 KERN_PROC_ALL 384 KERN_PROC_PID pid 385 KERN_PROC_PGRP pgrp 386 KERN_PROC_SESSION sess 387 KERN_PROC_TTY tty 388 KERN_PROC_UID uid 389 KERN_PROC_RUID uid 390 KERN_PROC_GID gid 391 KERN_PROC_RGID gid 392 393 all in all, probably not worth the effort... 394 */ 395 } 396 397 /* 398 * Initialize process 0. 399 */ 400 void 401 proc0_init(void) 402 { 403 struct proc *p; 404 struct pgrp *pg; 405 rlim_t lim; 406 int i; 407 408 p = &proc0; 409 pg = &pgrp0; 410 411 mutex_init(&p->p_stmutex, MUTEX_DEFAULT, IPL_HIGH); 412 mutex_init(&p->p_auxlock, MUTEX_DEFAULT, IPL_NONE); 413 p->p_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE); 414 415 rw_init(&p->p_reflock); 416 cv_init(&p->p_waitcv, "wait"); 417 cv_init(&p->p_lwpcv, "lwpwait"); 418 419 LIST_INSERT_HEAD(&p->p_lwps, &lwp0, l_sibling); 420 421 pid_table[0].pt_proc = p; 422 LIST_INSERT_HEAD(&allproc, p, p_list); 423 424 pid_table[0].pt_pgrp = pg; 425 LIST_INSERT_HEAD(&pg->pg_members, p, p_pglist); 426 427 #ifdef __HAVE_SYSCALL_INTERN 428 (*p->p_emul->e_syscall_intern)(p); 429 #endif 430 431 /* Create credentials. */ 432 cred0 = kauth_cred_alloc(); 433 p->p_cred = cred0; 434 435 /* Create the CWD info. */ 436 rw_init(&cwdi0.cwdi_lock); 437 438 /* Create the limits structures. */ 439 mutex_init(&limit0.pl_lock, MUTEX_DEFAULT, IPL_NONE); 440 for (i = 0; i < __arraycount(limit0.pl_rlimit); i++) 441 limit0.pl_rlimit[i].rlim_cur = 442 limit0.pl_rlimit[i].rlim_max = RLIM_INFINITY; 443 444 limit0.pl_rlimit[RLIMIT_NOFILE].rlim_max = maxfiles; 445 limit0.pl_rlimit[RLIMIT_NOFILE].rlim_cur = 446 maxfiles < nofile ? maxfiles : nofile; 447 448 limit0.pl_rlimit[RLIMIT_NPROC].rlim_max = maxproc; 449 limit0.pl_rlimit[RLIMIT_NPROC].rlim_cur = 450 maxproc < maxuprc ? maxproc : maxuprc; 451 452 lim = MIN(VM_MAXUSER_ADDRESS, ctob((rlim_t)uvmexp.free)); 453 limit0.pl_rlimit[RLIMIT_RSS].rlim_max = lim; 454 limit0.pl_rlimit[RLIMIT_MEMLOCK].rlim_max = lim; 455 limit0.pl_rlimit[RLIMIT_MEMLOCK].rlim_cur = lim / 3; 456 limit0.pl_corename = defcorename; 457 limit0.pl_refcnt = 1; 458 limit0.pl_sv_limit = NULL; 459 460 /* Configure virtual memory system, set vm rlimits. */ 461 uvm_init_limits(p); 462 463 /* Initialize file descriptor table for proc0. */ 464 fd_init(&filedesc0); 465 466 /* 467 * Initialize proc0's vmspace, which uses the kernel pmap. 468 * All kernel processes (which never have user space mappings) 469 * share proc0's vmspace, and thus, the kernel pmap. 470 */ 471 uvmspace_init(&vmspace0, pmap_kernel(), round_page(VM_MIN_ADDRESS), 472 trunc_page(VM_MAX_ADDRESS)); 473 474 /* Initialize signal state for proc0. XXX IPL_SCHED */ 475 mutex_init(&p->p_sigacts->sa_mutex, MUTEX_DEFAULT, IPL_SCHED); 476 siginit(p); 477 478 proc_initspecific(p); 479 kdtrace_proc_ctor(NULL, p); 480 } 481 482 /* 483 * Session reference counting. 484 */ 485 486 void 487 proc_sesshold(struct session *ss) 488 { 489 490 KASSERT(mutex_owned(proc_lock)); 491 ss->s_count++; 492 } 493 494 void 495 proc_sessrele(struct session *ss) 496 { 497 498 KASSERT(mutex_owned(proc_lock)); 499 /* 500 * We keep the pgrp with the same id as the session in order to 501 * stop a process being given the same pid. Since the pgrp holds 502 * a reference to the session, it must be a 'zombie' pgrp by now. 503 */ 504 if (--ss->s_count == 0) { 505 struct pgrp *pg; 506 507 pg = pg_remove(ss->s_sid); 508 mutex_exit(proc_lock); 509 510 kmem_free(pg, sizeof(struct pgrp)); 511 kmem_free(ss, sizeof(struct session)); 512 } else { 513 mutex_exit(proc_lock); 514 } 515 } 516 517 /* 518 * Check that the specified process group is in the session of the 519 * specified process. 520 * Treats -ve ids as process ids. 521 * Used to validate TIOCSPGRP requests. 522 */ 523 int 524 pgid_in_session(struct proc *p, pid_t pg_id) 525 { 526 struct pgrp *pgrp; 527 struct session *session; 528 int error; 529 530 mutex_enter(proc_lock); 531 if (pg_id < 0) { 532 struct proc *p1 = proc_find(-pg_id); 533 if (p1 == NULL) { 534 error = EINVAL; 535 goto fail; 536 } 537 pgrp = p1->p_pgrp; 538 } else { 539 pgrp = pgrp_find(pg_id); 540 if (pgrp == NULL) { 541 error = EINVAL; 542 goto fail; 543 } 544 } 545 session = pgrp->pg_session; 546 error = (session != p->p_pgrp->pg_session) ? EPERM : 0; 547 fail: 548 mutex_exit(proc_lock); 549 return error; 550 } 551 552 /* 553 * p_inferior: is p an inferior of q? 554 */ 555 static inline bool 556 p_inferior(struct proc *p, struct proc *q) 557 { 558 559 KASSERT(mutex_owned(proc_lock)); 560 561 for (; p != q; p = p->p_pptr) 562 if (p->p_pid == 0) 563 return false; 564 return true; 565 } 566 567 /* 568 * proc_find: locate a process by the ID. 569 * 570 * => Must be called with proc_lock held. 571 */ 572 proc_t * 573 proc_find_raw(pid_t pid) 574 { 575 struct pid_table *pt; 576 proc_t *p; 577 578 KASSERT(mutex_owned(proc_lock)); 579 pt = &pid_table[pid & pid_tbl_mask]; 580 p = pt->pt_proc; 581 if (__predict_false(!P_VALID(p) || pt->pt_pid != pid)) { 582 return NULL; 583 } 584 return p; 585 } 586 587 proc_t * 588 proc_find(pid_t pid) 589 { 590 proc_t *p; 591 592 p = proc_find_raw(pid); 593 if (__predict_false(p == NULL)) { 594 return NULL; 595 } 596 597 /* 598 * Only allow live processes to be found by PID. 599 * XXX: p_stat might change, since unlocked. 600 */ 601 if (__predict_true(p->p_stat == SACTIVE || p->p_stat == SSTOP)) { 602 return p; 603 } 604 return NULL; 605 } 606 607 /* 608 * pgrp_find: locate a process group by the ID. 609 * 610 * => Must be called with proc_lock held. 611 */ 612 struct pgrp * 613 pgrp_find(pid_t pgid) 614 { 615 struct pgrp *pg; 616 617 KASSERT(mutex_owned(proc_lock)); 618 619 pg = pid_table[pgid & pid_tbl_mask].pt_pgrp; 620 621 /* 622 * Cannot look up a process group that only exists because the 623 * session has not died yet (traditional). 624 */ 625 if (pg == NULL || pg->pg_id != pgid || LIST_EMPTY(&pg->pg_members)) { 626 return NULL; 627 } 628 return pg; 629 } 630 631 static void 632 expand_pid_table(void) 633 { 634 size_t pt_size, tsz; 635 struct pid_table *n_pt, *new_pt; 636 struct proc *proc; 637 struct pgrp *pgrp; 638 pid_t pid, rpid; 639 u_int i; 640 uint new_pt_mask; 641 642 pt_size = pid_tbl_mask + 1; 643 tsz = pt_size * 2 * sizeof(struct pid_table); 644 new_pt = kmem_alloc(tsz, KM_SLEEP); 645 new_pt_mask = pt_size * 2 - 1; 646 647 mutex_enter(proc_lock); 648 if (pt_size != pid_tbl_mask + 1) { 649 /* Another process beat us to it... */ 650 mutex_exit(proc_lock); 651 kmem_free(new_pt, tsz); 652 return; 653 } 654 655 /* 656 * Copy entries from old table into new one. 657 * If 'pid' is 'odd' we need to place in the upper half, 658 * even pid's to the lower half. 659 * Free items stay in the low half so we don't have to 660 * fixup the reference to them. 661 * We stuff free items on the front of the freelist 662 * because we can't write to unmodified entries. 663 * Processing the table backwards maintains a semblance 664 * of issuing pid numbers that increase with time. 665 */ 666 i = pt_size - 1; 667 n_pt = new_pt + i; 668 for (; ; i--, n_pt--) { 669 proc = pid_table[i].pt_proc; 670 pgrp = pid_table[i].pt_pgrp; 671 if (!P_VALID(proc)) { 672 /* Up 'use count' so that link is valid */ 673 pid = (P_NEXT(proc) + pt_size) & ~pt_size; 674 rpid = 0; 675 proc = P_FREE(pid); 676 if (pgrp) 677 pid = pgrp->pg_id; 678 } else { 679 pid = pid_table[i].pt_pid; 680 rpid = pid; 681 } 682 683 /* Save entry in appropriate half of table */ 684 n_pt[pid & pt_size].pt_proc = proc; 685 n_pt[pid & pt_size].pt_pgrp = pgrp; 686 n_pt[pid & pt_size].pt_pid = rpid; 687 688 /* Put other piece on start of free list */ 689 pid = (pid ^ pt_size) & ~pid_tbl_mask; 690 n_pt[pid & pt_size].pt_proc = 691 P_FREE((pid & ~pt_size) | next_free_pt); 692 n_pt[pid & pt_size].pt_pgrp = 0; 693 n_pt[pid & pt_size].pt_pid = 0; 694 695 next_free_pt = i | (pid & pt_size); 696 if (i == 0) 697 break; 698 } 699 700 /* Save old table size and switch tables */ 701 tsz = pt_size * sizeof(struct pid_table); 702 n_pt = pid_table; 703 pid_table = new_pt; 704 pid_tbl_mask = new_pt_mask; 705 706 /* 707 * pid_max starts as PID_MAX (= 30000), once we have 16384 708 * allocated pids we need it to be larger! 709 */ 710 if (pid_tbl_mask > PID_MAX) { 711 pid_max = pid_tbl_mask * 2 + 1; 712 pid_alloc_lim |= pid_alloc_lim << 1; 713 } else 714 pid_alloc_lim <<= 1; /* doubles number of free slots... */ 715 716 mutex_exit(proc_lock); 717 kmem_free(n_pt, tsz); 718 } 719 720 struct proc * 721 proc_alloc(void) 722 { 723 struct proc *p; 724 725 p = pool_cache_get(proc_cache, PR_WAITOK); 726 p->p_stat = SIDL; /* protect against others */ 727 proc_initspecific(p); 728 kdtrace_proc_ctor(NULL, p); 729 p->p_pid = -1; 730 proc_alloc_pid(p); 731 return p; 732 } 733 734 pid_t 735 proc_alloc_pid(struct proc *p) 736 { 737 struct pid_table *pt; 738 pid_t pid; 739 int nxt; 740 741 for (;;expand_pid_table()) { 742 if (__predict_false(pid_alloc_cnt >= pid_alloc_lim)) 743 /* ensure pids cycle through 2000+ values */ 744 continue; 745 mutex_enter(proc_lock); 746 pt = &pid_table[next_free_pt]; 747 #ifdef DIAGNOSTIC 748 if (__predict_false(P_VALID(pt->pt_proc) || pt->pt_pgrp)) 749 panic("proc_alloc: slot busy"); 750 #endif 751 nxt = P_NEXT(pt->pt_proc); 752 if (nxt & pid_tbl_mask) 753 break; 754 /* Table full - expand (NB last entry not used....) */ 755 mutex_exit(proc_lock); 756 } 757 758 /* pid is 'saved use count' + 'size' + entry */ 759 pid = (nxt & ~pid_tbl_mask) + pid_tbl_mask + 1 + next_free_pt; 760 if ((uint)pid > (uint)pid_max) 761 pid &= pid_tbl_mask; 762 next_free_pt = nxt & pid_tbl_mask; 763 764 /* Grab table slot */ 765 pt->pt_proc = p; 766 767 KASSERT(pt->pt_pid == 0); 768 pt->pt_pid = pid; 769 if (p->p_pid == -1) { 770 p->p_pid = pid; 771 } 772 pid_alloc_cnt++; 773 mutex_exit(proc_lock); 774 775 return pid; 776 } 777 778 /* 779 * Free a process id - called from proc_free (in kern_exit.c) 780 * 781 * Called with the proc_lock held. 782 */ 783 void 784 proc_free_pid(pid_t pid) 785 { 786 struct pid_table *pt; 787 788 KASSERT(mutex_owned(proc_lock)); 789 790 pt = &pid_table[pid & pid_tbl_mask]; 791 792 /* save pid use count in slot */ 793 pt->pt_proc = P_FREE(pid & ~pid_tbl_mask); 794 KASSERT(pt->pt_pid == pid); 795 pt->pt_pid = 0; 796 797 if (pt->pt_pgrp == NULL) { 798 /* link last freed entry onto ours */ 799 pid &= pid_tbl_mask; 800 pt = &pid_table[last_free_pt]; 801 pt->pt_proc = P_FREE(P_NEXT(pt->pt_proc) | pid); 802 pt->pt_pid = 0; 803 last_free_pt = pid; 804 pid_alloc_cnt--; 805 } 806 807 atomic_dec_uint(&nprocs); 808 } 809 810 void 811 proc_free_mem(struct proc *p) 812 { 813 814 kdtrace_proc_dtor(NULL, p); 815 pool_cache_put(proc_cache, p); 816 } 817 818 /* 819 * proc_enterpgrp: move p to a new or existing process group (and session). 820 * 821 * If we are creating a new pgrp, the pgid should equal 822 * the calling process' pid. 823 * If is only valid to enter a process group that is in the session 824 * of the process. 825 * Also mksess should only be set if we are creating a process group 826 * 827 * Only called from sys_setsid and sys_setpgid. 828 */ 829 int 830 proc_enterpgrp(struct proc *curp, pid_t pid, pid_t pgid, bool mksess) 831 { 832 struct pgrp *new_pgrp, *pgrp; 833 struct session *sess; 834 struct proc *p; 835 int rval; 836 pid_t pg_id = NO_PGID; 837 838 sess = mksess ? kmem_alloc(sizeof(*sess), KM_SLEEP) : NULL; 839 840 /* Allocate data areas we might need before doing any validity checks */ 841 mutex_enter(proc_lock); /* Because pid_table might change */ 842 if (pid_table[pgid & pid_tbl_mask].pt_pgrp == 0) { 843 mutex_exit(proc_lock); 844 new_pgrp = kmem_alloc(sizeof(*new_pgrp), KM_SLEEP); 845 mutex_enter(proc_lock); 846 } else 847 new_pgrp = NULL; 848 rval = EPERM; /* most common error (to save typing) */ 849 850 /* Check pgrp exists or can be created */ 851 pgrp = pid_table[pgid & pid_tbl_mask].pt_pgrp; 852 if (pgrp != NULL && pgrp->pg_id != pgid) 853 goto done; 854 855 /* Can only set another process under restricted circumstances. */ 856 if (pid != curp->p_pid) { 857 /* Must exist and be one of our children... */ 858 p = proc_find(pid); 859 if (p == NULL || !p_inferior(p, curp)) { 860 rval = ESRCH; 861 goto done; 862 } 863 /* ... in the same session... */ 864 if (sess != NULL || p->p_session != curp->p_session) 865 goto done; 866 /* ... existing pgid must be in same session ... */ 867 if (pgrp != NULL && pgrp->pg_session != p->p_session) 868 goto done; 869 /* ... and not done an exec. */ 870 if (p->p_flag & PK_EXEC) { 871 rval = EACCES; 872 goto done; 873 } 874 } else { 875 /* ... setsid() cannot re-enter a pgrp */ 876 if (mksess && (curp->p_pgid == curp->p_pid || 877 pgrp_find(curp->p_pid))) 878 goto done; 879 p = curp; 880 } 881 882 /* Changing the process group/session of a session 883 leader is definitely off limits. */ 884 if (SESS_LEADER(p)) { 885 if (sess == NULL && p->p_pgrp == pgrp) 886 /* unless it's a definite noop */ 887 rval = 0; 888 goto done; 889 } 890 891 /* Can only create a process group with id of process */ 892 if (pgrp == NULL && pgid != pid) 893 goto done; 894 895 /* Can only create a session if creating pgrp */ 896 if (sess != NULL && pgrp != NULL) 897 goto done; 898 899 /* Check we allocated memory for a pgrp... */ 900 if (pgrp == NULL && new_pgrp == NULL) 901 goto done; 902 903 /* Don't attach to 'zombie' pgrp */ 904 if (pgrp != NULL && LIST_EMPTY(&pgrp->pg_members)) 905 goto done; 906 907 /* Expect to succeed now */ 908 rval = 0; 909 910 if (pgrp == p->p_pgrp) 911 /* nothing to do */ 912 goto done; 913 914 /* Ok all setup, link up required structures */ 915 916 if (pgrp == NULL) { 917 pgrp = new_pgrp; 918 new_pgrp = NULL; 919 if (sess != NULL) { 920 sess->s_sid = p->p_pid; 921 sess->s_leader = p; 922 sess->s_count = 1; 923 sess->s_ttyvp = NULL; 924 sess->s_ttyp = NULL; 925 sess->s_flags = p->p_session->s_flags & ~S_LOGIN_SET; 926 memcpy(sess->s_login, p->p_session->s_login, 927 sizeof(sess->s_login)); 928 p->p_lflag &= ~PL_CONTROLT; 929 } else { 930 sess = p->p_pgrp->pg_session; 931 proc_sesshold(sess); 932 } 933 pgrp->pg_session = sess; 934 sess = NULL; 935 936 pgrp->pg_id = pgid; 937 LIST_INIT(&pgrp->pg_members); 938 #ifdef DIAGNOSTIC 939 if (__predict_false(pid_table[pgid & pid_tbl_mask].pt_pgrp)) 940 panic("enterpgrp: pgrp table slot in use"); 941 if (__predict_false(mksess && p != curp)) 942 panic("enterpgrp: mksession and p != curproc"); 943 #endif 944 pid_table[pgid & pid_tbl_mask].pt_pgrp = pgrp; 945 pgrp->pg_jobc = 0; 946 } 947 948 /* 949 * Adjust eligibility of affected pgrps to participate in job control. 950 * Increment eligibility counts before decrementing, otherwise we 951 * could reach 0 spuriously during the first call. 952 */ 953 fixjobc(p, pgrp, 1); 954 fixjobc(p, p->p_pgrp, 0); 955 956 /* Interlock with ttread(). */ 957 mutex_spin_enter(&tty_lock); 958 959 /* Move process to requested group. */ 960 LIST_REMOVE(p, p_pglist); 961 if (LIST_EMPTY(&p->p_pgrp->pg_members)) 962 /* defer delete until we've dumped the lock */ 963 pg_id = p->p_pgrp->pg_id; 964 p->p_pgrp = pgrp; 965 LIST_INSERT_HEAD(&pgrp->pg_members, p, p_pglist); 966 967 /* Done with the swap; we can release the tty mutex. */ 968 mutex_spin_exit(&tty_lock); 969 970 done: 971 if (pg_id != NO_PGID) { 972 /* Releases proc_lock. */ 973 pg_delete(pg_id); 974 } else { 975 mutex_exit(proc_lock); 976 } 977 if (sess != NULL) 978 kmem_free(sess, sizeof(*sess)); 979 if (new_pgrp != NULL) 980 kmem_free(new_pgrp, sizeof(*new_pgrp)); 981 #ifdef DEBUG_PGRP 982 if (__predict_false(rval)) 983 printf("enterpgrp(%d,%d,%d), curproc %d, rval %d\n", 984 pid, pgid, mksess, curp->p_pid, rval); 985 #endif 986 return rval; 987 } 988 989 /* 990 * proc_leavepgrp: remove a process from its process group. 991 * => must be called with the proc_lock held, which will be released; 992 */ 993 void 994 proc_leavepgrp(struct proc *p) 995 { 996 struct pgrp *pgrp; 997 998 KASSERT(mutex_owned(proc_lock)); 999 1000 /* Interlock with ttread() */ 1001 mutex_spin_enter(&tty_lock); 1002 pgrp = p->p_pgrp; 1003 LIST_REMOVE(p, p_pglist); 1004 p->p_pgrp = NULL; 1005 mutex_spin_exit(&tty_lock); 1006 1007 if (LIST_EMPTY(&pgrp->pg_members)) { 1008 /* Releases proc_lock. */ 1009 pg_delete(pgrp->pg_id); 1010 } else { 1011 mutex_exit(proc_lock); 1012 } 1013 } 1014 1015 /* 1016 * pg_remove: remove a process group from the table. 1017 * => must be called with the proc_lock held; 1018 * => returns process group to free; 1019 */ 1020 static struct pgrp * 1021 pg_remove(pid_t pg_id) 1022 { 1023 struct pgrp *pgrp; 1024 struct pid_table *pt; 1025 1026 KASSERT(mutex_owned(proc_lock)); 1027 1028 pt = &pid_table[pg_id & pid_tbl_mask]; 1029 pgrp = pt->pt_pgrp; 1030 1031 KASSERT(pgrp != NULL); 1032 KASSERT(pgrp->pg_id == pg_id); 1033 KASSERT(LIST_EMPTY(&pgrp->pg_members)); 1034 1035 pt->pt_pgrp = NULL; 1036 1037 if (!P_VALID(pt->pt_proc)) { 1038 /* Orphaned pgrp, put slot onto free list. */ 1039 KASSERT((P_NEXT(pt->pt_proc) & pid_tbl_mask) == 0); 1040 pg_id &= pid_tbl_mask; 1041 pt = &pid_table[last_free_pt]; 1042 pt->pt_proc = P_FREE(P_NEXT(pt->pt_proc) | pg_id); 1043 KASSERT(pt->pt_pid == 0); 1044 last_free_pt = pg_id; 1045 pid_alloc_cnt--; 1046 } 1047 return pgrp; 1048 } 1049 1050 /* 1051 * pg_delete: delete and free a process group. 1052 * => must be called with the proc_lock held, which will be released. 1053 */ 1054 static void 1055 pg_delete(pid_t pg_id) 1056 { 1057 struct pgrp *pg; 1058 struct tty *ttyp; 1059 struct session *ss; 1060 1061 KASSERT(mutex_owned(proc_lock)); 1062 1063 pg = pid_table[pg_id & pid_tbl_mask].pt_pgrp; 1064 if (pg == NULL || pg->pg_id != pg_id || !LIST_EMPTY(&pg->pg_members)) { 1065 mutex_exit(proc_lock); 1066 return; 1067 } 1068 1069 ss = pg->pg_session; 1070 1071 /* Remove reference (if any) from tty to this process group */ 1072 mutex_spin_enter(&tty_lock); 1073 ttyp = ss->s_ttyp; 1074 if (ttyp != NULL && ttyp->t_pgrp == pg) { 1075 ttyp->t_pgrp = NULL; 1076 KASSERT(ttyp->t_session == ss); 1077 } 1078 mutex_spin_exit(&tty_lock); 1079 1080 /* 1081 * The leading process group in a session is freed by proc_sessrele(), 1082 * if last reference. Note: proc_sessrele() releases proc_lock. 1083 */ 1084 pg = (ss->s_sid != pg->pg_id) ? pg_remove(pg_id) : NULL; 1085 proc_sessrele(ss); 1086 1087 if (pg != NULL) { 1088 /* Free it, if was not done by proc_sessrele(). */ 1089 kmem_free(pg, sizeof(struct pgrp)); 1090 } 1091 } 1092 1093 /* 1094 * Adjust pgrp jobc counters when specified process changes process group. 1095 * We count the number of processes in each process group that "qualify" 1096 * the group for terminal job control (those with a parent in a different 1097 * process group of the same session). If that count reaches zero, the 1098 * process group becomes orphaned. Check both the specified process' 1099 * process group and that of its children. 1100 * entering == 0 => p is leaving specified group. 1101 * entering == 1 => p is entering specified group. 1102 * 1103 * Call with proc_lock held. 1104 */ 1105 void 1106 fixjobc(struct proc *p, struct pgrp *pgrp, int entering) 1107 { 1108 struct pgrp *hispgrp; 1109 struct session *mysession = pgrp->pg_session; 1110 struct proc *child; 1111 1112 KASSERT(mutex_owned(proc_lock)); 1113 1114 /* 1115 * Check p's parent to see whether p qualifies its own process 1116 * group; if so, adjust count for p's process group. 1117 */ 1118 hispgrp = p->p_pptr->p_pgrp; 1119 if (hispgrp != pgrp && hispgrp->pg_session == mysession) { 1120 if (entering) { 1121 pgrp->pg_jobc++; 1122 p->p_lflag &= ~PL_ORPHANPG; 1123 } else if (--pgrp->pg_jobc == 0) 1124 orphanpg(pgrp); 1125 } 1126 1127 /* 1128 * Check this process' children to see whether they qualify 1129 * their process groups; if so, adjust counts for children's 1130 * process groups. 1131 */ 1132 LIST_FOREACH(child, &p->p_children, p_sibling) { 1133 hispgrp = child->p_pgrp; 1134 if (hispgrp != pgrp && hispgrp->pg_session == mysession && 1135 !P_ZOMBIE(child)) { 1136 if (entering) { 1137 child->p_lflag &= ~PL_ORPHANPG; 1138 hispgrp->pg_jobc++; 1139 } else if (--hispgrp->pg_jobc == 0) 1140 orphanpg(hispgrp); 1141 } 1142 } 1143 } 1144 1145 /* 1146 * A process group has become orphaned; 1147 * if there are any stopped processes in the group, 1148 * hang-up all process in that group. 1149 * 1150 * Call with proc_lock held. 1151 */ 1152 static void 1153 orphanpg(struct pgrp *pg) 1154 { 1155 struct proc *p; 1156 1157 KASSERT(mutex_owned(proc_lock)); 1158 1159 LIST_FOREACH(p, &pg->pg_members, p_pglist) { 1160 if (p->p_stat == SSTOP) { 1161 p->p_lflag |= PL_ORPHANPG; 1162 psignal(p, SIGHUP); 1163 psignal(p, SIGCONT); 1164 } 1165 } 1166 } 1167 1168 #ifdef DDB 1169 #include <ddb/db_output.h> 1170 void pidtbl_dump(void); 1171 void 1172 pidtbl_dump(void) 1173 { 1174 struct pid_table *pt; 1175 struct proc *p; 1176 struct pgrp *pgrp; 1177 int id; 1178 1179 db_printf("pid table %p size %x, next %x, last %x\n", 1180 pid_table, pid_tbl_mask+1, 1181 next_free_pt, last_free_pt); 1182 for (pt = pid_table, id = 0; id <= pid_tbl_mask; id++, pt++) { 1183 p = pt->pt_proc; 1184 if (!P_VALID(p) && !pt->pt_pgrp) 1185 continue; 1186 db_printf(" id %x: ", id); 1187 if (P_VALID(p)) 1188 db_printf("slotpid %d proc %p id %d (0x%x) %s\n", 1189 pt->pt_pid, p, p->p_pid, p->p_pid, p->p_comm); 1190 else 1191 db_printf("next %x use %x\n", 1192 P_NEXT(p) & pid_tbl_mask, 1193 P_NEXT(p) & ~pid_tbl_mask); 1194 if ((pgrp = pt->pt_pgrp)) { 1195 db_printf("\tsession %p, sid %d, count %d, login %s\n", 1196 pgrp->pg_session, pgrp->pg_session->s_sid, 1197 pgrp->pg_session->s_count, 1198 pgrp->pg_session->s_login); 1199 db_printf("\tpgrp %p, pg_id %d, pg_jobc %d, members %p\n", 1200 pgrp, pgrp->pg_id, pgrp->pg_jobc, 1201 LIST_FIRST(&pgrp->pg_members)); 1202 LIST_FOREACH(p, &pgrp->pg_members, p_pglist) { 1203 db_printf("\t\tpid %d addr %p pgrp %p %s\n", 1204 p->p_pid, p, p->p_pgrp, p->p_comm); 1205 } 1206 } 1207 } 1208 } 1209 #endif /* DDB */ 1210 1211 #ifdef KSTACK_CHECK_MAGIC 1212 1213 #define KSTACK_MAGIC 0xdeadbeaf 1214 1215 /* XXX should be per process basis? */ 1216 static int kstackleftmin = KSTACK_SIZE; 1217 static int kstackleftthres = KSTACK_SIZE / 8; 1218 1219 void 1220 kstack_setup_magic(const struct lwp *l) 1221 { 1222 uint32_t *ip; 1223 uint32_t const *end; 1224 1225 KASSERT(l != NULL); 1226 KASSERT(l != &lwp0); 1227 1228 /* 1229 * fill all the stack with magic number 1230 * so that later modification on it can be detected. 1231 */ 1232 ip = (uint32_t *)KSTACK_LOWEST_ADDR(l); 1233 end = (uint32_t *)((char *)KSTACK_LOWEST_ADDR(l) + KSTACK_SIZE); 1234 for (; ip < end; ip++) { 1235 *ip = KSTACK_MAGIC; 1236 } 1237 } 1238 1239 void 1240 kstack_check_magic(const struct lwp *l) 1241 { 1242 uint32_t const *ip, *end; 1243 int stackleft; 1244 1245 KASSERT(l != NULL); 1246 1247 /* don't check proc0 */ /*XXX*/ 1248 if (l == &lwp0) 1249 return; 1250 1251 #ifdef __MACHINE_STACK_GROWS_UP 1252 /* stack grows upwards (eg. hppa) */ 1253 ip = (uint32_t *)((void *)KSTACK_LOWEST_ADDR(l) + KSTACK_SIZE); 1254 end = (uint32_t *)KSTACK_LOWEST_ADDR(l); 1255 for (ip--; ip >= end; ip--) 1256 if (*ip != KSTACK_MAGIC) 1257 break; 1258 1259 stackleft = (void *)KSTACK_LOWEST_ADDR(l) + KSTACK_SIZE - (void *)ip; 1260 #else /* __MACHINE_STACK_GROWS_UP */ 1261 /* stack grows downwards (eg. i386) */ 1262 ip = (uint32_t *)KSTACK_LOWEST_ADDR(l); 1263 end = (uint32_t *)((char *)KSTACK_LOWEST_ADDR(l) + KSTACK_SIZE); 1264 for (; ip < end; ip++) 1265 if (*ip != KSTACK_MAGIC) 1266 break; 1267 1268 stackleft = ((const char *)ip) - (const char *)KSTACK_LOWEST_ADDR(l); 1269 #endif /* __MACHINE_STACK_GROWS_UP */ 1270 1271 if (kstackleftmin > stackleft) { 1272 kstackleftmin = stackleft; 1273 if (stackleft < kstackleftthres) 1274 printf("warning: kernel stack left %d bytes" 1275 "(pid %u:lid %u)\n", stackleft, 1276 (u_int)l->l_proc->p_pid, (u_int)l->l_lid); 1277 } 1278 1279 if (stackleft <= 0) { 1280 panic("magic on the top of kernel stack changed for " 1281 "pid %u, lid %u: maybe kernel stack overflow", 1282 (u_int)l->l_proc->p_pid, (u_int)l->l_lid); 1283 } 1284 } 1285 #endif /* KSTACK_CHECK_MAGIC */ 1286 1287 int 1288 proclist_foreach_call(struct proclist *list, 1289 int (*callback)(struct proc *, void *arg), void *arg) 1290 { 1291 struct proc marker; 1292 struct proc *p; 1293 int ret = 0; 1294 1295 marker.p_flag = PK_MARKER; 1296 mutex_enter(proc_lock); 1297 for (p = LIST_FIRST(list); ret == 0 && p != NULL;) { 1298 if (p->p_flag & PK_MARKER) { 1299 p = LIST_NEXT(p, p_list); 1300 continue; 1301 } 1302 LIST_INSERT_AFTER(p, &marker, p_list); 1303 ret = (*callback)(p, arg); 1304 KASSERT(mutex_owned(proc_lock)); 1305 p = LIST_NEXT(&marker, p_list); 1306 LIST_REMOVE(&marker, p_list); 1307 } 1308 mutex_exit(proc_lock); 1309 1310 return ret; 1311 } 1312 1313 int 1314 proc_vmspace_getref(struct proc *p, struct vmspace **vm) 1315 { 1316 1317 /* XXXCDC: how should locking work here? */ 1318 1319 /* curproc exception is for coredump. */ 1320 1321 if ((p != curproc && (p->p_sflag & PS_WEXIT) != 0) || 1322 (p->p_vmspace->vm_refcnt < 1)) { /* XXX */ 1323 return EFAULT; 1324 } 1325 1326 uvmspace_addref(p->p_vmspace); 1327 *vm = p->p_vmspace; 1328 1329 return 0; 1330 } 1331 1332 /* 1333 * Acquire a write lock on the process credential. 1334 */ 1335 void 1336 proc_crmod_enter(void) 1337 { 1338 struct lwp *l = curlwp; 1339 struct proc *p = l->l_proc; 1340 struct plimit *lim; 1341 kauth_cred_t oc; 1342 char *cn; 1343 1344 /* Reset what needs to be reset in plimit. */ 1345 if (p->p_limit->pl_corename != defcorename) { 1346 lim_privatise(p, false); 1347 lim = p->p_limit; 1348 mutex_enter(&lim->pl_lock); 1349 cn = lim->pl_corename; 1350 lim->pl_corename = defcorename; 1351 mutex_exit(&lim->pl_lock); 1352 if (cn != defcorename) 1353 free(cn, M_TEMP); 1354 } 1355 1356 mutex_enter(p->p_lock); 1357 1358 /* Ensure the LWP cached credentials are up to date. */ 1359 if ((oc = l->l_cred) != p->p_cred) { 1360 kauth_cred_hold(p->p_cred); 1361 l->l_cred = p->p_cred; 1362 kauth_cred_free(oc); 1363 } 1364 1365 } 1366 1367 /* 1368 * Set in a new process credential, and drop the write lock. The credential 1369 * must have a reference already. Optionally, free a no-longer required 1370 * credential. The scheduler also needs to inspect p_cred, so we also 1371 * briefly acquire the sched state mutex. 1372 */ 1373 void 1374 proc_crmod_leave(kauth_cred_t scred, kauth_cred_t fcred, bool sugid) 1375 { 1376 struct lwp *l = curlwp, *l2; 1377 struct proc *p = l->l_proc; 1378 kauth_cred_t oc; 1379 1380 KASSERT(mutex_owned(p->p_lock)); 1381 1382 /* Is there a new credential to set in? */ 1383 if (scred != NULL) { 1384 p->p_cred = scred; 1385 LIST_FOREACH(l2, &p->p_lwps, l_sibling) { 1386 if (l2 != l) 1387 l2->l_prflag |= LPR_CRMOD; 1388 } 1389 1390 /* Ensure the LWP cached credentials are up to date. */ 1391 if ((oc = l->l_cred) != scred) { 1392 kauth_cred_hold(scred); 1393 l->l_cred = scred; 1394 } 1395 } else 1396 oc = NULL; /* XXXgcc */ 1397 1398 if (sugid) { 1399 /* 1400 * Mark process as having changed credentials, stops 1401 * tracing etc. 1402 */ 1403 p->p_flag |= PK_SUGID; 1404 } 1405 1406 mutex_exit(p->p_lock); 1407 1408 /* If there is a credential to be released, free it now. */ 1409 if (fcred != NULL) { 1410 KASSERT(scred != NULL); 1411 kauth_cred_free(fcred); 1412 if (oc != scred) 1413 kauth_cred_free(oc); 1414 } 1415 } 1416 1417 /* 1418 * proc_specific_key_create -- 1419 * Create a key for subsystem proc-specific data. 1420 */ 1421 int 1422 proc_specific_key_create(specificdata_key_t *keyp, specificdata_dtor_t dtor) 1423 { 1424 1425 return (specificdata_key_create(proc_specificdata_domain, keyp, dtor)); 1426 } 1427 1428 /* 1429 * proc_specific_key_delete -- 1430 * Delete a key for subsystem proc-specific data. 1431 */ 1432 void 1433 proc_specific_key_delete(specificdata_key_t key) 1434 { 1435 1436 specificdata_key_delete(proc_specificdata_domain, key); 1437 } 1438 1439 /* 1440 * proc_initspecific -- 1441 * Initialize a proc's specificdata container. 1442 */ 1443 void 1444 proc_initspecific(struct proc *p) 1445 { 1446 int error; 1447 1448 error = specificdata_init(proc_specificdata_domain, &p->p_specdataref); 1449 KASSERT(error == 0); 1450 } 1451 1452 /* 1453 * proc_finispecific -- 1454 * Finalize a proc's specificdata container. 1455 */ 1456 void 1457 proc_finispecific(struct proc *p) 1458 { 1459 1460 specificdata_fini(proc_specificdata_domain, &p->p_specdataref); 1461 } 1462 1463 /* 1464 * proc_getspecific -- 1465 * Return proc-specific data corresponding to the specified key. 1466 */ 1467 void * 1468 proc_getspecific(struct proc *p, specificdata_key_t key) 1469 { 1470 1471 return (specificdata_getspecific(proc_specificdata_domain, 1472 &p->p_specdataref, key)); 1473 } 1474 1475 /* 1476 * proc_setspecific -- 1477 * Set proc-specific data corresponding to the specified key. 1478 */ 1479 void 1480 proc_setspecific(struct proc *p, specificdata_key_t key, void *data) 1481 { 1482 1483 specificdata_setspecific(proc_specificdata_domain, 1484 &p->p_specdataref, key, data); 1485 } 1486 1487 int 1488 proc_uidmatch(kauth_cred_t cred, kauth_cred_t target) 1489 { 1490 int r = 0; 1491 1492 if (kauth_cred_getuid(cred) != kauth_cred_getuid(target) || 1493 kauth_cred_getuid(cred) != kauth_cred_getsvuid(target)) { 1494 /* 1495 * suid proc of ours or proc not ours 1496 */ 1497 r = EPERM; 1498 } else if (kauth_cred_getgid(target) != kauth_cred_getsvgid(target)) { 1499 /* 1500 * sgid proc has sgid back to us temporarily 1501 */ 1502 r = EPERM; 1503 } else { 1504 /* 1505 * our rgid must be in target's group list (ie, 1506 * sub-processes started by a sgid process) 1507 */ 1508 int ismember = 0; 1509 1510 if (kauth_cred_ismember_gid(cred, 1511 kauth_cred_getgid(target), &ismember) != 0 || 1512 !ismember) 1513 r = EPERM; 1514 } 1515 1516 return (r); 1517 } 1518 1519 /* 1520 * sysctl stuff 1521 */ 1522 1523 #define KERN_PROCSLOP (5 * sizeof(struct kinfo_proc)) 1524 1525 static const u_int sysctl_flagmap[] = { 1526 PK_ADVLOCK, P_ADVLOCK, 1527 PK_EXEC, P_EXEC, 1528 PK_NOCLDWAIT, P_NOCLDWAIT, 1529 PK_32, P_32, 1530 PK_CLDSIGIGN, P_CLDSIGIGN, 1531 PK_SUGID, P_SUGID, 1532 0 1533 }; 1534 1535 static const u_int sysctl_sflagmap[] = { 1536 PS_NOCLDSTOP, P_NOCLDSTOP, 1537 PS_WEXIT, P_WEXIT, 1538 PS_STOPFORK, P_STOPFORK, 1539 PS_STOPEXEC, P_STOPEXEC, 1540 PS_STOPEXIT, P_STOPEXIT, 1541 0 1542 }; 1543 1544 static const u_int sysctl_slflagmap[] = { 1545 PSL_TRACED, P_TRACED, 1546 PSL_FSTRACE, P_FSTRACE, 1547 PSL_CHTRACED, P_CHTRACED, 1548 PSL_SYSCALL, P_SYSCALL, 1549 0 1550 }; 1551 1552 static const u_int sysctl_lflagmap[] = { 1553 PL_CONTROLT, P_CONTROLT, 1554 PL_PPWAIT, P_PPWAIT, 1555 0 1556 }; 1557 1558 static const u_int sysctl_stflagmap[] = { 1559 PST_PROFIL, P_PROFIL, 1560 0 1561 1562 }; 1563 1564 /* used by kern_lwp also */ 1565 const u_int sysctl_lwpflagmap[] = { 1566 LW_SINTR, L_SINTR, 1567 LW_SYSTEM, L_SYSTEM, 1568 LW_SA, L_SA, /* WRS ??? */ 1569 0 1570 }; 1571 1572 /* 1573 * Find the most ``active'' lwp of a process and return it for ps display 1574 * purposes 1575 */ 1576 static struct lwp * 1577 proc_active_lwp(struct proc *p) 1578 { 1579 static const int ostat[] = { 1580 0, 1581 2, /* LSIDL */ 1582 6, /* LSRUN */ 1583 5, /* LSSLEEP */ 1584 4, /* LSSTOP */ 1585 0, /* LSZOMB */ 1586 1, /* LSDEAD */ 1587 7, /* LSONPROC */ 1588 3 /* LSSUSPENDED */ 1589 }; 1590 1591 struct lwp *l, *lp = NULL; 1592 LIST_FOREACH(l, &p->p_lwps, l_sibling) { 1593 KASSERT(l->l_stat >= 0 && l->l_stat < __arraycount(ostat)); 1594 if (lp == NULL || 1595 ostat[l->l_stat] > ostat[lp->l_stat] || 1596 (ostat[l->l_stat] == ostat[lp->l_stat] && 1597 l->l_cpticks > lp->l_cpticks)) { 1598 lp = l; 1599 continue; 1600 } 1601 } 1602 return lp; 1603 } 1604 1605 static int 1606 sysctl_doeproc(SYSCTLFN_ARGS) 1607 { 1608 union { 1609 struct kinfo_proc kproc; 1610 struct kinfo_proc2 kproc2; 1611 } *kbuf; 1612 struct proc *p, *next, *marker; 1613 char *where, *dp; 1614 int type, op, arg, error; 1615 u_int elem_size, kelem_size, elem_count; 1616 size_t buflen, needed; 1617 bool match, zombie, mmmbrains; 1618 1619 if (namelen == 1 && name[0] == CTL_QUERY) 1620 return (sysctl_query(SYSCTLFN_CALL(rnode))); 1621 1622 dp = where = oldp; 1623 buflen = where != NULL ? *oldlenp : 0; 1624 error = 0; 1625 needed = 0; 1626 type = rnode->sysctl_num; 1627 1628 if (type == KERN_PROC) { 1629 if (namelen != 2 && !(namelen == 1 && name[0] == KERN_PROC_ALL)) 1630 return (EINVAL); 1631 op = name[0]; 1632 if (op != KERN_PROC_ALL) 1633 arg = name[1]; 1634 else 1635 arg = 0; /* Quell compiler warning */ 1636 elem_count = 0; /* Ditto */ 1637 kelem_size = elem_size = sizeof(kbuf->kproc); 1638 } else { 1639 if (namelen != 4) 1640 return (EINVAL); 1641 op = name[0]; 1642 arg = name[1]; 1643 elem_size = name[2]; 1644 elem_count = name[3]; 1645 kelem_size = sizeof(kbuf->kproc2); 1646 } 1647 1648 sysctl_unlock(); 1649 1650 kbuf = kmem_alloc(sizeof(*kbuf), KM_SLEEP); 1651 marker = kmem_alloc(sizeof(*marker), KM_SLEEP); 1652 marker->p_flag = PK_MARKER; 1653 1654 mutex_enter(proc_lock); 1655 mmmbrains = false; 1656 for (p = LIST_FIRST(&allproc);; p = next) { 1657 if (p == NULL) { 1658 if (!mmmbrains) { 1659 p = LIST_FIRST(&zombproc); 1660 mmmbrains = true; 1661 } 1662 if (p == NULL) 1663 break; 1664 } 1665 next = LIST_NEXT(p, p_list); 1666 if ((p->p_flag & PK_MARKER) != 0) 1667 continue; 1668 1669 /* 1670 * Skip embryonic processes. 1671 */ 1672 if (p->p_stat == SIDL) 1673 continue; 1674 1675 mutex_enter(p->p_lock); 1676 error = kauth_authorize_process(l->l_cred, 1677 KAUTH_PROCESS_CANSEE, p, 1678 KAUTH_ARG(KAUTH_REQ_PROCESS_CANSEE_ENTRY), NULL, NULL); 1679 if (error != 0) { 1680 mutex_exit(p->p_lock); 1681 continue; 1682 } 1683 1684 /* 1685 * TODO - make more efficient (see notes below). 1686 * do by session. 1687 */ 1688 switch (op) { 1689 case KERN_PROC_PID: 1690 /* could do this with just a lookup */ 1691 match = (p->p_pid == (pid_t)arg); 1692 break; 1693 1694 case KERN_PROC_PGRP: 1695 /* could do this by traversing pgrp */ 1696 match = (p->p_pgrp->pg_id == (pid_t)arg); 1697 break; 1698 1699 case KERN_PROC_SESSION: 1700 match = (p->p_session->s_sid == (pid_t)arg); 1701 break; 1702 1703 case KERN_PROC_TTY: 1704 match = true; 1705 if (arg == (int) KERN_PROC_TTY_REVOKE) { 1706 if ((p->p_lflag & PL_CONTROLT) == 0 || 1707 p->p_session->s_ttyp == NULL || 1708 p->p_session->s_ttyvp != NULL) { 1709 match = false; 1710 } 1711 } else if ((p->p_lflag & PL_CONTROLT) == 0 || 1712 p->p_session->s_ttyp == NULL) { 1713 if ((dev_t)arg != KERN_PROC_TTY_NODEV) { 1714 match = false; 1715 } 1716 } else if (p->p_session->s_ttyp->t_dev != (dev_t)arg) { 1717 match = false; 1718 } 1719 break; 1720 1721 case KERN_PROC_UID: 1722 match = (kauth_cred_geteuid(p->p_cred) == (uid_t)arg); 1723 break; 1724 1725 case KERN_PROC_RUID: 1726 match = (kauth_cred_getuid(p->p_cred) == (uid_t)arg); 1727 break; 1728 1729 case KERN_PROC_GID: 1730 match = (kauth_cred_getegid(p->p_cred) == (uid_t)arg); 1731 break; 1732 1733 case KERN_PROC_RGID: 1734 match = (kauth_cred_getgid(p->p_cred) == (uid_t)arg); 1735 break; 1736 1737 case KERN_PROC_ALL: 1738 match = true; 1739 /* allow everything */ 1740 break; 1741 1742 default: 1743 error = EINVAL; 1744 mutex_exit(p->p_lock); 1745 goto cleanup; 1746 } 1747 if (!match) { 1748 mutex_exit(p->p_lock); 1749 continue; 1750 } 1751 1752 /* 1753 * Grab a hold on the process. 1754 */ 1755 if (mmmbrains) { 1756 zombie = true; 1757 } else { 1758 zombie = !rw_tryenter(&p->p_reflock, RW_READER); 1759 } 1760 if (zombie) { 1761 LIST_INSERT_AFTER(p, marker, p_list); 1762 } 1763 1764 if (buflen >= elem_size && 1765 (type == KERN_PROC || elem_count > 0)) { 1766 if (type == KERN_PROC) { 1767 kbuf->kproc.kp_proc = *p; 1768 fill_eproc(p, &kbuf->kproc.kp_eproc, zombie); 1769 } else { 1770 fill_kproc2(p, &kbuf->kproc2, zombie); 1771 elem_count--; 1772 } 1773 mutex_exit(p->p_lock); 1774 mutex_exit(proc_lock); 1775 /* 1776 * Copy out elem_size, but not larger than kelem_size 1777 */ 1778 error = sysctl_copyout(l, kbuf, dp, 1779 min(kelem_size, elem_size)); 1780 mutex_enter(proc_lock); 1781 if (error) { 1782 goto bah; 1783 } 1784 dp += elem_size; 1785 buflen -= elem_size; 1786 } else { 1787 mutex_exit(p->p_lock); 1788 } 1789 needed += elem_size; 1790 1791 /* 1792 * Release reference to process. 1793 */ 1794 if (zombie) { 1795 next = LIST_NEXT(marker, p_list); 1796 LIST_REMOVE(marker, p_list); 1797 } else { 1798 rw_exit(&p->p_reflock); 1799 next = LIST_NEXT(p, p_list); 1800 } 1801 } 1802 mutex_exit(proc_lock); 1803 1804 if (where != NULL) { 1805 *oldlenp = dp - where; 1806 if (needed > *oldlenp) { 1807 error = ENOMEM; 1808 goto out; 1809 } 1810 } else { 1811 needed += KERN_PROCSLOP; 1812 *oldlenp = needed; 1813 } 1814 if (kbuf) 1815 kmem_free(kbuf, sizeof(*kbuf)); 1816 if (marker) 1817 kmem_free(marker, sizeof(*marker)); 1818 sysctl_relock(); 1819 return 0; 1820 bah: 1821 if (zombie) 1822 LIST_REMOVE(marker, p_list); 1823 else 1824 rw_exit(&p->p_reflock); 1825 cleanup: 1826 mutex_exit(proc_lock); 1827 out: 1828 if (kbuf) 1829 kmem_free(kbuf, sizeof(*kbuf)); 1830 if (marker) 1831 kmem_free(marker, sizeof(*marker)); 1832 sysctl_relock(); 1833 return error; 1834 } 1835 1836 /* 1837 * sysctl helper routine for kern.proc_args pseudo-subtree. 1838 */ 1839 static int 1840 sysctl_kern_proc_args(SYSCTLFN_ARGS) 1841 { 1842 struct ps_strings pss; 1843 struct proc *p; 1844 size_t len, i; 1845 struct uio auio; 1846 struct iovec aiov; 1847 pid_t pid; 1848 int nargv, type, error, argvlen; 1849 char *arg; 1850 char **argv = NULL; 1851 char *tmp; 1852 struct vmspace *vmspace; 1853 vaddr_t psstr_addr; 1854 vaddr_t offsetn; 1855 vaddr_t offsetv; 1856 1857 if (namelen == 1 && name[0] == CTL_QUERY) 1858 return (sysctl_query(SYSCTLFN_CALL(rnode))); 1859 1860 if (newp != NULL || namelen != 2) 1861 return (EINVAL); 1862 pid = name[0]; 1863 type = name[1]; 1864 argv = NULL; 1865 argvlen = 0; 1866 1867 switch (type) { 1868 case KERN_PROC_ARGV: 1869 case KERN_PROC_NARGV: 1870 case KERN_PROC_ENV: 1871 case KERN_PROC_NENV: 1872 /* ok */ 1873 break; 1874 default: 1875 return (EINVAL); 1876 } 1877 1878 sysctl_unlock(); 1879 1880 /* check pid */ 1881 mutex_enter(proc_lock); 1882 if ((p = proc_find(pid)) == NULL) { 1883 error = EINVAL; 1884 goto out_locked; 1885 } 1886 mutex_enter(p->p_lock); 1887 1888 /* Check permission. */ 1889 if (type == KERN_PROC_ARGV || type == KERN_PROC_NARGV) 1890 error = kauth_authorize_process(l->l_cred, KAUTH_PROCESS_CANSEE, 1891 p, KAUTH_ARG(KAUTH_REQ_PROCESS_CANSEE_ARGS), NULL, NULL); 1892 else if (type == KERN_PROC_ENV || type == KERN_PROC_NENV) 1893 error = kauth_authorize_process(l->l_cred, KAUTH_PROCESS_CANSEE, 1894 p, KAUTH_ARG(KAUTH_REQ_PROCESS_CANSEE_ENV), NULL, NULL); 1895 else 1896 error = EINVAL; /* XXXGCC */ 1897 if (error) { 1898 mutex_exit(p->p_lock); 1899 goto out_locked; 1900 } 1901 1902 if (oldp == NULL) { 1903 if (type == KERN_PROC_NARGV || type == KERN_PROC_NENV) 1904 *oldlenp = sizeof (int); 1905 else 1906 *oldlenp = ARG_MAX; /* XXX XXX XXX */ 1907 error = 0; 1908 mutex_exit(p->p_lock); 1909 goto out_locked; 1910 } 1911 1912 /* 1913 * Zombies don't have a stack, so we can't read their psstrings. 1914 * System processes also don't have a user stack. 1915 */ 1916 if (P_ZOMBIE(p) || (p->p_flag & PK_SYSTEM) != 0) { 1917 error = EINVAL; 1918 mutex_exit(p->p_lock); 1919 goto out_locked; 1920 } 1921 1922 /* 1923 * Lock the process down in memory. 1924 */ 1925 psstr_addr = (vaddr_t)p->p_psstr; 1926 if (type == KERN_PROC_ARGV || type == KERN_PROC_NARGV) { 1927 offsetn = p->p_psnargv; 1928 offsetv = p->p_psargv; 1929 } else { 1930 offsetn = p->p_psnenv; 1931 offsetv = p->p_psenv; 1932 } 1933 vmspace = p->p_vmspace; 1934 uvmspace_addref(vmspace); 1935 mutex_exit(p->p_lock); 1936 mutex_exit(proc_lock); 1937 1938 /* 1939 * Allocate a temporary buffer to hold the arguments. 1940 */ 1941 arg = kmem_alloc(PAGE_SIZE, KM_SLEEP); 1942 1943 /* 1944 * Read in the ps_strings structure. 1945 */ 1946 aiov.iov_base = &pss; 1947 aiov.iov_len = sizeof(pss); 1948 auio.uio_iov = &aiov; 1949 auio.uio_iovcnt = 1; 1950 auio.uio_offset = psstr_addr; 1951 auio.uio_resid = sizeof(pss); 1952 auio.uio_rw = UIO_READ; 1953 UIO_SETUP_SYSSPACE(&auio); 1954 error = uvm_io(&vmspace->vm_map, &auio); 1955 if (error) 1956 goto done; 1957 1958 memcpy(&nargv, (char *)&pss + offsetn, sizeof(nargv)); 1959 if (type == KERN_PROC_NARGV || type == KERN_PROC_NENV) { 1960 error = sysctl_copyout(l, &nargv, oldp, sizeof(nargv)); 1961 *oldlenp = sizeof(nargv); 1962 goto done; 1963 } 1964 /* 1965 * Now read the address of the argument vector. 1966 */ 1967 switch (type) { 1968 case KERN_PROC_ARGV: 1969 /* FALLTHROUGH */ 1970 case KERN_PROC_ENV: 1971 memcpy(&tmp, (char *)&pss + offsetv, sizeof(tmp)); 1972 break; 1973 default: 1974 error = EINVAL; 1975 goto done; 1976 } 1977 1978 #ifdef COMPAT_NETBSD32 1979 if (p->p_flag & PK_32) 1980 len = sizeof(netbsd32_charp) * nargv; 1981 else 1982 #endif 1983 len = sizeof(char *) * nargv; 1984 1985 if ((argvlen = len) != 0) 1986 argv = kmem_alloc(len, KM_SLEEP); 1987 1988 aiov.iov_base = argv; 1989 aiov.iov_len = len; 1990 auio.uio_iov = &aiov; 1991 auio.uio_iovcnt = 1; 1992 auio.uio_offset = (off_t)(unsigned long)tmp; 1993 auio.uio_resid = len; 1994 auio.uio_rw = UIO_READ; 1995 UIO_SETUP_SYSSPACE(&auio); 1996 error = uvm_io(&vmspace->vm_map, &auio); 1997 if (error) 1998 goto done; 1999 2000 /* 2001 * Now copy each string. 2002 */ 2003 len = 0; /* bytes written to user buffer */ 2004 for (i = 0; i < nargv; i++) { 2005 int finished = 0; 2006 vaddr_t base; 2007 size_t xlen; 2008 int j; 2009 2010 #ifdef COMPAT_NETBSD32 2011 if (p->p_flag & PK_32) { 2012 netbsd32_charp *argv32; 2013 2014 argv32 = (netbsd32_charp *)argv; 2015 base = (vaddr_t)NETBSD32PTR64(argv32[i]); 2016 } else 2017 #endif 2018 base = (vaddr_t)argv[i]; 2019 2020 /* 2021 * The program has messed around with its arguments, 2022 * possibly deleting some, and replacing them with 2023 * NULL's. Treat this as the last argument and not 2024 * a failure. 2025 */ 2026 if (base == 0) 2027 break; 2028 2029 while (!finished) { 2030 xlen = PAGE_SIZE - (base & PAGE_MASK); 2031 2032 aiov.iov_base = arg; 2033 aiov.iov_len = PAGE_SIZE; 2034 auio.uio_iov = &aiov; 2035 auio.uio_iovcnt = 1; 2036 auio.uio_offset = base; 2037 auio.uio_resid = xlen; 2038 auio.uio_rw = UIO_READ; 2039 UIO_SETUP_SYSSPACE(&auio); 2040 error = uvm_io(&vmspace->vm_map, &auio); 2041 if (error) 2042 goto done; 2043 2044 /* Look for the end of the string */ 2045 for (j = 0; j < xlen; j++) { 2046 if (arg[j] == '\0') { 2047 xlen = j + 1; 2048 finished = 1; 2049 break; 2050 } 2051 } 2052 2053 /* Check for user buffer overflow */ 2054 if (len + xlen > *oldlenp) { 2055 finished = 1; 2056 if (len > *oldlenp) 2057 xlen = 0; 2058 else 2059 xlen = *oldlenp - len; 2060 } 2061 2062 /* Copyout the page */ 2063 error = sysctl_copyout(l, arg, (char*)oldp + len, xlen); 2064 if (error) 2065 goto done; 2066 2067 len += xlen; 2068 base += xlen; 2069 } 2070 } 2071 *oldlenp = len; 2072 2073 done: 2074 if (argvlen != 0) 2075 kmem_free(argv, argvlen); 2076 uvmspace_free(vmspace); 2077 kmem_free(arg, PAGE_SIZE); 2078 sysctl_relock(); 2079 return error; 2080 2081 out_locked: 2082 mutex_exit(proc_lock); 2083 sysctl_relock(); 2084 return error; 2085 } 2086 2087 /* 2088 * Fill in an eproc structure for the specified process. 2089 */ 2090 void 2091 fill_eproc(struct proc *p, struct eproc *ep, bool zombie) 2092 { 2093 struct tty *tp; 2094 struct lwp *l; 2095 2096 KASSERT(mutex_owned(proc_lock)); 2097 KASSERT(mutex_owned(p->p_lock)); 2098 2099 memset(ep, 0, sizeof(*ep)); 2100 2101 ep->e_paddr = p; 2102 ep->e_sess = p->p_session; 2103 if (p->p_cred) { 2104 kauth_cred_topcred(p->p_cred, &ep->e_pcred); 2105 kauth_cred_toucred(p->p_cred, &ep->e_ucred); 2106 } 2107 if (p->p_stat != SIDL && !P_ZOMBIE(p) && !zombie) { 2108 struct vmspace *vm = p->p_vmspace; 2109 2110 ep->e_vm.vm_rssize = vm_resident_count(vm); 2111 ep->e_vm.vm_tsize = vm->vm_tsize; 2112 ep->e_vm.vm_dsize = vm->vm_dsize; 2113 ep->e_vm.vm_ssize = vm->vm_ssize; 2114 ep->e_vm.vm_map.size = vm->vm_map.size; 2115 2116 /* Pick the primary (first) LWP */ 2117 l = proc_active_lwp(p); 2118 KASSERT(l != NULL); 2119 lwp_lock(l); 2120 if (l->l_wchan) 2121 strncpy(ep->e_wmesg, l->l_wmesg, WMESGLEN); 2122 lwp_unlock(l); 2123 } 2124 if (p->p_pptr) 2125 ep->e_ppid = p->p_pptr->p_pid; 2126 if (p->p_pgrp && p->p_session) { 2127 ep->e_pgid = p->p_pgrp->pg_id; 2128 ep->e_jobc = p->p_pgrp->pg_jobc; 2129 ep->e_sid = p->p_session->s_sid; 2130 if ((p->p_lflag & PL_CONTROLT) && 2131 (tp = ep->e_sess->s_ttyp)) { 2132 ep->e_tdev = tp->t_dev; 2133 ep->e_tpgid = tp->t_pgrp ? tp->t_pgrp->pg_id : NO_PGID; 2134 ep->e_tsess = tp->t_session; 2135 } else 2136 ep->e_tdev = (uint32_t)NODEV; 2137 ep->e_flag = ep->e_sess->s_ttyvp ? EPROC_CTTY : 0; 2138 if (SESS_LEADER(p)) 2139 ep->e_flag |= EPROC_SLEADER; 2140 strncpy(ep->e_login, ep->e_sess->s_login, MAXLOGNAME); 2141 } 2142 ep->e_xsize = ep->e_xrssize = 0; 2143 ep->e_xccount = ep->e_xswrss = 0; 2144 } 2145 2146 /* 2147 * Fill in a kinfo_proc2 structure for the specified process. 2148 */ 2149 static void 2150 fill_kproc2(struct proc *p, struct kinfo_proc2 *ki, bool zombie) 2151 { 2152 struct tty *tp; 2153 struct lwp *l, *l2; 2154 struct timeval ut, st, rt; 2155 sigset_t ss1, ss2; 2156 struct rusage ru; 2157 struct vmspace *vm; 2158 2159 KASSERT(mutex_owned(proc_lock)); 2160 KASSERT(mutex_owned(p->p_lock)); 2161 2162 sigemptyset(&ss1); 2163 sigemptyset(&ss2); 2164 memset(ki, 0, sizeof(*ki)); 2165 2166 ki->p_paddr = PTRTOUINT64(p); 2167 ki->p_fd = PTRTOUINT64(p->p_fd); 2168 ki->p_cwdi = PTRTOUINT64(p->p_cwdi); 2169 ki->p_stats = PTRTOUINT64(p->p_stats); 2170 ki->p_limit = PTRTOUINT64(p->p_limit); 2171 ki->p_vmspace = PTRTOUINT64(p->p_vmspace); 2172 ki->p_sigacts = PTRTOUINT64(p->p_sigacts); 2173 ki->p_sess = PTRTOUINT64(p->p_session); 2174 ki->p_tsess = 0; /* may be changed if controlling tty below */ 2175 ki->p_ru = PTRTOUINT64(&p->p_stats->p_ru); 2176 ki->p_eflag = 0; 2177 ki->p_exitsig = p->p_exitsig; 2178 ki->p_flag = L_INMEM; /* Process never swapped out */ 2179 ki->p_flag |= sysctl_map_flags(sysctl_flagmap, p->p_flag); 2180 ki->p_flag |= sysctl_map_flags(sysctl_sflagmap, p->p_sflag); 2181 ki->p_flag |= sysctl_map_flags(sysctl_slflagmap, p->p_slflag); 2182 ki->p_flag |= sysctl_map_flags(sysctl_lflagmap, p->p_lflag); 2183 ki->p_flag |= sysctl_map_flags(sysctl_stflagmap, p->p_stflag); 2184 ki->p_pid = p->p_pid; 2185 if (p->p_pptr) 2186 ki->p_ppid = p->p_pptr->p_pid; 2187 else 2188 ki->p_ppid = 0; 2189 ki->p_uid = kauth_cred_geteuid(p->p_cred); 2190 ki->p_ruid = kauth_cred_getuid(p->p_cred); 2191 ki->p_gid = kauth_cred_getegid(p->p_cred); 2192 ki->p_rgid = kauth_cred_getgid(p->p_cred); 2193 ki->p_svuid = kauth_cred_getsvuid(p->p_cred); 2194 ki->p_svgid = kauth_cred_getsvgid(p->p_cred); 2195 ki->p_ngroups = kauth_cred_ngroups(p->p_cred); 2196 kauth_cred_getgroups(p->p_cred, ki->p_groups, 2197 min(ki->p_ngroups, sizeof(ki->p_groups) / sizeof(ki->p_groups[0])), 2198 UIO_SYSSPACE); 2199 2200 ki->p_uticks = p->p_uticks; 2201 ki->p_sticks = p->p_sticks; 2202 ki->p_iticks = p->p_iticks; 2203 ki->p_tpgid = NO_PGID; /* may be changed if controlling tty below */ 2204 ki->p_tracep = PTRTOUINT64(p->p_tracep); 2205 ki->p_traceflag = p->p_traceflag; 2206 2207 memcpy(&ki->p_sigignore, &p->p_sigctx.ps_sigignore,sizeof(ki_sigset_t)); 2208 memcpy(&ki->p_sigcatch, &p->p_sigctx.ps_sigcatch, sizeof(ki_sigset_t)); 2209 2210 ki->p_cpticks = 0; 2211 ki->p_pctcpu = p->p_pctcpu; 2212 ki->p_estcpu = 0; 2213 ki->p_stat = p->p_stat; /* Will likely be overridden by LWP status */ 2214 ki->p_realstat = p->p_stat; 2215 ki->p_nice = p->p_nice; 2216 ki->p_xstat = p->p_xstat; 2217 ki->p_acflag = p->p_acflag; 2218 2219 strncpy(ki->p_comm, p->p_comm, 2220 min(sizeof(ki->p_comm), sizeof(p->p_comm))); 2221 strncpy(ki->p_ename, p->p_emul->e_name, sizeof(ki->p_ename)); 2222 2223 ki->p_nlwps = p->p_nlwps; 2224 ki->p_realflag = ki->p_flag; 2225 2226 if (p->p_stat != SIDL && !P_ZOMBIE(p) && !zombie) { 2227 vm = p->p_vmspace; 2228 ki->p_vm_rssize = vm_resident_count(vm); 2229 ki->p_vm_tsize = vm->vm_tsize; 2230 ki->p_vm_dsize = vm->vm_dsize; 2231 ki->p_vm_ssize = vm->vm_ssize; 2232 ki->p_vm_vsize = vm->vm_map.size; 2233 /* 2234 * Since the stack is initially mapped mostly with 2235 * PROT_NONE and grown as needed, adjust the "mapped size" 2236 * to skip the unused stack portion. 2237 */ 2238 ki->p_vm_msize = 2239 atop(vm->vm_map.size) - vm->vm_issize + vm->vm_ssize; 2240 2241 /* Pick the primary (first) LWP */ 2242 l = proc_active_lwp(p); 2243 KASSERT(l != NULL); 2244 lwp_lock(l); 2245 ki->p_nrlwps = p->p_nrlwps; 2246 ki->p_forw = 0; 2247 ki->p_back = 0; 2248 ki->p_addr = PTRTOUINT64(l->l_addr); 2249 ki->p_stat = l->l_stat; 2250 ki->p_flag |= sysctl_map_flags(sysctl_lwpflagmap, l->l_flag); 2251 ki->p_swtime = l->l_swtime; 2252 ki->p_slptime = l->l_slptime; 2253 if (l->l_stat == LSONPROC) 2254 ki->p_schedflags = l->l_cpu->ci_schedstate.spc_flags; 2255 else 2256 ki->p_schedflags = 0; 2257 ki->p_priority = lwp_eprio(l); 2258 ki->p_usrpri = l->l_priority; 2259 if (l->l_wchan) 2260 strncpy(ki->p_wmesg, l->l_wmesg, sizeof(ki->p_wmesg)); 2261 ki->p_wchan = PTRTOUINT64(l->l_wchan); 2262 ki->p_cpuid = cpu_index(l->l_cpu); 2263 lwp_unlock(l); 2264 LIST_FOREACH(l, &p->p_lwps, l_sibling) { 2265 /* This is hardly correct, but... */ 2266 sigplusset(&l->l_sigpend.sp_set, &ss1); 2267 sigplusset(&l->l_sigmask, &ss2); 2268 ki->p_cpticks += l->l_cpticks; 2269 ki->p_pctcpu += l->l_pctcpu; 2270 ki->p_estcpu += l->l_estcpu; 2271 } 2272 } 2273 sigplusset(&p->p_sigpend.sp_set, &ss2); 2274 memcpy(&ki->p_siglist, &ss1, sizeof(ki_sigset_t)); 2275 memcpy(&ki->p_sigmask, &ss2, sizeof(ki_sigset_t)); 2276 2277 if (p->p_session != NULL) { 2278 ki->p_sid = p->p_session->s_sid; 2279 ki->p__pgid = p->p_pgrp->pg_id; 2280 if (p->p_session->s_ttyvp) 2281 ki->p_eflag |= EPROC_CTTY; 2282 if (SESS_LEADER(p)) 2283 ki->p_eflag |= EPROC_SLEADER; 2284 strncpy(ki->p_login, p->p_session->s_login, 2285 min(sizeof ki->p_login - 1, sizeof p->p_session->s_login)); 2286 ki->p_jobc = p->p_pgrp->pg_jobc; 2287 if ((p->p_lflag & PL_CONTROLT) && (tp = p->p_session->s_ttyp)) { 2288 ki->p_tdev = tp->t_dev; 2289 ki->p_tpgid = tp->t_pgrp ? tp->t_pgrp->pg_id : NO_PGID; 2290 ki->p_tsess = PTRTOUINT64(tp->t_session); 2291 } else { 2292 ki->p_tdev = (int32_t)NODEV; 2293 } 2294 } 2295 2296 if (!P_ZOMBIE(p) && !zombie) { 2297 ki->p_uvalid = 1; 2298 ki->p_ustart_sec = p->p_stats->p_start.tv_sec; 2299 ki->p_ustart_usec = p->p_stats->p_start.tv_usec; 2300 2301 calcru(p, &ut, &st, NULL, &rt); 2302 ki->p_rtime_sec = rt.tv_sec; 2303 ki->p_rtime_usec = rt.tv_usec; 2304 ki->p_uutime_sec = ut.tv_sec; 2305 ki->p_uutime_usec = ut.tv_usec; 2306 ki->p_ustime_sec = st.tv_sec; 2307 ki->p_ustime_usec = st.tv_usec; 2308 2309 memcpy(&ru, &p->p_stats->p_ru, sizeof(ru)); 2310 ki->p_uru_nvcsw = 0; 2311 ki->p_uru_nivcsw = 0; 2312 LIST_FOREACH(l2, &p->p_lwps, l_sibling) { 2313 ki->p_uru_nvcsw += (l2->l_ncsw - l2->l_nivcsw); 2314 ki->p_uru_nivcsw += l2->l_nivcsw; 2315 ruadd(&ru, &l2->l_ru); 2316 } 2317 ki->p_uru_maxrss = ru.ru_maxrss; 2318 ki->p_uru_ixrss = ru.ru_ixrss; 2319 ki->p_uru_idrss = ru.ru_idrss; 2320 ki->p_uru_isrss = ru.ru_isrss; 2321 ki->p_uru_minflt = ru.ru_minflt; 2322 ki->p_uru_majflt = ru.ru_majflt; 2323 ki->p_uru_nswap = ru.ru_nswap; 2324 ki->p_uru_inblock = ru.ru_inblock; 2325 ki->p_uru_oublock = ru.ru_oublock; 2326 ki->p_uru_msgsnd = ru.ru_msgsnd; 2327 ki->p_uru_msgrcv = ru.ru_msgrcv; 2328 ki->p_uru_nsignals = ru.ru_nsignals; 2329 2330 timeradd(&p->p_stats->p_cru.ru_utime, 2331 &p->p_stats->p_cru.ru_stime, &ut); 2332 ki->p_uctime_sec = ut.tv_sec; 2333 ki->p_uctime_usec = ut.tv_usec; 2334 } 2335 } 2336
Indexes created Sun Sep 21 20:09:37 GMT 2025