1 /*- 2 * Copyright (C) 2001 Julian Elischer <julian@freebsd.org>. 3 * All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice(s), this list of conditions and the following disclaimer as 10 * the first lines of this file unmodified other than the possible 11 * addition of one or more copyright notices. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice(s), this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY 17 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 18 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 19 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY 20 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 21 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 22 * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 23 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH 26 * DAMAGE. 27 */ 28 29 #include "opt_witness.h" 30 #include "opt_hwpmc_hooks.h" 31 32 #include <sys/cdefs.h> 33 __FBSDID("$FreeBSD$"); 34 35 #include <sys/param.h> 36 #include <sys/systm.h> 37 #include <sys/kernel.h> 38 #include <sys/lock.h> 39 #include <sys/mutex.h> 40 #include <sys/proc.h> 41 #include <sys/rangelock.h> 42 #include <sys/resourcevar.h> 43 #include <sys/sdt.h> 44 #include <sys/smp.h> 45 #include <sys/sched.h> 46 #include <sys/sleepqueue.h> 47 #include <sys/selinfo.h> 48 #include <sys/sysent.h> 49 #include <sys/turnstile.h> 50 #include <sys/ktr.h> 51 #include <sys/rwlock.h> 52 #include <sys/umtx.h> 53 #include <sys/cpuset.h> 54 #ifdef HWPMC_HOOKS 55 #include <sys/pmckern.h> 56 #endif 57 58 #include <security/audit/audit.h> 59 60 #include <vm/vm.h> 61 #include <vm/vm_extern.h> 62 #include <vm/uma.h> 63 #include <vm/vm_domain.h> 64 #include <sys/eventhandler.h> 65 66 SDT_PROVIDER_DECLARE(proc); 67 SDT_PROBE_DEFINE(proc, , , lwp__exit); 68 69 /* 70 * thread related storage. 71 */ 72 static uma_zone_t thread_zone; 73 74 TAILQ_HEAD(, thread) zombie_threads = TAILQ_HEAD_INITIALIZER(zombie_threads); 75 static struct mtx zombie_lock; 76 MTX_SYSINIT(zombie_lock, &zombie_lock, "zombie lock", MTX_SPIN); 77 78 static void thread_zombie(struct thread *); 79 static int thread_unsuspend_one(struct thread *td, struct proc *p, 80 bool boundary); 81 82 #define TID_BUFFER_SIZE 1024 83 84 struct mtx tid_lock; 85 static struct unrhdr *tid_unrhdr; 86 static lwpid_t tid_buffer[TID_BUFFER_SIZE]; 87 static int tid_head, tid_tail; 88 static MALLOC_DEFINE(M_TIDHASH, "tidhash", "thread hash"); 89 90 struct tidhashhead *tidhashtbl; 91 u_long tidhash; 92 struct rwlock tidhash_lock; 93 94 static lwpid_t 95 tid_alloc(void) 96 { 97 lwpid_t tid; 98 99 tid = alloc_unr(tid_unrhdr); 100 if (tid != -1) 101 return (tid); 102 mtx_lock(&tid_lock); 103 if (tid_head == tid_tail) { 104 mtx_unlock(&tid_lock); 105 return (-1); 106 } 107 tid = tid_buffer[tid_head]; 108 tid_head = (tid_head + 1) % TID_BUFFER_SIZE; 109 mtx_unlock(&tid_lock); 110 return (tid); 111 } 112 113 static void 114 tid_free(lwpid_t tid) 115 { 116 lwpid_t tmp_tid = -1; 117 118 mtx_lock(&tid_lock); 119 if ((tid_tail + 1) % TID_BUFFER_SIZE == tid_head) { 120 tmp_tid = tid_buffer[tid_head]; 121 tid_head = (tid_head + 1) % TID_BUFFER_SIZE; 122 } 123 tid_buffer[tid_tail] = tid; 124 tid_tail = (tid_tail + 1) % TID_BUFFER_SIZE; 125 mtx_unlock(&tid_lock); 126 if (tmp_tid != -1) 127 free_unr(tid_unrhdr, tmp_tid); 128 } 129 130 /* 131 * Prepare a thread for use. 132 */ 133 static int 134 thread_ctor(void *mem, int size, void *arg, int flags) 135 { 136 struct thread *td; 137 138 td = (struct thread *)mem; 139 td->td_state = TDS_INACTIVE; 140 td->td_oncpu = NOCPU; 141 142 td->td_tid = tid_alloc(); 143 144 /* 145 * Note that td_critnest begins life as 1 because the thread is not 146 * running and is thereby implicitly waiting to be on the receiving 147 * end of a context switch. 148 */ 149 td->td_critnest = 1; 150 td->td_lend_user_pri = PRI_MAX; 151 EVENTHANDLER_INVOKE(thread_ctor, td); 152 #ifdef AUDIT 153 audit_thread_alloc(td); 154 #endif 155 umtx_thread_alloc(td); 156 return (0); 157 } 158 159 /* 160 * Reclaim a thread after use. 161 */ 162 static void 163 thread_dtor(void *mem, int size, void *arg) 164 { 165 struct thread *td; 166 167 td = (struct thread *)mem; 168 169 #ifdef INVARIANTS 170 /* Verify that this thread is in a safe state to free. */ 171 switch (td->td_state) { 172 case TDS_INHIBITED: 173 case TDS_RUNNING: 174 case TDS_CAN_RUN: 175 case TDS_RUNQ: 176 /* 177 * We must never unlink a thread that is in one of 178 * these states, because it is currently active. 179 */ 180 panic("bad state for thread unlinking"); 181 /* NOTREACHED */ 182 case TDS_INACTIVE: 183 break; 184 default: 185 panic("bad thread state"); 186 /* NOTREACHED */ 187 } 188 #endif 189 #ifdef AUDIT 190 audit_thread_free(td); 191 #endif 192 /* Free all OSD associated to this thread. */ 193 osd_thread_exit(td); 194 195 EVENTHANDLER_INVOKE(thread_dtor, td); 196 tid_free(td->td_tid); 197 } 198 199 /* 200 * Initialize type-stable parts of a thread (when newly created). 201 */ 202 static int 203 thread_init(void *mem, int size, int flags) 204 { 205 struct thread *td; 206 207 td = (struct thread *)mem; 208 209 td->td_sleepqueue = sleepq_alloc(); 210 td->td_turnstile = turnstile_alloc(); 211 td->td_rlqe = NULL; 212 EVENTHANDLER_INVOKE(thread_init, td); 213 td->td_sched = (struct td_sched *)&td[1]; 214 umtx_thread_init(td); 215 td->td_kstack = 0; 216 td->td_sel = NULL; 217 return (0); 218 } 219 220 /* 221 * Tear down type-stable parts of a thread (just before being discarded). 222 */ 223 static void 224 thread_fini(void *mem, int size) 225 { 226 struct thread *td; 227 228 td = (struct thread *)mem; 229 EVENTHANDLER_INVOKE(thread_fini, td); 230 rlqentry_free(td->td_rlqe); 231 turnstile_free(td->td_turnstile); 232 sleepq_free(td->td_sleepqueue); 233 umtx_thread_fini(td); 234 seltdfini(td); 235 } 236 237 /* 238 * For a newly created process, 239 * link up all the structures and its initial threads etc. 240 * called from: 241 * {arch}/{arch}/machdep.c {arch}_init(), init386() etc. 242 * proc_dtor() (should go away) 243 * proc_init() 244 */ 245 void 246 proc_linkup0(struct proc *p, struct thread *td) 247 { 248 TAILQ_INIT(&p->p_threads); /* all threads in proc */ 249 proc_linkup(p, td); 250 } 251 252 void 253 proc_linkup(struct proc *p, struct thread *td) 254 { 255 256 sigqueue_init(&p->p_sigqueue, p); 257 p->p_ksi = ksiginfo_alloc(1); 258 if (p->p_ksi != NULL) { 259 /* XXX p_ksi may be null if ksiginfo zone is not ready */ 260 p->p_ksi->ksi_flags = KSI_EXT | KSI_INS; 261 } 262 LIST_INIT(&p->p_mqnotifier); 263 p->p_numthreads = 0; 264 thread_link(td, p); 265 } 266 267 /* 268 * Initialize global thread allocation resources. 269 */ 270 void 271 threadinit(void) 272 { 273 274 mtx_init(&tid_lock, "TID lock", NULL, MTX_DEF); 275 276 /* 277 * pid_max cannot be greater than PID_MAX. 278 * leave one number for thread0. 279 */ 280 tid_unrhdr = new_unrhdr(PID_MAX + 2, INT_MAX, &tid_lock); 281 282 thread_zone = uma_zcreate("THREAD", sched_sizeof_thread(), 283 thread_ctor, thread_dtor, thread_init, thread_fini, 284 16 - 1, 0); 285 tidhashtbl = hashinit(maxproc / 2, M_TIDHASH, &tidhash); 286 rw_init(&tidhash_lock, "tidhash"); 287 } 288 289 /* 290 * Place an unused thread on the zombie list. 291 * Use the slpq as that must be unused by now. 292 */ 293 void 294 thread_zombie(struct thread *td) 295 { 296 mtx_lock_spin(&zombie_lock); 297 TAILQ_INSERT_HEAD(&zombie_threads, td, td_slpq); 298 mtx_unlock_spin(&zombie_lock); 299 } 300 301 /* 302 * Release a thread that has exited after cpu_throw(). 303 */ 304 void 305 thread_stash(struct thread *td) 306 { 307 atomic_subtract_rel_int(&td->td_proc->p_exitthreads, 1); 308 thread_zombie(td); 309 } 310 311 /* 312 * Reap zombie resources. 313 */ 314 void 315 thread_reap(void) 316 { 317 struct thread *td_first, *td_next; 318 319 /* 320 * Don't even bother to lock if none at this instant, 321 * we really don't care about the next instant.. 322 */ 323 if (!TAILQ_EMPTY(&zombie_threads)) { 324 mtx_lock_spin(&zombie_lock); 325 td_first = TAILQ_FIRST(&zombie_threads); 326 if (td_first) 327 TAILQ_INIT(&zombie_threads); 328 mtx_unlock_spin(&zombie_lock); 329 while (td_first) { 330 td_next = TAILQ_NEXT(td_first, td_slpq); 331 thread_cow_free(td_first); 332 thread_free(td_first); 333 td_first = td_next; 334 } 335 } 336 } 337 338 /* 339 * Allocate a thread. 340 */ 341 struct thread * 342 thread_alloc(int pages) 343 { 344 struct thread *td; 345 346 thread_reap(); /* check if any zombies to get */ 347 348 td = (struct thread *)uma_zalloc(thread_zone, M_WAITOK); 349 KASSERT(td->td_kstack == 0, ("thread_alloc got thread with kstack")); 350 if (!vm_thread_new(td, pages)) { 351 uma_zfree(thread_zone, td); 352 return (NULL); 353 } 354 cpu_thread_alloc(td); 355 vm_domain_policy_init(&td->td_vm_dom_policy); 356 return (td); 357 } 358 359 int 360 thread_alloc_stack(struct thread *td, int pages) 361 { 362 363 KASSERT(td->td_kstack == 0, 364 ("thread_alloc_stack called on a thread with kstack")); 365 if (!vm_thread_new(td, pages)) 366 return (0); 367 cpu_thread_alloc(td); 368 return (1); 369 } 370 371 /* 372 * Deallocate a thread. 373 */ 374 void 375 thread_free(struct thread *td) 376 { 377 378 lock_profile_thread_exit(td); 379 if (td->td_cpuset) 380 cpuset_rel(td->td_cpuset); 381 td->td_cpuset = NULL; 382 cpu_thread_free(td); 383 if (td->td_kstack != 0) 384 vm_thread_dispose(td); 385 vm_domain_policy_cleanup(&td->td_vm_dom_policy); 386 uma_zfree(thread_zone, td); 387 } 388 389 void 390 thread_cow_get_proc(struct thread *newtd, struct proc *p) 391 { 392 393 PROC_LOCK_ASSERT(p, MA_OWNED); 394 newtd->td_ucred = crhold(p->p_ucred); 395 newtd->td_limit = lim_hold(p->p_limit); 396 newtd->td_cowgen = p->p_cowgen; 397 } 398 399 void 400 thread_cow_get(struct thread *newtd, struct thread *td) 401 { 402 403 newtd->td_ucred = crhold(td->td_ucred); 404 newtd->td_limit = lim_hold(td->td_limit); 405 newtd->td_cowgen = td->td_cowgen; 406 } 407 408 void 409 thread_cow_free(struct thread *td) 410 { 411 412 if (td->td_ucred != NULL) 413 crfree(td->td_ucred); 414 if (td->td_limit != NULL) 415 lim_free(td->td_limit); 416 } 417 418 void 419 thread_cow_update(struct thread *td) 420 { 421 struct proc *p; 422 struct ucred *oldcred; 423 struct plimit *oldlimit; 424 425 p = td->td_proc; 426 oldcred = NULL; 427 oldlimit = NULL; 428 PROC_LOCK(p); 429 if (td->td_ucred != p->p_ucred) { 430 oldcred = td->td_ucred; 431 td->td_ucred = crhold(p->p_ucred); 432 } 433 if (td->td_limit != p->p_limit) { 434 oldlimit = td->td_limit; 435 td->td_limit = lim_hold(p->p_limit); 436 } 437 td->td_cowgen = p->p_cowgen; 438 PROC_UNLOCK(p); 439 if (oldcred != NULL) 440 crfree(oldcred); 441 if (oldlimit != NULL) 442 lim_free(oldlimit); 443 } 444 445 /* 446 * Discard the current thread and exit from its context. 447 * Always called with scheduler locked. 448 * 449 * Because we can't free a thread while we're operating under its context, 450 * push the current thread into our CPU's deadthread holder. This means 451 * we needn't worry about someone else grabbing our context before we 452 * do a cpu_throw(). 453 */ 454 void 455 thread_exit(void) 456 { 457 uint64_t runtime, new_switchtime; 458 struct thread *td; 459 struct thread *td2; 460 struct proc *p; 461 int wakeup_swapper; 462 463 td = curthread; 464 p = td->td_proc; 465 466 PROC_SLOCK_ASSERT(p, MA_OWNED); 467 mtx_assert(&Giant, MA_NOTOWNED); 468 469 PROC_LOCK_ASSERT(p, MA_OWNED); 470 KASSERT(p != NULL, ("thread exiting without a process")); 471 CTR3(KTR_PROC, "thread_exit: thread %p (pid %ld, %s)", td, 472 (long)p->p_pid, td->td_name); 473 KASSERT(TAILQ_EMPTY(&td->td_sigqueue.sq_list), ("signal pending")); 474 475 #ifdef AUDIT 476 AUDIT_SYSCALL_EXIT(0, td); 477 #endif 478 /* 479 * drop FPU & debug register state storage, or any other 480 * architecture specific resources that 481 * would not be on a new untouched process. 482 */ 483 cpu_thread_exit(td); /* XXXSMP */ 484 485 /* 486 * The last thread is left attached to the process 487 * So that the whole bundle gets recycled. Skip 488 * all this stuff if we never had threads. 489 * EXIT clears all sign of other threads when 490 * it goes to single threading, so the last thread always 491 * takes the short path. 492 */ 493 if (p->p_flag & P_HADTHREADS) { 494 if (p->p_numthreads > 1) { 495 atomic_add_int(&td->td_proc->p_exitthreads, 1); 496 thread_unlink(td); 497 td2 = FIRST_THREAD_IN_PROC(p); 498 sched_exit_thread(td2, td); 499 500 /* 501 * The test below is NOT true if we are the 502 * sole exiting thread. P_STOPPED_SINGLE is unset 503 * in exit1() after it is the only survivor. 504 */ 505 if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE) { 506 if (p->p_numthreads == p->p_suspcount) { 507 thread_lock(p->p_singlethread); 508 wakeup_swapper = thread_unsuspend_one( 509 p->p_singlethread, p, false); 510 thread_unlock(p->p_singlethread); 511 if (wakeup_swapper) 512 kick_proc0(); 513 } 514 } 515 516 PCPU_SET(deadthread, td); 517 } else { 518 /* 519 * The last thread is exiting.. but not through exit() 520 */ 521 panic ("thread_exit: Last thread exiting on its own"); 522 } 523 } 524 #ifdef HWPMC_HOOKS 525 /* 526 * If this thread is part of a process that is being tracked by hwpmc(4), 527 * inform the module of the thread's impending exit. 528 */ 529 if (PMC_PROC_IS_USING_PMCS(td->td_proc)) 530 PMC_SWITCH_CONTEXT(td, PMC_FN_CSW_OUT); 531 #endif 532 PROC_UNLOCK(p); 533 PROC_STATLOCK(p); 534 thread_lock(td); 535 PROC_SUNLOCK(p); 536 537 /* Do the same timestamp bookkeeping that mi_switch() would do. */ 538 new_switchtime = cpu_ticks(); 539 runtime = new_switchtime - PCPU_GET(switchtime); 540 td->td_runtime += runtime; 541 td->td_incruntime += runtime; 542 PCPU_SET(switchtime, new_switchtime); 543 PCPU_SET(switchticks, ticks); 544 PCPU_INC(cnt.v_swtch); 545 546 /* Save our resource usage in our process. */ 547 td->td_ru.ru_nvcsw++; 548 ruxagg(p, td); 549 rucollect(&p->p_ru, &td->td_ru); 550 PROC_STATUNLOCK(p); 551 552 td->td_state = TDS_INACTIVE; 553 #ifdef WITNESS 554 witness_thread_exit(td); 555 #endif 556 CTR1(KTR_PROC, "thread_exit: cpu_throw() thread %p", td); 557 sched_throw(td); 558 panic("I'm a teapot!"); 559 /* NOTREACHED */ 560 } 561 562 /* 563 * Do any thread specific cleanups that may be needed in wait() 564 * called with Giant, proc and schedlock not held. 565 */ 566 void 567 thread_wait(struct proc *p) 568 { 569 struct thread *td; 570 571 mtx_assert(&Giant, MA_NOTOWNED); 572 KASSERT(p->p_numthreads == 1, ("multiple threads in thread_wait()")); 573 KASSERT(p->p_exitthreads == 0, ("p_exitthreads leaking")); 574 td = FIRST_THREAD_IN_PROC(p); 575 /* Lock the last thread so we spin until it exits cpu_throw(). */ 576 thread_lock(td); 577 thread_unlock(td); 578 lock_profile_thread_exit(td); 579 cpuset_rel(td->td_cpuset); 580 td->td_cpuset = NULL; 581 cpu_thread_clean(td); 582 thread_cow_free(td); 583 thread_reap(); /* check for zombie threads etc. */ 584 } 585 586 /* 587 * Link a thread to a process. 588 * set up anything that needs to be initialized for it to 589 * be used by the process. 590 */ 591 void 592 thread_link(struct thread *td, struct proc *p) 593 { 594 595 /* 596 * XXX This can't be enabled because it's called for proc0 before 597 * its lock has been created. 598 * PROC_LOCK_ASSERT(p, MA_OWNED); 599 */ 600 td->td_state = TDS_INACTIVE; 601 td->td_proc = p; 602 td->td_flags = TDF_INMEM; 603 604 LIST_INIT(&td->td_contested); 605 LIST_INIT(&td->td_lprof[0]); 606 LIST_INIT(&td->td_lprof[1]); 607 sigqueue_init(&td->td_sigqueue, p); 608 callout_init(&td->td_slpcallout, 1); 609 TAILQ_INSERT_TAIL(&p->p_threads, td, td_plist); 610 p->p_numthreads++; 611 } 612 613 /* 614 * Called from: 615 * thread_exit() 616 */ 617 void 618 thread_unlink(struct thread *td) 619 { 620 struct proc *p = td->td_proc; 621 622 PROC_LOCK_ASSERT(p, MA_OWNED); 623 TAILQ_REMOVE(&p->p_threads, td, td_plist); 624 p->p_numthreads--; 625 /* could clear a few other things here */ 626 /* Must NOT clear links to proc! */ 627 } 628 629 static int 630 calc_remaining(struct proc *p, int mode) 631 { 632 int remaining; 633 634 PROC_LOCK_ASSERT(p, MA_OWNED); 635 PROC_SLOCK_ASSERT(p, MA_OWNED); 636 if (mode == SINGLE_EXIT) 637 remaining = p->p_numthreads; 638 else if (mode == SINGLE_BOUNDARY) 639 remaining = p->p_numthreads - p->p_boundary_count; 640 else if (mode == SINGLE_NO_EXIT || mode == SINGLE_ALLPROC) 641 remaining = p->p_numthreads - p->p_suspcount; 642 else 643 panic("calc_remaining: wrong mode %d", mode); 644 return (remaining); 645 } 646 647 static int 648 remain_for_mode(int mode) 649 { 650 651 return (mode == SINGLE_ALLPROC ? 0 : 1); 652 } 653 654 static int 655 weed_inhib(int mode, struct thread *td2, struct proc *p) 656 { 657 int wakeup_swapper; 658 659 PROC_LOCK_ASSERT(p, MA_OWNED); 660 PROC_SLOCK_ASSERT(p, MA_OWNED); 661 THREAD_LOCK_ASSERT(td2, MA_OWNED); 662 663 wakeup_swapper = 0; 664 switch (mode) { 665 case SINGLE_EXIT: 666 if (TD_IS_SUSPENDED(td2)) 667 wakeup_swapper |= thread_unsuspend_one(td2, p, true); 668 if (TD_ON_SLEEPQ(td2) && (td2->td_flags & TDF_SINTR) != 0) 669 wakeup_swapper |= sleepq_abort(td2, EINTR); 670 break; 671 case SINGLE_BOUNDARY: 672 if (TD_IS_SUSPENDED(td2) && (td2->td_flags & TDF_BOUNDARY) == 0) 673 wakeup_swapper |= thread_unsuspend_one(td2, p, false); 674 if (TD_ON_SLEEPQ(td2) && (td2->td_flags & TDF_SINTR) != 0) 675 wakeup_swapper |= sleepq_abort(td2, ERESTART); 676 break; 677 case SINGLE_NO_EXIT: 678 if (TD_IS_SUSPENDED(td2) && (td2->td_flags & TDF_BOUNDARY) == 0) 679 wakeup_swapper |= thread_unsuspend_one(td2, p, false); 680 if (TD_ON_SLEEPQ(td2) && (td2->td_flags & TDF_SINTR) != 0) 681 wakeup_swapper |= sleepq_abort(td2, ERESTART); 682 break; 683 case SINGLE_ALLPROC: 684 /* 685 * ALLPROC suspend tries to avoid spurious EINTR for 686 * threads sleeping interruptable, by suspending the 687 * thread directly, similarly to sig_suspend_threads(). 688 * Since such sleep is not performed at the user 689 * boundary, TDF_BOUNDARY flag is not set, and TDF_ALLPROCSUSP 690 * is used to avoid immediate un-suspend. 691 */ 692 if (TD_IS_SUSPENDED(td2) && (td2->td_flags & (TDF_BOUNDARY | 693 TDF_ALLPROCSUSP)) == 0) 694 wakeup_swapper |= thread_unsuspend_one(td2, p, false); 695 if (TD_ON_SLEEPQ(td2) && (td2->td_flags & TDF_SINTR) != 0) { 696 if ((td2->td_flags & TDF_SBDRY) == 0) { 697 thread_suspend_one(td2); 698 td2->td_flags |= TDF_ALLPROCSUSP; 699 } else { 700 wakeup_swapper |= sleepq_abort(td2, ERESTART); 701 } 702 } 703 break; 704 } 705 return (wakeup_swapper); 706 } 707 708 /* 709 * Enforce single-threading. 710 * 711 * Returns 1 if the caller must abort (another thread is waiting to 712 * exit the process or similar). Process is locked! 713 * Returns 0 when you are successfully the only thread running. 714 * A process has successfully single threaded in the suspend mode when 715 * There are no threads in user mode. Threads in the kernel must be 716 * allowed to continue until they get to the user boundary. They may even 717 * copy out their return values and data before suspending. They may however be 718 * accelerated in reaching the user boundary as we will wake up 719 * any sleeping threads that are interruptable. (PCATCH). 720 */ 721 int 722 thread_single(struct proc *p, int mode) 723 { 724 struct thread *td; 725 struct thread *td2; 726 int remaining, wakeup_swapper; 727 728 td = curthread; 729 KASSERT(mode == SINGLE_EXIT || mode == SINGLE_BOUNDARY || 730 mode == SINGLE_ALLPROC || mode == SINGLE_NO_EXIT, 731 ("invalid mode %d", mode)); 732 /* 733 * If allowing non-ALLPROC singlethreading for non-curproc 734 * callers, calc_remaining() and remain_for_mode() should be 735 * adjusted to also account for td->td_proc != p. For now 736 * this is not implemented because it is not used. 737 */ 738 KASSERT((mode == SINGLE_ALLPROC && td->td_proc != p) || 739 (mode != SINGLE_ALLPROC && td->td_proc == p), 740 ("mode %d proc %p curproc %p", mode, p, td->td_proc)); 741 mtx_assert(&Giant, MA_NOTOWNED); 742 PROC_LOCK_ASSERT(p, MA_OWNED); 743 744 if ((p->p_flag & P_HADTHREADS) == 0 && mode != SINGLE_ALLPROC) 745 return (0); 746 747 /* Is someone already single threading? */ 748 if (p->p_singlethread != NULL && p->p_singlethread != td) 749 return (1); 750 751 if (mode == SINGLE_EXIT) { 752 p->p_flag |= P_SINGLE_EXIT; 753 p->p_flag &= ~P_SINGLE_BOUNDARY; 754 } else { 755 p->p_flag &= ~P_SINGLE_EXIT; 756 if (mode == SINGLE_BOUNDARY) 757 p->p_flag |= P_SINGLE_BOUNDARY; 758 else 759 p->p_flag &= ~P_SINGLE_BOUNDARY; 760 } 761 if (mode == SINGLE_ALLPROC) 762 p->p_flag |= P_TOTAL_STOP; 763 p->p_flag |= P_STOPPED_SINGLE; 764 PROC_SLOCK(p); 765 p->p_singlethread = td; 766 remaining = calc_remaining(p, mode); 767 while (remaining != remain_for_mode(mode)) { 768 if (P_SHOULDSTOP(p) != P_STOPPED_SINGLE) 769 goto stopme; 770 wakeup_swapper = 0; 771 FOREACH_THREAD_IN_PROC(p, td2) { 772 if (td2 == td) 773 continue; 774 thread_lock(td2); 775 td2->td_flags |= TDF_ASTPENDING | TDF_NEEDSUSPCHK; 776 if (TD_IS_INHIBITED(td2)) { 777 wakeup_swapper |= weed_inhib(mode, td2, p); 778 #ifdef SMP 779 } else if (TD_IS_RUNNING(td2) && td != td2) { 780 forward_signal(td2); 781 #endif 782 } 783 thread_unlock(td2); 784 } 785 if (wakeup_swapper) 786 kick_proc0(); 787 remaining = calc_remaining(p, mode); 788 789 /* 790 * Maybe we suspended some threads.. was it enough? 791 */ 792 if (remaining == remain_for_mode(mode)) 793 break; 794 795 stopme: 796 /* 797 * Wake us up when everyone else has suspended. 798 * In the mean time we suspend as well. 799 */ 800 thread_suspend_switch(td, p); 801 remaining = calc_remaining(p, mode); 802 } 803 if (mode == SINGLE_EXIT) { 804 /* 805 * Convert the process to an unthreaded process. The 806 * SINGLE_EXIT is called by exit1() or execve(), in 807 * both cases other threads must be retired. 808 */ 809 KASSERT(p->p_numthreads == 1, ("Unthreading with >1 threads")); 810 p->p_singlethread = NULL; 811 p->p_flag &= ~(P_STOPPED_SINGLE | P_SINGLE_EXIT | P_HADTHREADS); 812 813 /* 814 * Wait for any remaining threads to exit cpu_throw(). 815 */ 816 while (p->p_exitthreads != 0) { 817 PROC_SUNLOCK(p); 818 PROC_UNLOCK(p); 819 sched_relinquish(td); 820 PROC_LOCK(p); 821 PROC_SLOCK(p); 822 } 823 } else if (mode == SINGLE_BOUNDARY) { 824 /* 825 * Wait until all suspended threads are removed from 826 * the processors. The thread_suspend_check() 827 * increments p_boundary_count while it is still 828 * running, which makes it possible for the execve() 829 * to destroy vmspace while our other threads are 830 * still using the address space. 831 * 832 * We lock the thread, which is only allowed to 833 * succeed after context switch code finished using 834 * the address space. 835 */ 836 FOREACH_THREAD_IN_PROC(p, td2) { 837 if (td2 == td) 838 continue; 839 thread_lock(td2); 840 KASSERT((td2->td_flags & TDF_BOUNDARY) != 0, 841 ("td %p not on boundary", td2)); 842 KASSERT(TD_IS_SUSPENDED(td2), 843 ("td %p is not suspended", td2)); 844 thread_unlock(td2); 845 } 846 } 847 PROC_SUNLOCK(p); 848 return (0); 849 } 850 851 bool 852 thread_suspend_check_needed(void) 853 { 854 struct proc *p; 855 struct thread *td; 856 857 td = curthread; 858 p = td->td_proc; 859 PROC_LOCK_ASSERT(p, MA_OWNED); 860 return (P_SHOULDSTOP(p) || ((p->p_flag & P_TRACED) != 0 && 861 (td->td_dbgflags & TDB_SUSPEND) != 0)); 862 } 863 864 /* 865 * Called in from locations that can safely check to see 866 * whether we have to suspend or at least throttle for a 867 * single-thread event (e.g. fork). 868 * 869 * Such locations include userret(). 870 * If the "return_instead" argument is non zero, the thread must be able to 871 * accept 0 (caller may continue), or 1 (caller must abort) as a result. 872 * 873 * The 'return_instead' argument tells the function if it may do a 874 * thread_exit() or suspend, or whether the caller must abort and back 875 * out instead. 876 * 877 * If the thread that set the single_threading request has set the 878 * P_SINGLE_EXIT bit in the process flags then this call will never return 879 * if 'return_instead' is false, but will exit. 880 * 881 * P_SINGLE_EXIT | return_instead == 0| return_instead != 0 882 *---------------+--------------------+--------------------- 883 * 0 | returns 0 | returns 0 or 1 884 * | when ST ends | immediately 885 *---------------+--------------------+--------------------- 886 * 1 | thread exits | returns 1 887 * | | immediately 888 * 0 = thread_exit() or suspension ok, 889 * other = return error instead of stopping the thread. 890 * 891 * While a full suspension is under effect, even a single threading 892 * thread would be suspended if it made this call (but it shouldn't). 893 * This call should only be made from places where 894 * thread_exit() would be safe as that may be the outcome unless 895 * return_instead is set. 896 */ 897 int 898 thread_suspend_check(int return_instead) 899 { 900 struct thread *td; 901 struct proc *p; 902 int wakeup_swapper; 903 904 td = curthread; 905 p = td->td_proc; 906 mtx_assert(&Giant, MA_NOTOWNED); 907 PROC_LOCK_ASSERT(p, MA_OWNED); 908 while (thread_suspend_check_needed()) { 909 if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE) { 910 KASSERT(p->p_singlethread != NULL, 911 ("singlethread not set")); 912 /* 913 * The only suspension in action is a 914 * single-threading. Single threader need not stop. 915 * XXX Should be safe to access unlocked 916 * as it can only be set to be true by us. 917 */ 918 if (p->p_singlethread == td) 919 return (0); /* Exempt from stopping. */ 920 } 921 if ((p->p_flag & P_SINGLE_EXIT) && return_instead) 922 return (EINTR); 923 924 /* Should we goto user boundary if we didn't come from there? */ 925 if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE && 926 (p->p_flag & P_SINGLE_BOUNDARY) && return_instead) 927 return (ERESTART); 928 929 /* 930 * Ignore suspend requests if they are deferred. 931 */ 932 if ((td->td_flags & TDF_SBDRY) != 0) { 933 KASSERT(return_instead, 934 ("TDF_SBDRY set for unsafe thread_suspend_check")); 935 return (0); 936 } 937 938 /* 939 * If the process is waiting for us to exit, 940 * this thread should just suicide. 941 * Assumes that P_SINGLE_EXIT implies P_STOPPED_SINGLE. 942 */ 943 if ((p->p_flag & P_SINGLE_EXIT) && (p->p_singlethread != td)) { 944 PROC_UNLOCK(p); 945 tidhash_remove(td); 946 947 /* 948 * Allow Linux emulation layer to do some work 949 * before thread suicide. 950 */ 951 if (__predict_false(p->p_sysent->sv_thread_detach != NULL)) 952 (p->p_sysent->sv_thread_detach)(td); 953 954 PROC_LOCK(p); 955 tdsigcleanup(td); 956 umtx_thread_exit(td); 957 PROC_SLOCK(p); 958 thread_stopped(p); 959 thread_exit(); 960 } 961 962 PROC_SLOCK(p); 963 thread_stopped(p); 964 if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE) { 965 if (p->p_numthreads == p->p_suspcount + 1) { 966 thread_lock(p->p_singlethread); 967 wakeup_swapper = thread_unsuspend_one( 968 p->p_singlethread, p, false); 969 thread_unlock(p->p_singlethread); 970 if (wakeup_swapper) 971 kick_proc0(); 972 } 973 } 974 PROC_UNLOCK(p); 975 thread_lock(td); 976 /* 977 * When a thread suspends, it just 978 * gets taken off all queues. 979 */ 980 thread_suspend_one(td); 981 if (return_instead == 0) { 982 p->p_boundary_count++; 983 td->td_flags |= TDF_BOUNDARY; 984 } 985 PROC_SUNLOCK(p); 986 mi_switch(SW_INVOL | SWT_SUSPEND, NULL); 987 thread_unlock(td); 988 PROC_LOCK(p); 989 } 990 return (0); 991 } 992 993 void 994 thread_suspend_switch(struct thread *td, struct proc *p) 995 { 996 997 KASSERT(!TD_IS_SUSPENDED(td), ("already suspended")); 998 PROC_LOCK_ASSERT(p, MA_OWNED); 999 PROC_SLOCK_ASSERT(p, MA_OWNED); 1000 /* 1001 * We implement thread_suspend_one in stages here to avoid 1002 * dropping the proc lock while the thread lock is owned. 1003 */ 1004 if (p == td->td_proc) { 1005 thread_stopped(p); 1006 p->p_suspcount++; 1007 } 1008 PROC_UNLOCK(p); 1009 thread_lock(td); 1010 td->td_flags &= ~TDF_NEEDSUSPCHK; 1011 TD_SET_SUSPENDED(td); 1012 sched_sleep(td, 0); 1013 PROC_SUNLOCK(p); 1014 DROP_GIANT(); 1015 mi_switch(SW_VOL | SWT_SUSPEND, NULL); 1016 thread_unlock(td); 1017 PICKUP_GIANT(); 1018 PROC_LOCK(p); 1019 PROC_SLOCK(p); 1020 } 1021 1022 void 1023 thread_suspend_one(struct thread *td) 1024 { 1025 struct proc *p; 1026 1027 p = td->td_proc; 1028 PROC_SLOCK_ASSERT(p, MA_OWNED); 1029 THREAD_LOCK_ASSERT(td, MA_OWNED); 1030 KASSERT(!TD_IS_SUSPENDED(td), ("already suspended")); 1031 p->p_suspcount++; 1032 td->td_flags &= ~TDF_NEEDSUSPCHK; 1033 TD_SET_SUSPENDED(td); 1034 sched_sleep(td, 0); 1035 } 1036 1037 static int 1038 thread_unsuspend_one(struct thread *td, struct proc *p, bool boundary) 1039 { 1040 1041 THREAD_LOCK_ASSERT(td, MA_OWNED); 1042 KASSERT(TD_IS_SUSPENDED(td), ("Thread not suspended")); 1043 TD_CLR_SUSPENDED(td); 1044 td->td_flags &= ~TDF_ALLPROCSUSP; 1045 if (td->td_proc == p) { 1046 PROC_SLOCK_ASSERT(p, MA_OWNED); 1047 p->p_suspcount--; 1048 if (boundary && (td->td_flags & TDF_BOUNDARY) != 0) { 1049 td->td_flags &= ~TDF_BOUNDARY; 1050 p->p_boundary_count--; 1051 } 1052 } 1053 return (setrunnable(td)); 1054 } 1055 1056 /* 1057 * Allow all threads blocked by single threading to continue running. 1058 */ 1059 void 1060 thread_unsuspend(struct proc *p) 1061 { 1062 struct thread *td; 1063 int wakeup_swapper; 1064 1065 PROC_LOCK_ASSERT(p, MA_OWNED); 1066 PROC_SLOCK_ASSERT(p, MA_OWNED); 1067 wakeup_swapper = 0; 1068 if (!P_SHOULDSTOP(p)) { 1069 FOREACH_THREAD_IN_PROC(p, td) { 1070 thread_lock(td); 1071 if (TD_IS_SUSPENDED(td)) { 1072 wakeup_swapper |= thread_unsuspend_one(td, p, 1073 true); 1074 } 1075 thread_unlock(td); 1076 } 1077 } else if (P_SHOULDSTOP(p) == P_STOPPED_SINGLE && 1078 p->p_numthreads == p->p_suspcount) { 1079 /* 1080 * Stopping everything also did the job for the single 1081 * threading request. Now we've downgraded to single-threaded, 1082 * let it continue. 1083 */ 1084 if (p->p_singlethread->td_proc == p) { 1085 thread_lock(p->p_singlethread); 1086 wakeup_swapper = thread_unsuspend_one( 1087 p->p_singlethread, p, false); 1088 thread_unlock(p->p_singlethread); 1089 } 1090 } 1091 if (wakeup_swapper) 1092 kick_proc0(); 1093 } 1094 1095 /* 1096 * End the single threading mode.. 1097 */ 1098 void 1099 thread_single_end(struct proc *p, int mode) 1100 { 1101 struct thread *td; 1102 int wakeup_swapper; 1103 1104 KASSERT(mode == SINGLE_EXIT || mode == SINGLE_BOUNDARY || 1105 mode == SINGLE_ALLPROC || mode == SINGLE_NO_EXIT, 1106 ("invalid mode %d", mode)); 1107 PROC_LOCK_ASSERT(p, MA_OWNED); 1108 KASSERT((mode == SINGLE_ALLPROC && (p->p_flag & P_TOTAL_STOP) != 0) || 1109 (mode != SINGLE_ALLPROC && (p->p_flag & P_TOTAL_STOP) == 0), 1110 ("mode %d does not match P_TOTAL_STOP", mode)); 1111 KASSERT(mode == SINGLE_ALLPROC || p->p_singlethread == curthread, 1112 ("thread_single_end from other thread %p %p", 1113 curthread, p->p_singlethread)); 1114 KASSERT(mode != SINGLE_BOUNDARY || 1115 (p->p_flag & P_SINGLE_BOUNDARY) != 0, 1116 ("mis-matched SINGLE_BOUNDARY flags %x", p->p_flag)); 1117 p->p_flag &= ~(P_STOPPED_SINGLE | P_SINGLE_EXIT | P_SINGLE_BOUNDARY | 1118 P_TOTAL_STOP); 1119 PROC_SLOCK(p); 1120 p->p_singlethread = NULL; 1121 wakeup_swapper = 0; 1122 /* 1123 * If there are other threads they may now run, 1124 * unless of course there is a blanket 'stop order' 1125 * on the process. The single threader must be allowed 1126 * to continue however as this is a bad place to stop. 1127 */ 1128 if (p->p_numthreads != remain_for_mode(mode) && !P_SHOULDSTOP(p)) { 1129 FOREACH_THREAD_IN_PROC(p, td) { 1130 thread_lock(td); 1131 if (TD_IS_SUSPENDED(td)) { 1132 wakeup_swapper |= thread_unsuspend_one(td, p, 1133 mode == SINGLE_BOUNDARY); 1134 } 1135 thread_unlock(td); 1136 } 1137 } 1138 KASSERT(mode != SINGLE_BOUNDARY || p->p_boundary_count == 0, 1139 ("inconsistent boundary count %d", p->p_boundary_count)); 1140 PROC_SUNLOCK(p); 1141 if (wakeup_swapper) 1142 kick_proc0(); 1143 } 1144 1145 struct thread * 1146 thread_find(struct proc *p, lwpid_t tid) 1147 { 1148 struct thread *td; 1149 1150 PROC_LOCK_ASSERT(p, MA_OWNED); 1151 FOREACH_THREAD_IN_PROC(p, td) { 1152 if (td->td_tid == tid) 1153 break; 1154 } 1155 return (td); 1156 } 1157 1158 /* Locate a thread by number; return with proc lock held. */ 1159 struct thread * 1160 tdfind(lwpid_t tid, pid_t pid) 1161 { 1162 #define RUN_THRESH 16 1163 struct thread *td; 1164 int run = 0; 1165 1166 rw_rlock(&tidhash_lock); 1167 LIST_FOREACH(td, TIDHASH(tid), td_hash) { 1168 if (td->td_tid == tid) { 1169 if (pid != -1 && td->td_proc->p_pid != pid) { 1170 td = NULL; 1171 break; 1172 } 1173 PROC_LOCK(td->td_proc); 1174 if (td->td_proc->p_state == PRS_NEW) { 1175 PROC_UNLOCK(td->td_proc); 1176 td = NULL; 1177 break; 1178 } 1179 if (run > RUN_THRESH) { 1180 if (rw_try_upgrade(&tidhash_lock)) { 1181 LIST_REMOVE(td, td_hash); 1182 LIST_INSERT_HEAD(TIDHASH(td->td_tid), 1183 td, td_hash); 1184 rw_wunlock(&tidhash_lock); 1185 return (td); 1186 } 1187 } 1188 break; 1189 } 1190 run++; 1191 } 1192 rw_runlock(&tidhash_lock); 1193 return (td); 1194 } 1195 1196 void 1197 tidhash_add(struct thread *td) 1198 { 1199 rw_wlock(&tidhash_lock); 1200 LIST_INSERT_HEAD(TIDHASH(td->td_tid), td, td_hash); 1201 rw_wunlock(&tidhash_lock); 1202 } 1203 1204 void 1205 tidhash_remove(struct thread *td) 1206 { 1207 rw_wlock(&tidhash_lock); 1208 LIST_REMOVE(td, td_hash); 1209 rw_wunlock(&tidhash_lock); 1210 } 1211