cpus.c 23.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
/*
 * QEMU System Emulator
 *
 * Copyright (c) 2003-2008 Fabrice Bellard
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */

/* Needed early for CONFIG_BSD etc. */
#include "config-host.h"

#include "monitor.h"
#include "sysemu.h"
#include "gdbstub.h"
#include "dma.h"
#include "kvm.h"

P
Paolo Bonzini 已提交
34
#include "qemu-thread.h"
35
#include "cpus.h"
J
Jan Kiszka 已提交
36 37

#ifndef _WIN32
M
Marcelo Tosatti 已提交
38
#include "compatfd.h"
J
Jan Kiszka 已提交
39
#endif
40

41 42 43 44 45 46
#ifdef SIGRTMIN
#define SIG_IPI (SIGRTMIN+4)
#else
#define SIG_IPI SIGUSR1
#endif

47 48 49 50
#ifdef CONFIG_LINUX

#include <sys/prctl.h>

M
Marcelo Tosatti 已提交
51 52 53 54
#ifndef PR_MCE_KILL
#define PR_MCE_KILL 33
#endif

55 56 57 58 59 60 61 62 63 64
#ifndef PR_MCE_KILL_SET
#define PR_MCE_KILL_SET 1
#endif

#ifndef PR_MCE_KILL_EARLY
#define PR_MCE_KILL_EARLY 1
#endif

#endif /* CONFIG_LINUX */

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
static CPUState *next_cpu;

/***********************************************************/
void hw_error(const char *fmt, ...)
{
    va_list ap;
    CPUState *env;

    va_start(ap, fmt);
    fprintf(stderr, "qemu: hardware error: ");
    vfprintf(stderr, fmt, ap);
    fprintf(stderr, "\n");
    for(env = first_cpu; env != NULL; env = env->next_cpu) {
        fprintf(stderr, "CPU #%d:\n", env->cpu_index);
#ifdef TARGET_I386
        cpu_dump_state(env, stderr, fprintf, X86_DUMP_FPU);
#else
        cpu_dump_state(env, stderr, fprintf, 0);
#endif
    }
    va_end(ap);
    abort();
}

void cpu_synchronize_all_states(void)
{
    CPUState *cpu;

    for (cpu = first_cpu; cpu; cpu = cpu->next_cpu) {
        cpu_synchronize_state(cpu);
    }
}

void cpu_synchronize_all_post_reset(void)
{
    CPUState *cpu;

    for (cpu = first_cpu; cpu; cpu = cpu->next_cpu) {
        cpu_synchronize_post_reset(cpu);
    }
}

void cpu_synchronize_all_post_init(void)
{
    CPUState *cpu;

    for (cpu = first_cpu; cpu; cpu = cpu->next_cpu) {
        cpu_synchronize_post_init(cpu);
    }
}

M
Marcelo Tosatti 已提交
116 117
int cpu_is_stopped(CPUState *env)
{
118
    return !runstate_is_running() || env->stopped;
M
Marcelo Tosatti 已提交
119 120
}

121
static void do_vm_stop(RunState state)
122
{
123
    if (runstate_is_running()) {
124 125
        cpu_disable_ticks();
        pause_all_vcpus();
126
        runstate_set(state);
127
        vm_state_notify(0, state);
128 129
        qemu_aio_flush();
        bdrv_flush_all();
130 131 132 133 134 135
        monitor_protocol_event(QEVENT_STOP, NULL);
    }
}

static int cpu_can_run(CPUState *env)
{
136
    if (env->stop) {
137
        return 0;
138
    }
139
    if (env->stopped || !runstate_is_running()) {
140
        return 0;
141
    }
142 143 144
    return 1;
}

145
static bool cpu_thread_is_idle(CPUState *env)
146
{
147 148 149
    if (env->stop || env->queued_work_first) {
        return false;
    }
150
    if (env->stopped || !runstate_is_running()) {
151 152
        return true;
    }
153 154
    if (!env->halted || qemu_cpu_has_work(env) ||
        (kvm_enabled() && kvm_irqchip_in_kernel())) {
155 156 157
        return false;
    }
    return true;
158 159
}

160
bool all_cpu_threads_idle(void)
161 162 163
{
    CPUState *env;

164 165 166 167 168 169
    for (env = first_cpu; env != NULL; env = env->next_cpu) {
        if (!cpu_thread_is_idle(env)) {
            return false;
        }
    }
    return true;
170 171
}

172
static void cpu_handle_guest_debug(CPUState *env)
173
{
174
    gdb_set_stop_cpu(env);
175
    qemu_system_debug_request();
176
    env->stopped = 1;
177 178
}

179 180 181 182 183 184 185 186
static void cpu_signal(int sig)
{
    if (cpu_single_env) {
        cpu_exit(cpu_single_env);
    }
    exit_request = 1;
}

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
#ifdef CONFIG_LINUX
static void sigbus_reraise(void)
{
    sigset_t set;
    struct sigaction action;

    memset(&action, 0, sizeof(action));
    action.sa_handler = SIG_DFL;
    if (!sigaction(SIGBUS, &action, NULL)) {
        raise(SIGBUS);
        sigemptyset(&set);
        sigaddset(&set, SIGBUS);
        sigprocmask(SIG_UNBLOCK, &set, NULL);
    }
    perror("Failed to re-raise SIGBUS!\n");
    abort();
}

static void sigbus_handler(int n, struct qemu_signalfd_siginfo *siginfo,
                           void *ctx)
{
    if (kvm_on_sigbus(siginfo->ssi_code,
                      (void *)(intptr_t)siginfo->ssi_addr)) {
        sigbus_reraise();
    }
}

static void qemu_init_sigbus(void)
{
    struct sigaction action;

    memset(&action, 0, sizeof(action));
    action.sa_flags = SA_SIGINFO;
    action.sa_sigaction = (void (*)(int, siginfo_t*, void*))sigbus_handler;
    sigaction(SIGBUS, &action, NULL);

    prctl(PR_MCE_KILL, PR_MCE_KILL_SET, PR_MCE_KILL_EARLY, 0, 0);
}

226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262
static void qemu_kvm_eat_signals(CPUState *env)
{
    struct timespec ts = { 0, 0 };
    siginfo_t siginfo;
    sigset_t waitset;
    sigset_t chkset;
    int r;

    sigemptyset(&waitset);
    sigaddset(&waitset, SIG_IPI);
    sigaddset(&waitset, SIGBUS);

    do {
        r = sigtimedwait(&waitset, &siginfo, &ts);
        if (r == -1 && !(errno == EAGAIN || errno == EINTR)) {
            perror("sigtimedwait");
            exit(1);
        }

        switch (r) {
        case SIGBUS:
            if (kvm_on_sigbus_vcpu(env, siginfo.si_code, siginfo.si_addr)) {
                sigbus_reraise();
            }
            break;
        default:
            break;
        }

        r = sigpending(&chkset);
        if (r == -1) {
            perror("sigpending");
            exit(1);
        }
    } while (sigismember(&chkset, SIG_IPI) || sigismember(&chkset, SIGBUS));
}

263 264 265 266 267
#else /* !CONFIG_LINUX */

static void qemu_init_sigbus(void)
{
}
268 269 270 271

static void qemu_kvm_eat_signals(CPUState *env)
{
}
272 273
#endif /* !CONFIG_LINUX */

274 275 276 277 278 279
#ifndef _WIN32
static int io_thread_fd = -1;

static void qemu_event_increment(void)
{
    /* Write 8 bytes to be compatible with eventfd.  */
B
Blue Swirl 已提交
280
    static const uint64_t val = 1;
281 282
    ssize_t ret;

283
    if (io_thread_fd == -1) {
284
        return;
285
    }
286 287 288 289 290 291
    do {
        ret = write(io_thread_fd, &val, sizeof(val));
    } while (ret < 0 && errno == EINTR);

    /* EAGAIN is fine, a read must be pending.  */
    if (ret < 0 && errno != EAGAIN) {
A
Alexandre Raymond 已提交
292
        fprintf(stderr, "qemu_event_increment: write() failed: %s\n",
293 294 295 296 297 298 299
                strerror(errno));
        exit (1);
    }
}

static void qemu_event_read(void *opaque)
{
300
    int fd = (intptr_t)opaque;
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315
    ssize_t len;
    char buffer[512];

    /* Drain the notify pipe.  For eventfd, only 8 bytes will be read.  */
    do {
        len = read(fd, buffer, sizeof(buffer));
    } while ((len == -1 && errno == EINTR) || len == sizeof(buffer));
}

static int qemu_event_init(void)
{
    int err;
    int fds[2];

    err = qemu_eventfd(fds);
316
    if (err == -1) {
317
        return -errno;
318
    }
319
    err = fcntl_setfl(fds[0], O_NONBLOCK);
320
    if (err < 0) {
321
        goto fail;
322
    }
323
    err = fcntl_setfl(fds[1], O_NONBLOCK);
324
    if (err < 0) {
325
        goto fail;
326
    }
327
    qemu_set_fd_handler2(fds[0], NULL, qemu_event_read, NULL,
328
                         (void *)(intptr_t)fds[0]);
329 330 331 332 333 334 335 336 337

    io_thread_fd = fds[1];
    return 0;

fail:
    close(fds[0]);
    close(fds[1]);
    return err;
}
338 339 340 341 342

static void dummy_signal(int sig)
{
}

343 344 345 346 347 348
/* If we have signalfd, we mask out the signals we want to handle and then
 * use signalfd to listen for them.  We rely on whatever the current signal
 * handler is to dispatch the signals when we receive them.
 */
static void sigfd_handler(void *opaque)
{
349
    int fd = (intptr_t)opaque;
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377
    struct qemu_signalfd_siginfo info;
    struct sigaction action;
    ssize_t len;

    while (1) {
        do {
            len = read(fd, &info, sizeof(info));
        } while (len == -1 && errno == EINTR);

        if (len == -1 && errno == EAGAIN) {
            break;
        }

        if (len != sizeof(info)) {
            printf("read from sigfd returned %zd: %m\n", len);
            return;
        }

        sigaction(info.ssi_signo, NULL, &action);
        if ((action.sa_flags & SA_SIGINFO) && action.sa_sigaction) {
            action.sa_sigaction(info.ssi_signo,
                                (siginfo_t *)&info, NULL);
        } else if (action.sa_handler) {
            action.sa_handler(info.ssi_signo);
        }
    }
}

378
static int qemu_signal_init(void)
379 380
{
    int sigfd;
381
    sigset_t set;
382

383 384 385 386 387 388 389 390 391
    /*
     * SIG_IPI must be blocked in the main thread and must not be caught
     * by sigwait() in the signal thread. Otherwise, the cpu thread will
     * not catch it reliably.
     */
    sigemptyset(&set);
    sigaddset(&set, SIG_IPI);
    pthread_sigmask(SIG_BLOCK, &set, NULL);

392 393 394 395
    sigemptyset(&set);
    sigaddset(&set, SIGIO);
    sigaddset(&set, SIGALRM);
    sigaddset(&set, SIGBUS);
396
    pthread_sigmask(SIG_BLOCK, &set, NULL);
397 398

    sigfd = qemu_signalfd(&set);
399 400 401 402 403 404 405 406
    if (sigfd == -1) {
        fprintf(stderr, "failed to create signalfd\n");
        return -errno;
    }

    fcntl_setfl(sigfd, O_NONBLOCK);

    qemu_set_fd_handler2(sigfd, NULL, sigfd_handler, NULL,
407
                         (void *)(intptr_t)sigfd);
408 409 410 411

    return 0;
}

412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
static void qemu_kvm_init_cpu_signals(CPUState *env)
{
    int r;
    sigset_t set;
    struct sigaction sigact;

    memset(&sigact, 0, sizeof(sigact));
    sigact.sa_handler = dummy_signal;
    sigaction(SIG_IPI, &sigact, NULL);

    pthread_sigmask(SIG_BLOCK, NULL, &set);
    sigdelset(&set, SIG_IPI);
    sigdelset(&set, SIGBUS);
    r = kvm_set_signal_mask(env, &set);
    if (r) {
        fprintf(stderr, "kvm_set_signal_mask: %s\n", strerror(-r));
        exit(1);
    }

    sigdelset(&set, SIG_IPI);
    sigdelset(&set, SIGBUS);
    r = kvm_set_signal_mask(env, &set);
    if (r) {
        fprintf(stderr, "kvm_set_signal_mask: %s\n", strerror(-r));
        exit(1);
    }
}

static void qemu_tcg_init_cpu_signals(void)
{
    sigset_t set;
    struct sigaction sigact;

    memset(&sigact, 0, sizeof(sigact));
    sigact.sa_handler = cpu_signal;
    sigaction(SIG_IPI, &sigact, NULL);

    sigemptyset(&set);
    sigaddset(&set, SIG_IPI);
    pthread_sigmask(SIG_UNBLOCK, &set, NULL);
}

454 455
#else /* _WIN32 */

456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480
HANDLE qemu_event_handle;

static void dummy_event_handler(void *opaque)
{
}

static int qemu_event_init(void)
{
    qemu_event_handle = CreateEvent(NULL, FALSE, FALSE, NULL);
    if (!qemu_event_handle) {
        fprintf(stderr, "Failed CreateEvent: %ld\n", GetLastError());
        return -1;
    }
    qemu_add_wait_object(qemu_event_handle, dummy_event_handler, NULL);
    return 0;
}

static void qemu_event_increment(void)
{
    if (!SetEvent(qemu_event_handle)) {
        fprintf(stderr, "qemu_event_increment: SetEvent failed: %ld\n",
                GetLastError());
        exit (1);
    }
}
481

482 483 484 485 486
static int qemu_signal_init(void)
{
    return 0;
}

487 488
static void qemu_kvm_init_cpu_signals(CPUState *env)
{
489 490
    abort();
}
491

492 493
static void qemu_tcg_init_cpu_signals(void)
{
494
}
495
#endif /* _WIN32 */
496

497
QemuMutex qemu_global_mutex;
498 499
static QemuCond qemu_io_proceeded_cond;
static bool iothread_requesting_mutex;
500 501 502 503 504 505 506 507 508 509

static QemuThread io_thread;

static QemuThread *tcg_cpu_thread;
static QemuCond *tcg_halt_cond;

/* cpu creation */
static QemuCond qemu_cpu_cond;
/* system init */
static QemuCond qemu_pause_cond;
M
Marcelo Tosatti 已提交
510
static QemuCond qemu_work_cond;
511 512 513 514 515

int qemu_init_main_loop(void)
{
    int ret;

516
    qemu_init_sigbus();
517

518
    ret = qemu_signal_init();
519
    if (ret) {
M
Marcelo Tosatti 已提交
520
        return ret;
521
    }
M
Marcelo Tosatti 已提交
522 523

    /* Note eventfd must be drained before signalfd handlers run */
524
    ret = qemu_event_init();
525
    if (ret) {
526
        return ret;
527
    }
528

529 530 531
    qemu_cond_init(&qemu_cpu_cond);
    qemu_cond_init(&qemu_pause_cond);
    qemu_cond_init(&qemu_work_cond);
532
    qemu_cond_init(&qemu_io_proceeded_cond);
533 534 535
    qemu_mutex_init(&qemu_global_mutex);
    qemu_mutex_lock(&qemu_global_mutex);

J
Jan Kiszka 已提交
536
    qemu_thread_get_self(&io_thread);
537 538 539 540

    return 0;
}

541 542
void qemu_main_loop_start(void)
{
543
    resume_all_vcpus();
544 545
}

M
Marcelo Tosatti 已提交
546 547 548 549
void run_on_cpu(CPUState *env, void (*func)(void *data), void *data)
{
    struct qemu_work_item wi;

J
Jan Kiszka 已提交
550
    if (qemu_cpu_is_self(env)) {
M
Marcelo Tosatti 已提交
551 552 553 554 555 556
        func(data);
        return;
    }

    wi.func = func;
    wi.data = data;
557
    if (!env->queued_work_first) {
M
Marcelo Tosatti 已提交
558
        env->queued_work_first = &wi;
559
    } else {
M
Marcelo Tosatti 已提交
560
        env->queued_work_last->next = &wi;
561
    }
M
Marcelo Tosatti 已提交
562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
    env->queued_work_last = &wi;
    wi.next = NULL;
    wi.done = false;

    qemu_cpu_kick(env);
    while (!wi.done) {
        CPUState *self_env = cpu_single_env;

        qemu_cond_wait(&qemu_work_cond, &qemu_global_mutex);
        cpu_single_env = self_env;
    }
}

static void flush_queued_work(CPUState *env)
{
    struct qemu_work_item *wi;

579
    if (!env->queued_work_first) {
M
Marcelo Tosatti 已提交
580
        return;
581
    }
M
Marcelo Tosatti 已提交
582 583 584 585 586 587 588 589 590 591

    while ((wi = env->queued_work_first)) {
        env->queued_work_first = wi->next;
        wi->func(wi->data);
        wi->done = true;
    }
    env->queued_work_last = NULL;
    qemu_cond_broadcast(&qemu_work_cond);
}

592 593 594 595 596 597 598
static void qemu_wait_io_event_common(CPUState *env)
{
    if (env->stop) {
        env->stop = 0;
        env->stopped = 1;
        qemu_cond_signal(&qemu_pause_cond);
    }
M
Marcelo Tosatti 已提交
599
    flush_queued_work(env);
600
    env->thread_kicked = false;
601 602
}

603
static void qemu_tcg_wait_io_event(void)
604
{
605 606
    CPUState *env;

607
    while (all_cpu_threads_idle()) {
608 609 610
       /* Start accounting real time to the virtual clock if the CPUs
          are idle.  */
        qemu_clock_warp(vm_clock);
611
        qemu_cond_wait(tcg_halt_cond, &qemu_global_mutex);
612
    }
613

614 615 616
    while (iothread_requesting_mutex) {
        qemu_cond_wait(&qemu_io_proceeded_cond, &qemu_global_mutex);
    }
617 618 619 620

    for (env = first_cpu; env != NULL; env = env->next_cpu) {
        qemu_wait_io_event_common(env);
    }
621 622 623 624
}

static void qemu_kvm_wait_io_event(CPUState *env)
{
625
    while (cpu_thread_is_idle(env)) {
626
        qemu_cond_wait(env->halt_cond, &qemu_global_mutex);
627
    }
628

J
Jan Kiszka 已提交
629
    qemu_kvm_eat_signals(env);
630 631 632
    qemu_wait_io_event_common(env);
}

633
static void *qemu_kvm_cpu_thread_fn(void *arg)
634 635
{
    CPUState *env = arg;
J
Jan Kiszka 已提交
636
    int r;
637

638
    qemu_mutex_lock(&qemu_global_mutex);
J
Jan Kiszka 已提交
639
    qemu_thread_get_self(env->thread);
J
Jan Kiszka 已提交
640
    env->thread_id = qemu_get_thread_id();
641

J
Jan Kiszka 已提交
642 643 644 645 646
    r = kvm_init_vcpu(env);
    if (r < 0) {
        fprintf(stderr, "kvm_init_vcpu failed: %s\n", strerror(-r));
        exit(1);
    }
647

648
    qemu_kvm_init_cpu_signals(env);
649 650 651 652 653 654

    /* signal CPU creation */
    env->created = 1;
    qemu_cond_signal(&qemu_cpu_cond);

    while (1) {
655
        if (cpu_can_run(env)) {
656
            r = kvm_cpu_exec(env);
657
            if (r == EXCP_DEBUG) {
658
                cpu_handle_guest_debug(env);
659
            }
660
        }
661 662 663 664 665 666
        qemu_kvm_wait_io_event(env);
    }

    return NULL;
}

667
static void *qemu_tcg_cpu_thread_fn(void *arg)
668 669 670
{
    CPUState *env = arg;

671
    qemu_tcg_init_cpu_signals();
J
Jan Kiszka 已提交
672
    qemu_thread_get_self(env->thread);
673 674 675

    /* signal CPU creation */
    qemu_mutex_lock(&qemu_global_mutex);
676
    for (env = first_cpu; env != NULL; env = env->next_cpu) {
J
Jan Kiszka 已提交
677
        env->thread_id = qemu_get_thread_id();
678
        env->created = 1;
679
    }
680 681
    qemu_cond_signal(&qemu_cpu_cond);

682 683 684
    /* wait for initial kick-off after machine start */
    while (first_cpu->stopped) {
        qemu_cond_wait(tcg_halt_cond, &qemu_global_mutex);
685
    }
686 687

    while (1) {
688
        cpu_exec_all();
689
        if (use_icount && qemu_next_icount_deadline() <= 0) {
690 691
            qemu_notify_event();
        }
692
        qemu_tcg_wait_io_event();
693 694 695 696 697
    }

    return NULL;
}

P
Paolo Bonzini 已提交
698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716
static void qemu_cpu_kick_thread(CPUState *env)
{
#ifndef _WIN32
    int err;

    err = pthread_kill(env->thread->thread, SIG_IPI);
    if (err) {
        fprintf(stderr, "qemu:%s: %s", __func__, strerror(err));
        exit(1);
    }
#else /* _WIN32 */
    if (!qemu_cpu_is_self(env)) {
        SuspendThread(env->thread->thread);
        cpu_signal(0);
        ResumeThread(env->thread->thread);
    }
#endif
}

717 718 719 720 721
void qemu_cpu_kick(void *_env)
{
    CPUState *env = _env;

    qemu_cond_broadcast(env->halt_cond);
J
Jan Kiszka 已提交
722
    if (kvm_enabled() && !env->thread_kicked) {
P
Paolo Bonzini 已提交
723
        qemu_cpu_kick_thread(env);
724 725
        env->thread_kicked = true;
    }
726 727
}

728
void qemu_cpu_kick_self(void)
729
{
730
#ifndef _WIN32
731
    assert(cpu_single_env);
732

733
    if (!cpu_single_env->thread_kicked) {
P
Paolo Bonzini 已提交
734
        qemu_cpu_kick_thread(cpu_single_env);
735
        cpu_single_env->thread_kicked = true;
736
    }
737 738 739
#else
    abort();
#endif
740 741
}

J
Jan Kiszka 已提交
742
int qemu_cpu_is_self(void *_env)
743 744
{
    CPUState *env = _env;
M
Marcelo Tosatti 已提交
745

J
Jan Kiszka 已提交
746
    return qemu_thread_is_self(env->thread);
747 748 749 750 751 752
}

void qemu_mutex_lock_iothread(void)
{
    if (kvm_enabled()) {
        qemu_mutex_lock(&qemu_global_mutex);
753
    } else {
754
        iothread_requesting_mutex = true;
755
        if (qemu_mutex_trylock(&qemu_global_mutex)) {
P
Paolo Bonzini 已提交
756
            qemu_cpu_kick_thread(first_cpu);
757 758
            qemu_mutex_lock(&qemu_global_mutex);
        }
759 760
        iothread_requesting_mutex = false;
        qemu_cond_broadcast(&qemu_io_proceeded_cond);
761
    }
762 763 764 765 766 767 768 769 770 771 772 773
}

void qemu_mutex_unlock_iothread(void)
{
    qemu_mutex_unlock(&qemu_global_mutex);
}

static int all_vcpus_paused(void)
{
    CPUState *penv = first_cpu;

    while (penv) {
774
        if (!penv->stopped) {
775
            return 0;
776
        }
777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793
        penv = (CPUState *)penv->next_cpu;
    }

    return 1;
}

void pause_all_vcpus(void)
{
    CPUState *penv = first_cpu;

    while (penv) {
        penv->stop = 1;
        qemu_cpu_kick(penv);
        penv = (CPUState *)penv->next_cpu;
    }

    while (!all_vcpus_paused()) {
794
        qemu_cond_wait(&qemu_pause_cond, &qemu_global_mutex);
795 796
        penv = first_cpu;
        while (penv) {
797
            qemu_cpu_kick(penv);
798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814
            penv = (CPUState *)penv->next_cpu;
        }
    }
}

void resume_all_vcpus(void)
{
    CPUState *penv = first_cpu;

    while (penv) {
        penv->stop = 0;
        penv->stopped = 0;
        qemu_cpu_kick(penv);
        penv = (CPUState *)penv->next_cpu;
    }
}

815
static void qemu_tcg_init_vcpu(void *_env)
816 817
{
    CPUState *env = _env;
818

819 820
    /* share a single thread for all cpus with TCG */
    if (!tcg_cpu_thread) {
821 822
        env->thread = g_malloc0(sizeof(QemuThread));
        env->halt_cond = g_malloc0(sizeof(QemuCond));
823
        qemu_cond_init(env->halt_cond);
824
        tcg_halt_cond = env->halt_cond;
825
        qemu_thread_create(env->thread, qemu_tcg_cpu_thread_fn, env);
826
        while (env->created == 0) {
827
            qemu_cond_wait(&qemu_cpu_cond, &qemu_global_mutex);
828
        }
829 830 831 832 833 834 835
        tcg_cpu_thread = env->thread;
    } else {
        env->thread = tcg_cpu_thread;
        env->halt_cond = tcg_halt_cond;
    }
}

836
static void qemu_kvm_start_vcpu(CPUState *env)
837
{
838 839
    env->thread = g_malloc0(sizeof(QemuThread));
    env->halt_cond = g_malloc0(sizeof(QemuCond));
840
    qemu_cond_init(env->halt_cond);
841
    qemu_thread_create(env->thread, qemu_kvm_cpu_thread_fn, env);
842
    while (env->created == 0) {
843
        qemu_cond_wait(&qemu_cpu_cond, &qemu_global_mutex);
844
    }
845 846 847 848 849 850 851 852
}

void qemu_init_vcpu(void *_env)
{
    CPUState *env = _env;

    env->nr_cores = smp_cores;
    env->nr_threads = smp_threads;
853
    env->stopped = 1;
854
    if (kvm_enabled()) {
855
        qemu_kvm_start_vcpu(env);
856
    } else {
857
        qemu_tcg_init_vcpu(env);
858
    }
859 860 861 862 863 864 865
}

void qemu_notify_event(void)
{
    qemu_event_increment();
}

866
void cpu_stop_current(void)
867
{
868
    if (cpu_single_env) {
869
        cpu_single_env->stop = 0;
870 871
        cpu_single_env->stopped = 1;
        cpu_exit(cpu_single_env);
872
        qemu_cond_signal(&qemu_pause_cond);
873
    }
874 875
}

876
void vm_stop(RunState state)
877
{
J
Jan Kiszka 已提交
878
    if (!qemu_thread_is_self(&io_thread)) {
879
        qemu_system_vmstop_request(state);
880 881 882 883
        /*
         * FIXME: should not return to device code in case
         * vm_stop() has been requested.
         */
884
        cpu_stop_current();
885 886
        return;
    }
887
    do_vm_stop(state);
888 889
}

890 891 892 893 894 895 896 897 898 899 900
/* does a state transition even if the VM is already stopped,
   current state is forgotten forever */
void vm_stop_force_state(RunState state)
{
    if (runstate_is_running()) {
        vm_stop(state);
    } else {
        runstate_set(state);
    }
}

901
static int tcg_cpu_exec(CPUState *env)
902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
{
    int ret;
#ifdef CONFIG_PROFILER
    int64_t ti;
#endif

#ifdef CONFIG_PROFILER
    ti = profile_getclock();
#endif
    if (use_icount) {
        int64_t count;
        int decr;
        qemu_icount -= (env->icount_decr.u16.low + env->icount_extra);
        env->icount_decr.u16.low = 0;
        env->icount_extra = 0;
917
        count = qemu_icount_round(qemu_next_icount_deadline());
918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938
        qemu_icount += count;
        decr = (count > 0xffff) ? 0xffff : count;
        count -= decr;
        env->icount_decr.u16.low = decr;
        env->icount_extra = count;
    }
    ret = cpu_exec(env);
#ifdef CONFIG_PROFILER
    qemu_time += profile_getclock() - ti;
#endif
    if (use_icount) {
        /* Fold pending instructions back into the
           instruction counter, and clear the interrupt flag.  */
        qemu_icount -= (env->icount_decr.u16.low
                        + env->icount_extra);
        env->icount_decr.u32 = 0;
        env->icount_extra = 0;
    }
    return ret;
}

939
bool cpu_exec_all(void)
940
{
941 942
    int r;

943 944 945
    /* Account partial waits to the vm_clock.  */
    qemu_clock_warp(vm_clock);

946
    if (next_cpu == NULL) {
947
        next_cpu = first_cpu;
948
    }
J
Jan Kiszka 已提交
949
    for (; next_cpu != NULL && !exit_request; next_cpu = next_cpu->next_cpu) {
950
        CPUState *env = next_cpu;
951 952

        qemu_clock_enable(vm_clock,
953
                          (env->singlestep_enabled & SSTEP_NOTIMER) == 0);
954

955
        if (cpu_can_run(env)) {
956
            if (kvm_enabled()) {
957
                r = kvm_cpu_exec(env);
958
                qemu_kvm_eat_signals(env);
959 960
            } else {
                r = tcg_cpu_exec(env);
961 962
            }
            if (r == EXCP_DEBUG) {
963
                cpu_handle_guest_debug(env);
964 965
                break;
            }
966
        } else if (env->stop || env->stopped) {
967 968 969
            break;
        }
    }
J
Jan Kiszka 已提交
970
    exit_request = 0;
971
    return !all_cpu_threads_idle();
972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002
}

void set_numa_modes(void)
{
    CPUState *env;
    int i;

    for (env = first_cpu; env != NULL; env = env->next_cpu) {
        for (i = 0; i < nb_numa_nodes; i++) {
            if (node_cpumask[i] & (1 << env->cpu_index)) {
                env->numa_node = i;
            }
        }
    }
}

void set_cpu_log(const char *optarg)
{
    int mask;
    const CPULogItem *item;

    mask = cpu_str_to_log_mask(optarg);
    if (!mask) {
        printf("Log items (comma separated):\n");
        for (item = cpu_log_items; item->mask != 0; item++) {
            printf("%-10s %s\n", item->name, item->help);
        }
        exit(1);
    }
    cpu_set_log(mask);
}
B
Blue Swirl 已提交
1003

1004 1005 1006 1007 1008
void set_cpu_log_filename(const char *optarg)
{
    cpu_set_log_filename(optarg);
}

B
Blue Swirl 已提交
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
/* Return the virtual CPU time, based on the instruction counter.  */
int64_t cpu_get_icount(void)
{
    int64_t icount;
    CPUState *env = cpu_single_env;;

    icount = qemu_icount;
    if (env) {
        if (!can_do_io(env)) {
            fprintf(stderr, "Bad clock read\n");
        }
        icount -= (env->icount_decr.u16.low + env->icount_extra);
    }
    return qemu_icount_bias + (icount << icount_time_shift);
}
1024

1025
void list_cpus(FILE *f, fprintf_function cpu_fprintf, const char *optarg)
1026 1027 1028 1029 1030 1031 1032 1033
{
    /* XXX: implement xxx_cpu_list for targets that still miss it */
#if defined(cpu_list_id)
    cpu_list_id(f, cpu_fprintf, optarg);
#elif defined(cpu_list)
    cpu_list(f, cpu_fprintf); /* deprecated */
#endif
}