vl.c 157.5 KB
Newer Older
1
/*
B
bellard 已提交
2
 * QEMU System Emulator
3
 *
B
bellard 已提交
4
 * Copyright (c) 2003-2008 Fabrice Bellard
5
 *
B
bellard 已提交
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
 * 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.
23 24 25 26 27 28
 */
#include <unistd.h>
#include <fcntl.h>
#include <signal.h>
#include <time.h>
#include <errno.h>
B
bellard 已提交
29
#include <sys/time.h>
30
#include <zlib.h>
B
bellard 已提交
31

J
Juan Quintela 已提交
32
/* Needed early for CONFIG_BSD etc. */
B
blueswir1 已提交
33 34
#include "config-host.h"

B
bellard 已提交
35
#ifndef _WIN32
P
Paul Brook 已提交
36
#include <libgen.h>
37
#include <pwd.h>
B
bellard 已提交
38
#include <sys/times.h>
B
bellard 已提交
39
#include <sys/wait.h>
B
bellard 已提交
40 41
#include <termios.h>
#include <sys/mman.h>
B
bellard 已提交
42
#include <sys/ioctl.h>
B
blueswir1 已提交
43
#include <sys/resource.h>
B
bellard 已提交
44
#include <sys/socket.h>
B
bellard 已提交
45
#include <netinet/in.h>
B
blueswir1 已提交
46 47
#include <net/if.h>
#include <arpa/inet.h>
B
bellard 已提交
48
#include <dirent.h>
B
bellard 已提交
49
#include <netdb.h>
50
#include <sys/select.h>
J
Juan Quintela 已提交
51
#ifdef CONFIG_BSD
B
bellard 已提交
52
#include <sys/stat.h>
A
Aurelien Jarno 已提交
53
#if defined(__FreeBSD__) || defined(__FreeBSD_kernel__) || defined(__DragonFly__)
B
bellard 已提交
54
#include <libutil.h>
B
blueswir1 已提交
55 56
#else
#include <util.h>
57
#endif
58
#else
59
#ifdef __linux__
B
bellard 已提交
60 61
#include <pty.h>
#include <malloc.h>
62
#include <linux/rtc.h>
A
Andi Kleen 已提交
63
#include <sys/prctl.h>
T
ths 已提交
64 65 66 67 68 69

/* For the benefit of older linux systems which don't supply it,
   we use a local copy of hpet.h. */
/* #include <linux/hpet.h> */
#include "hpet.h"

B
bellard 已提交
70
#include <linux/ppdev.h>
71
#include <linux/parport.h>
72 73
#endif
#ifdef __sun__
74 75 76 77 78 79 80 81 82 83 84 85 86
#include <sys/stat.h>
#include <sys/ethernet.h>
#include <sys/sockio.h>
#include <netinet/arp.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_icmp.h> // must come after ip.h
#include <netinet/udp.h>
#include <netinet/tcp.h>
#include <net/if.h>
#include <syslog.h>
#include <stropts.h>
87 88 89
/* See MySQL bug #7156 (http://bugs.mysql.com/bug.php?id=7156) for
   discussion about Solaris header problems */
extern int madvise(caddr_t, size_t, int);
B
bellard 已提交
90
#endif
B
bellard 已提交
91
#endif
B
bellard 已提交
92
#endif
B
bellard 已提交
93

94 95 96 97
#if defined(__OpenBSD__)
#include <util.h>
#endif

98 99 100 101
#if defined(CONFIG_VDE)
#include <libvdeplug.h>
#endif

B
bellard 已提交
102
#ifdef _WIN32
103
#include <windows.h>
T
ths 已提交
104
#include <mmsystem.h>
B
bellard 已提交
105 106
#endif

B
bellard 已提交
107
#ifdef CONFIG_SDL
S
Stefan Weil 已提交
108
#if defined(__APPLE__) || defined(main)
S
Stefan Weil 已提交
109
#include <SDL.h>
M
malc 已提交
110 111 112
int qemu_main(int argc, char **argv, char **envp);
int main(int argc, char **argv)
{
S
Stefan Weil 已提交
113
    return qemu_main(argc, argv, NULL);
M
malc 已提交
114 115 116
}
#undef main
#define main qemu_main
B
bellard 已提交
117
#endif
B
bellard 已提交
118
#endif /* CONFIG_SDL */
119

120 121 122 123 124
#ifdef CONFIG_COCOA
#undef main
#define main qemu_main
#endif /* CONFIG_COCOA */

125 126 127 128 129 130 131 132 133
#include "hw/hw.h"
#include "hw/boards.h"
#include "hw/usb.h"
#include "hw/pcmcia.h"
#include "hw/pc.h"
#include "hw/audiodev.h"
#include "hw/isa.h"
#include "hw/baum.h"
#include "hw/bt.h"
R
Richard W.M. Jones 已提交
134
#include "hw/watchdog.h"
135
#include "hw/smbios.h"
136
#include "hw/xen.h"
137
#include "hw/qdev.h"
138
#include "hw/loader.h"
A
aurel32 已提交
139
#include "bt-host.h"
140
#include "net.h"
141
#include "net/slirp.h"
142 143 144 145 146 147 148 149
#include "monitor.h"
#include "console.h"
#include "sysemu.h"
#include "gdbstub.h"
#include "qemu-timer.h"
#include "qemu-char.h"
#include "cache-utils.h"
#include "block.h"
L
lirans@il.ibm.com 已提交
150 151
#include "block_int.h"
#include "block-migration.h"
B
blueswir1 已提交
152
#include "dma.h"
153 154 155 156
#include "audio/audio.h"
#include "migration.h"
#include "kvm.h"
#include "balloon.h"
K
Kevin Wolf 已提交
157
#include "qemu-option.h"
G
Gerd Hoffmann 已提交
158
#include "qemu-config.h"
159
#include "qemu-objects.h"
160

161
#include "disas.h"
B
bellard 已提交
162

163
#include "exec-all.h"
164

165 166
#include "qemu_socket.h"

J
Jan Kiszka 已提交
167
#include "slirp/libslirp.h"
168

B
Blue Swirl 已提交
169 170
#include "qemu-queue.h"

171 172
//#define DEBUG_NET
//#define DEBUG_SLIRP
173

174
#define DEFAULT_RAM_SIZE 128
175

176 177
#define MAX_VIRTIO_CONSOLES 1

P
Paul Brook 已提交
178
static const char *data_dir;
179
const char *bios_name = NULL;
T
ths 已提交
180
/* Note: drives_table[MAX_DRIVES] is a dummy block driver if none available
B
bellard 已提交
181
   to store the VM snapshots */
B
Blue Swirl 已提交
182 183
struct drivelist drives = QTAILQ_HEAD_INITIALIZER(drives);
struct driveoptlist driveopts = QTAILQ_HEAD_INITIALIZER(driveopts);
M
malc 已提交
184
enum vga_retrace_method vga_retrace_method = VGA_RETRACE_DUMB;
185
DisplayType display_type = DT_DEFAULT;
186
const char* keyboard_layout = NULL;
A
Anthony Liguori 已提交
187
ram_addr_t ram_size;
188 189 190 191
const char *mem_path = NULL;
#ifdef MAP_POPULATE
int mem_prealloc = 0; /* force preallocation of physical target memory */
#endif
192
int nb_nics;
B
bellard 已提交
193
NICInfo nd_table[MAX_NICS];
194
int vm_running;
P
Paolo Bonzini 已提交
195
int autostart;
196 197
static int rtc_utc = 1;
static int rtc_date_offset = -1; /* -1 means no change */
J
Jan Kiszka 已提交
198
QEMUClock *rtc_clock;
G
Gerd Hoffmann 已提交
199
int vga_interface_type = VGA_NONE;
200 201 202
#ifdef TARGET_SPARC
int graphic_width = 1024;
int graphic_height = 768;
B
blueswir1 已提交
203
int graphic_depth = 8;
204
#else
205 206
int graphic_width = 800;
int graphic_height = 600;
207
int graphic_depth = 15;
B
blueswir1 已提交
208
#endif
B
blueswir1 已提交
209
static int full_screen = 0;
210
#ifdef CONFIG_SDL
B
blueswir1 已提交
211
static int no_frame = 0;
212
#endif
T
ths 已提交
213
int no_quit = 0;
214
CharDriverState *serial_hds[MAX_SERIAL_PORTS];
215
CharDriverState *parallel_hds[MAX_PARALLEL_PORTS];
216
CharDriverState *virtcon_hds[MAX_VIRTIO_CONSOLES];
217 218
#ifdef TARGET_I386
int win2k_install_hack = 0;
219
int rtc_td_hack = 0;
220
#endif
B
bellard 已提交
221
int usb_enabled = 0;
222
int singlestep = 0;
B
bellard 已提交
223
int smp_cpus = 1;
224
int max_cpus = 0;
225 226
int smp_cores = 1;
int smp_threads = 1;
227
const char *vnc_display;
B
bellard 已提交
228
int acpi_enabled = 1;
A
aliguori 已提交
229
int no_hpet = 0;
B
bellard 已提交
230
int fd_bootchk = 1;
B
bellard 已提交
231
int no_reboot = 0;
A
aurel32 已提交
232
int no_shutdown = 0;
233
int cursor_hide = 1;
234
int graphic_rotate = 0;
J
Jes Sorensen 已提交
235
uint8_t irq0override = 1;
B
blueswir1 已提交
236
#ifndef _WIN32
T
ths 已提交
237
int daemonize = 0;
B
blueswir1 已提交
238
#endif
M
Markus Armbruster 已提交
239
const char *watchdog;
240 241
const char *option_rom[MAX_OPTION_ROMS];
int nb_option_roms;
242
int semihosting_enabled = 0;
243 244 245
#ifdef TARGET_ARM
int old_param = 0;
#endif
T
ths 已提交
246
const char *qemu_name;
T
ths 已提交
247
int alt_grab = 0;
248
int ctrl_grab = 0;
249
#if defined(TARGET_SPARC) || defined(TARGET_PPC)
B
blueswir1 已提交
250 251 252
unsigned int nb_prom_envs = 0;
const char *prom_envs[MAX_PROM_ENVS];
#endif
253
int boot_menu;
254

255 256 257 258
int nb_numa_nodes;
uint64_t node_mem[MAX_NODES];
uint64_t node_cpumask[MAX_NODES];

259 260
static CPUState *cur_cpu;
static CPUState *next_cpu;
T
ths 已提交
261
/* Conversion factor from emulated instructions to virtual clock ticks.  */
P
pbrook 已提交
262
static int icount_time_shift;
T
ths 已提交
263
/* Arbitrarily pick 1MIPS as the minimum allowable speed.  */
P
pbrook 已提交
264 265 266
#define MAX_ICOUNT_SHIFT 10
/* Compensate for varying guest execution speed.  */
static int64_t qemu_icount_bias;
B
blueswir1 已提交
267 268
static QEMUTimer *icount_rt_timer;
static QEMUTimer *icount_vm_timer;
B
blueswir1 已提交
269
static QEMUTimer *nographic_timer;
270

271 272
uint8_t qemu_uuid[16];

J
Jan Kiszka 已提交
273 274 275
static QEMUBootSetHandler *boot_set_handler;
static void *boot_set_opaque;

276 277 278 279 280 281
#ifdef SIGRTMIN
#define SIG_IPI (SIGRTMIN+4)
#else
#define SIG_IPI SIGUSR1
#endif

282
static int default_serial = 1;
G
Gerd Hoffmann 已提交
283
static int default_parallel = 1;
284
static int default_virtcon = 1;
G
Gerd Hoffmann 已提交
285
static int default_monitor = 1;
G
Gerd Hoffmann 已提交
286
static int default_vga = 1;
G
Gerd Hoffmann 已提交
287 288 289
static int default_floppy = 1;
static int default_cdrom = 1;
static int default_sdcard = 1;
290 291 292 293 294

static struct {
    const char *driver;
    int *flag;
} default_list[] = {
G
Gerd Hoffmann 已提交
295 296
    { .driver = "isa-serial",           .flag = &default_serial    },
    { .driver = "isa-parallel",         .flag = &default_parallel  },
G
Gerd Hoffmann 已提交
297 298
    { .driver = "isa-fdc",              .flag = &default_floppy    },
    { .driver = "ide-drive",            .flag = &default_cdrom     },
299 300 301
    { .driver = "virtio-serial-pci",    .flag = &default_virtcon   },
    { .driver = "virtio-serial-s390",   .flag = &default_virtcon   },
    { .driver = "virtio-serial",        .flag = &default_virtcon   },
G
Gerd Hoffmann 已提交
302
    { .driver = "VGA",                  .flag = &default_vga       },
G
Gerd Hoffmann 已提交
303 304
    { .driver = "cirrus-vga",           .flag = &default_vga       },
    { .driver = "vmware-svga",          .flag = &default_vga       },
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321
};

static int default_driver_check(QemuOpts *opts, void *opaque)
{
    const char *driver = qemu_opt_get(opts, "driver");
    int i;

    if (!driver)
        return 0;
    for (i = 0; i < ARRAY_SIZE(default_list); i++) {
        if (strcmp(default_list[i].driver, driver) != 0)
            continue;
        *(default_list[i].flag) = 0;
    }
    return 0;
}

322
/***********************************************************/
323 324
/* x86 ISA bus support */

A
Anthony Liguori 已提交
325
target_phys_addr_t isa_mem_base = 0;
B
bellard 已提交
326
PicState2 *isa_pic;
327 328 329 330 331

/***********************************************************/
void hw_error(const char *fmt, ...)
{
    va_list ap;
B
bellard 已提交
332
    CPUState *env;
333 334 335 336 337

    va_start(ap, fmt);
    fprintf(stderr, "qemu: hardware error: ");
    vfprintf(stderr, fmt, ap);
    fprintf(stderr, "\n");
B
bellard 已提交
338 339
    for(env = first_cpu; env != NULL; env = env->next_cpu) {
        fprintf(stderr, "CPU #%d:\n", env->cpu_index);
340
#ifdef TARGET_I386
B
bellard 已提交
341
        cpu_dump_state(env, stderr, fprintf, X86_DUMP_FPU);
342
#else
B
bellard 已提交
343
        cpu_dump_state(env, stderr, fprintf, 0);
344
#endif
B
bellard 已提交
345
    }
346 347 348
    va_end(ap);
    abort();
}
A
Andi Kleen 已提交
349 350 351

static void set_proc_name(const char *s)
{
352
#if defined(__linux__) && defined(PR_SET_NAME)
A
Andi Kleen 已提交
353 354 355 356 357 358 359 360 361 362
    char name[16];
    if (!s)
        return;
    name[sizeof(name) - 1] = 0;
    strncpy(name, s, sizeof(name));
    /* Could rewrite argv[0] too, but that's a bit more complicated.
       This simple way is enough for `top'. */
    prctl(PR_SET_NAME, name);
#endif    	
}
A
aliguori 已提交
363 364 365 366 367 368 369 370 371 372 373 374 375
 
/***************/
/* ballooning */

static QEMUBalloonEvent *qemu_balloon_event;
void *qemu_balloon_event_opaque;

void qemu_add_balloon_handler(QEMUBalloonEvent *func, void *opaque)
{
    qemu_balloon_event = func;
    qemu_balloon_event_opaque = opaque;
}

376
int qemu_balloon(ram_addr_t target, MonitorCompletion cb, void *opaque)
A
aliguori 已提交
377
{
378 379 380 381 382 383
    if (qemu_balloon_event) {
        qemu_balloon_event(qemu_balloon_event_opaque, target, cb, opaque);
        return 1;
    } else {
        return 0;
    }
A
aliguori 已提交
384 385
}

386
int qemu_balloon_status(MonitorCompletion cb, void *opaque)
A
aliguori 已提交
387
{
388 389 390 391 392 393
    if (qemu_balloon_event) {
        qemu_balloon_event(qemu_balloon_event_opaque, 0, cb, opaque);
        return 1;
    } else {
        return 0;
    }
A
aliguori 已提交
394
}
395

396

397 398
/***********************************************************/
/* real time host monotonic timer */
399

400 401
/* compute with 96 bit intermediate result: (a*b)/c */
uint64_t muldiv64(uint64_t a, uint32_t b, uint32_t c)
402
{
403 404 405
    union {
        uint64_t ll;
        struct {
406
#ifdef HOST_WORDS_BIGENDIAN
407 408 409
            uint32_t high, low;
#else
            uint32_t low, high;
410
#endif
411 412 413
        } l;
    } u, res;
    uint64_t rl, rh;
414

415 416 417 418 419 420 421
    u.ll = a;
    rl = (uint64_t)u.l.low * (uint64_t)b;
    rh = (uint64_t)u.l.high * (uint64_t)b;
    rh += (rl >> 32);
    res.l.high = rh / c;
    res.l.low = (((rh % c) << 32) + (rl & 0xffffffff)) / c;
    return res.ll;
422 423
}

J
Jan Kiszka 已提交
424 425 426 427 428 429 430 431
static int64_t get_clock_realtime(void)
{
    struct timeval tv;

    gettimeofday(&tv, NULL);
    return tv.tv_sec * 1000000000LL + (tv.tv_usec * 1000);
}

432
#ifdef WIN32
433

434
static int64_t clock_freq;
B
bellard 已提交
435

436
static void init_get_clock(void)
B
bellard 已提交
437
{
438 439
    LARGE_INTEGER freq;
    int ret;
440 441 442 443 444 445
    ret = QueryPerformanceFrequency(&freq);
    if (ret == 0) {
        fprintf(stderr, "Could not calibrate ticks\n");
        exit(1);
    }
    clock_freq = freq.QuadPart;
B
bellard 已提交
446 447
}

448
static int64_t get_clock(void)
B
bellard 已提交
449
{
450 451
    LARGE_INTEGER ti;
    QueryPerformanceCounter(&ti);
452
    return muldiv64(ti.QuadPart, get_ticks_per_sec(), clock_freq);
B
bellard 已提交
453 454
}

455
#else
B
bellard 已提交
456

457 458 459
static int use_rt_clock;

static void init_get_clock(void)
B
bellard 已提交
460
{
461
    use_rt_clock = 0;
462
#if defined(__linux__) || (defined(__FreeBSD__) && __FreeBSD_version >= 500000) \
A
Aurelien Jarno 已提交
463
    || defined(__DragonFly__) || defined(__FreeBSD_kernel__)
464 465 466 467 468 469 470
    {
        struct timespec ts;
        if (clock_gettime(CLOCK_MONOTONIC, &ts) == 0) {
            use_rt_clock = 1;
        }
    }
#endif
B
bellard 已提交
471 472
}

473
static int64_t get_clock(void)
B
bellard 已提交
474
{
475
#if defined(__linux__) || (defined(__FreeBSD__) && __FreeBSD_version >= 500000) \
A
Aurelien Jarno 已提交
476
	|| defined(__DragonFly__) || defined(__FreeBSD_kernel__)
477 478 479 480
    if (use_rt_clock) {
        struct timespec ts;
        clock_gettime(CLOCK_MONOTONIC, &ts);
        return ts.tv_sec * 1000000000LL + ts.tv_nsec;
481
    } else
B
bellard 已提交
482
#endif
483 484 485
    {
        /* XXX: using gettimeofday leads to problems if the date
           changes, so it should be avoided. */
J
Jan Kiszka 已提交
486
        return get_clock_realtime();
487
    }
B
bellard 已提交
488
}
489 490
#endif

P
pbrook 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503 504
/* Return the virtual CPU time, based on the instruction counter.  */
static 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);
}

505 506 507
/***********************************************************/
/* guest cycle counter */

508 509 510 511 512
typedef struct TimersState {
    int64_t cpu_ticks_prev;
    int64_t cpu_ticks_offset;
    int64_t cpu_clock_offset;
    int32_t cpu_ticks_enabled;
513
    int64_t dummy;
514 515 516
} TimersState;

TimersState timers_state;
517

518 519
/* return the host CPU cycle counter and handle stop/restart */
int64_t cpu_get_ticks(void)
520
{
P
pbrook 已提交
521 522 523
    if (use_icount) {
        return cpu_get_icount();
    }
524 525
    if (!timers_state.cpu_ticks_enabled) {
        return timers_state.cpu_ticks_offset;
526
    } else {
527 528
        int64_t ticks;
        ticks = cpu_get_real_ticks();
529
        if (timers_state.cpu_ticks_prev > ticks) {
530 531
            /* Note: non increasing ticks may happen if the host uses
               software suspend */
532
            timers_state.cpu_ticks_offset += timers_state.cpu_ticks_prev - ticks;
533
        }
534 535
        timers_state.cpu_ticks_prev = ticks;
        return ticks + timers_state.cpu_ticks_offset;
536
    }
537 538
}

539 540 541 542
/* return the host CPU monotonic timer and handle stop/restart */
static int64_t cpu_get_clock(void)
{
    int64_t ti;
543 544
    if (!timers_state.cpu_ticks_enabled) {
        return timers_state.cpu_clock_offset;
545 546
    } else {
        ti = get_clock();
547
        return ti + timers_state.cpu_clock_offset;
548 549 550
    }
}

551 552 553
/* enable cpu_get_ticks() */
void cpu_enable_ticks(void)
{
554 555 556 557
    if (!timers_state.cpu_ticks_enabled) {
        timers_state.cpu_ticks_offset -= cpu_get_real_ticks();
        timers_state.cpu_clock_offset -= get_clock();
        timers_state.cpu_ticks_enabled = 1;
558
    }
559 560 561 562 563 564
}

/* disable cpu_get_ticks() : the clock is stopped. You must not call
   cpu_get_ticks() after that.  */
void cpu_disable_ticks(void)
{
565 566 567 568
    if (timers_state.cpu_ticks_enabled) {
        timers_state.cpu_ticks_offset = cpu_get_ticks();
        timers_state.cpu_clock_offset = cpu_get_clock();
        timers_state.cpu_ticks_enabled = 0;
569
    }
570 571
}

572 573
/***********************************************************/
/* timers */
574

J
Jan Kiszka 已提交
575 576
#define QEMU_CLOCK_REALTIME 0
#define QEMU_CLOCK_VIRTUAL  1
J
Jan Kiszka 已提交
577
#define QEMU_CLOCK_HOST     2
578 579 580

struct QEMUClock {
    int type;
581
    int enabled;
582 583 584 585 586 587 588 589 590 591 592
    /* XXX: add frequency */
};

struct QEMUTimer {
    QEMUClock *clock;
    int64_t expire_time;
    QEMUTimerCB *cb;
    void *opaque;
    struct QEMUTimer *next;
};

593 594 595 596
struct qemu_alarm_timer {
    char const *name;
    int (*start)(struct qemu_alarm_timer *t);
    void (*stop)(struct qemu_alarm_timer *t);
597
    void (*rearm)(struct qemu_alarm_timer *t);
598 599
    void *priv;

P
Paolo Bonzini 已提交
600 601
    char expired;
    char pending;
602
};
603 604 605

static inline int alarm_has_dynticks(struct qemu_alarm_timer *t)
{
P
Paolo Bonzini 已提交
606
    return !!t->rearm;
607 608 609 610 611 612 613 614 615 616 617 618 619
}

static void qemu_rearm_alarm_timer(struct qemu_alarm_timer *t)
{
    if (!alarm_has_dynticks(t))
        return;

    t->rearm(t);
}

/* TODO: MIN_TIMER_REARM_US should be optimized */
#define MIN_TIMER_REARM_US 250

620
static struct qemu_alarm_timer *alarm_timer;
621

B
bellard 已提交
622
#ifdef _WIN32
623 624 625 626

struct qemu_alarm_win32 {
    MMRESULT timerId;
    unsigned int period;
627
} alarm_win32_data = {0, 0};
628 629 630

static int win32_start_timer(struct qemu_alarm_timer *t);
static void win32_stop_timer(struct qemu_alarm_timer *t);
631
static void win32_rearm_timer(struct qemu_alarm_timer *t);
632

B
bellard 已提交
633
#else
634 635 636 637

static int unix_start_timer(struct qemu_alarm_timer *t);
static void unix_stop_timer(struct qemu_alarm_timer *t);

638 639
#ifdef __linux__

640 641 642 643
static int dynticks_start_timer(struct qemu_alarm_timer *t);
static void dynticks_stop_timer(struct qemu_alarm_timer *t);
static void dynticks_rearm_timer(struct qemu_alarm_timer *t);

T
ths 已提交
644 645 646
static int hpet_start_timer(struct qemu_alarm_timer *t);
static void hpet_stop_timer(struct qemu_alarm_timer *t);

647 648 649
static int rtc_start_timer(struct qemu_alarm_timer *t);
static void rtc_stop_timer(struct qemu_alarm_timer *t);

650
#endif /* __linux__ */
651

652 653
#endif /* _WIN32 */

P
pbrook 已提交
654
/* Correlation between real and virtual time is always going to be
T
ths 已提交
655
   fairly approximate, so ignore small variation.
P
pbrook 已提交
656 657
   When the guest is idle real and virtual time will be aligned in
   the IO wait loop.  */
658
#define ICOUNT_WOBBLE (get_ticks_per_sec() / 10)
P
pbrook 已提交
659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699

static void icount_adjust(void)
{
    int64_t cur_time;
    int64_t cur_icount;
    int64_t delta;
    static int64_t last_delta;
    /* If the VM is not running, then do nothing.  */
    if (!vm_running)
        return;

    cur_time = cpu_get_clock();
    cur_icount = qemu_get_clock(vm_clock);
    delta = cur_icount - cur_time;
    /* FIXME: This is a very crude algorithm, somewhat prone to oscillation.  */
    if (delta > 0
        && last_delta + ICOUNT_WOBBLE < delta * 2
        && icount_time_shift > 0) {
        /* The guest is getting too far ahead.  Slow time down.  */
        icount_time_shift--;
    }
    if (delta < 0
        && last_delta - ICOUNT_WOBBLE > delta * 2
        && icount_time_shift < MAX_ICOUNT_SHIFT) {
        /* The guest is getting too far behind.  Speed time up.  */
        icount_time_shift++;
    }
    last_delta = delta;
    qemu_icount_bias = cur_icount - (qemu_icount << icount_time_shift);
}

static void icount_adjust_rt(void * opaque)
{
    qemu_mod_timer(icount_rt_timer,
                   qemu_get_clock(rt_clock) + 1000);
    icount_adjust();
}

static void icount_adjust_vm(void * opaque)
{
    qemu_mod_timer(icount_vm_timer,
700
                   qemu_get_clock(vm_clock) + get_ticks_per_sec() / 10);
P
pbrook 已提交
701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
    icount_adjust();
}

static void init_icount_adjust(void)
{
    /* Have both realtime and virtual time triggers for speed adjustment.
       The realtime trigger catches emulated time passing too slowly,
       the virtual time trigger catches emulated time passing too fast.
       Realtime triggers occur even when idle, so use them less frequently
       than VM triggers.  */
    icount_rt_timer = qemu_new_timer(rt_clock, icount_adjust_rt, NULL);
    qemu_mod_timer(icount_rt_timer,
                   qemu_get_clock(rt_clock) + 1000);
    icount_vm_timer = qemu_new_timer(vm_clock, icount_adjust_vm, NULL);
    qemu_mod_timer(icount_vm_timer,
716
                   qemu_get_clock(vm_clock) + get_ticks_per_sec() / 10);
P
pbrook 已提交
717 718
}

719
static struct qemu_alarm_timer alarm_timers[] = {
720
#ifndef _WIN32
721
#ifdef __linux__
722
    {"dynticks", dynticks_start_timer,
723
     dynticks_stop_timer, dynticks_rearm_timer, NULL},
T
ths 已提交
724
    /* HPET - if available - is preferred */
725
    {"hpet", hpet_start_timer, hpet_stop_timer, NULL, NULL},
T
ths 已提交
726
    /* ...otherwise try RTC */
727
    {"rtc", rtc_start_timer, rtc_stop_timer, NULL, NULL},
728
#endif
729
    {"unix", unix_start_timer, unix_stop_timer, NULL, NULL},
730
#else
731
    {"dynticks", win32_start_timer,
732
     win32_stop_timer, win32_rearm_timer, &alarm_win32_data},
733
    {"win32", win32_start_timer,
734
     win32_stop_timer, NULL, &alarm_win32_data},
735 736 737 738
#endif
    {NULL, }
};

739
static void show_available_alarms(void)
740 741 742 743 744 745 746 747 748 749 750 751
{
    int i;

    printf("Available alarm timers, in order of precedence:\n");
    for (i = 0; alarm_timers[i].name; i++)
        printf("%s\n", alarm_timers[i].name);
}

static void configure_alarms(char const *opt)
{
    int i;
    int cur = 0;
752
    int count = ARRAY_SIZE(alarm_timers) - 1;
753 754
    char *arg;
    char *name;
P
pbrook 已提交
755
    struct qemu_alarm_timer tmp;
756

757
    if (!strcmp(opt, "?")) {
758 759 760 761
        show_available_alarms();
        exit(0);
    }

762
    arg = qemu_strdup(opt);
763 764 765 766

    /* Reorder the array */
    name = strtok(arg, ",");
    while (name) {
767
        for (i = 0; i < count && alarm_timers[i].name; i++) {
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790
            if (!strcmp(alarm_timers[i].name, name))
                break;
        }

        if (i == count) {
            fprintf(stderr, "Unknown clock %s\n", name);
            goto next;
        }

        if (i < cur)
            /* Ignore */
            goto next;

	/* Swap */
        tmp = alarm_timers[i];
        alarm_timers[i] = alarm_timers[cur];
        alarm_timers[cur] = tmp;

        cur++;
next:
        name = strtok(NULL, ",");
    }

791
    qemu_free(arg);
792 793

    if (cur) {
P
pbrook 已提交
794
        /* Disable remaining timers */
795 796
        for (i = cur; i < count; i++)
            alarm_timers[i].name = NULL;
797 798 799
    } else {
        show_available_alarms();
        exit(1);
800 801 802
    }
}

J
Jan Kiszka 已提交
803 804
#define QEMU_NUM_CLOCKS 3

805 806
QEMUClock *rt_clock;
QEMUClock *vm_clock;
J
Jan Kiszka 已提交
807
QEMUClock *host_clock;
808

J
Jan Kiszka 已提交
809
static QEMUTimer *active_timers[QEMU_NUM_CLOCKS];
810

811
static QEMUClock *qemu_new_clock(int type)
812 813 814 815
{
    QEMUClock *clock;
    clock = qemu_mallocz(sizeof(QEMUClock));
    clock->type = type;
816
    clock->enabled = 1;
817 818 819
    return clock;
}

820 821 822 823 824
static void qemu_clock_enable(QEMUClock *clock, int enabled)
{
    clock->enabled = enabled;
}

825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
QEMUTimer *qemu_new_timer(QEMUClock *clock, QEMUTimerCB *cb, void *opaque)
{
    QEMUTimer *ts;

    ts = qemu_mallocz(sizeof(QEMUTimer));
    ts->clock = clock;
    ts->cb = cb;
    ts->opaque = opaque;
    return ts;
}

void qemu_free_timer(QEMUTimer *ts)
{
    qemu_free(ts);
}

/* stop a timer, but do not dealloc it */
void qemu_del_timer(QEMUTimer *ts)
{
    QEMUTimer **pt, *t;

    /* NOTE: this code must be signal safe because
       qemu_timer_expired() can be called from a signal. */
    pt = &active_timers[ts->clock->type];
    for(;;) {
        t = *pt;
        if (!t)
            break;
        if (t == ts) {
            *pt = t->next;
            break;
        }
        pt = &t->next;
    }
}

/* modify the current timer so that it will be fired when current_time
   >= expire_time. The corresponding callback will be called. */
void qemu_mod_timer(QEMUTimer *ts, int64_t expire_time)
{
    QEMUTimer **pt, *t;

    qemu_del_timer(ts);

    /* add the timer in the sorted list */
    /* NOTE: this code must be signal safe because
       qemu_timer_expired() can be called from a signal. */
    pt = &active_timers[ts->clock->type];
    for(;;) {
        t = *pt;
        if (!t)
            break;
877
        if (t->expire_time > expire_time)
878 879 880 881 882 883
            break;
        pt = &t->next;
    }
    ts->expire_time = expire_time;
    ts->next = *pt;
    *pt = ts;
884 885

    /* Rearm if necessary  */
P
pbrook 已提交
886
    if (pt == &active_timers[ts->clock->type]) {
P
Paolo Bonzini 已提交
887
        if (!alarm_timer->pending) {
P
pbrook 已提交
888 889 890
            qemu_rearm_alarm_timer(alarm_timer);
        }
        /* Interrupt execution to force deadline recalculation.  */
891 892
        if (use_icount)
            qemu_notify_event();
P
pbrook 已提交
893
    }
894 895 896 897 898 899 900 901 902 903 904 905
}

int qemu_timer_pending(QEMUTimer *ts)
{
    QEMUTimer *t;
    for(t = active_timers[ts->clock->type]; t != NULL; t = t->next) {
        if (t == ts)
            return 1;
    }
    return 0;
}

S
Stefano Stabellini 已提交
906
int qemu_timer_expired(QEMUTimer *timer_head, int64_t current_time)
907 908 909 910 911 912
{
    if (!timer_head)
        return 0;
    return (timer_head->expire_time <= current_time);
}

P
Paolo Bonzini 已提交
913
static void qemu_run_timers(QEMUClock *clock)
914
{
P
Paolo Bonzini 已提交
915 916
    QEMUTimer **ptimer_head, *ts;
    int64_t current_time;
917 918 919
   
    if (!clock->enabled)
        return;
920

P
Paolo Bonzini 已提交
921 922
    current_time = qemu_get_clock (clock);
    ptimer_head = &active_timers[clock->type];
923 924
    for(;;) {
        ts = *ptimer_head;
925
        if (!ts || ts->expire_time > current_time)
926 927 928 929
            break;
        /* remove timer from the list before calling the callback */
        *ptimer_head = ts->next;
        ts->next = NULL;
930

931 932 933 934 935 936 937 938
        /* run the callback (the timer list can be modified) */
        ts->cb(ts->opaque);
    }
}

int64_t qemu_get_clock(QEMUClock *clock)
{
    switch(clock->type) {
J
Jan Kiszka 已提交
939
    case QEMU_CLOCK_REALTIME:
940
        return get_clock() / 1000000;
941
    default:
J
Jan Kiszka 已提交
942
    case QEMU_CLOCK_VIRTUAL:
P
pbrook 已提交
943 944 945 946 947
        if (use_icount) {
            return cpu_get_icount();
        } else {
            return cpu_get_clock();
        }
J
Jan Kiszka 已提交
948 949
    case QEMU_CLOCK_HOST:
        return get_clock_realtime();
950 951 952
    }
}

P
Paolo Bonzini 已提交
953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969
int64_t qemu_get_clock_ns(QEMUClock *clock)
{
    switch(clock->type) {
    case QEMU_CLOCK_REALTIME:
        return get_clock();
    default:
    case QEMU_CLOCK_VIRTUAL:
        if (use_icount) {
            return cpu_get_icount();
        } else {
            return cpu_get_clock();
        }
    case QEMU_CLOCK_HOST:
        return get_clock_realtime();
    }
}

J
Jan Kiszka 已提交
970
static void init_clocks(void)
971 972
{
    init_get_clock();
J
Jan Kiszka 已提交
973 974
    rt_clock = qemu_new_clock(QEMU_CLOCK_REALTIME);
    vm_clock = qemu_new_clock(QEMU_CLOCK_VIRTUAL);
J
Jan Kiszka 已提交
975
    host_clock = qemu_new_clock(QEMU_CLOCK_HOST);
J
Jan Kiszka 已提交
976 977

    rtc_clock = host_clock;
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 1003 1004
/* save a timer */
void qemu_put_timer(QEMUFile *f, QEMUTimer *ts)
{
    uint64_t expire_time;

    if (qemu_timer_pending(ts)) {
        expire_time = ts->expire_time;
    } else {
        expire_time = -1;
    }
    qemu_put_be64(f, expire_time);
}

void qemu_get_timer(QEMUFile *f, QEMUTimer *ts)
{
    uint64_t expire_time;

    expire_time = qemu_get_be64(f);
    if (expire_time != -1) {
        qemu_mod_timer(ts, expire_time);
    } else {
        qemu_del_timer(ts);
    }
}

J
Juan Quintela 已提交
1005 1006 1007 1008 1009 1010 1011
static const VMStateDescription vmstate_timers = {
    .name = "timer",
    .version_id = 2,
    .minimum_version_id = 1,
    .minimum_version_id_old = 1,
    .fields      = (VMStateField []) {
        VMSTATE_INT64(cpu_ticks_offset, TimersState),
1012
        VMSTATE_INT64(dummy, TimersState),
J
Juan Quintela 已提交
1013 1014
        VMSTATE_INT64_V(cpu_clock_offset, TimersState, 2),
        VMSTATE_END_OF_LIST()
1015
    }
J
Juan Quintela 已提交
1016
};
1017

1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029
static void qemu_run_all_timers(void)
{
    /* rearm timer, if not periodic */
    if (alarm_timer->expired) {
        alarm_timer->expired = 0;
        qemu_rearm_alarm_timer(alarm_timer);
    }

    alarm_timer->pending = 0;

    /* vm time timers */
    if (vm_running) {
1030
        qemu_run_timers(vm_clock);
1031 1032
    }

P
Paolo Bonzini 已提交
1033 1034
    qemu_run_timers(rt_clock);
    qemu_run_timers(host_clock);
1035
}
1036

B
bellard 已提交
1037
#ifdef _WIN32
B
blueswir1 已提交
1038 1039 1040
static void CALLBACK host_alarm_handler(UINT uTimerID, UINT uMsg,
                                        DWORD_PTR dwUser, DWORD_PTR dw1,
                                        DWORD_PTR dw2)
B
bellard 已提交
1041
#else
1042
static void host_alarm_handler(int host_signum)
B
bellard 已提交
1043
#endif
1044
{
P
Paolo Bonzini 已提交
1045 1046 1047 1048
    struct qemu_alarm_timer *t = alarm_timer;
    if (!t)
	return;

1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063
#if 0
#define DISP_FREQ 1000
    {
        static int64_t delta_min = INT64_MAX;
        static int64_t delta_max, delta_cum, last_clock, delta, ti;
        static int count;
        ti = qemu_get_clock(vm_clock);
        if (last_clock != 0) {
            delta = ti - last_clock;
            if (delta < delta_min)
                delta_min = delta;
            if (delta > delta_max)
                delta_max = delta;
            delta_cum += delta;
            if (++count == DISP_FREQ) {
B
bellard 已提交
1064
                printf("timer: min=%" PRId64 " us max=%" PRId64 " us avg=%" PRId64 " us avg_freq=%0.3f Hz\n",
1065 1066 1067 1068
                       muldiv64(delta_min, 1000000, get_ticks_per_sec()),
                       muldiv64(delta_max, 1000000, get_ticks_per_sec()),
                       muldiv64(delta_cum, 1000000 / DISP_FREQ, get_ticks_per_sec()),
                       (double)get_ticks_per_sec() / ((double)delta_cum / DISP_FREQ));
1069 1070 1071 1072 1073 1074 1075 1076 1077
                count = 0;
                delta_min = INT64_MAX;
                delta_max = 0;
                delta_cum = 0;
            }
        }
        last_clock = ti;
    }
#endif
P
Paolo Bonzini 已提交
1078
    if (alarm_has_dynticks(t) ||
P
pbrook 已提交
1079
        (!use_icount &&
J
Jan Kiszka 已提交
1080
            qemu_timer_expired(active_timers[QEMU_CLOCK_VIRTUAL],
P
pbrook 已提交
1081
                               qemu_get_clock(vm_clock))) ||
J
Jan Kiszka 已提交
1082
        qemu_timer_expired(active_timers[QEMU_CLOCK_REALTIME],
J
Jan Kiszka 已提交
1083 1084 1085
                           qemu_get_clock(rt_clock)) ||
        qemu_timer_expired(active_timers[QEMU_CLOCK_HOST],
                           qemu_get_clock(host_clock))) {
1086

P
Paolo Bonzini 已提交
1087 1088
        t->expired = alarm_has_dynticks(t);
        t->pending = 1;
1089
        qemu_notify_event();
1090 1091 1092
    }
}

P
pbrook 已提交
1093
static int64_t qemu_next_deadline(void)
1094
{
J
Jan Kiszka 已提交
1095 1096
    /* To avoid problems with overflow limit this to 2^32.  */
    int64_t delta = INT32_MAX;
1097

J
Jan Kiszka 已提交
1098 1099
    if (active_timers[QEMU_CLOCK_VIRTUAL]) {
        delta = active_timers[QEMU_CLOCK_VIRTUAL]->expire_time -
P
pbrook 已提交
1100
                     qemu_get_clock(vm_clock);
J
Jan Kiszka 已提交
1101 1102 1103 1104 1105 1106
    }
    if (active_timers[QEMU_CLOCK_HOST]) {
        int64_t hdelta = active_timers[QEMU_CLOCK_HOST]->expire_time -
                 qemu_get_clock(host_clock);
        if (hdelta < delta)
            delta = hdelta;
1107 1108
    }

P
pbrook 已提交
1109 1110
    if (delta < 0)
        delta = 0;
1111

P
pbrook 已提交
1112 1113 1114
    return delta;
}

J
Jan Kiszka 已提交
1115
#if defined(__linux__)
P
pbrook 已提交
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125
static uint64_t qemu_next_deadline_dyntick(void)
{
    int64_t delta;
    int64_t rtdelta;

    if (use_icount)
        delta = INT32_MAX;
    else
        delta = (qemu_next_deadline() + 999) / 1000;

J
Jan Kiszka 已提交
1126 1127
    if (active_timers[QEMU_CLOCK_REALTIME]) {
        rtdelta = (active_timers[QEMU_CLOCK_REALTIME]->expire_time -
P
pbrook 已提交
1128 1129 1130 1131 1132 1133 1134 1135 1136
                 qemu_get_clock(rt_clock))*1000;
        if (rtdelta < delta)
            delta = rtdelta;
    }

    if (delta < MIN_TIMER_REARM_US)
        delta = MIN_TIMER_REARM_US;

    return delta;
1137
}
1138
#endif
1139

1140 1141
#ifndef _WIN32

1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156
/* Sets a specific flag */
static int fcntl_setfl(int fd, int flag)
{
    int flags;

    flags = fcntl(fd, F_GETFL);
    if (flags == -1)
        return -errno;

    if (fcntl(fd, F_SETFL, flags | flag) == -1)
        return -errno;

    return 0;
}

B
BSD fix  
bellard 已提交
1157 1158
#if defined(__linux__)

1159 1160
#define RTC_FREQ 1024

1161
static void enable_sigio_timer(int fd)
1162 1163 1164 1165 1166 1167 1168 1169 1170
{
    struct sigaction act;

    /* timer signal */
    sigfillset(&act.sa_mask);
    act.sa_flags = 0;
    act.sa_handler = host_alarm_handler;

    sigaction(SIGIO, &act, NULL);
1171
    fcntl_setfl(fd, O_ASYNC);
1172 1173
    fcntl(fd, F_SETOWN, getpid());
}
B
BSD fix  
bellard 已提交
1174

T
ths 已提交
1175 1176 1177 1178 1179
static int hpet_start_timer(struct qemu_alarm_timer *t)
{
    struct hpet_info info;
    int r, fd;

K
Kevin Wolf 已提交
1180
    fd = qemu_open("/dev/hpet", O_RDONLY);
T
ths 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208
    if (fd < 0)
        return -1;

    /* Set frequency */
    r = ioctl(fd, HPET_IRQFREQ, RTC_FREQ);
    if (r < 0) {
        fprintf(stderr, "Could not configure '/dev/hpet' to have a 1024Hz timer. This is not a fatal\n"
                "error, but for better emulation accuracy type:\n"
                "'echo 1024 > /proc/sys/dev/hpet/max-user-freq' as root.\n");
        goto fail;
    }

    /* Check capabilities */
    r = ioctl(fd, HPET_INFO, &info);
    if (r < 0)
        goto fail;

    /* Enable periodic mode */
    r = ioctl(fd, HPET_EPI, 0);
    if (info.hi_flags && (r < 0))
        goto fail;

    /* Enable interrupt */
    r = ioctl(fd, HPET_IE_ON, 0);
    if (r < 0)
        goto fail;

    enable_sigio_timer(fd);
1209
    t->priv = (void *)(long)fd;
T
ths 已提交
1210 1211 1212 1213 1214 1215 1216 1217 1218

    return 0;
fail:
    close(fd);
    return -1;
}

static void hpet_stop_timer(struct qemu_alarm_timer *t)
{
1219
    int fd = (long)t->priv;
T
ths 已提交
1220 1221 1222 1223

    close(fd);
}

1224
static int rtc_start_timer(struct qemu_alarm_timer *t)
1225
{
1226
    int rtc_fd;
1227
    unsigned long current_rtc_freq = 0;
1228

K
Kevin Wolf 已提交
1229
    TFR(rtc_fd = qemu_open("/dev/rtc", O_RDONLY));
1230 1231
    if (rtc_fd < 0)
        return -1;
1232 1233 1234
    ioctl(rtc_fd, RTC_IRQP_READ, &current_rtc_freq);
    if (current_rtc_freq != RTC_FREQ &&
        ioctl(rtc_fd, RTC_IRQP_SET, RTC_FREQ) < 0) {
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
        fprintf(stderr, "Could not configure '/dev/rtc' to have a 1024 Hz timer. This is not a fatal\n"
                "error, but for better emulation accuracy either use a 2.6 host Linux kernel or\n"
                "type 'echo 1024 > /proc/sys/dev/rtc/max-user-freq' as root.\n");
        goto fail;
    }
    if (ioctl(rtc_fd, RTC_PIE_ON, 0) < 0) {
    fail:
        close(rtc_fd);
        return -1;
    }
1245 1246 1247

    enable_sigio_timer(rtc_fd);

1248
    t->priv = (void *)(long)rtc_fd;
1249

1250 1251 1252
    return 0;
}

1253
static void rtc_stop_timer(struct qemu_alarm_timer *t)
B
BSD fix  
bellard 已提交
1254
{
1255
    int rtc_fd = (long)t->priv;
1256 1257

    close(rtc_fd);
B
BSD fix  
bellard 已提交
1258 1259
}

1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271
static int dynticks_start_timer(struct qemu_alarm_timer *t)
{
    struct sigevent ev;
    timer_t host_timer;
    struct sigaction act;

    sigfillset(&act.sa_mask);
    act.sa_flags = 0;
    act.sa_handler = host_alarm_handler;

    sigaction(SIGALRM, &act, NULL);

1272 1273 1274 1275 1276
    /* 
     * Initialize ev struct to 0 to avoid valgrind complaining
     * about uninitialized data in timer_create call
     */
    memset(&ev, 0, sizeof(ev));
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289
    ev.sigev_value.sival_int = 0;
    ev.sigev_notify = SIGEV_SIGNAL;
    ev.sigev_signo = SIGALRM;

    if (timer_create(CLOCK_REALTIME, &ev, &host_timer)) {
        perror("timer_create");

        /* disable dynticks */
        fprintf(stderr, "Dynamic Ticks disabled\n");

        return -1;
    }

1290
    t->priv = (void *)(long)host_timer;
1291 1292 1293 1294 1295 1296

    return 0;
}

static void dynticks_stop_timer(struct qemu_alarm_timer *t)
{
1297
    timer_t host_timer = (timer_t)(long)t->priv;
1298 1299 1300 1301 1302 1303

    timer_delete(host_timer);
}

static void dynticks_rearm_timer(struct qemu_alarm_timer *t)
{
1304
    timer_t host_timer = (timer_t)(long)t->priv;
1305 1306 1307 1308
    struct itimerspec timeout;
    int64_t nearest_delta_us = INT64_MAX;
    int64_t current_us;

1309
    assert(alarm_has_dynticks(t));
J
Jan Kiszka 已提交
1310
    if (!active_timers[QEMU_CLOCK_REALTIME] &&
J
Jan Kiszka 已提交
1311 1312
        !active_timers[QEMU_CLOCK_VIRTUAL] &&
        !active_timers[QEMU_CLOCK_HOST])
1313
        return;
1314

P
pbrook 已提交
1315
    nearest_delta_us = qemu_next_deadline_dyntick();
1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337

    /* check whether a timer is already running */
    if (timer_gettime(host_timer, &timeout)) {
        perror("gettime");
        fprintf(stderr, "Internal timer error: aborting\n");
        exit(1);
    }
    current_us = timeout.it_value.tv_sec * 1000000 + timeout.it_value.tv_nsec/1000;
    if (current_us && current_us <= nearest_delta_us)
        return;

    timeout.it_interval.tv_sec = 0;
    timeout.it_interval.tv_nsec = 0; /* 0 for one-shot timer */
    timeout.it_value.tv_sec =  nearest_delta_us / 1000000;
    timeout.it_value.tv_nsec = (nearest_delta_us % 1000000) * 1000;
    if (timer_settime(host_timer, 0 /* RELATIVE */, &timeout, NULL)) {
        perror("settime");
        fprintf(stderr, "Internal timer error: aborting\n");
        exit(1);
    }
}

T
ths 已提交
1338
#endif /* defined(__linux__) */
1339

1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373
static int unix_start_timer(struct qemu_alarm_timer *t)
{
    struct sigaction act;
    struct itimerval itv;
    int err;

    /* timer signal */
    sigfillset(&act.sa_mask);
    act.sa_flags = 0;
    act.sa_handler = host_alarm_handler;

    sigaction(SIGALRM, &act, NULL);

    itv.it_interval.tv_sec = 0;
    /* for i386 kernel 2.6 to get 1 ms */
    itv.it_interval.tv_usec = 999;
    itv.it_value.tv_sec = 0;
    itv.it_value.tv_usec = 10 * 1000;

    err = setitimer(ITIMER_REAL, &itv, NULL);
    if (err)
        return -1;

    return 0;
}

static void unix_stop_timer(struct qemu_alarm_timer *t)
{
    struct itimerval itv;

    memset(&itv, 0, sizeof(itv));
    setitimer(ITIMER_REAL, &itv, NULL);
}

B
BSD fix  
bellard 已提交
1374
#endif /* !defined(_WIN32) */
1375

1376

1377 1378 1379 1380 1381 1382
#ifdef _WIN32

static int win32_start_timer(struct qemu_alarm_timer *t)
{
    TIMECAPS tc;
    struct qemu_alarm_win32 *data = t->priv;
1383
    UINT flags;
1384 1385 1386 1387

    memset(&tc, 0, sizeof(tc));
    timeGetDevCaps(&tc, sizeof(tc));

1388
    data->period = tc.wPeriodMin;
1389 1390
    timeBeginPeriod(data->period);

1391 1392 1393 1394 1395 1396
    flags = TIME_CALLBACK_FUNCTION;
    if (alarm_has_dynticks(t))
        flags |= TIME_ONESHOT;
    else
        flags |= TIME_PERIODIC;

1397 1398 1399 1400
    data->timerId = timeSetEvent(1,         // interval (ms)
                        data->period,       // resolution
                        host_alarm_handler, // function
                        (DWORD)t,           // parameter
1401
                        flags);
1402 1403

    if (!data->timerId) {
B
Blue Swirl 已提交
1404
        fprintf(stderr, "Failed to initialize win32 alarm timer: %ld\n",
1405
                GetLastError());
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
        timeEndPeriod(data->period);
        return -1;
    }

    return 0;
}

static void win32_stop_timer(struct qemu_alarm_timer *t)
{
    struct qemu_alarm_win32 *data = t->priv;

    timeKillEvent(data->timerId);
    timeEndPeriod(data->period);
}

1421 1422 1423 1424
static void win32_rearm_timer(struct qemu_alarm_timer *t)
{
    struct qemu_alarm_win32 *data = t->priv;

1425
    assert(alarm_has_dynticks(t));
J
Jan Kiszka 已提交
1426
    if (!active_timers[QEMU_CLOCK_REALTIME] &&
J
Jan Kiszka 已提交
1427 1428
        !active_timers[QEMU_CLOCK_VIRTUAL] &&
        !active_timers[QEMU_CLOCK_HOST])
1429
        return;
1430 1431 1432 1433 1434 1435 1436

    timeKillEvent(data->timerId);

    data->timerId = timeSetEvent(1,
                        data->period,
                        host_alarm_handler,
                        (DWORD)t,
P
Paolo Bonzini 已提交
1437
                        TIME_ONESHOT | TIME_CALLBACK_FUNCTION);
1438 1439

    if (!data->timerId) {
B
Blue Swirl 已提交
1440
        fprintf(stderr, "Failed to re-arm win32 alarm timer %ld\n",
1441
                GetLastError());
1442 1443 1444 1445 1446 1447

        timeEndPeriod(data->period);
        exit(1);
    }
}

1448 1449
#endif /* _WIN32 */

1450 1451 1452 1453 1454 1455
static void alarm_timer_on_change_state_rearm(void *opaque, int running, int reason)
{
    if (running)
        qemu_rearm_alarm_timer((struct qemu_alarm_timer *) opaque);
}

1456
static int init_timer_alarm(void)
1457
{
1458
    struct qemu_alarm_timer *t = NULL;
1459
    int i, err = -1;
1460

1461 1462 1463 1464 1465 1466
    for (i = 0; alarm_timers[i].name; i++) {
        t = &alarm_timers[i];

        err = t->start(t);
        if (!err)
            break;
B
bellard 已提交
1467
    }
1468

1469
    if (err) {
1470 1471
        err = -ENOENT;
        goto fail;
B
bellard 已提交
1472
    }
1473

P
Paolo Bonzini 已提交
1474 1475
    /* first event is at time 0 */
    t->pending = 1;
1476
    alarm_timer = t;
1477
    qemu_add_vm_change_state_handler(alarm_timer_on_change_state_rearm, t);
1478

1479
    return 0;
1480 1481 1482

fail:
    return err;
1483 1484
}

1485
static void quit_timers(void)
B
bellard 已提交
1486
{
P
Paolo Bonzini 已提交
1487
    struct qemu_alarm_timer *t = alarm_timer;
1488
    alarm_timer = NULL;
P
Paolo Bonzini 已提交
1489
    t->stop(t);
B
bellard 已提交
1490 1491
}

1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528
/***********************************************************/
/* host time/date access */
void qemu_get_timedate(struct tm *tm, int offset)
{
    time_t ti;
    struct tm *ret;

    time(&ti);
    ti += offset;
    if (rtc_date_offset == -1) {
        if (rtc_utc)
            ret = gmtime(&ti);
        else
            ret = localtime(&ti);
    } else {
        ti -= rtc_date_offset;
        ret = gmtime(&ti);
    }

    memcpy(tm, ret, sizeof(struct tm));
}

int qemu_timedate_diff(struct tm *tm)
{
    time_t seconds;

    if (rtc_date_offset == -1)
        if (rtc_utc)
            seconds = mktimegm(tm);
        else
            seconds = mktime(tm);
    else
        seconds = mktimegm(tm) + rtc_date_offset;

    return seconds - time(NULL);
}

1529 1530 1531 1532 1533 1534 1535 1536 1537
void rtc_change_mon_event(struct tm *tm)
{
    QObject *data;

    data = qobject_from_jsonf("{ 'offset': %d }", qemu_timedate_diff(tm));
    monitor_protocol_event(QEVENT_RTC_CHANGE, data);
    qobject_decref(data);
}

J
Jan Kiszka 已提交
1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
static void configure_rtc_date_offset(const char *startdate, int legacy)
{
    time_t rtc_start_date;
    struct tm tm;

    if (!strcmp(startdate, "now") && legacy) {
        rtc_date_offset = -1;
    } else {
        if (sscanf(startdate, "%d-%d-%dT%d:%d:%d",
                   &tm.tm_year,
                   &tm.tm_mon,
                   &tm.tm_mday,
                   &tm.tm_hour,
                   &tm.tm_min,
                   &tm.tm_sec) == 6) {
            /* OK */
        } else if (sscanf(startdate, "%d-%d-%d",
                          &tm.tm_year,
                          &tm.tm_mon,
                          &tm.tm_mday) == 3) {
            tm.tm_hour = 0;
            tm.tm_min = 0;
            tm.tm_sec = 0;
        } else {
            goto date_fail;
        }
        tm.tm_year -= 1900;
        tm.tm_mon--;
        rtc_start_date = mktimegm(&tm);
        if (rtc_start_date == -1) {
        date_fail:
            fprintf(stderr, "Invalid date format. Valid formats are:\n"
                            "'2006-06-17T16:01:21' or '2006-06-17'\n");
            exit(1);
        }
        rtc_date_offset = time(NULL) - rtc_start_date;
    }
}

static void configure_rtc(QemuOpts *opts)
{
    const char *value;

    value = qemu_opt_get(opts, "base");
    if (value) {
        if (!strcmp(value, "utc")) {
            rtc_utc = 1;
        } else if (!strcmp(value, "localtime")) {
            rtc_utc = 0;
        } else {
            configure_rtc_date_offset(value, 0);
        }
    }
J
Jan Kiszka 已提交
1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601
    value = qemu_opt_get(opts, "clock");
    if (value) {
        if (!strcmp(value, "host")) {
            rtc_clock = host_clock;
        } else if (!strcmp(value, "vm")) {
            rtc_clock = vm_clock;
        } else {
            fprintf(stderr, "qemu: invalid option value '%s'\n", value);
            exit(1);
        }
    }
J
Jan Kiszka 已提交
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
#ifdef CONFIG_TARGET_I386
    value = qemu_opt_get(opts, "driftfix");
    if (value) {
        if (!strcmp(buf, "slew")) {
            rtc_td_hack = 1;
        } else if (!strcmp(buf, "none")) {
            rtc_td_hack = 0;
        } else {
            fprintf(stderr, "qemu: invalid option value '%s'\n", value);
            exit(1);
        }
    }
#endif
}

B
bellard 已提交
1617 1618 1619 1620 1621
#ifdef _WIN32
static void socket_cleanup(void)
{
    WSACleanup();
}
B
bellard 已提交
1622

B
bellard 已提交
1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636
static int socket_init(void)
{
    WSADATA Data;
    int ret, err;

    ret = WSAStartup(MAKEWORD(2,2), &Data);
    if (ret != 0) {
        err = WSAGetLastError();
        fprintf(stderr, "WSAStartup: %d\n", err);
        return -1;
    }
    atexit(socket_cleanup);
    return 0;
}
1637 1638
#endif

1639 1640 1641 1642 1643
/***********************************************************/
/* Bluetooth support */
static int nb_hcis;
static int cur_hci;
static struct HCIInfo *hci_table[MAX_NICS];
1644

1645 1646 1647 1648 1649 1650 1651
static struct bt_vlan_s {
    struct bt_scatternet_s net;
    int id;
    struct bt_vlan_s *next;
} *first_bt_vlan;

/* find or alloc a new bluetooth "VLAN" */
B
blueswir1 已提交
1652
static struct bt_scatternet_s *qemu_find_bt_vlan(int id)
1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691
{
    struct bt_vlan_s **pvlan, *vlan;
    for (vlan = first_bt_vlan; vlan != NULL; vlan = vlan->next) {
        if (vlan->id == id)
            return &vlan->net;
    }
    vlan = qemu_mallocz(sizeof(struct bt_vlan_s));
    vlan->id = id;
    pvlan = &first_bt_vlan;
    while (*pvlan != NULL)
        pvlan = &(*pvlan)->next;
    *pvlan = vlan;
    return &vlan->net;
}

static void null_hci_send(struct HCIInfo *hci, const uint8_t *data, int len)
{
}

static int null_hci_addr_set(struct HCIInfo *hci, const uint8_t *bd_addr)
{
    return -ENOTSUP;
}

static struct HCIInfo null_hci = {
    .cmd_send = null_hci_send,
    .sco_send = null_hci_send,
    .acl_send = null_hci_send,
    .bdaddr_set = null_hci_addr_set,
};

struct HCIInfo *qemu_next_hci(void)
{
    if (cur_hci == nb_hcis)
        return &null_hci;

    return hci_table[cur_hci++];
}

1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724
static struct HCIInfo *hci_init(const char *str)
{
    char *endp;
    struct bt_scatternet_s *vlan = 0;

    if (!strcmp(str, "null"))
        /* null */
        return &null_hci;
    else if (!strncmp(str, "host", 4) && (str[4] == '\0' || str[4] == ':'))
        /* host[:hciN] */
        return bt_host_hci(str[4] ? str + 5 : "hci0");
    else if (!strncmp(str, "hci", 3)) {
        /* hci[,vlan=n] */
        if (str[3]) {
            if (!strncmp(str + 3, ",vlan=", 6)) {
                vlan = qemu_find_bt_vlan(strtol(str + 9, &endp, 0));
                if (*endp)
                    vlan = 0;
            }
        } else
            vlan = qemu_find_bt_vlan(0);
        if (vlan)
           return bt_new_hci(vlan);
    }

    fprintf(stderr, "qemu: Unknown bluetooth HCI `%s'.\n", str);

    return 0;
}

static int bt_hci_parse(const char *str)
{
    struct HCIInfo *hci;
A
Anthony Liguori 已提交
1725
    bdaddr_t bdaddr;
1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829

    if (nb_hcis >= MAX_NICS) {
        fprintf(stderr, "qemu: Too many bluetooth HCIs (max %i).\n", MAX_NICS);
        return -1;
    }

    hci = hci_init(str);
    if (!hci)
        return -1;

    bdaddr.b[0] = 0x52;
    bdaddr.b[1] = 0x54;
    bdaddr.b[2] = 0x00;
    bdaddr.b[3] = 0x12;
    bdaddr.b[4] = 0x34;
    bdaddr.b[5] = 0x56 + nb_hcis;
    hci->bdaddr_set(hci, bdaddr.b);

    hci_table[nb_hcis++] = hci;

    return 0;
}

static void bt_vhci_add(int vlan_id)
{
    struct bt_scatternet_s *vlan = qemu_find_bt_vlan(vlan_id);

    if (!vlan->slave)
        fprintf(stderr, "qemu: warning: adding a VHCI to "
                        "an empty scatternet %i\n", vlan_id);

    bt_vhci_init(bt_new_hci(vlan));
}

static struct bt_device_s *bt_device_add(const char *opt)
{
    struct bt_scatternet_s *vlan;
    int vlan_id = 0;
    char *endp = strstr(opt, ",vlan=");
    int len = (endp ? endp - opt : strlen(opt)) + 1;
    char devname[10];

    pstrcpy(devname, MIN(sizeof(devname), len), opt);

    if (endp) {
        vlan_id = strtol(endp + 6, &endp, 0);
        if (*endp) {
            fprintf(stderr, "qemu: unrecognised bluetooth vlan Id\n");
            return 0;
        }
    }

    vlan = qemu_find_bt_vlan(vlan_id);

    if (!vlan->slave)
        fprintf(stderr, "qemu: warning: adding a slave device to "
                        "an empty scatternet %i\n", vlan_id);

    if (!strcmp(devname, "keyboard"))
        return bt_keyboard_init(vlan);

    fprintf(stderr, "qemu: unsupported bluetooth device `%s'\n", devname);
    return 0;
}

static int bt_parse(const char *opt)
{
    const char *endp, *p;
    int vlan;

    if (strstart(opt, "hci", &endp)) {
        if (!*endp || *endp == ',') {
            if (*endp)
                if (!strstart(endp, ",vlan=", 0))
                    opt = endp + 1;

            return bt_hci_parse(opt);
       }
    } else if (strstart(opt, "vhci", &endp)) {
        if (!*endp || *endp == ',') {
            if (*endp) {
                if (strstart(endp, ",vlan=", &p)) {
                    vlan = strtol(p, (char **) &endp, 0);
                    if (*endp) {
                        fprintf(stderr, "qemu: bad scatternet '%s'\n", p);
                        return 1;
                    }
                } else {
                    fprintf(stderr, "qemu: bad parameter '%s'\n", endp + 1);
                    return 1;
                }
            } else
                vlan = 0;

            bt_vhci_add(vlan);
            return 0;
        }
    } else if (strstart(opt, "device:", &endp))
        return !bt_device_add(endp);

    fprintf(stderr, "qemu: bad bluetooth parameter '%s'\n", opt);
    return 1;
}

1830 1831 1832
/***********************************************************/
/* QEMU Block devices */

1833
#define HD_ALIAS "index=%d,media=disk"
T
ths 已提交
1834 1835
#define CDROM_ALIAS "index=2,media=cdrom"
#define FD_ALIAS "index=%d,if=floppy"
1836 1837
#define PFLASH_ALIAS "if=pflash"
#define MTD_ALIAS "if=mtd"
1838
#define SD_ALIAS "index=0,if=sd"
T
ths 已提交
1839

G
Gerd Hoffmann 已提交
1840
QemuOpts *drive_add(const char *file, const char *fmt, ...)
T
ths 已提交
1841 1842
{
    va_list ap;
G
Gerd Hoffmann 已提交
1843 1844
    char optstr[1024];
    QemuOpts *opts;
T
ths 已提交
1845 1846

    va_start(ap, fmt);
G
Gerd Hoffmann 已提交
1847
    vsnprintf(optstr, sizeof(optstr), fmt, ap);
T
ths 已提交
1848 1849
    va_end(ap);

1850
    opts = qemu_opts_parse(&qemu_drive_opts, optstr, 0);
G
Gerd Hoffmann 已提交
1851 1852 1853 1854 1855 1856 1857 1858
    if (!opts) {
        fprintf(stderr, "%s: huh? duplicate? (%s)\n",
                __FUNCTION__, optstr);
        return NULL;
    }
    if (file)
        qemu_opt_set(opts, "file", file);
    return opts;
1859 1860
}

G
Gerd Hoffmann 已提交
1861
DriveInfo *drive_get(BlockInterfaceType type, int bus, int unit)
T
ths 已提交
1862
{
G
Gerd Hoffmann 已提交
1863
    DriveInfo *dinfo;
T
ths 已提交
1864 1865 1866

    /* seek interface, bus and unit */

B
Blue Swirl 已提交
1867
    QTAILQ_FOREACH(dinfo, &drives, next) {
G
Gerd Hoffmann 已提交
1868 1869 1870 1871 1872
        if (dinfo->type == type &&
	    dinfo->bus == bus &&
	    dinfo->unit == unit)
            return dinfo;
    }
T
ths 已提交
1873

G
Gerd Hoffmann 已提交
1874
    return NULL;
T
ths 已提交
1875 1876
}

G
Gerd Hoffmann 已提交
1877
DriveInfo *drive_get_by_id(const char *id)
G
Gerd Hoffmann 已提交
1878 1879 1880
{
    DriveInfo *dinfo;

B
Blue Swirl 已提交
1881
    QTAILQ_FOREACH(dinfo, &drives, next) {
G
Gerd Hoffmann 已提交
1882 1883 1884 1885 1886 1887 1888
        if (strcmp(id, dinfo->id))
            continue;
        return dinfo;
    }
    return NULL;
}

1889
int drive_get_max_bus(BlockInterfaceType type)
T
ths 已提交
1890 1891
{
    int max_bus;
G
Gerd Hoffmann 已提交
1892
    DriveInfo *dinfo;
T
ths 已提交
1893 1894

    max_bus = -1;
B
Blue Swirl 已提交
1895
    QTAILQ_FOREACH(dinfo, &drives, next) {
G
Gerd Hoffmann 已提交
1896 1897 1898
        if(dinfo->type == type &&
           dinfo->bus > max_bus)
            max_bus = dinfo->bus;
T
ths 已提交
1899 1900 1901 1902
    }
    return max_bus;
}

1903 1904
const char *drive_get_serial(BlockDriverState *bdrv)
{
G
Gerd Hoffmann 已提交
1905
    DriveInfo *dinfo;
1906

B
Blue Swirl 已提交
1907
    QTAILQ_FOREACH(dinfo, &drives, next) {
G
Gerd Hoffmann 已提交
1908 1909 1910
        if (dinfo->bdrv == bdrv)
            return dinfo->serial;
    }
1911 1912 1913 1914

    return "\0";
}

1915 1916
BlockInterfaceErrorAction drive_get_on_error(
    BlockDriverState *bdrv, int is_read)
1917
{
G
Gerd Hoffmann 已提交
1918
    DriveInfo *dinfo;
1919

B
Blue Swirl 已提交
1920
    QTAILQ_FOREACH(dinfo, &drives, next) {
G
Gerd Hoffmann 已提交
1921
        if (dinfo->bdrv == bdrv)
K
Kevin Wolf 已提交
1922
            return is_read ? dinfo->on_read_error : dinfo->on_write_error;
G
Gerd Hoffmann 已提交
1923
    }
1924

K
Kevin Wolf 已提交
1925
    return is_read ? BLOCK_ERR_REPORT : BLOCK_ERR_STOP_ENOSPC;
1926 1927
}

1928 1929 1930 1931 1932
static void bdrv_format_print(void *opaque, const char *name)
{
    fprintf(stderr, " %s", name);
}

G
Gerd Hoffmann 已提交
1933
void drive_uninit(DriveInfo *dinfo)
1934
{
G
Gerd Hoffmann 已提交
1935 1936 1937 1938
    qemu_opts_del(dinfo->opts);
    bdrv_delete(dinfo->bdrv);
    QTAILQ_REMOVE(&drives, dinfo, next);
    qemu_free(dinfo);
1939 1940
}

K
Kevin Wolf 已提交
1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957
static int parse_block_error_action(const char *buf, int is_read)
{
    if (!strcmp(buf, "ignore")) {
        return BLOCK_ERR_IGNORE;
    } else if (!is_read && !strcmp(buf, "enospc")) {
        return BLOCK_ERR_STOP_ENOSPC;
    } else if (!strcmp(buf, "stop")) {
        return BLOCK_ERR_STOP_ANY;
    } else if (!strcmp(buf, "report")) {
        return BLOCK_ERR_REPORT;
    } else {
        fprintf(stderr, "qemu: '%s' invalid %s error action\n",
            buf, is_read ? "read" : "write");
        return -1;
    }
}

G
Gerd Hoffmann 已提交
1958
DriveInfo *drive_init(QemuOpts *opts, void *opaque,
G
Gerd Hoffmann 已提交
1959
                      int *fatal_error)
T
ths 已提交
1960
{
G
Gerd Hoffmann 已提交
1961 1962
    const char *buf;
    const char *file = NULL;
1963
    char devname[128];
G
Gerd Hoffmann 已提交
1964
    const char *serial;
1965
    const char *mediastr = "";
1966
    BlockInterfaceType type;
T
ths 已提交
1967 1968 1969
    enum { MEDIA_DISK, MEDIA_CDROM } media;
    int bus_id, unit_id;
    int cyls, heads, secs, translation;
1970
    BlockDriver *drv = NULL;
1971
    QEMUMachine *machine = opaque;
T
ths 已提交
1972 1973
    int max_devs;
    int index;
1974
    int cache;
1975
    int aio = 0;
1976
    int ro = 0;
K
Kevin Wolf 已提交
1977 1978
    int bdrv_flags;
    int on_read_error, on_write_error;
1979
    const char *devaddr;
G
Gerd Hoffmann 已提交
1980
    DriveInfo *dinfo;
G
Gerd Hoffmann 已提交
1981
    int snapshot = 0;
T
ths 已提交
1982

G
Gerd Hoffmann 已提交
1983
    *fatal_error = 1;
T
ths 已提交
1984 1985

    translation = BIOS_ATA_TRANSLATION_AUTO;
1986
    cache = 1;
T
ths 已提交
1987

1988
    if (machine && machine->use_scsi) {
1989
        type = IF_SCSI;
T
ths 已提交
1990
        max_devs = MAX_SCSI_DEVS;
B
blueswir1 已提交
1991
        pstrcpy(devname, sizeof(devname), "scsi");
T
ths 已提交
1992
    } else {
1993
        type = IF_IDE;
T
ths 已提交
1994
        max_devs = MAX_IDE_DEVS;
B
blueswir1 已提交
1995
        pstrcpy(devname, sizeof(devname), "ide");
T
ths 已提交
1996 1997 1998 1999
    }
    media = MEDIA_DISK;

    /* extract parameters */
G
Gerd Hoffmann 已提交
2000 2001 2002
    bus_id  = qemu_opt_get_number(opts, "bus", 0);
    unit_id = qemu_opt_get_number(opts, "unit", -1);
    index   = qemu_opt_get_number(opts, "index", -1);
T
ths 已提交
2003

G
Gerd Hoffmann 已提交
2004 2005 2006
    cyls  = qemu_opt_get_number(opts, "cyls", 0);
    heads = qemu_opt_get_number(opts, "heads", 0);
    secs  = qemu_opt_get_number(opts, "secs", 0);
T
ths 已提交
2007

G
Gerd Hoffmann 已提交
2008
    snapshot = qemu_opt_get_bool(opts, "snapshot", 0);
2009
    ro = qemu_opt_get_bool(opts, "readonly", 0);
T
ths 已提交
2010

G
Gerd Hoffmann 已提交
2011 2012 2013 2014
    file = qemu_opt_get(opts, "file");
    serial = qemu_opt_get(opts, "serial");

    if ((buf = qemu_opt_get(opts, "if")) != NULL) {
B
bellard 已提交
2015
        pstrcpy(devname, sizeof(devname), buf);
T
ths 已提交
2016
        if (!strcmp(buf, "ide")) {
2017
	    type = IF_IDE;
T
ths 已提交
2018 2019
            max_devs = MAX_IDE_DEVS;
        } else if (!strcmp(buf, "scsi")) {
2020
	    type = IF_SCSI;
T
ths 已提交
2021 2022
            max_devs = MAX_SCSI_DEVS;
        } else if (!strcmp(buf, "floppy")) {
2023
	    type = IF_FLOPPY;
T
ths 已提交
2024 2025
            max_devs = 0;
        } else if (!strcmp(buf, "pflash")) {
2026
	    type = IF_PFLASH;
T
ths 已提交
2027 2028
            max_devs = 0;
	} else if (!strcmp(buf, "mtd")) {
2029
	    type = IF_MTD;
T
ths 已提交
2030 2031
            max_devs = 0;
	} else if (!strcmp(buf, "sd")) {
2032
	    type = IF_SD;
T
ths 已提交
2033
            max_devs = 0;
A
aliguori 已提交
2034 2035 2036
        } else if (!strcmp(buf, "virtio")) {
            type = IF_VIRTIO;
            max_devs = 0;
2037 2038 2039
	} else if (!strcmp(buf, "xen")) {
	    type = IF_XEN;
            max_devs = 0;
G
Gerd Hoffmann 已提交
2040 2041 2042
	} else if (!strcmp(buf, "none")) {
	    type = IF_NONE;
            max_devs = 0;
2043
	} else {
G
Gerd Hoffmann 已提交
2044
            fprintf(stderr, "qemu: unsupported bus type '%s'\n", buf);
G
Gerd Hoffmann 已提交
2045
            return NULL;
T
ths 已提交
2046 2047 2048 2049
	}
    }

    if (cyls || heads || secs) {
2050
        if (cyls < 1 || (type == IF_IDE && cyls > 16383)) {
G
Gerd Hoffmann 已提交
2051
            fprintf(stderr, "qemu: '%s' invalid physical cyls number\n", buf);
G
Gerd Hoffmann 已提交
2052
	    return NULL;
T
ths 已提交
2053
	}
2054
        if (heads < 1 || (type == IF_IDE && heads > 16)) {
G
Gerd Hoffmann 已提交
2055
            fprintf(stderr, "qemu: '%s' invalid physical heads number\n", buf);
G
Gerd Hoffmann 已提交
2056
	    return NULL;
T
ths 已提交
2057
	}
2058
        if (secs < 1 || (type == IF_IDE && secs > 63)) {
G
Gerd Hoffmann 已提交
2059
            fprintf(stderr, "qemu: '%s' invalid physical secs number\n", buf);
G
Gerd Hoffmann 已提交
2060
	    return NULL;
T
ths 已提交
2061 2062 2063
	}
    }

G
Gerd Hoffmann 已提交
2064
    if ((buf = qemu_opt_get(opts, "trans")) != NULL) {
T
ths 已提交
2065 2066 2067
        if (!cyls) {
            fprintf(stderr,
                    "qemu: '%s' trans must be used with cyls,heads and secs\n",
G
Gerd Hoffmann 已提交
2068
                    buf);
G
Gerd Hoffmann 已提交
2069
            return NULL;
T
ths 已提交
2070 2071 2072 2073 2074 2075 2076 2077
        }
        if (!strcmp(buf, "none"))
            translation = BIOS_ATA_TRANSLATION_NONE;
        else if (!strcmp(buf, "lba"))
            translation = BIOS_ATA_TRANSLATION_LBA;
        else if (!strcmp(buf, "auto"))
            translation = BIOS_ATA_TRANSLATION_AUTO;
	else {
G
Gerd Hoffmann 已提交
2078
            fprintf(stderr, "qemu: '%s' invalid translation type\n", buf);
G
Gerd Hoffmann 已提交
2079
	    return NULL;
T
ths 已提交
2080 2081 2082
	}
    }

G
Gerd Hoffmann 已提交
2083
    if ((buf = qemu_opt_get(opts, "media")) != NULL) {
T
ths 已提交
2084 2085 2086 2087 2088
        if (!strcmp(buf, "disk")) {
	    media = MEDIA_DISK;
	} else if (!strcmp(buf, "cdrom")) {
            if (cyls || secs || heads) {
                fprintf(stderr,
G
Gerd Hoffmann 已提交
2089
                        "qemu: '%s' invalid physical CHS format\n", buf);
G
Gerd Hoffmann 已提交
2090
	        return NULL;
T
ths 已提交
2091 2092 2093
            }
	    media = MEDIA_CDROM;
	} else {
G
Gerd Hoffmann 已提交
2094
	    fprintf(stderr, "qemu: '%s' invalid media\n", buf);
G
Gerd Hoffmann 已提交
2095
	    return NULL;
T
ths 已提交
2096 2097 2098
	}
    }

G
Gerd Hoffmann 已提交
2099
    if ((buf = qemu_opt_get(opts, "cache")) != NULL) {
2100
        if (!strcmp(buf, "off") || !strcmp(buf, "none"))
2101
            cache = 0;
2102
        else if (!strcmp(buf, "writethrough"))
2103
            cache = 1;
2104 2105
        else if (!strcmp(buf, "writeback"))
            cache = 2;
2106 2107
        else {
           fprintf(stderr, "qemu: invalid cache option\n");
G
Gerd Hoffmann 已提交
2108
           return NULL;
2109 2110 2111
        }
    }

2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124
#ifdef CONFIG_LINUX_AIO
    if ((buf = qemu_opt_get(opts, "aio")) != NULL) {
        if (!strcmp(buf, "threads"))
            aio = 0;
        else if (!strcmp(buf, "native"))
            aio = 1;
        else {
           fprintf(stderr, "qemu: invalid aio option\n");
           return NULL;
        }
    }
#endif

G
Gerd Hoffmann 已提交
2125
    if ((buf = qemu_opt_get(opts, "format")) != NULL) {
2126 2127 2128 2129
       if (strcmp(buf, "?") == 0) {
            fprintf(stderr, "qemu: Supported formats:");
            bdrv_iterate_format(bdrv_format_print, NULL);
            fprintf(stderr, "\n");
G
Gerd Hoffmann 已提交
2130
	    return NULL;
2131
        }
2132
        drv = bdrv_find_whitelisted_format(buf);
2133 2134
        if (!drv) {
            fprintf(stderr, "qemu: '%s' invalid format\n", buf);
G
Gerd Hoffmann 已提交
2135
            return NULL;
2136 2137 2138
        }
    }

K
Kevin Wolf 已提交
2139
    on_write_error = BLOCK_ERR_STOP_ENOSPC;
G
Gerd Hoffmann 已提交
2140
    if ((buf = qemu_opt_get(opts, "werror")) != NULL) {
2141
        if (type != IF_IDE && type != IF_SCSI && type != IF_VIRTIO) {
2142
            fprintf(stderr, "werror is no supported by this format\n");
G
Gerd Hoffmann 已提交
2143
            return NULL;
2144
        }
K
Kevin Wolf 已提交
2145 2146 2147 2148 2149 2150 2151 2152 2153

        on_write_error = parse_block_error_action(buf, 0);
        if (on_write_error < 0) {
            return NULL;
        }
    }

    on_read_error = BLOCK_ERR_REPORT;
    if ((buf = qemu_opt_get(opts, "rerror")) != NULL) {
K
Kevin Wolf 已提交
2154
        if (type != IF_IDE && type != IF_VIRTIO) {
K
Kevin Wolf 已提交
2155 2156 2157 2158 2159 2160
            fprintf(stderr, "rerror is no supported by this format\n");
            return NULL;
        }

        on_read_error = parse_block_error_action(buf, 1);
        if (on_read_error < 0) {
G
Gerd Hoffmann 已提交
2161
            return NULL;
2162 2163 2164
        }
    }

G
Gerd Hoffmann 已提交
2165
    if ((devaddr = qemu_opt_get(opts, "addr")) != NULL) {
2166
        if (type != IF_VIRTIO) {
G
Gerd Hoffmann 已提交
2167
            fprintf(stderr, "addr is not supported\n");
G
Gerd Hoffmann 已提交
2168
            return NULL;
2169 2170 2171
        }
    }

T
ths 已提交
2172 2173 2174 2175 2176
    /* compute bus and unit according index */

    if (index != -1) {
        if (bus_id != 0 || unit_id != -1) {
            fprintf(stderr,
G
Gerd Hoffmann 已提交
2177
                    "qemu: index cannot be used with bus and unit\n");
G
Gerd Hoffmann 已提交
2178
            return NULL;
T
ths 已提交
2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195
        }
        if (max_devs == 0)
        {
            unit_id = index;
            bus_id = 0;
        } else {
            unit_id = index % max_devs;
            bus_id = index / max_devs;
        }
    }

    /* if user doesn't specify a unit_id,
     * try to find the first free
     */

    if (unit_id == -1) {
       unit_id = 0;
G
Gerd Hoffmann 已提交
2196
       while (drive_get(type, bus_id, unit_id) != NULL) {
T
ths 已提交
2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207
           unit_id++;
           if (max_devs && unit_id >= max_devs) {
               unit_id -= max_devs;
               bus_id++;
           }
       }
    }

    /* check unit id */

    if (max_devs && unit_id >= max_devs) {
G
Gerd Hoffmann 已提交
2208 2209
        fprintf(stderr, "qemu: unit %d too big (max is %d)\n",
                unit_id, max_devs - 1);
G
Gerd Hoffmann 已提交
2210
        return NULL;
T
ths 已提交
2211 2212 2213 2214 2215 2216
    }

    /*
     * ignore multiple definitions
     */

G
Gerd Hoffmann 已提交
2217 2218 2219 2220
    if (drive_get(type, bus_id, unit_id) != NULL) {
        *fatal_error = 0;
        return NULL;
    }
T
ths 已提交
2221 2222 2223

    /* init */

G
Gerd Hoffmann 已提交
2224
    dinfo = qemu_mallocz(sizeof(*dinfo));
2225
    if ((buf = qemu_opts_id(opts)) != NULL) {
G
Gerd Hoffmann 已提交
2226 2227
        dinfo->id = qemu_strdup(buf);
    } else {
G
Gerd Hoffmann 已提交
2228
        /* no id supplied -> create one */
G
Gerd Hoffmann 已提交
2229
        dinfo->id = qemu_mallocz(32);
G
Gerd Hoffmann 已提交
2230 2231 2232
        if (type == IF_IDE || type == IF_SCSI)
            mediastr = (media == MEDIA_CDROM) ? "-cd" : "-hd";
        if (max_devs)
G
Gerd Hoffmann 已提交
2233
            snprintf(dinfo->id, 32, "%s%i%s%i",
G
Gerd Hoffmann 已提交
2234 2235
                     devname, bus_id, mediastr, unit_id);
        else
G
Gerd Hoffmann 已提交
2236
            snprintf(dinfo->id, 32, "%s%s%i",
G
Gerd Hoffmann 已提交
2237 2238 2239
                     devname, mediastr, unit_id);
    }
    dinfo->bdrv = bdrv_new(dinfo->id);
G
Gerd Hoffmann 已提交
2240 2241 2242 2243
    dinfo->devaddr = devaddr;
    dinfo->type = type;
    dinfo->bus = bus_id;
    dinfo->unit = unit_id;
K
Kevin Wolf 已提交
2244 2245
    dinfo->on_read_error = on_read_error;
    dinfo->on_write_error = on_write_error;
G
Gerd Hoffmann 已提交
2246 2247 2248
    dinfo->opts = opts;
    if (serial)
        strncpy(dinfo->serial, serial, sizeof(serial));
B
Blue Swirl 已提交
2249
    QTAILQ_INSERT_TAIL(&drives, dinfo, next);
T
ths 已提交
2250

2251
    switch(type) {
T
ths 已提交
2252 2253
    case IF_IDE:
    case IF_SCSI:
2254
    case IF_XEN:
G
Gerd Hoffmann 已提交
2255
    case IF_NONE:
T
ths 已提交
2256 2257 2258
        switch(media) {
	case MEDIA_DISK:
            if (cyls != 0) {
G
Gerd Hoffmann 已提交
2259 2260
                bdrv_set_geometry_hint(dinfo->bdrv, cyls, heads, secs);
                bdrv_set_translation_hint(dinfo->bdrv, translation);
T
ths 已提交
2261 2262 2263
            }
	    break;
	case MEDIA_CDROM:
G
Gerd Hoffmann 已提交
2264
            bdrv_set_type_hint(dinfo->bdrv, BDRV_TYPE_CDROM);
T
ths 已提交
2265 2266 2267 2268 2269 2270 2271
	    break;
	}
        break;
    case IF_SD:
        /* FIXME: This isn't really a floppy, but it's a reasonable
           approximation.  */
    case IF_FLOPPY:
G
Gerd Hoffmann 已提交
2272
        bdrv_set_type_hint(dinfo->bdrv, BDRV_TYPE_FLOPPY);
T
ths 已提交
2273 2274 2275 2276
        break;
    case IF_PFLASH:
    case IF_MTD:
        break;
G
Gerd Hoffmann 已提交
2277 2278 2279 2280 2281 2282 2283 2284
    case IF_VIRTIO:
        /* add virtio block device */
        opts = qemu_opts_create(&qemu_device_opts, NULL, 0);
        qemu_opt_set(opts, "driver", "virtio-blk-pci");
        qemu_opt_set(opts, "drive", dinfo->id);
        if (devaddr)
            qemu_opt_set(opts, "addr", devaddr);
        break;
P
Paul Brook 已提交
2285 2286
    case IF_COUNT:
        abort();
T
ths 已提交
2287
    }
G
Gerd Hoffmann 已提交
2288
    if (!file) {
G
Gerd Hoffmann 已提交
2289 2290 2291
        *fatal_error = 0;
        return NULL;
    }
2292
    bdrv_flags = 0;
2293
    if (snapshot) {
2294
        bdrv_flags |= BDRV_O_SNAPSHOT;
2295 2296 2297 2298 2299 2300
        cache = 2; /* always use write-back with snapshot */
    }
    if (cache == 0) /* no caching */
        bdrv_flags |= BDRV_O_NOCACHE;
    else if (cache == 2) /* write-back */
        bdrv_flags |= BDRV_O_CACHE_WB;
2301 2302 2303 2304 2305 2306 2307

    if (aio == 1) {
        bdrv_flags |= BDRV_O_NATIVE_AIO;
    } else {
        bdrv_flags &= ~BDRV_O_NATIVE_AIO;
    }

2308
    if (ro == 1) {
2309 2310
        if (type != IF_SCSI && type != IF_VIRTIO && type != IF_FLOPPY) {
            fprintf(stderr, "qemu: readonly flag not supported for drive with this interface\n");
2311 2312
            return NULL;
        }
N
Naphtali Sprei 已提交
2313 2314 2315 2316 2317 2318
    }
    /* 
     * cdrom is read-only. Set it now, after above interface checking
     * since readonly attribute not explicitly required, so no error.
     */
    if (media == MEDIA_CDROM) {
2319
        ro = 1;
2320
    }
2321
    bdrv_flags |= ro ? 0 : BDRV_O_RDWR;
2322

G
Gerd Hoffmann 已提交
2323
    if (bdrv_open2(dinfo->bdrv, file, bdrv_flags, drv) < 0) {
2324 2325
        fprintf(stderr, "qemu: could not open disk image %s: %s\n",
                        file, strerror(errno));
G
Gerd Hoffmann 已提交
2326
        return NULL;
T
ths 已提交
2327
    }
2328

G
Gerd Hoffmann 已提交
2329
    if (bdrv_key_required(dinfo->bdrv))
2330
        autostart = 0;
G
Gerd Hoffmann 已提交
2331 2332
    *fatal_error = 0;
    return dinfo;
T
ths 已提交
2333 2334
}

G
Gerd Hoffmann 已提交
2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354
static int drive_init_func(QemuOpts *opts, void *opaque)
{
    QEMUMachine *machine = opaque;
    int fatal_error = 0;

    if (drive_init(opts, machine, &fatal_error) == NULL) {
        if (fatal_error)
            return 1;
    }
    return 0;
}

static int drive_enable_snapshot(QemuOpts *opts, void *opaque)
{
    if (NULL == qemu_opt_get(opts, "snapshot")) {
        qemu_opt_set(opts, "snapshot", "on");
    }
    return 0;
}

J
Jan Kiszka 已提交
2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368
void qemu_register_boot_set(QEMUBootSetHandler *func, void *opaque)
{
    boot_set_handler = func;
    boot_set_opaque = opaque;
}

int qemu_boot_set(const char *boot_devices)
{
    if (!boot_set_handler) {
        return -EINVAL;
    }
    return boot_set_handler(boot_set_opaque, boot_devices);
}

J
Jan Kiszka 已提交
2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397
static int parse_bootdevices(char *devices)
{
    /* We just do some generic consistency checks */
    const char *p;
    int bitmap = 0;

    for (p = devices; *p != '\0'; p++) {
        /* Allowed boot devices are:
         * a-b: floppy disk drives
         * c-f: IDE disk drives
         * g-m: machine implementation dependant drives
         * n-p: network devices
         * It's up to each machine implementation to check if the given boot
         * devices match the actual hardware implementation and firmware
         * features.
         */
        if (*p < 'a' || *p > 'p') {
            fprintf(stderr, "Invalid boot device '%c'\n", *p);
            exit(1);
        }
        if (bitmap & (1 << (*p - 'a'))) {
            fprintf(stderr, "Boot device '%c' was given twice\n", *p);
            exit(1);
        }
        bitmap |= 1 << (*p - 'a');
    }
    return bitmap;
}

J
Jan Kiszka 已提交
2398 2399 2400 2401 2402 2403 2404 2405 2406 2407
static void restore_boot_devices(void *opaque)
{
    char *standard_boot_devices = opaque;

    qemu_boot_set(standard_boot_devices);

    qemu_unregister_reset(restore_boot_devices, standard_boot_devices);
    qemu_free(standard_boot_devices);
}

2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451
static void numa_add(const char *optarg)
{
    char option[128];
    char *endptr;
    unsigned long long value, endvalue;
    int nodenr;

    optarg = get_opt_name(option, 128, optarg, ',') + 1;
    if (!strcmp(option, "node")) {
        if (get_param_value(option, 128, "nodeid", optarg) == 0) {
            nodenr = nb_numa_nodes;
        } else {
            nodenr = strtoull(option, NULL, 10);
        }

        if (get_param_value(option, 128, "mem", optarg) == 0) {
            node_mem[nodenr] = 0;
        } else {
            value = strtoull(option, &endptr, 0);
            switch (*endptr) {
            case 0: case 'M': case 'm':
                value <<= 20;
                break;
            case 'G': case 'g':
                value <<= 30;
                break;
            }
            node_mem[nodenr] = value;
        }
        if (get_param_value(option, 128, "cpus", optarg) == 0) {
            node_cpumask[nodenr] = 0;
        } else {
            value = strtoull(option, &endptr, 10);
            if (value >= 64) {
                value = 63;
                fprintf(stderr, "only 64 CPUs in NUMA mode supported.\n");
            } else {
                if (*endptr == '-') {
                    endvalue = strtoull(endptr+1, &endptr, 10);
                    if (endvalue >= 63) {
                        endvalue = 62;
                        fprintf(stderr,
                            "only 63 CPUs in NUMA mode supported.\n");
                    }
P
Paolo Bonzini 已提交
2452
                    value = (2ULL << endvalue) - (1ULL << value);
2453
                } else {
P
Paolo Bonzini 已提交
2454
                    value = 1ULL << value;
2455 2456 2457 2458 2459 2460 2461 2462 2463
                }
            }
            node_cpumask[nodenr] = value;
        }
        nb_numa_nodes++;
    }
    return;
}

2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497
static void smp_parse(const char *optarg)
{
    int smp, sockets = 0, threads = 0, cores = 0;
    char *endptr;
    char option[128];

    smp = strtoul(optarg, &endptr, 10);
    if (endptr != optarg) {
        if (*endptr == ',') {
            endptr++;
        }
    }
    if (get_param_value(option, 128, "sockets", endptr) != 0)
        sockets = strtoull(option, NULL, 10);
    if (get_param_value(option, 128, "cores", endptr) != 0)
        cores = strtoull(option, NULL, 10);
    if (get_param_value(option, 128, "threads", endptr) != 0)
        threads = strtoull(option, NULL, 10);
    if (get_param_value(option, 128, "maxcpus", endptr) != 0)
        max_cpus = strtoull(option, NULL, 10);

    /* compute missing values, prefer sockets over cores over threads */
    if (smp == 0 || sockets == 0) {
        sockets = sockets > 0 ? sockets : 1;
        cores = cores > 0 ? cores : 1;
        threads = threads > 0 ? threads : 1;
        if (smp == 0) {
            smp = cores * threads * sockets;
        }
    } else {
        if (cores == 0) {
            threads = threads > 0 ? threads : 1;
            cores = smp / (sockets * threads);
        } else {
A
Amit Shah 已提交
2498
            if (sockets) {
2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509
                threads = smp / (cores * sockets);
            }
        }
    }
    smp_cpus = smp;
    smp_cores = cores > 0 ? cores : 1;
    smp_threads = threads > 0 ? threads : 1;
    if (max_cpus == 0)
        max_cpus = smp_cpus;
}

B
bellard 已提交
2510 2511 2512
/***********************************************************/
/* USB devices */

2513
static int usb_device_add(const char *devname, int is_hotplug)
B
bellard 已提交
2514 2515
{
    const char *p;
2516
    USBDevice *dev = NULL;
B
bellard 已提交
2517

2518
    if (!usb_enabled)
B
bellard 已提交
2519 2520
        return -1;

G
Gerd Hoffmann 已提交
2521 2522 2523 2524 2525
    /* drivers with .usbdevice_name entry in USBDeviceInfo */
    dev = usbdevice_create(devname);
    if (dev)
        goto done;

2526
    /* the other ones */
B
bellard 已提交
2527 2528
    if (strstart(devname, "host:", &p)) {
        dev = usb_host_device_open(p);
2529 2530 2531
    } else if (!strcmp(devname, "bt") || strstart(devname, "bt:", &p)) {
        dev = usb_bt_init(devname[2] ? hci_init(p) :
                        bt_new_hci(qemu_find_bt_vlan(0)));
B
bellard 已提交
2532 2533 2534
    } else {
        return -1;
    }
P
pbrook 已提交
2535 2536 2537
    if (!dev)
        return -1;

2538
done:
B
bellard 已提交
2539 2540 2541
    return 0;
}

2542 2543 2544 2545 2546
static int usb_device_del(const char *devname)
{
    int bus_num, addr;
    const char *p;

2547 2548 2549
    if (strstart(devname, "host:", &p))
        return usb_host_device_close(p);

2550
    if (!usb_enabled)
2551 2552 2553 2554 2555 2556 2557 2558
        return -1;

    p = strchr(devname, '.');
    if (!p)
        return -1;
    bus_num = strtoul(devname, NULL, 0);
    addr = strtoul(p + 1, NULL, 0);

2559
    return usb_device_delete_addr(bus_num, addr);
2560 2561
}

2562 2563
static int usb_parse(const char *cmdline)
{
S
Scott Tsai 已提交
2564 2565 2566 2567 2568 2569
    int r;
    r = usb_device_add(cmdline, 0);
    if (r < 0) {
        fprintf(stderr, "qemu: could not add USB device '%s'\n", cmdline);
    }
    return r;
2570 2571
}

2572
void do_usb_add(Monitor *mon, const QDict *qdict)
B
bellard 已提交
2573
{
S
Scott Tsai 已提交
2574 2575
    const char *devname = qdict_get_str(qdict, "devname");
    if (usb_device_add(devname, 1) < 0) {
2576
        error_report("could not add USB device '%s'", devname);
S
Scott Tsai 已提交
2577
    }
B
bellard 已提交
2578 2579
}

2580
void do_usb_del(Monitor *mon, const QDict *qdict)
B
bellard 已提交
2581
{
S
Scott Tsai 已提交
2582 2583
    const char *devname = qdict_get_str(qdict, "devname");
    if (usb_device_del(devname) < 0) {
2584
        error_report("could not delete USB device '%s'", devname);
S
Scott Tsai 已提交
2585
    }
B
bellard 已提交
2586 2587
}

2588 2589 2590 2591
/***********************************************************/
/* PCMCIA/Cardbus */

static struct pcmcia_socket_entry_s {
P
Paul Brook 已提交
2592
    PCMCIASocket *socket;
2593 2594 2595
    struct pcmcia_socket_entry_s *next;
} *pcmcia_sockets = 0;

P
Paul Brook 已提交
2596
void pcmcia_socket_register(PCMCIASocket *socket)
2597 2598 2599 2600 2601 2602 2603 2604 2605
{
    struct pcmcia_socket_entry_s *entry;

    entry = qemu_malloc(sizeof(struct pcmcia_socket_entry_s));
    entry->socket = socket;
    entry->next = pcmcia_sockets;
    pcmcia_sockets = entry;
}

P
Paul Brook 已提交
2606
void pcmcia_socket_unregister(PCMCIASocket *socket)
2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617
{
    struct pcmcia_socket_entry_s *entry, **ptr;

    ptr = &pcmcia_sockets;
    for (entry = *ptr; entry; ptr = &entry->next, entry = *ptr)
        if (entry->socket == socket) {
            *ptr = entry->next;
            qemu_free(entry);
        }
}

A
aliguori 已提交
2618
void pcmcia_info(Monitor *mon)
2619 2620
{
    struct pcmcia_socket_entry_s *iter;
A
aliguori 已提交
2621

2622
    if (!pcmcia_sockets)
A
aliguori 已提交
2623
        monitor_printf(mon, "No PCMCIA sockets\n");
2624 2625

    for (iter = pcmcia_sockets; iter; iter = iter->next)
A
aliguori 已提交
2626 2627 2628
        monitor_printf(mon, "%s: %s\n", iter->socket->slot_string,
                       iter->socket->attached ? iter->socket->card_string :
                       "Empty");
2629 2630
}

2631
/***********************************************************/
2632
/* I/O handling */
2633

2634 2635
typedef struct IOHandlerRecord {
    int fd;
B
bellard 已提交
2636 2637 2638
    IOCanRWHandler *fd_read_poll;
    IOHandler *fd_read;
    IOHandler *fd_write;
2639
    int deleted;
2640 2641 2642
    void *opaque;
    /* temporary data */
    struct pollfd *ufd;
2643
    struct IOHandlerRecord *next;
2644 2645
} IOHandlerRecord;

2646
static IOHandlerRecord *first_io_handler;
2647

B
bellard 已提交
2648 2649
/* XXX: fd_read_poll should be suppressed, but an API change is
   necessary in the character devices to suppress fd_can_read(). */
2650 2651 2652 2653
int qemu_set_fd_handler2(int fd,
                         IOCanRWHandler *fd_read_poll,
                         IOHandler *fd_read,
                         IOHandler *fd_write,
B
bellard 已提交
2654
                         void *opaque)
2655
{
B
bellard 已提交
2656
    IOHandlerRecord **pioh, *ioh;
2657

B
bellard 已提交
2658 2659 2660 2661 2662 2663 2664
    if (!fd_read && !fd_write) {
        pioh = &first_io_handler;
        for(;;) {
            ioh = *pioh;
            if (ioh == NULL)
                break;
            if (ioh->fd == fd) {
2665
                ioh->deleted = 1;
B
bellard 已提交
2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683
                break;
            }
            pioh = &ioh->next;
        }
    } else {
        for(ioh = first_io_handler; ioh != NULL; ioh = ioh->next) {
            if (ioh->fd == fd)
                goto found;
        }
        ioh = qemu_mallocz(sizeof(IOHandlerRecord));
        ioh->next = first_io_handler;
        first_io_handler = ioh;
    found:
        ioh->fd = fd;
        ioh->fd_read_poll = fd_read_poll;
        ioh->fd_read = fd_read;
        ioh->fd_write = fd_write;
        ioh->opaque = opaque;
2684
        ioh->deleted = 0;
B
bellard 已提交
2685
    }
2686 2687 2688
    return 0;
}

2689 2690 2691
int qemu_set_fd_handler(int fd,
                        IOHandler *fd_read,
                        IOHandler *fd_write,
B
bellard 已提交
2692
                        void *opaque)
2693
{
B
bellard 已提交
2694
    return qemu_set_fd_handler2(fd, NULL, fd_read, fd_write, opaque);
2695 2696
}

A
aliguori 已提交
2697
#ifdef _WIN32
2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732
/***********************************************************/
/* Polling handling */

typedef struct PollingEntry {
    PollingFunc *func;
    void *opaque;
    struct PollingEntry *next;
} PollingEntry;

static PollingEntry *first_polling_entry;

int qemu_add_polling_cb(PollingFunc *func, void *opaque)
{
    PollingEntry **ppe, *pe;
    pe = qemu_mallocz(sizeof(PollingEntry));
    pe->func = func;
    pe->opaque = opaque;
    for(ppe = &first_polling_entry; *ppe != NULL; ppe = &(*ppe)->next);
    *ppe = pe;
    return 0;
}

void qemu_del_polling_cb(PollingFunc *func, void *opaque)
{
    PollingEntry **ppe, *pe;
    for(ppe = &first_polling_entry; *ppe != NULL; ppe = &(*ppe)->next) {
        pe = *ppe;
        if (pe->func == func && pe->opaque == opaque) {
            *ppe = pe->next;
            qemu_free(pe);
            break;
        }
    }
}

2733 2734 2735 2736 2737 2738 2739 2740 2741 2742
/***********************************************************/
/* Wait objects support */
typedef struct WaitObjects {
    int num;
    HANDLE events[MAXIMUM_WAIT_OBJECTS + 1];
    WaitObjectFunc *func[MAXIMUM_WAIT_OBJECTS + 1];
    void *opaque[MAXIMUM_WAIT_OBJECTS + 1];
} WaitObjects;

static WaitObjects wait_objects = {0};
2743

2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769
int qemu_add_wait_object(HANDLE handle, WaitObjectFunc *func, void *opaque)
{
    WaitObjects *w = &wait_objects;

    if (w->num >= MAXIMUM_WAIT_OBJECTS)
        return -1;
    w->events[w->num] = handle;
    w->func[w->num] = func;
    w->opaque[w->num] = opaque;
    w->num++;
    return 0;
}

void qemu_del_wait_object(HANDLE handle, WaitObjectFunc *func, void *opaque)
{
    int i, found;
    WaitObjects *w = &wait_objects;

    found = 0;
    for (i = 0; i < w->num; i++) {
        if (w->events[i] == handle)
            found = 1;
        if (found) {
            w->events[i] = w->events[i + 1];
            w->func[i] = w->func[i + 1];
            w->opaque[i] = w->opaque[i + 1];
2770
        }
2771 2772 2773 2774 2775 2776
    }
    if (found)
        w->num--;
}
#endif

2777 2778 2779
/***********************************************************/
/* ram save/restore */

2780
#define RAM_SAVE_FLAG_FULL	0x01 /* Obsolete, not used anymore */
2781 2782 2783 2784 2785 2786
#define RAM_SAVE_FLAG_COMPRESS	0x02
#define RAM_SAVE_FLAG_MEM_SIZE	0x04
#define RAM_SAVE_FLAG_PAGE	0x08
#define RAM_SAVE_FLAG_EOS	0x10

static int is_dup_page(uint8_t *page, uint8_t ch)
2787
{
2788 2789 2790
    uint32_t val = ch << 24 | ch << 16 | ch << 8 | ch;
    uint32_t *array = (uint32_t *)page;
    int i;
2791

2792 2793 2794 2795 2796 2797 2798 2799 2800 2801
    for (i = 0; i < (TARGET_PAGE_SIZE / 4); i++) {
        if (array[i] != val)
            return 0;
    }

    return 1;
}

static int ram_save_block(QEMUFile *f)
{
A
Anthony Liguori 已提交
2802 2803 2804
    static ram_addr_t current_addr = 0;
    ram_addr_t saved_addr = current_addr;
    ram_addr_t addr = 0;
2805 2806
    int found = 0;

P
pbrook 已提交
2807
    while (addr < last_ram_offset) {
2808
        if (cpu_physical_memory_get_dirty(current_addr, MIGRATION_DIRTY_FLAG)) {
P
pbrook 已提交
2809
            uint8_t *p;
2810 2811 2812 2813 2814

            cpu_physical_memory_reset_dirty(current_addr,
                                            current_addr + TARGET_PAGE_SIZE,
                                            MIGRATION_DIRTY_FLAG);

P
pbrook 已提交
2815
            p = qemu_get_ram_ptr(current_addr);
2816

P
pbrook 已提交
2817
            if (is_dup_page(p, *p)) {
2818
                qemu_put_be64(f, current_addr | RAM_SAVE_FLAG_COMPRESS);
P
pbrook 已提交
2819
                qemu_put_byte(f, *p);
2820 2821
            } else {
                qemu_put_be64(f, current_addr | RAM_SAVE_FLAG_PAGE);
P
pbrook 已提交
2822
                qemu_put_buffer(f, p, TARGET_PAGE_SIZE);
2823
            }
2824 2825 2826

            found = 1;
            break;
2827
        }
2828
        addr += TARGET_PAGE_SIZE;
P
pbrook 已提交
2829
        current_addr = (saved_addr + addr) % last_ram_offset;
2830
    }
2831 2832

    return found;
2833 2834
}

2835
static uint64_t bytes_transferred;
2836

A
Anthony Liguori 已提交
2837
static ram_addr_t ram_save_remaining(void)
2838
{
A
Anthony Liguori 已提交
2839 2840
    ram_addr_t addr;
    ram_addr_t count = 0;
2841

P
pbrook 已提交
2842
    for (addr = 0; addr < last_ram_offset; addr += TARGET_PAGE_SIZE) {
2843 2844 2845 2846 2847 2848 2849
        if (cpu_physical_memory_get_dirty(addr, MIGRATION_DIRTY_FLAG))
            count++;
    }

    return count;
}

2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864
uint64_t ram_bytes_remaining(void)
{
    return ram_save_remaining() * TARGET_PAGE_SIZE;
}

uint64_t ram_bytes_transferred(void)
{
    return bytes_transferred;
}

uint64_t ram_bytes_total(void)
{
    return last_ram_offset;
}

2865
static int ram_save_live(Monitor *mon, QEMUFile *f, int stage, void *opaque)
2866
{
A
Anthony Liguori 已提交
2867
    ram_addr_t addr;
2868 2869 2870
    uint64_t bytes_transferred_last;
    double bwidth = 0;
    uint64_t expected_time = 0;
2871

2872 2873 2874 2875 2876
    if (stage < 0) {
        cpu_physical_memory_set_dirty_tracking(0);
        return 0;
    }

2877
    if (cpu_physical_sync_dirty_bitmap(0, TARGET_PHYS_ADDR_MAX) != 0) {
2878 2879 2880 2881
        qemu_file_set_error(f);
        return 0;
    }

2882
    if (stage == 1) {
2883 2884
        bytes_transferred = 0;

2885
        /* Make sure all dirty bits are set */
P
pbrook 已提交
2886
        for (addr = 0; addr < last_ram_offset; addr += TARGET_PAGE_SIZE) {
2887 2888 2889
            if (!cpu_physical_memory_get_dirty(addr, MIGRATION_DIRTY_FLAG))
                cpu_physical_memory_set_dirty(addr);
        }
2890

2891 2892 2893
        /* Enable dirty memory tracking */
        cpu_physical_memory_set_dirty_tracking(1);

P
pbrook 已提交
2894
        qemu_put_be64(f, last_ram_offset | RAM_SAVE_FLAG_MEM_SIZE);
2895 2896
    }

2897
    bytes_transferred_last = bytes_transferred;
P
Paolo Bonzini 已提交
2898
    bwidth = qemu_get_clock_ns(rt_clock);
2899

2900 2901 2902 2903
    while (!qemu_file_rate_limit(f)) {
        int ret;

        ret = ram_save_block(f);
2904
        bytes_transferred += ret * TARGET_PAGE_SIZE;
2905 2906 2907 2908
        if (ret == 0) /* no more blocks */
            break;
    }

P
Paolo Bonzini 已提交
2909
    bwidth = qemu_get_clock_ns(rt_clock) - bwidth;
2910 2911 2912 2913 2914 2915 2916
    bwidth = (bytes_transferred - bytes_transferred_last) / bwidth;

    /* if we haven't transferred anything this round, force expected_time to a
     * a very high value, but without crashing */
    if (bwidth == 0)
        bwidth = 0.000001;

2917 2918 2919
    /* try transferring iterative blocks of memory */
    if (stage == 3) {
        /* flush all remaining blocks regardless of rate limiting */
2920 2921 2922
        while (ram_save_block(f) != 0) {
            bytes_transferred += TARGET_PAGE_SIZE;
        }
2923
        cpu_physical_memory_set_dirty_tracking(0);
2924 2925 2926 2927
    }

    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);

2928 2929 2930
    expected_time = ram_save_remaining() * TARGET_PAGE_SIZE / bwidth;

    return (stage == 2) && (expected_time <= migrate_max_downtime());
2931 2932 2933 2934
}

static int ram_load(QEMUFile *f, void *opaque, int version_id)
{
A
Anthony Liguori 已提交
2935
    ram_addr_t addr;
2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947
    int flags;

    if (version_id != 3)
        return -EINVAL;

    do {
        addr = qemu_get_be64(f);

        flags = addr & ~TARGET_PAGE_MASK;
        addr &= TARGET_PAGE_MASK;

        if (flags & RAM_SAVE_FLAG_MEM_SIZE) {
P
pbrook 已提交
2948
            if (addr != last_ram_offset)
2949 2950 2951 2952 2953
                return -EINVAL;
        }

        if (flags & RAM_SAVE_FLAG_COMPRESS) {
            uint8_t ch = qemu_get_byte(f);
2954 2955
            memset(qemu_get_ram_ptr(addr), ch, TARGET_PAGE_SIZE);
#ifndef _WIN32
2956 2957 2958
            if (ch == 0 &&
                (!kvm_enabled() || kvm_has_sync_mmu())) {
                madvise(qemu_get_ram_ptr(addr), TARGET_PAGE_SIZE, MADV_DONTNEED);
2959
            }
2960
#endif
J
Jan Kiszka 已提交
2961
        } else if (flags & RAM_SAVE_FLAG_PAGE) {
P
pbrook 已提交
2962
            qemu_get_buffer(f, qemu_get_ram_ptr(addr), TARGET_PAGE_SIZE);
J
Jan Kiszka 已提交
2963 2964 2965 2966
        }
        if (qemu_file_has_error(f)) {
            return -EIO;
        }
2967 2968
    } while (!(flags & RAM_SAVE_FLAG_EOS));

2969 2970 2971
    return 0;
}

A
aliguori 已提交
2972 2973
void qemu_service_io(void)
{
2974
    qemu_notify_event();
A
aliguori 已提交
2975 2976
}

2977 2978 2979
/***********************************************************/
/* machine registration */

B
blueswir1 已提交
2980
static QEMUMachine *first_machine = NULL;
2981
QEMUMachine *current_machine = NULL;
2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993

int qemu_register_machine(QEMUMachine *m)
{
    QEMUMachine **pm;
    pm = &first_machine;
    while (*pm != NULL)
        pm = &(*pm)->next;
    m->next = NULL;
    *pm = m;
    return 0;
}

2994
static QEMUMachine *find_machine(const char *name)
2995 2996 2997 2998 2999 3000
{
    QEMUMachine *m;

    for(m = first_machine; m != NULL; m = m->next) {
        if (!strcmp(m->name, name))
            return m;
M
Mark McLoughlin 已提交
3001 3002
        if (m->alias && !strcmp(m->alias, name))
            return m;
3003 3004 3005 3006
    }
    return NULL;
}

3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018
static QEMUMachine *find_default_machine(void)
{
    QEMUMachine *m;

    for(m = first_machine; m != NULL; m = m->next) {
        if (m->is_default) {
            return m;
        }
    }
    return NULL;
}

3019 3020 3021
/***********************************************************/
/* main execution loop */

3022
static void gui_update(void *opaque)
3023
{
3024
    uint64_t interval = GUI_REFRESH_INTERVAL;
3025
    DisplayState *ds = opaque;
3026 3027
    DisplayChangeListener *dcl = ds->listeners;

3028
    qemu_flush_coalesced_mmio_buffer();
3029 3030 3031 3032 3033 3034 3035 3036 3037
    dpy_refresh(ds);

    while (dcl != NULL) {
        if (dcl->gui_timer_interval &&
            dcl->gui_timer_interval < interval)
            interval = dcl->gui_timer_interval;
        dcl = dcl->next;
    }
    qemu_mod_timer(ds->gui_timer, interval + qemu_get_clock(rt_clock));
3038 3039
}

B
blueswir1 已提交
3040 3041 3042 3043
static void nographic_update(void *opaque)
{
    uint64_t interval = GUI_REFRESH_INTERVAL;

3044
    qemu_flush_coalesced_mmio_buffer();
B
blueswir1 已提交
3045 3046 3047
    qemu_mod_timer(nographic_timer, interval + qemu_get_clock(rt_clock));
}

3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074
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);
    }
}

B
bellard 已提交
3075 3076 3077
struct vm_change_state_entry {
    VMChangeStateHandler *cb;
    void *opaque;
B
Blue Swirl 已提交
3078
    QLIST_ENTRY (vm_change_state_entry) entries;
B
bellard 已提交
3079 3080
};

B
Blue Swirl 已提交
3081
static QLIST_HEAD(vm_change_state_head, vm_change_state_entry) vm_change_state_head;
B
bellard 已提交
3082 3083 3084 3085 3086 3087 3088 3089 3090 3091

VMChangeStateEntry *qemu_add_vm_change_state_handler(VMChangeStateHandler *cb,
                                                     void *opaque)
{
    VMChangeStateEntry *e;

    e = qemu_mallocz(sizeof (*e));

    e->cb = cb;
    e->opaque = opaque;
B
Blue Swirl 已提交
3092
    QLIST_INSERT_HEAD(&vm_change_state_head, e, entries);
B
bellard 已提交
3093 3094 3095 3096 3097
    return e;
}

void qemu_del_vm_change_state_handler(VMChangeStateEntry *e)
{
B
Blue Swirl 已提交
3098
    QLIST_REMOVE (e, entries);
B
bellard 已提交
3099 3100 3101
    qemu_free (e);
}

3102
static void vm_state_notify(int running, int reason)
B
bellard 已提交
3103 3104 3105 3106
{
    VMChangeStateEntry *e;

    for (e = vm_change_state_head.lh_first; e; e = e->entries.le_next) {
3107
        e->cb(e->opaque, running, reason);
B
bellard 已提交
3108 3109 3110
    }
}

3111 3112 3113
static void resume_all_vcpus(void);
static void pause_all_vcpus(void);

3114 3115 3116 3117 3118
void vm_start(void)
{
    if (!vm_running) {
        cpu_enable_ticks();
        vm_running = 1;
3119
        vm_state_notify(1, 0);
3120
        resume_all_vcpus();
3121 3122 3123
    }
}

3124 3125 3126
/* reset/shutdown handler */

typedef struct QEMUResetEntry {
B
Blue Swirl 已提交
3127
    QTAILQ_ENTRY(QEMUResetEntry) entry;
3128 3129 3130 3131
    QEMUResetHandler *func;
    void *opaque;
} QEMUResetEntry;

B
Blue Swirl 已提交
3132 3133
static QTAILQ_HEAD(reset_handlers, QEMUResetEntry) reset_handlers =
    QTAILQ_HEAD_INITIALIZER(reset_handlers);
3134 3135
static int reset_requested;
static int shutdown_requested;
B
bellard 已提交
3136
static int powerdown_requested;
3137
static int debug_requested;
3138
static int vmstop_requested;
3139

A
aurel32 已提交
3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160
int qemu_shutdown_requested(void)
{
    int r = shutdown_requested;
    shutdown_requested = 0;
    return r;
}

int qemu_reset_requested(void)
{
    int r = reset_requested;
    reset_requested = 0;
    return r;
}

int qemu_powerdown_requested(void)
{
    int r = powerdown_requested;
    powerdown_requested = 0;
    return r;
}

3161 3162 3163 3164 3165 3166 3167
static int qemu_debug_requested(void)
{
    int r = debug_requested;
    debug_requested = 0;
    return r;
}

3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179
static int qemu_vmstop_requested(void)
{
    int r = vmstop_requested;
    vmstop_requested = 0;
    return r;
}

static void do_vm_stop(int reason)
{
    if (vm_running) {
        cpu_disable_ticks();
        vm_running = 0;
3180
        pause_all_vcpus();
3181
        vm_state_notify(0, reason);
3182
        monitor_protocol_event(QEVENT_STOP, NULL);
3183 3184 3185
    }
}

3186
void qemu_register_reset(QEMUResetHandler *func, void *opaque)
3187
{
J
Jan Kiszka 已提交
3188
    QEMUResetEntry *re = qemu_mallocz(sizeof(QEMUResetEntry));
3189 3190 3191

    re->func = func;
    re->opaque = opaque;
B
Blue Swirl 已提交
3192
    QTAILQ_INSERT_TAIL(&reset_handlers, re, entry);
3193 3194
}

J
Jan Kiszka 已提交
3195
void qemu_unregister_reset(QEMUResetHandler *func, void *opaque)
3196 3197 3198
{
    QEMUResetEntry *re;

B
Blue Swirl 已提交
3199
    QTAILQ_FOREACH(re, &reset_handlers, entry) {
J
Jan Kiszka 已提交
3200
        if (re->func == func && re->opaque == opaque) {
B
Blue Swirl 已提交
3201
            QTAILQ_REMOVE(&reset_handlers, re, entry);
J
Jan Kiszka 已提交
3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212
            qemu_free(re);
            return;
        }
    }
}

void qemu_system_reset(void)
{
    QEMUResetEntry *re, *nre;

    /* reset all devices */
B
Blue Swirl 已提交
3213
    QTAILQ_FOREACH_SAFE(re, &reset_handlers, entry, nre) {
3214 3215
        re->func(re->opaque);
    }
3216
    monitor_protocol_event(QEVENT_RESET, NULL);
3217
    cpu_synchronize_all_post_reset();
3218 3219 3220 3221
}

void qemu_system_reset_request(void)
{
B
bellard 已提交
3222 3223 3224 3225 3226
    if (no_reboot) {
        shutdown_requested = 1;
    } else {
        reset_requested = 1;
    }
3227
    qemu_notify_event();
3228 3229 3230 3231 3232
}

void qemu_system_shutdown_request(void)
{
    shutdown_requested = 1;
3233
    qemu_notify_event();
3234 3235
}

B
bellard 已提交
3236 3237 3238
void qemu_system_powerdown_request(void)
{
    powerdown_requested = 1;
3239 3240 3241
    qemu_notify_event();
}

3242 3243
#ifdef CONFIG_IOTHREAD
static void qemu_system_vmstop_request(int reason)
3244
{
3245 3246
    vmstop_requested = reason;
    qemu_notify_event();
3247
}
3248
#endif
3249

3250 3251 3252 3253
#ifndef _WIN32
static int io_thread_fd = -1;

static void qemu_event_increment(void)
3254
{
P
Paolo Bonzini 已提交
3255 3256
    /* Write 8 bytes to be compatible with eventfd.  */
    static uint64_t val = 1;
3257
    ssize_t ret;
3258 3259 3260 3261

    if (io_thread_fd == -1)
        return;

3262
    do {
P
Paolo Bonzini 已提交
3263
        ret = write(io_thread_fd, &val, sizeof(val));
3264 3265 3266 3267
    } while (ret < 0 && errno == EINTR);

    /* EAGAIN is fine, a read must be pending.  */
    if (ret < 0 && errno != EAGAIN) {
3268 3269 3270 3271
        fprintf(stderr, "qemu_event_increment: write() filed: %s\n",
                strerror(errno));
        exit (1);
    }
3272 3273 3274 3275 3276 3277
}

static void qemu_event_read(void *opaque)
{
    int fd = (unsigned long)opaque;
    ssize_t len;
3278
    char buffer[512];
3279

P
Paolo Bonzini 已提交
3280
    /* Drain the notify pipe.  For eventfd, only 8 bytes will be read.  */
3281 3282
    do {
        len = read(fd, buffer, sizeof(buffer));
3283
    } while ((len == -1 && errno == EINTR) || len == sizeof(buffer));
3284 3285 3286 3287 3288 3289 3290
}

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

P
Paolo Bonzini 已提交
3291
    err = qemu_eventfd(fds);
3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306
    if (err == -1)
        return -errno;

    err = fcntl_setfl(fds[0], O_NONBLOCK);
    if (err < 0)
        goto fail;

    err = fcntl_setfl(fds[1], O_NONBLOCK);
    if (err < 0)
        goto fail;

    qemu_set_fd_handler2(fds[0], NULL, qemu_event_read, NULL,
                         (void *)(unsigned long)fds[0]);

    io_thread_fd = fds[1];
J
Jan Kiszka 已提交
3307 3308
    return 0;

3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324
fail:
    close(fds[0]);
    close(fds[1]);
    return err;
}
#else
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) {
B
Blue Swirl 已提交
3325
        fprintf(stderr, "Failed CreateEvent: %ld\n", GetLastError());
3326 3327 3328
        return -1;
    }
    qemu_add_wait_object(qemu_event_handle, dummy_event_handler, NULL);
3329 3330 3331
    return 0;
}

3332 3333
static void qemu_event_increment(void)
{
M
malc 已提交
3334
    if (!SetEvent(qemu_event_handle)) {
B
Blue Swirl 已提交
3335
        fprintf(stderr, "qemu_event_increment: SetEvent failed: %ld\n",
M
malc 已提交
3336 3337 3338
                GetLastError());
        exit (1);
    }
3339 3340 3341
}
#endif

3342 3343 3344 3345 3346 3347
static int cpu_can_run(CPUState *env)
{
    if (env->stop)
        return 0;
    if (env->stopped)
        return 0;
3348 3349
    if (!vm_running)
        return 0;
3350 3351 3352 3353
    return 1;
}

#ifndef CONFIG_IOTHREAD
3354 3355 3356 3357 3358
static int qemu_init_main_loop(void)
{
    return qemu_event_init();
}

3359 3360 3361 3362
void qemu_init_vcpu(void *_env)
{
    CPUState *env = _env;

3363 3364
    env->nr_cores = smp_cores;
    env->nr_threads = smp_threads;
J
Jiri Denemark 已提交
3365 3366
    if (kvm_enabled())
        kvm_init_vcpu(env);
3367 3368 3369
    return;
}

3370 3371 3372 3373 3374
int qemu_cpu_self(void *env)
{
    return 1;
}

3375 3376 3377 3378 3379 3380 3381 3382
static void resume_all_vcpus(void)
{
}

static void pause_all_vcpus(void)
{
}

3383 3384 3385 3386 3387
void qemu_cpu_kick(void *env)
{
    return;
}

3388 3389 3390 3391
void qemu_notify_event(void)
{
    CPUState *env = cpu_single_env;

3392
    qemu_event_increment ();
3393 3394
    if (env) {
        cpu_exit(env);
3395
    }
P
Paolo Bonzini 已提交
3396 3397 3398
    if (next_cpu && env != next_cpu) {
	cpu_exit(next_cpu);
    }
3399 3400
}

3401 3402
void qemu_mutex_lock_iothread(void) {}
void qemu_mutex_unlock_iothread(void) {}
3403

3404 3405 3406 3407 3408
void vm_stop(int reason)
{
    do_vm_stop(reason);
}

3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427
#else /* CONFIG_IOTHREAD */

#include "qemu-thread.h"

QemuMutex qemu_global_mutex;
static QemuMutex qemu_fair_mutex;

static QemuThread io_thread;

static QemuThread *tcg_cpu_thread;
static QemuCond *tcg_halt_cond;

static int qemu_system_ready;
/* cpu creation */
static QemuCond qemu_cpu_cond;
/* system init */
static QemuCond qemu_system_cond;
static QemuCond qemu_pause_cond;

3428 3429
static void tcg_block_io_signals(void);
static void kvm_block_io_signals(CPUState *env);
3430 3431
static void unblock_io_signals(void);
static int tcg_has_work(void);
M
Marcelo Tosatti 已提交
3432
static int cpu_has_work(CPUState *env);
3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452

static int qemu_init_main_loop(void)
{
    int ret;

    ret = qemu_event_init();
    if (ret)
        return ret;

    qemu_cond_init(&qemu_pause_cond);
    qemu_mutex_init(&qemu_fair_mutex);
    qemu_mutex_init(&qemu_global_mutex);
    qemu_mutex_lock(&qemu_global_mutex);

    unblock_io_signals();
    qemu_thread_self(&io_thread);

    return 0;
}

M
Marcelo Tosatti 已提交
3453 3454 3455 3456 3457 3458 3459 3460 3461
static void qemu_wait_io_event_common(CPUState *env)
{
    if (env->stop) {
        env->stop = 0;
        env->stopped = 1;
        qemu_cond_signal(&qemu_pause_cond);
    }
}

3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477
static void qemu_wait_io_event(CPUState *env)
{
    while (!tcg_has_work())
        qemu_cond_timedwait(env->halt_cond, &qemu_global_mutex, 1000);

    qemu_mutex_unlock(&qemu_global_mutex);

    /*
     * Users of qemu_global_mutex can be starved, having no chance
     * to acquire it since this path will get to it first.
     * So use another lock to provide fairness.
     */
    qemu_mutex_lock(&qemu_fair_mutex);
    qemu_mutex_unlock(&qemu_fair_mutex);

    qemu_mutex_lock(&qemu_global_mutex);
M
Marcelo Tosatti 已提交
3478 3479 3480
    qemu_wait_io_event_common(env);
}

3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501
static void qemu_kvm_eat_signal(CPUState *env, int timeout)
{
    struct timespec ts;
    int r, e;
    siginfo_t siginfo;
    sigset_t waitset;

    ts.tv_sec = timeout / 1000;
    ts.tv_nsec = (timeout % 1000) * 1000000;

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

    qemu_mutex_unlock(&qemu_global_mutex);
    r = sigtimedwait(&waitset, &siginfo, &ts);
    e = errno;
    qemu_mutex_lock(&qemu_global_mutex);

    if (r == -1 && !(e == EAGAIN || e == EINTR)) {
        fprintf(stderr, "sigtimedwait: %s\n", strerror(e));
        exit(1);
3502 3503 3504
    }
}

M
Marcelo Tosatti 已提交
3505 3506 3507 3508 3509
static void qemu_kvm_wait_io_event(CPUState *env)
{
    while (!cpu_has_work(env))
        qemu_cond_timedwait(env->halt_cond, &qemu_global_mutex, 1000);

3510
    qemu_kvm_eat_signal(env, 0);
M
Marcelo Tosatti 已提交
3511
    qemu_wait_io_event_common(env);
3512 3513 3514 3515 3516 3517 3518 3519 3520
}

static int qemu_cpu_exec(CPUState *env);

static void *kvm_cpu_thread_fn(void *arg)
{
    CPUState *env = arg;

    qemu_thread_self(env->thread);
3521 3522
    if (kvm_enabled())
        kvm_init_vcpu(env);
3523

3524 3525
    kvm_block_io_signals(env);

3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537
    /* signal CPU creation */
    qemu_mutex_lock(&qemu_global_mutex);
    env->created = 1;
    qemu_cond_signal(&qemu_cpu_cond);

    /* and wait for machine initialization */
    while (!qemu_system_ready)
        qemu_cond_timedwait(&qemu_system_cond, &qemu_global_mutex, 100);

    while (1) {
        if (cpu_can_run(env))
            qemu_cpu_exec(env);
M
Marcelo Tosatti 已提交
3538
        qemu_kvm_wait_io_event(env);
3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549
    }

    return NULL;
}

static void tcg_cpu_exec(void);

static void *tcg_cpu_thread_fn(void *arg)
{
    CPUState *env = arg;

3550
    tcg_block_io_signals();
3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575
    qemu_thread_self(env->thread);

    /* signal CPU creation */
    qemu_mutex_lock(&qemu_global_mutex);
    for (env = first_cpu; env != NULL; env = env->next_cpu)
        env->created = 1;
    qemu_cond_signal(&qemu_cpu_cond);

    /* and wait for machine initialization */
    while (!qemu_system_ready)
        qemu_cond_timedwait(&qemu_system_cond, &qemu_global_mutex, 100);

    while (1) {
        tcg_cpu_exec();
        qemu_wait_io_event(cur_cpu);
    }

    return NULL;
}

void qemu_cpu_kick(void *_env)
{
    CPUState *env = _env;
    qemu_cond_broadcast(env->halt_cond);
    if (kvm_enabled())
3576
        qemu_thread_signal(env->thread, SIG_IPI);
3577 3578
}

G
Glauber Costa 已提交
3579
int qemu_cpu_self(void *_env)
3580
{
G
Glauber Costa 已提交
3581 3582 3583 3584 3585 3586
    CPUState *env = _env;
    QemuThread this;
 
    qemu_thread_self(&this);
 
    return qemu_thread_equal(&this, env->thread);
3587 3588 3589 3590 3591 3592 3593 3594
}

static void cpu_signal(int sig)
{
    if (cpu_single_env)
        cpu_exit(cpu_single_env);
}

3595
static void tcg_block_io_signals(void)
3596 3597 3598 3599 3600 3601 3602 3603
{
    sigset_t set;
    struct sigaction sigact;

    sigemptyset(&set);
    sigaddset(&set, SIGUSR2);
    sigaddset(&set, SIGIO);
    sigaddset(&set, SIGALRM);
3604
    sigaddset(&set, SIGCHLD);
3605 3606 3607
    pthread_sigmask(SIG_BLOCK, &set, NULL);

    sigemptyset(&set);
3608
    sigaddset(&set, SIG_IPI);
3609 3610 3611 3612
    pthread_sigmask(SIG_UNBLOCK, &set, NULL);

    memset(&sigact, 0, sizeof(sigact));
    sigact.sa_handler = cpu_signal;
3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645
    sigaction(SIG_IPI, &sigact, NULL);
}

static void dummy_signal(int sig)
{
}

static void kvm_block_io_signals(CPUState *env)
{
    int r;
    sigset_t set;
    struct sigaction sigact;

    sigemptyset(&set);
    sigaddset(&set, SIGUSR2);
    sigaddset(&set, SIGIO);
    sigaddset(&set, SIGALRM);
    sigaddset(&set, SIGCHLD);
    sigaddset(&set, SIG_IPI);
    pthread_sigmask(SIG_BLOCK, &set, NULL);

    pthread_sigmask(SIG_BLOCK, NULL, &set);
    sigdelset(&set, SIG_IPI);

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

    r = kvm_set_signal_mask(env, &set);
    if (r) {
        fprintf(stderr, "kvm_set_signal_mask: %s\n", strerror(r));
        exit(1);
    }
3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658
}

static void unblock_io_signals(void)
{
    sigset_t set;

    sigemptyset(&set);
    sigaddset(&set, SIGUSR2);
    sigaddset(&set, SIGIO);
    sigaddset(&set, SIGALRM);
    pthread_sigmask(SIG_UNBLOCK, &set, NULL);

    sigemptyset(&set);
3659
    sigaddset(&set, SIG_IPI);
3660 3661 3662 3663 3664 3665 3666 3667
    pthread_sigmask(SIG_BLOCK, &set, NULL);
}

static void qemu_signal_lock(unsigned int msecs)
{
    qemu_mutex_lock(&qemu_fair_mutex);

    while (qemu_mutex_trylock(&qemu_global_mutex)) {
3668
        qemu_thread_signal(tcg_cpu_thread, SIG_IPI);
3669 3670 3671 3672 3673 3674
        if (!qemu_mutex_timedlock(&qemu_global_mutex, msecs))
            break;
    }
    qemu_mutex_unlock(&qemu_fair_mutex);
}

3675
void qemu_mutex_lock_iothread(void)
3676 3677 3678 3679 3680 3681 3682 3683 3684
{
    if (kvm_enabled()) {
        qemu_mutex_lock(&qemu_fair_mutex);
        qemu_mutex_lock(&qemu_global_mutex);
        qemu_mutex_unlock(&qemu_fair_mutex);
    } else
        qemu_signal_lock(100);
}

3685
void qemu_mutex_unlock_iothread(void)
3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708
{
    qemu_mutex_unlock(&qemu_global_mutex);
}

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

    while (penv) {
        if (!penv->stopped)
            return 0;
        penv = (CPUState *)penv->next_cpu;
    }

    return 1;
}

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

    while (penv) {
        penv->stop = 1;
3709
        qemu_thread_signal(penv->thread, SIG_IPI);
3710 3711 3712 3713 3714 3715 3716 3717
        qemu_cpu_kick(penv);
        penv = (CPUState *)penv->next_cpu;
    }

    while (!all_vcpus_paused()) {
        qemu_cond_timedwait(&qemu_pause_cond, &qemu_global_mutex, 100);
        penv = first_cpu;
        while (penv) {
3718
            qemu_thread_signal(penv->thread, SIG_IPI);
3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730
            penv = (CPUState *)penv->next_cpu;
        }
    }
}

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

    while (penv) {
        penv->stop = 0;
        penv->stopped = 0;
3731
        qemu_thread_signal(penv->thread, SIG_IPI);
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769
        qemu_cpu_kick(penv);
        penv = (CPUState *)penv->next_cpu;
    }
}

static void tcg_init_vcpu(void *_env)
{
    CPUState *env = _env;
    /* share a single thread for all cpus with TCG */
    if (!tcg_cpu_thread) {
        env->thread = qemu_mallocz(sizeof(QemuThread));
        env->halt_cond = qemu_mallocz(sizeof(QemuCond));
        qemu_cond_init(env->halt_cond);
        qemu_thread_create(env->thread, tcg_cpu_thread_fn, env);
        while (env->created == 0)
            qemu_cond_timedwait(&qemu_cpu_cond, &qemu_global_mutex, 100);
        tcg_cpu_thread = env->thread;
        tcg_halt_cond = env->halt_cond;
    } else {
        env->thread = tcg_cpu_thread;
        env->halt_cond = tcg_halt_cond;
    }
}

static void kvm_start_vcpu(CPUState *env)
{
    env->thread = qemu_mallocz(sizeof(QemuThread));
    env->halt_cond = qemu_mallocz(sizeof(QemuCond));
    qemu_cond_init(env->halt_cond);
    qemu_thread_create(env->thread, kvm_cpu_thread_fn, env);
    while (env->created == 0)
        qemu_cond_timedwait(&qemu_cpu_cond, &qemu_global_mutex, 100);
}

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

J
Jiri Denemark 已提交
3770 3771
    env->nr_cores = smp_cores;
    env->nr_threads = smp_threads;
3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805
    if (kvm_enabled())
        kvm_start_vcpu(env);
    else
        tcg_init_vcpu(env);
}

void qemu_notify_event(void)
{
    qemu_event_increment();
}

void vm_stop(int reason)
{
    QemuThread me;
    qemu_thread_self(&me);

    if (!qemu_thread_equal(&me, &io_thread)) {
        qemu_system_vmstop_request(reason);
        /*
         * FIXME: should not return to device code in case
         * vm_stop() has been requested.
         */
        if (cpu_single_env) {
            cpu_exit(cpu_single_env);
            cpu_single_env->stop = 1;
        }
        return;
    }
    do_vm_stop(reason);
}

#endif


T
ths 已提交
3806
#ifdef _WIN32
3807
static void host_main_loop_wait(int *timeout)
A
aliguori 已提交
3808 3809
{
    int ret, ret2, i;
3810 3811
    PollingEntry *pe;

3812

3813 3814 3815 3816 3817
    /* XXX: need to suppress polling by better using win32 events */
    ret = 0;
    for(pe = first_polling_entry; pe != NULL; pe = pe->next) {
        ret |= pe->func(pe->opaque);
    }
3818
    if (ret == 0) {
3819 3820
        int err;
        WaitObjects *w = &wait_objects;
3821

A
aliguori 已提交
3822
        ret = WaitForMultipleObjects(w->num, w->events, FALSE, *timeout);
3823 3824 3825
        if (WAIT_OBJECT_0 + 0 <= ret && ret <= WAIT_OBJECT_0 + w->num - 1) {
            if (w->func[ret - WAIT_OBJECT_0])
                w->func[ret - WAIT_OBJECT_0](w->opaque[ret - WAIT_OBJECT_0]);
3826

3827
            /* Check for additional signaled events */
3828
            for(i = (ret - WAIT_OBJECT_0 + 1); i < w->num; i++) {
3829

3830 3831 3832 3833 3834 3835 3836 3837 3838
                /* Check if event is signaled */
                ret2 = WaitForSingleObject(w->events[i], 0);
                if(ret2 == WAIT_OBJECT_0) {
                    if (w->func[i])
                        w->func[i](w->opaque[i]);
                } else if (ret2 == WAIT_TIMEOUT) {
                } else {
                    err = GetLastError();
                    fprintf(stderr, "WaitForSingleObject error %d %d\n", i, err);
3839 3840
                }
            }
3841 3842 3843
        } else if (ret == WAIT_TIMEOUT) {
        } else {
            err = GetLastError();
3844
            fprintf(stderr, "WaitForMultipleObjects error %d %d\n", ret, err);
3845
        }
3846
    }
A
aliguori 已提交
3847 3848 3849 3850

    *timeout = 0;
}
#else
3851
static void host_main_loop_wait(int *timeout)
A
aliguori 已提交
3852 3853
{
}
B
bellard 已提交
3854
#endif
A
aliguori 已提交
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866

void main_loop_wait(int timeout)
{
    IOHandlerRecord *ioh;
    fd_set rfds, wfds, xfds;
    int ret, nfds;
    struct timeval tv;

    qemu_bh_update_timeout(&timeout);

    host_main_loop_wait(&timeout);

B
bellard 已提交
3867 3868
    /* poll any events */
    /* XXX: separate device handlers from system ones */
3869
    nfds = -1;
B
bellard 已提交
3870 3871
    FD_ZERO(&rfds);
    FD_ZERO(&wfds);
3872
    FD_ZERO(&xfds);
B
bellard 已提交
3873
    for(ioh = first_io_handler; ioh != NULL; ioh = ioh->next) {
3874 3875
        if (ioh->deleted)
            continue;
B
bellard 已提交
3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888
        if (ioh->fd_read &&
            (!ioh->fd_read_poll ||
             ioh->fd_read_poll(ioh->opaque) != 0)) {
            FD_SET(ioh->fd, &rfds);
            if (ioh->fd > nfds)
                nfds = ioh->fd;
        }
        if (ioh->fd_write) {
            FD_SET(ioh->fd, &wfds);
            if (ioh->fd > nfds)
                nfds = ioh->fd;
        }
    }
3889

A
aliguori 已提交
3890 3891 3892
    tv.tv_sec = timeout / 1000;
    tv.tv_usec = (timeout % 1000) * 1000;

J
Jan Kiszka 已提交
3893 3894
    slirp_select_fill(&nfds, &rfds, &wfds, &xfds);

3895
    qemu_mutex_unlock_iothread();
3896
    ret = select(nfds + 1, &rfds, &wfds, &xfds, &tv);
3897
    qemu_mutex_lock_iothread();
B
bellard 已提交
3898
    if (ret > 0) {
3899 3900 3901
        IOHandlerRecord **pioh;

        for(ioh = first_io_handler; ioh != NULL; ioh = ioh->next) {
3902
            if (!ioh->deleted && ioh->fd_read && FD_ISSET(ioh->fd, &rfds)) {
B
bellard 已提交
3903
                ioh->fd_read(ioh->opaque);
B
bellard 已提交
3904
            }
3905
            if (!ioh->deleted && ioh->fd_write && FD_ISSET(ioh->fd, &wfds)) {
B
bellard 已提交
3906
                ioh->fd_write(ioh->opaque);
3907
            }
B
bellard 已提交
3908
        }
3909 3910 3911 3912 3913 3914 3915 3916

	/* remove deleted IO handlers */
	pioh = &first_io_handler;
	while (*pioh) {
            ioh = *pioh;
            if (ioh->deleted) {
                *pioh = ioh->next;
                qemu_free(ioh);
3917
            } else
3918 3919
                pioh = &ioh->next;
        }
B
bellard 已提交
3920
    }
J
Jan Kiszka 已提交
3921 3922

    slirp_select_poll(&rfds, &wfds, &xfds, (ret < 0));
B
bellard 已提交
3923

3924
    qemu_run_all_timers();
J
Jan Kiszka 已提交
3925

P
pbrook 已提交
3926 3927 3928
    /* Check bottom-halves last in case any of the earlier events triggered
       them.  */
    qemu_bh_poll();
3929

B
bellard 已提交
3930 3931
}

3932
static int qemu_cpu_exec(CPUState *env)
B
bellard 已提交
3933
{
3934
    int ret;
3935 3936 3937
#ifdef CONFIG_PROFILER
    int64_t ti;
#endif
B
bellard 已提交
3938

3939
#ifdef CONFIG_PROFILER
3940
    ti = profile_getclock();
3941
#endif
3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957
    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;
        count = qemu_next_deadline();
        count = (count + (1 << icount_time_shift) - 1)
                >> icount_time_shift;
        qemu_icount += count;
        decr = (count > 0xffff) ? 0xffff : count;
        count -= decr;
        env->icount_decr.u16.low = decr;
        env->icount_extra = count;
    }
    ret = cpu_exec(env);
3958
#ifdef CONFIG_PROFILER
3959
    qemu_time += profile_getclock() - ti;
3960
#endif
3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971
    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;
}

3972 3973
static void tcg_cpu_exec(void)
{
3974
    int ret = 0;
3975 3976 3977 3978 3979 3980

    if (next_cpu == NULL)
        next_cpu = first_cpu;
    for (; next_cpu != NULL; next_cpu = next_cpu->next_cpu) {
        CPUState *env = cur_cpu = next_cpu;

3981 3982 3983
        qemu_clock_enable(vm_clock,
                          (cur_cpu->singlestep_enabled & SSTEP_NOTIMER) == 0);

P
Paolo Bonzini 已提交
3984
        if (alarm_timer->pending)
3985
            break;
3986 3987
        if (cpu_can_run(env))
            ret = qemu_cpu_exec(env);
3988 3989 3990
        else if (env->stop)
            break;

3991 3992 3993 3994 3995 3996 3997 3998
        if (ret == EXCP_DEBUG) {
            gdb_set_stop_cpu(env);
            debug_requested = 1;
            break;
        }
    }
}

3999 4000
static int cpu_has_work(CPUState *env)
{
4001 4002 4003 4004
    if (env->stop)
        return 1;
    if (env->stopped)
        return 0;
4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023
    if (!env->halted)
        return 1;
    if (qemu_cpu_has_work(env))
        return 1;
    return 0;
}

static int tcg_has_work(void)
{
    CPUState *env;

    for (env = first_cpu; env != NULL; env = env->next_cpu)
        if (cpu_has_work(env))
            return 1;
    return 0;
}

static int qemu_calculate_timeout(void)
{
4024
#ifndef CONFIG_IOTHREAD
4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069
    int timeout;

    if (!vm_running)
        timeout = 5000;
    else if (tcg_has_work())
        timeout = 0;
    else if (!use_icount)
        timeout = 5000;
    else {
     /* XXX: use timeout computed from timers */
        int64_t add;
        int64_t delta;
        /* Advance virtual time to the next event.  */
        if (use_icount == 1) {
            /* When not using an adaptive execution frequency
               we tend to get badly out of sync with real time,
               so just delay for a reasonable amount of time.  */
            delta = 0;
        } else {
            delta = cpu_get_icount() - cpu_get_clock();
        }
        if (delta > 0) {
            /* If virtual time is ahead of real time then just
               wait for IO.  */
            timeout = (delta / 1000000) + 1;
        } else {
            /* Wait for either IO to occur or the next
               timer event.  */
            add = qemu_next_deadline();
            /* We advance the timer before checking for IO.
               Limit the amount we advance so that early IO
               activity won't get the guest too far ahead.  */
            if (add > 10000000)
                add = 10000000;
            delta += add;
            add = (add + (1 << icount_time_shift) - 1)
                  >> icount_time_shift;
            qemu_icount += add;
            timeout = delta / 1000000;
            if (timeout < 0)
                timeout = 0;
        }
    }

    return timeout;
4070 4071 4072
#else /* CONFIG_IOTHREAD */
    return 1000;
#endif
4073 4074 4075 4076 4077 4078 4079 4080 4081 4082
}

static int vm_can_run(void)
{
    if (powerdown_requested)
        return 0;
    if (reset_requested)
        return 0;
    if (shutdown_requested)
        return 0;
4083 4084
    if (debug_requested)
        return 0;
4085 4086 4087
    return 1;
}

B
Blue Swirl 已提交
4088 4089
qemu_irq qemu_system_powerdown;

4090 4091
static void main_loop(void)
{
4092
    int r;
4093

4094 4095 4096 4097 4098
#ifdef CONFIG_IOTHREAD
    qemu_system_ready = 1;
    qemu_cond_broadcast(&qemu_system_cond);
#endif

4099
    for (;;) {
4100
        do {
4101 4102 4103
#ifdef CONFIG_PROFILER
            int64_t ti;
#endif
4104
#ifndef CONFIG_IOTHREAD
4105
            tcg_cpu_exec();
4106
#endif
4107
#ifdef CONFIG_PROFILER
4108
            ti = profile_getclock();
4109
#endif
4110
            main_loop_wait(qemu_calculate_timeout());
4111
#ifdef CONFIG_PROFILER
4112
            dev_time += profile_getclock() - ti;
4113
#endif
4114
        } while (vm_can_run());
4115

4116
        if (qemu_debug_requested()) {
4117
            vm_stop(EXCP_DEBUG);
4118
        }
4119
        if (qemu_shutdown_requested()) {
4120
            monitor_protocol_event(QEVENT_SHUTDOWN, NULL);
4121 4122 4123 4124 4125 4126
            if (no_shutdown) {
                vm_stop(0);
                no_shutdown = 0;
            } else
                break;
        }
4127 4128
        if (qemu_reset_requested()) {
            pause_all_vcpus();
4129
            qemu_system_reset();
4130 4131
            resume_all_vcpus();
        }
B
Blue Swirl 已提交
4132
        if (qemu_powerdown_requested()) {
4133
            monitor_protocol_event(QEVENT_POWERDOWN, NULL);
B
Blue Swirl 已提交
4134 4135
            qemu_irq_raise(qemu_system_powerdown);
        }
4136
        if ((r = qemu_vmstop_requested())) {
4137
            vm_stop(r);
4138
        }
B
bellard 已提交
4139
    }
4140
    pause_all_vcpus();
B
bellard 已提交
4141 4142
}

P
pbrook 已提交
4143 4144
static void version(void)
{
P
pbrook 已提交
4145
    printf("QEMU PC emulator version " QEMU_VERSION QEMU_PKGVERSION ", Copyright (c) 2003-2008 Fabrice Bellard\n");
P
pbrook 已提交
4146 4147
}

4148
static void help(int exitcode)
4149
{
4150
    const char *options_help =
4151 4152 4153 4154 4155 4156 4157
#define DEF(option, opt_arg, opt_enum, opt_help)        \
           opt_help
#define DEFHEADING(text) stringify(text) "\n"
#include "qemu-options.h"
#undef DEF
#undef DEFHEADING
#undef GEN_DOCS
4158 4159 4160 4161 4162
        ;
    version();
    printf("usage: %s [options] [disk_image]\n"
           "\n"
           "'disk_image' is a raw hard image image for IDE hard disk 0\n"
4163
           "\n"
4164
           "%s\n"
4165 4166 4167 4168 4169
           "During emulation, the following keys are useful:\n"
           "ctrl-alt-f      toggle full screen\n"
           "ctrl-alt-n      switch to virtual console 'n'\n"
           "ctrl-alt        toggle mouse and keyboard grab\n"
           "\n"
4170 4171 4172
           "When using -nographic, press 'ctrl-a h' to get some help.\n",
           "qemu",
           options_help);
4173
    exit(exitcode);
4174 4175
}

4176 4177 4178
#define HAS_ARG 0x0001

enum {
4179 4180 4181 4182 4183 4184 4185
#define DEF(option, opt_arg, opt_enum, opt_help)        \
    opt_enum,
#define DEFHEADING(text)
#include "qemu-options.h"
#undef DEF
#undef DEFHEADING
#undef GEN_DOCS
4186 4187 4188 4189 4190 4191 4192 4193
};

typedef struct QEMUOption {
    const char *name;
    int flags;
    int index;
} QEMUOption;

B
blueswir1 已提交
4194
static const QEMUOption qemu_options[] = {
4195
    { "h", 0, QEMU_OPTION_h },
4196 4197 4198 4199 4200 4201 4202
#define DEF(option, opt_arg, opt_enum, opt_help)        \
    { option, opt_arg, opt_enum },
#define DEFHEADING(text)
#include "qemu-options.h"
#undef DEF
#undef DEFHEADING
#undef GEN_DOCS
4203
    { NULL },
B
bellard 已提交
4204 4205
};

4206
#ifdef HAS_AUDIO
4207
struct soundhw soundhw[] = {
4208
#ifdef HAS_AUDIO_CHOICE
A
aurel32 已提交
4209
#if defined(TARGET_I386) || defined(TARGET_MIPS)
B
bellard 已提交
4210 4211 4212 4213 4214 4215 4216 4217
    {
        "pcspk",
        "PC speaker",
        0,
        1,
        { .init_isa = pcspk_audio_init }
    },
#endif
M
malc 已提交
4218 4219

#ifdef CONFIG_SB16
4220 4221 4222 4223 4224 4225 4226
    {
        "sb16",
        "Creative Sound Blaster 16",
        0,
        1,
        { .init_isa = SB16_init }
    },
M
malc 已提交
4227
#endif
4228

M
malc 已提交
4229 4230 4231 4232 4233 4234 4235 4236 4237 4238
#ifdef CONFIG_CS4231A
    {
        "cs4231a",
        "CS4231A",
        0,
        1,
        { .init_isa = cs4231a_init }
    },
#endif

4239
#ifdef CONFIG_ADLIB
4240 4241
    {
        "adlib",
4242
#ifdef HAS_YMF262
4243
        "Yamaha YMF262 (OPL3)",
4244
#else
4245
        "Yamaha YM3812 (OPL2)",
4246
#endif
4247 4248 4249 4250
        0,
        1,
        { .init_isa = Adlib_init }
    },
4251
#endif
4252

4253
#ifdef CONFIG_GUS
4254 4255 4256 4257 4258 4259 4260
    {
        "gus",
        "Gravis Ultrasound GF1",
        0,
        1,
        { .init_isa = GUS_init }
    },
4261
#endif
4262

M
malc 已提交
4263
#ifdef CONFIG_AC97
B
balrog 已提交
4264 4265 4266 4267 4268 4269 4270
    {
        "ac97",
        "Intel 82801AA AC97 Audio",
        0,
        0,
        { .init_pci = ac97_init }
    },
M
malc 已提交
4271
#endif
B
balrog 已提交
4272

M
malc 已提交
4273
#ifdef CONFIG_ES1370
4274 4275 4276 4277 4278 4279 4280
    {
        "es1370",
        "ENSONIQ AudioPCI ES1370",
        0,
        0,
        { .init_pci = es1370_init }
    },
4281
#endif
4282

M
malc 已提交
4283 4284
#endif /* HAS_AUDIO_CHOICE */

4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299
    { NULL, NULL, 0, 0, { NULL } }
};

static void select_soundhw (const char *optarg)
{
    struct soundhw *c;

    if (*optarg == '?') {
    show_valid_cards:

        printf ("Valid sound card names (comma separated):\n");
        for (c = soundhw; c->name; ++c) {
            printf ("%-11s %s\n", c->name, c->descr);
        }
        printf ("\n-soundhw all will enable all of the above\n");
4300 4301 4302
        exit (*optarg != '?');
    }
    else {
4303
        size_t l;
4304 4305 4306 4307
        const char *p;
        char *e;
        int bad_card = 0;

4308 4309 4310 4311 4312 4313
        if (!strcmp (optarg, "all")) {
            for (c = soundhw; c->name; ++c) {
                c->enabled = 1;
            }
            return;
        }
4314

4315
        p = optarg;
4316 4317 4318
        while (*p) {
            e = strchr (p, ',');
            l = !e ? strlen (p) : (size_t) (e - p);
4319 4320

            for (c = soundhw; c->name; ++c) {
4321
                if (!strncmp (c->name, p, l) && !c->name[l]) {
4322
                    c->enabled = 1;
4323 4324 4325
                    break;
                }
            }
4326 4327

            if (!c->name) {
4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346
                if (l > 80) {
                    fprintf (stderr,
                             "Unknown sound card name (too big to show)\n");
                }
                else {
                    fprintf (stderr, "Unknown sound card name `%.*s'\n",
                             (int) l, p);
                }
                bad_card = 1;
            }
            p += l + (e != NULL);
        }

        if (bad_card)
            goto show_valid_cards;
    }
}
#endif

4347 4348 4349 4350
static void select_vgahw (const char *p)
{
    const char *opts;

G
Gerd Hoffmann 已提交
4351
    default_vga = 0;
4352
    vga_interface_type = VGA_NONE;
4353
    if (strstart(p, "std", &opts)) {
4354
        vga_interface_type = VGA_STD;
4355
    } else if (strstart(p, "cirrus", &opts)) {
4356
        vga_interface_type = VGA_CIRRUS;
4357
    } else if (strstart(p, "vmware", &opts)) {
4358
        vga_interface_type = VGA_VMWARE;
4359
    } else if (strstart(p, "xenfb", &opts)) {
4360
        vga_interface_type = VGA_XENFB;
4361
    } else if (!strstart(p, "none", &opts)) {
4362 4363 4364 4365
    invalid_vga:
        fprintf(stderr, "Unknown vga type: %s\n", p);
        exit(1);
    }
M
malc 已提交
4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378
    while (*opts) {
        const char *nextopt;

        if (strstart(opts, ",retrace=", &nextopt)) {
            opts = nextopt;
            if (strstart(opts, "dumb", &nextopt))
                vga_retrace_method = VGA_RETRACE_DUMB;
            else if (strstart(opts, "precise", &nextopt))
                vga_retrace_method = VGA_RETRACE_PRECISE;
            else goto invalid_vga;
        } else goto invalid_vga;
        opts = nextopt;
    }
4379 4380
}

4381 4382 4383
#ifdef TARGET_I386
static int balloon_parse(const char *arg)
{
4384
    QemuOpts *opts;
4385

4386 4387 4388 4389 4390 4391 4392
    if (strcmp(arg, "none") == 0) {
        return 0;
    }

    if (!strncmp(arg, "virtio", 6)) {
        if (arg[6] == ',') {
            /* have params -> parse them */
4393
            opts = qemu_opts_parse(&qemu_device_opts, arg+7, 0);
4394 4395 4396 4397 4398
            if (!opts)
                return  -1;
        } else {
            /* create empty opts */
            opts = qemu_opts_create(&qemu_device_opts, NULL, 0);
4399
        }
4400 4401
        qemu_opt_set(opts, "driver", "virtio-balloon-pci");
        return 0;
4402
    }
4403 4404

    return -1;
4405 4406 4407
}
#endif

4408 4409 4410 4411 4412 4413 4414 4415
#ifdef _WIN32
static BOOL WINAPI qemu_ctrl_handler(DWORD type)
{
    exit(STATUS_CONTROL_C_EXIT);
    return TRUE;
}
#endif

4416
int qemu_uuid_parse(const char *str, uint8_t *uuid)
4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429
{
    int ret;

    if(strlen(str) != 36)
        return -1;

    ret = sscanf(str, UUID_FMT, &uuid[0], &uuid[1], &uuid[2], &uuid[3],
            &uuid[4], &uuid[5], &uuid[6], &uuid[7], &uuid[8], &uuid[9],
            &uuid[10], &uuid[11], &uuid[12], &uuid[13], &uuid[14], &uuid[15]);

    if(ret != 16)
        return -1;

4430 4431 4432 4433
#ifdef TARGET_I386
    smbios_add_field(1, offsetof(struct smbios_type_1, uuid), 16, uuid);
#endif

4434 4435 4436
    return 0;
}

4437 4438 4439 4440 4441 4442 4443
#ifndef _WIN32

static void termsig_handler(int signal)
{
    qemu_system_shutdown_request();
}

4444 4445 4446 4447 4448 4449
static void sigchld_handler(int signal)
{
    waitpid(-1, NULL, WNOHANG);
}

static void sighandler_setup(void)
4450 4451 4452 4453 4454 4455 4456 4457
{
    struct sigaction act;

    memset(&act, 0, sizeof(act));
    act.sa_handler = termsig_handler;
    sigaction(SIGINT,  &act, NULL);
    sigaction(SIGHUP,  &act, NULL);
    sigaction(SIGTERM, &act, NULL);
4458 4459 4460 4461

    act.sa_handler = sigchld_handler;
    act.sa_flags = SA_NOCLDSTOP;
    sigaction(SIGCHLD, &act, NULL);
4462 4463 4464 4465
}

#endif

P
Paul Brook 已提交
4466 4467 4468 4469 4470 4471 4472 4473 4474 4475
#ifdef _WIN32
/* Look for support files in the same directory as the executable.  */
static char *find_datadir(const char *argv0)
{
    char *p;
    char buf[MAX_PATH];
    DWORD len;

    len = GetModuleFileName(NULL, buf, sizeof(buf) - 1);
    if (len == 0) {
B
Blue Swirl 已提交
4476
        return NULL;
P
Paul Brook 已提交
4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501
    }

    buf[len] = 0;
    p = buf + len - 1;
    while (p != buf && *p != '\\')
        p--;
    *p = 0;
    if (access(buf, R_OK) == 0) {
        return qemu_strdup(buf);
    }
    return NULL;
}
#else /* !_WIN32 */

/* Find a likely location for support files using the location of the binary.
   For installed binaries this will be "$bindir/../share/qemu".  When
   running from the build tree this will be "$bindir/../pc-bios".  */
#define SHARE_SUFFIX "/share/qemu"
#define BUILD_SUFFIX "/pc-bios"
static char *find_datadir(const char *argv0)
{
    char *dir;
    char *p = NULL;
    char *res;
    char buf[PATH_MAX];
4502
    size_t max_len;
P
Paul Brook 已提交
4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525

#if defined(__linux__)
    {
        int len;
        len = readlink("/proc/self/exe", buf, sizeof(buf) - 1);
        if (len > 0) {
            buf[len] = 0;
            p = buf;
        }
    }
#elif defined(__FreeBSD__)
    {
        int len;
        len = readlink("/proc/curproc/file", buf, sizeof(buf) - 1);
        if (len > 0) {
            buf[len] = 0;
            p = buf;
        }
    }
#endif
    /* If we don't have any way of figuring out the actual executable
       location then try argv[0].  */
    if (!p) {
4526
        p = realpath(argv0, buf);
P
Paul Brook 已提交
4527 4528 4529 4530 4531 4532 4533
        if (!p) {
            return NULL;
        }
    }
    dir = dirname(p);
    dir = dirname(dir);

4534 4535 4536 4537
    max_len = strlen(dir) +
        MAX(strlen(SHARE_SUFFIX), strlen(BUILD_SUFFIX)) + 1;
    res = qemu_mallocz(max_len);
    snprintf(res, max_len, "%s%s", dir, SHARE_SUFFIX);
P
Paul Brook 已提交
4538
    if (access(res, R_OK)) {
4539
        snprintf(res, max_len, "%s%s", dir, BUILD_SUFFIX);
P
Paul Brook 已提交
4540 4541 4542 4543 4544
        if (access(res, R_OK)) {
            qemu_free(res);
            res = NULL;
        }
    }
4545

P
Paul Brook 已提交
4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560
    return res;
}
#undef SHARE_SUFFIX
#undef BUILD_SUFFIX
#endif

char *qemu_find_file(int type, const char *name)
{
    int len;
    const char *subdir;
    char *buf;

    /* If name contains path separators then try it as a straight path.  */
    if ((strchr(name, '/') || strchr(name, '\\'))
        && access(name, R_OK) == 0) {
4561
        return qemu_strdup(name);
P
Paul Brook 已提交
4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574
    }
    switch (type) {
    case QEMU_FILE_TYPE_BIOS:
        subdir = "";
        break;
    case QEMU_FILE_TYPE_KEYMAP:
        subdir = "keymaps/";
        break;
    default:
        abort();
    }
    len = strlen(data_dir) + strlen(name) + strlen(subdir) + 2;
    buf = qemu_mallocz(len);
4575
    snprintf(buf, len, "%s/%s%s", data_dir, subdir, name);
P
Paul Brook 已提交
4576 4577 4578 4579 4580 4581 4582
    if (access(buf, R_OK)) {
        qemu_free(buf);
        return NULL;
    }
    return buf;
}

4583 4584 4585 4586 4587
static int device_help_func(QemuOpts *opts, void *opaque)
{
    return qdev_device_help(opts);
}

G
Gerd Hoffmann 已提交
4588 4589 4590 4591 4592 4593 4594 4595 4596 4597
static int device_init_func(QemuOpts *opts, void *opaque)
{
    DeviceState *dev;

    dev = qdev_device_add(opts);
    if (!dev)
        return -1;
    return 0;
}

4598 4599 4600 4601 4602 4603 4604 4605 4606 4607
static int chardev_init_func(QemuOpts *opts, void *opaque)
{
    CharDriverState *chr;

    chr = qemu_chr_open_opts(opts, NULL);
    if (!chr)
        return -1;
    return 0;
}

4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641
static int mon_init_func(QemuOpts *opts, void *opaque)
{
    CharDriverState *chr;
    const char *chardev;
    const char *mode;
    int flags;

    mode = qemu_opt_get(opts, "mode");
    if (mode == NULL) {
        mode = "readline";
    }
    if (strcmp(mode, "readline") == 0) {
        flags = MONITOR_USE_READLINE;
    } else if (strcmp(mode, "control") == 0) {
        flags = MONITOR_USE_CONTROL;
    } else {
        fprintf(stderr, "unknown monitor mode \"%s\"\n", mode);
        exit(1);
    }

    if (qemu_opt_get_bool(opts, "default", 0))
        flags |= MONITOR_IS_DEFAULT;

    chardev = qemu_opt_get(opts, "chardev");
    chr = qemu_chr_find(chardev);
    if (chr == NULL) {
        fprintf(stderr, "chardev \"%s\" not found\n", chardev);
        exit(1);
    }

    monitor_init(chr, flags);
    return 0;
}

G
Gerd Hoffmann 已提交
4642
static void monitor_parse(const char *optarg, const char *mode)
4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671
{
    static int monitor_device_index = 0;
    QemuOpts *opts;
    const char *p;
    char label[32];
    int def = 0;

    if (strstart(optarg, "chardev:", &p)) {
        snprintf(label, sizeof(label), "%s", p);
    } else {
        if (monitor_device_index) {
            snprintf(label, sizeof(label), "monitor%d",
                     monitor_device_index);
        } else {
            snprintf(label, sizeof(label), "monitor");
            def = 1;
        }
        opts = qemu_chr_parse_compat(label, optarg);
        if (!opts) {
            fprintf(stderr, "parse error: %s\n", optarg);
            exit(1);
        }
    }

    opts = qemu_opts_create(&qemu_mon_opts, label, 1);
    if (!opts) {
        fprintf(stderr, "duplicate chardev: %s\n", label);
        exit(1);
    }
G
Gerd Hoffmann 已提交
4672
    qemu_opt_set(opts, "mode", mode);
4673 4674 4675 4676 4677 4678
    qemu_opt_set(opts, "chardev", label);
    if (def)
        qemu_opt_set(opts, "default", "on");
    monitor_device_index++;
}

4679 4680
struct device_config {
    enum {
G
Gerd Hoffmann 已提交
4681 4682 4683 4684 4685
        DEV_USB,       /* -usbdevice     */
        DEV_BT,        /* -bt            */
        DEV_SERIAL,    /* -serial        */
        DEV_PARALLEL,  /* -parallel      */
        DEV_VIRTCON,   /* -virtioconsole */
4686
        DEV_DEBUGCON,  /* -debugcon */
4687 4688
    } type;
    const char *cmdline;
B
Blue Swirl 已提交
4689
    QTAILQ_ENTRY(device_config) next;
4690
};
B
Blue Swirl 已提交
4691
QTAILQ_HEAD(, device_config) device_configs = QTAILQ_HEAD_INITIALIZER(device_configs);
4692 4693 4694 4695 4696 4697 4698 4699

static void add_device_config(int type, const char *cmdline)
{
    struct device_config *conf;

    conf = qemu_mallocz(sizeof(*conf));
    conf->type = type;
    conf->cmdline = cmdline;
B
Blue Swirl 已提交
4700
    QTAILQ_INSERT_TAIL(&device_configs, conf, next);
4701 4702 4703 4704 4705 4706 4707
}

static int foreach_device_config(int type, int (*func)(const char *cmdline))
{
    struct device_config *conf;
    int rc;

B
Blue Swirl 已提交
4708
    QTAILQ_FOREACH(conf, &device_configs, next) {
4709 4710 4711 4712 4713 4714 4715 4716 4717
        if (conf->type != type)
            continue;
        rc = func(conf->cmdline);
        if (0 != rc)
            return rc;
    }
    return 0;
}

4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739
static int serial_parse(const char *devname)
{
    static int index = 0;
    char label[32];

    if (strcmp(devname, "none") == 0)
        return 0;
    if (index == MAX_SERIAL_PORTS) {
        fprintf(stderr, "qemu: too many serial ports\n");
        exit(1);
    }
    snprintf(label, sizeof(label), "serial%d", index);
    serial_hds[index] = qemu_chr_open(label, devname, NULL);
    if (!serial_hds[index]) {
        fprintf(stderr, "qemu: could not open serial device '%s': %s\n",
                devname, strerror(errno));
        return -1;
    }
    index++;
    return 0;
}

G
Gerd Hoffmann 已提交
4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761
static int parallel_parse(const char *devname)
{
    static int index = 0;
    char label[32];

    if (strcmp(devname, "none") == 0)
        return 0;
    if (index == MAX_PARALLEL_PORTS) {
        fprintf(stderr, "qemu: too many parallel ports\n");
        exit(1);
    }
    snprintf(label, sizeof(label), "parallel%d", index);
    parallel_hds[index] = qemu_chr_open(label, devname, NULL);
    if (!parallel_hds[index]) {
        fprintf(stderr, "qemu: could not open parallel device '%s': %s\n",
                devname, strerror(errno));
        return -1;
    }
    index++;
    return 0;
}

G
Gerd Hoffmann 已提交
4762 4763 4764 4765
static int virtcon_parse(const char *devname)
{
    static int index = 0;
    char label[32];
4766
    QemuOpts *bus_opts, *dev_opts;
G
Gerd Hoffmann 已提交
4767 4768 4769 4770 4771 4772 4773

    if (strcmp(devname, "none") == 0)
        return 0;
    if (index == MAX_VIRTIO_CONSOLES) {
        fprintf(stderr, "qemu: too many virtio consoles\n");
        exit(1);
    }
4774 4775 4776 4777 4778 4779 4780

    bus_opts = qemu_opts_create(&qemu_device_opts, NULL, 0);
    qemu_opt_set(bus_opts, "driver", "virtio-serial");

    dev_opts = qemu_opts_create(&qemu_device_opts, NULL, 0);
    qemu_opt_set(dev_opts, "driver", "virtconsole");

G
Gerd Hoffmann 已提交
4781 4782 4783 4784 4785 4786 4787
    snprintf(label, sizeof(label), "virtcon%d", index);
    virtcon_hds[index] = qemu_chr_open(label, devname, NULL);
    if (!virtcon_hds[index]) {
        fprintf(stderr, "qemu: could not open virtio console '%s': %s\n",
                devname, strerror(errno));
        return -1;
    }
4788 4789
    qemu_opt_set(dev_opts, "chardev", label);

G
Gerd Hoffmann 已提交
4790 4791 4792 4793
    index++;
    return 0;
}

4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810
static int debugcon_parse(const char *devname)
{   
    QemuOpts *opts;

    if (!qemu_chr_open("debugcon", devname, NULL)) {
        exit(1);
    }
    opts = qemu_opts_create(&qemu_device_opts, "debugcon", 1);
    if (!opts) {
        fprintf(stderr, "qemu: already have a debugcon device\n");
        exit(1);
    }
    qemu_opt_set(opts, "driver", "isa-debugcon");
    qemu_opt_set(opts, "chardev", "debugcon");
    return 0;
}

4811 4812 4813 4814 4815 4816 4817 4818
static const QEMUOption *lookup_opt(int argc, char **argv,
                                    const char **poptarg, int *poptind)
{
    const QEMUOption *popt;
    int optind = *poptind;
    char *r = argv[optind];
    const char *optarg;

4819
    loc_set_cmdline(argv, optind, 1);
4820 4821 4822 4823 4824 4825 4826
    optind++;
    /* Treat --foo the same as -foo.  */
    if (r[1] == '-')
        r++;
    popt = qemu_options;
    for(;;) {
        if (!popt->name) {
4827
            error_report("invalid option");
4828 4829 4830 4831 4832 4833 4834 4835
            exit(1);
        }
        if (!strcmp(popt->name, r + 1))
            break;
        popt++;
    }
    if (popt->flags & HAS_ARG) {
        if (optind >= argc) {
4836
            error_report("requires an argument");
4837 4838 4839
            exit(1);
        }
        optarg = argv[optind++];
4840
        loc_set_cmdline(argv, optind - 2, 2);
4841 4842 4843 4844 4845 4846 4847 4848 4849 4850
    } else {
        optarg = NULL;
    }

    *poptarg = optarg;
    *poptind = optind;

    return popt;
}

M
malc 已提交
4851
int main(int argc, char **argv, char **envp)
4852
{
4853
    const char *gdbstub_dev = NULL;
4854
    uint32_t boot_devices_bitmap = 0;
T
ths 已提交
4855
    int i;
4856
    int snapshot, linux_boot, net_boot;
B
bellard 已提交
4857
    const char *initrd_filename;
4858
    const char *kernel_filename, *kernel_cmdline;
4859
    char boot_devices[33] = "cad"; /* default to HD->floppy->CD-ROM */
4860
    DisplayState *ds;
4861
    DisplayChangeListener *dcl;
B
bellard 已提交
4862
    int cyls, heads, secs, translation;
G
Gerd Hoffmann 已提交
4863
    QemuOpts *hda_opts = NULL, *opts;
4864
    int optind;
4865
    const char *optarg;
B
bellard 已提交
4866
    const char *loadvm = NULL;
4867
    QEMUMachine *machine;
4868
    const char *cpu_model;
B
blueswir1 已提交
4869
#ifndef _WIN32
T
ths 已提交
4870
    int fds[2];
B
blueswir1 已提交
4871
#endif
4872
    int tb_size;
4873
    const char *pid_file = NULL;
A
aliguori 已提交
4874
    const char *incoming = NULL;
B
blueswir1 已提交
4875
#ifndef _WIN32
4876 4877
    int fd = 0;
    struct passwd *pwd = NULL;
4878 4879
    const char *chroot_dir = NULL;
    const char *run_as = NULL;
B
blueswir1 已提交
4880
#endif
4881
    CPUState *env;
4882
    int show_vnc_port = 0;
4883
    int defconfig = 1;
B
bellard 已提交
4884

4885 4886
    error_set_progname(argv[0]);

J
Jan Kiszka 已提交
4887 4888
    init_clocks();

M
malc 已提交
4889 4890
    qemu_cache_utils_init(envp);

B
Blue Swirl 已提交
4891
    QLIST_INIT (&vm_change_state_head);
4892 4893 4894 4895 4896 4897 4898 4899
#ifndef _WIN32
    {
        struct sigaction act;
        sigfillset(&act.sa_mask);
        act.sa_flags = 0;
        act.sa_handler = SIG_IGN;
        sigaction(SIGPIPE, &act, NULL);
    }
4900 4901
#else
    SetConsoleCtrlHandler(qemu_ctrl_handler, TRUE);
4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919
    /* Note: cpu_interrupt() is currently not SMP safe, so we force
       QEMU to run on a single CPU */
    {
        HANDLE h;
        DWORD mask, smask;
        int i;
        h = GetCurrentProcess();
        if (GetProcessAffinityMask(h, &mask, &smask)) {
            for(i = 0; i < 32; i++) {
                if (mask & (1 << i))
                    break;
            }
            if (i != 32) {
                mask = 1 << i;
                SetProcessAffinityMask(h, mask);
            }
        }
    }
B
bellard 已提交
4920
#endif
4921

4922
    module_call_init(MODULE_INIT_MACHINE);
4923
    machine = find_default_machine();
4924
    cpu_model = NULL;
B
bellard 已提交
4925
    initrd_filename = NULL;
4926
    ram_size = 0;
4927
    snapshot = 0;
4928 4929
    kernel_filename = NULL;
    kernel_cmdline = "";
4930
    cyls = heads = secs = 0;
B
bellard 已提交
4931
    translation = BIOS_ATA_TRANSLATION_AUTO;
4932

4933 4934 4935 4936 4937 4938
    for (i = 0; i < MAX_NODES; i++) {
        node_mem[i] = 0;
        node_cpumask[i] = 0;
    }

    nb_numa_nodes = 0;
B
bellard 已提交
4939
    nb_nics = 0;
4940

4941
    tb_size = 0;
4942 4943
    autostart= 1;

4944 4945 4946 4947 4948
    /* first pass of option parsing */
    optind = 1;
    while (optind < argc) {
        if (argv[optind][0] != '-') {
            /* disk image */
4949
            optind++;
4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963
            continue;
        } else {
            const QEMUOption *popt;

            popt = lookup_opt(argc, argv, &optarg, &optind);
            switch (popt->index) {
            case QEMU_OPTION_nodefconfig:
                defconfig=0;
                break;
            }
        }
    }

    if (defconfig) {
4964
        const char *fname;
4965
        FILE *fp;
4966 4967 4968

        fname = CONFIG_QEMU_CONFDIR "/qemu.conf";
        fp = fopen(fname, "r");
4969
        if (fp) {
4970
            if (qemu_config_parse(fp, fname) != 0) {
4971 4972 4973 4974 4975
                exit(1);
            }
            fclose(fp);
        }

4976 4977
        fname = CONFIG_QEMU_CONFDIR "/target-" TARGET_ARCH ".conf";
        fp = fopen(fname, "r");
4978
        if (fp) {
4979
            if (qemu_config_parse(fp, fname) != 0) {
4980 4981 4982 4983 4984
                exit(1);
            }
            fclose(fp);
        }
    }
4985 4986 4987
#if defined(cpudef_setup)
    cpudef_setup(); /* parse cpu definitions in target config file */
#endif
4988 4989

    /* second pass of option parsing */
4990
    optind = 1;
4991
    for(;;) {
4992
        if (optind >= argc)
4993
            break;
4994
        if (argv[optind][0] != '-') {
G
Gerd Hoffmann 已提交
4995
	    hda_opts = drive_add(argv[optind++], HD_ALIAS, 0);
4996 4997 4998
        } else {
            const QEMUOption *popt;

4999
            popt = lookup_opt(argc, argv, &optarg, &optind);
5000
            switch(popt->index) {
5001 5002 5003 5004 5005 5006
            case QEMU_OPTION_M:
                machine = find_machine(optarg);
                if (!machine) {
                    QEMUMachine *m;
                    printf("Supported machines are:\n");
                    for(m = first_machine; m != NULL; m = m->next) {
M
Mark McLoughlin 已提交
5007 5008 5009
                        if (m->alias)
                            printf("%-10s %s (alias of %s)\n",
                                   m->alias, m->desc, m->name);
5010
                        printf("%-10s %s%s\n",
5011
                               m->name, m->desc,
5012
                               m->is_default ? " (default)" : "");
5013
                    }
5014
                    exit(*optarg != '?');
5015 5016
                }
                break;
5017 5018
            case QEMU_OPTION_cpu:
                /* hw initialization will check this */
5019
                if (*optarg == '?') {
J
j_mayer 已提交
5020
/* XXX: implement xxx_cpu_list for targets that still miss it */
5021 5022 5023 5024
#if defined(cpu_list_id)
                    cpu_list_id(stdout, &fprintf, optarg);
#elif defined(cpu_list)
                    cpu_list(stdout, &fprintf);	        /* deprecated */
5025
#endif
5026
                    exit(0);
5027 5028 5029 5030
                } else {
                    cpu_model = optarg;
                }
                break;
5031
            case QEMU_OPTION_initrd:
B
bellard 已提交
5032 5033
                initrd_filename = optarg;
                break;
5034
            case QEMU_OPTION_hda:
T
ths 已提交
5035
                if (cyls == 0)
G
Gerd Hoffmann 已提交
5036
                    hda_opts = drive_add(optarg, HD_ALIAS, 0);
T
ths 已提交
5037
                else
G
Gerd Hoffmann 已提交
5038
                    hda_opts = drive_add(optarg, HD_ALIAS
T
ths 已提交
5039
			     ",cyls=%d,heads=%d,secs=%d%s",
5040
                             0, cyls, heads, secs,
T
ths 已提交
5041 5042 5043 5044 5045
                             translation == BIOS_ATA_TRANSLATION_LBA ?
                                 ",trans=lba" :
                             translation == BIOS_ATA_TRANSLATION_NONE ?
                                 ",trans=none" : "");
                 break;
5046
            case QEMU_OPTION_hdb:
5047 5048
            case QEMU_OPTION_hdc:
            case QEMU_OPTION_hdd:
5049
                drive_add(optarg, HD_ALIAS, popt->index - QEMU_OPTION_hda);
B
bellard 已提交
5050
                break;
T
ths 已提交
5051
            case QEMU_OPTION_drive:
5052
                drive_add(NULL, "%s", optarg);
T
ths 已提交
5053
	        break;
G
Gerd Hoffmann 已提交
5054 5055 5056 5057
            case QEMU_OPTION_set:
                if (qemu_set_option(optarg) != 0)
                    exit(1);
	        break;
5058 5059 5060 5061
            case QEMU_OPTION_global:
                if (qemu_global_option(optarg) != 0)
                    exit(1);
	        break;
5062
            case QEMU_OPTION_mtdblock:
5063
                drive_add(optarg, MTD_ALIAS);
5064
                break;
5065
            case QEMU_OPTION_sd:
5066
                drive_add(optarg, SD_ALIAS);
5067
                break;
5068
            case QEMU_OPTION_pflash:
5069
                drive_add(optarg, PFLASH_ALIAS);
5070
                break;
5071
            case QEMU_OPTION_snapshot:
5072 5073
                snapshot = 1;
                break;
5074
            case QEMU_OPTION_hdachs:
5075 5076 5077 5078
                {
                    const char *p;
                    p = optarg;
                    cyls = strtol(p, (char **)&p, 0);
B
bellard 已提交
5079 5080
                    if (cyls < 1 || cyls > 16383)
                        goto chs_fail;
5081 5082 5083 5084
                    if (*p != ',')
                        goto chs_fail;
                    p++;
                    heads = strtol(p, (char **)&p, 0);
B
bellard 已提交
5085 5086
                    if (heads < 1 || heads > 16)
                        goto chs_fail;
5087 5088 5089 5090
                    if (*p != ',')
                        goto chs_fail;
                    p++;
                    secs = strtol(p, (char **)&p, 0);
B
bellard 已提交
5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103
                    if (secs < 1 || secs > 63)
                        goto chs_fail;
                    if (*p == ',') {
                        p++;
                        if (!strcmp(p, "none"))
                            translation = BIOS_ATA_TRANSLATION_NONE;
                        else if (!strcmp(p, "lba"))
                            translation = BIOS_ATA_TRANSLATION_LBA;
                        else if (!strcmp(p, "auto"))
                            translation = BIOS_ATA_TRANSLATION_AUTO;
                        else
                            goto chs_fail;
                    } else if (*p != '\0') {
5104
                    chs_fail:
B
bellard 已提交
5105 5106
                        fprintf(stderr, "qemu: invalid physical CHS format\n");
                        exit(1);
5107
                    }
G
Gerd Hoffmann 已提交
5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120
		    if (hda_opts != NULL) {
                        char num[16];
                        snprintf(num, sizeof(num), "%d", cyls);
                        qemu_opt_set(hda_opts, "cyls", num);
                        snprintf(num, sizeof(num), "%d", heads);
                        qemu_opt_set(hda_opts, "heads", num);
                        snprintf(num, sizeof(num), "%d", secs);
                        qemu_opt_set(hda_opts, "secs", num);
                        if (translation == BIOS_ATA_TRANSLATION_LBA)
                            qemu_opt_set(hda_opts, "trans", "lba");
                        if (translation == BIOS_ATA_TRANSLATION_NONE)
                            qemu_opt_set(hda_opts, "trans", "none");
                    }
5121 5122
                }
                break;
5123 5124 5125 5126 5127 5128 5129
            case QEMU_OPTION_numa:
                if (nb_numa_nodes >= MAX_NODES) {
                    fprintf(stderr, "qemu: too many NUMA nodes\n");
                    exit(1);
                }
                numa_add(optarg);
                break;
5130
            case QEMU_OPTION_nographic:
5131
                display_type = DT_NOGRAPHIC;
5132
                break;
B
balrog 已提交
5133 5134
#ifdef CONFIG_CURSES
            case QEMU_OPTION_curses:
5135
                display_type = DT_CURSES;
B
balrog 已提交
5136 5137
                break;
#endif
5138 5139 5140
            case QEMU_OPTION_portrait:
                graphic_rotate = 1;
                break;
5141
            case QEMU_OPTION_kernel:
5142 5143
                kernel_filename = optarg;
                break;
5144
            case QEMU_OPTION_append:
5145
                kernel_cmdline = optarg;
5146
                break;
5147
            case QEMU_OPTION_cdrom:
5148
                drive_add(optarg, CDROM_ALIAS);
5149
                break;
5150
            case QEMU_OPTION_boot:
5151
                {
J
Jan Kiszka 已提交
5152
                    static const char * const params[] = {
5153
                        "order", "once", "menu", NULL
J
Jan Kiszka 已提交
5154 5155
                    };
                    char buf[sizeof(boot_devices)];
J
Jan Kiszka 已提交
5156
                    char *standard_boot_devices;
J
Jan Kiszka 已提交
5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172
                    int legacy = 0;

                    if (!strchr(optarg, '=')) {
                        legacy = 1;
                        pstrcpy(buf, sizeof(buf), optarg);
                    } else if (check_params(buf, sizeof(buf), params, optarg) < 0) {
                        fprintf(stderr,
                                "qemu: unknown boot parameter '%s' in '%s'\n",
                                buf, optarg);
                        exit(1);
                    }

                    if (legacy ||
                        get_param_value(buf, sizeof(buf), "order", optarg)) {
                        boot_devices_bitmap = parse_bootdevices(buf);
                        pstrcpy(boot_devices, sizeof(boot_devices), buf);
5173
                    }
J
Jan Kiszka 已提交
5174 5175 5176 5177 5178 5179 5180 5181 5182
                    if (!legacy) {
                        if (get_param_value(buf, sizeof(buf),
                                            "once", optarg)) {
                            boot_devices_bitmap |= parse_bootdevices(buf);
                            standard_boot_devices = qemu_strdup(boot_devices);
                            pstrcpy(boot_devices, sizeof(boot_devices), buf);
                            qemu_register_reset(restore_boot_devices,
                                                standard_boot_devices);
                        }
5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195
                        if (get_param_value(buf, sizeof(buf),
                                            "menu", optarg)) {
                            if (!strcmp(buf, "on")) {
                                boot_menu = 1;
                            } else if (!strcmp(buf, "off")) {
                                boot_menu = 0;
                            } else {
                                fprintf(stderr,
                                        "qemu: invalid option value '%s'\n",
                                        buf);
                                exit(1);
                            }
                        }
J
Jan Kiszka 已提交
5196
                    }
5197 5198
                }
                break;
5199 5200
            case QEMU_OPTION_fda:
            case QEMU_OPTION_fdb:
5201
                drive_add(optarg, FD_ALIAS, popt->index - QEMU_OPTION_fda);
5202
                break;
B
bellard 已提交
5203 5204 5205 5206 5207
#ifdef TARGET_I386
            case QEMU_OPTION_no_fd_bootchk:
                fd_bootchk = 0;
                break;
#endif
M
Mark McLoughlin 已提交
5208 5209 5210 5211 5212
            case QEMU_OPTION_netdev:
                if (net_client_parse(&qemu_netdev_opts, optarg) == -1) {
                    exit(1);
                }
                break;
B
bellard 已提交
5213
            case QEMU_OPTION_net:
5214
                if (net_client_parse(&qemu_net_opts, optarg) == -1) {
5215 5216
                    exit(1);
                }
B
bellard 已提交
5217
                break;
B
bellard 已提交
5218 5219
#ifdef CONFIG_SLIRP
            case QEMU_OPTION_tftp:
5220
                legacy_tftp_prefix = optarg;
B
bellard 已提交
5221
                break;
5222
            case QEMU_OPTION_bootp:
5223
                legacy_bootp_filename = optarg;
5224
                break;
B
bellard 已提交
5225
#ifndef _WIN32
B
bellard 已提交
5226
            case QEMU_OPTION_smb:
M
Markus Armbruster 已提交
5227 5228
                if (net_slirp_smb(optarg) < 0)
                    exit(1);
B
bellard 已提交
5229
                break;
B
bellard 已提交
5230
#endif
B
bellard 已提交
5231
            case QEMU_OPTION_redir:
M
Markus Armbruster 已提交
5232 5233
                if (net_slirp_redir(optarg) < 0)
                    exit(1);
B
bellard 已提交
5234
                break;
B
bellard 已提交
5235
#endif
5236
            case QEMU_OPTION_bt:
5237
                add_device_config(DEV_BT, optarg);
5238
                break;
5239 5240 5241 5242 5243 5244 5245 5246 5247
#ifdef HAS_AUDIO
            case QEMU_OPTION_audio_help:
                AUD_help ();
                exit (0);
                break;
            case QEMU_OPTION_soundhw:
                select_soundhw (optarg);
                break;
#endif
5248
            case QEMU_OPTION_h:
5249
                help(0);
5250
                break;
P
pbrook 已提交
5251 5252 5253 5254
            case QEMU_OPTION_version:
                version();
                exit(0);
                break;
5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268
            case QEMU_OPTION_m: {
                uint64_t value;
                char *ptr;

                value = strtoul(optarg, &ptr, 10);
                switch (*ptr) {
                case 0: case 'M': case 'm':
                    value <<= 20;
                    break;
                case 'G': case 'g':
                    value <<= 30;
                    break;
                default:
                    fprintf(stderr, "qemu: invalid ram size: %s\n", optarg);
5269 5270
                    exit(1);
                }
5271 5272

                /* On 32-bit hosts, QEMU is limited by virtual address space */
5273
                if (value > (2047 << 20) && HOST_LONG_BITS == 32) {
5274 5275 5276
                    fprintf(stderr, "qemu: at most 2047 MB RAM can be simulated\n");
                    exit(1);
                }
A
Anthony Liguori 已提交
5277
                if (value != (uint64_t)(ram_addr_t)value) {
5278 5279 5280 5281
                    fprintf(stderr, "qemu: ram size too large\n");
                    exit(1);
                }
                ram_size = value;
5282
                break;
5283
            }
5284 5285 5286 5287 5288 5289 5290 5291
            case QEMU_OPTION_mempath:
                mem_path = optarg;
                break;
#ifdef MAP_POPULATE
            case QEMU_OPTION_mem_prealloc:
                mem_prealloc = 1;
                break;
#endif
5292 5293 5294
            case QEMU_OPTION_d:
                {
                    int mask;
B
blueswir1 已提交
5295
                    const CPULogItem *item;
5296

5297 5298 5299
                    mask = cpu_str_to_log_mask(optarg);
                    if (!mask) {
                        printf("Log items (comma separated):\n");
5300 5301 5302 5303
                    for(item = cpu_log_items; item->mask != 0; item++) {
                        printf("%-10s %s\n", item->name, item->help);
                    }
                    exit(1);
5304 5305
                    }
                    cpu_set_log(mask);
5306
                }
5307 5308
                break;
            case QEMU_OPTION_s:
5309
                gdbstub_dev = "tcp::" DEFAULT_GDBSTUB_PORT;
5310
                break;
5311 5312
            case QEMU_OPTION_gdb:
                gdbstub_dev = optarg;
5313 5314
                break;
            case QEMU_OPTION_L:
P
Paul Brook 已提交
5315
                data_dir = optarg;
5316
                break;
5317 5318 5319
            case QEMU_OPTION_bios:
                bios_name = optarg;
                break;
5320 5321 5322
            case QEMU_OPTION_singlestep:
                singlestep = 1;
                break;
5323
            case QEMU_OPTION_S:
5324
                autostart = 0;
5325
                break;
5326 5327 5328
	    case QEMU_OPTION_k:
		keyboard_layout = optarg;
		break;
B
bellard 已提交
5329 5330 5331
            case QEMU_OPTION_localtime:
                rtc_utc = 0;
                break;
5332 5333
            case QEMU_OPTION_vga:
                select_vgahw (optarg);
5334
                break;
5335
#if defined(TARGET_PPC) || defined(TARGET_SPARC)
5336 5337 5338 5339 5340 5341 5342 5343 5344 5345 5346 5347 5348 5349 5350 5351 5352 5353 5354 5355
            case QEMU_OPTION_g:
                {
                    const char *p;
                    int w, h, depth;
                    p = optarg;
                    w = strtol(p, (char **)&p, 10);
                    if (w <= 0) {
                    graphic_error:
                        fprintf(stderr, "qemu: invalid resolution or depth\n");
                        exit(1);
                    }
                    if (*p != 'x')
                        goto graphic_error;
                    p++;
                    h = strtol(p, (char **)&p, 10);
                    if (h <= 0)
                        goto graphic_error;
                    if (*p == 'x') {
                        p++;
                        depth = strtol(p, (char **)&p, 10);
5356
                        if (depth != 8 && depth != 15 && depth != 16 &&
5357 5358 5359 5360 5361 5362 5363
                            depth != 24 && depth != 32)
                            goto graphic_error;
                    } else if (*p == '\0') {
                        depth = graphic_depth;
                    } else {
                        goto graphic_error;
                    }
5364

5365 5366 5367 5368 5369
                    graphic_width = w;
                    graphic_height = h;
                    graphic_depth = depth;
                }
                break;
5370
#endif
T
ths 已提交
5371 5372 5373 5374 5375 5376 5377 5378
            case QEMU_OPTION_echr:
                {
                    char *r;
                    term_escape_char = strtol(optarg, &r, 0);
                    if (r == optarg)
                        printf("Bad argument to echr\n");
                    break;
                }
B
bellard 已提交
5379
            case QEMU_OPTION_monitor:
G
Gerd Hoffmann 已提交
5380 5381 5382 5383 5384
                monitor_parse(optarg, "readline");
                default_monitor = 0;
                break;
            case QEMU_OPTION_qmp:
                monitor_parse(optarg, "control");
G
Gerd Hoffmann 已提交
5385
                default_monitor = 0;
B
bellard 已提交
5386
                break;
G
Gerd Hoffmann 已提交
5387
            case QEMU_OPTION_mon:
5388
                opts = qemu_opts_parse(&qemu_mon_opts, optarg, 1);
G
Gerd Hoffmann 已提交
5389 5390 5391 5392 5393 5394
                if (!opts) {
                    fprintf(stderr, "parse error: %s\n", optarg);
                    exit(1);
                }
                default_monitor = 0;
                break;
5395
            case QEMU_OPTION_chardev:
5396
                opts = qemu_opts_parse(&qemu_chardev_opts, optarg, 1);
5397 5398 5399 5400 5401
                if (!opts) {
                    fprintf(stderr, "parse error: %s\n", optarg);
                    exit(1);
                }
                break;
B
bellard 已提交
5402
            case QEMU_OPTION_serial:
5403 5404
                add_device_config(DEV_SERIAL, optarg);
                default_serial = 0;
5405 5406 5407
                if (strncmp(optarg, "mon:", 4) == 0) {
                    default_monitor = 0;
                }
B
bellard 已提交
5408
                break;
R
Richard W.M. Jones 已提交
5409
            case QEMU_OPTION_watchdog:
M
Markus Armbruster 已提交
5410 5411 5412 5413 5414 5415
                if (watchdog) {
                    fprintf(stderr,
                            "qemu: only one watchdog option may be given\n");
                    return 1;
                }
                watchdog = optarg;
R
Richard W.M. Jones 已提交
5416 5417 5418 5419 5420 5421 5422
                break;
            case QEMU_OPTION_watchdog_action:
                if (select_watchdog_action(optarg) == -1) {
                    fprintf(stderr, "Unknown -watchdog-action parameter\n");
                    exit(1);
                }
                break;
5423
            case QEMU_OPTION_virtiocon:
G
Gerd Hoffmann 已提交
5424 5425
                add_device_config(DEV_VIRTCON, optarg);
                default_virtcon = 0;
5426 5427 5428
                if (strncmp(optarg, "mon:", 4) == 0) {
                    default_monitor = 0;
                }
5429
                break;
5430
            case QEMU_OPTION_parallel:
G
Gerd Hoffmann 已提交
5431 5432
                add_device_config(DEV_PARALLEL, optarg);
                default_parallel = 0;
5433 5434 5435
                if (strncmp(optarg, "mon:", 4) == 0) {
                    default_monitor = 0;
                }
5436
                break;
5437 5438 5439
            case QEMU_OPTION_debugcon:
                add_device_config(DEV_DEBUGCON, optarg);
                break;
B
bellard 已提交
5440 5441 5442 5443 5444 5445
	    case QEMU_OPTION_loadvm:
		loadvm = optarg;
		break;
            case QEMU_OPTION_full_screen:
                full_screen = 1;
                break;
T
ths 已提交
5446
#ifdef CONFIG_SDL
5447 5448 5449
            case QEMU_OPTION_no_frame:
                no_frame = 1;
                break;
T
ths 已提交
5450 5451 5452
            case QEMU_OPTION_alt_grab:
                alt_grab = 1;
                break;
5453 5454 5455
            case QEMU_OPTION_ctrl_grab:
                ctrl_grab = 1;
                break;
T
ths 已提交
5456 5457 5458
            case QEMU_OPTION_no_quit:
                no_quit = 1;
                break;
5459
            case QEMU_OPTION_sdl:
5460
                display_type = DT_SDL;
5461
                break;
T
ths 已提交
5462
#endif
B
bellard 已提交
5463
            case QEMU_OPTION_pidfile:
5464
                pid_file = optarg;
B
bellard 已提交
5465
                break;
5466 5467 5468 5469
#ifdef TARGET_I386
            case QEMU_OPTION_win2k_hack:
                win2k_install_hack = 1;
                break;
5470 5471 5472
            case QEMU_OPTION_rtc_td_hack:
                rtc_td_hack = 1;
                break;
5473 5474 5475 5476 5477 5478
            case QEMU_OPTION_acpitable:
                if(acpi_table_add(optarg) < 0) {
                    fprintf(stderr, "Wrong acpi table provided\n");
                    exit(1);
                }
                break;
5479 5480 5481 5482 5483 5484
            case QEMU_OPTION_smbios:
                if(smbios_entry_add(optarg) < 0) {
                    fprintf(stderr, "Wrong smbios provided\n");
                    exit(1);
                }
                break;
5485
#endif
A
aliguori 已提交
5486 5487 5488 5489
#ifdef CONFIG_KVM
            case QEMU_OPTION_enable_kvm:
                kvm_allowed = 1;
                break;
B
bellard 已提交
5490
#endif
B
bellard 已提交
5491 5492 5493
            case QEMU_OPTION_usb:
                usb_enabled = 1;
                break;
B
bellard 已提交
5494 5495
            case QEMU_OPTION_usbdevice:
                usb_enabled = 1;
5496 5497 5498
                add_device_config(DEV_USB, optarg);
                break;
            case QEMU_OPTION_device:
5499
                if (!qemu_opts_parse(&qemu_device_opts, optarg, 1)) {
G
Gerd Hoffmann 已提交
5500 5501
                    exit(1);
                }
B
bellard 已提交
5502
                break;
B
bellard 已提交
5503
            case QEMU_OPTION_smp:
5504
                smp_parse(optarg);
5505
                if (smp_cpus < 1) {
B
bellard 已提交
5506 5507 5508
                    fprintf(stderr, "Invalid number of CPUs\n");
                    exit(1);
                }
5509 5510 5511 5512 5513 5514 5515 5516 5517
                if (max_cpus < smp_cpus) {
                    fprintf(stderr, "maxcpus must be equal to or greater than "
                            "smp\n");
                    exit(1);
                }
                if (max_cpus > 255) {
                    fprintf(stderr, "Unsupported number of maxcpus\n");
                    exit(1);
                }
B
bellard 已提交
5518
                break;
B
bellard 已提交
5519
	    case QEMU_OPTION_vnc:
5520
                display_type = DT_VNC;
5521
		vnc_display = optarg;
B
bellard 已提交
5522
		break;
5523
#ifdef TARGET_I386
B
bellard 已提交
5524 5525 5526
            case QEMU_OPTION_no_acpi:
                acpi_enabled = 0;
                break;
A
aliguori 已提交
5527 5528 5529
            case QEMU_OPTION_no_hpet:
                no_hpet = 1;
                break;
5530 5531 5532 5533 5534
            case QEMU_OPTION_balloon:
                if (balloon_parse(optarg) < 0) {
                    fprintf(stderr, "Unknown -balloon argument %s\n", optarg);
                    exit(1);
                }
5535
                break;
5536
#endif
B
bellard 已提交
5537 5538 5539
            case QEMU_OPTION_no_reboot:
                no_reboot = 1;
                break;
A
aurel32 已提交
5540 5541 5542
            case QEMU_OPTION_no_shutdown:
                no_shutdown = 1;
                break;
5543 5544 5545
            case QEMU_OPTION_show_cursor:
                cursor_hide = 0;
                break;
5546 5547 5548 5549 5550 5551 5552
            case QEMU_OPTION_uuid:
                if(qemu_uuid_parse(optarg, qemu_uuid) < 0) {
                    fprintf(stderr, "Fail to parse UUID string."
                            " Wrong format.\n");
                    exit(1);
                }
                break;
5553
#ifndef _WIN32
T
ths 已提交
5554 5555 5556
	    case QEMU_OPTION_daemonize:
		daemonize = 1;
		break;
5557
#endif
5558 5559 5560 5561 5562 5563 5564 5565
	    case QEMU_OPTION_option_rom:
		if (nb_option_roms >= MAX_OPTION_ROMS) {
		    fprintf(stderr, "Too many option ROMs\n");
		    exit(1);
		}
		option_rom[nb_option_roms] = optarg;
		nb_option_roms++;
		break;
5566
#if defined(TARGET_ARM) || defined(TARGET_M68K)
5567 5568 5569
            case QEMU_OPTION_semihosting:
                semihosting_enabled = 1;
                break;
5570
#endif
T
ths 已提交
5571
            case QEMU_OPTION_name:
A
Andi Kleen 已提交
5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584
                qemu_name = qemu_strdup(optarg);
		 {
		     char *p = strchr(qemu_name, ',');
		     if (p != NULL) {
		        *p++ = 0;
			if (strncmp(p, "process=", 8)) {
			    fprintf(stderr, "Unknown subargument %s to -name", p);
			    exit(1);
			}
			p += 8;
			set_proc_name(p);
		     }	
		 }	
T
ths 已提交
5585
                break;
5586
#if defined(TARGET_SPARC) || defined(TARGET_PPC)
B
blueswir1 已提交
5587 5588 5589 5590 5591 5592 5593 5594
            case QEMU_OPTION_prom_env:
                if (nb_prom_envs >= MAX_PROM_ENVS) {
                    fprintf(stderr, "Too many prom variables\n");
                    exit(1);
                }
                prom_envs[nb_prom_envs] = optarg;
                nb_prom_envs++;
                break;
5595 5596 5597 5598
#endif
#ifdef TARGET_ARM
            case QEMU_OPTION_old_param:
                old_param = 1;
5599
                break;
B
blueswir1 已提交
5600
#endif
5601 5602 5603
            case QEMU_OPTION_clock:
                configure_alarms(optarg);
                break;
B
bellard 已提交
5604
            case QEMU_OPTION_startdate:
J
Jan Kiszka 已提交
5605 5606 5607
                configure_rtc_date_offset(optarg, 1);
                break;
            case QEMU_OPTION_rtc:
5608
                opts = qemu_opts_parse(&qemu_rtc_opts, optarg, 0);
J
Jan Kiszka 已提交
5609 5610 5611
                if (!opts) {
                    fprintf(stderr, "parse error: %s\n", optarg);
                    exit(1);
B
bellard 已提交
5612
                }
J
Jan Kiszka 已提交
5613
                configure_rtc(opts);
B
bellard 已提交
5614
                break;
5615 5616 5617 5618 5619
            case QEMU_OPTION_tb_size:
                tb_size = strtol(optarg, NULL, 0);
                if (tb_size < 0)
                    tb_size = 0;
                break;
P
pbrook 已提交
5620 5621 5622 5623 5624 5625 5626 5627
            case QEMU_OPTION_icount:
                use_icount = 1;
                if (strcmp(optarg, "auto") == 0) {
                    icount_time_shift = -1;
                } else {
                    icount_time_shift = strtol(optarg, NULL, 0);
                }
                break;
A
aliguori 已提交
5628 5629 5630
            case QEMU_OPTION_incoming:
                incoming = optarg;
                break;
5631 5632 5633
            case QEMU_OPTION_nodefaults:
                default_serial = 0;
                default_parallel = 0;
G
Gerd Hoffmann 已提交
5634
                default_virtcon = 0;
5635 5636
                default_monitor = 0;
                default_vga = 0;
G
Gerd Hoffmann 已提交
5637
                default_net = 0;
G
Gerd Hoffmann 已提交
5638 5639 5640
                default_floppy = 0;
                default_cdrom = 0;
                default_sdcard = 0;
5641
                break;
5642
#ifndef _WIN32
5643 5644 5645 5646 5647 5648
            case QEMU_OPTION_chroot:
                chroot_dir = optarg;
                break;
            case QEMU_OPTION_runas:
                run_as = optarg;
                break;
5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659
#endif
#ifdef CONFIG_XEN
            case QEMU_OPTION_xen_domid:
                xen_domid = atoi(optarg);
                break;
            case QEMU_OPTION_xen_create:
                xen_mode = XEN_CREATE;
                break;
            case QEMU_OPTION_xen_attach:
                xen_mode = XEN_ATTACH;
                break;
5660
#endif
5661 5662 5663 5664 5665 5666 5667 5668
            case QEMU_OPTION_readconfig:
                {
                    FILE *fp;
                    fp = fopen(optarg, "r");
                    if (fp == NULL) {
                        fprintf(stderr, "open %s: %s\n", optarg, strerror(errno));
                        exit(1);
                    }
5669
                    if (qemu_config_parse(fp, optarg) != 0) {
5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690
                        exit(1);
                    }
                    fclose(fp);
                    break;
                }
            case QEMU_OPTION_writeconfig:
                {
                    FILE *fp;
                    if (strcmp(optarg, "-") == 0) {
                        fp = stdout;
                    } else {
                        fp = fopen(optarg, "w");
                        if (fp == NULL) {
                            fprintf(stderr, "open %s: %s\n", optarg, strerror(errno));
                            exit(1);
                        }
                    }
                    qemu_config_write(fp);
                    fclose(fp);
                    break;
                }
5691
            }
5692 5693
        }
    }
5694
    loc_set_none();
5695

P
Paul Brook 已提交
5696 5697 5698 5699 5700 5701 5702 5703 5704 5705
    /* If no data_dir is specified then try to find it relative to the
       executable path.  */
    if (!data_dir) {
        data_dir = find_datadir(argv[0]);
    }
    /* If all else fails use the install patch specified when building.  */
    if (!data_dir) {
        data_dir = CONFIG_QEMU_SHAREDIR;
    }

5706 5707 5708 5709 5710 5711 5712
    /*
     * Default to max_cpus = smp_cpus, in case the user doesn't
     * specify a max_cpus value.
     */
    if (!max_cpus)
        max_cpus = smp_cpus;

B
balrog 已提交
5713
    machine->max_cpus = machine->max_cpus ?: 1; /* Default to UP */
5714 5715 5716 5717 5718 5719 5720
    if (smp_cpus > machine->max_cpus) {
        fprintf(stderr, "Number of SMP cpus requested (%d), exceeds max cpus "
                "supported by machine `%s' (%d)\n", smp_cpus,  machine->name,
                machine->max_cpus);
        exit(1);
    }

5721
    qemu_opts_foreach(&qemu_device_opts, default_driver_check, NULL, 0);
G
Gerd Hoffmann 已提交
5722
    qemu_opts_foreach(&qemu_global_opts, default_driver_check, NULL, 0);
5723

5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735
    if (machine->no_serial) {
        default_serial = 0;
    }
    if (machine->no_parallel) {
        default_parallel = 0;
    }
    if (!machine->use_virtcon) {
        default_virtcon = 0;
    }
    if (machine->no_vga) {
        default_vga = 0;
    }
G
Gerd Hoffmann 已提交
5736 5737 5738 5739 5740 5741 5742 5743 5744
    if (machine->no_floppy) {
        default_floppy = 0;
    }
    if (machine->no_cdrom) {
        default_cdrom = 0;
    }
    if (machine->no_sdcard) {
        default_sdcard = 0;
    }
5745

5746
    if (display_type == DT_NOGRAPHIC) {
G
Gerd Hoffmann 已提交
5747 5748
        if (default_parallel)
            add_device_config(DEV_PARALLEL, "null");
G
Gerd Hoffmann 已提交
5749 5750
        if (default_serial && default_monitor) {
            add_device_config(DEV_SERIAL, "mon:stdio");
5751 5752
        } else if (default_virtcon && default_monitor) {
            add_device_config(DEV_VIRTCON, "mon:stdio");
G
Gerd Hoffmann 已提交
5753 5754 5755
        } else {
            if (default_serial)
                add_device_config(DEV_SERIAL, "stdio");
5756 5757
            if (default_virtcon)
                add_device_config(DEV_VIRTCON, "stdio");
G
Gerd Hoffmann 已提交
5758
            if (default_monitor)
G
Gerd Hoffmann 已提交
5759
                monitor_parse("stdio", "readline");
G
Gerd Hoffmann 已提交
5760
        }
5761 5762 5763
    } else {
        if (default_serial)
            add_device_config(DEV_SERIAL, "vc:80Cx24C");
G
Gerd Hoffmann 已提交
5764 5765
        if (default_parallel)
            add_device_config(DEV_PARALLEL, "vc:80Cx24C");
G
Gerd Hoffmann 已提交
5766
        if (default_monitor)
G
Gerd Hoffmann 已提交
5767
            monitor_parse("vc:80Cx24C", "readline");
5768 5769
        if (default_virtcon)
            add_device_config(DEV_VIRTCON, "vc:80Cx24C");
5770
    }
G
Gerd Hoffmann 已提交
5771 5772
    if (default_vga)
        vga_interface_type = VGA_CIRRUS;
5773

5774 5775 5776
    if (qemu_opts_foreach(&qemu_chardev_opts, chardev_init_func, NULL, 1) != 0)
        exit(1);

T
ths 已提交
5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791
#ifndef _WIN32
    if (daemonize) {
	pid_t pid;

	if (pipe(fds) == -1)
	    exit(1);

	pid = fork();
	if (pid > 0) {
	    uint8_t status;
	    ssize_t len;

	    close(fds[1]);

	again:
5792 5793 5794 5795 5796 5797 5798
            len = read(fds[0], &status, 1);
            if (len == -1 && (errno == EINTR))
                goto again;

            if (len != 1)
                exit(1);
            else if (status == 1) {
5799
                fprintf(stderr, "Could not acquire pidfile: %s\n", strerror(errno));
5800 5801 5802
                exit(1);
            } else
                exit(0);
T
ths 已提交
5803
	} else if (pid < 0)
5804
            exit(1);
T
ths 已提交
5805

K
Kevin Wolf 已提交
5806 5807 5808
	close(fds[0]);
	qemu_set_cloexec(fds[1]);

T
ths 已提交
5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819 5820 5821 5822
	setsid();

	pid = fork();
	if (pid > 0)
	    exit(0);
	else if (pid < 0)
	    exit(1);

	umask(027);

        signal(SIGTSTP, SIG_IGN);
        signal(SIGTTOU, SIG_IGN);
        signal(SIGTTIN, SIG_IGN);
    }
5823
#endif
T
ths 已提交
5824

5825
    if (pid_file && qemu_create_pidfile(pid_file) != 0) {
5826
#ifndef _WIN32
5827 5828
        if (daemonize) {
            uint8_t status = 1;
5829 5830 5831
            if (write(fds[1], &status, 1) != 1) {
                perror("daemonize. Writing to pipe\n");
            }
5832
        } else
5833
#endif
5834
            fprintf(stderr, "Could not acquire pid file: %s\n", strerror(errno));
5835 5836 5837
        exit(1);
    }

M
Marcelo Tosatti 已提交
5838 5839 5840 5841 5842 5843 5844 5845 5846 5847
    if (kvm_enabled()) {
        int ret;

        ret = kvm_init(smp_cpus);
        if (ret < 0) {
            fprintf(stderr, "failed to initialize KVM\n");
            exit(1);
        }
    }

5848 5849 5850 5851
    if (qemu_init_main_loop()) {
        fprintf(stderr, "qemu_init_main_loop failed\n");
        exit(1);
    }
5852
    linux_boot = (kernel_filename != NULL);
B
balrog 已提交
5853

5854 5855 5856 5857 5858 5859 5860 5861 5862 5863
    if (!linux_boot && *kernel_cmdline != '\0') {
        fprintf(stderr, "-append only allowed with -kernel option\n");
        exit(1);
    }

    if (!linux_boot && initrd_filename != NULL) {
        fprintf(stderr, "-initrd only allowed with -kernel option\n");
        exit(1);
    }

5864 5865
#ifndef _WIN32
    /* Win32 doesn't support line-buffering and requires size >= 2 */
B
bellard 已提交
5866
    setvbuf(stdout, NULL, _IOLBF, 0);
5867
#endif
5868

5869 5870 5871 5872
    if (init_timer_alarm() < 0) {
        fprintf(stderr, "could not initialize alarm timer\n");
        exit(1);
    }
P
pbrook 已提交
5873 5874 5875 5876 5877 5878 5879
    if (use_icount && icount_time_shift < 0) {
        use_icount = 2;
        /* 125MIPS seems a reasonable initial guess at the guest speed.
           It will be corrected fairly quickly anyway.  */
        icount_time_shift = 3;
        init_icount_adjust();
    }
5880

B
bellard 已提交
5881 5882 5883 5884
#ifdef _WIN32
    socket_init();
#endif

5885 5886
    if (net_init_clients() < 0) {
        exit(1);
B
bellard 已提交
5887
    }
B
bellard 已提交
5888

5889 5890 5891
    net_boot = (boot_devices_bitmap >> ('n' - 'a')) & 0xF;
    net_set_boot_mask(net_boot);

5892
    /* init the bluetooth world */
5893 5894
    if (foreach_device_config(DEV_BT, bt_parse))
        exit(1);
5895

5896
    /* init the memory */
P
pbrook 已提交
5897 5898
    if (ram_size == 0)
        ram_size = DEFAULT_RAM_SIZE * 1024 * 1024;
5899

5900 5901 5902
    /* init the dynamic translator */
    cpu_exec_init_all(tb_size * 1024 * 1024);

5903
    bdrv_init_with_whitelist();
5904

L
lirans@il.ibm.com 已提交
5905 5906
    blk_mig_init();

G
Gerd Hoffmann 已提交
5907
    if (default_cdrom) {
G
Gerd Hoffmann 已提交
5908 5909
        /* we always create the cdrom drive, even if no disk is there */
        drive_add(NULL, CDROM_ALIAS);
G
Gerd Hoffmann 已提交
5910
    }
5911

G
Gerd Hoffmann 已提交
5912
    if (default_floppy) {
G
Gerd Hoffmann 已提交
5913 5914
        /* we always create at least one floppy */
        drive_add(NULL, FD_ALIAS, 0);
G
Gerd Hoffmann 已提交
5915
    }
5916

G
Gerd Hoffmann 已提交
5917
    if (default_sdcard) {
G
Gerd Hoffmann 已提交
5918 5919 5920
        /* we always create one sd slot, even if no card is in it */
        drive_add(NULL, SD_ALIAS);
    }
5921

T
ths 已提交
5922
    /* open the virtual block devices */
G
Gerd Hoffmann 已提交
5923
    if (snapshot)
G
Gerd Hoffmann 已提交
5924 5925
        qemu_opts_foreach(&qemu_drive_opts, drive_enable_snapshot, NULL, 0);
    if (qemu_opts_foreach(&qemu_drive_opts, drive_init_func, machine, 1) != 0)
G
Gerd Hoffmann 已提交
5926
        exit(1);
5927

J
Juan Quintela 已提交
5928
    vmstate_register(0, &vmstate_timers ,&timers_state);
L
lirans@il.ibm.com 已提交
5929 5930
    register_savevm_live("ram", 0, 3, NULL, ram_save_live, NULL, 
                         ram_load, NULL);
5931

5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973
    if (nb_numa_nodes > 0) {
        int i;

        if (nb_numa_nodes > smp_cpus) {
            nb_numa_nodes = smp_cpus;
        }

        /* If no memory size if given for any node, assume the default case
         * and distribute the available memory equally across all nodes
         */
        for (i = 0; i < nb_numa_nodes; i++) {
            if (node_mem[i] != 0)
                break;
        }
        if (i == nb_numa_nodes) {
            uint64_t usedmem = 0;

            /* On Linux, the each node's border has to be 8MB aligned,
             * the final node gets the rest.
             */
            for (i = 0; i < nb_numa_nodes - 1; i++) {
                node_mem[i] = (ram_size / nb_numa_nodes) & ~((1 << 23UL) - 1);
                usedmem += node_mem[i];
            }
            node_mem[i] = ram_size - usedmem;
        }

        for (i = 0; i < nb_numa_nodes; i++) {
            if (node_cpumask[i] != 0)
                break;
        }
        /* assigning the VCPUs round-robin is easier to implement, guest OSes
         * must cope with this anyway, because there are BIOSes out there in
         * real machines which also use this scheme.
         */
        if (i == nb_numa_nodes) {
            for (i = 0; i < smp_cpus; i++) {
                node_cpumask[i % nb_numa_nodes] |= 1 << i;
            }
        }
    }

5974 5975
    if (foreach_device_config(DEV_SERIAL, serial_parse) < 0)
        exit(1);
G
Gerd Hoffmann 已提交
5976 5977
    if (foreach_device_config(DEV_PARALLEL, parallel_parse) < 0)
        exit(1);
G
Gerd Hoffmann 已提交
5978 5979
    if (foreach_device_config(DEV_VIRTCON, virtcon_parse) < 0)
        exit(1);
5980 5981
    if (foreach_device_config(DEV_DEBUGCON, debugcon_parse) < 0)
        exit(1);
5982

P
Paul Brook 已提交
5983 5984
    module_call_init(MODULE_INIT_DEVICE);

5985 5986 5987
    if (qemu_opts_foreach(&qemu_device_opts, device_help_func, NULL, 0) != 0)
        exit(0);

M
Markus Armbruster 已提交
5988 5989 5990 5991 5992 5993
    if (watchdog) {
        i = select_watchdog(watchdog);
        if (i > 0)
            exit (i == 1 ? 1 : 0);
    }

5994
    if (machine->compat_props) {
5995
        qdev_prop_register_global_list(machine->compat_props);
5996
    }
5997 5998
    qemu_add_globals();

P
Paul Brook 已提交
5999
    machine->init(ram_size, boot_devices,
6000 6001
                  kernel_filename, kernel_cmdline, initrd_filename, cpu_model);

6002
    cpu_synchronize_all_post_init();
6003

J
Juan Quintela 已提交
6004 6005 6006 6007 6008
#ifndef _WIN32
    /* must be after terminal init, SDL library changes signal handlers */
    sighandler_setup();
#endif

6009 6010 6011 6012 6013 6014 6015 6016
    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;
            }
        }
    }

6017 6018
    current_machine = machine;

6019 6020
    /* init USB devices */
    if (usb_enabled) {
M
Markus Armbruster 已提交
6021 6022
        if (foreach_device_config(DEV_USB, usb_parse) < 0)
            exit(1);
6023 6024
    }

6025
    /* init generic devices */
G
Gerd Hoffmann 已提交
6026
    if (qemu_opts_foreach(&qemu_device_opts, device_init_func, NULL, 1) != 0)
6027 6028
        exit(1);

6029 6030
    net_check_clients();

6031
    /* just use the first displaystate for the moment */
6032
    ds = get_displaystate();
6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047

    if (display_type == DT_DEFAULT) {
#if defined(CONFIG_SDL) || defined(CONFIG_COCOA)
        display_type = DT_SDL;
#else
        display_type = DT_VNC;
        vnc_display = "localhost:0,to=99";
        show_vnc_port = 1;
#endif
    }
        

    switch (display_type) {
    case DT_NOGRAPHIC:
        break;
B
balrog 已提交
6048
#if defined(CONFIG_CURSES)
6049 6050 6051
    case DT_CURSES:
        curses_display_init(ds, full_screen);
        break;
B
balrog 已提交
6052
#endif
6053
#if defined(CONFIG_SDL)
6054 6055 6056
    case DT_SDL:
        sdl_display_init(ds, full_screen, no_frame);
        break;
6057
#elif defined(CONFIG_COCOA)
6058 6059 6060
    case DT_SDL:
        cocoa_display_init(ds, full_screen);
        break;
6061
#endif
6062 6063 6064 6065
    case DT_VNC:
        vnc_display_init(ds);
        if (vnc_display_open(ds, vnc_display) < 0)
            exit(1);
6066

6067 6068
        if (show_vnc_port) {
            printf("VNC server running on `%s'\n", vnc_display_local_addr(ds));
6069
        }
6070 6071 6072
        break;
    default:
        break;
6073
    }
6074
    dpy_resize(ds);
6075

6076 6077 6078 6079 6080
    dcl = ds->listeners;
    while (dcl != NULL) {
        if (dcl->dpy_refresh != NULL) {
            ds->gui_timer = qemu_new_timer(rt_clock, gui_update, ds);
            qemu_mod_timer(ds->gui_timer, qemu_get_clock(rt_clock));
T
ths 已提交
6081
        }
6082
        dcl = dcl->next;
T
ths 已提交
6083
    }
6084

6085
    if (display_type == DT_NOGRAPHIC || display_type == DT_VNC) {
B
blueswir1 已提交
6086 6087 6088 6089
        nographic_timer = qemu_new_timer(rt_clock, nographic_update, NULL);
        qemu_mod_timer(nographic_timer, qemu_get_clock(rt_clock));
    }

6090
    text_consoles_set_display(ds);
6091

6092 6093
    if (qemu_opts_foreach(&qemu_mon_opts, mon_init_func, NULL, 1) != 0)
        exit(1);
B
bellard 已提交
6094

6095 6096 6097 6098
    if (gdbstub_dev && gdbserver_start(gdbstub_dev) < 0) {
        fprintf(stderr, "qemu: could not open gdbserver on device '%s'\n",
                gdbstub_dev);
        exit(1);
6099 6100
    }

6101 6102
    qdev_machine_creation_done();

G
Gerd Hoffmann 已提交
6103 6104 6105 6106
    if (rom_load_all() != 0) {
        fprintf(stderr, "rom loading failed\n");
        exit(1);
    }
6107

6108
    qemu_system_reset();
6109
    if (loadvm) {
6110
        if (load_vmstate(loadvm) < 0) {
6111 6112 6113
            autostart = 0;
        }
    }
B
bellard 已提交
6114

G
Glauber Costa 已提交
6115
    if (incoming) {
A
aliguori 已提交
6116
        qemu_start_incoming_migration(incoming);
6117
    } else if (autostart) {
6118
        vm_start();
6119
    }
6120

B
blueswir1 已提交
6121
#ifndef _WIN32
T
ths 已提交
6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133
    if (daemonize) {
	uint8_t status = 0;
	ssize_t len;

    again1:
	len = write(fds[1], &status, 1);
	if (len == -1 && (errno == EINTR))
	    goto again1;

	if (len != 1)
	    exit(1);

6134 6135 6136 6137
        if (chdir("/")) {
            perror("not able to chdir to /");
            exit(1);
        }
K
Kevin Wolf 已提交
6138
	TFR(fd = qemu_open("/dev/null", O_RDWR));
T
ths 已提交
6139 6140
	if (fd == -1)
	    exit(1);
6141
    }
T
ths 已提交
6142

6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155
    if (run_as) {
        pwd = getpwnam(run_as);
        if (!pwd) {
            fprintf(stderr, "User \"%s\" doesn't exist\n", run_as);
            exit(1);
        }
    }

    if (chroot_dir) {
        if (chroot(chroot_dir) < 0) {
            fprintf(stderr, "chroot failed\n");
            exit(1);
        }
6156 6157 6158 6159
        if (chdir("/")) {
            perror("not able to chdir to /");
            exit(1);
        }
6160 6161 6162 6163 6164 6165 6166 6167 6168 6169 6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180
    }

    if (run_as) {
        if (setgid(pwd->pw_gid) < 0) {
            fprintf(stderr, "Failed to setgid(%d)\n", pwd->pw_gid);
            exit(1);
        }
        if (setuid(pwd->pw_uid) < 0) {
            fprintf(stderr, "Failed to setuid(%d)\n", pwd->pw_uid);
            exit(1);
        }
        if (setuid(0) != -1) {
            fprintf(stderr, "Dropping privileges failed\n");
            exit(1);
        }
    }

    if (daemonize) {
        dup2(fd, 0);
        dup2(fd, 1);
        dup2(fd, 2);
T
ths 已提交
6181

6182
        close(fd);
T
ths 已提交
6183
    }
B
blueswir1 已提交
6184
#endif
T
ths 已提交
6185

6186
    main_loop();
B
bellard 已提交
6187
    quit_timers();
6188
    net_cleanup();
T
ths 已提交
6189

6190 6191
    return 0;
}