arch_init.c 29.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
/*
 * QEMU System Emulator
 *
 * Copyright (c) 2003-2008 Fabrice Bellard
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */
#include <stdint.h>
#include <stdarg.h>
26
#include <stdlib.h>
27
#ifndef _WIN32
28
#include <sys/types.h>
29 30 31
#include <sys/mman.h>
#endif
#include "config.h"
32
#include "monitor/monitor.h"
33
#include "sysemu/sysemu.h"
34 35
#include "qemu/bitops.h"
#include "qemu/bitmap.h"
36
#include "sysemu/arch_init.h"
37
#include "audio/audio.h"
P
Paolo Bonzini 已提交
38
#include "hw/i386/pc.h"
39
#include "hw/pci/pci.h"
P
Paolo Bonzini 已提交
40
#include "hw/audio/audio.h"
41
#include "sysemu/kvm.h"
42
#include "migration/migration.h"
P
Paolo Bonzini 已提交
43
#include "hw/i386/smbios.h"
44
#include "exec/address-spaces.h"
P
Paolo Bonzini 已提交
45
#include "hw/audio/pcspk.h"
46
#include "migration/page_cache.h"
47
#include "qemu/config-file.h"
48
#include "qmp-commands.h"
49
#include "trace.h"
50
#include "exec/cpu-all.h"
51
#include "hw/acpi/acpi.h"
52

O
Orit Wasserman 已提交
53 54 55 56 57 58 59 60
#ifdef DEBUG_ARCH_INIT
#define DPRINTF(fmt, ...) \
    do { fprintf(stdout, "arch_init: " fmt, ## __VA_ARGS__); } while (0)
#else
#define DPRINTF(fmt, ...) \
    do { } while (0)
#endif

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
#ifdef TARGET_SPARC
int graphic_width = 1024;
int graphic_height = 768;
int graphic_depth = 8;
#else
int graphic_width = 800;
int graphic_height = 600;
int graphic_depth = 15;
#endif


#if defined(TARGET_ALPHA)
#define QEMU_ARCH QEMU_ARCH_ALPHA
#elif defined(TARGET_ARM)
#define QEMU_ARCH QEMU_ARCH_ARM
#elif defined(TARGET_CRIS)
#define QEMU_ARCH QEMU_ARCH_CRIS
#elif defined(TARGET_I386)
#define QEMU_ARCH QEMU_ARCH_I386
#elif defined(TARGET_M68K)
#define QEMU_ARCH QEMU_ARCH_M68K
M
Michael Walle 已提交
82 83
#elif defined(TARGET_LM32)
#define QEMU_ARCH QEMU_ARCH_LM32
84 85 86 87
#elif defined(TARGET_MICROBLAZE)
#define QEMU_ARCH QEMU_ARCH_MICROBLAZE
#elif defined(TARGET_MIPS)
#define QEMU_ARCH QEMU_ARCH_MIPS
A
Anthony Green 已提交
88 89
#elif defined(TARGET_MOXIE)
#define QEMU_ARCH QEMU_ARCH_MOXIE
90 91
#elif defined(TARGET_OPENRISC)
#define QEMU_ARCH QEMU_ARCH_OPENRISC
92 93 94 95 96 97 98 99
#elif defined(TARGET_PPC)
#define QEMU_ARCH QEMU_ARCH_PPC
#elif defined(TARGET_S390X)
#define QEMU_ARCH QEMU_ARCH_S390X
#elif defined(TARGET_SH4)
#define QEMU_ARCH QEMU_ARCH_SH4
#elif defined(TARGET_SPARC)
#define QEMU_ARCH QEMU_ARCH_SPARC
M
Max Filippov 已提交
100 101
#elif defined(TARGET_XTENSA)
#define QEMU_ARCH QEMU_ARCH_XTENSA
102 103
#elif defined(TARGET_UNICORE32)
#define QEMU_ARCH QEMU_ARCH_UNICORE32
104 105 106 107 108 109 110
#endif

const uint32_t arch_type = QEMU_ARCH;

/***********************************************************/
/* ram save/restore */

111 112 113 114 115 116
#define RAM_SAVE_FLAG_FULL     0x01 /* Obsolete, not used anymore */
#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
#define RAM_SAVE_FLAG_CONTINUE 0x20
117
#define RAM_SAVE_FLAG_XBZRLE   0x40
118

119

120 121
static struct defconfig_file {
    const char *filename;
122 123
    /* Indicates it is an user config file (disabled by -no-user-config) */
    bool userconfig;
124
} default_config_files[] = {
125
    { CONFIG_QEMU_CONFDIR "/qemu.conf",                   true },
126
    { CONFIG_QEMU_CONFDIR "/target-" TARGET_NAME ".conf", true },
127 128 129 130
    { NULL }, /* end of list */
};


131
int qemu_read_default_config_files(bool userconfig)
132 133
{
    int ret;
134
    struct defconfig_file *f;
135

136
    for (f = default_config_files; f->filename; f++) {
137 138 139
        if (!userconfig && f->userconfig) {
            continue;
        }
140 141 142 143
        ret = qemu_read_config_file(f->filename);
        if (ret < 0 && ret != -ENOENT) {
            return ret;
        }
144
    }
L
Laszlo Ersek 已提交
145

146 147 148
    return 0;
}

149
static inline bool is_zero_page(uint8_t *p)
150
{
151 152
    return buffer_find_nonzero_offset(p, TARGET_PAGE_SIZE) ==
        TARGET_PAGE_SIZE;
153 154
}

155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
/* struct contains XBZRLE cache and a static page
   used by the compression */
static struct {
    /* buffer used for XBZRLE encoding */
    uint8_t *encoded_buf;
    /* buffer for storing page content */
    uint8_t *current_buf;
    /* buffer used for XBZRLE decoding */
    uint8_t *decoded_buf;
    /* Cache for XBZRLE */
    PageCache *cache;
} XBZRLE = {
    .encoded_buf = NULL,
    .current_buf = NULL,
    .decoded_buf = NULL,
    .cache = NULL,
};

173 174 175 176 177 178 179 180 181 182

int64_t xbzrle_cache_resize(int64_t new_size)
{
    if (XBZRLE.cache != NULL) {
        return cache_resize(XBZRLE.cache, new_size / TARGET_PAGE_SIZE) *
            TARGET_PAGE_SIZE;
    }
    return pow2floor(new_size);
}

183 184 185
/* accounting for migration statistics */
typedef struct AccountingInfo {
    uint64_t dup_pages;
186
    uint64_t skipped_pages;
187 188
    uint64_t norm_pages;
    uint64_t iterations;
O
Orit Wasserman 已提交
189 190 191 192
    uint64_t xbzrle_bytes;
    uint64_t xbzrle_pages;
    uint64_t xbzrle_cache_miss;
    uint64_t xbzrle_overflows;
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
} AccountingInfo;

static AccountingInfo acct_info;

static void acct_clear(void)
{
    memset(&acct_info, 0, sizeof(acct_info));
}

uint64_t dup_mig_bytes_transferred(void)
{
    return acct_info.dup_pages * TARGET_PAGE_SIZE;
}

uint64_t dup_mig_pages_transferred(void)
{
    return acct_info.dup_pages;
}

212 213 214 215 216 217 218 219 220 221
uint64_t skipped_mig_bytes_transferred(void)
{
    return acct_info.skipped_pages * TARGET_PAGE_SIZE;
}

uint64_t skipped_mig_pages_transferred(void)
{
    return acct_info.skipped_pages;
}

222 223 224 225 226 227 228 229 230 231
uint64_t norm_mig_bytes_transferred(void)
{
    return acct_info.norm_pages * TARGET_PAGE_SIZE;
}

uint64_t norm_mig_pages_transferred(void)
{
    return acct_info.norm_pages;
}

O
Orit Wasserman 已提交
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251
uint64_t xbzrle_mig_bytes_transferred(void)
{
    return acct_info.xbzrle_bytes;
}

uint64_t xbzrle_mig_pages_transferred(void)
{
    return acct_info.xbzrle_pages;
}

uint64_t xbzrle_mig_pages_cache_miss(void)
{
    return acct_info.xbzrle_cache_miss;
}

uint64_t xbzrle_mig_pages_overflow(void)
{
    return acct_info.xbzrle_overflows;
}

252 253
static size_t save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
                             int cont, int flag)
O
Orit Wasserman 已提交
254
{
255 256 257 258
    size_t size;

    qemu_put_be64(f, offset | cont | flag);
    size = 8;
O
Orit Wasserman 已提交
259

260 261 262 263 264 265 266
    if (!cont) {
        qemu_put_byte(f, strlen(block->idstr));
        qemu_put_buffer(f, (uint8_t *)block->idstr,
                        strlen(block->idstr));
        size += 1 + strlen(block->idstr);
    }
    return size;
O
Orit Wasserman 已提交
267 268
}

269 270 271 272
#define ENCODING_FLAG_XBZRLE 0x1

static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
                            ram_addr_t current_addr, RAMBlock *block,
273
                            ram_addr_t offset, int cont, bool last_stage)
274 275 276 277 278
{
    int encoded_len = 0, bytes_sent = -1;
    uint8_t *prev_cached_page;

    if (!cache_is_cached(XBZRLE.cache, current_addr)) {
279
        if (!last_stage) {
P
Peter Lieven 已提交
280
            cache_insert(XBZRLE.cache, current_addr, current_data);
281
        }
O
Orit Wasserman 已提交
282
        acct_info.xbzrle_cache_miss++;
283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299
        return -1;
    }

    prev_cached_page = get_cached_data(XBZRLE.cache, current_addr);

    /* save current buffer into memory */
    memcpy(XBZRLE.current_buf, current_data, TARGET_PAGE_SIZE);

    /* XBZRLE encoding (if there is no overflow) */
    encoded_len = xbzrle_encode_buffer(prev_cached_page, XBZRLE.current_buf,
                                       TARGET_PAGE_SIZE, XBZRLE.encoded_buf,
                                       TARGET_PAGE_SIZE);
    if (encoded_len == 0) {
        DPRINTF("Skipping unmodified page\n");
        return 0;
    } else if (encoded_len == -1) {
        DPRINTF("Overflow\n");
O
Orit Wasserman 已提交
300
        acct_info.xbzrle_overflows++;
301 302 303 304 305 306
        /* update data in the cache */
        memcpy(prev_cached_page, current_data, TARGET_PAGE_SIZE);
        return -1;
    }

    /* we need to update the data in the cache, in order to get the same data */
307 308 309
    if (!last_stage) {
        memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
    }
310 311

    /* Send XBZRLE based compressed page */
312
    bytes_sent = save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_XBZRLE);
313 314 315
    qemu_put_byte(f, ENCODING_FLAG_XBZRLE);
    qemu_put_be16(f, encoded_len);
    qemu_put_buffer(f, XBZRLE.encoded_buf, encoded_len);
316
    bytes_sent += encoded_len + 1 + 2;
O
Orit Wasserman 已提交
317 318
    acct_info.xbzrle_pages++;
    acct_info.xbzrle_bytes += bytes_sent;
319 320 321 322

    return bytes_sent;
}

323 324 325 326

/* This is the last block that we have visited serching for dirty pages
 */
static RAMBlock *last_seen_block;
J
Juan Quintela 已提交
327 328
/* This is the last block from where we have sent data */
static RAMBlock *last_sent_block;
329
static ram_addr_t last_offset;
J
Juan Quintela 已提交
330 331
static unsigned long *migration_bitmap;
static uint64_t migration_dirty_pages;
U
Umesh Deshpande 已提交
332
static uint32_t last_version;
333
static bool ram_bulk_stage;
334

335 336 337
static inline
ram_addr_t migration_bitmap_find_and_reset_dirty(MemoryRegion *mr,
                                                 ram_addr_t start)
338
{
339 340 341
    unsigned long base = mr->ram_addr >> TARGET_PAGE_BITS;
    unsigned long nr = base + (start >> TARGET_PAGE_BITS);
    unsigned long size = base + (int128_get64(mr->size) >> TARGET_PAGE_BITS);
J
Juan Quintela 已提交
342

343 344 345 346 347 348 349
    unsigned long next;

    if (ram_bulk_stage && nr > base) {
        next = nr + 1;
    } else {
        next = find_next_bit(migration_bitmap, size, nr);
    }
350

351 352
    if (next < size) {
        clear_bit(next, migration_bitmap);
J
Juan Quintela 已提交
353
        migration_dirty_pages--;
354
    }
355
    return (next - base) << TARGET_PAGE_BITS;
356 357
}

J
Juan Quintela 已提交
358 359
static inline bool migration_bitmap_set_dirty(MemoryRegion *mr,
                                              ram_addr_t offset)
360
{
J
Juan Quintela 已提交
361 362
    bool ret;
    int nr = (mr->ram_addr + offset) >> TARGET_PAGE_BITS;
363

J
Juan Quintela 已提交
364 365 366 367
    ret = test_and_set_bit(nr, migration_bitmap);

    if (!ret) {
        migration_dirty_pages++;
368
    }
J
Juan Quintela 已提交
369
    return ret;
370 371
}

372 373
/* Needs iothread lock! */

374 375
static void migration_bitmap_sync(void)
{
J
Juan Quintela 已提交
376 377 378
    RAMBlock *block;
    ram_addr_t addr;
    uint64_t num_dirty_pages_init = migration_dirty_pages;
379 380 381 382 383 384 385 386
    MigrationState *s = migrate_get_current();
    static int64_t start_time;
    static int64_t num_dirty_pages_period;
    int64_t end_time;

    if (!start_time) {
        start_time = qemu_get_clock_ms(rt_clock);
    }
387 388

    trace_migration_bitmap_sync_start();
389
    address_space_sync_dirty_bitmap(&address_space_memory);
J
Juan Quintela 已提交
390

P
Paolo Bonzini 已提交
391
    QTAILQ_FOREACH(block, &ram_list.blocks, next) {
J
Juan Quintela 已提交
392
        for (addr = 0; addr < block->length; addr += TARGET_PAGE_SIZE) {
393 394 395
            if (memory_region_test_and_clear_dirty(block->mr,
                                                   addr, TARGET_PAGE_SIZE,
                                                   DIRTY_MEMORY_MIGRATION)) {
J
Juan Quintela 已提交
396 397 398 399 400
                migration_bitmap_set_dirty(block->mr, addr);
            }
        }
    }
    trace_migration_bitmap_sync_end(migration_dirty_pages
401
                                    - num_dirty_pages_init);
402 403 404 405 406 407 408
    num_dirty_pages_period += migration_dirty_pages - num_dirty_pages_init;
    end_time = qemu_get_clock_ms(rt_clock);

    /* more than 1 second = 1000 millisecons */
    if (end_time > start_time + 1000) {
        s->dirty_pages_rate = num_dirty_pages_period * 1000
            / (end_time - start_time);
409
        s->dirty_bytes_rate = s->dirty_pages_rate * TARGET_PAGE_SIZE;
410 411 412
        start_time = end_time;
        num_dirty_pages_period = 0;
    }
413 414
}

415 416 417
/*
 * ram_save_block: Writes a page of memory to the stream f
 *
418 419
 * Returns:  The number of bytes written.
 *           0 means no dirty pages
420 421
 */

422
static int ram_save_block(QEMUFile *f, bool last_stage)
423
{
424
    RAMBlock *block = last_seen_block;
425
    ram_addr_t offset = last_offset;
426
    bool complete_round = false;
427
    int bytes_sent = 0;
A
Avi Kivity 已提交
428
    MemoryRegion *mr;
429
    ram_addr_t current_addr;
430

431
    if (!block)
P
Paolo Bonzini 已提交
432
        block = QTAILQ_FIRST(&ram_list.blocks);
433

434
    while (true) {
A
Avi Kivity 已提交
435
        mr = block->mr;
436 437 438 439 440 441 442 443 444 445 446
        offset = migration_bitmap_find_and_reset_dirty(mr, offset);
        if (complete_round && block == last_seen_block &&
            offset >= last_offset) {
            break;
        }
        if (offset >= block->length) {
            offset = 0;
            block = QTAILQ_NEXT(block, next);
            if (!block) {
                block = QTAILQ_FIRST(&ram_list.blocks);
                complete_round = true;
447
                ram_bulk_stage = false;
448 449
            }
        } else {
450
            uint8_t *p;
J
Juan Quintela 已提交
451
            int cont = (block == last_sent_block) ?
452
                RAM_SAVE_FLAG_CONTINUE : 0;
453

A
Avi Kivity 已提交
454
            p = memory_region_get_ram_ptr(mr) + offset;
455

456 457
            /* In doubt sent page as normal */
            bytes_sent = -1;
458
            if (is_zero_page(p)) {
459
                acct_info.dup_pages++;
460 461 462 463 464 465 466 467 468
                if (!ram_bulk_stage) {
                    bytes_sent = save_block_hdr(f, block, offset, cont,
                                                RAM_SAVE_FLAG_COMPRESS);
                    qemu_put_byte(f, 0);
                    bytes_sent++;
                } else {
                    acct_info.skipped_pages++;
                    bytes_sent = 0;
                }
469
            } else if (!ram_bulk_stage && migrate_use_xbzrle()) {
470 471
                current_addr = block->offset + offset;
                bytes_sent = save_xbzrle_page(f, p, current_addr, block,
472 473 474 475
                                              offset, cont, last_stage);
                if (!last_stage) {
                    p = get_cached_data(XBZRLE.cache, current_addr);
                }
476 477
            }

478
            /* XBZRLE overflow or normal page */
479
            if (bytes_sent == -1) {
480
                bytes_sent = save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_PAGE);
481
                qemu_put_buffer_async(f, p, TARGET_PAGE_SIZE);
482
                bytes_sent += TARGET_PAGE_SIZE;
483
                acct_info.norm_pages++;
484 485
            }

486
            /* if page is unmodified, continue to the next */
487
            if (bytes_sent > 0) {
J
Juan Quintela 已提交
488
                last_sent_block = block;
489 490
                break;
            }
491
        }
492
    }
493
    last_seen_block = block;
494
    last_offset = offset;
495

496
    return bytes_sent;
497 498 499 500 501 502
}

static uint64_t bytes_transferred;

static ram_addr_t ram_save_remaining(void)
{
J
Juan Quintela 已提交
503
    return migration_dirty_pages;
504 505 506 507 508 509 510 511 512 513 514 515 516 517
}

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)
{
518 519 520
    RAMBlock *block;
    uint64_t total = 0;

P
Paolo Bonzini 已提交
521
    QTAILQ_FOREACH(block, &ram_list.blocks, next)
522 523 524
        total += block->length;

    return total;
525 526
}

O
Orit Wasserman 已提交
527 528
static void migration_end(void)
{
529 530 531 532 533
    if (migration_bitmap) {
        memory_global_dirty_log_stop();
        g_free(migration_bitmap);
        migration_bitmap = NULL;
    }
534

535
    if (XBZRLE.cache) {
536 537 538 539 540 541 542
        cache_fini(XBZRLE.cache);
        g_free(XBZRLE.cache);
        g_free(XBZRLE.encoded_buf);
        g_free(XBZRLE.current_buf);
        g_free(XBZRLE.decoded_buf);
        XBZRLE.cache = NULL;
    }
O
Orit Wasserman 已提交
543 544
}

545 546 547 548 549
static void ram_migration_cancel(void *opaque)
{
    migration_end();
}

550 551
static void reset_ram_globals(void)
{
552
    last_seen_block = NULL;
J
Juan Quintela 已提交
553
    last_sent_block = NULL;
554
    last_offset = 0;
U
Umesh Deshpande 已提交
555
    last_version = ram_list.version;
556
    ram_bulk_stage = true;
557 558
}

559 560
#define MAX_WAIT 50 /* ms, half buffered_file limit */

561
static int ram_save_setup(QEMUFile *f, void *opaque)
562
{
563
    RAMBlock *block;
J
Juan Quintela 已提交
564 565 566
    int64_t ram_pages = last_ram_offset() >> TARGET_PAGE_BITS;

    migration_bitmap = bitmap_new(ram_pages);
567
    bitmap_set(migration_bitmap, 0, ram_pages);
J
Juan Quintela 已提交
568
    migration_dirty_pages = ram_pages;
569

570 571 572 573 574 575 576 577 578 579
    if (migrate_use_xbzrle()) {
        XBZRLE.cache = cache_init(migrate_xbzrle_cache_size() /
                                  TARGET_PAGE_SIZE,
                                  TARGET_PAGE_SIZE);
        if (!XBZRLE.cache) {
            DPRINTF("Error creating cache\n");
            return -1;
        }
        XBZRLE.encoded_buf = g_malloc0(TARGET_PAGE_SIZE);
        XBZRLE.current_buf = g_malloc(TARGET_PAGE_SIZE);
580
        acct_clear();
581 582
    }

583 584 585 586 587
    qemu_mutex_lock_iothread();
    qemu_mutex_lock_ramlist();
    bytes_transferred = 0;
    reset_ram_globals();

588
    memory_global_dirty_log_start();
J
Juan Quintela 已提交
589
    migration_bitmap_sync();
590
    qemu_mutex_unlock_iothread();
591

592
    qemu_put_be64(f, ram_bytes_total() | RAM_SAVE_FLAG_MEM_SIZE);
593

P
Paolo Bonzini 已提交
594
    QTAILQ_FOREACH(block, &ram_list.blocks, next) {
595 596 597
        qemu_put_byte(f, strlen(block->idstr));
        qemu_put_buffer(f, (uint8_t *)block->idstr, strlen(block->idstr));
        qemu_put_be64(f, block->length);
598 599
    }

600
    qemu_mutex_unlock_ramlist();
601 602 603 604 605
    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);

    return 0;
}

606
static int ram_save_iterate(QEMUFile *f, void *opaque)
607 608 609
{
    int ret;
    int i;
610
    int64_t t0;
611
    int total_sent = 0;
612

613 614
    qemu_mutex_lock_ramlist();

U
Umesh Deshpande 已提交
615 616 617 618
    if (ram_list.version != last_version) {
        reset_ram_globals();
    }

619
    t0 = qemu_get_clock_ns(rt_clock);
620
    i = 0;
621
    while ((ret = qemu_file_rate_limit(f)) == 0) {
622
        int bytes_sent;
623

624
        bytes_sent = ram_save_block(f, false);
625
        /* no more blocks to sent */
626
        if (bytes_sent == 0) {
627 628
            break;
        }
629
        total_sent += bytes_sent;
630
        acct_info.iterations++;
631 632 633 634 635 636
        /* we want to check in the 1st loop, just in case it was the 1st time
           and we had to sync the dirty bitmap.
           qemu_get_clock_ns() is a bit expensive, so we only check each some
           iterations
        */
        if ((i & 63) == 0) {
637
            uint64_t t1 = (qemu_get_clock_ns(rt_clock) - t0) / 1000000;
638
            if (t1 > MAX_WAIT) {
639
                DPRINTF("big wait: %" PRIu64 " milliseconds, %d iterations\n",
640 641 642 643 644
                        t1, i);
                break;
            }
        }
        i++;
645 646
    }

647 648
    qemu_mutex_unlock_ramlist();

649
    if (ret < 0) {
650
        bytes_transferred += total_sent;
651 652 653
        return ret;
    }

654
    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
655 656
    total_sent += 8;
    bytes_transferred += total_sent;
657

658
    return total_sent;
659 660 661 662
}

static int ram_save_complete(QEMUFile *f, void *opaque)
{
663
    qemu_mutex_lock_ramlist();
664
    migration_bitmap_sync();
665

666
    /* try transferring iterative blocks of memory */
O
Orit Wasserman 已提交
667

668
    /* flush all remaining blocks regardless of rate limiting */
669
    while (true) {
670 671
        int bytes_sent;

672
        bytes_sent = ram_save_block(f, true);
673
        /* no more blocks to sent */
674
        if (bytes_sent == 0) {
675
            break;
676
        }
677
        bytes_transferred += bytes_sent;
678
    }
679
    migration_end();
680

681
    qemu_mutex_unlock_ramlist();
682 683
    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);

684
    return 0;
685 686
}

687 688 689 690 691 692 693
static uint64_t ram_save_pending(QEMUFile *f, void *opaque, uint64_t max_size)
{
    uint64_t remaining_size;

    remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;

    if (remaining_size < max_size) {
694
        qemu_mutex_lock_iothread();
695
        migration_bitmap_sync();
696
        qemu_mutex_unlock_iothread();
697 698 699 700 701
        remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;
    }
    return remaining_size;
}

702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742
static int load_xbzrle(QEMUFile *f, ram_addr_t addr, void *host)
{
    int ret, rc = 0;
    unsigned int xh_len;
    int xh_flags;

    if (!XBZRLE.decoded_buf) {
        XBZRLE.decoded_buf = g_malloc(TARGET_PAGE_SIZE);
    }

    /* extract RLE header */
    xh_flags = qemu_get_byte(f);
    xh_len = qemu_get_be16(f);

    if (xh_flags != ENCODING_FLAG_XBZRLE) {
        fprintf(stderr, "Failed to load XBZRLE page - wrong compression!\n");
        return -1;
    }

    if (xh_len > TARGET_PAGE_SIZE) {
        fprintf(stderr, "Failed to load XBZRLE page - len overflow!\n");
        return -1;
    }
    /* load data and decode */
    qemu_get_buffer(f, XBZRLE.decoded_buf, xh_len);

    /* decode RLE */
    ret = xbzrle_decode_buffer(XBZRLE.decoded_buf, xh_len, host,
                               TARGET_PAGE_SIZE);
    if (ret == -1) {
        fprintf(stderr, "Failed to load XBZRLE page - decode error!\n");
        rc = -1;
    } else  if (ret > TARGET_PAGE_SIZE) {
        fprintf(stderr, "Failed to load XBZRLE page - size %d exceeds %d!\n",
                ret, TARGET_PAGE_SIZE);
        abort();
    }

    return rc;
}

743 744 745 746 747 748 749 750 751 752 753 754 755 756
static inline void *host_from_stream_offset(QEMUFile *f,
                                            ram_addr_t offset,
                                            int flags)
{
    static RAMBlock *block = NULL;
    char id[256];
    uint8_t len;

    if (flags & RAM_SAVE_FLAG_CONTINUE) {
        if (!block) {
            fprintf(stderr, "Ack, bad migration stream!\n");
            return NULL;
        }

757
        return memory_region_get_ram_ptr(block->mr) + offset;
758 759 760 761 762 763
    }

    len = qemu_get_byte(f);
    qemu_get_buffer(f, (uint8_t *)id, len);
    id[len] = 0;

P
Paolo Bonzini 已提交
764
    QTAILQ_FOREACH(block, &ram_list.blocks, next) {
765
        if (!strncmp(id, block->idstr, sizeof(id)))
766
            return memory_region_get_ram_ptr(block->mr) + offset;
767 768 769 770 771 772
    }

    fprintf(stderr, "Can't find block %s!\n", id);
    return NULL;
}

773
static int ram_load(QEMUFile *f, void *opaque, int version_id)
774 775
{
    ram_addr_t addr;
O
Orit Wasserman 已提交
776
    int flags, ret = 0;
777
    int error;
O
Orit Wasserman 已提交
778 779 780
    static uint64_t seq_iter;

    seq_iter++;
781

782
    if (version_id < 4 || version_id > 4) {
783 784 785 786 787 788 789 790 791 792
        return -EINVAL;
    }

    do {
        addr = qemu_get_be64(f);

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

        if (flags & RAM_SAVE_FLAG_MEM_SIZE) {
793
            if (version_id == 4) {
794 795 796 797 798 799 800 801 802 803 804 805 806 807
                /* Synchronize RAM block list */
                char id[256];
                ram_addr_t length;
                ram_addr_t total_ram_bytes = addr;

                while (total_ram_bytes) {
                    RAMBlock *block;
                    uint8_t len;

                    len = qemu_get_byte(f);
                    qemu_get_buffer(f, (uint8_t *)id, len);
                    id[len] = 0;
                    length = qemu_get_be64(f);

P
Paolo Bonzini 已提交
808
                    QTAILQ_FOREACH(block, &ram_list.blocks, next) {
809
                        if (!strncmp(id, block->idstr, sizeof(id))) {
O
Orit Wasserman 已提交
810
                            if (block->length != length) {
811 812 813
                                fprintf(stderr, "Length mismatch: %s: %ld "
                                        "in != " RAM_ADDR_FMT "\n", id, length,
                                        block->length);
O
Orit Wasserman 已提交
814 815 816
                                ret =  -EINVAL;
                                goto done;
                            }
817 818 819 820 821
                            break;
                        }
                    }

                    if (!block) {
822 823
                        fprintf(stderr, "Unknown ramblock \"%s\", cannot "
                                "accept migration\n", id);
O
Orit Wasserman 已提交
824 825
                        ret = -EINVAL;
                        goto done;
826 827 828 829
                    }

                    total_ram_bytes -= length;
                }
830 831 832 833
            }
        }

        if (flags & RAM_SAVE_FLAG_COMPRESS) {
834 835 836
            void *host;
            uint8_t ch;

837
            host = host_from_stream_offset(f, addr, flags);
838 839 840
            if (!host) {
                return -EINVAL;
            }
841 842 843

            ch = qemu_get_byte(f);
            memset(host, ch, TARGET_PAGE_SIZE);
844 845
#ifndef _WIN32
            if (ch == 0 &&
846 847
                (!kvm_enabled() || kvm_has_sync_mmu()) &&
                getpagesize() <= TARGET_PAGE_SIZE) {
A
Andreas Färber 已提交
848
                qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
849 850 851
            }
#endif
        } else if (flags & RAM_SAVE_FLAG_PAGE) {
852 853
            void *host;

854
            host = host_from_stream_offset(f, addr, flags);
855 856 857
            if (!host) {
                return -EINVAL;
            }
858 859

            qemu_get_buffer(f, host, TARGET_PAGE_SIZE);
860 861 862 863 864 865 866 867 868 869
        } else if (flags & RAM_SAVE_FLAG_XBZRLE) {
            void *host = host_from_stream_offset(f, addr, flags);
            if (!host) {
                return -EINVAL;
            }

            if (load_xbzrle(f, addr, host) < 0) {
                ret = -EINVAL;
                goto done;
            }
870
        }
871 872
        error = qemu_file_get_error(f);
        if (error) {
O
Orit Wasserman 已提交
873 874
            ret = error;
            goto done;
875 876 877
        }
    } while (!(flags & RAM_SAVE_FLAG_EOS));

O
Orit Wasserman 已提交
878
done:
879 880
    DPRINTF("Completed load of VM with exit code %d seq iteration "
            "%" PRIu64 "\n", ret, seq_iter);
O
Orit Wasserman 已提交
881
    return ret;
882 883
}

884
SaveVMHandlers savevm_ram_handlers = {
885
    .save_live_setup = ram_save_setup,
886 887
    .save_live_iterate = ram_save_iterate,
    .save_live_complete = ram_save_complete,
888
    .save_live_pending = ram_save_pending,
889
    .load_state = ram_load,
890
    .cancel = ram_migration_cancel,
891 892
};

I
Isaku Yamahata 已提交
893 894 895 896 897 898
struct soundhw {
    const char *name;
    const char *descr;
    int enabled;
    int isa;
    union {
899
        int (*init_isa) (ISABus *bus);
I
Isaku Yamahata 已提交
900 901 902 903
        int (*init_pci) (PCIBus *bus);
    } init;
};

904 905
static struct soundhw soundhw[9];
static int soundhw_count;
906

907 908 909 910 911 912 913 914 915 916
void isa_register_soundhw(const char *name, const char *descr,
                          int (*init_isa)(ISABus *bus))
{
    assert(soundhw_count < ARRAY_SIZE(soundhw) - 1);
    soundhw[soundhw_count].name = name;
    soundhw[soundhw_count].descr = descr;
    soundhw[soundhw_count].isa = 1;
    soundhw[soundhw_count].init.init_isa = init_isa;
    soundhw_count++;
}
917

918 919 920 921 922 923 924 925 926 927
void pci_register_soundhw(const char *name, const char *descr,
                          int (*init_pci)(PCIBus *bus))
{
    assert(soundhw_count < ARRAY_SIZE(soundhw) - 1);
    soundhw[soundhw_count].name = name;
    soundhw[soundhw_count].descr = descr;
    soundhw[soundhw_count].isa = 0;
    soundhw[soundhw_count].init.init_pci = init_pci;
    soundhw_count++;
}
928 929 930 931 932

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

933
    if (is_help_option(optarg)) {
934 935
    show_valid_cards:

936 937 938 939 940 941 942 943 944
        if (soundhw_count) {
             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");
        } else {
             printf("Machine has no user-selectable audio hardware "
                    "(it may or may not have always-present audio hardware).\n");
945
        }
946
        exit(!is_help_option(optarg));
947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991
    }
    else {
        size_t l;
        const char *p;
        char *e;
        int bad_card = 0;

        if (!strcmp(optarg, "all")) {
            for (c = soundhw; c->name; ++c) {
                c->enabled = 1;
            }
            return;
        }

        p = optarg;
        while (*p) {
            e = strchr(p, ',');
            l = !e ? strlen(p) : (size_t) (e - p);

            for (c = soundhw; c->name; ++c) {
                if (!strncmp(c->name, p, l) && !c->name[l]) {
                    c->enabled = 1;
                    break;
                }
            }

            if (!c->name) {
                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;
        }
    }
}
I
Isaku Yamahata 已提交
992

993
void audio_init(void)
I
Isaku Yamahata 已提交
994 995
{
    struct soundhw *c;
996 997
    ISABus *isa_bus = (ISABus *) object_resolve_path_type("", TYPE_ISA_BUS, NULL);
    PCIBus *pci_bus = (PCIBus *) object_resolve_path_type("", TYPE_PCI_BUS, NULL);
I
Isaku Yamahata 已提交
998 999 1000 1001

    for (c = soundhw; c->name; ++c) {
        if (c->enabled) {
            if (c->isa) {
1002 1003 1004
                if (!isa_bus) {
                    fprintf(stderr, "ISA bus not available for %s\n", c->name);
                    exit(1);
I
Isaku Yamahata 已提交
1005
                }
1006
                c->init.init_isa(isa_bus);
I
Isaku Yamahata 已提交
1007
            } else {
1008 1009 1010
                if (!pci_bus) {
                    fprintf(stderr, "PCI bus not available for %s\n", c->name);
                    exit(1);
I
Isaku Yamahata 已提交
1011
                }
1012
                c->init.init_pci(pci_bus);
I
Isaku Yamahata 已提交
1013 1014 1015 1016
            }
        }
    }
}
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034

int qemu_uuid_parse(const char *str, uint8_t *uuid)
{
    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;
    }
#ifdef TARGET_I386
1035
    smbios_add_field(1, offsetof(struct smbios_type_1, uuid), uuid, 16);
1036 1037 1038 1039
#endif
    return 0;
}

1040
void do_acpitable_option(const QemuOpts *opts)
1041 1042
{
#ifdef TARGET_I386
1043 1044 1045 1046 1047 1048 1049
    Error *err = NULL;

    acpi_table_add(opts, &err);
    if (err) {
        fprintf(stderr, "Wrong acpi table provided: %s\n",
                error_get_pretty(err));
        error_free(err);
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
        exit(1);
    }
#endif
}

void do_smbios_option(const char *optarg)
{
#ifdef TARGET_I386
    if (smbios_entry_add(optarg) < 0) {
        exit(1);
    }
#endif
}

void cpudef_init(void)
{
#if defined(cpudef_setup)
    cpudef_setup(); /* parse cpu definitions in target config file */
#endif
}

1071 1072 1073 1074 1075
int tcg_available(void)
{
    return 1;
}

1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
int kvm_available(void)
{
#ifdef CONFIG_KVM
    return 1;
#else
    return 0;
#endif
}

int xen_available(void)
{
#ifdef CONFIG_XEN
    return 1;
#else
    return 0;
#endif
}
1093 1094 1095 1096 1097 1098


TargetInfo *qmp_query_target(Error **errp)
{
    TargetInfo *info = g_malloc0(sizeof(*info));

P
Paolo Bonzini 已提交
1099
    info->arch = g_strdup(TARGET_NAME);
1100 1101 1102

    return info;
}