block-qcow.c

qemu虚拟机代码

/*
 * Block driver for the QCOW format
 * 
 * Copyright (c) 2004 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 "vl.h"
#include "block_int.h"
#include <zlib.h>
#include "aes.h"

/**************************************************************/
/* QEMU COW block driver with compression and encryption support */

#define QCOW_MAGIC (('Q' << 24) | ('F' << 16) | ('I' << 8) | 0xfb)
#define QCOW_VERSION 1

#define QCOW_CRYPT_NONE 0
#define QCOW_CRYPT_AES  1

#define QCOW_OFLAG_COMPRESSED (1LL << 63)

typedef struct QCowHeader {
    uint32_t magic;
    uint32_t version;
    uint64_t backing_file_offset;
    uint32_t backing_file_size;
    uint32_t mtime;
    uint64_t size; /* in bytes */
    uint8_t cluster_bits;
    uint8_t l2_bits;
    uint32_t crypt_method;
    uint64_t l1_table_offset;
} QCowHeader;

#define L2_CACHE_SIZE 16

typedef struct BDRVQcowState {
    int fd;
    int cluster_bits;
    int cluster_size;
    int cluster_sectors;
    int l2_bits;
    int l2_size;
    int l1_size;
    uint64_t cluster_offset_mask;
    uint64_t l1_table_offset;
    uint64_t *l1_table;
    uint64_t *l2_cache;
    uint64_t l2_cache_offsets[L2_CACHE_SIZE];
    uint32_t l2_cache_counts[L2_CACHE_SIZE];
    uint8_t *cluster_cache;
    uint8_t *cluster_data;
    uint64_t cluster_cache_offset;
    uint32_t crypt_method; /* current crypt method, 0 if no key yet */
    uint32_t crypt_method_header;
    AES_KEY aes_encrypt_key;
    AES_KEY aes_decrypt_key;
} BDRVQcowState;

static int decompress_cluster(BDRVQcowState *s, uint64_t cluster_offset);

static int qcow_probe(const uint8_t *buf, int buf_size, const char *filename)
{
    const QCowHeader *cow_header = (const void *)buf;
    
    if (buf_size >= sizeof(QCowHeader) &&
        be32_to_cpu(cow_header->magic) == QCOW_MAGIC &&
        be32_to_cpu(cow_header->version) == QCOW_VERSION) 
        return 100;
    else
        return 0;
}

static int qcow_open(BlockDriverState *bs, const char *filename)
{
    BDRVQcowState *s = bs->opaque;
    int fd, len, i, shift;
    QCowHeader header;
    
    fd = open(filename, O_RDWR | O_BINARY | O_LARGEFILE);
    if (fd < 0) {
        fd = open(filename, O_RDONLY | O_BINARY | O_LARGEFILE);
        if (fd < 0)
            return -1;
    }
    s->fd = fd;
    if (read(fd, &header, sizeof(header)) != sizeof(header))
        goto fail;
    be32_to_cpus(&header.magic);
    be32_to_cpus(&header.version);
    be64_to_cpus(&header.backing_file_offset);
    be32_to_cpus(&header.backing_file_size);
    be32_to_cpus(&header.mtime);
    be64_to_cpus(&header.size);
    be32_to_cpus(&header.crypt_method);
    be64_to_cpus(&header.l1_table_offset);
    
    if (header.magic != QCOW_MAGIC || header.version != QCOW_VERSION)
        goto fail;
    if (header.size <= 1 || header.cluster_bits < 9)
        goto fail;
    if (header.crypt_method > QCOW_CRYPT_AES)
        goto fail;
    s->crypt_method_header = header.crypt_method;
    if (s->crypt_method_header)
        bs->encrypted = 1;
    s->cluster_bits = header.cluster_bits;
    s->cluster_size = 1 << s->cluster_bits;
    s->cluster_sectors = 1 << (s->cluster_bits - 9);
    s->l2_bits = header.l2_bits;
    s->l2_size = 1 << s->l2_bits;
    bs->total_sectors = header.size / 512;
    s->cluster_offset_mask = (1LL << (63 - s->cluster_bits)) - 1;

    /* read the level 1 table */
    shift = s->cluster_bits + s->l2_bits;
    s->l1_size = (header.size + (1LL << shift) - 1) >> shift;

    s->l1_table_offset = header.l1_table_offset;
    s->l1_table = qemu_malloc(s->l1_size * sizeof(uint64_t));
    if (!s->l1_table)
        goto fail;
    lseek(fd, s->l1_table_offset, SEEK_SET);
    if (read(fd, s->l1_table, s->l1_size * sizeof(uint64_t)) != 
        s->l1_size * sizeof(uint64_t))
        goto fail;
    for(i = 0;i < s->l1_size; i++) {
        be64_to_cpus(&s->l1_table[i]);
    }
    /* alloc L2 cache */
    s->l2_cache = qemu_malloc(s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
    if (!s->l2_cache)
        goto fail;
    s->cluster_cache = qemu_malloc(s->cluster_size);
    if (!s->cluster_cache)
        goto fail;
    s->cluster_data = qemu_malloc(s->cluster_size);
    if (!s->cluster_data)
        goto fail;
    s->cluster_cache_offset = -1;
    
    /* read the backing file name */
    if (header.backing_file_offset != 0) {
        len = header.backing_file_size;
        if (len > 1023)
            len = 1023;
        lseek(fd, header.backing_file_offset, SEEK_SET);
        if (read(fd, bs->backing_file, len) != len)
            goto fail;
        bs->backing_file[len] = '\0';
    }
    return 0;

 fail:
    qemu_free(s->l1_table);
    qemu_free(s->l2_cache);
    qemu_free(s->cluster_cache);
    qemu_free(s->cluster_data);
    close(fd);
    return -1;
}

static int qcow_set_key(BlockDriverState *bs, const char *key)
{
    BDRVQcowState *s = bs->opaque;
    uint8_t keybuf[16];
    int len, i;
    
    memset(keybuf, 0, 16);
    len = strlen(key);
    if (len > 16)
        len = 16;
    /* XXX: we could compress the chars to 7 bits to increase
       entropy */
    for(i = 0;i < len;i++) {
        keybuf[i] = key[i];
    }
    s->crypt_method = s->crypt_method_header;

    if (AES_set_encrypt_key(keybuf, 128, &s->aes_encrypt_key) != 0)
        return -1;
    if (AES_set_decrypt_key(keybuf, 128, &s->aes_decrypt_key) != 0)
        return -1;
#if 0
    /* test */
    {
        uint8_t in[16];
        uint8_t out[16];
        uint8_t tmp[16];
        for(i=0;i<16;i++)
            in[i] = i;
        AES_encrypt(in, tmp, &s->aes_encrypt_key);
        AES_decrypt(tmp, out, &s->aes_decrypt_key);
        for(i = 0; i < 16; i++)
            printf(" %02x", tmp[i]);
        printf("\n");
        for(i = 0; i < 16; i++)
            printf(" %02x", out[i]);
        printf("\n");
    }
#endif
    return 0;
}

/* The crypt function is compatible with the linux cryptoloop
   algorithm for < 4 GB images. NOTE: out_buf == in_buf is
   supported */
static void encrypt_sectors(BDRVQcowState *s, int64_t sector_num,
                            uint8_t *out_buf, const uint8_t *in_buf,
                            int nb_sectors, int enc,
                            const AES_KEY *key)
{
    union {
        uint64_t ll[2];
        uint8_t b[16];
    } ivec;
    int i;

    for(i = 0; i < nb_sectors; i++) {
        ivec.ll[0] = cpu_to_le64(sector_num);
        ivec.ll[1] = 0;
        AES_cbc_encrypt(in_buf, out_buf, 512, key, 
                        ivec.b, enc);
        sector_num++;
        in_buf += 512;
        out_buf += 512;
    }
}

/* 'allocate' is:
 *
 * 0 to not allocate.
 *
 * 1 to allocate a normal cluster (for sector indexes 'n_start' to
 * 'n_end')
 *
 * 2 to allocate a compressed cluster of size
 * 'compressed_size'. 'compressed_size' must be > 0 and <
 * cluster_size 
 *
 * return 0 if not allocated.
 */
static uint64_t get_cluster_offset(BlockDriverState *bs,
                                   uint64_t offset, int allocate,
                                   int compressed_size,
                                   int n_start, int n_end)
{
    BDRVQcowState *s = bs->opaque;
    int min_index, i, j, l1_index, l2_index;
    uint64_t l2_offset, *l2_table, cluster_offset, tmp;
    uint32_t min_count;
    int new_l2_table;
    
    l1_index = offset >> (s->l2_bits + s->cluster_bits);
    l2_offset = s->l1_table[l1_index];
    new_l2_table = 0;
    if (!l2_offset) {
        if (!allocate)
            return 0;
        /* allocate a new l2 entry */
        l2_offset = lseek(s->fd, 0, SEEK_END);
        /* round to cluster size */
        l2_offset = (l2_offset + s->cluster_size - 1) & ~(s->cluster_size - 1);
        /* update the L1 entry */
        s->l1_table[l1_index] = l2_offset;
        tmp = cpu_to_be64(l2_offset);
        lseek(s->fd, s->l1_table_offset + l1_index * sizeof(tmp), SEEK_SET);
        if (write(s->fd, &tmp, sizeof(tmp)) != sizeof(tmp))
            return 0;
        new_l2_table = 1;
    }
    for(i = 0; i < L2_CACHE_SIZE; i++) {
        if (l2_offset == s->l2_cache_offsets[i]) {
            /* increment the hit count */
            if (++s->l2_cache_counts[i] == 0xffffffff) {
                for(j = 0; j < L2_CACHE_SIZE; j++) {
                    s->l2_cache_counts[j] >>= 1;
                }
            }
            l2_table = s->l2_cache + (i << s->l2_bits);
            goto found;
        }
    }
    /* not found: load a new entry in the least used one */
    min_index = 0;
    min_count = 0xffffffff;
    for(i = 0; i < L2_CACHE_SIZE; i++) {
        if (s->l2_cache_counts[i] < min_count) {
            min_count = s->l2_cache_counts[i];
            min_index = i;
        }
    }
    l2_table = s->l2_cache + (min_index << s->l2_bits);
    lseek(s->fd, l2_offset, SEEK_SET);
    if (new_l2_table) {
        memset(l2_table, 0, s->l2_size * sizeof(uint64_t));
        if (write(s->fd, l2_table, s->l2_size * sizeof(uint64_t)) !=
            s->l2_size * sizeof(uint64_t))
            return 0;
    } else {
        if (read(s->fd, l2_table, s->l2_size * sizeof(uint64_t)) != 
            s->l2_size * sizeof(uint64_t))
            return 0;
    }
    s->l2_cache_offsets[min_index] = l2_offset;
    s->l2_cache_counts[min_index] = 1;
 found:
    l2_index = (offset >> s->cluster_bits) & (s->l2_size - 1);
    cluster_offset = be64_to_cpu(l2_table[l2_index]);
    if (!cluster_offset || 
        ((cluster_offset & QCOW_OFLAG_COMPRESSED) && allocate == 1)) {
        if (!allocate)
            return 0;
        /* allocate a new cluster */
        if ((cluster_offset & QCOW_OFLAG_COMPRESSED) &&
            (n_end - n_start) < s->cluster_sectors) {
            /* if the cluster is already compressed, we must
               decompress it in the case it is not completely
               overwritten */
            if (decompress_cluster(s, cluster_offset) < 0)
                return 0;
            cluster_offset = lseek(s->fd, 0, SEEK_END);
            cluster_offset = (cluster_offset + s->cluster_size - 1) & 
                ~(s->cluster_size - 1);
            /* write the cluster content */
            lseek(s->fd, cluster_offset, SEEK_SET);
            if (write(s->fd, s->cluster_cache, s->cluster_size) != 
                s->cluster_size)
                return -1;
        } else {
            cluster_offset = lseek(s->fd, 0, SEEK_END);
            if (allocate == 1) {
                /* round to cluster size */
                cluster_offset = (cluster_offset + s->cluster_size - 1) & 
                    ~(s->cluster_size - 1);
                ftruncate(s->fd, cluster_offset + s->cluster_size);
                /* if encrypted, we must initialize the cluster
                   content which won't be written */