source: mainline/uspace/lib/block/block.c@ f77c1c9

/*
 * Copyright (c) 2008 Jakub Jermar
 * Copyright (c) 2008 Martin Decky
 * Copyright (c) 2011 Martin Sucha
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * - Redistributions of source code must retain the above copyright
 *   notice, this list of conditions and the following disclaimer.
 * - Redistributions in binary form must reproduce the above copyright
 *   notice, this list of conditions and the following disclaimer in the
 *   documentation and/or other materials provided with the distribution.
 * - The name of the author may not be used to endorse or promote products
 *   derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

/** @addtogroup libblock
 * @{
 */
/**
 * @file
 * @brief
 */

#include <ipc/loc.h>
#include <ipc/services.h>
#include <errno.h>
#include <async.h>
#include <as.h>
#include <assert.h>
#include <bd.h>
#include <fibril_synch.h>
#include <adt/list.h>
#include <adt/hash_table.h>
#include <macros.h>
#include <mem.h>
#include <stdlib.h>
#include <stdio.h>
#include <stacktrace.h>
#include <str_error.h>
#include <offset.h>
#include <inttypes.h>
#include "block.h"

#define MAX_WRITE_RETRIES 10

/** Lock protecting the device connection list */
static FIBRIL_MUTEX_INITIALIZE(dcl_lock);
/** Device connection list head. */
static LIST_INITIALIZE(dcl);


typedef struct {
        fibril_mutex_t lock;
        size_t lblock_size;       /**< Logical block size. */
        unsigned blocks_cluster;  /**< Physical blocks per block_t */
        unsigned block_count;     /**< Total number of blocks. */
        unsigned blocks_cached;   /**< Number of cached blocks. */
        hash_table_t block_hash;
        list_t free_list;
        enum cache_mode mode;
} cache_t;

typedef struct {
        link_t link;
        service_id_t service_id;
        async_sess_t *sess;
        bd_t *bd;
        void *bb_buf;
        aoff64_t bb_addr;
        aoff64_t pblocks;    /**< Number of physical blocks */
        size_t pblock_size;  /**< Physical block size. */
        cache_t *cache;
} devcon_t;

static int read_blocks(devcon_t *, aoff64_t, size_t, void *, size_t);
static int write_blocks(devcon_t *, aoff64_t, size_t, void *, size_t);
static aoff64_t ba_ltop(devcon_t *, aoff64_t);

static devcon_t *devcon_search(service_id_t service_id)
{
        fibril_mutex_lock(&dcl_lock);
        
        list_foreach(dcl, link, devcon_t, devcon) {
                if (devcon->service_id == service_id) {
                        fibril_mutex_unlock(&dcl_lock);
                        return devcon;
                }
        }
        
        fibril_mutex_unlock(&dcl_lock);
        return NULL;
}

static int devcon_add(service_id_t service_id, async_sess_t *sess,
    size_t bsize, aoff64_t dev_size, bd_t *bd)
{
        devcon_t *devcon;
        
        devcon = malloc(sizeof(devcon_t));
        if (!devcon)
                return ENOMEM;
        
        link_initialize(&devcon->link);
        devcon->service_id = service_id;
        devcon->sess = sess;
        devcon->bd = bd;
        devcon->bb_buf = NULL;
        devcon->bb_addr = 0;
        devcon->pblock_size = bsize;
        devcon->pblocks = dev_size;
        devcon->cache = NULL;
        
        fibril_mutex_lock(&dcl_lock);
        list_foreach(dcl, link, devcon_t, d) {
                if (d->service_id == service_id) {
                        fibril_mutex_unlock(&dcl_lock);
                        free(devcon);
                        return EEXIST;
                }
        }
        list_append(&devcon->link, &dcl);
        fibril_mutex_unlock(&dcl_lock);
        return EOK;
}

static void devcon_remove(devcon_t *devcon)
{
        fibril_mutex_lock(&dcl_lock);
        list_remove(&devcon->link);
        fibril_mutex_unlock(&dcl_lock);
}

int block_init(service_id_t service_id, size_t comm_size)
{
        bd_t *bd;

        async_sess_t *sess = loc_service_connect(service_id, INTERFACE_BLOCK,
            IPC_FLAG_BLOCKING);
        if (!sess) {
                return ENOENT;
        }
        
        int rc = bd_open(sess, &bd);
        if (rc != EOK) {
                async_hangup(sess);
                return rc;
        }
        
        size_t bsize;
        rc = bd_get_block_size(bd, &bsize);
        if (rc != EOK) {
                bd_close(bd);
                async_hangup(sess);
                return rc;
        }

        aoff64_t dev_size;
        rc = bd_get_num_blocks(bd, &dev_size);
        if (rc != EOK) {
                bd_close(bd);
                async_hangup(sess);
                return rc;
        }
        
        rc = devcon_add(service_id, sess, bsize, dev_size, bd);
        if (rc != EOK) {
                bd_close(bd);
                async_hangup(sess);
                return rc;
        }
        
        return EOK;
}

void block_fini(service_id_t service_id)
{
        devcon_t *devcon = devcon_search(service_id);
        assert(devcon);
        
        if (devcon->cache)
                (void) block_cache_fini(service_id);
        
        (void)bd_sync_cache(devcon->bd, 0, 0);
        
        devcon_remove(devcon);
        
        if (devcon->bb_buf)
                free(devcon->bb_buf);
        
        bd_close(devcon->bd);
        async_hangup(devcon->sess);
        
        free(devcon);
}

int block_bb_read(service_id_t service_id, aoff64_t ba)
{
        void *bb_buf;
        int rc;

        devcon_t *devcon = devcon_search(service_id);
        if (!devcon)
                return ENOENT;
        if (devcon->bb_buf)
                return EEXIST;
        bb_buf = malloc(devcon->pblock_size);
        if (!bb_buf)
                return ENOMEM;

        rc = read_blocks(devcon, 0, 1, bb_buf, devcon->pblock_size);
        if (rc != EOK) {
                free(bb_buf);
                return rc;
        }

        devcon->bb_buf = bb_buf;
        devcon->bb_addr = ba;

        return EOK;
}

void *block_bb_get(service_id_t service_id)
{
        devcon_t *devcon = devcon_search(service_id);
        assert(devcon);
        return devcon->bb_buf;
}

static size_t cache_key_hash(void *key)
{
        aoff64_t *lba = (aoff64_t*)key;
        return *lba;
}

static size_t cache_hash(const ht_link_t *item)
{
        block_t *b = hash_table_get_inst(item, block_t, hash_link);
        return b->lba;
}

static bool cache_key_equal(void *key, const ht_link_t *item)
{
        aoff64_t *lba = (aoff64_t*)key;
        block_t *b = hash_table_get_inst(item, block_t, hash_link);
        return b->lba == *lba;
}


static hash_table_ops_t cache_ops = {
        .hash = cache_hash,
        .key_hash = cache_key_hash,
        .key_equal = cache_key_equal,
        .equal = NULL,
        .remove_callback = NULL
};

int block_cache_init(service_id_t service_id, size_t size, unsigned blocks,
    enum cache_mode mode)
{
        devcon_t *devcon = devcon_search(service_id);
        cache_t *cache;
        if (!devcon)
                return ENOENT;
        if (devcon->cache)
                return EEXIST;
        cache = malloc(sizeof(cache_t));
        if (!cache)
                return ENOMEM;
        
        fibril_mutex_initialize(&cache->lock);
        list_initialize(&cache->free_list);
        cache->lblock_size = size;
        cache->block_count = blocks;
        cache->blocks_cached = 0;
        cache->mode = mode;

        /* Allow 1:1 or small-to-large block size translation */
        if (cache->lblock_size % devcon->pblock_size != 0) {
                free(cache);
                return ENOTSUP;
        }

        cache->blocks_cluster = cache->lblock_size / devcon->pblock_size;

        if (!hash_table_create(&cache->block_hash, 0, 0, &cache_ops)) {
                free(cache);
                return ENOMEM;
        }

        devcon->cache = cache;
        return EOK;
}

int block_cache_fini(service_id_t service_id)
{
        devcon_t *devcon = devcon_search(service_id);
        cache_t *cache;
        int rc;

        if (!devcon)
                return ENOENT;
        if (!devcon->cache)
                return EOK;
        cache = devcon->cache;
        
        /*
         * We are expecting to find all blocks for this device handle on the
         * free list, i.e. the block reference count should be zero. Do not
         * bother with the cache and block locks because we are single-threaded.
         */
        while (!list_empty(&cache->free_list)) {
                block_t *b = list_get_instance(list_first(&cache->free_list),
                    block_t, free_link);

                list_remove(&b->free_link);
                if (b->dirty) {
                        rc = write_blocks(devcon, b->pba, cache->blocks_cluster,
                            b->data, b->size);
                        if (rc != EOK)
                                return rc;
                }

                hash_table_remove_item(&cache->block_hash, &b->hash_link);
                
                free(b->data);
                free(b);
        }

        hash_table_destroy(&cache->block_hash);
        devcon->cache = NULL;
        free(cache);

        return EOK;
}

#define CACHE_LO_WATERMARK      10      
#define CACHE_HI_WATERMARK      20      
static bool cache_can_grow(cache_t *cache)
{
        if (cache->blocks_cached < CACHE_LO_WATERMARK)
                return true;
        if (!list_empty(&cache->free_list))
                return false;
        return true;
}

static void block_initialize(block_t *b)
{
        fibril_mutex_initialize(&b->lock);
        b->refcnt = 1;
        b->write_failures = 0;
        b->dirty = false;
        b->toxic = false;
        fibril_rwlock_initialize(&b->contents_lock);
        link_initialize(&b->free_link);
}

/** Instantiate a block in memory and get a reference to it.
 *
 * @param block                 Pointer to where the function will store the
 *                              block pointer on success.
 * @param service_id            Service ID of the block device.
 * @param ba                    Block address (logical).
 * @param flags                 If BLOCK_FLAGS_NOREAD is specified, block_get()
 *                              will not read the contents of the block from the
 *                              device.
 *
 * @return                      EOK on success or a negative error code.
 */
int block_get(block_t **block, service_id_t service_id, aoff64_t ba, int flags)
{
        devcon_t *devcon;
        cache_t *cache;
        block_t *b;
        link_t *link;
        aoff64_t p_ba;
        int rc;
        
        devcon = devcon_search(service_id);

        assert(devcon);
        assert(devcon->cache);
        
        cache = devcon->cache;

        /* Check whether the logical block (or part of it) is beyond
         * the end of the device or not.
         */
        p_ba = ba_ltop(devcon, ba);
        p_ba += cache->blocks_cluster;
        if (p_ba >= devcon->pblocks) {
                /* This request cannot be satisfied */
                return EIO;
        }


retry:
        rc = EOK;
        b = NULL;

        fibril_mutex_lock(&cache->lock);
        ht_link_t *hlink = hash_table_find(&cache->block_hash, &ba);
        if (hlink) {
found:
                /*
                 * We found the block in the cache.
                 */
                b = hash_table_get_inst(hlink, block_t, hash_link);
                fibril_mutex_lock(&b->lock);
                if (b->refcnt++ == 0)
                        list_remove(&b->free_link);
                if (b->toxic)
                        rc = EIO;
                fibril_mutex_unlock(&b->lock);
                fibril_mutex_unlock(&cache->lock);
        } else {
                /*
                 * The block was not found in the cache.
                 */
                if (cache_can_grow(cache)) {
                        /*
                         * We can grow the cache by allocating new blocks.
                         * Should the allocation fail, we fail over and try to
                         * recycle a block from the cache.
                         */
                        b = malloc(sizeof(block_t));
                        if (!b)
                                goto recycle;
                        b->data = malloc(cache->lblock_size);
                        if (!b->data) {
                                free(b);
                                b = NULL;
                                goto recycle;
                        }
                        cache->blocks_cached++;
                } else {
                        /*
                         * Try to recycle a block from the free list.
                         */
recycle:
                        if (list_empty(&cache->free_list)) {
                                fibril_mutex_unlock(&cache->lock);
                                rc = ENOMEM;
                                goto out;
                        }
                        link = list_first(&cache->free_list);
                        b = list_get_instance(link, block_t, free_link);

                        fibril_mutex_lock(&b->lock);
                        if (b->dirty) {
                                /*
                                 * The block needs to be written back to the
                                 * device before it changes identity. Do this
                                 * while not holding the cache lock so that
                                 * concurrency is not impeded. Also move the
                                 * block to the end of the free list so that we
                                 * do not slow down other instances of
                                 * block_get() draining the free list.
                                 */
                                list_remove(&b->free_link);
                                list_append(&b->free_link, &cache->free_list);
                                fibril_mutex_unlock(&cache->lock);
                                rc = write_blocks(devcon, b->pba,
                                    cache->blocks_cluster, b->data, b->size);
                                if (rc != EOK) {
                                        /*
                                         * We did not manage to write the block
                                         * to the device. Keep it around for
                                         * another try. Hopefully, we will grab
                                         * another block next time.
                                         */
                                        if (b->write_failures < MAX_WRITE_RETRIES) {
                                                b->write_failures++;
                                                fibril_mutex_unlock(&b->lock);
                                                goto retry;
                                        } else {
                                                printf("Too many errors writing block %"
                                                    PRIuOFF64 "from device handle %" PRIun "\n"
                                                    "SEVERE DATA LOSS POSSIBLE\n",
                                                    b->lba, devcon->service_id);
                                        }
                                } else
                                        b->write_failures = 0;

                                b->dirty = false;
                                if (!fibril_mutex_trylock(&cache->lock)) {
                                        /*
                                         * Somebody is probably racing with us.
                                         * Unlock the block and retry.
                                         */
                                        fibril_mutex_unlock(&b->lock);
                                        goto retry;
                                }
                                hlink = hash_table_find(&cache->block_hash, &ba);
                                if (hlink) {
                                        /*
                                         * Someone else must have already
                                         * instantiated the block while we were
                                         * not holding the cache lock.
                                         * Leave the recycled block on the
                                         * freelist and continue as if we
                                         * found the block of interest during
                                         * the first try.
                                         */
                                        fibril_mutex_unlock(&b->lock);
                                        goto found;
                                }

                        }
                        fibril_mutex_unlock(&b->lock);

                        /*
                         * Unlink the block from the free list and the hash
                         * table.
                         */
                        list_remove(&b->free_link);
                        hash_table_remove_item(&cache->block_hash, &b->hash_link);
                }

                block_initialize(b);
                b->service_id = service_id;
                b->size = cache->lblock_size;
                b->lba = ba;
                b->pba = ba_ltop(devcon, b->lba);
                hash_table_insert(&cache->block_hash, &b->hash_link);

                /*
                 * Lock the block before releasing the cache lock. Thus we don't
                 * kill concurrent operations on the cache while doing I/O on
                 * the block.
                 */
                fibril_mutex_lock(&b->lock);
                fibril_mutex_unlock(&cache->lock);

                if (!(flags & BLOCK_FLAGS_NOREAD)) {
                        /*
                         * The block contains old or no data. We need to read
                         * the new contents from the device.
                         */
                        rc = read_blocks(devcon, b->pba, cache->blocks_cluster,
                            b->data, cache->lblock_size);
                        if (rc != EOK) 
                                b->toxic = true;
                } else
                        rc = EOK;

                fibril_mutex_unlock(&b->lock);
        }
out:
        if ((rc != EOK) && b) {
                assert(b->toxic);
                (void) block_put(b);
                b = NULL;
        }
        *block = b;
        return rc;
}

/** Release a reference to a block.
 *
 * If the last reference is dropped, the block is put on the free list.
 *
 * @param block         Block of which a reference is to be released.
 *
 * @return              EOK on success or a negative error code.
 */
int block_put(block_t *block)
{
        devcon_t *devcon = devcon_search(block->service_id);
        cache_t *cache;
        unsigned blocks_cached;
        enum cache_mode mode;
        int rc = EOK;

        assert(devcon);
        assert(devcon->cache);
        assert(block->refcnt >= 1);

        cache = devcon->cache;

retry:
        fibril_mutex_lock(&cache->lock);
        blocks_cached = cache->blocks_cached;
        mode = cache->mode;
        fibril_mutex_unlock(&cache->lock);

        /*
         * Determine whether to sync the block. Syncing the block is best done
         * when not holding the cache lock as it does not impede concurrency.
         * Since the situation may have changed when we unlocked the cache, the
         * blocks_cached and mode variables are mere hints. We will recheck the
         * conditions later when the cache lock is held again.
         */
        fibril_mutex_lock(&block->lock);
        if (block->toxic)
                block->dirty = false;   /* will not write back toxic block */
        if (block->dirty && (block->refcnt == 1) &&
            (blocks_cached > CACHE_HI_WATERMARK || mode != CACHE_MODE_WB)) {
                rc = write_blocks(devcon, block->pba, cache->blocks_cluster,
                    block->data, block->size);
                if (rc == EOK)
                        block->write_failures = 0;
                block->dirty = false;
        }
        fibril_mutex_unlock(&block->lock);

        fibril_mutex_lock(&cache->lock);
        fibril_mutex_lock(&block->lock);
        if (!--block->refcnt) {
                /*
                 * Last reference to the block was dropped. Either free the
                 * block or put it on the free list. In case of an I/O error,
                 * free the block.
                 */
                if ((cache->blocks_cached > CACHE_HI_WATERMARK) ||
                    (rc != EOK)) {
                        /*
                         * Currently there are too many cached blocks or there
                         * was an I/O error when writing the block back to the
                         * device.
                         */
                        if (block->dirty) {
                                /*
                                 * We cannot sync the block while holding the
                                 * cache lock. Release everything and retry.
                                 */
                                block->refcnt++;

                                if (block->write_failures < MAX_WRITE_RETRIES) {
                                        block->write_failures++;
                                        fibril_mutex_unlock(&block->lock);
                                        fibril_mutex_unlock(&cache->lock);
                                        goto retry;
                                } else {
                                        printf("Too many errors writing block %"
                                            PRIuOFF64 "from device handle %" PRIun "\n"
                                            "SEVERE DATA LOSS POSSIBLE\n",
                                            block->lba, devcon->service_id);
                                }
                        }
                        /*
                         * Take the block out of the cache and free it.
                         */
                        hash_table_remove_item(&cache->block_hash, &block->hash_link);
                        fibril_mutex_unlock(&block->lock);
                        free(block->data);
                        free(block);
                        cache->blocks_cached--;
                        fibril_mutex_unlock(&cache->lock);
                        return rc;
                }
                /*
                 * Put the block on the free list.
                 */
                if (cache->mode != CACHE_MODE_WB && block->dirty) {
                        /*
                         * We cannot sync the block while holding the cache
                         * lock. Release everything and retry.
                         */
                        block->refcnt++;
                        fibril_mutex_unlock(&block->lock);
                        fibril_mutex_unlock(&cache->lock);
                        goto retry;
                }
                list_append(&block->free_link, &cache->free_list);
        }
        fibril_mutex_unlock(&block->lock);
        fibril_mutex_unlock(&cache->lock);

        return rc;
}

/** Read sequential data from a block device.
 *
 * @param service_id    Service ID of the block device.
 * @param buf           Buffer for holding one block
 * @param bufpos        Pointer to the first unread valid offset within the
 *                      communication buffer.
 * @param buflen        Pointer to the number of unread bytes that are ready in
 *                      the communication buffer.
 * @param pos           Device position to be read.
 * @param dst           Destination buffer.
 * @param size          Size of the destination buffer.
 * @param block_size    Block size to be used for the transfer.
 *
 * @return              EOK on success or a negative return code on failure.
 */
int block_seqread(service_id_t service_id, void *buf, size_t *bufpos,
    size_t *buflen, aoff64_t *pos, void *dst, size_t size)
{
        size_t offset = 0;
        size_t left = size;
        size_t block_size;
        devcon_t *devcon;

        devcon = devcon_search(service_id);
        assert(devcon);
        block_size = devcon->pblock_size;
        
        while (left > 0) {
                size_t rd;
                
                if (*bufpos + left < *buflen)
                        rd = left;
                else
                        rd = *buflen - *bufpos;
                
                if (rd > 0) {
                        /*
                         * Copy the contents of the communication buffer to the
                         * destination buffer.
                         */
                        memcpy(dst + offset, buf + *bufpos, rd);
                        offset += rd;
                        *bufpos += rd;
                        *pos += rd;
                        left -= rd;
                }
                
                if (*bufpos == *buflen) {
                        /* Refill the communication buffer with a new block. */
                        int rc;

                        rc = read_blocks(devcon, *pos / block_size, 1, buf,
                            devcon->pblock_size);
                        if (rc != EOK) {
                                return rc;
                        }
                        
                        *bufpos = 0;
                        *buflen = block_size;
                }
        }
        
        return EOK;
}

/** Read blocks directly from device (bypass cache).
 *
 * @param service_id    Service ID of the block device.
 * @param ba            Address of first block (physical).
 * @param cnt           Number of blocks.
 * @param src           Buffer for storing the data.
 *
 * @return              EOK on success or negative error code on failure.
 */
int block_read_direct(service_id_t service_id, aoff64_t ba, size_t cnt, void *buf)
{
        devcon_t *devcon;

        devcon = devcon_search(service_id);
        assert(devcon);

        return read_blocks(devcon, ba, cnt, buf, devcon->pblock_size * cnt);
}

/** Write blocks directly to device (bypass cache).
 *
 * @param service_id    Service ID of the block device.
 * @param ba            Address of first block (physical).
 * @param cnt           Number of blocks.
 * @param src           The data to be written.
 *
 * @return              EOK on success or negative error code on failure.
 */
int block_write_direct(service_id_t service_id, aoff64_t ba, size_t cnt,
    const void *data)
{
        devcon_t *devcon;

        devcon = devcon_search(service_id);
        assert(devcon);

        return write_blocks(devcon, ba, cnt, (void *)data, devcon->pblock_size * cnt);
}

/** Synchronize blocks to persistent storage.
 *
 * @param service_id    Service ID of the block device.
 * @param ba            Address of first block (physical).
 * @param cnt           Number of blocks.
 *
 * @return              EOK on success or negative error code on failure.
 */
int block_sync_cache(service_id_t service_id, aoff64_t ba, size_t cnt)
{
        devcon_t *devcon;

        devcon = devcon_search(service_id);
        assert(devcon);

        return bd_sync_cache(devcon->bd, ba, cnt);
}

/** Get device block size.
 *
 * @param service_id    Service ID of the block device.
 * @param bsize         Output block size.
 *
 * @return              EOK on success or negative error code on failure.
 */
int block_get_bsize(service_id_t service_id, size_t *bsize)
{
        devcon_t *devcon;

        devcon = devcon_search(service_id);
        assert(devcon);

        return bd_get_block_size(devcon->bd, bsize);
}

/** Get number of blocks on device.
 *
 * @param service_id    Service ID of the block device.
 * @param nblocks       Output number of blocks.
 *
 * @return              EOK on success or negative error code on failure.
 */
int block_get_nblocks(service_id_t service_id, aoff64_t *nblocks)
{
        devcon_t *devcon = devcon_search(service_id);
        assert(devcon);

        return bd_get_num_blocks(devcon->bd, nblocks);
}

/** Read bytes directly from the device (bypass cache)
 * 
 * @param service_id    Service ID of the block device.
 * @param abs_offset    Absolute offset in bytes where to start reading
 * @param bytes                 Number of bytes to read
 * @param data                  Buffer that receives the data
 * 
 * @return              EOK on success or negative error code on failure.
 */
int block_read_bytes_direct(service_id_t service_id, aoff64_t abs_offset,
    size_t bytes, void *data)
{
        int rc;
        size_t phys_block_size;
        size_t buf_size;
        void *buffer;
        aoff64_t first_block;
        aoff64_t last_block;
        size_t blocks;
        size_t offset;
        
        rc = block_get_bsize(service_id, &phys_block_size);
        if (rc != EOK) {
                return rc;
        }
        
        /* calculate data position and required space */
        first_block = abs_offset / phys_block_size;
        offset = abs_offset % phys_block_size;
        last_block = (abs_offset + bytes - 1) / phys_block_size;
        blocks = last_block - first_block + 1;
        buf_size = blocks * phys_block_size;
        
        /* read the data into memory */
        buffer = malloc(buf_size);
        if (buffer == NULL) {
                return ENOMEM;
        }
        
        rc = block_read_direct(service_id, first_block, blocks, buffer);
        if (rc != EOK) {
                free(buffer);
                return rc;
        }
        
        /* copy the data from the buffer */
        memcpy(data, buffer + offset, bytes);
        free(buffer);
        
        return EOK;
}

/** Get TOC from device.
 *
 * @param service_id Service ID of the block device.
 * @param session    Starting session.
 *
 * @return Allocated TOC structure.
 * @return EOK on success or negative error code.
 *
 */
int block_read_toc(service_id_t service_id, uint8_t session, void *buf,
    size_t bufsize)
{
        devcon_t *devcon = devcon_search(service_id);
        
        assert(devcon);
        return bd_read_toc(devcon->bd, session, buf, bufsize);
}

/** Read blocks from block device.
 *
 * @param devcon        Device connection.
 * @param ba            Address of first block.
 * @param cnt           Number of blocks.
 * @param src           Buffer for storing the data.
 *
 * @return              EOK on success or negative error code on failure.
 */
static int read_blocks(devcon_t *devcon, aoff64_t ba, size_t cnt, void *buf,
    size_t size)
{
        assert(devcon);
        
        int rc = bd_read_blocks(devcon->bd, ba, cnt, buf, size);
        if (rc != EOK) {
                printf("Error %s reading %zu blocks starting at block %" PRIuOFF64
                    " from device handle %" PRIun "\n", str_error_name(rc), cnt, ba,
                    devcon->service_id);
#ifndef NDEBUG
                stacktrace_print();
#endif
        }
        
        return rc;
}

/** Write block to block device.
 *
 * @param devcon        Device connection.
 * @param ba            Address of first block.
 * @param cnt           Number of blocks.
 * @param src           Buffer containing the data to write.
 *
 * @return              EOK on success or negative error code on failure.
 */
static int write_blocks(devcon_t *devcon, aoff64_t ba, size_t cnt, void *data,
    size_t size)
{
        assert(devcon);
        
        int rc = bd_write_blocks(devcon->bd, ba, cnt, data, size);
        if (rc != EOK) {
                printf("Error %s writing %zu blocks starting at block %" PRIuOFF64
                    " to device handle %" PRIun "\n", str_error_name(rc), cnt, ba, devcon->service_id);
#ifndef NDEBUG
                stacktrace_print();
#endif
        }
        
        return rc;
}

/** Convert logical block address to physical block address. */
static aoff64_t ba_ltop(devcon_t *devcon, aoff64_t lba)
{
        assert(devcon->cache != NULL);
        return lba * devcon->cache->blocks_cluster;
}

/** @}
 */