source: rtems/cpukit/libblock/src/bdbuf.c @ 0ebfac19

/*
 * Disk I/O buffering
 * Buffer managment
 *
 * Copyright (C) 2001 OKTET Ltd., St.-Peterburg, Russia
 * Author: Andrey G. Ivanov <Andrey.Ivanov@oktet.ru>
 *         Victor V. Vengerov <vvv@oktet.ru>
 *         Alexander Kukuta <kam@oktet.ru>
 *
 * @(#) $Id$
 */

#if HAVE_CONFIG_H
#include "config.h"
#endif

#define __RTEMS_VIOLATE_KERNEL_VISIBILITY__
#include <rtems.h>
#include <limits.h>
#include <errno.h>
#include <assert.h>

#include "rtems/bdbuf.h"

/* Fatal errors: */
#define BLKDEV_FATAL_ERROR(n) (((uint32_t)'B' << 24) | ((uint32_t)(n) & (uint32_t)0x00FFFFFF))
#define BLKDEV_FATAL_BDBUF_CONSISTENCY BLKDEV_FATAL_ERROR(1)
#define BLKDEV_FATAL_BDBUF_SWAPOUT     BLKDEV_FATAL_ERROR(2)


#define SWAPOUT_PRIORITY 15
#define SWAPOUT_STACK_SIZE (RTEMS_MINIMUM_STACK_SIZE * 2)

#define READ_MULTIPLE

#if defined(READ_MULTIPLE)
#define READ_AHEAD_MAX_BLK_CNT 32
typedef struct {
  blkdev_request   req;
  blkdev_sg_buffer sg[READ_AHEAD_MAX_BLK_CNT];
} blkdev_request_read_ahead;
/*
 * list of bd_bufs related to one transfer request
 */
typedef struct {
  int cnt;
  bdbuf_buffer *bd_bufs[READ_AHEAD_MAX_BLK_CNT];
} read_ahead_bd_buf_group;
#endif

typedef struct {
  blkdev_request *req;
  bdbuf_buffer   **write_store;
} write_tfer_done_arg_t;

static rtems_task bdbuf_swapout_task(rtems_task_argument unused);

static rtems_status_code bdbuf_release(bdbuf_buffer *bd_buf);
/*
 * The groups of the blocks with the same size are collected in the
 * bd_pool. Note that a several of the buffer's groups with the
 * same size can exists.
 */
typedef struct bdbuf_pool
{
    bdbuf_buffer *tree;         /* Buffer descriptor lookup AVL tree root */

    Chain_Control free;         /* Free buffers list */
    Chain_Control lru;          /* Last recently used list */

    int           blksize;      /* The size of the blocks (in bytes) */
    int           nblks;        /* Number of blocks in this pool */
    rtems_id      bufget_sema;  /* Buffer obtain counting semaphore */
    void         *mallocd_bufs; /* Pointer to the malloc'd buffer memory,
                                   or NULL, if buffer memory provided in
                                   buffer configuration */
    bdbuf_buffer *bdbufs;       /* Pointer to table of buffer descriptors
                                   allocated for this buffer pool. */
} bdbuf_pool;

/* Buffering layer context definition */
struct bdbuf_context {
    bdbuf_pool    *pool;         /* Table of buffer pools */
    int            npools;       /* Number of entries in pool table */

    Chain_Control  mod;          /* Modified buffers list */
    rtems_id       flush_sema;   /* Buffer flush semaphore; counting
                                    semaphore; incremented when buffer
                                    flushed to the disk; decremented when
                                    buffer modified */
    rtems_id       swapout_task; /* Swapout task ID */
};
/*
 * maximum number of blocks that might be chained together to one
 * write driver call
 */
#define SWAP_OUT_MAX_BLK_CNT 32 
/*#define SWAP_OUT_MAX_BLK_CNT 1 */
/*
 * XXX: this is a global buffer. It is used in the swapout task
 * and currently will be reused only after it is no longer in use
 * 
 */
static bdbuf_buffer *bd_buf_write_store[SWAP_OUT_MAX_BLK_CNT];

/* Block device request with a single buffer provided */
typedef struct blkdev_request1 {
    blkdev_request   req;
    blkdev_sg_buffer sg[1];
} blkdev_request1;

/* The static context of buffering layer */
static struct bdbuf_context bd_ctx;

#define SAFE
#ifdef SAFE
typedef rtems_mode preemption_key;

#define DISABLE_PREEMPTION(key) \
    do {                                                               \
        rtems_task_mode(RTEMS_NO_PREEMPT, RTEMS_PREEMPT_MASK, &(key)); \
    } while (0)

#define ENABLE_PREEMPTION(key) \
    do {                                                        \
        rtems_mode temp;                                        \
        rtems_task_mode((key), RTEMS_PREEMPT_MASK, &temp);      \
    } while (0)

#else

typedef boolean preemption_key;

#define DISABLE_PREEMPTION(key) \
    do {                                             \
        (key) = _Thread_Executing->is_preemptible;   \
        _Thread_Executing->is_preemptible = 0;       \
    } while (0)

#define ENABLE_PREEMPTION(key) \
    do {                                             \
        _Thread_Executing->is_preemptible = (key);   \
        if (_Thread_Evaluate_mode())                 \
            _Thread_Dispatch();                      \
    } while (0)

#endif


/* The default maximum height of 32 allows for AVL trees having
   between 5,704,880 and 4,294,967,295 nodes, depending on order of
   insertion.  You may change this compile-time constant as you
   wish. */
#ifndef AVL_MAX_HEIGHT
#define AVL_MAX_HEIGHT  32
#endif

/*
 * avl_search --
 *     Searches for the node with specified dev/block.
 *
 * PARAMETERS:
 *     root - pointer to the root node of the AVL-Tree.
 *     dev, block - search key
 *
 * RETURNS:
 *     NULL if node with specified dev/block not found
 *     non-NULL - pointer to the node with specified dev/block
 */
static bdbuf_buffer *
avl_search(bdbuf_buffer **root, dev_t dev, blkdev_bnum block)
{
    bdbuf_buffer *p = *root;

    while ((p != NULL) && ((p->dev != dev) || (p->block != block)))
    {
        if ((p->dev < dev) || ((p->dev == dev) && (p->block < block)))
        {
            p = p->avl.right;
        }
        else
        {
            p = p->avl.left;
        }
    }

    return p;
}


/* avl_search_for_sync --
 *     Search in AVL tree for first modified buffer belongs to specified
 *     disk device.
 *
 * PARAMETERS:
 *     root - pointer to tree root
 *     dd - disk device descriptor
 *
 * RETURNS:
 *     Block buffer, or NULL if no modified blocks on specified device
 *     exists.
 */
static bdbuf_buffer *
avl_search_for_sync(bdbuf_buffer **root, disk_device *dd)
{
    dev_t dev = dd->phys_dev->dev;
    blkdev_bnum block_start = dd->start;
    blkdev_bnum block_end = dd->start + dd->size - 1;

    bdbuf_buffer *buf_stack[AVL_MAX_HEIGHT];
    bdbuf_buffer **buf_prev = buf_stack;
    bdbuf_buffer *p = *root;

    while (p != NULL)
    {
        if ((p->dev < dev) || ((p->dev == dev) && (p->block < block_start)))
        {
            p = p->avl.right;
        }
        else if ((p->dev > dev) || ((p->dev == dev) && (p->block > block_end)))
        {
            p = p->avl.left;
        }
        else if (p->modified)
        {
            return p;
        }
        else
        {
            if (p->avl.right != NULL)
            {
                *buf_prev++ = p->avl.right;
            }
            p = p->avl.left;
        }

        if ((p == NULL) && (buf_prev > buf_stack))
        {
            p = *--buf_prev;
        }
    }

    return p;
}


/*
 * avl_insert --
 *     Inserts the specified node to the AVl-Tree.
 *
 * PARAMETERS:
 *     root - Pointer to pointer to the root node
 *     node - Pointer to the node to add.
 *
 * RETURNS:
 *     0 - The node added successfully
 *    -1 - An error occured
 */
static int
avl_insert(bdbuf_buffer **root, bdbuf_buffer *node)
{
    dev_t dev = node->dev;
    blkdev_bnum block = node->block;

    bdbuf_buffer *p = *root;
    bdbuf_buffer *q, *p1, *p2;
    bdbuf_buffer *buf_stack[AVL_MAX_HEIGHT];
    bdbuf_buffer **buf_prev = buf_stack;

    boolean modified = FALSE;

    if (p == NULL)
    {
        *root = node;
        node->avl.left = NULL;
        node->avl.right = NULL;
        node->avl.bal = 0;
        return 0;
    }

    while (p != NULL)
    {
        *buf_prev++ = p;

        if ((p->dev < dev) || ((p->dev == dev) && (p->block < block)))
        {
            p->avl.cache = 1;
            q = p->avl.right;
            if (q == NULL)
            {
                q = node;
                p->avl.right = q = node;
                break;
            }
        }
        else if ((p->dev != dev) || (p->block != block))
        {
            p->avl.cache = -1;
            q = p->avl.left;
            if (q == NULL)
            {
                q = node;
                p->avl.left = q;
                break;
            }
        }
        else
        {
            return -1;
        }

        p = q;
    }

    q->avl.left = q->avl.right = NULL;
    q->avl.bal = 0;
    modified = TRUE;
    buf_prev--;

    while (modified)
    {
        if (p->avl.cache == -1)
        {
            switch (p->avl.bal)
            {
                case 1:
                    p->avl.bal = 0;
                    modified = FALSE;
                    break;

                case 0:
                    p->avl.bal = -1;
                    break;

                case -1:
                    p1 = p->avl.left;
                    if (p1->avl.bal == -1) /* simple LL-turn */
                    {
                        p->avl.left = p1->avl.right;
                        p1->avl.right = p;
                        p->avl.bal = 0;
                        p = p1;
                    }
                    else /* double LR-turn */
                    {
                        p2 = p1->avl.right;
                        p1->avl.right = p2->avl.left;
                        p2->avl.left = p1;
                        p->avl.left = p2->avl.right;
                        p2->avl.right = p;
                        if (p2->avl.bal == -1) p->avl.bal = +1; else p->avl.bal = 0;
                        if (p2->avl.bal == +1) p1->avl.bal = -1; else p1->avl.bal = 0;
                        p = p2;
                    }
                    p->avl.bal = 0;
                    modified = FALSE;
                    break;

                default:
                    break;
            }
        }
        else
        {
            switch (p->avl.bal)
            {
                case -1:
                    p->avl.bal = 0;
                    modified = FALSE;
                    break;

                case 0:
                    p->avl.bal = 1;
                    break;

                case 1:
                    p1 = p->avl.right;
                    if (p1->avl.bal == 1) /* simple RR-turn */
                    {
                        p->avl.right = p1->avl.left;
                        p1->avl.left = p;
                        p->avl.bal = 0;
                        p = p1;
                    }
                    else /* double RL-turn */
                    {
                        p2 = p1->avl.left;
                        p1->avl.left = p2->avl.right;
                        p2->avl.right = p1;
                        p->avl.right = p2->avl.left;
                        p2->avl.left = p;
                        if (p2->avl.bal == +1) p->avl.bal = -1; else p->avl.bal = 0;
                        if (p2->avl.bal == -1) p1->avl.bal = +1; else p1->avl.bal = 0;
                        p = p2;
                    }
                    p->avl.bal = 0;
                    modified = FALSE;
                    break;

                default:
                    break;
            }
        }
        q = p;
        if (buf_prev > buf_stack)
        {
            p = *--buf_prev;

            if (p->avl.cache == -1)
            {
                p->avl.left = q;
            }
            else
            {
                p->avl.right = q;
            }
        }
        else
        {
            *root = p;
            break;
        }
    };

    return 0;
}


/* avl_remove --
 *     removes the node from the tree.
 *
 * PARAMETERS:
 *     root_addr - Pointer to pointer to the root node
 *     node      - Pointer to the node to remove
 *
 * RETURNS:
 *     0 - Item removed
 *    -1 - No such item found
 */
static int
avl_remove(bdbuf_buffer **root, const bdbuf_buffer *node)
{
    dev_t dev = node->dev;
    blkdev_bnum block = node->block;

    bdbuf_buffer *p = *root;
    bdbuf_buffer *q, *r, *s, *p1, *p2;
    bdbuf_buffer *buf_stack[AVL_MAX_HEIGHT];
    bdbuf_buffer **buf_prev = buf_stack;

    boolean modified = FALSE;

    memset(buf_stack, 0, sizeof(buf_stack));

    while (p != NULL)
    {
        *buf_prev++ = p;

        if ((p->dev < dev) || ((p->dev == dev) && (p->block < block)))
        {
            p->avl.cache = 1;
            p = p->avl.right;
        }
        else if ((p->dev != dev) || (p->block != block))
        {
            p->avl.cache = -1;
            p = p->avl.left;
        }
        else
        {
            /* node found */
            break;
        }
    }

    if (p == NULL)
    {
        /* there is no such node */
        return -1;
    }

    q = p;

    buf_prev--;
    if (buf_prev > buf_stack)
    {
        p = *(buf_prev - 1);
    }
    else
    {
        p = NULL;
    }

    /* at this moment q - is a node to delete, p is q's parent */
    if (q->avl.right == NULL)
    {
        r = q->avl.left;
        if (r != NULL)
        {
            r->avl.bal = 0;
        }
        q = r;
    }
    else
    {
        bdbuf_buffer **t;

        r = q->avl.right;

        if (r->avl.left == NULL)
        {
            r->avl.left = q->avl.left;
            r->avl.bal = q->avl.bal;
            r->avl.cache = 1;
            *buf_prev++ = q = r;
        }
        else
        {
            t = buf_prev++;
            s = r;

            while (s->avl.left != NULL)
            {
                *buf_prev++ = r = s;
                s = r->avl.left;
                r->avl.cache = -1;
            }

            s->avl.left = q->avl.left;
            r->avl.left = s->avl.right;
            s->avl.right = q->avl.right;
            s->avl.bal = q->avl.bal;
            s->avl.cache = 1;

            *t = q = s;
        }
    }

    if (p != NULL)
    {
        if (p->avl.cache == -1)
        {
            p->avl.left = q;
        }
        else
        {
            p->avl.right = q;
        }
    }
    else
    {
        *root = q;
    }

    modified = TRUE;

    while (modified)
    {
        if (buf_prev > buf_stack)
        {
            p = *--buf_prev;
        }
        else
        {
            break;
        }

        if (p->avl.cache == -1)
        {
            /* rebalance left branch */
            switch (p->avl.bal)
            {
                case -1:
                    p->avl.bal = 0;
                    break;
                case  0:
                    p->avl.bal = 1;
                    modified = FALSE;
                    break;

                case +1:
                    p1 = p->avl.right;

                    if (p1->avl.bal >= 0) /* simple RR-turn */
                    {
                        p->avl.right = p1->avl.left;
                        p1->avl.left = p;

                        if (p1->avl.bal == 0)
                        {
                            p1->avl.bal = -1;
                            modified = FALSE;
                        }
                        else
                        {
                            p->avl.bal = 0;
                            p1->avl.bal = 0;
                        }
                        p = p1;
                    }
                    else /* double RL-turn */
                    {
                        p2 = p1->avl.left;

                        p1->avl.left = p2->avl.right;
                        p2->avl.right = p1;
                        p->avl.right = p2->avl.left;
                        p2->avl.left = p;

                        if (p2->avl.bal == +1) p->avl.bal = -1; else p->avl.bal = 0;
                        if (p2->avl.bal == -1) p1->avl.bal = 1; else p1->avl.bal = 0;

                        p = p2;
                        p2->avl.bal = 0;
                    }
                    break;

                default:
                    break;
            }
        }
        else
        {
            /* rebalance right branch */
            switch (p->avl.bal)
            {
                case +1:
                    p->avl.bal = 0;
                    break;

                case  0:
                    p->avl.bal = -1;
                    modified = FALSE;
                    break;

                case -1:
                    p1 = p->avl.left;

                    if (p1->avl.bal <= 0) /* simple LL-turn */
                    {
                        p->avl.left = p1->avl.right;
                        p1->avl.right = p;
                        if (p1->avl.bal == 0)
                        {
                            p1->avl.bal = 1;
                            modified = FALSE;
                        }
                        else
                        {
                            p->avl.bal = 0;
                            p1->avl.bal = 0;
                        }
                        p = p1;
                    }
                    else /* double LR-turn */
                    {
                        p2 = p1->avl.right;

                        p1->avl.right = p2->avl.left;
                        p2->avl.left = p1;
                        p->avl.left = p2->avl.right;
                        p2->avl.right = p;

                        if (p2->avl.bal == -1) p->avl.bal = 1; else p->avl.bal = 0;
                        if (p2->avl.bal == +1) p1->avl.bal = -1; else p1->avl.bal = 0;

                        p = p2;
                        p2->avl.bal = 0;
                    }
                    break;

                default:
                    break;
            }
        }

        if (buf_prev > buf_stack)
        {
            q = *(buf_prev - 1);

            if (q->avl.cache == -1)
            {
                q->avl.left = p;
            }
            else
            {
                q->avl.right = p;
            }
        }
        else
        {
            *root = p;
            break;
        }

    }

    return 0;
}

/* bdbuf_initialize_pool --
 *      Initialize single buffer pool.
 *
 * PARAMETERS:
 *     config - buffer pool configuration
 *     pool   - pool number
 *
 * RETURNS:
 *     RTEMS_SUCCESSFUL, if buffer pool initialized successfully, or error
 *     code if error occured.
 */
static rtems_status_code
bdbuf_initialize_pool(rtems_bdbuf_config *config, int pool)
{
    bdbuf_pool *p = bd_ctx.pool + pool;
    unsigned char *bufs;
    bdbuf_buffer *b;
    rtems_status_code rc;
    int i;

    p->blksize = config->size;
    p->nblks = config->num;
    p->tree = NULL;

    Chain_Initialize_empty(&p->free);
    Chain_Initialize_empty(&p->lru);

    /* Allocate memory for buffer descriptors */
    p->bdbufs = calloc(config->num, sizeof(bdbuf_buffer));
    if (p->bdbufs == NULL)
    {
        return RTEMS_NO_MEMORY;
    }

    /* Allocate memory for buffers if required */
    if (config->mem_area == NULL)
    {
        bufs = p->mallocd_bufs = malloc(config->num * config->size);
        if (bufs == NULL)
        {
            free(p->bdbufs);
            return RTEMS_NO_MEMORY;
        }
    }
    else
    {
        bufs = config->mem_area;
        p->mallocd_bufs = NULL;
    }

    for (i = 0, b = p->bdbufs; i < p->nblks; i++, b++, bufs += p->blksize)
    {
        b->dev = -1; b->block = 0;
        b->buffer = bufs;
        b->actual = b->modified = b->in_progress = FALSE;
        b->use_count = 0;
        b->pool = pool;
        _Chain_Append(&p->free, &b->link);
    }

    rc = rtems_semaphore_create(
        rtems_build_name('B', 'U', 'F', 'G'),
        p->nblks,
        RTEMS_FIFO | RTEMS_COUNTING_SEMAPHORE | RTEMS_NO_INHERIT_PRIORITY |
        RTEMS_NO_PRIORITY_CEILING | RTEMS_LOCAL,
        0,
        &p->bufget_sema);

    if (rc != RTEMS_SUCCESSFUL)
    {
        free(p->bdbufs);
        free(p->mallocd_bufs);
        return rc;
    }

    return RTEMS_SUCCESSFUL;
}

/* bdbuf_release_pool --
 *     Free resources allocated for buffer pool with specified number.
 *
 * PARAMETERS:
 *     pool - buffer pool number
 *
 * RETURNS:
 *     RTEMS_SUCCESSFUL
 */
static rtems_status_code
bdbuf_release_pool(rtems_bdpool_id pool)
{
    bdbuf_pool *p = bd_ctx.pool + pool;
    rtems_semaphore_delete(p->bufget_sema);
    free(p->bdbufs);
    free(p->mallocd_bufs);
    return RTEMS_SUCCESSFUL;
}

/* rtems_bdbuf_init --
 *     Prepare buffering layer to work - initialize buffer descritors
 *     and (if it is neccessary)buffers. Buffers will be allocated accoriding
 *     to the configuration table, each entry describes kind of block and
 *     amount requested. After initialization all blocks is placed into
 *     free elements lists.
 *
 * PARAMETERS:
 *     conf_table - pointer to the buffers configuration table
 *     size       - number of entries in configuration table
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 */
rtems_status_code
rtems_bdbuf_init(rtems_bdbuf_config *conf_table, int size)
{
    rtems_bdpool_id i;
    rtems_status_code rc;

    if (size <= 0)
        return RTEMS_INVALID_SIZE;

    bd_ctx.npools = size;

    /*
     * Allocate memory for buffer pool descriptors
     */
    bd_ctx.pool = calloc(size, sizeof(bdbuf_pool));
    if (bd_ctx.pool == NULL)
    {
        return RTEMS_NO_MEMORY;
    }

    Chain_Initialize_empty(&bd_ctx.mod);

    /* Initialize buffer pools and roll out if something failed */
    for (i = 0; i < size; i++)
    {
        rc = bdbuf_initialize_pool(conf_table + i, i);
        if (rc != RTEMS_SUCCESSFUL)
        {
             rtems_bdpool_id j;
             for (j = 0; j < i - 1; j++)
             {
                 bdbuf_release_pool(j);
             }
             return rc;
        }
    }

    /* Create buffer flush semaphore */
    rc = rtems_semaphore_create(
        rtems_build_name('B', 'F', 'L', 'U'), 0,
        RTEMS_FIFO | RTEMS_COUNTING_SEMAPHORE | RTEMS_NO_INHERIT_PRIORITY |
        RTEMS_NO_PRIORITY_CEILING | RTEMS_LOCAL, 0,
        &bd_ctx.flush_sema);
    if (rc != RTEMS_SUCCESSFUL)
    {
        for (i = 0; i < size; i++)
            bdbuf_release_pool(i);
        free(bd_ctx.pool);
        return rc;
    }

    /* Create and start swapout task */
    rc = rtems_task_create(
        rtems_build_name('B', 'S', 'W', 'P'),
        SWAPOUT_PRIORITY,
        SWAPOUT_STACK_SIZE,
        RTEMS_DEFAULT_MODES | RTEMS_NO_PREEMPT,
        RTEMS_DEFAULT_ATTRIBUTES,
        &bd_ctx.swapout_task);
    if (rc != RTEMS_SUCCESSFUL)
    {
        rtems_semaphore_delete(bd_ctx.flush_sema);
        for (i = 0; i < size; i++)
            bdbuf_release_pool(i);
        free(bd_ctx.pool);
        return rc;
    }

    rc = rtems_task_start(bd_ctx.swapout_task, bdbuf_swapout_task, 0);
    if (rc != RTEMS_SUCCESSFUL)
    {
        rtems_task_delete(bd_ctx.swapout_task);
        rtems_semaphore_delete(bd_ctx.flush_sema);
        for (i = 0; i < size; i++)
            bdbuf_release_pool(i);
        free(bd_ctx.pool);
        return rc;
    }

    return RTEMS_SUCCESSFUL;
}

/* find_or_assign_buffer --
 *     Looks for buffer already assigned for this dev/block. If one is found
 *     obtain block buffer. If specified block already cached (i.e. there's
 *     block in the _modified_, or _recently_used_), return address
 *     of appropriate buffer descriptor and increment reference counter to 1.
 *     If block is not cached, allocate new buffer and return it. Data
 *     shouldn't be read to the buffer from media; buffer contains arbitrary
 *     data. This primitive may be blocked if there are no free buffer
 *     descriptors available and there are no unused non-modified (or
 *     synchronized with media) buffers available.
 *
 * PARAMETERS:
 *     device - device number (constructed of major and minor device number
 *     block  - linear media block number
 *     ret_buf - address of the variable to store address of found descriptor
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFEECTS:
 *     bufget_sema may be obtained by this primitive
 *
 * NOTE:
 *     It is assumed that primitive invoked when thread preemption is disabled.
 */
static rtems_status_code
find_or_assign_buffer(disk_device *dd,
                      blkdev_bnum block,
                      bdbuf_buffer **ret_buf)
{
    bdbuf_buffer *bd_buf;
    bdbuf_pool   *bd_pool;
    rtems_status_code rc;
    dev_t         device;
    ISR_Level     level;

    int blksize;

    device = dd->dev;
    bd_pool = bd_ctx.pool + dd->pool;
    blksize = dd->block_size;

again:
    /* Looking for buffer descriptor used for this dev/block. */
    bd_buf = avl_search(&bd_pool->tree, device, block);

    if (bd_buf == NULL)
    {
        /* Try to obtain semaphore without waiting first. It is the most
           frequent case when reasonable number of buffers configured. If
           it is failed, obtain semaphore blocking on it. In this case
           it should be checked that appropriate buffer hasn't been loaded
           by another thread, because this thread is preempted */
        rc = rtems_semaphore_obtain(bd_pool->bufget_sema, RTEMS_NO_WAIT, 0);
        if (rc == RTEMS_UNSATISFIED)
        {
            rc = rtems_semaphore_obtain(bd_pool->bufget_sema,
                                        RTEMS_WAIT, RTEMS_NO_TIMEOUT);
            bd_buf = avl_search(&bd_pool->tree, device, block);
            if (bd_buf != NULL)
                rtems_semaphore_release(bd_pool->bufget_sema);
        }
    }

    if (bd_buf == NULL)
    {
        /* Assign new buffer descriptor */
        if (_Chain_Is_empty(&bd_pool->free))
        {
            bd_buf = (bdbuf_buffer *)Chain_Get(&bd_pool->lru);
            if (bd_buf != NULL)
            {
                int avl_result;
                avl_result = avl_remove(&bd_pool->tree, bd_buf);
                if (avl_result != 0)
                {
                    rtems_fatal_error_occurred(BLKDEV_FATAL_BDBUF_CONSISTENCY);
                    return RTEMS_INTERNAL_ERROR;
                }
            }
        }
        else
        {
            bd_buf = (bdbuf_buffer *)Chain_Get(&(bd_pool->free));
        }

        if (bd_buf == NULL)
        {
            goto again;
        }
        else
        {
            bd_buf->dev = device;
            bd_buf->block = block;
#ifdef AVL_GPL
            bd_buf->avl.link[0] = NULL;
            bd_buf->avl.link[1] = NULL;
#else
            bd_buf->avl.left = NULL;
            bd_buf->avl.right = NULL;
#endif
            bd_buf->use_count = 1;
            bd_buf->modified = bd_buf->actual = bd_buf->in_progress = FALSE;
            bd_buf->status = RTEMS_SUCCESSFUL;

            if (avl_insert(&bd_pool->tree, bd_buf) != 0)
            {
                rtems_fatal_error_occurred(BLKDEV_FATAL_BDBUF_CONSISTENCY);
                return RTEMS_INTERNAL_ERROR;
            }

            *ret_buf = bd_buf;

            return RTEMS_SUCCESSFUL;
        }
    }
    else
    {
        /* Buffer descriptor already assigned for this dev/block */
        if (bd_buf->use_count == 0)
        {
            /* If we are removing from lru list, obtain the bufget_sema
             * first. If we are removing from mod list, obtain flush sema.
             * It should be obtained without blocking because we know
             * that our buffer descriptor is in the list. */
            if (bd_buf->modified)
            {
                rc = rtems_semaphore_obtain(bd_ctx.flush_sema,
                                            RTEMS_NO_WAIT, 0);
            }
            else
            {
                rc = rtems_semaphore_obtain(bd_pool->bufget_sema,
                                            RTEMS_NO_WAIT, 0);
            }
            /* It is possible that we couldn't obtain flush or bufget sema
             * although buffer in the appropriate chain is available:
             * semaphore may be released to swapout task, but this task
             * actually did not start to process it. */
            if (rc == RTEMS_UNSATISFIED)
                rc = RTEMS_SUCCESSFUL;
            if (rc != RTEMS_SUCCESSFUL)
            {
                rtems_fatal_error_occurred(BLKDEV_FATAL_BDBUF_CONSISTENCY);
                return RTEMS_INTERNAL_ERROR;
            }

            /* Buffer descriptor is linked to the lru or mod chain. Remove
               it from there. */
            Chain_Extract(&bd_buf->link);
        }
        bd_buf->use_count++;
        while (bd_buf->in_progress != 0)
        {
            rtems_interrupt_disable(level);
            _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE,
                              WATCHDOG_NO_TIMEOUT, level);
        }

        *ret_buf = bd_buf;
        return RTEMS_SUCCESSFUL;
    }
}

/* rtems_bdbuf_get --
 *     Obtain block buffer. If specified block already cached (i.e. there's
 *     block in the _modified_, or _recently_used_), return address
 *     of appropriate buffer descriptor and increment reference counter to 1.
 *     If block is not cached, allocate new buffer and return it. Data
 *     shouldn't be read to the buffer from media; buffer may contains
 *     arbitrary data. This primitive may be blocked if there are no free
 *     buffer descriptors available and there are no unused non-modified
 *     (or synchronized with media) buffers available.
 *
 * PARAMETERS:
 *     device - device number (constructed of major and minor device number)
 *     block  - linear media block number
 *     bd     - address of variable to store pointer to the buffer descriptor
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFECTS:
 *     bufget_sema semaphore obtained by this primitive.
 */
rtems_status_code
rtems_bdbuf_get(dev_t device, blkdev_bnum block, bdbuf_buffer **bd)
{
    rtems_status_code rc;
    disk_device *dd;
    disk_device *pdd;
    preemption_key key;

    /*
     * Convert logical dev/block to physical one
     */
    dd = rtems_disk_lookup(device);
    if (dd == NULL)
        return RTEMS_INVALID_ID;

    if (block >= dd->size)
    {
        rtems_disk_release(dd);
        return RTEMS_INVALID_NUMBER;
    }

    pdd = dd->phys_dev;
    block += dd->start;
    rtems_disk_release(dd);

    DISABLE_PREEMPTION(key);
    rc = find_or_assign_buffer(pdd, block, bd);
    ENABLE_PREEMPTION(key);

    if (rc != RTEMS_SUCCESSFUL)
        return rc;

    return RTEMS_SUCCESSFUL;
}

/* bdbuf_initialize_transfer_sema --
 *     Initialize transfer_sema mutex semaphore associated with buffer
 *     descriptor.
 */
static inline void
bdbuf_initialize_transfer_sema(bdbuf_buffer *bd_buf)
{
    CORE_mutex_Attributes mutex_attr;
    mutex_attr.lock_nesting_behavior = CORE_MUTEX_NESTING_BLOCKS;
    mutex_attr.only_owner_release = FALSE;
    mutex_attr.discipline = CORE_MUTEX_DISCIPLINES_FIFO;
    mutex_attr.priority_ceiling = 0;

    _CORE_mutex_Initialize(&bd_buf->transfer_sema,
                           &mutex_attr, CORE_MUTEX_LOCKED);
}

/* bdbuf_write_transfer_done --
 *     Callout function. Invoked by block device driver when data transfer
 *     to device (write) is completed. This function may be invoked from
 *     interrupt handler.
 *
 * PARAMETERS:
 *     arg    - arbitrary argument specified in block device request
 *              structure (in this case - pointer to the appropriate
 *              bdbuf_buffer buffer descriptor structure).
 *     status - I/O completion status
 *     error  - errno error code if status != RTEMS_SUCCESSFUL
 *
 * RETURNS:
 *     none
 */
static void
bdbuf_write_transfer_done(void *arg, rtems_status_code status, int error)
{
    int i;
    write_tfer_done_arg_t *wtd_arg = arg;
    blkdev_request *req = wtd_arg->req;
    bdbuf_buffer **bd_buf_write_store = wtd_arg->write_store;
    bdbuf_buffer *bd_buf;
    for (i = 0;i < req->count;i++) {
      bd_buf = bd_buf_write_store[i];
      bd_buf->status = status;
      bd_buf->error = RTEMS_IO_ERROR;
      
      bd_buf->in_progress = FALSE;
      _CORE_mutex_Surrender(&bd_buf->transfer_sema, 0, NULL);
      _CORE_mutex_Flush(&bd_buf->transfer_sema, NULL, 
                        CORE_MUTEX_STATUS_SUCCESSFUL);
    }
}

/* bdbuf_read_transfer_done --
 *     Callout function. Invoked by block device driver when data transfer
 *     from device (read) is completed. This function may be invoked from
 *     interrupt handler.
 *
 * PARAMETERS:
 *     arg    - arbitrary argument specified in block device request
 *              structure (in this case - pointer to the appropriate
 *              bdbuf_buffer buffer descriptor structure).
 *     status - I/O completion status
 *     error  - errno error code if status != RTEMS_SUCCESSFUL
 *
 * RETURNS:
 *     none
 */
static void
bdbuf_read_transfer_done(void *arg, rtems_status_code status, int error)
{
#if defined(READ_MULTIPLE)  
  
  read_ahead_bd_buf_group *bd_buf_group = arg;
  bdbuf_buffer *bd_buf;
  int i;
  for (i = 0;i < bd_buf_group->cnt;i++) {
    bd_buf = bd_buf_group->bd_bufs[i];

    bd_buf->status = status;
    bd_buf->error = RTEMS_IO_ERROR;
    _CORE_mutex_Surrender(&bd_buf->transfer_sema, 0, NULL);
    _CORE_mutex_Flush(&bd_buf->transfer_sema, NULL, 
                      CORE_MUTEX_STATUS_SUCCESSFUL);
  }
#else
    bdbuf_buffer *bd_buf = arg;
    bd_buf->status = status;
    bd_buf->error = RTEMS_IO_ERROR;
    _CORE_mutex_Surrender(&bd_buf->transfer_sema, 0, NULL);
    _CORE_mutex_Flush(&bd_buf->transfer_sema, NULL,
                      CORE_MUTEX_STATUS_SUCCESSFUL);
#endif
}

/* rtems_bdbuf_read --
 *     (Similar to the rtems_bdbuf_get, except reading data from media)
 *     Obtain block buffer. If specified block already cached, return address
 *     of appropriate buffer and increment reference counter to 1. If block is
 *     not cached, allocate new buffer and read data to it from the media.
 *     This primitive may be blocked on waiting until data to be read from
 *     media, if there are no free buffer descriptors available and there are
 *     no unused non-modified (or synchronized with media) buffers available.
 *
 * PARAMETERS:
 *     device - device number (consists of major and minor device number)
 *     block  - linear media block number
 *     bd     - address of variable to store pointer to the buffer descriptor
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFECTS:
 *     bufget_sema and transfer_sema semaphores obtained by this primitive.
 */
#if !defined(READ_MULTIPLE)
rtems_status_code
rtems_bdbuf_read(dev_t device,
                 blkdev_bnum block,
                 bdbuf_buffer **bd)
{
    preemption_key key;
    ISR_Level level;

    bdbuf_buffer *bd_buf;
    rtems_status_code rc;
    int result;
    disk_device *dd;
    disk_device *pdd;
    blkdev_request1 req;

    dd = rtems_disk_lookup(device);
    if (dd == NULL)
        return RTEMS_INVALID_ID;

    if (block >= dd->size)
    {
        rtems_disk_release(dd);
        return RTEMS_INVALID_NUMBER;
    }

    pdd = dd->phys_dev;
    block += dd->start;

    DISABLE_PREEMPTION(key);
    rc = find_or_assign_buffer(pdd, block, &bd_buf);

    if (rc != RTEMS_SUCCESSFUL)
    {
        ENABLE_PREEMPTION(key);
        rtems_disk_release(dd);
        return rc;
    }

    if (!bd_buf->actual)
    {
        bd_buf->in_progress = 1;

        req.req.req = BLKDEV_REQ_READ;
        req.req.req_done = bdbuf_read_transfer_done;
        req.req.done_arg = bd_buf;
        req.req.start = block;
        req.req.count = 1;
        req.req.bufnum = 1;
        req.req.bufs[0].length = dd->block_size;
        req.req.bufs[0].buffer = bd_buf->buffer;

        bdbuf_initialize_transfer_sema(bd_buf);
        result = dd->ioctl(pdd->dev, BLKIO_REQUEST, &req);
        if (result == -1)
        {
            bd_buf->status = RTEMS_IO_ERROR;
            bd_buf->error = errno;
            bd_buf->actual = FALSE;
        }
        else
        {
            rtems_interrupt_disable(level);
            _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE,
                              WATCHDOG_NO_TIMEOUT, level);
            bd_buf->actual = TRUE;
        }
        bd_buf->in_progress = FALSE;
    }
    rtems_disk_release(dd);

    ENABLE_PREEMPTION(key);

    *bd = bd_buf;
           
    return RTEMS_SUCCESSFUL;
}
#else /* READ_MULTIPLE */
rtems_status_code
rtems_bdbuf_read(dev_t device, 
                 blkdev_bnum block,
                 bdbuf_buffer **bd)
{
    preemption_key key;
    ISR_Level level;

    bdbuf_buffer *bd_buf,*first_bd_buf;
    rtems_status_code rc;
    int result;
    disk_device *dd;
    disk_device *pdd;
    blkdev_request_read_ahead req;
    read_ahead_bd_buf_group bd_buf_group;
    boolean find_more_buffers;
    int i;

    dd = rtems_disk_lookup(device);
    if (dd == NULL)
        return RTEMS_INVALID_ID;
     
    if (block >= dd->size)
    {
        rtems_disk_release(dd);
        return RTEMS_INVALID_NUMBER;
    }
    
    pdd = dd->phys_dev;
    block += dd->start;

    DISABLE_PREEMPTION(key);
    rc = find_or_assign_buffer(pdd, block, &first_bd_buf);

    if (rc != RTEMS_SUCCESSFUL)
    {
        ENABLE_PREEMPTION(key);
        rtems_disk_release(dd);
        return rc;
    }
    if (!first_bd_buf->actual)
    {

        bd_buf_group.bd_bufs[0] = first_bd_buf;
        bd_buf_group.cnt = 1;

        first_bd_buf->in_progress = TRUE;

        req.req.req = BLKDEV_REQ_READ;
        req.req.req_done = bdbuf_read_transfer_done;
        req.req.done_arg = &bd_buf_group;
        req.req.start = block;
        req.req.count = 1;
        req.req.bufnum = 1;
        req.req.bufs[0].length = dd->block_size;
        req.req.bufs[0].buffer = first_bd_buf->buffer;
        
        bdbuf_initialize_transfer_sema(first_bd_buf);
        /*
         * FIXME: check for following blocks to be:
         *   - still in range of partition size
         *   - not yet assigned
         *   - buffer available
         * allocate for read call, if possible
         */
        find_more_buffers = TRUE;
        while (find_more_buffers) {       
          block++;
          /*
           * still bd_buf_group entries free and 
           * still in range of this disk? 
           */
          if ((bd_buf_group.cnt >= READ_AHEAD_MAX_BLK_CNT) ||
              (block >= dd->size)) {
            find_more_buffers = FALSE;
          }
          if (find_more_buffers) {
            rc = find_or_assign_buffer(pdd, block, &bd_buf);
            if (rc != RTEMS_SUCCESSFUL) {
              find_more_buffers = FALSE;
            }      
            else if (bd_buf->actual) {
              find_more_buffers = FALSE;
          bdbuf_release(bd_buf);
            }      
          }
          if (find_more_buffers) {
            bdbuf_initialize_transfer_sema(bd_buf);
            bd_buf->in_progress = TRUE;

            req.req.bufs[req.req.count].length = dd->block_size;
            req.req.bufs[req.req.count].buffer = bd_buf->buffer;
            req.req.count++;
            req.req.bufnum++;
            bd_buf_group.bd_bufs[bd_buf_group.cnt] = bd_buf;
            bd_buf_group.cnt++;
          }
        }           

        /* do the actual read call here 
         */
        result = dd->ioctl(pdd->dev, BLKIO_REQUEST, &req);

        /*
         * cleanup: 
         * wait, until all bd_bufs are processed
         * set status in all bd_bufs
         */
        for (i = 0;i < bd_buf_group.cnt;i++) {
          bd_buf = bd_buf_group.bd_bufs[i];
          if (result == -1)
            {
              bd_buf->status = RTEMS_IO_ERROR;
              bd_buf->error = errno;
              bd_buf->actual = FALSE;
            }
          else
            {
              rtems_interrupt_disable(level);
              _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE,
                                WATCHDOG_NO_TIMEOUT, level);
              bd_buf->actual = TRUE;
            }   
          bd_buf->in_progress = FALSE;
          /* release any pre-read buffers */
          if (i > 0) {
            bdbuf_release(bd_buf);
          }
        }
    }
    rtems_disk_release(dd);
    
    ENABLE_PREEMPTION(key);

    *bd = first_bd_buf;
           
    return RTEMS_SUCCESSFUL;
}
#endif /* READ_MULTIPLE */


/* bdbuf_release --
 *     Release buffer. Decrease buffer usage counter. If it is zero, further
 *     processing depends on modified attribute. If buffer was modified, it
 *     is inserted into mod chain and swapout task waken up. If buffer was
 *     not modified, it is returned to the end of lru chain making it available
 *     for further use.
 *
 * PARAMETERS:
 *     bd_buf - pointer to the released buffer descriptor.
 *
 * RETURNS:
 *     RTEMS_SUCCESSFUL if buffer released successfully, or error code if
 *     error occured.
 *
 * NOTE:
 *     This is internal function. It is assumed that task made non-preemptive
 *     before its invocation.
 */
static rtems_status_code
bdbuf_release(bdbuf_buffer *bd_buf)
{
    bdbuf_pool *bd_pool;
    rtems_status_code rc = RTEMS_SUCCESSFUL;

    if (bd_buf->use_count <= 0)
        return RTEMS_INTERNAL_ERROR;

    bd_pool = bd_ctx.pool + bd_buf->pool;

    bd_buf->use_count--;

    if (bd_buf->use_count == 0)
    {
        if (bd_buf->modified)
        {

            /* Buffer was modified. Insert buffer to the modified buffers
             * list and initiate flushing. */
            Chain_Append(&bd_ctx.mod, &bd_buf->link);

            /* Release the flush_sema */
            rc = rtems_semaphore_release(bd_ctx.flush_sema);
        }
        else
        {
            /* Buffer was not modified. Add this descriptor to the
             * end of lru chain and make it available for reuse. */
            Chain_Append(&bd_pool->lru, &bd_buf->link);
            rc = rtems_semaphore_release(bd_pool->bufget_sema);
        }
    }
    return rc;
}


/* rtems_bdbuf_release --
 *     Release buffer allocated before. This primitive decrease the
 *     usage counter. If it is zero, further destiny of buffer depends on
 *     'modified' status. If buffer was modified, it is placed to the end of
 *     mod list and flush task waken up. If buffer was not modified,
 *     it is placed to the end of lru list, and bufget_sema released, allowing
 *     to reuse this buffer.
 *
 * PARAMETERS:
 *     bd_buf - pointer to the bdbuf_buffer structure previously obtained using
 *              get/read primitive.
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFECTS:
 *     flush_sema and bufget_sema semaphores may be released by this primitive.
 */
rtems_status_code
rtems_bdbuf_release(bdbuf_buffer *bd_buf)
{
    preemption_key key;
    rtems_status_code rc = RTEMS_SUCCESSFUL;

    if (bd_buf == NULL)
        return RTEMS_INVALID_ADDRESS;

    DISABLE_PREEMPTION(key);

    rc = bdbuf_release(bd_buf);

    ENABLE_PREEMPTION(key);

    return rc;
}

/* rtems_bdbuf_release_modified --
 *     Release buffer allocated before, assuming that it is _modified_ by
 *     it's owner. This primitive decrease usage counter for buffer, mark
 *     buffer descriptor as modified. If usage counter is 0, insert it at
 *     end of mod chain and release flush_sema semaphore to activate the
 *     flush task.
 *
 * PARAMETERS:
 *     bd_buf - pointer to the bdbuf_buffer structure previously obtained using
 *              get/read primitive.
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFECTS:
 *     flush_sema semaphore may be released by this primitive.
 */
rtems_status_code
rtems_bdbuf_release_modified(bdbuf_buffer *bd_buf)
{
    preemption_key key;
    rtems_status_code rc = RTEMS_SUCCESSFUL;

    if (bd_buf == NULL)
        return RTEMS_INVALID_ADDRESS;

    DISABLE_PREEMPTION(key);

    if (!bd_buf->modified)
    {
        bdbuf_initialize_transfer_sema(bd_buf);
    }
    bd_buf->modified = TRUE;
    bd_buf->actual = TRUE;
    rc = bdbuf_release(bd_buf);

    ENABLE_PREEMPTION(key);

    return rc;
}

/* rtems_bdbuf_sync --
 *     Wait until specified buffer synchronized with disk. Invoked on exchanges
 *     critical for data consistency on the media. This primitive mark owned
 *     block as modified, decrease usage counter. If usage counter is 0,
 *     block inserted to the mod chain and flush_sema semaphore released.
 *     Finally, primitives blocked on transfer_sema semaphore.
 *
 * PARAMETERS:
 *     bd_buf - pointer to the bdbuf_buffer structure previously obtained using
 *              get/read primitive.
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 *
 * SIDE EFFECTS:
 *     Primitive may be blocked on transfer_sema semaphore.
 */
rtems_status_code
rtems_bdbuf_sync(bdbuf_buffer *bd_buf)
{
    preemption_key key;
    ISR_Level level;
    rtems_status_code rc = RTEMS_SUCCESSFUL;

    if (bd_buf == NULL)
        return RTEMS_INVALID_ADDRESS;

    DISABLE_PREEMPTION(key);

    if (!bd_buf->modified)
    {
        bdbuf_initialize_transfer_sema(bd_buf);
    }
    bd_buf->modified = TRUE;
    bd_buf->actual = TRUE;

    rc = bdbuf_release(bd_buf);

    if (rc == RTEMS_SUCCESSFUL)
    {
        rtems_interrupt_disable(level);
        _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE,
                          WATCHDOG_NO_TIMEOUT, level);
    }

    ENABLE_PREEMPTION(key);

    return rc;
}

/* rtems_bdbuf_syncdev --
 *     Synchronize with disk all buffers containing the blocks belonging to
 *     specified device.
 *
 * PARAMETERS:
 *     dev - block device number
 *
 * RETURNS:
 *     RTEMS status code (RTEMS_SUCCESSFUL if operation completed successfully
 *     or error code if error is occured)
 */
rtems_status_code
rtems_bdbuf_syncdev(dev_t dev)
{
    preemption_key key;
    ISR_Level level;

    bdbuf_buffer *bd_buf;
    disk_device *dd;
    bdbuf_pool  *pool;

    dd = rtems_disk_lookup(dev);
    if (dd == NULL)
        return RTEMS_INVALID_ID;

    pool = bd_ctx.pool + dd->pool;

    DISABLE_PREEMPTION(key);
    do {
        bd_buf = avl_search_for_sync(&pool->tree, dd);
        if (bd_buf != NULL /* && bd_buf->modified */)
        {
            rtems_interrupt_disable(level);
            _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE,
                              WATCHDOG_NO_TIMEOUT, level);
        }
    } while (bd_buf != NULL);
    ENABLE_PREEMPTION(key);
    return rtems_disk_release(dd);
}

/* bdbuf_swapout_task --
 *     Body of task which take care on flushing modified buffers to the
 *     disk.
 */
static rtems_task
bdbuf_swapout_task(rtems_task_argument unused)
{
    rtems_status_code rc;
    int result;
    int i;
    ISR_Level level;
    bdbuf_buffer *bd_buf;
    bdbuf_buffer *nxt_bd_buf;
    bdbuf_pool *bd_pool = NULL;
    disk_device *dd = NULL;
    struct {
      blkdev_request   req;
      blkdev_sg_buffer sg[SWAP_OUT_MAX_BLK_CNT];
    } req;
    write_tfer_done_arg_t write_tfer_done_arg;

    /*
     * provide info needed for write_transfer_done function
     */
    write_tfer_done_arg.req         = (blkdev_request *)&req.req;
    write_tfer_done_arg.write_store =  bd_buf_write_store;
    nxt_bd_buf = NULL;
    while (1)
    {
        req.req.req = BLKDEV_REQ_WRITE;
        req.req.req_done = bdbuf_write_transfer_done;
        req.req.done_arg = &write_tfer_done_arg;
        req.req.count = 0;
        req.req.bufnum = 0;
        bd_buf = NULL;
        do {
          /*
           * if a buffer was left over from last loop, then use this buffer
           * otherwise fetch new buffer from chain. 
           * Wait for buffer, if this is the first one of the request,
           * otherwise do not wait, if no buffer available
           */
          if (nxt_bd_buf == NULL) {
            rc = rtems_semaphore_obtain(bd_ctx.flush_sema, 
                                        (req.req.count == 0)
                                        ? RTEMS_WAIT
                                        : RTEMS_NO_WAIT, 
                                        0);
            if (rc == RTEMS_SUCCESSFUL) {
              nxt_bd_buf = (bdbuf_buffer *)Chain_Get(&bd_ctx.mod);
          if (nxt_bd_buf != NULL) {
                nxt_bd_buf->in_progress = TRUE;
            /* IMD try: clear "modified" bit early             */
            /* (and not in bdbuf_write_transfer_done) to allow */
            /* another modification during write processing    */
            nxt_bd_buf->modified    = FALSE;

                nxt_bd_buf->use_count++;
                  }
            }
            else if ((rc != RTEMS_UNSATISFIED) && 
                     (rc != RTEMS_TIMEOUT)) {
              rtems_fatal_error_occurred(BLKDEV_FATAL_BDBUF_SWAPOUT);
            }
          }
          /* 
           * It is possible that flush_sema semaphore will be released, but
           * buffer to be removed from mod chain before swapout task start
           * its processing. 
           */
          if ((req.req.count == 0)  || /* first bd_buf for this request */
              ((nxt_bd_buf        != NULL)            && 
               (nxt_bd_buf->dev   == bd_buf->dev)     && /* same device */
               (nxt_bd_buf->block == bd_buf->block+1))) {/* next block  */
            bd_buf     = nxt_bd_buf;
            nxt_bd_buf = NULL;
          }
          else {
            bd_buf = NULL;
          }
          /*
           * here we have three possible states:
           * bd_buf == NULL, nxt_bd_buf == NULL: no further block available
           * bd_buf != NULL, nxt_bd_buf == NULL: append bd_buf to request
           * bd_buf == NULL, nxt_bd_buf != NULL: nxt_bd_buf canot be appended 
           *                                     to current request, keep it
           *                                     for next main loop
           */
          if (bd_buf != NULL) {
            bd_pool = bd_ctx.pool + bd_buf->pool;
            if (req.req.count == 0) {
              /*
               * this is the first block, so use its address
               */
              dd = rtems_disk_lookup(bd_buf->dev);
              req.req.start = bd_buf->block + dd->start;
            }
            req.req.bufs[req.req.bufnum].length = dd->block_size;
            req.req.bufs[req.req.bufnum].buffer = bd_buf->buffer;
            /*
             * keep bd_buf for postprocessing
             */
            bd_buf_write_store[req.req.bufnum] = bd_buf;
            req.req.count++;
            req.req.bufnum++;
          }
        } while ((bd_buf != NULL) &&
                 (req.req.count < SWAP_OUT_MAX_BLK_CNT));
                 
        /* transfer_sema initialized when bd_buf inserted in the mod chain 
           first time */
        result = dd->ioctl(dd->phys_dev->dev, BLKIO_REQUEST, &req);

        rtems_disk_release(dd);
        
        for (i = 0;i < req.req.count;i++) {
          bd_buf = bd_buf_write_store[i];
          if (result == -1)
            {
              
              bd_buf->status = RTEMS_IO_ERROR;
              bd_buf->error = errno;
              /* Release tasks waiting on syncing this buffer */
              _CORE_mutex_Flush(&bd_buf->transfer_sema, NULL,
                                CORE_MUTEX_STATUS_SUCCESSFUL);
            }
          else
            {
              if (bd_buf->in_progress)
                {
                  rtems_interrupt_disable(level);
                  _CORE_mutex_Seize(&bd_buf->transfer_sema, 0, TRUE, 0, level);
                }
            }
          bd_buf->use_count--;

          /* Another task have chance to use this buffer, or even
           * modify it. If buffer is not in use, insert it in appropriate chain
           * and release semaphore */
          if (bd_buf->use_count == 0)
            {
              if (bd_buf->modified)
                {
                  Chain_Append(&bd_ctx.mod, &bd_buf->link);
                  rc = rtems_semaphore_release(bd_ctx.flush_sema);
                }
              else
                {
                  Chain_Append(&bd_pool->lru, &bd_buf->link);
                  rc = rtems_semaphore_release(bd_pool->bufget_sema);
                }
            }
        }
    }
}

/* rtems_bdbuf_find_pool --
 *     Find first appropriate buffer pool. This primitive returns the index
 *     of first buffer pool which block size is greater than or equal to
 *     specified size.
 *
 * PARAMETERS:
 *     block_size - requested block size
 *     pool       - placeholder for result
 *
 * RETURNS:
 *     RTEMS status code: RTEMS_SUCCESSFUL if operation completed successfully,
 *     RTEMS_INVALID_SIZE if specified block size is invalid (not a power
 *     of 2), RTEMS_NOT_DEFINED if buffer pool for this or greater block size
 *     is not configured.
 */
rtems_status_code
rtems_bdbuf_find_pool(int block_size, rtems_bdpool_id *pool)
{
    rtems_bdpool_id i;
    bdbuf_pool *p;
    int cursize = INT_MAX;
    rtems_bdpool_id curid = -1;
    rtems_boolean found = FALSE;
    int j;

    for (j = block_size; (j != 0) && ((j & 1) == 0); j >>= 1);
    if (j != 1)
        return RTEMS_INVALID_SIZE;

    for (i = 0, p = bd_ctx.pool; i < bd_ctx.npools; i++, p++)
    {
        if ((p->blksize >= block_size) &&
            (p->blksize < cursize))
        {
            curid = i;
            cursize = p->blksize;
            found = TRUE;
        }
    }

    if (found)
    {
        if (pool != NULL)
            *pool = curid;
        return RTEMS_SUCCESSFUL;
    }
    else
    {
        return RTEMS_NOT_DEFINED;
    }
}

/* rtems_bdbuf_get_pool_info --
 *     Obtain characteristics of buffer pool with specified number.
 *
 * PARAMETERS:
 *     pool       - buffer pool number
 *     block_size - block size for which buffer pool is configured returned
 *                  there
 *     blocks     - number of buffers in buffer pool returned there
 *
 * RETURNS:
 *     RTEMS status code: RTEMS_SUCCESSFUL if operation completed successfully,
 *     RTEMS_INVALID_NUMBER if appropriate buffer pool is not configured.
 *
 * NOTE:
 *     Buffer pools enumerated contiguously starting from 0.
 */
rtems_status_code
rtems_bdbuf_get_pool_info(rtems_bdpool_id pool, int *block_size,
                          int *blocks)
{
    if (pool >= bd_ctx.npools)
        return RTEMS_INVALID_NUMBER;

    if (block_size != NULL)
    {
        *block_size = bd_ctx.pool[pool].blksize;
    }

    if (blocks != NULL)
    {
        *blocks = bd_ctx.pool[pool].nblks;
    }

    return RTEMS_SUCCESSFUL;
}