block_dev.c

ARM 嵌入式 系统 设计与实例开发 实验教材 二源码

/*
 *  linux/fs/block_dev.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *  Copyright (C) 2001  Andrea Arcangeli <[e-mail]> SuSE
 */

#include <linux/config.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/locks.h>
#include <linux/fcntl.h>
#include <linux/slab.h>
#include <linux/kmod.h>
#include <linux/major.h>
#include <linux/devfs_fs_kernel.h>
#include <linux/smp_lock.h>
#include <linux/iobuf.h>
#include <linux/highmem.h>
#include <linux/blkdev.h>
#include <linux/module.h>

#include <asm/uaccess.h>

#define MAX_BUF_PER_PAGE (PAGE_CACHE_SIZE / 512)

static unsigned long max_block(kdev_t dev)
{
	unsigned int retval = ~0U;
	int major = MAJOR(dev);

	if (blk_size[major]) {
		int minor = MINOR(dev);
		unsigned int blocks = blk_size[major][minor];
		if (blocks) {
			unsigned int size = block_size(dev);
			unsigned int sizebits = blksize_bits(size);
			blocks += (size-1) >> BLOCK_SIZE_BITS;
			retval = blocks << (BLOCK_SIZE_BITS - sizebits);
			if (sizebits > BLOCK_SIZE_BITS)
				retval = blocks >> (sizebits - BLOCK_SIZE_BITS);
		}
	}
	return retval;
}

static loff_t blkdev_size(kdev_t dev)
{
	unsigned int blocks = ~0U;
	int major = MAJOR(dev);

	if (blk_size[major]) {
		int minor = MINOR(dev);
		blocks = blk_size[major][minor];
	}
	return (loff_t) blocks << BLOCK_SIZE_BITS;
}

/* Kill _all_ buffers, dirty or not.. */
static void kill_bdev(struct block_device *bdev)
{
	invalidate_bdev(bdev, 1);
	truncate_inode_pages(bdev->bd_inode->i_mapping, 0);
}	

int set_blocksize(kdev_t dev, int size)
{
	int oldsize;
	struct block_device *bdev;

	/* Size must be a power of two, and between 512 and PAGE_SIZE */
	if (size > PAGE_SIZE || size < 512 || (size & (size-1)))
		return -EINVAL;

	/* Size cannot be smaller than the size supported by the device */
	if (size < get_hardsect_size(dev))
		return -EINVAL;

	/* No blocksize array? Implies hardcoded BLOCK_SIZE */
	if (!blksize_size[MAJOR(dev)]) {
		if (size == BLOCK_SIZE)
			return 0;
		return -EINVAL;
	}

	oldsize = blksize_size[MAJOR(dev)][MINOR(dev)];
	if (oldsize == size)
		return 0;

	if (!oldsize && size == BLOCK_SIZE) {
		blksize_size[MAJOR(dev)][MINOR(dev)] = size;
		return 0;
	}

	/* Ok, we're actually changing the blocksize.. */
	bdev = bdget(dev);
	sync_buffers(dev, 2);
	blksize_size[MAJOR(dev)][MINOR(dev)] = size;
	bdev->bd_inode->i_blkbits = blksize_bits(size);
	kill_bdev(bdev);
	bdput(bdev);
	return 0;
}

static int blkdev_get_block(struct inode * inode, long iblock, struct buffer_head * bh, int create)
{
	if (iblock >= max_block(inode->i_rdev))
		return -EIO;

	bh->b_dev = inode->i_rdev;
	bh->b_blocknr = iblock;
	bh->b_state |= 1UL << BH_Mapped;
	return 0;
}

static int blkdev_direct_IO(int rw, struct inode * inode, struct kiobuf * iobuf, unsigned long blocknr, int blocksize)
{
	return generic_direct_IO(rw, inode, iobuf, blocknr, blocksize, blkdev_get_block);
}

static int blkdev_writepage(struct page * page)
{
	return block_write_full_page(page, blkdev_get_block);
}

static int blkdev_readpage(struct file * file, struct page * page)
{
	return block_read_full_page(page, blkdev_get_block);
}

static int blkdev_prepare_write(struct file *file, struct page *page, unsigned from, unsigned to)
{
	return block_prepare_write(page, from, to, blkdev_get_block);
}

static int blkdev_commit_write(struct file *file, struct page *page, unsigned from, unsigned to)
{
	return block_commit_write(page, from, to);
}

/*
 * private llseek:
 * for a block special file file->f_dentry->d_inode->i_size is zero
 * so we compute the size by hand (just as in block_read/write above)
 */
static loff_t block_llseek(struct file *file, loff_t offset, int origin)
{
	/* ewww */
	loff_t size = file->f_dentry->d_inode->i_bdev->bd_inode->i_size;
	loff_t retval;

	switch (origin) {
		case 2:
			offset += size;
			break;
		case 1:
			offset += file->f_pos;
	}
	retval = -EINVAL;
	if (offset >= 0 && offset <= size) {
		if (offset != file->f_pos) {
			file->f_pos = offset;
			file->f_reada = 0;
			file->f_version = ++event;
		}
		retval = offset;
	}
	return retval;
}
	

static int __block_fsync(struct inode * inode)
{
	int ret, err;

	ret = filemap_fdatasync(inode->i_mapping);
	err = sync_buffers(inode->i_rdev, 1);
	if (err && !ret)
		ret = err;
	err = filemap_fdatawait(inode->i_mapping);
	if (err && !ret)
		ret = err;

	return ret;
}

/*
 *	Filp may be NULL when we are called by an msync of a vma
 *	since the vma has no handle.
 */
 
static int block_fsync(struct file *filp, struct dentry *dentry, int datasync)
{
	struct inode * inode = dentry->d_inode;

	return __block_fsync(inode);
}

/*
 * pseudo-fs
 */

static struct super_block *bd_read_super(struct super_block *sb, void *data, int silent)
{
	static struct super_operations sops = {};
	struct inode *root = new_inode(sb);
	if (!root)
		return NULL;
	root->i_mode = S_IFDIR | S_IRUSR | S_IWUSR;
	root->i_uid = root->i_gid = 0;
	root->i_atime = root->i_mtime = root->i_ctime = CURRENT_TIME;
	sb->s_maxbytes = ~0ULL;
	sb->s_blocksize = 1024;
	sb->s_blocksize_bits = 10;
	sb->s_magic = 0x62646576;
	sb->s_op = &sops;
	sb->s_root = d_alloc(NULL, &(const struct qstr) { "bdev:", 5, 0 });
	if (!sb->s_root) {
		iput(root);
		return NULL;
	}
	sb->s_root->d_sb = sb;
	sb->s_root->d_parent = sb->s_root;
	d_instantiate(sb->s_root, root);
	return sb;
}

static DECLARE_FSTYPE(bd_type, "bdev", bd_read_super, FS_NOMOUNT);

static struct vfsmount *bd_mnt;

/*
 * bdev cache handling - shamelessly stolen from inode.c
 * We use smaller hashtable, though.
 */

#define HASH_BITS	6
#define HASH_SIZE	(1UL << HASH_BITS)
#define HASH_MASK	(HASH_SIZE-1)
static struct list_head bdev_hashtable[HASH_SIZE];
static spinlock_t bdev_lock __cacheline_aligned_in_smp = SPIN_LOCK_UNLOCKED;
static kmem_cache_t * bdev_cachep;

#define alloc_bdev() \
	 ((struct block_device *) kmem_cache_alloc(bdev_cachep, SLAB_KERNEL))
#define destroy_bdev(bdev) kmem_cache_free(bdev_cachep, (bdev))

static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
{
	struct block_device * bdev = (struct block_device *) foo;

	if ((flags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) ==
	    SLAB_CTOR_CONSTRUCTOR)
	{
		memset(bdev, 0, sizeof(*bdev));
		sema_init(&bdev->bd_sem, 1);
		INIT_LIST_HEAD(&bdev->bd_inodes);
	}
}

void __init bdev_cache_init(void)
{
	int i, err;
	struct list_head *head = bdev_hashtable;

	i = HASH_SIZE;
	do {
		INIT_LIST_HEAD(head);
		head++;
		i--;
	} while (i);

	bdev_cachep = kmem_cache_create("bdev_cache",
					 sizeof(struct block_device),
					 0, SLAB_HWCACHE_ALIGN, init_once,
					 NULL);
	if (!bdev_cachep)
		panic("Cannot create bdev_cache SLAB cache");
	err = register_filesystem(&bd_type);
	if (err)
		panic("Cannot register bdev pseudo-fs");
	bd_mnt = kern_mount(&bd_type);
	err = PTR_ERR(bd_mnt);
	if (IS_ERR(bd_mnt))
		panic("Cannot create bdev pseudo-fs");
}

/*
 * Most likely _very_ bad one - but then it's hardly critical for small
 * /dev and can be fixed when somebody will need really large one.
 */
static inline unsigned long hash(dev_t dev)
{
	unsigned long tmp = dev;
	tmp = tmp + (tmp >> HASH_BITS) + (tmp >> HASH_BITS*2);
	return tmp & HASH_MASK;
}

static struct block_device *bdfind(dev_t dev, struct list_head *head)
{
	struct list_head *p;
	struct block_device *bdev;
	for (p=head->next; p!=head; p=p->next) {
		bdev = list_entry(p, struct block_device, bd_hash);
		if (bdev->bd_dev != dev)
			continue;
		atomic_inc(&bdev->bd_count);
		return bdev;
	}
	return NULL;
}

struct block_device *bdget(dev_t dev)
{
	struct list_head * head = bdev_hashtable + hash(dev);
	struct block_device *bdev, *new_bdev;
	spin_lock(&bdev_lock);
	bdev = bdfind(dev, head);
	spin_unlock(&bdev_lock);
	if (bdev)
		return bdev;
	new_bdev = alloc_bdev();
	if (new_bdev) {
		struct inode *inode = new_inode(bd_mnt->mnt_sb);
		if (inode) {
			kdev_t kdev = to_kdev_t(dev);
			atomic_set(&new_bdev->bd_count,1);
			new_bdev->bd_dev = dev;
			new_bdev->bd_op = NULL;
			new_bdev->bd_inode = inode;
			inode->i_rdev = kdev;
			inode->i_dev = kdev;
			inode->i_bdev = new_bdev;
			inode->i_data.a_ops = &def_blk_aops;
			inode->i_data.gfp_mask = GFP_USER;
			inode->i_mode = S_IFBLK;
			spin_lock(&bdev_lock);
			bdev = bdfind(dev, head);
			if (!bdev) {
				list_add(&new_bdev->bd_hash, head);
				spin_unlock(&bdev_lock);