}
static ssize_t
-rw_raw_dev(int rw, struct file *filp, char *buf, size_t size, loff_t *offp)
+rw_raw_dev(int rw, struct file *filp, const struct iovec *iov, unsigned long nr_segs, loff_t *offp)
{
const int minor = minor(filp->f_dentry->d_inode->i_rdev);
struct block_device *bdev = raw_devices[minor].binding;
struct inode *inode = bdev->bd_inode;
+ size_t count = iov_length(iov, nr_segs);
ssize_t ret = 0;
- if (size == 0)
- goto out;
- ret = -EINVAL;
- if (size < 0)
- goto out;
- ret = -ENXIO;
- if (*offp >= inode->i_size)
- goto out;
+ if (count == 0)
+ goto out;
+
+ if ((ssize_t)count < 0)
+ return -EINVAL;
+
+ if (*offp >= inode->i_size)
+ return -ENXIO;
+
+ if (count + *offp > inode->i_size) {
+ count = inode->i_size - *offp;
+ nr_segs = iov_shorten((struct iovec *)iov, nr_segs, count);
+ }
+ ret = generic_file_direct_IO(rw, inode, iov, *offp, nr_segs);
- if (size + *offp > inode->i_size)
- size = inode->i_size - *offp;
- ret = generic_file_direct_IO(rw, inode, buf, *offp, size);
if (ret > 0)
*offp += ret;
out:
}
static ssize_t
-raw_read(struct file *filp, char * buf, size_t size, loff_t *offp)
+raw_read(struct file *filp, char *buf, size_t size, loff_t *offp)
{
- return rw_raw_dev(READ, filp, buf, size, offp);
+ struct iovec local_iov = { .iov_base = buf, .iov_len = size};
+
+ return rw_raw_dev(READ, filp, &local_iov, 1, offp);
}
static ssize_t
raw_write(struct file *filp, const char *buf, size_t size, loff_t *offp)
{
- return rw_raw_dev(WRITE, filp, (char *)buf, size, offp);
+ struct iovec local_iov = { .iov_base = buf, .iov_len = size};
+
+ return rw_raw_dev(WRITE, filp, &local_iov, 1, offp);
+}
+
+static ssize_t
+raw_readv(struct file *filp, const struct iovec *iov, unsigned long nr_segs, loff_t *offp)
+{
+ return rw_raw_dev(READ, filp, iov, nr_segs, offp);
+}
+
+static ssize_t
+raw_writev(struct file *filp, const struct iovec *iov, unsigned long nr_segs, loff_t *offp)
+{
+ return rw_raw_dev(WRITE, filp, iov, nr_segs, offp);
}
static struct file_operations raw_fops = {
.open = raw_open,
.release= raw_release,
.ioctl = raw_ioctl,
+ .readv = raw_readv,
+ .writev = raw_writev,
.owner = THIS_MODULE,
};
}
static int
-blkdev_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count)
+blkdev_direct_IO(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs)
{
- return generic_direct_IO(rw, inode, buf, offset,
- count, blkdev_get_blocks);
+ return generic_direct_IO(rw, inode, iov, offset,
+ nr_segs, blkdev_get_blocks);
}
static int blkdev_writepage(struct page * page)
return res;
}
+static ssize_t blkdev_file_write(struct file *file, const char *buf,
+ size_t count, loff_t *ppos)
+{
+ struct iovec local_iov = { .iov_base = (void *)buf, .iov_len = count };
+
+ return generic_file_write_nolock(file, &local_iov, 1, ppos);
+}
+
static int blkdev_ioctl(struct inode *inode, struct file *file, unsigned cmd,
unsigned long arg)
{
}
struct address_space_operations def_blk_aops = {
- readpage: blkdev_readpage,
- writepage: blkdev_writepage,
- sync_page: block_sync_page,
- prepare_write: blkdev_prepare_write,
- commit_write: blkdev_commit_write,
- writepages: generic_writepages,
- vm_writeback: generic_vm_writeback,
- direct_IO: blkdev_direct_IO,
+ .readpage = blkdev_readpage,
+ .writepage = blkdev_writepage,
+ .sync_page = block_sync_page,
+ .prepare_write = blkdev_prepare_write,
+ .commit_write = blkdev_commit_write,
+ .writepages = generic_writepages,
+ .vm_writeback = generic_vm_writeback,
+ .direct_IO = blkdev_direct_IO,
};
struct file_operations def_blk_fops = {
- open: blkdev_open,
- release: blkdev_close,
- llseek: block_llseek,
- read: generic_file_read,
- write: generic_file_write_nolock,
- mmap: generic_file_mmap,
- fsync: block_fsync,
- ioctl: blkdev_ioctl,
- sendfile: generic_file_sendfile,
+ .open = blkdev_open,
+ .release = blkdev_close,
+ .llseek = block_llseek,
+ .read = generic_file_read,
+ .write = blkdev_file_write,
+ .mmap = generic_file_mmap,
+ .fsync = block_fsync,
+ .ioctl = blkdev_ioctl,
+ .readv = generic_file_readv,
+ .writev = generic_file_writev,
+ .sendfile = generic_file_sendfile,
};
int ioctl_by_bdev(struct block_device *bdev, unsigned cmd, unsigned long arg)
*/
static inline unsigned dio_pages_present(struct dio *dio)
{
- return dio->head - dio->tail;
+ return dio->tail - dio->head;
}
/*
static int dio_await_completion(struct dio *dio)
{
int ret = 0;
+
+ if (dio->bio)
+ dio_bio_submit(dio);
+
while (atomic_read(&dio->bio_count)) {
struct bio *bio = dio_await_one(dio);
int ret2;
return ret;
}
-/*
- * The main direct-IO function. This is a library function for use by
- * filesystem drivers.
- */
int
-generic_direct_IO(int rw, struct inode *inode, char *buf, loff_t offset,
- size_t count, get_blocks_t get_blocks)
+direct_io_worker(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs, get_blocks_t get_blocks)
{
const unsigned blkbits = inode->i_blkbits;
- const unsigned blocksize_mask = (1 << blkbits) - 1;
- const unsigned long user_addr = (unsigned long)buf;
- int ret;
- int ret2;
+ unsigned long user_addr;
+ int seg, ret2, ret = 0;
struct dio dio;
- size_t bytes;
+ size_t bytes, tot_bytes = 0;
- /* Check the memory alignment. Blocks cannot straddle pages */
- if ((user_addr & blocksize_mask) || (count & blocksize_mask)) {
- ret = -EINVAL;
- goto out;
- }
-
- /* BIO submission state */
dio.bio = NULL;
dio.bvec = NULL;
dio.inode = inode;
dio.blkbits = blkbits;
dio.block_in_file = offset >> blkbits;
dio.blocks_available = 0;
- dio.final_block_in_request = (offset + count) >> blkbits;
- /* Index into the first page of the first block */
- dio.first_block_in_page = (user_addr & (PAGE_SIZE - 1)) >> blkbits;
dio.boundary = 0;
dio.reap_counter = 0;
dio.get_blocks = get_blocks;
dio.last_block_in_bio = -1;
dio.next_block_in_bio = -1;
- /* Page fetching state */
- dio.curr_page = 0;
- bytes = count;
- dio.total_pages = 0;
- if (user_addr & (PAGE_SIZE - 1)) {
- dio.total_pages++;
- bytes -= PAGE_SIZE - (user_addr & (PAGE_SIZE - 1));
- }
-
- dio.total_pages += (bytes + PAGE_SIZE - 1) / PAGE_SIZE;
- dio.curr_user_address = user_addr;
-
- /* Page queue */
- dio.head = 0;
- dio.tail = 0;
dio.page_errors = 0;
/* BIO completion state */
dio.bio_list = NULL;
dio.waiter = NULL;
- ret = do_direct_IO(&dio);
+ for (seg = 0; seg < nr_segs; seg++) {
+ user_addr = (unsigned long)iov[seg].iov_base;
+ bytes = iov[seg].iov_len;
+
+ /* Index into the first page of the first block */
+ dio.first_block_in_page = (user_addr & (PAGE_SIZE - 1)) >> blkbits;
+ dio.final_block_in_request = dio.block_in_file + (bytes >> blkbits);
+ /* Page fetching state */
+ dio.head = 0;
+ dio.tail = 0;
+ dio.curr_page = 0;
+
+ dio.total_pages = 0;
+ if (user_addr & (PAGE_SIZE-1)) {
+ dio.total_pages++;
+ bytes -= PAGE_SIZE - (user_addr & (PAGE_SIZE - 1));
+ }
+ dio.total_pages += (bytes + PAGE_SIZE - 1) / PAGE_SIZE;
+ dio.curr_user_address = user_addr;
+
+ ret = do_direct_IO(&dio);
+
+ if (ret) {
+ dio_cleanup(&dio);
+ break;
+ }
+
+ tot_bytes += iov[seg].iov_len - ((dio.final_block_in_request -
+ dio.block_in_file) << blkbits);
+
+ } /* end iovec loop */
- if (dio.bio)
- dio_bio_submit(&dio);
- if (ret)
- dio_cleanup(&dio);
ret2 = dio_await_completion(&dio);
if (ret == 0)
ret = ret2;
if (ret == 0)
ret = dio.page_errors;
if (ret == 0)
- ret = count - ((dio.final_block_in_request -
- dio.block_in_file) << blkbits);
-out:
+ ret = tot_bytes;
+
return ret;
}
-ssize_t
-generic_file_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count)
+/*
+ * This is a library function for use by filesystem drivers.
+ */
+int
+generic_direct_IO(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs, get_blocks_t get_blocks)
{
+ int seg;
+ size_t size;
+ unsigned long addr;
struct address_space *mapping = inode->i_mapping;
- unsigned blocksize_mask;
- ssize_t retval;
+ unsigned blocksize_mask = (1 << inode->i_blkbits) - 1;
+ ssize_t retval = -EINVAL;
- blocksize_mask = (1 << inode->i_blkbits) - 1;
- if ((offset & blocksize_mask) || (count & blocksize_mask)) {
- retval = -EINVAL;
+ if (offset & blocksize_mask) {
goto out;
}
+ /* Check the memory alignment. Blocks cannot straddle pages */
+ for (seg = 0; seg < nr_segs; seg++) {
+ addr = (unsigned long)iov[seg].iov_base;
+ size = iov[seg].iov_len;
+ if ((addr & blocksize_mask) || (size & blocksize_mask))
+ goto out;
+ }
+
if (mapping->nrpages) {
retval = filemap_fdatawrite(mapping);
if (retval == 0)
if (retval)
goto out;
}
- retval = mapping->a_ops->direct_IO(rw, inode, buf, offset, count);
+
+ retval = direct_io_worker(rw, inode, iov, offset, nr_segs, get_blocks);
+out:
+ return retval;
+}
+
+ssize_t
+generic_file_direct_IO(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs)
+{
+ struct address_space *mapping = inode->i_mapping;
+ ssize_t retval;
+
+ retval = mapping->a_ops->direct_IO(rw, inode, iov, offset, nr_segs);
if (inode->i_mapping->nrpages)
invalidate_inode_pages2(inode->i_mapping);
-out:
return retval;
}
.open = generic_file_open,
.release = ext2_release_file,
.fsync = ext2_sync_file,
+ .readv = generic_file_readv,
+ .writev = generic_file_writev,
.sendfile = generic_file_sendfile,
};
}
static int
-ext2_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count)
+ext2_direct_IO(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs)
{
- return generic_direct_IO(rw, inode, buf,
- offset, count, ext2_get_blocks);
+ return generic_direct_IO(rw, inode, iov,
+ offset, nr_segs, ext2_get_blocks);
}
static int
}
struct file_operations ext3_file_operations = {
- .llseek = generic_file_llseek, /* BKL held */
- .read = generic_file_read, /* BKL not held. Don't need */
- .write = ext3_file_write, /* BKL not held. Don't need */
- .ioctl = ext3_ioctl, /* BKL held */
+ .llseek = generic_file_llseek,
+ .read = generic_file_read,
+ .write = ext3_file_write,
+ .readv = generic_file_readv,
+ .writev = generic_file_writev,
+ .ioctl = ext3_ioctl,
.mmap = generic_file_mmap,
- .open = ext3_open_file, /* BKL not held. Don't need */
- .release = ext3_release_file, /* BKL not held. Don't need */
- .fsync = ext3_sync_file, /* BKL held */
- .sendfile = generic_file_sendfile, /* BKL not held. Don't need */
+ .open = ext3_open_file,
+ .release = ext3_release_file,
+ .fsync = ext3_sync_file,
+ .sendfile = generic_file_sendfile,
};
struct inode_operations ext3_file_inode_operations = {
- .truncate = ext3_truncate, /* BKL held */
- .setattr = ext3_setattr, /* BKL held */
+ .truncate = ext3_truncate,
+ .setattr = ext3_setattr,
};
* If the O_DIRECT write is intantiating holes inside i_size and the machine
* crashes then stale disk data _may_ be exposed inside the file.
*/
-static int ext3_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count)
+static int ext3_direct_IO(int rw, struct inode *inode,
+ const struct iovec *iov, loff_t offset,
+ unsigned long nr_segs)
{
struct ext3_inode_info *ei = EXT3_I(inode);
handle_t *handle = NULL;
int ret;
int orphan = 0;
+ size_t count = iov_length(iov, nr_segs);
if (rw == WRITE) {
loff_t final_size = offset + count;
}
}
- ret = generic_direct_IO(rw, inode, buf, offset,
- count, ext3_direct_io_get_blocks);
+ ret = generic_direct_IO(rw, inode, iov, offset,
+ nr_segs, ext3_direct_io_get_blocks);
out_stop:
if (handle) {
.write = generic_file_write,
.read = generic_file_read,
.mmap = generic_file_mmap,
+ .readv = generic_file_readv,
+ .writev = generic_file_writev,
.sendfile = generic_file_sendfile,
.fsync = jfs_fsync,
};
return generic_block_bmap(mapping, block, jfs_get_block);
}
-static int jfs_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count)
+static int jfs_direct_IO(int rw, struct inode *inode, const struct iovec *iov,
+ loff_t offset, unsigned long nr_segs)
{
- return generic_direct_IO(rw, inode, buf,
- offset, count, jfs_get_blocks);
+ return generic_direct_IO(rw, inode, iov,
+ offset, nr_segs, jfs_get_blocks);
}
struct address_space_operations jfs_aops = {
return ret;
}
+/*
+ * Reduce an iovec's length in-place. Return the resulting number of segments
+ */
+unsigned long iov_shorten(struct iovec *iov, unsigned long nr_segs, size_t to)
+{
+ unsigned long seg = 0;
+ size_t len = 0;
+
+ while (seg < nr_segs) {
+ seg++;
+ if (len + iov->iov_len >= to) {
+ iov->iov_len = to - len;
+ break;
+ }
+ len += iov->iov_len;
+ iov++;
+ }
+ return seg;
+}
+
static ssize_t do_readv_writev(int type, struct file *file,
const struct iovec * vector,
- unsigned long count)
+ unsigned long nr_segs)
{
typedef ssize_t (*io_fn_t)(struct file *, char *, size_t, loff_t *);
typedef ssize_t (*iov_fn_t)(struct file *, const struct iovec *, unsigned long, loff_t *);
size_t tot_len;
struct iovec iovstack[UIO_FASTIOV];
struct iovec *iov=iovstack;
- ssize_t ret, i;
+ ssize_t ret = -EINVAL;
+ int seg;
io_fn_t fn;
iov_fn_t fnv;
struct inode *inode;
+ /*
+ * SuS says "The readv() function *may* fail if the iovcnt argument
+ * was less than or equal to 0, or greater than {IOV_MAX}. Linux has
+ * traditionally returned -EINVAL for zero segments, so...
+ */
+ if (nr_segs == 0)
+ goto out;
+
/*
* First get the "struct iovec" from user memory and
* verify all the pointers
*/
- ret = 0;
- if (!count)
- goto out_nofree;
- ret = -EINVAL;
- if (count > UIO_MAXIOV)
- goto out_nofree;
+ if ((nr_segs > UIO_MAXIOV) || (nr_segs <= 0))
+ goto out;
if (!file->f_op)
- goto out_nofree;
- if (count > UIO_FASTIOV) {
+ goto out;
+ if (nr_segs > UIO_FASTIOV) {
ret = -ENOMEM;
- iov = kmalloc(count*sizeof(struct iovec), GFP_KERNEL);
+ iov = kmalloc(nr_segs*sizeof(struct iovec), GFP_KERNEL);
if (!iov)
- goto out_nofree;
+ goto out;
}
ret = -EFAULT;
- if (copy_from_user(iov, vector, count*sizeof(*vector)))
+ if (copy_from_user(iov, vector, nr_segs*sizeof(*vector)))
goto out;
/*
* Single unix specification:
- * We should -EINVAL if an element length is not >= 0 and fitting an ssize_t
- * The total length is fitting an ssize_t
+ * We should -EINVAL if an element length is not >= 0 and fitting an
+ * ssize_t. The total length is fitting an ssize_t
*
* Be careful here because iov_len is a size_t not an ssize_t
*/
-
tot_len = 0;
ret = -EINVAL;
- for (i = 0 ; i < count ; i++) {
+ for (seg = 0 ; seg < nr_segs; seg++) {
ssize_t tmp = tot_len;
- ssize_t len = (ssize_t)iov[i].iov_len;
+ ssize_t len = (ssize_t)iov[seg].iov_len;
if (len < 0) /* size_t not fitting an ssize_t .. */
goto out;
tot_len += len;
if (tot_len < tmp) /* maths overflow on the ssize_t */
goto out;
}
+ if (tot_len == 0) {
+ ret = 0;
+ goto out;
+ }
inode = file->f_dentry->d_inode;
/* VERIFY_WRITE actually means a read, as we write to user space */
- ret = locks_verify_area((type == VERIFY_WRITE
+ ret = locks_verify_area((type == READ
? FLOCK_VERIFY_READ : FLOCK_VERIFY_WRITE),
inode, file, file->f_pos, tot_len);
- if (ret) goto out;
+ if (ret)
+ goto out;
- fnv = (type == VERIFY_WRITE ? file->f_op->readv : file->f_op->writev);
+ fnv = NULL;
+ if (type == READ) {
+ fn = file->f_op->read;
+ fnv = file->f_op->readv;
+ } else {
+ fn = (io_fn_t)file->f_op->write;
+ fnv = file->f_op->writev;
+ }
if (fnv) {
- ret = fnv(file, iov, count, &file->f_pos);
+ ret = fnv(file, iov, nr_segs, &file->f_pos);
goto out;
}
- /* VERIFY_WRITE actually means a read, as we write to user space */
- fn = (type == VERIFY_WRITE ? file->f_op->read :
- (io_fn_t) file->f_op->write);
-
+ /* Do it by hand, with file-ops */
ret = 0;
vector = iov;
- while (count > 0) {
+ while (nr_segs > 0) {
void * base;
size_t len;
ssize_t nr;
base = vector->iov_base;
len = vector->iov_len;
vector++;
- count--;
+ nr_segs--;
nr = fn(file, base, len, &file->f_pos);
if (nr != len)
break;
}
-
out:
if (iov != iovstack)
kfree(iov);
-out_nofree:
- /* VERIFY_WRITE actually means a read, as we write to user space */
- if ((ret + (type == VERIFY_WRITE)) > 0)
+ if ((ret + (type == READ)) > 0)
dnotify_parent(file->f_dentry,
- (type == VERIFY_WRITE) ? DN_MODIFY : DN_ACCESS);
+ (type == READ) ? DN_MODIFY : DN_ACCESS);
return ret;
}
-asmlinkage ssize_t sys_readv(unsigned long fd, const struct iovec * vector,
- unsigned long count)
+
+asmlinkage ssize_t
+sys_readv(unsigned long fd, const struct iovec *vector, unsigned long nr_segs)
{
struct file * file;
ssize_t ret;
(file->f_op->readv || file->f_op->read)) {
ret = security_ops->file_permission (file, MAY_READ);
if (!ret)
- ret = do_readv_writev(VERIFY_WRITE, file, vector, count);
+ ret = do_readv_writev(READ, file, vector, nr_segs);
}
fput(file);
return ret;
}
-asmlinkage ssize_t sys_writev(unsigned long fd, const struct iovec * vector,
- unsigned long count)
+asmlinkage ssize_t
+sys_writev(unsigned long fd, const struct iovec * vector, unsigned long nr_segs)
{
struct file * file;
ssize_t ret;
(file->f_op->writev || file->f_op->write)) {
ret = security_ops->file_permission (file, MAY_WRITE);
if (!ret)
- ret = do_readv_writev(VERIFY_READ, file, vector, count);
+ ret = do_readv_writev(WRITE, file, vector, nr_segs);
}
fput(file);
int (*bmap)(struct address_space *, long);
int (*invalidatepage) (struct page *, unsigned long);
int (*releasepage) (struct page *, int);
- int (*direct_IO)(int, struct inode *, char *buf,
- loff_t offset, size_t count);
+ int (*direct_IO)(int, struct inode *, const struct iovec *iov, loff_t offset, unsigned long nr_segs);
};
struct backing_dev_info;
extern int file_read_actor(read_descriptor_t * desc, struct page *page, unsigned long offset, unsigned long size);
extern ssize_t generic_file_read(struct file *, char *, size_t, loff_t *);
extern ssize_t generic_file_write(struct file *, const char *, size_t, loff_t *);
-extern ssize_t generic_file_write_nolock(struct file *, const char *, size_t, loff_t *);
+ssize_t generic_file_write_nolock(struct file *file, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos);
extern ssize_t generic_file_sendfile(struct file *, struct file *, loff_t *, size_t);
extern void do_generic_file_read(struct file *, loff_t *, read_descriptor_t *, read_actor_t);
-ssize_t generic_file_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count);
-int generic_direct_IO(int rw, struct inode *inode, char *buf,
- loff_t offset, size_t count, get_blocks_t *get_blocks);
-
+extern ssize_t generic_file_direct_IO(int rw, struct inode *inode,
+ const struct iovec *iov, loff_t offset, unsigned long nr_segs);
+extern int generic_direct_IO(int rw, struct inode *inode, const struct iovec
+ *iov, loff_t offset, unsigned long nr_segs, get_blocks_t *get_blocks);
+extern ssize_t generic_file_readv(struct file *filp, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos);
+ssize_t generic_file_writev(struct file *filp, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos);
extern loff_t no_llseek(struct file *file, loff_t offset, int origin);
extern loff_t generic_file_llseek(struct file *file, loff_t offset, int origin);
extern loff_t remote_llseek(struct file *file, loff_t offset, int origin);
/* Beg pardon: BSD has 1024 --ANK */
#endif
+/*
+ * Total number of bytes covered by an iovec
+ */
+static inline size_t iov_length(const struct iovec *iov, unsigned long nr_segs)
+{
+ unsigned long seg;
+ size_t ret = 0;
+
+ for (seg = 0; seg < nr_segs; seg++)
+ ret += iov[seg].iov_len;
+ return ret;
+}
+
+unsigned long iov_shorten(struct iovec *iov, unsigned long nr_segs, size_t to);
+
#endif
#include <linux/highuid.h>
#include <linux/brlock.h>
#include <linux/fs.h>
+#include <linux/uio.h>
#include <linux/tty.h>
#include <linux/in6.h>
#include <linux/completion.h>
EXPORT_SYMBOL(read_dev_sector);
EXPORT_SYMBOL(init_buffer);
EXPORT_SYMBOL_GPL(generic_file_direct_IO);
+EXPORT_SYMBOL(generic_file_readv);
+EXPORT_SYMBOL(generic_file_writev);
+EXPORT_SYMBOL(iov_shorten);
/* tty routines */
EXPORT_SYMBOL(tty_hangup);
#include <linux/mman.h>
#include <linux/pagemap.h>
#include <linux/file.h>
+#include <linux/uio.h>
#include <linux/iobuf.h>
#include <linux/hash.h>
#include <linux/writeback.h>
* This is the "read()" routine for all filesystems
* that can use the page cache directly.
*/
-ssize_t
-generic_file_read(struct file *filp, char *buf, size_t count, loff_t *ppos)
+static ssize_t
+__generic_file_read(struct file *filp, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos)
{
ssize_t retval;
+ unsigned long seg;
+ size_t count = iov_length(iov, nr_segs);
if ((ssize_t) count < 0)
return -EINVAL;
+ /* coalesce the iovecs and go direct-to-BIO for O_DIRECT */
if (filp->f_flags & O_DIRECT) {
loff_t pos = *ppos, size;
struct address_space *mapping;
goto out; /* skip atime */
size = inode->i_size;
if (pos < size) {
- if (pos + count > size)
+ if (pos + count > size) {
count = size - pos;
- retval = generic_file_direct_IO(READ, inode,
- buf, pos, count);
+ nr_segs = iov_shorten((struct iovec *)iov,
+ nr_segs, count);
+ }
+ retval = generic_file_direct_IO(READ, inode,
+ iov, pos, nr_segs);
if (retval > 0)
*ppos = pos + retval;
}
goto out;
}
- retval = -EFAULT;
- if (access_ok(VERIFY_WRITE, buf, count)) {
- retval = 0;
+ for (seg = 0; seg < nr_segs; seg++) {
+ if (!access_ok(VERIFY_WRITE,iov[seg].iov_base,iov[seg].iov_len))
+ return -EFAULT;
+ }
- if (count) {
+ retval = 0;
+ if (count) {
+ for (seg = 0; seg < nr_segs; seg++) {
read_descriptor_t desc;
desc.written = 0;
- desc.count = count;
- desc.buf = buf;
+ desc.buf = iov[seg].iov_base;
+ desc.count = iov[seg].iov_len;
+ if (desc.count == 0)
+ continue;
desc.error = 0;
do_generic_file_read(filp,ppos,&desc,file_read_actor);
- retval = desc.written;
- if (!retval)
+ retval += desc.written;
+ if (!retval) {
retval = desc.error;
+ break;
+ }
}
}
out:
return retval;
}
+ssize_t
+generic_file_read(struct file *filp, char *buf, size_t count, loff_t *ppos)
+{
+ struct iovec local_iov = { .iov_base = buf, .iov_len = count };
+
+ return __generic_file_read(filp, &local_iov, 1, ppos);
+}
+
static int file_send_actor(read_descriptor_t * desc, struct page *page, unsigned long offset, unsigned long size)
{
ssize_t written;
* it for writing by marking it dirty.
* okir@monad.swb.de
*/
-ssize_t generic_file_write_nolock(struct file *file, const char *buf,
- size_t count, loff_t *ppos)
+ssize_t
+generic_file_write_nolock(struct file *file, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos)
{
struct address_space * mapping = file->f_dentry->d_inode->i_mapping;
struct address_space_operations *a_ops = mapping->a_ops;
+ const size_t ocount = iov_length(iov, nr_segs);
+ size_t count = ocount;
struct inode *inode = mapping->host;
unsigned long limit = current->rlim[RLIMIT_FSIZE].rlim_cur;
long status = 0;
unsigned bytes;
time_t time_now;
struct pagevec lru_pvec;
+ struct iovec *cur_iov;
+ unsigned iov_bytes; /* Cumulative count to the end of the
+ current iovec */
+ unsigned long seg;
+ char *buf;
if (unlikely((ssize_t)count < 0))
return -EINVAL;
- if (unlikely(!access_ok(VERIFY_READ, buf, count)))
- return -EFAULT;
+ for (seg = 0; seg < nr_segs; seg++) {
+ if (!access_ok(VERIFY_READ,iov[seg].iov_base,iov[seg].iov_len))
+ return -EFAULT;
+ }
pos = *ppos;
if (unlikely(pos < 0))
mark_inode_dirty_sync(inode);
}
+ /* coalesce the iovecs and go direct-to-BIO for O_DIRECT */
if (unlikely(file->f_flags & O_DIRECT)) {
- written = generic_file_direct_IO(WRITE, inode,
- (char *)buf, pos, count);
+ if (count != ocount)
+ nr_segs = iov_shorten((struct iovec *)iov,
+ nr_segs, count);
+ written = generic_file_direct_IO(WRITE, inode,
+ iov, pos, nr_segs);
if (written > 0) {
loff_t end = pos + written;
if (end > inode->i_size && !S_ISBLK(inode->i_mode)) {
goto out_status;
}
+ cur_iov = (struct iovec *)iov;
+ iov_bytes = cur_iov->iov_len;
+ buf = cur_iov->iov_base;
do {
unsigned long index;
unsigned long offset;
bytes = PAGE_CACHE_SIZE - offset;
if (bytes > count)
bytes = count;
+ if (bytes + written > iov_bytes)
+ bytes = iov_bytes - written;
/*
* Bring in the user page that we will copy from _first_.
*/
fault_in_pages_readable(buf, bytes);
- page = __grab_cache_page(mapping, index, &cached_page, &lru_pvec);
+ page = __grab_cache_page(mapping,index,&cached_page,&lru_pvec);
if (!page) {
status = -ENOMEM;
break;
count -= status;
pos += status;
buf += status;
+ if (written == iov_bytes && count) {
+ cur_iov++;
+ iov_bytes += cur_iov->iov_len;
+ buf = cur_iov->iov_base;
+ }
}
}
if (!PageReferenced(page))
{
struct inode *inode = file->f_dentry->d_inode->i_mapping->host;
int err;
+ struct iovec local_iov = { .iov_base = (void *)buf, .iov_len = count };
down(&inode->i_sem);
- err = generic_file_write_nolock(file, buf, count, ppos);
+ err = generic_file_write_nolock(file, &local_iov, 1, ppos);
up(&inode->i_sem);
return err;
}
+
+ssize_t generic_file_readv(struct file *filp, const struct iovec *iov,
+ unsigned long nr_segs, loff_t *ppos)
+{
+ return __generic_file_read(filp, iov, nr_segs, ppos);
+}
+
+ssize_t generic_file_writev(struct file *file, const struct iovec *iov,
+ unsigned long nr_segs, loff_t * ppos)
+{
+ struct inode *inode = file->f_dentry->d_inode;
+ ssize_t ret;
+
+ down(&inode->i_sem);
+ ret = generic_file_write_nolock(file, iov, nr_segs, ppos);
+ up(&inode->i_sem);
+ return ret;
+}
projects / linux / commitdiff