kevin/linux
1
Fork 0
Code
70524490ee
linux/fs/pipe.c
Jens Axboe 70524490ee [PATCH] splice: add support for sys_tee() 
Basically an in-kernel implementation of tee, which uses splice and the
pipe buffers as an intelligent way to pass data around by reference.

Where the user space tee consumes the input and produces a stdout and
file output, this syscall merely duplicates the data inside a pipe to
another pipe. No data is copied, the output just grabs a reference to the
input pipe data.

Signed-off-by: Jens Axboe <axboe@suse.de>
2006-04-11 15:51:17 +02:00

912 lines
19 KiB
C
Raw Blame History

/*
* linux/fs/pipe.c
*
* Copyright (C) 1991, 1992, 1999 Linus Torvalds
*/
#include <linux/mm.h>
#include <linux/file.h>
#include <linux/poll.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/mount.h>
#include <linux/pipe_fs_i.h>
#include <linux/uio.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
/*
* We use a start+len construction, which provides full use of the
* allocated memory.
* -- Florian Coosmann (FGC)
*
* Reads with count = 0 should always return 0.
* -- Julian Bradfield 1999-06-07.
*
* FIFOs and Pipes now generate SIGIO for both readers and writers.
* -- Jeremy Elson <jelson@circlemud.org> 2001-08-16
*
* pipe_read & write cleanup
* -- Manfred Spraul <manfred@colorfullife.com> 2002-05-09
*/
/* Drop the inode semaphore and wait for a pipe event, atomically */
void pipe_wait(struct pipe_inode_info *pipe)
{
DEFINE_WAIT(wait);
/*
* Pipes are system-local resources, so sleeping on them
* is considered a noninteractive wait:
*/
prepare_to_wait(&pipe->wait, &wait,
TASK_INTERRUPTIBLE | TASK_NONINTERACTIVE);
if (pipe->inode)
mutex_unlock(&pipe->inode->i_mutex);
schedule();
finish_wait(&pipe->wait, &wait);
if (pipe->inode)
mutex_lock(&pipe->inode->i_mutex);
}
static int
pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len)
{
unsigned long copy;
while (len > 0) {
while (!iov->iov_len)
iov++;
copy = min_t(unsigned long, len, iov->iov_len);
if (copy_from_user(to, iov->iov_base, copy))
return -EFAULT;
to += copy;
len -= copy;
iov->iov_base += copy;
iov->iov_len -= copy;
}
return 0;
}
static int
pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len)
{
unsigned long copy;
while (len > 0) {
while (!iov->iov_len)
iov++;
copy = min_t(unsigned long, len, iov->iov_len);
if (copy_to_user(iov->iov_base, from, copy))
return -EFAULT;
from += copy;
len -= copy;
iov->iov_base += copy;
iov->iov_len -= copy;
}
return 0;
}
static void anon_pipe_buf_release(struct pipe_inode_info *pipe,
struct pipe_buffer *buf)
{
struct page *page = buf->page;
buf->flags &= ~PIPE_BUF_FLAG_STOLEN;
/*
* If nobody else uses this page, and we don't already have a
* temporary page, let's keep track of it as a one-deep
* allocation cache. (Otherwise just release our reference to it)
*/
if (page_count(page) == 1 && !pipe->tmp_page)
pipe->tmp_page = page;
else
page_cache_release(page);
}
static void * anon_pipe_buf_map(struct file *file, struct pipe_inode_info *pipe,
struct pipe_buffer *buf)
{
return kmap(buf->page);
}
static void anon_pipe_buf_unmap(struct pipe_inode_info *pipe,
struct pipe_buffer *buf)
{
kunmap(buf->page);
}
static int anon_pipe_buf_steal(struct pipe_inode_info *pipe,
struct pipe_buffer *buf)
{
buf->flags |= PIPE_BUF_FLAG_STOLEN;
return 0;
}
static void anon_pipe_buf_get(struct pipe_inode_info *info,
struct pipe_buffer *buf)
{
page_cache_get(buf->page);
}
static struct pipe_buf_operations anon_pipe_buf_ops = {
.can_merge = 1,
.map = anon_pipe_buf_map,
.unmap = anon_pipe_buf_unmap,
.release = anon_pipe_buf_release,
.steal = anon_pipe_buf_steal,
.get = anon_pipe_buf_get,
};
static ssize_t
pipe_readv(struct file *filp, const struct iovec *_iov,
unsigned long nr_segs, loff_t *ppos)
{
struct inode *inode = filp->f_dentry->d_inode;
struct pipe_inode_info *pipe;
int do_wakeup;
ssize_t ret;
struct iovec *iov = (struct iovec *)_iov;
size_t total_len;
total_len = iov_length(iov, nr_segs);
/* Null read succeeds. */
if (unlikely(total_len == 0))
return 0;
do_wakeup = 0;
ret = 0;
mutex_lock(&inode->i_mutex);
pipe = inode->i_pipe;
for (;;) {
int bufs = pipe->nrbufs;
if (bufs) {
int curbuf = pipe->curbuf;
struct pipe_buffer *buf = pipe->bufs + curbuf;
struct pipe_buf_operations *ops = buf->ops;
void *addr;
size_t chars = buf->len;
int error;
if (chars > total_len)
chars = total_len;
addr = ops->map(filp, pipe, buf);
if (IS_ERR(addr)) {
if (!ret)
ret = PTR_ERR(addr);
break;
}
error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars);
ops->unmap(pipe, buf);
if (unlikely(error)) {
if (!ret)
ret = -EFAULT;
break;
}
ret += chars;
buf->offset += chars;
buf->len -= chars;
if (!buf->len) {
buf->ops = NULL;
ops->release(pipe, buf);
curbuf = (curbuf + 1) & (PIPE_BUFFERS-1);
pipe->curbuf = curbuf;
pipe->nrbufs = --bufs;
do_wakeup = 1;
}
total_len -= chars;
if (!total_len)
break; /* common path: read succeeded */
}
if (bufs) /* More to do? */
continue;
if (!pipe->writers)
break;
if (!pipe->waiting_writers) {
/* syscall merging: Usually we must not sleep
* if O_NONBLOCK is set, or if we got some data.
* But if a writer sleeps in kernel space, then
* we can wait for that data without violating POSIX.
*/
if (ret)
break;
if (filp->f_flags & O_NONBLOCK) {
ret = -EAGAIN;
break;
}
}
if (signal_pending(current)) {
if (!ret)
ret = -ERESTARTSYS;
break;
}
if (do_wakeup) {
wake_up_interruptible_sync(&pipe->wait);
kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
}
pipe_wait(pipe);
}
mutex_unlock(&inode->i_mutex);
/* Signal writers asynchronously that there is more room. */
if (do_wakeup) {
wake_up_interruptible(&pipe->wait);
kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
}
if (ret > 0)
file_accessed(filp);
return ret;
}
static ssize_t
pipe_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos)
{
struct iovec iov = { .iov_base = buf, .iov_len = count };
return pipe_readv(filp, &iov, 1, ppos);
}
static ssize_t
pipe_writev(struct file *filp, const struct iovec *_iov,
unsigned long nr_segs, loff_t *ppos)
{
struct inode *inode = filp->f_dentry->d_inode;
struct pipe_inode_info *pipe;
ssize_t ret;
int do_wakeup;
struct iovec *iov = (struct iovec *)_iov;
size_t total_len;
ssize_t chars;
total_len = iov_length(iov, nr_segs);
/* Null write succeeds. */
if (unlikely(total_len == 0))
return 0;
do_wakeup = 0;
ret = 0;
mutex_lock(&inode->i_mutex);
pipe = inode->i_pipe;
if (!pipe->readers) {
send_sig(SIGPIPE, current, 0);
ret = -EPIPE;
goto out;
}
/* We try to merge small writes */
chars = total_len & (PAGE_SIZE-1); /* size of the last buffer */
if (pipe->nrbufs && chars != 0) {
int lastbuf = (pipe->curbuf + pipe->nrbufs - 1) &
(PIPE_BUFFERS-1);
struct pipe_buffer *buf = pipe->bufs + lastbuf;
struct pipe_buf_operations *ops = buf->ops;
int offset = buf->offset + buf->len;
if (ops->can_merge && offset + chars <= PAGE_SIZE) {
void *addr;
int error;
addr = ops->map(filp, pipe, buf);
if (IS_ERR(addr)) {
error = PTR_ERR(addr);
goto out;
}
error = pipe_iov_copy_from_user(offset + addr, iov,
chars);
ops->unmap(pipe, buf);
ret = error;
do_wakeup = 1;
if (error)
goto out;
buf->len += chars;
total_len -= chars;
ret = chars;
if (!total_len)
goto out;
}
}
for (;;) {
int bufs;
if (!pipe->readers) {
send_sig(SIGPIPE, current, 0);
if (!ret)
ret = -EPIPE;
break;
}
bufs = pipe->nrbufs;
if (bufs < PIPE_BUFFERS) {
int newbuf = (pipe->curbuf + bufs) & (PIPE_BUFFERS-1);
struct pipe_buffer *buf = pipe->bufs + newbuf;
struct page *page = pipe->tmp_page;
int error;
if (!page) {
page = alloc_page(GFP_HIGHUSER);
if (unlikely(!page)) {
ret = ret ? : -ENOMEM;
break;
}
pipe->tmp_page = page;
}
/* Always wake up, even if the copy fails. Otherwise
* we lock up (O_NONBLOCK-)readers that sleep due to
* syscall merging.
* FIXME! Is this really true?
*/
do_wakeup = 1;
chars = PAGE_SIZE;
if (chars > total_len)
chars = total_len;
error = pipe_iov_copy_from_user(kmap(page), iov, chars);
kunmap(page);
if (unlikely(error)) {
if (!ret)
ret = -EFAULT;
break;
}
ret += chars;
/* Insert it into the buffer array */
buf->page = page;
buf->ops = &anon_pipe_buf_ops;
buf->offset = 0;
buf->len = chars;
pipe->nrbufs = ++bufs;
pipe->tmp_page = NULL;
total_len -= chars;
if (!total_len)
break;
}
if (bufs < PIPE_BUFFERS)
continue;
if (filp->f_flags & O_NONBLOCK) {
if (!ret)
ret = -EAGAIN;
break;
}
if (signal_pending(current)) {
if (!ret)
ret = -ERESTARTSYS;
break;
}
if (do_wakeup) {
wake_up_interruptible_sync(&pipe->wait);
kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN);
do_wakeup = 0;
}
pipe->waiting_writers++;
pipe_wait(pipe);
pipe->waiting_writers--;
}
out:
mutex_unlock(&inode->i_mutex);
if (do_wakeup) {
wake_up_interruptible(&pipe->wait);
kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN);
}
if (ret > 0)
file_update_time(filp);
return ret;
}
static ssize_t
pipe_write(struct file *filp, const char __user *buf,
size_t count, loff_t *ppos)
{
struct iovec iov = { .iov_base = (void __user *)buf, .iov_len = count };
return pipe_writev(filp, &iov, 1, ppos);
}
static ssize_t
bad_pipe_r(struct file *filp, char __user *buf, size_t count, loff_t *ppos)
{
return -EBADF;
}
static ssize_t
bad_pipe_w(struct file *filp, const char __user *buf, size_t count,
loff_t *ppos)
{
return -EBADF;
}
static int
pipe_ioctl(struct inode *pino, struct file *filp,
unsigned int cmd, unsigned long arg)
{
struct inode *inode = filp->f_dentry->d_inode;
struct pipe_inode_info *pipe;
int count, buf, nrbufs;
switch (cmd) {
case FIONREAD:
mutex_lock(&inode->i_mutex);
pipe = inode->i_pipe;
count = 0;
buf = pipe->curbuf;
nrbufs = pipe->nrbufs;
while (--nrbufs >= 0) {
count += pipe->bufs[buf].len;
buf = (buf+1) & (PIPE_BUFFERS-1);
}
mutex_unlock(&inode->i_mutex);
return put_user(count, (int __user *)arg);
default:
return -EINVAL;
}
}
/* No kernel lock held - fine */
static unsigned int
pipe_poll(struct file *filp, poll_table *wait)
{
unsigned int mask;
struct inode *inode = filp->f_dentry->d_inode;
struct pipe_inode_info *pipe = inode->i_pipe;
int nrbufs;
poll_wait(filp, &pipe->wait, wait);
/* Reading only -- no need for acquiring the semaphore. */
nrbufs = pipe->nrbufs;
mask = 0;
if (filp->f_mode & FMODE_READ) {
mask = (nrbufs > 0) ? POLLIN | POLLRDNORM : 0;
if (!pipe->writers && filp->f_version != pipe->w_counter)
mask |= POLLHUP;
}
if (filp->f_mode & FMODE_WRITE) {
mask |= (nrbufs < PIPE_BUFFERS) ? POLLOUT | POLLWRNORM : 0;
/*
* Most Unices do not set POLLERR for FIFOs but on Linux they
* behave exactly like pipes for poll().
*/
if (!pipe->readers)
mask |= POLLERR;
}
return mask;
}
static int
pipe_release(struct inode *inode, int decr, int decw)
{
struct pipe_inode_info *pipe;
mutex_lock(&inode->i_mutex);
pipe = inode->i_pipe;
pipe->readers -= decr;
pipe->writers -= decw;
if (!pipe->readers && !pipe->writers) {
free_pipe_info(inode);
} else {
wake_up_interruptible(&pipe->wait);
kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN);
kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
}
mutex_unlock(&inode->i_mutex);
return 0;
}
static int
pipe_read_fasync(int fd, struct file *filp, int on)
{
struct inode *inode = filp->f_dentry->d_inode;
int retval;
mutex_lock(&inode->i_mutex);
retval = fasync_helper(fd, filp, on, &inode->i_pipe->fasync_readers);
mutex_unlock(&inode->i_mutex);
if (retval < 0)
return retval;
return 0;
}
static int
pipe_write_fasync(int fd, struct file *filp, int on)
{
struct inode *inode = filp->f_dentry->d_inode;
int retval;
mutex_lock(&inode->i_mutex);
retval = fasync_helper(fd, filp, on, &inode->i_pipe->fasync_writers);
mutex_unlock(&inode->i_mutex);
if (retval < 0)
return retval;
return 0;
}
static int
pipe_rdwr_fasync(int fd, struct file *filp, int on)
{
struct inode *inode = filp->f_dentry->d_inode;
struct pipe_inode_info *pipe = inode->i_pipe;
int retval;
mutex_lock(&inode->i_mutex);
retval = fasync_helper(fd, filp, on, &pipe->fasync_readers);
if (retval >= 0)
retval = fasync_helper(fd, filp, on, &pipe->fasync_writers);
mutex_unlock(&inode->i_mutex);
if (retval < 0)
return retval;
return 0;
}
static int
pipe_read_release(struct inode *inode, struct file *filp)
{
pipe_read_fasync(-1, filp, 0);
return pipe_release(inode, 1, 0);
}
static int
pipe_write_release(struct inode *inode, struct file *filp)
{
pipe_write_fasync(-1, filp, 0);
return pipe_release(inode, 0, 1);
}
static int
pipe_rdwr_release(struct inode *inode, struct file *filp)
{
int decr, decw;
pipe_rdwr_fasync(-1, filp, 0);
decr = (filp->f_mode & FMODE_READ) != 0;
decw = (filp->f_mode & FMODE_WRITE) != 0;
return pipe_release(inode, decr, decw);
}
static int
pipe_read_open(struct inode *inode, struct file *filp)
{
/* We could have perhaps used atomic_t, but this and friends
below are the only places. So it doesn't seem worthwhile. */
mutex_lock(&inode->i_mutex);
inode->i_pipe->readers++;
mutex_unlock(&inode->i_mutex);
return 0;
}
static int
pipe_write_open(struct inode *inode, struct file *filp)
{
mutex_lock(&inode->i_mutex);
inode->i_pipe->writers++;
mutex_unlock(&inode->i_mutex);
return 0;
}
static int
pipe_rdwr_open(struct inode *inode, struct file *filp)
{
mutex_lock(&inode->i_mutex);
if (filp->f_mode & FMODE_READ)
inode->i_pipe->readers++;
if (filp->f_mode & FMODE_WRITE)
inode->i_pipe->writers++;
mutex_unlock(&inode->i_mutex);
return 0;
}
/*
* The file_operations structs are not static because they
* are also used in linux/fs/fifo.c to do operations on FIFOs.
*/
const struct file_operations read_fifo_fops = {
.llseek = no_llseek,
.read = pipe_read,
.readv = pipe_readv,
.write = bad_pipe_w,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_read_open,
.release = pipe_read_release,
.fasync = pipe_read_fasync,
};
const struct file_operations write_fifo_fops = {
.llseek = no_llseek,
.read = bad_pipe_r,
.write = pipe_write,
.writev = pipe_writev,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_write_open,
.release = pipe_write_release,
.fasync = pipe_write_fasync,
};
const struct file_operations rdwr_fifo_fops = {
.llseek = no_llseek,
.read = pipe_read,
.readv = pipe_readv,
.write = pipe_write,
.writev = pipe_writev,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_rdwr_open,
.release = pipe_rdwr_release,
.fasync = pipe_rdwr_fasync,
};
static struct file_operations read_pipe_fops = {
.llseek = no_llseek,
.read = pipe_read,
.readv = pipe_readv,
.write = bad_pipe_w,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_read_open,
.release = pipe_read_release,
.fasync = pipe_read_fasync,
};
static struct file_operations write_pipe_fops = {
.llseek = no_llseek,
.read = bad_pipe_r,
.write = pipe_write,
.writev = pipe_writev,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_write_open,
.release = pipe_write_release,
.fasync = pipe_write_fasync,
};
static struct file_operations rdwr_pipe_fops = {
.llseek = no_llseek,
.read = pipe_read,
.readv = pipe_readv,
.write = pipe_write,
.writev = pipe_writev,
.poll = pipe_poll,
.ioctl = pipe_ioctl,
.open = pipe_rdwr_open,
.release = pipe_rdwr_release,
.fasync = pipe_rdwr_fasync,
};
struct pipe_inode_info * alloc_pipe_info(struct inode *inode)
{
struct pipe_inode_info *pipe;
pipe = kzalloc(sizeof(struct pipe_inode_info), GFP_KERNEL);
if (pipe) {
init_waitqueue_head(&pipe->wait);
pipe->r_counter = pipe->w_counter = 1;
pipe->inode = inode;
}
return pipe;
}
void __free_pipe_info(struct pipe_inode_info *pipe)
{
int i;
for (i = 0; i < PIPE_BUFFERS; i++) {
struct pipe_buffer *buf = pipe->bufs + i;
if (buf->ops)
buf->ops->release(pipe, buf);
}
if (pipe->tmp_page)
__free_page(pipe->tmp_page);
kfree(pipe);
}
void free_pipe_info(struct inode *inode)
{
__free_pipe_info(inode->i_pipe);
inode->i_pipe = NULL;
}
static struct vfsmount *pipe_mnt __read_mostly;
static int pipefs_delete_dentry(struct dentry *dentry)
{
return 1;
}
static struct dentry_operations pipefs_dentry_operations = {
.d_delete = pipefs_delete_dentry,
};
static struct inode * get_pipe_inode(void)
{
struct inode *inode = new_inode(pipe_mnt->mnt_sb);
struct pipe_inode_info *pipe;
if (!inode)
goto fail_inode;
pipe = alloc_pipe_info(inode);
if (!pipe)
goto fail_iput;
inode->i_pipe = pipe;
pipe->readers = pipe->writers = 1;
inode->i_fop = &rdwr_pipe_fops;
/*
* Mark the inode dirty from the very beginning,
* that way it will never be moved to the dirty
* list because "mark_inode_dirty()" will think
* that it already _is_ on the dirty list.
*/
inode->i_state = I_DIRTY;
inode->i_mode = S_IFIFO | S_IRUSR | S_IWUSR;
inode->i_uid = current->fsuid;
inode->i_gid = current->fsgid;
inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
inode->i_blksize = PAGE_SIZE;
return inode;
fail_iput:
iput(inode);
fail_inode:
return NULL;
}
int do_pipe(int *fd)
{
struct qstr this;
char name[32];
struct dentry *dentry;
struct inode * inode;
struct file *f1, *f2;
int error;
int i, j;
error = -ENFILE;
f1 = get_empty_filp();
if (!f1)
goto no_files;
f2 = get_empty_filp();
if (!f2)
goto close_f1;
inode = get_pipe_inode();
if (!inode)
goto close_f12;
error = get_unused_fd();
if (error < 0)
goto close_f12_inode;
i = error;
error = get_unused_fd();
if (error < 0)
goto close_f12_inode_i;
j = error;
error = -ENOMEM;
sprintf(name, "[%lu]", inode->i_ino);
this.name = name;
this.len = strlen(name);
this.hash = inode->i_ino; /* will go */
dentry = d_alloc(pipe_mnt->mnt_sb->s_root, &this);
if (!dentry)
goto close_f12_inode_i_j;
dentry->d_op = &pipefs_dentry_operations;
d_add(dentry, inode);
f1->f_vfsmnt = f2->f_vfsmnt = mntget(mntget(pipe_mnt));
f1->f_dentry = f2->f_dentry = dget(dentry);
f1->f_mapping = f2->f_mapping = inode->i_mapping;
/* read file */
f1->f_pos = f2->f_pos = 0;
f1->f_flags = O_RDONLY;
f1->f_op = &read_pipe_fops;
f1->f_mode = FMODE_READ;
f1->f_version = 0;
/* write file */
f2->f_flags = O_WRONLY;
f2->f_op = &write_pipe_fops;
f2->f_mode = FMODE_WRITE;
f2->f_version = 0;
fd_install(i, f1);
fd_install(j, f2);
fd[0] = i;
fd[1] = j;
return 0;
close_f12_inode_i_j:
put_unused_fd(j);
close_f12_inode_i:
put_unused_fd(i);
close_f12_inode:
free_pipe_info(inode);
iput(inode);
close_f12:
put_filp(f2);
close_f1:
put_filp(f1);
no_files:
return error;
}
/*
* pipefs should _never_ be mounted by userland - too much of security hassle,
* no real gain from having the whole whorehouse mounted. So we don't need
* any operations on the root directory. However, we need a non-trivial
* d_name - pipe: will go nicely and kill the special-casing in procfs.
*/
static struct super_block *
pipefs_get_sb(struct file_system_type *fs_type, int flags,
const char *dev_name, void *data)
{
return get_sb_pseudo(fs_type, "pipe:", NULL, PIPEFS_MAGIC);
}
static struct file_system_type pipe_fs_type = {
.name = "pipefs",
.get_sb = pipefs_get_sb,
.kill_sb = kill_anon_super,
};
static int __init init_pipe_fs(void)
{
int err = register_filesystem(&pipe_fs_type);
if (!err) {
pipe_mnt = kern_mount(&pipe_fs_type);
if (IS_ERR(pipe_mnt)) {
err = PTR_ERR(pipe_mnt);
unregister_filesystem(&pipe_fs_type);
}
}
return err;
}
static void __exit exit_pipe_fs(void)
{
unregister_filesystem(&pipe_fs_type);
mntput(pipe_mnt);
}
fs_initcall(init_pipe_fs);
module_exit(exit_pipe_fs);
View Git Blame Copy Permalink