struct inode* pipe_new(struct inode* inode)
{
unsigned long page;
page = __get_free_page(GFP_USER);
if (!page)
return NULL;
inode->i_pipe = kmalloc(sizeof(struct pipe_inode_info), GFP_KERNEL);
if (!inode->i_pipe)
goto fail_page;
init_waitqueue_head(PIPE_WAIT(*inode));
PIPE_BASE(*inode) = (char*) page;
PIPE_START(*inode) = PIPE_LEN(*inode) = 0;
PIPE_READERS(*inode) = PIPE_WRITERS(*inode) = 0;
PIPE_WAITING_WRITERS(*inode) = 0;
PIPE_RCOUNTER(*inode) = PIPE_WCOUNTER(*inode) = 1;
*PIPE_FASYNC_READERS(*inode) = *PIPE_FASYNC_WRITERS(*inode) = NULL;
return inode;
fail_page:
free_page(page);
return NULL;
}
static int
pipe_write_fasync(int fd, struct file *filp, int on)
{
struct inode *inode = filp->f_dentry->d_inode;
int retval;
down(PIPE_SEM(*inode));
retval = fasync_helper(fd, filp, on, PIPE_FASYNC_WRITERS(*inode));
up(PIPE_SEM(*inode));
if (retval < 0)
return retval;
return 0;
}
static int
pipe_release(struct inode *inode, int decr, int decw)
{
down(PIPE_SEM(*inode));
PIPE_READERS(*inode) -= decr;
PIPE_WRITERS(*inode) -= decw;
if (!PIPE_READERS(*inode) && !PIPE_WRITERS(*inode)) {
struct pipe_inode_info *info = inode->i_pipe;
inode->i_pipe = NULL;
free_page((unsigned long) info->base);
kfree(info);
} else {
wake_up_interruptible(PIPE_WAIT(*inode));
kill_fasync(PIPE_FASYNC_READERS(*inode), SIGIO, POLL_IN);
kill_fasync(PIPE_FASYNC_WRITERS(*inode), SIGIO, POLL_OUT);
}
up(PIPE_SEM(*inode));
return 0;
}
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;
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;
down(PIPE_SEM(*inode));
for (;;) {
int size = PIPE_LEN(*inode);
if (size) {
char *pipebuf = PIPE_BASE(*inode) + PIPE_START(*inode);
ssize_t chars = PIPE_MAX_RCHUNK(*inode);
if (chars > total_len)
chars = total_len;
if (chars > size)
chars = size;
if (pipe_iov_copy_to_user(iov, pipebuf, chars)) {
if (!ret) ret = -EFAULT;
break;
}
ret += chars;
PIPE_START(*inode) += chars;
PIPE_START(*inode) &= (PIPE_SIZE - 1);
PIPE_LEN(*inode) -= chars;
total_len -= chars;
do_wakeup = 1;
if (!total_len)
break; /* common path: read succeeded */
}
if (PIPE_LEN(*inode)) /* test for cyclic buffers */
continue;
if (!PIPE_WRITERS(*inode))
break;
if (!PIPE_WAITING_WRITERS(*inode)) {
/* 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(*inode));
kill_fasync(PIPE_FASYNC_WRITERS(*inode), SIGIO, POLL_OUT);
}
pipe_wait(inode);
}
up(PIPE_SEM(*inode));
/* Signal writers asynchronously that there is more room. */
if (do_wakeup) {
wake_up_interruptible(PIPE_WAIT(*inode));
kill_fasync(PIPE_FASYNC_WRITERS(*inode), SIGIO, POLL_OUT);
}
/*
* Hack: we turn off atime updates for -RT kernels.
* Who uses them on pipes anyway?
*/
#ifndef CONFIG_PREEMPT_RT
if (ret > 0)
file_accessed(filp);
#endif
return ret;
}
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 *info;
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;
down(PIPE_SEM(*inode));
info = inode->i_pipe;
for (;;) {
int bufs = info->nrbufs;
if (bufs) {
int curbuf = info->curbuf;
struct pipe_buffer *buf = info->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, info, buf);
error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars);
ops->unmap(info, 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(info, buf);
curbuf = (curbuf + 1) & (PIPE_BUFFERS-1);
info->curbuf = curbuf;
info->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(*inode))
break;
if (!PIPE_WAITING_WRITERS(*inode)) {
/* 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(*inode));
kill_fasync(PIPE_FASYNC_WRITERS(*inode), SIGIO, POLL_OUT);
}
pipe_wait(inode);
}
up(PIPE_SEM(*inode));
/* Signal writers asynchronously that there is more room. */
if (do_wakeup) {
wake_up_interruptible(PIPE_WAIT(*inode));
kill_fasync(PIPE_FASYNC_WRITERS(*inode), SIGIO, POLL_OUT);
}
if (ret > 0)
file_accessed(filp);
return ret;
}