X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=fs%2Fpipe.c;h=68090e84f589305ebe80ee57e723c86494ca6143;hb=5ccac88eeb5659c716af8e695e2943509c80d172;hp=e984beb93a0ea88c436037f6e17c0edfbad37d54;hpb=341b446bc5aa36d1d5b8159c1e66716b5d89024d;p=safe%2Fjmp%2Flinux-2.6 diff --git a/fs/pipe.c b/fs/pipe.c index e984beb..68090e8 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -55,7 +55,8 @@ void pipe_wait(struct pipe_inode_info *pipe) } static int -pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) +pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len, + int atomic) { unsigned long copy; @@ -64,8 +65,13 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) iov++; copy = min_t(unsigned long, len, iov->iov_len); - if (copy_from_user(to, iov->iov_base, copy)) - return -EFAULT; + if (atomic) { + if (__copy_from_user_inatomic(to, iov->iov_base, copy)) + return -EFAULT; + } else { + if (copy_from_user(to, iov->iov_base, copy)) + return -EFAULT; + } to += copy; len -= copy; iov->iov_base += copy; @@ -75,7 +81,8 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len) } static int -pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) +pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len, + int atomic) { unsigned long copy; @@ -84,8 +91,13 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) iov++; copy = min_t(unsigned long, len, iov->iov_len); - if (copy_to_user(iov->iov_base, from, copy)) - return -EFAULT; + if (atomic) { + if (__copy_to_user_inatomic(iov->iov_base, from, copy)) + return -EFAULT; + } else { + if (copy_to_user(iov->iov_base, from, copy)) + return -EFAULT; + } from += copy; len -= copy; iov->iov_base += copy; @@ -94,13 +106,52 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len) return 0; } +/* + * Attempt to pre-fault in the user memory, so we can use atomic copies. + * Returns the number of bytes not faulted in. + */ +static int iov_fault_in_pages_write(struct iovec *iov, unsigned long len) +{ + while (!iov->iov_len) + iov++; + + while (len > 0) { + unsigned long this_len; + + this_len = min_t(unsigned long, len, iov->iov_len); + if (fault_in_pages_writeable(iov->iov_base, this_len)) + break; + + len -= this_len; + iov++; + } + + return len; +} + +/* + * Pre-fault in the user memory, so we can use atomic copies. + */ +static void iov_fault_in_pages_read(struct iovec *iov, unsigned long len) +{ + while (!iov->iov_len) + iov++; + + while (len > 0) { + unsigned long this_len; + + this_len = min_t(unsigned long, len, iov->iov_len); + fault_in_pages_readable(iov->iov_base, this_len); + len -= this_len; + iov++; + } +} + 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 @@ -112,38 +163,66 @@ static void anon_pipe_buf_release(struct pipe_inode_info *pipe, page_cache_release(page); } -static void * anon_pipe_buf_map(struct file *file, struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +void *generic_pipe_buf_map(struct pipe_inode_info *pipe, + struct pipe_buffer *buf, int atomic) { + if (atomic) { + buf->flags |= PIPE_BUF_FLAG_ATOMIC; + return kmap_atomic(buf->page, KM_USER0); + } + return kmap(buf->page); } -static void anon_pipe_buf_unmap(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +void generic_pipe_buf_unmap(struct pipe_inode_info *pipe, + struct pipe_buffer *buf, void *map_data) { - kunmap(buf->page); + if (buf->flags & PIPE_BUF_FLAG_ATOMIC) { + buf->flags &= ~PIPE_BUF_FLAG_ATOMIC; + kunmap_atomic(map_data, KM_USER0); + } else + kunmap(buf->page); } -static int anon_pipe_buf_steal(struct pipe_inode_info *pipe, - struct pipe_buffer *buf) +int generic_pipe_buf_steal(struct pipe_inode_info *pipe, + struct pipe_buffer *buf) +{ + struct page *page = buf->page; + + if (page_count(page) == 1) { + lock_page(page); + return 0; + } + + return 1; +} + +void generic_pipe_buf_get(struct pipe_inode_info *info, struct pipe_buffer *buf) +{ + page_cache_get(buf->page); +} + +int generic_pipe_buf_pin(struct pipe_inode_info *info, struct pipe_buffer *buf) { - buf->flags |= PIPE_BUF_FLAG_STOLEN; return 0; } -static struct pipe_buf_operations anon_pipe_buf_ops = { +static const struct pipe_buf_operations anon_pipe_buf_ops = { .can_merge = 1, - .map = anon_pipe_buf_map, - .unmap = anon_pipe_buf_unmap, + .map = generic_pipe_buf_map, + .unmap = generic_pipe_buf_unmap, + .pin = generic_pipe_buf_pin, .release = anon_pipe_buf_release, - .steal = anon_pipe_buf_steal, + .steal = generic_pipe_buf_steal, + .get = generic_pipe_buf_get, }; static ssize_t -pipe_readv(struct file *filp, const struct iovec *_iov, - unsigned long nr_segs, loff_t *ppos) +pipe_read(struct kiocb *iocb, const struct iovec *_iov, + unsigned long nr_segs, loff_t pos) { - struct inode *inode = filp->f_dentry->d_inode; + struct file *filp = iocb->ki_filp; + struct inode *inode = filp->f_path.dentry->d_inode; struct pipe_inode_info *pipe; int do_wakeup; ssize_t ret; @@ -164,25 +243,36 @@ pipe_readv(struct file *filp, const struct iovec *_iov, if (bufs) { int curbuf = pipe->curbuf; struct pipe_buffer *buf = pipe->bufs + curbuf; - struct pipe_buf_operations *ops = buf->ops; + const struct pipe_buf_operations *ops = buf->ops; void *addr; size_t chars = buf->len; - int error; + int error, atomic; if (chars > total_len) chars = total_len; - addr = ops->map(filp, pipe, buf); - if (IS_ERR(addr)) { + error = ops->pin(pipe, buf); + if (error) { if (!ret) - ret = PTR_ERR(addr); + error = ret; break; } - error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars); - ops->unmap(pipe, buf); + + atomic = !iov_fault_in_pages_write(iov, chars); +redo: + addr = ops->map(pipe, buf, atomic); + error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars, atomic); + ops->unmap(pipe, buf, addr); if (unlikely(error)) { + /* + * Just retry with the slow path if we failed. + */ + if (atomic) { + atomic = 0; + goto redo; + } if (!ret) - ret = -EFAULT; + ret = error; break; } ret += chars; @@ -241,18 +331,11 @@ pipe_readv(struct file *filp, const struct iovec *_iov, } static ssize_t -pipe_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos) +pipe_write(struct kiocb *iocb, const struct iovec *_iov, + unsigned long nr_segs, 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 file *filp = iocb->ki_filp; + struct inode *inode = filp->f_path.dentry->d_inode; struct pipe_inode_info *pipe; ssize_t ret; int do_wakeup; @@ -282,25 +365,32 @@ pipe_writev(struct file *filp, const struct iovec *_iov, int lastbuf = (pipe->curbuf + pipe->nrbufs - 1) & (PIPE_BUFFERS-1); struct pipe_buffer *buf = pipe->bufs + lastbuf; - struct pipe_buf_operations *ops = buf->ops; + const struct pipe_buf_operations *ops = buf->ops; int offset = buf->offset + buf->len; if (ops->can_merge && offset + chars <= PAGE_SIZE) { + int error, atomic = 1; void *addr; - int error; - addr = ops->map(filp, pipe, buf); - if (IS_ERR(addr)) { - error = PTR_ERR(addr); + error = ops->pin(pipe, buf); + if (error) goto out; - } + + iov_fault_in_pages_read(iov, chars); +redo1: + addr = ops->map(pipe, buf, atomic); error = pipe_iov_copy_from_user(offset + addr, iov, - chars); - ops->unmap(pipe, buf); + chars, atomic); + ops->unmap(pipe, buf, addr); ret = error; do_wakeup = 1; - if (error) + if (error) { + if (atomic) { + atomic = 0; + goto redo1; + } goto out; + } buf->len += chars; total_len -= chars; ret = chars; @@ -323,7 +413,8 @@ pipe_writev(struct file *filp, const struct iovec *_iov, int newbuf = (pipe->curbuf + bufs) & (PIPE_BUFFERS-1); struct pipe_buffer *buf = pipe->bufs + newbuf; struct page *page = pipe->tmp_page; - int error; + char *src; + int error, atomic = 1; if (!page) { page = alloc_page(GFP_HIGHUSER); @@ -343,11 +434,27 @@ pipe_writev(struct file *filp, const struct iovec *_iov, if (chars > total_len) chars = total_len; - error = pipe_iov_copy_from_user(kmap(page), iov, chars); - kunmap(page); + iov_fault_in_pages_read(iov, chars); +redo2: + if (atomic) + src = kmap_atomic(page, KM_USER0); + else + src = kmap(page); + + error = pipe_iov_copy_from_user(src, iov, chars, + atomic); + if (atomic) + kunmap_atomic(src, KM_USER0); + else + kunmap(page); + if (unlikely(error)) { + if (atomic) { + atomic = 0; + goto redo2; + } if (!ret) - ret = -EFAULT; + ret = error; break; } ret += chars; @@ -397,15 +504,6 @@ out: } 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; @@ -422,7 +520,7 @@ 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 inode *inode = filp->f_path.dentry->d_inode; struct pipe_inode_info *pipe; int count, buf, nrbufs; @@ -450,7 +548,7 @@ static unsigned int pipe_poll(struct file *filp, poll_table *wait) { unsigned int mask; - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = filp->f_path.dentry->d_inode; struct pipe_inode_info *pipe = inode->i_pipe; int nrbufs; @@ -503,7 +601,7 @@ pipe_release(struct inode *inode, int decr, int decw) static int pipe_read_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = filp->f_path.dentry->d_inode; int retval; mutex_lock(&inode->i_mutex); @@ -520,7 +618,7 @@ pipe_read_fasync(int fd, struct file *filp, int on) static int pipe_write_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = filp->f_path.dentry->d_inode; int retval; mutex_lock(&inode->i_mutex); @@ -537,7 +635,7 @@ pipe_write_fasync(int fd, struct file *filp, int on) static int pipe_rdwr_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = filp->f_path.dentry->d_inode; struct pipe_inode_info *pipe = inode->i_pipe; int retval; @@ -623,8 +721,8 @@ pipe_rdwr_open(struct inode *inode, struct file *filp) */ const struct file_operations read_fifo_fops = { .llseek = no_llseek, - .read = pipe_read, - .readv = pipe_readv, + .read = do_sync_read, + .aio_read = pipe_read, .write = bad_pipe_w, .poll = pipe_poll, .ioctl = pipe_ioctl, @@ -636,8 +734,8 @@ const struct file_operations read_fifo_fops = { const struct file_operations write_fifo_fops = { .llseek = no_llseek, .read = bad_pipe_r, - .write = pipe_write, - .writev = pipe_writev, + .write = do_sync_write, + .aio_write = pipe_write, .poll = pipe_poll, .ioctl = pipe_ioctl, .open = pipe_write_open, @@ -647,10 +745,10 @@ const struct file_operations write_fifo_fops = { const struct file_operations rdwr_fifo_fops = { .llseek = no_llseek, - .read = pipe_read, - .readv = pipe_readv, - .write = pipe_write, - .writev = pipe_writev, + .read = do_sync_read, + .aio_read = pipe_read, + .write = do_sync_write, + .aio_write = pipe_write, .poll = pipe_poll, .ioctl = pipe_ioctl, .open = pipe_rdwr_open, @@ -658,10 +756,10 @@ const struct file_operations rdwr_fifo_fops = { .fasync = pipe_rdwr_fasync, }; -static struct file_operations read_pipe_fops = { +static const struct file_operations read_pipe_fops = { .llseek = no_llseek, - .read = pipe_read, - .readv = pipe_readv, + .read = do_sync_read, + .aio_read = pipe_read, .write = bad_pipe_w, .poll = pipe_poll, .ioctl = pipe_ioctl, @@ -670,11 +768,11 @@ static struct file_operations read_pipe_fops = { .fasync = pipe_read_fasync, }; -static struct file_operations write_pipe_fops = { +static const struct file_operations write_pipe_fops = { .llseek = no_llseek, .read = bad_pipe_r, - .write = pipe_write, - .writev = pipe_writev, + .write = do_sync_write, + .aio_write = pipe_write, .poll = pipe_poll, .ioctl = pipe_ioctl, .open = pipe_write_open, @@ -682,12 +780,12 @@ static struct file_operations write_pipe_fops = { .fasync = pipe_write_fasync, }; -static struct file_operations rdwr_pipe_fops = { +static const struct file_operations rdwr_pipe_fops = { .llseek = no_llseek, - .read = pipe_read, - .readv = pipe_readv, - .write = pipe_write, - .writev = pipe_writev, + .read = do_sync_read, + .aio_read = pipe_read, + .write = do_sync_write, + .aio_write = pipe_write, .poll = pipe_poll, .ioctl = pipe_ioctl, .open = pipe_rdwr_open, @@ -732,7 +830,14 @@ void free_pipe_info(struct inode *inode) static struct vfsmount *pipe_mnt __read_mostly; static int pipefs_delete_dentry(struct dentry *dentry) { - return 1; + /* + * At creation time, we pretended this dentry was hashed + * (by clearing DCACHE_UNHASHED bit in d_flags) + * At delete time, we restore the truth : not hashed. + * (so that dput() can proceed correctly) + */ + dentry->d_flags |= DCACHE_UNHASHED; + return 0; } static struct dentry_operations pipefs_dentry_operations = { @@ -766,7 +871,6 @@ static struct inode * get_pipe_inode(void) 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; @@ -777,87 +881,126 @@ fail_inode: return NULL; } -int do_pipe(int *fd) +struct file *create_write_pipe(void) { - struct qstr this; - char name[32]; + int err; + struct inode *inode; + struct file *f; 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; + char name[32]; + struct qstr this; + f = get_empty_filp(); + if (!f) + return ERR_PTR(-ENFILE); + err = -ENFILE; inode = get_pipe_inode(); if (!inode) - goto close_f12; + goto err_file; - 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.len = sprintf(name, "[%lu]", inode->i_ino); this.name = name; - this.len = strlen(name); - this.hash = inode->i_ino; /* will go */ + this.hash = 0; + err = -ENOMEM; dentry = d_alloc(pipe_mnt->mnt_sb->s_root, &this); if (!dentry) - goto close_f12_inode_i_j; + goto err_inode; 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; + /* + * We dont want to publish this dentry into global dentry hash table. + * We pretend dentry is already hashed, by unsetting DCACHE_UNHASHED + * This permits a working /proc/$pid/fd/XXX on pipes + */ + dentry->d_flags &= ~DCACHE_UNHASHED; + d_instantiate(dentry, inode); + f->f_path.mnt = mntget(pipe_mnt); + f->f_path.dentry = dentry; + f->f_mapping = inode->i_mapping; - return 0; + f->f_flags = O_WRONLY; + f->f_op = &write_pipe_fops; + f->f_mode = FMODE_WRITE; + f->f_version = 0; -close_f12_inode_i_j: - put_unused_fd(j); -close_f12_inode_i: - put_unused_fd(i); -close_f12_inode: + return f; + + err_inode: free_pipe_info(inode); iput(inode); -close_f12: - put_filp(f2); -close_f1: - put_filp(f1); -no_files: - return error; + err_file: + put_filp(f); + return ERR_PTR(err); +} + +void free_write_pipe(struct file *f) +{ + free_pipe_info(f->f_dentry->d_inode); + dput(f->f_path.dentry); + mntput(f->f_path.mnt); + put_filp(f); +} + +struct file *create_read_pipe(struct file *wrf) +{ + struct file *f = get_empty_filp(); + if (!f) + return ERR_PTR(-ENFILE); + + /* Grab pipe from the writer */ + f->f_path.mnt = mntget(wrf->f_path.mnt); + f->f_path.dentry = dget(wrf->f_path.dentry); + f->f_mapping = wrf->f_path.dentry->d_inode->i_mapping; + + f->f_pos = 0; + f->f_flags = O_RDONLY; + f->f_op = &read_pipe_fops; + f->f_mode = FMODE_READ; + f->f_version = 0; + + return f; +} + +int do_pipe(int *fd) +{ + struct file *fw, *fr; + int error; + int fdw, fdr; + + fw = create_write_pipe(); + if (IS_ERR(fw)) + return PTR_ERR(fw); + fr = create_read_pipe(fw); + error = PTR_ERR(fr); + if (IS_ERR(fr)) + goto err_write_pipe; + + error = get_unused_fd(); + if (error < 0) + goto err_read_pipe; + fdr = error; + + error = get_unused_fd(); + if (error < 0) + goto err_fdr; + fdw = error; + + fd_install(fdr, fr); + fd_install(fdw, fw); + fd[0] = fdr; + fd[1] = fdw; + + return 0; + + err_fdr: + put_unused_fd(fdr); + err_read_pipe: + dput(fr->f_dentry); + mntput(fr->f_vfsmnt); + put_filp(fr); + err_write_pipe: + free_write_pipe(fw); + return error; } /* @@ -866,12 +1009,11 @@ no_files: * 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) +static int pipefs_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data, + struct vfsmount *mnt) { - return get_sb_pseudo(fs_type, "pipe:", NULL, PIPEFS_MAGIC); + return get_sb_pseudo(fs_type, "pipe:", NULL, PIPEFS_MAGIC, mnt); } static struct file_system_type pipe_fs_type = {