switch generic_write_checks() to iocb and iter

... returning -E... upon error and amount of data left in iter after
(possible) truncation upon success. Note, that normal case gives
a non-zero (positive) return value, so any tests for != 0 _must_ be
updated.

Signed-off-by: Al Viro <[e-mail]>

Conflicts:
fs/ext4/file.c

fs/9p/vfs_file.c

···
 v9fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 {
 	struct file *file = iocb->ki_filp;
-	ssize_t retval = 0;
-	loff_t origin = iocb->ki_pos;
-	size_t count = iov_iter_count(from);
+	ssize_t retval;
+	loff_t origin;
 	int err = 0;
 
-	retval = generic_write_checks(file, &origin, &count);
-	if (retval)
+	retval = generic_write_checks(iocb, from);
+	if (retval <= 0)
 		return retval;
 
-	iov_iter_truncate(from, count);
-
-	if (!count)
-		return 0;
-
-	retval = p9_client_write(file->private_data, origin, from, &err);
+	origin = iocb->ki_pos;
+	retval = p9_client_write(file->private_data, iocb->ki_pos, from, &err);
 	if (retval > 0) {
 		struct inode *inode = file_inode(file);
 		loff_t i_size;
···
 		if (inode->i_mapping && inode->i_mapping->nrpages)
 			invalidate_inode_pages2_range(inode->i_mapping,
 						      pg_start, pg_end);
-		origin += retval;
+		iocb->ki_pos += retval;
 		i_size = i_size_read(inode);
-		iocb->ki_pos = origin;
-		if (origin > i_size) {
-			inode_add_bytes(inode, origin - i_size);
-			i_size_write(inode, origin);
+		if (iocb->ki_pos > i_size) {
+			inode_add_bytes(inode, iocb->ki_pos - i_size);
+			i_size_write(inode, iocb->ki_pos);
 		}
 		return retval;
 	}

fs/btrfs/file.c

···
 	u64 start_pos;
 	u64 end_pos;
 	ssize_t num_written = 0;
-	ssize_t err = 0;
-	size_t count = iov_iter_count(from);
 	bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);
-	loff_t pos = iocb->ki_pos;
+	ssize_t err;
+	loff_t pos;
+	size_t count;
 
 	mutex_lock(&inode->i_mutex);
+	err = generic_write_checks(iocb, from);
+	if (err <= 0) {
+		mutex_unlock(&inode->i_mutex);
+		return err;
+	}
 
 	current->backing_dev_info = inode_to_bdi(inode);
-	err = generic_write_checks(file, &pos, &count);
-	if (err) {
-		mutex_unlock(&inode->i_mutex);
-		goto out;
-	}
-
-	if (count == 0) {
-		mutex_unlock(&inode->i_mutex);
-		goto out;
-	}
-
-	iov_iter_truncate(from, count);
-
 	err = file_remove_suid(file);
 	if (err) {
 		mutex_unlock(&inode->i_mutex);
···
 	 */
 	update_time_for_write(inode);
 
+	pos = iocb->ki_pos;
+	count = iov_iter_count(from);
 	start_pos = round_down(pos, root->sectorsize);
 	if (start_pos > i_size_read(inode)) {
 		/* Expand hole size to cover write data, preventing empty gap */

fs/ceph/file.c

···
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_osd_client *osdc =
 		&ceph_sb_to_client(inode->i_sb)->client->osdc;
-	ssize_t count = iov_iter_count(from), written = 0;
+	ssize_t count, written = 0;
 	int err, want, got;
-	loff_t pos = iocb->ki_pos;
+	loff_t pos;
 
 	if (ceph_snap(inode) != CEPH_NOSNAP)
 		return -EROFS;
···
 	/* We can write back this queue in page reclaim */
 	current->backing_dev_info = inode_to_bdi(inode);
 
-	err = generic_write_checks(file, &pos, &count);
-	if (err)
+	err = generic_write_checks(iocb, from);
+	if (err <= 0)
 		goto out;
 
-	if (count == 0)
-		goto out;
-	iov_iter_truncate(from, count);
-
+	pos = iocb->ki_pos;
+	count = iov_iter_count(from);
 	err = file_remove_suid(file);
 	if (err)
 		goto out;

fs/cifs/file.c

···
 ssize_t cifs_user_writev(struct kiocb *iocb, struct iov_iter *from)
 {
 	struct file *file = iocb->ki_filp;
-	size_t len;
 	ssize_t total_written = 0;
 	struct cifsFileInfo *open_file;
 	struct cifs_tcon *tcon;
···
 	 * write request.
 	 */
 
-	len = iov_iter_count(from);
-	rc = generic_write_checks(file, &iocb->ki_pos, &len);
-	if (rc)
+	rc = generic_write_checks(iocb, from);
+	if (rc <= 0)
 		return rc;
-
-	if (!len)
-		return 0;
-
-	iov_iter_truncate(from, len);
 
 	INIT_LIST_HEAD(&wdata_list);
 	cifs_sb = CIFS_FILE_SB(file);
···
 
 	memcpy(&saved_from, from, sizeof(struct iov_iter));
 
-	rc = cifs_write_from_iter(iocb->ki_pos, len, from, open_file, cifs_sb,
-				  &wdata_list);
+	rc = cifs_write_from_iter(iocb->ki_pos, iov_iter_count(from), from,
+				  open_file, cifs_sb, &wdata_list);
 
 	/*
 	 * If at least one write was successfully sent, then discard any rc
···
 	struct cifsInodeInfo *cinode = CIFS_I(inode);
 	struct TCP_Server_Info *server = tlink_tcon(cfile->tlink)->ses->server;
 	ssize_t rc;
-	size_t count;
 
 	/*
 	 * We need to hold the sem to be sure nobody modifies lock list
···
 	down_read(&cinode->lock_sem);
 	mutex_lock(&inode->i_mutex);
 
-	count = iov_iter_count(from);
-	rc = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (rc)
+	rc = generic_write_checks(iocb, from);
+	if (rc <= 0)
 		goto out;
-
-	if (count == 0)
-		goto out;
-
-	iov_iter_truncate(from, count);
 
 	if (!cifs_find_lock_conflict(cfile, iocb->ki_pos, iov_iter_count(from),
 				     server->vals->exclusive_lock_type, NULL,

fs/ext4/file.c

···
 	struct blk_plug plug;
 	int o_direct = io_is_direct(file);
 	int overwrite = 0;
-	size_t length = iov_iter_count(from);
 	ssize_t ret;
-	loff_t pos;
 
 	/*
 	 * Unaligned direct AIO must be serialized; see comment above
···
 	}
 
 	mutex_lock(&inode->i_mutex);
-	ret = generic_write_checks(file, &iocb->ki_pos, &length);
-	if (ret)
+	ret = generic_write_checks(iocb, from);
+	if (ret <= 0)
 		goto out;
-
-	if (length == 0)
-		goto out;
-
-	iov_iter_truncate(from, length);
-	pos = iocb->ki_pos;
 
 	/*
 	 * If we have encountered a bitmap-format file, the size limit
···
 	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 		struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 
-		if (pos >= sbi->s_bitmap_maxbytes) {
+		if (iocb->ki_pos >= sbi->s_bitmap_maxbytes) {
 			ret = -EFBIG;
 			goto out;
 		}
-		iov_iter_truncate(from, sbi->s_bitmap_maxbytes - pos);
+		iov_iter_truncate(from, sbi->s_bitmap_maxbytes - iocb->ki_pos);
 	}
 
 	iocb->private = &overwrite;
 	if (o_direct) {
-		length = iov_iter_count(from);
+		size_t length = iov_iter_count(from);
+		loff_t pos = iocb->ki_pos;
 		blk_start_plug(&plug);
-
 
 		/* check whether we do a DIO overwrite or not */
 		if (ext4_should_dioread_nolock(inode) && !aio_mutex &&

fs/fuse/file.c

···
 {
 	struct file *file = iocb->ki_filp;
 	struct address_space *mapping = file->f_mapping;
-	size_t count = iov_iter_count(from);
 	ssize_t written = 0;
 	ssize_t written_buffered = 0;
 	struct inode *inode = mapping->host;
 	ssize_t err;
 	loff_t endbyte = 0;
-	loff_t pos = iocb->ki_pos;
 
 	if (get_fuse_conn(inode)->writeback_cache) {
 		/* Update size (EOF optimization) and mode (SUID clearing) */
···
 	/* We can write back this queue in page reclaim */
 	current->backing_dev_info = inode_to_bdi(inode);
 
-	err = generic_write_checks(file, &pos, &count);
-	if (err)
+	err = generic_write_checks(iocb, from);
+	if (err <= 0)
 		goto out;
 
-	if (count == 0)
-		goto out;
-
-	iov_iter_truncate(from, count);
 	err = file_remove_suid(file);
 	if (err)
 		goto out;
···
 		goto out;
 
 	if (file->f_flags & O_DIRECT) {
+		loff_t pos = iocb->ki_pos;
 		written = generic_file_direct_write(iocb, from, pos);
 		if (written < 0 || !iov_iter_count(from))
 			goto out;
···
 		written += written_buffered;
 		iocb->ki_pos = pos + written_buffered;
 	} else {
-		written = fuse_perform_write(file, mapping, from, pos);
+		written = fuse_perform_write(file, mapping, from, iocb->ki_pos);
 		if (written >= 0)
-			iocb->ki_pos = pos + written;
+			iocb->ki_pos += written;
 	}
 out:
 	current->backing_dev_info = NULL;
···
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file_inode(file);
 	struct fuse_io_priv io = { .async = 0, .file = file };
-	size_t count = iov_iter_count(from);
 	ssize_t res;
 
 	if (is_bad_inode(inode))
···
 
 	/* Don't allow parallel writes to the same file */
 	mutex_lock(&inode->i_mutex);
-	res = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (!res) {
-		iov_iter_truncate(from, count);
+	res = generic_write_checks(iocb, from);
+	if (res > 0)
 		res = fuse_direct_io(&io, from, &iocb->ki_pos, FUSE_DIO_WRITE);
-	}
 	fuse_invalidate_attr(inode);
 	if (res > 0)
 		fuse_write_update_size(inode, iocb->ki_pos);

fs/ncpfs/file.c

···
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file_inode(file);
 	size_t already_written = 0;
-	loff_t pos = iocb->ki_pos;
-	size_t count = iov_iter_count(from);
 	size_t bufsize;
 	int errno;
 	void *bouncebuffer;
+	off_t pos;
 
 	ncp_dbg(1, "enter %pD2\n", file);
-	errno = generic_write_checks(file, &pos, &count);
-	if (errno)
+	errno = generic_write_checks(iocb, from);
+	if (errno <= 0)
 		return errno;
-	iov_iter_truncate(from, count);
-	
-	if (!count)
-		return 0;
 
 	errno = ncp_make_open(inode, O_WRONLY);
 	if (errno) {
···
 		errno = -EIO;	/* -ENOMEM */
 		goto outrel;
 	}
+	pos = iocb->ki_pos;
 	while (iov_iter_count(from)) {
 		int written_this_time;
 		size_t to_write = min_t(size_t,
-				      bufsize - ((off_t)pos % bufsize),
+				      bufsize - (pos % bufsize),
 				      iov_iter_count(from));
 
 		if (copy_from_iter(bouncebuffer, to_write, from) != to_write) {

fs/nfs/direct.c

···
 	struct nfs_direct_req *dreq;
 	struct nfs_lock_context *l_ctx;
 	loff_t end;
-	size_t count = iov_iter_count(iter);
-	end = (pos + count - 1) >> PAGE_CACHE_SHIFT;
-
-	nfs_add_stats(mapping->host, NFSIOS_DIRECTWRITTENBYTES, count);
 
 	dfprintk(FILE, "NFS: direct write(%pD2, %zd@%Ld)\n",
-		file, count, (long long) pos);
+		file, iov_iter_count(iter), (long long) iocb->ki_pos);
 
-	result = generic_write_checks(file, &pos, &count);
-	if (result)
+	result = generic_write_checks(iocb, iter);
+	if (result <= 0)
 		goto out;
 
-	result = -EINVAL;
-	if ((ssize_t) count < 0)
-		goto out;
-	result = 0;
-	if (!count)
-		goto out;
+	nfs_add_stats(mapping->host, NFSIOS_DIRECTWRITTENBYTES,
+		      iov_iter_count(iter));
+
+	pos = iocb->ki_pos;
+	end = (pos + iov_iter_count(iter) - 1) >> PAGE_CACHE_SHIFT;
 
 	mutex_lock(&inode->i_mutex);
 
···
 			goto out_unlock;
 	}
 
-	task_io_account_write(count);
+	task_io_account_write(iov_iter_count(iter));
 
 	result = -ENOMEM;
 	dreq = nfs_direct_req_alloc();
···
 		goto out_unlock;
 
 	dreq->inode = inode;
-	dreq->bytes_left = count;
+	dreq->bytes_left = iov_iter_count(iter);
 	dreq->io_start = pos;
 	dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp));
 	l_ctx = nfs_get_lock_context(dreq->ctx);

fs/ntfs/file.c

···
 	struct inode *vi = file_inode(file);
 	ntfs_inode *base_ni, *ni = NTFS_I(vi);
 	ntfs_volume *vol = ni->vol;
-	size_t count = iov_iter_count(from);
 
 	ntfs_debug("Entering for i_ino 0x%lx, attribute type 0x%x, pos "
 			"0x%llx, count 0x%zx.", vi->i_ino,
 			(unsigned)le32_to_cpu(ni->type),
-			(unsigned long long)iocb->ki_pos, count);
-	err = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (unlikely(err))
-		goto out;
-	iov_iter_truncate(from, count);
-	if (count == 0)
+			(unsigned long long)iocb->ki_pos,
+			iov_iter_count(from));
+	err = generic_write_checks(iocb, from);
+	if (unlikely(err <= 0))
 		goto out;
 	/*
 	 * All checks have passed.  Before we start doing any writing we want

fs/ocfs2/file.c

···
 static ssize_t ocfs2_file_write_iter(struct kiocb *iocb,
 				    struct iov_iter *from)
 {
-	int ret, direct_io, appending, rw_level, have_alloc_sem  = 0;
+	int direct_io, appending, rw_level, have_alloc_sem  = 0;
 	int can_do_direct, has_refcount = 0;
 	ssize_t written = 0;
+	ssize_t ret;
 	size_t count = iov_iter_count(from), orig_count;
 	loff_t old_size;
 	u32 old_clusters;
···
 		ocfs2_inode_unlock(inode, 1);
 	}
 
-	orig_count = count;
-	ret = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (ret < 0) {
-		mlog_errno(ret);
+	orig_count = iov_iter_count(from);
+	ret = generic_write_checks(iocb, from);
+	if (ret <= 0) {
+		if (ret)
+			mlog_errno(ret);
 		goto out;
 	}
-	iov_iter_truncate(from, count);
+	count = ret;
 
 	can_do_direct = direct_io;
 	ret = ocfs2_prepare_inode_for_write(file, iocb->ki_pos, count, appending,
···
 		rw_level = -1;
 
 		direct_io = 0;
-		iov_iter_reexpand(from, count = orig_count);
+		iov_iter_reexpand(from, orig_count);
 		goto relock;
 	}
 
···
 		loff_t endbyte;
 		ssize_t written_buffered;
 		written = generic_file_direct_write(iocb, from, iocb->ki_pos);
-		if (written < 0 || written == count) {
+		if (written < 0 || !iov_iter_count(from)) {
 			ret = written;
 			goto out_dio;
 		}
···
 		/*
 		 * for completing the rest of the request.
 		 */
-		count -= written;
 		written_buffered = generic_perform_write(file, from, iocb->ki_pos);
 		/*
 		 * If generic_file_buffered_write() returned a synchronous error

fs/udf/file.c

···
 	ssize_t retval;
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file_inode(file);
-	size_t count = iov_iter_count(from);
 	struct udf_inode_info *iinfo = UDF_I(inode);
 	int err;
 
 	mutex_lock(&inode->i_mutex);
 
-	retval = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (retval)
+	retval = generic_write_checks(iocb, from);
+	if (retval <= 0)
 		goto out;
-
-	if (count == 0)
-		goto out;
-
-	iov_iter_truncate(from, count);
 
 	down_write(&iinfo->i_data_sem);
 	if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) {

fs/xfs/xfs_file.c

···
 	struct file		*file = iocb->ki_filp;
 	struct inode		*inode = file->f_mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
-	int			error = 0;
+	ssize_t			error = 0;
 	size_t			count = iov_iter_count(from);
 
 restart:
-	error = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (error)
+	error = generic_write_checks(iocb, from);
+	if (error <= 0)
 		return error;
 
 	error = xfs_break_layouts(inode, iolock);
···
 			xfs_rw_iunlock(ip, *iolock);
 			*iolock = XFS_IOLOCK_EXCL;
 			xfs_rw_ilock(ip, *iolock);
+			iov_iter_reexpand(from, count);
 			goto restart;
 		}
 		error = xfs_zero_eof(ip, iocb->ki_pos, i_size_read(inode), &zero);
 		if (error)
 			return error;
 	}
-	iov_iter_truncate(from, count);
 
 	/*
 	 * Updating the timestamps will grab the ilock again from

include/linux/fs.h

···
 
 extern int generic_file_mmap(struct file *, struct vm_area_struct *);
 extern int generic_file_readonly_mmap(struct file *, struct vm_area_struct *);
-int generic_write_checks(struct file *file, loff_t *pos, size_t *count);
+extern ssize_t generic_write_checks(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_read_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t __generic_file_write_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_write_iter(struct kiocb *, struct iov_iter *);

mm/filemap.c

···
  * Returns appropriate error code that caller should return or
  * zero in case that write should be allowed.
  */
-inline int generic_write_checks(struct file *file, loff_t *pos, size_t *count)
+inline ssize_t generic_write_checks(struct kiocb *iocb, struct iov_iter *from)
 {
+	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
 	unsigned long limit = rlimit(RLIMIT_FSIZE);
+	loff_t pos;
 
-        if (unlikely(*pos < 0))
-                return -EINVAL;
+	if (!iov_iter_count(from))
+		return 0;
 
 	/* FIXME: this is for backwards compatibility with 2.4 */
 	if (file->f_flags & O_APPEND)
-		*pos = i_size_read(inode);
+		iocb->ki_pos = i_size_read(inode);
+
+	pos = iocb->ki_pos;
 
 	if (limit != RLIM_INFINITY) {
-		if (*pos >= limit) {
+		if (iocb->ki_pos >= limit) {
 			send_sig(SIGXFSZ, current, 0);
 			return -EFBIG;
 		}
-		if (*count > limit - (typeof(limit))*pos)
-			*count = limit - (typeof(limit))*pos;
+		iov_iter_truncate(from, limit - (unsigned long)pos);
 	}
 
 	/*
 	 * LFS rule
 	 */
-	if (unlikely(*pos + *count > MAX_NON_LFS &&
+	if (unlikely(pos + iov_iter_count(from) > MAX_NON_LFS &&
 				!(file->f_flags & O_LARGEFILE))) {
-		if (*pos >= MAX_NON_LFS)
+		if (pos >= MAX_NON_LFS)
 			return -EFBIG;
-		if (*count > MAX_NON_LFS - (unsigned long)*pos)
-			*count = MAX_NON_LFS - (unsigned long)*pos;
+		iov_iter_truncate(from, MAX_NON_LFS - (unsigned long)pos);
 	}
 
 	/*
···
 	 * exceeded without writing data we send a signal and return EFBIG.
 	 * Linus frestrict idea will clean these up nicely..
 	 */
-	if (unlikely(*pos >= inode->i_sb->s_maxbytes)) {
-		if (*count || *pos > inode->i_sb->s_maxbytes) {
-			return -EFBIG;
-		}
-		/* zero-length writes at ->s_maxbytes are OK */
-	}
+	if (unlikely(pos >= inode->i_sb->s_maxbytes))
+		return -EFBIG;
 
-	if (unlikely(*pos + *count > inode->i_sb->s_maxbytes))
-		*count = inode->i_sb->s_maxbytes - *pos;
-	return 0;
+	iov_iter_truncate(from, inode->i_sb->s_maxbytes - pos);
+	return iov_iter_count(from);
 }
 EXPORT_SYMBOL(generic_write_checks);
 
···
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
-	size_t count = iov_iter_count(from);
 
 	mutex_lock(&inode->i_mutex);
-	ret = generic_write_checks(file, &iocb->ki_pos, &count);
-	if (!ret && count) {
-		iov_iter_truncate(from, count);
+	ret = generic_write_checks(iocb, from);
+	if (ret > 0)
 		ret = __generic_file_write_iter(iocb, from);
-	}
 	mutex_unlock(&inode->i_mutex);
 
 	if (ret > 0) {