Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 9ea7df53 authored by Jan Kara's avatar Jan Kara Committed by Al Viro
Browse files

ext4: Rewrite ext4_page_mkwrite() to use generic helpers 



Rewrite ext4_page_mkwrite() to use __block_page_mkwrite() helper. This
removes the need of using i_alloc_sem to avoid races with truncate which
seems to be the wrong locking order according to lock ordering documented in
mm/rmap.c. Also calling ext4_da_write_begin() as used by the old code seems to
be problematic because we can decide to flush delay-allocated blocks which
will acquire s_umount semaphore - again creating unpleasant lock dependency
if not directly a deadlock.

Also add a check for frozen filesystem so that we don't busyloop in page fault
when the filesystem is frozen.

Signed-off-by: default avatarJan Kara <jack@suse.cz>
Signed-off-by: default avatarAl Viro <viro@zeniv.linux.org.uk>
parent 58268691

fs/ext4/inode.c

+55 −51
Original line number Original line Diff line number Diff line
@@ -5843,80 +5843,84 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) 
	struct page *page = vmf->page;

	struct page *page = vmf->page;

	loff_t size;

	loff_t size;

	unsigned long len;

	unsigned long len;

	int ret = -EINVAL;

	int ret;

	void *fsdata;

	struct file *file = vma->vm_file;

	struct file *file = vma->vm_file;

	struct inode *inode = file->f_path.dentry->d_inode;

	struct inode *inode = file->f_path.dentry->d_inode;

	struct address_space *mapping = inode->i_mapping;

	struct address_space *mapping = inode->i_mapping;

	handle_t *handle;

	get_block_t *get_block;

	int retries = 0;





	/*

	/*

	 * Get i_alloc_sem to stop truncates messing with the inode. We cannot

	 * This check is racy but catches the common case. We rely on

	 * get i_mutex because we are already holding mmap_sem.

	 * __block_page_mkwrite() to do a reliable check.

	 */

	 */

	down_read(&inode->i_alloc_sem);

	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);

	size = i_size_read(inode);

	/* Delalloc case is easy... */

	if (page->mapping != mapping || size <= page_offset(page)

	if (test_opt(inode->i_sb, DELALLOC) &&

	    || !PageUptodate(page)) {

	    !ext4_should_journal_data(inode) &&

		/* page got truncated from under us? */

	    !ext4_nonda_switch(inode->i_sb)) {

		goto out_unlock;

		do {

			ret = __block_page_mkwrite(vma, vmf,

						   ext4_da_get_block_prep);

		} while (ret == -ENOSPC &&

		       ext4_should_retry_alloc(inode->i_sb, &retries));

		goto out_ret;

	}

	}

	ret = 0;





	lock_page(page);

	lock_page(page);

	wait_on_page_writeback(page);

	size = i_size_read(inode);

	if (PageMappedToDisk(page)) {

	/* Page got truncated from under us? */

		up_read(&inode->i_alloc_sem);

	if (page->mapping != mapping || page_offset(page) > size) {

		return VM_FAULT_LOCKED;

		unlock_page(page);

		ret = VM_FAULT_NOPAGE;

		goto out;

	}

	}





	if (page->index == size >> PAGE_CACHE_SHIFT)

	if (page->index == size >> PAGE_CACHE_SHIFT)

		len = size & ~PAGE_CACHE_MASK;

		len = size & ~PAGE_CACHE_MASK;

	else

	else

		len = PAGE_CACHE_SIZE;

		len = PAGE_CACHE_SIZE;



	/*

	/*

	 * return if we have all the buffers mapped. This avoid

	 * Return if we have all the buffers mapped. This avoids the need to do

	 * the need to call write_begin/write_end which does a

	 * journal_start/journal_stop which can block and take a long time

	 * journal_start/journal_stop which can block and take

	 * long time

	 */

	 */

	if (page_has_buffers(page)) {

	if (page_has_buffers(page)) {

		if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,

		if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,

					ext4_bh_unmapped)) {

					ext4_bh_unmapped)) {

			up_read(&inode->i_alloc_sem);

			/* Wait so that we don't change page under IO */

			return VM_FAULT_LOCKED;

			wait_on_page_writeback(page);

			ret = VM_FAULT_LOCKED;

			goto out;

		}

		}

	}

	}

	unlock_page(page);

	unlock_page(page);

	/*

	/* OK, we need to fill the hole... */

	 * OK, we need to fill the hole... Do write_begin write_end

	if (ext4_should_dioread_nolock(inode))

	 * to do block allocation/reservation.We are not holding

		get_block = ext4_get_block_write;

	 * inode.i__mutex here. That allow * parallel write_begin,

	else

	 * write_end call. lock_page prevent this from happening

		get_block = ext4_get_block;

	 * on the same page though

retry_alloc:

	 */

	handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode));

	ret = mapping->a_ops->write_begin(file, mapping, page_offset(page),

	if (IS_ERR(handle)) {

			len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata);

		ret = VM_FAULT_SIGBUS;

	if (ret < 0)

		goto out;

		goto out_unlock;

	}

	ret = mapping->a_ops->write_end(file, mapping, page_offset(page),

	ret = __block_page_mkwrite(vma, vmf, get_block);

			len, len, page, fsdata);

	if (!ret && ext4_should_journal_data(inode)) {

	if (ret < 0)

		if (walk_page_buffers(handle, page_buffers(page), 0,

		goto out_unlock;

			  PAGE_CACHE_SIZE, NULL, do_journal_get_write_access)) {

	ret = 0;

			unlock_page(page);



	/*

	 * write_begin/end might have created a dirty page and someone

	 * could wander in and start the IO.  Make sure that hasn't

	 * happened.

	 */

	lock_page(page);

	wait_on_page_writeback(page);

	up_read(&inode->i_alloc_sem);

	return VM_FAULT_LOCKED;

out_unlock:

	if (ret)

			ret = VM_FAULT_SIGBUS;

			ret = VM_FAULT_SIGBUS;

	up_read(&inode->i_alloc_sem);

			goto out;

		}

		ext4_set_inode_state(inode, EXT4_STATE_JDATA);

	}

	ext4_journal_stop(handle);

	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))

		goto retry_alloc;

out_ret:

	ret = block_page_mkwrite_return(ret);

out:

	return ret;

	return ret;

}
 
}