xfs_aops.c

linux 内核源代码

/*
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include "xfs.h"
#include "xfs_bit.h"
#include "xfs_log.h"
#include "xfs_inum.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_dir2.h"
#include "xfs_trans.h"
#include "xfs_dmapi.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_alloc.h"
#include "xfs_btree.h"
#include "xfs_error.h"
#include "xfs_rw.h"
#include "xfs_iomap.h"
#include "xfs_vnodeops.h"
#include <linux/mpage.h>
#include <linux/pagevec.h>
#include <linux/writeback.h>

STATIC void
xfs_count_page_state(
	struct page		*page,
	int			*delalloc,
	int			*unmapped,
	int			*unwritten)
{
	struct buffer_head	*bh, *head;

	*delalloc = *unmapped = *unwritten = 0;

	bh = head = page_buffers(page);
	do {
		if (buffer_uptodate(bh) && !buffer_mapped(bh))
			(*unmapped) = 1;
		else if (buffer_unwritten(bh))
			(*unwritten) = 1;
		else if (buffer_delay(bh))
			(*delalloc) = 1;
	} while ((bh = bh->b_this_page) != head);
}

#if defined(XFS_RW_TRACE)
void
xfs_page_trace(
	int		tag,
	struct inode	*inode,
	struct page	*page,
	unsigned long	pgoff)
{
	xfs_inode_t	*ip;
	bhv_vnode_t	*vp = vn_from_inode(inode);
	loff_t		isize = i_size_read(inode);
	loff_t		offset = page_offset(page);
	int		delalloc = -1, unmapped = -1, unwritten = -1;

	if (page_has_buffers(page))
		xfs_count_page_state(page, &delalloc, &unmapped, &unwritten);

	ip = xfs_vtoi(vp);
	if (!ip->i_rwtrace)
		return;

	ktrace_enter(ip->i_rwtrace,
		(void *)((unsigned long)tag),
		(void *)ip,
		(void *)inode,
		(void *)page,
		(void *)pgoff,
		(void *)((unsigned long)((ip->i_d.di_size >> 32) & 0xffffffff)),
		(void *)((unsigned long)(ip->i_d.di_size & 0xffffffff)),
		(void *)((unsigned long)((isize >> 32) & 0xffffffff)),
		(void *)((unsigned long)(isize & 0xffffffff)),
		(void *)((unsigned long)((offset >> 32) & 0xffffffff)),
		(void *)((unsigned long)(offset & 0xffffffff)),
		(void *)((unsigned long)delalloc),
		(void *)((unsigned long)unmapped),
		(void *)((unsigned long)unwritten),
		(void *)((unsigned long)current_pid()),
		(void *)NULL);
}
#else
#define xfs_page_trace(tag, inode, page, pgoff)
#endif

/*
 * Schedule IO completion handling on a xfsdatad if this was
 * the final hold on this ioend. If we are asked to wait,
 * flush the workqueue.
 */
STATIC void
xfs_finish_ioend(
	xfs_ioend_t	*ioend,
	int		wait)
{
	if (atomic_dec_and_test(&ioend->io_remaining)) {
		queue_work(xfsdatad_workqueue, &ioend->io_work);
		if (wait)
			flush_workqueue(xfsdatad_workqueue);
	}
}

/*
 * We're now finished for good with this ioend structure.
 * Update the page state via the associated buffer_heads,
 * release holds on the inode and bio, and finally free
 * up memory.  Do not use the ioend after this.
 */
STATIC void
xfs_destroy_ioend(
	xfs_ioend_t		*ioend)
{
	struct buffer_head	*bh, *next;

	for (bh = ioend->io_buffer_head; bh; bh = next) {
		next = bh->b_private;
		bh->b_end_io(bh, !ioend->io_error);
	}
	if (unlikely(ioend->io_error)) {
		vn_ioerror(XFS_I(ioend->io_inode), ioend->io_error,
				__FILE__,__LINE__);
	}
	vn_iowake(XFS_I(ioend->io_inode));
	mempool_free(ioend, xfs_ioend_pool);
}

/*
 * Update on-disk file size now that data has been written to disk.
 * The current in-memory file size is i_size.  If a write is beyond
 * eof io_new_size will be the intended file size until i_size is
 * updated.  If this write does not extend all the way to the valid
 * file size then restrict this update to the end of the write.
 */
STATIC void
xfs_setfilesize(
	xfs_ioend_t		*ioend)
{
	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
	xfs_fsize_t		isize;
	xfs_fsize_t		bsize;

	ASSERT((ip->i_d.di_mode & S_IFMT) == S_IFREG);
	ASSERT(ioend->io_type != IOMAP_READ);

	if (unlikely(ioend->io_error))
		return;

	bsize = ioend->io_offset + ioend->io_size;

	xfs_ilock(ip, XFS_ILOCK_EXCL);

	isize = MAX(ip->i_size, ip->i_iocore.io_new_size);
	isize = MIN(isize, bsize);

	if (ip->i_d.di_size < isize) {
		ip->i_d.di_size = isize;
		ip->i_update_core = 1;
		ip->i_update_size = 1;
		mark_inode_dirty_sync(ioend->io_inode);
	}

	xfs_iunlock(ip, XFS_ILOCK_EXCL);
}

/*
 * Buffered IO write completion for delayed allocate extents.
 */
STATIC void
xfs_end_bio_delalloc(
	struct work_struct	*work)
{
	xfs_ioend_t		*ioend =
		container_of(work, xfs_ioend_t, io_work);

	xfs_setfilesize(ioend);
	xfs_destroy_ioend(ioend);
}

/*
 * Buffered IO write completion for regular, written extents.
 */
STATIC void
xfs_end_bio_written(
	struct work_struct	*work)
{
	xfs_ioend_t		*ioend =
		container_of(work, xfs_ioend_t, io_work);

	xfs_setfilesize(ioend);
	xfs_destroy_ioend(ioend);
}

/*
 * IO write completion for unwritten extents.
 *
 * Issue transactions to convert a buffer range from unwritten
 * to written extents.
 */
STATIC void
xfs_end_bio_unwritten(
	struct work_struct	*work)
{
	xfs_ioend_t		*ioend =
		container_of(work, xfs_ioend_t, io_work);
	xfs_off_t		offset = ioend->io_offset;
	size_t			size = ioend->io_size;

	if (likely(!ioend->io_error)) {
		xfs_bmap(XFS_I(ioend->io_inode), offset, size,
				BMAPI_UNWRITTEN, NULL, NULL);
		xfs_setfilesize(ioend);
	}
	xfs_destroy_ioend(ioend);
}

/*
 * IO read completion for regular, written extents.
 */
STATIC void
xfs_end_bio_read(
	struct work_struct	*work)
{
	xfs_ioend_t		*ioend =
		container_of(work, xfs_ioend_t, io_work);

	xfs_destroy_ioend(ioend);
}

/*
 * Allocate and initialise an IO completion structure.
 * We need to track unwritten extent write completion here initially.
 * We'll need to extend this for updating the ondisk inode size later
 * (vs. incore size).
 */
STATIC xfs_ioend_t *
xfs_alloc_ioend(
	struct inode		*inode,
	unsigned int		type)
{
	xfs_ioend_t		*ioend;

	ioend = mempool_alloc(xfs_ioend_pool, GFP_NOFS);

	/*
	 * Set the count to 1 initially, which will prevent an I/O
	 * completion callback from happening before we have started
	 * all the I/O from calling the completion routine too early.
	 */
	atomic_set(&ioend->io_remaining, 1);
	ioend->io_error = 0;
	ioend->io_list = NULL;
	ioend->io_type = type;
	ioend->io_inode = inode;
	ioend->io_buffer_head = NULL;
	ioend->io_buffer_tail = NULL;
	atomic_inc(&XFS_I(ioend->io_inode)->i_iocount);
	ioend->io_offset = 0;
	ioend->io_size = 0;

	if (type == IOMAP_UNWRITTEN)
		INIT_WORK(&ioend->io_work, xfs_end_bio_unwritten);
	else if (type == IOMAP_DELAY)
		INIT_WORK(&ioend->io_work, xfs_end_bio_delalloc);
	else if (type == IOMAP_READ)
		INIT_WORK(&ioend->io_work, xfs_end_bio_read);
	else
		INIT_WORK(&ioend->io_work, xfs_end_bio_written);

	return ioend;
}

STATIC int
xfs_map_blocks(
	struct inode		*inode,
	loff_t			offset,
	ssize_t			count,
	xfs_iomap_t		*mapp,
	int			flags)
{
	xfs_inode_t		*ip = XFS_I(inode);
	int			error, nmaps = 1;

	error = xfs_bmap(ip, offset, count,
				flags, mapp, &nmaps);
	if (!error && (flags & (BMAPI_WRITE|BMAPI_ALLOCATE)))
		xfs_iflags_set(ip, XFS_IMODIFIED);
	return -error;
}

STATIC_INLINE int
xfs_iomap_valid(
	xfs_iomap_t		*iomapp,
	loff_t			offset)
{
	return offset >= iomapp->iomap_offset &&
		offset < iomapp->iomap_offset + iomapp->iomap_bsize;
}

/*
 * BIO completion handler for buffered IO.
 */
STATIC void
xfs_end_bio(
	struct bio		*bio,
	int			error)
{
	xfs_ioend_t		*ioend = bio->bi_private;

	ASSERT(atomic_read(&bio->bi_cnt) >= 1);
	ioend->io_error = test_bit(BIO_UPTODATE, &bio->bi_flags) ? 0 : error;

	/* Toss bio and pass work off to an xfsdatad thread */
	bio->bi_private = NULL;
	bio->bi_end_io = NULL;
	bio_put(bio);

	xfs_finish_ioend(ioend, 0);
}

STATIC void
xfs_submit_ioend_bio(
	xfs_ioend_t	*ioend,
	struct bio	*bio)
{
	atomic_inc(&ioend->io_remaining);

	bio->bi_private = ioend;
	bio->bi_end_io = xfs_end_bio;

	submit_bio(WRITE, bio);
	ASSERT(!bio_flagged(bio, BIO_EOPNOTSUPP));
	bio_put(bio);
}

STATIC struct bio *
xfs_alloc_ioend_bio(
	struct buffer_head	*bh)
{
	struct bio		*bio;
	int			nvecs = bio_get_nr_vecs(bh->b_bdev);

	do {
		bio = bio_alloc(GFP_NOIO, nvecs);
		nvecs >>= 1;
	} while (!bio);

	ASSERT(bio->bi_private == NULL);
	bio->bi_sector = bh->b_blocknr * (bh->b_size >> 9);
	bio->bi_bdev = bh->b_bdev;
	bio_get(bio);
	return bio;
}

STATIC void
xfs_start_buffer_writeback(
	struct buffer_head	*bh)
{
	ASSERT(buffer_mapped(bh));
	ASSERT(buffer_locked(bh));
	ASSERT(!buffer_delay(bh));
	ASSERT(!buffer_unwritten(bh));

	mark_buffer_async_write(bh);
	set_buffer_uptodate(bh);
	clear_buffer_dirty(bh);
}

STATIC void
xfs_start_page_writeback(
	struct page		*page,
	struct writeback_control *wbc,
	int			clear_dirty,
	int			buffers)
{
	ASSERT(PageLocked(page));
	ASSERT(!PageWriteback(page));
	if (clear_dirty)
		clear_page_dirty_for_io(page);
	set_page_writeback(page);
	unlock_page(page);
	/* If no buffers on the page are to be written, finish it here */
	if (!buffers)
		end_page_writeback(page);
}

static inline int bio_add_buffer(struct bio *bio, struct buffer_head *bh)
{
	return bio_add_page(bio, bh->b_page, bh->b_size, bh_offset(bh));
}

/*
 * Submit all of the bios for all of the ioends we have saved up, covering the
 * initial writepage page and also any probed pages.
 *
 * Because we may have multiple ioends spanning a page, we need to start
 * writeback on all the buffers before we submit them for I/O. If we mark the
 * buffers as we got, then we can end up with a page that only has buffers
 * marked async write and I/O complete on can occur before we mark the other
 * buffers async write.
 *
 * The end result of this is that we trip a bug in end_page_writeback() because
 * we call it twice for the one page as the code in end_buffer_async_write()
 * assumes that all buffers on the page are started at the same time.
 *
 * The fix is two passes across the ioend list - one to start writeback on the
 * buffer_heads, and then submit them for I/O on the second pass.
 */
STATIC void
xfs_submit_ioend(
	xfs_ioend_t		*ioend)
{
	xfs_ioend_t		*head = ioend;
	xfs_ioend_t		*next;
	struct buffer_head	*bh;
	struct bio		*bio;
	sector_t		lastblock = 0;

	/* Pass 1 - start writeback */
	do {
		next = ioend->io_list;
		for (bh = ioend->io_buffer_head; bh; bh = bh->b_private) {
			xfs_start_buffer_writeback(bh);
		}
	} while ((ioend = next) != NULL);

	/* Pass 2 - submit I/O */
	ioend = head;
	do {
		next = ioend->io_list;
		bio = NULL;

		for (bh = ioend->io_buffer_head; bh; bh = bh->b_private) {

			if (!bio) {
 retry:
				bio = xfs_alloc_ioend_bio(bh);
			} else if (bh->b_blocknr != lastblock + 1) {
				xfs_submit_ioend_bio(ioend, bio);
				goto retry;
			}

			if (bio_add_buffer(bio, bh) != bh->b_size) {
				xfs_submit_ioend_bio(ioend, bio);
				goto retry;
			}

			lastblock = bh->b_blocknr;
		}
		if (bio)
			xfs_submit_ioend_bio(ioend, bio);
		xfs_finish_ioend(ioend, 0);
	} while ((ioend = next) != NULL);
}

/*
 * Cancel submission of all buffer_heads so far in this endio.
 * Toss the endio too.  Only ever called for the initial page
 * in a writepage request, so only ever one page.
 */
STATIC void
xfs_cancel_ioend(
	xfs_ioend_t		*ioend)
{
	xfs_ioend_t		*next;
	struct buffer_head	*bh, *next_bh;

	do {
		next = ioend->io_list;
		bh = ioend->io_buffer_head;
		do {
			next_bh = bh->b_private;
			clear_buffer_async_write(bh);
			unlock_buffer(bh);
		} while ((bh = next_bh) != NULL);

		vn_iowake(XFS_I(ioend->io_inode));
		mempool_free(ioend, xfs_ioend_pool);
	} while ((ioend = next) != NULL);
}

/*
 * Test to see if we've been building up a completion structure for
 * earlier buffers -- if so, we try to append to this ioend if we
 * can, otherwise we finish off any current ioend and start another.
 * Return true if we've finished the given ioend.
 */
STATIC void
xfs_add_to_ioend(
	struct inode		*inode,
	struct buffer_head	*bh,
	xfs_off_t		offset,
	unsigned int		type,
	xfs_ioend_t		**result,
	int			need_ioend)
{
	xfs_ioend_t		*ioend = *result;

	if (!ioend || need_ioend || type != ioend->io_type) {
		xfs_ioend_t	*previous = *result;

		ioend = xfs_alloc_ioend(inode, type);
		ioend->io_offset = offset;