diff mbox series

[10/13] fs: factor out a direct_write_fallback helper

Message ID 20230519093521.133226-11-hch@lst.de
State New
Headers show
Series [01/13] iomap: update ki_pos a little later in iomap_dio_complete | expand

Commit Message

Christoph Hellwig May 19, 2023, 9:35 a.m. UTC
Add a helper dealing with handling the syncing of a buffered write fallback
for direct I/O.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 fs/libfs.c         | 36 ++++++++++++++++++++++++++++
 include/linux/fs.h |  2 ++
 mm/filemap.c       | 59 ++++++++++------------------------------------
 3 files changed, 50 insertions(+), 47 deletions(-)

Comments

Damien Le Moal May 22, 2023, 12:16 a.m. UTC | #1
On 5/19/23 18:35, Christoph Hellwig wrote:
> Add a helper dealing with handling the syncing of a buffered write fallback
> for direct I/O.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Looks OK. One comment below.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>

> +	/*
> +	 * We need to ensure that the page cache pages are written to disk and
> +	 * invalidated to preserve the expected O_DIRECT semantics.
> +	 */
> +	end = pos + buffered_written - 1;
> +	err = filemap_write_and_wait_range(mapping, pos, end);
> +	if (err < 0) {
> +		/*
> +		 * We don't know how much we wrote, so just return the number of
> +		 * bytes which were direct-written
> +		 */
> +		return err;
> +	}
> +	invalidate_mapping_pages(mapping, pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
> +	return direct_written + buffered_written;

Why not adding here something like:

	if (buffered_written != iov_iter_count(from))
		return -EIO;

	return direct_written + buffered_written;

to have the same semantic as plain DIO ?
Miklos Szeredi May 22, 2023, 2:19 p.m. UTC | #2
On Fri, 19 May 2023 at 11:36, Christoph Hellwig <hch@lst.de> wrote:
>
> Add a helper dealing with handling the syncing of a buffered write fallback
> for direct I/O.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  fs/libfs.c         | 36 ++++++++++++++++++++++++++++
>  include/linux/fs.h |  2 ++
>  mm/filemap.c       | 59 ++++++++++------------------------------------
>  3 files changed, 50 insertions(+), 47 deletions(-)
>
> diff --git a/fs/libfs.c b/fs/libfs.c
> index 89cf614a327158..9f3791fc6e0715 100644
> --- a/fs/libfs.c
> +++ b/fs/libfs.c
> @@ -1613,3 +1613,39 @@ u64 inode_query_iversion(struct inode *inode)
>         return cur >> I_VERSION_QUERIED_SHIFT;
>  }
>  EXPORT_SYMBOL(inode_query_iversion);
> +
> +ssize_t direct_write_fallback(struct kiocb *iocb, struct iov_iter *iter,
> +               ssize_t direct_written, ssize_t buffered_written)
> +{
> +       struct address_space *mapping = iocb->ki_filp->f_mapping;
> +       loff_t pos = iocb->ki_pos, end;

At this point pos will point after the end of the buffered write (as
per earlier patches), yes?

> +       int err;
> +
> +       /*
> +        * If the buffered write fallback returned an error, we want to return
> +        * the number of bytes which were written by direct I/O, or the error
> +        * code if that was zero.
> +        *
> +        * Note that this differs from normal direct-io semantics, which will
> +        * return -EFOO even if some bytes were written.
> +        */
> +       if (unlikely(buffered_written < 0))
> +               return buffered_written;
> +
> +       /*
> +        * We need to ensure that the page cache pages are written to disk and
> +        * invalidated to preserve the expected O_DIRECT semantics.
> +        */
> +       end = pos + buffered_written - 1;

So this calculation is wrong.

AFAICS this affects later patches as well.

Thanks,
Miklos
Christoph Hellwig May 23, 2023, 4:03 p.m. UTC | #3
On Mon, May 22, 2023 at 04:19:38PM +0200, Miklos Szeredi wrote:
> > +               ssize_t direct_written, ssize_t buffered_written)
> > +{
> > +       struct address_space *mapping = iocb->ki_filp->f_mapping;
> > +       loff_t pos = iocb->ki_pos, end;
> 
> At this point pos will point after the end of the buffered write (as
> per earlier patches), yes?

Yes.  I'll fix the pos and end calculation.
diff mbox series

Patch

diff --git a/fs/libfs.c b/fs/libfs.c
index 89cf614a327158..9f3791fc6e0715 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -1613,3 +1613,39 @@  u64 inode_query_iversion(struct inode *inode)
 	return cur >> I_VERSION_QUERIED_SHIFT;
 }
 EXPORT_SYMBOL(inode_query_iversion);
+
+ssize_t direct_write_fallback(struct kiocb *iocb, struct iov_iter *iter,
+		ssize_t direct_written, ssize_t buffered_written)
+{
+	struct address_space *mapping = iocb->ki_filp->f_mapping;
+	loff_t pos = iocb->ki_pos, end;
+	int err;
+
+	/*
+	 * If the buffered write fallback returned an error, we want to return
+	 * the number of bytes which were written by direct I/O, or the error
+	 * code if that was zero.
+	 *
+	 * Note that this differs from normal direct-io semantics, which will
+	 * return -EFOO even if some bytes were written.
+	 */
+	if (unlikely(buffered_written < 0))
+		return buffered_written;
+
+	/*
+	 * We need to ensure that the page cache pages are written to disk and
+	 * invalidated to preserve the expected O_DIRECT semantics.
+	 */
+	end = pos + buffered_written - 1;
+	err = filemap_write_and_wait_range(mapping, pos, end);
+	if (err < 0) {
+		/*
+		 * We don't know how much we wrote, so just return the number of
+		 * bytes which were direct-written
+		 */
+		return err;
+	}
+	invalidate_mapping_pages(mapping, pos >> PAGE_SHIFT, end >> PAGE_SHIFT);
+	return direct_written + buffered_written;
+}
+EXPORT_SYMBOL_GPL(direct_write_fallback);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index e4efc1792a877a..576a945db178ef 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2738,6 +2738,8 @@  extern ssize_t __generic_file_write_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_write_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_direct_write(struct kiocb *, struct iov_iter *);
 ssize_t generic_perform_write(struct kiocb *, struct iov_iter *);
+ssize_t direct_write_fallback(struct kiocb *iocb, struct iov_iter *iter,
+		ssize_t direct_written, ssize_t buffered_written);
 
 ssize_t vfs_iter_read(struct file *file, struct iov_iter *iter, loff_t *ppos,
 		rwf_t flags);
diff --git a/mm/filemap.c b/mm/filemap.c
index c1b988199aece5..875b2108d0a05f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -4008,25 +4008,21 @@  ssize_t __generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 {
 	struct file *file = iocb->ki_filp;
 	struct address_space *mapping = file->f_mapping;
-	struct inode 	*inode = mapping->host;
-	ssize_t		written = 0;
-	ssize_t		err;
-	ssize_t		status;
+	struct inode *inode = mapping->host;
+	ssize_t ret;
 
 	/* We can write back this queue in page reclaim */
 	current->backing_dev_info = inode_to_bdi(inode);
-	err = file_remove_privs(file);
-	if (err)
+	ret = file_remove_privs(file);
+	if (ret)
 		goto out;
 
-	err = file_update_time(file);
-	if (err)
+	ret = file_update_time(file);
+	if (ret)
 		goto out;
 
 	if (iocb->ki_flags & IOCB_DIRECT) {
-		loff_t pos, endbyte;
-
-		written = generic_file_direct_write(iocb, from);
+		ret = generic_file_direct_write(iocb, from);
 		/*
 		 * If the write stopped short of completing, fall back to
 		 * buffered writes.  Some filesystems do this for writes to
@@ -4034,46 +4030,15 @@  ssize_t __generic_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 		 * not succeed (even if it did, DAX does not handle dirty
 		 * page-cache pages correctly).
 		 */
-		if (written < 0 || !iov_iter_count(from) || IS_DAX(inode))
-			goto out;
-
-		pos = iocb->ki_pos;
-		status = generic_perform_write(iocb, from);
-		/*
-		 * If generic_perform_write() returned a synchronous error
-		 * then we want to return the number of bytes which were
-		 * direct-written, or the error code if that was zero.  Note
-		 * that this differs from normal direct-io semantics, which
-		 * will return -EFOO even if some bytes were written.
-		 */
-		if (unlikely(status < 0)) {
-			err = status;
-			goto out;
-		}
-		/*
-		 * We need to ensure that the page cache pages are written to
-		 * disk and invalidated to preserve the expected O_DIRECT
-		 * semantics.
-		 */
-		endbyte = pos + status - 1;
-		err = filemap_write_and_wait_range(mapping, pos, endbyte);
-		if (err == 0) {
-			written += status;
-			invalidate_mapping_pages(mapping,
-						 pos >> PAGE_SHIFT,
-						 endbyte >> PAGE_SHIFT);
-		} else {
-			/*
-			 * We don't know how much we wrote, so just return
-			 * the number of bytes which were direct-written
-			 */
-		}
+		if (ret >= 0 && iov_iter_count(from) && !IS_DAX(inode))
+			ret = direct_write_fallback(iocb, from, ret,
+					generic_perform_write(iocb, from));
 	} else {
-		written = generic_perform_write(iocb, from);
+		ret = generic_perform_write(iocb, from);
 	}
 out:
 	current->backing_dev_info = NULL;
-	return written ? written : err;
+	return ret;
 }
 EXPORT_SYMBOL(__generic_file_write_iter);