public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache
@ 2012-08-13 13:08 Jeff Liu
  2012-08-20 18:21 ` Mark Tinguely
  2012-08-20 23:16 ` Dave Chinner
  0 siblings, 2 replies; 4+ messages in thread
From: Jeff Liu @ 2012-08-13 13:08 UTC (permalink / raw)
  To: xfs

Improve xfs_seek_data() to search data offset from page cache for unwritten extents.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>

---
 fs/xfs/xfs_file.c |   70 +++++++++++++++++++++++++++++++++++++++-------------
 1 files changed, 52 insertions(+), 18 deletions(-)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 023dc9b..43b3718 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -1187,8 +1187,6 @@ xfs_seek_data(
 	struct inode		*inode = file->f_mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_bmbt_irec	map[2];
-	int			nmap = 2;
 	loff_t			uninitialized_var(offset);
 	xfs_fsize_t		isize;
 	xfs_fileoff_t		fsbno;
@@ -1204,36 +1202,72 @@ xfs_seek_data(
 		goto out_unlock;
 	}
 
-	fsbno = XFS_B_TO_FSBT(mp, start);
-
 	/*
 	 * Try to read extents from the first block indicated
 	 * by fsbno to the end block of the file.
 	 */
+	fsbno = XFS_B_TO_FSBT(mp, start);
 	end = XFS_B_TO_FSB(mp, isize);
+	for (;;) {
+		struct xfs_bmbt_irec	map[2];
+		int			nmap = 2;
+		unsigned int		i;
 
-	error = xfs_bmapi_read(ip, fsbno, end - fsbno, map, &nmap,
-			       XFS_BMAPI_ENTIRE);
-	if (error)
-		goto out_unlock;
+		error = xfs_bmapi_read(ip, fsbno, end - fsbno, map, &nmap,
+				       XFS_BMAPI_ENTIRE);
+		if (error)
+			goto out_unlock;
 
-	/*
-	 * Treat unwritten extent as data extent since it might
-	 * contains dirty data in page cache.
-	 */
-	if (map[0].br_startblock != HOLESTARTBLOCK) {
-		offset = max_t(loff_t, start,
-			       XFS_FSB_TO_B(mp, map[0].br_startoff));
-	} else {
+		/* No extents at given offset, must be beyond EOF */
+		if (nmap == 0) {
+			error = ENXIO;
+			goto out_unlock;
+		}
+
+		for (i = 0; i < nmap; i++) {
+			offset = max_t(loff_t, start,
+				       XFS_FSB_TO_B(mp, map[i].br_startoff));
+
+			/* Landed in a data extent */
+			if (map[i].br_startblock == DELAYSTARTBLOCK ||
+			    (map[i].br_state == XFS_EXT_NORM &&
+			     !isnullstartblock(map[i].br_startblock)))
+				goto out;
+
+			/*
+			 * Landed in an unwritten extent, try to search data
+			 * from page cache.
+			 */
+			if (map[i].br_state == XFS_EXT_UNWRITTEN) {
+				if (xfs_find_get_desired_pgoff(inode, &map[i],
+							DATA_OFF, &offset))
+					goto out;
+			}
+		}
+
+		/*
+		 * map[0] is hole or its an unwritten extent but
+		 * without data in page cache.  Probably means that
+		 * we are reading after EOF if nothing in map[1].
+		 */
 		if (nmap == 1) {
 			error = ENXIO;
 			goto out_unlock;
 		}
 
-		offset = max_t(loff_t, start,
-			       XFS_FSB_TO_B(mp, map[1].br_startoff));
+		/*
+		 * Nothing was found, proceed to the next round of search
+		 * if reading offset not beyond or hit EOF.
+		 */
+		fsbno = map[i - 1].br_startoff + map[i - 1].br_blockcount;
+		start = XFS_FSB_TO_B(mp, fsbno);
+		if (start >= isize) {
+			error = ENXIO;
+			goto out_unlock;
+		}
 	}
 
+out:
 	if (offset != file->f_pos)
 		file->f_pos = offset;
 
-- 
1.7.4.1

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache
  2012-08-13 13:08 [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache Jeff Liu
@ 2012-08-20 18:21 ` Mark Tinguely
  2012-08-20 23:16 ` Dave Chinner
  1 sibling, 0 replies; 4+ messages in thread
From: Mark Tinguely @ 2012-08-20 18:21 UTC (permalink / raw)
  To: jeff.liu; +Cc: xfs

On 08/13/12 08:08, Jeff Liu wrote:
> Improve xfs_seek_data() to search data offset from page cache for unwritten extents.
>
> Signed-off-by: Jie Liu<jeff.liu@oracle.com>
>
> ---

Looks good.

Reviewed-by: Mark Tinguely <tinguely@sgi.com>

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache
  2012-08-13 13:08 [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache Jeff Liu
  2012-08-20 18:21 ` Mark Tinguely
@ 2012-08-20 23:16 ` Dave Chinner
  2012-08-21  5:07   ` Jie Liu
  1 sibling, 1 reply; 4+ messages in thread
From: Dave Chinner @ 2012-08-20 23:16 UTC (permalink / raw)
  To: Jeff Liu; +Cc: xfs

On Mon, Aug 13, 2012 at 09:08:09PM +0800, Jeff Liu wrote:
> Improve xfs_seek_data() to search data offset from page cache for unwritten extents.
> 
> Signed-off-by: Jie Liu <jeff.liu@oracle.com>

One minor addition, otherwise

Reviewed-by: Dave Chinner <dchinner@redhat.com>

....
> +
> +		/*
> +		 * map[0] is hole or its an unwritten extent but
> +		 * without data in page cache.  Probably means that
> +		 * we are reading after EOF if nothing in map[1].
> +		 */
>  		if (nmap == 1) {
>  			error = ENXIO;
>  			goto out_unlock;
>  		}
>  
> -		offset = max_t(loff_t, start,
> -			       XFS_FSB_TO_B(mp, map[1].br_startoff));
> +		/*
> +		 * Nothing was found, proceed to the next round of search
> +		 * if reading offset not beyond or hit EOF.
> +		 */

		ASSERT(i > 1);

> +		fsbno = map[i - 1].br_startoff + map[i - 1].br_blockcount;
> +		start = XFS_FSB_TO_B(mp, fsbno);
> +		if (start >= isize) {
> +			error = ENXIO;
> +			goto out_unlock;
> +		}
>  	}
>  
> +out:
>  	if (offset != file->f_pos)
>  		file->f_pos = offset;

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache
  2012-08-20 23:16 ` Dave Chinner
@ 2012-08-21  5:07   ` Jie Liu
  0 siblings, 0 replies; 4+ messages in thread
From: Jie Liu @ 2012-08-21  5:07 UTC (permalink / raw)
  To: Dave Chinner; +Cc: xfs

On 08/21/12 07:16, Dave Chinner wrote:
> On Mon, Aug 13, 2012 at 09:08:09PM +0800, Jeff Liu wrote:
>> Improve xfs_seek_data() to search data offset from page cache for unwritten extents.
>>
>> Signed-off-by: Jie Liu <jeff.liu@oracle.com>
> One minor addition, otherwise
>
> Reviewed-by: Dave Chinner <dchinner@redhat.com>
>
> ....
>> +
>> +		/*
>> +		 * map[0] is hole or its an unwritten extent but
>> +		 * without data in page cache.  Probably means that
>> +		 * we are reading after EOF if nothing in map[1].
>> +		 */
>>  		if (nmap == 1) {
>>  			error = ENXIO;
>>  			goto out_unlock;
>>  		}
>>  
>> -		offset = max_t(loff_t, start,
>> -			       XFS_FSB_TO_B(mp, map[1].br_startoff));
>> +		/*
>> +		 * Nothing was found, proceed to the next round of search
>> +		 * if reading offset not beyond or hit EOF.
>> +		 */
> 		ASSERT(i > 1);
Will add this check up.

Thanks,
-Jeff
>
>> +		fsbno = map[i - 1].br_startoff + map[i - 1].br_blockcount;
>> +		start = XFS_FSB_TO_B(mp, fsbno);
>> +		if (start >= isize) {
>> +			error = ENXIO;
>> +			goto out_unlock;
>> +		}
>>  	}
>>  
>> +out:
>>  	if (offset != file->f_pos)
>>  		file->f_pos = offset;
> Cheers,
>
> Dave.

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2012-08-21  5:07 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-08-13 13:08 [PATCH v7 3/4] xfs: xfs_seek_data() refinement with unwritten extents check up from page cache Jeff Liu
2012-08-20 18:21 ` Mark Tinguely
2012-08-20 23:16 ` Dave Chinner
2012-08-21  5:07   ` Jie Liu

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox