[dm-devel] [PATCH 0/6] Support DAX for device-mapper dm-linear devices

Kani, Toshimitsu toshi.kani at hpe.com
Mon Jun 20 21:28:00 UTC 2016


On Mon, 2016-06-20 at 14:01 -0600, Kani, Toshimitsu wrote:
> On Mon, 2016-06-20 at 15:52 -0400, Mike Snitzer wrote:
> > 
> > On Mon, Jun 20 2016 at  3:40pm -0400,
> > Mike Snitzer <snitzer at redhat.com> wrote:
> >  
 :
> > > If I don't use XFS, and only issue IO directly to the /dev/pmem/lv, I
> > > don't see this corruption.
> >
> > I did the same test with ext4 instead of xfs and it resulted in the same
> > type of systemic corruption (lvm2 metadata corrupted too):
> > 
 :
> I will look into the issue.

Hi Mike,

Can you fold the following patch to the dm-linear patch?

Thanks,
-Tsohi

------
Subject: [PATCH] dm-linear: Fix partition handling for DAX

Partition handling was missing in linear_direct_access().
Call bdev_direct_access(), instead of directly calling
target direct_access function.

Signed-off-by: Toshi Kani <toshi.kani at hpe.com>
---
 drivers/md/dm-linear.c |   14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c
index 325aa06..38323e4 100644
--- a/drivers/md/dm-linear.c
+++ b/drivers/md/dm-linear.c
@@ -148,10 +148,16 @@ static long linear_direct_access(struct dm_target *ti,
sector_t sector,
 {
 	struct linear_c *lc = ti->private;
 	struct block_device *bdev = lc->dev->bdev;
-	const struct block_device_operations *bd_ops = bdev->bd_disk->fops;
-
-	return bd_ops->direct_access(bdev, linear_map_sector(ti, sector),
-				     kaddr, pfn, size);
+	struct blk_dax_ctl dax = {
+		.sector = linear_map_sector(ti, sector),
+		.size = size,
+	};
+	long ret;
+
+	ret = bdev_direct_access(bdev, &dax);
+	*kaddr = dax.addr;
+	*pfn = dax.pfn;
+	return ret;
 }
 
 static struct target_type linear_target = {




More information about the dm-devel mailing list