diff mbox

[v2,3/4] xfs: unmap dax at shutdown (force_failure)

Message ID 20160106045632.38788.84927.stgit@dwillia2-desk3.amr.corp.intel.com (mailing list archive)
State New, archived
Headers show

Commit Message

Dan Williams Jan. 6, 2016, 4:56 a.m. UTC
When an exceptional event triggers xfs_force_shutdown() tear down dax
mappings.  Quoting Dave,

    "The simple fact is that a /filesystem/ shutdown needs to do DAX
    mapping invalidation regardless of whether the block device has
    been unplugged or not. This is not a case of "this only happens
    when we unplug the device", this is a user data protection
    mechanism that we use to prevent corruption propagation once it
    has been detected. A device unplug is just one type of
    "corruption" that can occur."

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 fs/xfs/xfs_fsops.c |    9 +++++++++
 1 file changed, 9 insertions(+)

Comments

Dave Chinner Jan. 8, 2016, 12:16 a.m. UTC | #1
On Tue, Jan 05, 2016 at 08:56:32PM -0800, Dan Williams wrote:
> When an exceptional event triggers xfs_force_shutdown() tear down dax
> mappings.  Quoting Dave,
> 
>     "The simple fact is that a /filesystem/ shutdown needs to do DAX
>     mapping invalidation regardless of whether the block device has
>     been unplugged or not. This is not a case of "this only happens
>     when we unplug the device", this is a user data protection
>     mechanism that we use to prevent corruption propagation once it
>     has been detected. A device unplug is just one type of
>     "corruption" that can occur."
> 
> Signed-off-by: Dan Williams <dan.j.williams@intel.com>
> ---
>  fs/xfs/xfs_fsops.c |    9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
> index ee3aaa0a5317..0c6a52809dcc 100644
> --- a/fs/xfs/xfs_fsops.c
> +++ b/fs/xfs/xfs_fsops.c
> @@ -828,6 +828,15 @@ xfs_do_force_shutdown(
>  	if (xfs_log_force_umount(mp, logerror))
>  		return;
>  
> +	/*
> +	 * If DAX is in use, we have to unmap all direct access virtual
> +	 * mappings to ensure nothing more gets written directly from
> +	 * userspace. This will force them to refault and that will
> +	 * result in them detecting the shutdown condition and hence
> +	 * will fail appropriately.
> +	 */
> +	unmap_dax_inodes(mp->m_super);
> +
>  	if (flags & SHUTDOWN_CORRUPT_INCORE) {
>  		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_CORRUPT,
>      "Corruption of in-memory data detected.  Shutting down filesystem");

Looks fine.

Acked-by: Dave Chinner <dchinner@redhat.com>
diff mbox

Patch

diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index ee3aaa0a5317..0c6a52809dcc 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -828,6 +828,15 @@  xfs_do_force_shutdown(
 	if (xfs_log_force_umount(mp, logerror))
 		return;
 
+	/*
+	 * If DAX is in use, we have to unmap all direct access virtual
+	 * mappings to ensure nothing more gets written directly from
+	 * userspace. This will force them to refault and that will
+	 * result in them detecting the shutdown condition and hence
+	 * will fail appropriately.
+	 */
+	unmap_dax_inodes(mp->m_super);
+
 	if (flags & SHUTDOWN_CORRUPT_INCORE) {
 		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_CORRUPT,
     "Corruption of in-memory data detected.  Shutting down filesystem");