This is in preparation to support DAX for realtime device on XFS.
bdev_dax_supported() will be taking a block_device as input instead of
a superblock. The only place a super_block is used in this function is
providing the id for debug outputs. Also __bdev_dax_supported()
will be removed since it just directly calls bdev_dax_supported()
and is not reference by any other code.
Signed-off-by: Dave Jiang <[email protected]>
---
drivers/dax/super.c | 33 ++++++++++++++++-----------------
fs/ext2/super.c | 2 +-
fs/ext4/super.c | 2 +-
fs/xfs/xfs_ioctl.c | 2 +-
fs/xfs/xfs_super.c | 2 +-
include/linux/dax.h | 8 ++------
6 files changed, 22 insertions(+), 27 deletions(-)
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index 473af694ad1c..3bdf6787b6df 100644
--- a/drivers/dax/super.c
+++ b/drivers/dax/super.c
@@ -70,11 +70,10 @@ struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev)
return fs_dax_get_by_host(bdev->bd_disk->disk_name);
}
EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
-#endif
/**
- * __bdev_dax_supported() - Check if the device supports dax for filesystem
- * @sb: The superblock of the device
+ * bdev_dax_supported() - Check if the device supports dax for filesystem
+ * @sb: The block_device of the device
* @blocksize: The block size of the device
*
* This is a library function for filesystems to check if the block device
@@ -82,9 +81,8 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
*
* Return: negative errno if unsupported, 0 if supported.
*/
-int __bdev_dax_supported(struct super_block *sb, int blocksize)
+int bdev_dax_supported(struct block_device *bdev, int blocksize)
{
- struct block_device *bdev = sb->s_bdev;
struct dax_device *dax_dev;
pgoff_t pgoff;
int err, id;
@@ -93,22 +91,22 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
long len;
if (blocksize != PAGE_SIZE) {
- pr_debug("VFS (%s): error: unsupported blocksize for dax\n",
- sb->s_id);
+ pr_debug("bdev (%d:%d): error: unsupported blocksize for dax\n",
+ MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
return -EINVAL;
}
err = bdev_dax_pgoff(bdev, 0, PAGE_SIZE, &pgoff);
if (err) {
- pr_debug("VFS (%s): error: unaligned partition for dax\n",
- sb->s_id);
+ pr_debug("bdev (%d:%d): error: unaligned partition for dax\n",
+ MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
return err;
}
dax_dev = dax_get_by_host(bdev->bd_disk->disk_name);
if (!dax_dev) {
- pr_debug("VFS (%s): error: device does not support dax\n",
- sb->s_id);
+ pr_debug("bdev (%d:%d): error: device does not support dax\n",
+ MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
return -EOPNOTSUPP;
}
@@ -119,8 +117,8 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
put_dax(dax_dev);
if (len < 1) {
- pr_debug("VFS (%s): error: dax access failed (%ld)\n",
- sb->s_id, len);
+ pr_debug("bdev (%d:%d): error: dax access failed (%ld)\n",
+ MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev), len);
return len < 0 ? len : -EIO;
}
@@ -128,15 +126,16 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
|| pfn_t_devmap(pfn))
/* pass */;
else {
- pr_debug("VFS (%s): error: dax support not enabled\n",
- sb->s_id);
+ pr_debug("bdev (%d:%d): error: dax support not enabled\n",
+ MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
return -EOPNOTSUPP;
}
return 0;
}
-EXPORT_SYMBOL_GPL(__bdev_dax_supported);
-#endif
+EXPORT_SYMBOL_GPL(bdev_dax_supported);
+#endif /* CONFIG_FS_DAX */
+#endif /* CONFIG_BLOCK */
enum dax_device_flags {
/* !alive + rcu grace period == no new operations / mappings */
diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 38f9222606ee..a0de090b18d5 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -958,7 +958,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
if (sbi->s_mount_opt & EXT2_MOUNT_DAX) {
- err = bdev_dax_supported(sb, blocksize);
+ err = bdev_dax_supported(sb->s_bdev, blocksize);
if (err) {
ext2_msg(sb, KERN_ERR,
"DAX unsupported by block device. Turning off DAX.");
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 18873ea89e08..7ebc8d5cab8c 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -3712,7 +3712,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
" that may contain inline data");
sbi->s_mount_opt &= ~EXT4_MOUNT_DAX;
}
- err = bdev_dax_supported(sb, blocksize);
+ err = bdev_dax_supported(sb->s_bdev, blocksize);
if (err) {
ext4_msg(sb, KERN_ERR,
"DAX unsupported by block device. Turning off DAX.");
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 20dc65fef6a4..5fd4d50eb1c6 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1102,7 +1102,7 @@ xfs_ioctl_setattr_dax_invalidate(
if (fa->fsx_xflags & FS_XFLAG_DAX) {
if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
return -EINVAL;
- if (bdev_dax_supported(sb, sb->s_blocksize) < 0)
+ if (bdev_dax_supported(sb->s_bdev, sb->s_blocksize) < 0)
return -EINVAL;
}
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index 1dacccc367f8..e8a687232614 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1652,7 +1652,7 @@ xfs_fs_fill_super(
xfs_warn(mp,
"DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
- error = bdev_dax_supported(sb, sb->s_blocksize);
+ error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
if (error) {
xfs_alert(mp,
"DAX unsupported by block device. Turning off DAX.");
diff --git a/include/linux/dax.h b/include/linux/dax.h
index 5258346c558c..a31a7e3f929b 100644
--- a/include/linux/dax.h
+++ b/include/linux/dax.h
@@ -40,11 +40,7 @@ static inline void put_dax(struct dax_device *dax_dev)
int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff);
#if IS_ENABLED(CONFIG_FS_DAX)
-int __bdev_dax_supported(struct super_block *sb, int blocksize);
-static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
-{
- return __bdev_dax_supported(sb, blocksize);
-}
+int bdev_dax_supported(struct block_device *bdev, int blocksize);
static inline struct dax_device *fs_dax_get_by_host(const char *host)
{
@@ -58,7 +54,7 @@ static inline void fs_put_dax(struct dax_device *dax_dev)
struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev);
#else
-static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
+static inline int bdev_dax_supported(struct block_device *bdev, int blocksize)
{
return -EOPNOTSUPP;
}
When using realtime device (rtdev) with xfs where the data device is not
DAX capable, two issues arise. One is when data device is not DAX but the
realtime device is DAX capable, we currently disable DAX.
After passing this check, we are also not marking the inode as DAX capable.
This change will allow DAX enabled if the data device or the realtime
device is DAX capable. S_DAX will be marked for the inode if the file is
residing on a DAX capable device. This will prevent the case of rtdev is not
DAX and data device is DAX to create realtime files.
Signed-off-by: Dave Jiang <[email protected]>
Reported-by: Darrick Wong <[email protected]>
---
fs/xfs/xfs_iops.c | 3 ++-
fs/xfs/xfs_super.c | 9 ++++++++-
2 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 56475fcd76f2..ab352c325301 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
!xfs_is_reflink_inode(ip) &&
(ip->i_mount->m_flags & XFS_MOUNT_DAX ||
- ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
+ ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
+ blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
inode->i_flags |= S_DAX;
}
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index e8a687232614..5ac478924dce 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
sb->s_flags |= SB_I_VERSION;
if (mp->m_flags & XFS_MOUNT_DAX) {
+ bool rtdev_is_dax = false;
+
xfs_warn(mp,
"DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
+ if (mp->m_rtdev_targp->bt_daxdev)
+ if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
+ sb->s_blocksize) == 0)
+ rtdev_is_dax = true;
+
error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
- if (error) {
+ if (error && !rtdev_is_dax) {
xfs_alert(mp,
"DAX unsupported by block device. Turning off DAX.");
mp->m_flags &= ~XFS_MOUNT_DAX;
On Thu, Feb 01, 2018 at 01:32:59PM -0700, Dave Jiang wrote:
> This is in preparation to support DAX for realtime device on XFS.
> bdev_dax_supported() will be taking a block_device as input instead of
> a superblock. The only place a super_block is used in this function is
> providing the id for debug outputs. Also __bdev_dax_supported()
> will be removed since it just directly calls bdev_dax_supported()
> and is not reference by any other code.
>
> Signed-off-by: Dave Jiang <[email protected]>
> ---
> drivers/dax/super.c | 33 ++++++++++++++++-----------------
> fs/ext2/super.c | 2 +-
> fs/ext4/super.c | 2 +-
> fs/xfs/xfs_ioctl.c | 2 +-
> fs/xfs/xfs_super.c | 2 +-
> include/linux/dax.h | 8 ++------
> 6 files changed, 22 insertions(+), 27 deletions(-)
>
> diff --git a/drivers/dax/super.c b/drivers/dax/super.c
> index 473af694ad1c..3bdf6787b6df 100644
> --- a/drivers/dax/super.c
> +++ b/drivers/dax/super.c
> @@ -70,11 +70,10 @@ struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev)
> return fs_dax_get_by_host(bdev->bd_disk->disk_name);
> }
> EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> -#endif
>
> /**
> - * __bdev_dax_supported() - Check if the device supports dax for filesystem
> - * @sb: The superblock of the device
> + * bdev_dax_supported() - Check if the device supports dax for filesystem
> + * @sb: The block_device of the device
> * @blocksize: The block size of the device
> *
> * This is a library function for filesystems to check if the block device
> @@ -82,9 +81,8 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> *
> * Return: negative errno if unsupported, 0 if supported.
> */
> -int __bdev_dax_supported(struct super_block *sb, int blocksize)
> +int bdev_dax_supported(struct block_device *bdev, int blocksize)
Looking ahead to filesystems that can span multiple dax devices, perhaps
it would be better to be able topass in the bdev that we want to check?
int bdev_dax_supported(sb, bdev, blocksize)
{
/* same tests we do today... */
}
int sb_dax_supported(sb, blocksize)
{
return bdev_dax_supported(sb, sb->s_bdev, blocksize);
}
That way a single bdev filesystem (ext4) can retain the same api it uses
today, and two-bdev filesystems (xfs) can specify which bdev to check.
(Continuing review in the next patch...)
--D
> {
> - struct block_device *bdev = sb->s_bdev;
> struct dax_device *dax_dev;
> pgoff_t pgoff;
> int err, id;
> @@ -93,22 +91,22 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
> long len;
>
> if (blocksize != PAGE_SIZE) {
> - pr_debug("VFS (%s): error: unsupported blocksize for dax\n",
> - sb->s_id);
> + pr_debug("bdev (%d:%d): error: unsupported blocksize for dax\n",
> + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
> return -EINVAL;
> }
>
> err = bdev_dax_pgoff(bdev, 0, PAGE_SIZE, &pgoff);
> if (err) {
> - pr_debug("VFS (%s): error: unaligned partition for dax\n",
> - sb->s_id);
> + pr_debug("bdev (%d:%d): error: unaligned partition for dax\n",
> + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
> return err;
> }
>
> dax_dev = dax_get_by_host(bdev->bd_disk->disk_name);
> if (!dax_dev) {
> - pr_debug("VFS (%s): error: device does not support dax\n",
> - sb->s_id);
> + pr_debug("bdev (%d:%d): error: device does not support dax\n",
> + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
> return -EOPNOTSUPP;
> }
>
> @@ -119,8 +117,8 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
> put_dax(dax_dev);
>
> if (len < 1) {
> - pr_debug("VFS (%s): error: dax access failed (%ld)\n",
> - sb->s_id, len);
> + pr_debug("bdev (%d:%d): error: dax access failed (%ld)\n",
> + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev), len);
> return len < 0 ? len : -EIO;
> }
>
> @@ -128,15 +126,16 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
> || pfn_t_devmap(pfn))
> /* pass */;
> else {
> - pr_debug("VFS (%s): error: dax support not enabled\n",
> - sb->s_id);
> + pr_debug("bdev (%d:%d): error: dax support not enabled\n",
> + MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev));
> return -EOPNOTSUPP;
> }
>
> return 0;
> }
> -EXPORT_SYMBOL_GPL(__bdev_dax_supported);
> -#endif
> +EXPORT_SYMBOL_GPL(bdev_dax_supported);
> +#endif /* CONFIG_FS_DAX */
> +#endif /* CONFIG_BLOCK */
>
> enum dax_device_flags {
> /* !alive + rcu grace period == no new operations / mappings */
> diff --git a/fs/ext2/super.c b/fs/ext2/super.c
> index 38f9222606ee..a0de090b18d5 100644
> --- a/fs/ext2/super.c
> +++ b/fs/ext2/super.c
> @@ -958,7 +958,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
> blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
>
> if (sbi->s_mount_opt & EXT2_MOUNT_DAX) {
> - err = bdev_dax_supported(sb, blocksize);
> + err = bdev_dax_supported(sb->s_bdev, blocksize);
> if (err) {
> ext2_msg(sb, KERN_ERR,
> "DAX unsupported by block device. Turning off DAX.");
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 18873ea89e08..7ebc8d5cab8c 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -3712,7 +3712,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> " that may contain inline data");
> sbi->s_mount_opt &= ~EXT4_MOUNT_DAX;
> }
> - err = bdev_dax_supported(sb, blocksize);
> + err = bdev_dax_supported(sb->s_bdev, blocksize);
> if (err) {
> ext4_msg(sb, KERN_ERR,
> "DAX unsupported by block device. Turning off DAX.");
> diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
> index 20dc65fef6a4..5fd4d50eb1c6 100644
> --- a/fs/xfs/xfs_ioctl.c
> +++ b/fs/xfs/xfs_ioctl.c
> @@ -1102,7 +1102,7 @@ xfs_ioctl_setattr_dax_invalidate(
> if (fa->fsx_xflags & FS_XFLAG_DAX) {
> if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
> return -EINVAL;
> - if (bdev_dax_supported(sb, sb->s_blocksize) < 0)
> + if (bdev_dax_supported(sb->s_bdev, sb->s_blocksize) < 0)
> return -EINVAL;
> }
>
> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> index 1dacccc367f8..e8a687232614 100644
> --- a/fs/xfs/xfs_super.c
> +++ b/fs/xfs/xfs_super.c
> @@ -1652,7 +1652,7 @@ xfs_fs_fill_super(
> xfs_warn(mp,
> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>
> - error = bdev_dax_supported(sb, sb->s_blocksize);
> + error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
> if (error) {
> xfs_alert(mp,
> "DAX unsupported by block device. Turning off DAX.");
> diff --git a/include/linux/dax.h b/include/linux/dax.h
> index 5258346c558c..a31a7e3f929b 100644
> --- a/include/linux/dax.h
> +++ b/include/linux/dax.h
> @@ -40,11 +40,7 @@ static inline void put_dax(struct dax_device *dax_dev)
>
> int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff);
> #if IS_ENABLED(CONFIG_FS_DAX)
> -int __bdev_dax_supported(struct super_block *sb, int blocksize);
> -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> -{
> - return __bdev_dax_supported(sb, blocksize);
> -}
> +int bdev_dax_supported(struct block_device *bdev, int blocksize);
>
> static inline struct dax_device *fs_dax_get_by_host(const char *host)
> {
> @@ -58,7 +54,7 @@ static inline void fs_put_dax(struct dax_device *dax_dev)
>
> struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev);
> #else
> -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> +static inline int bdev_dax_supported(struct block_device *bdev, int blocksize)
> {
> return -EOPNOTSUPP;
> }
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> [PATCH 2 2/2] xfs: fix rt_dev usage for DAX
"[PATCH v2 2/2]" to distinguish the version number from the patch number
more explicitly.
On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> When using realtime device (rtdev) with xfs where the data device is not
> DAX capable, two issues arise. One is when data device is not DAX but the
> realtime device is DAX capable, we currently disable DAX.
> After passing this check, we are also not marking the inode as DAX capable.
> This change will allow DAX enabled if the data device or the realtime
> device is DAX capable. S_DAX will be marked for the inode if the file is
> residing on a DAX capable device. This will prevent the case of rtdev is not
> DAX and data device is DAX to create realtime files.
>
> Signed-off-by: Dave Jiang <[email protected]>
> Reported-by: Darrick Wong <[email protected]>
> ---
> fs/xfs/xfs_iops.c | 3 ++-
> fs/xfs/xfs_super.c | 9 ++++++++-
> 2 files changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index 56475fcd76f2..ab352c325301 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> !xfs_is_reflink_inode(ip) &&
> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
inode->i_sb->s_bdev is the data device bdev, so if the inode is a
realtime file, we're checking the wrong device for daxiness, I think.
Maybe this whole ugly switch statement should get turned into a helper
function?
xfs_ioctl_setattr_dax_invalidate needs to pick the right bdev to check.
> inode->i_flags |= S_DAX;
> }
>
> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> index e8a687232614..5ac478924dce 100644
> --- a/fs/xfs/xfs_super.c
> +++ b/fs/xfs/xfs_super.c
> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
> sb->s_flags |= SB_I_VERSION;
>
> if (mp->m_flags & XFS_MOUNT_DAX) {
> + bool rtdev_is_dax = false;
> +
> xfs_warn(mp,
> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>
> + if (mp->m_rtdev_targp->bt_daxdev)
> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> + sb->s_blocksize) == 0)
> + rtdev_is_dax = true;
> +
> error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
> - if (error) {
> + if (error && !rtdev_is_dax) {
> xfs_alert(mp,
> "DAX unsupported by block device. Turning off DAX.");
> mp->m_flags &= ~XFS_MOUNT_DAX;
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
Does the following patch fix everything for you?
(Note that we can't switch S_DAX on a running fs so you have to remount
the whole fs after setting the dax flag...)
--D
--------------------
fs: allow per-device dax status checking for filesystems
Refactor __bdev_dax_supported into a sb_dax_supported helper for
single-bdev filesystems and a regular bdev_dax_supported that takes a
bdev parameter. This enables multi-device filesystems like xfs to check
that a dax device can work for the particular filesystem. Once that's
in place, actually fix all the parts of XFS where we need to be able to
distinguish between datadev and rtdev.
This patch fixes the problem where we screw up the dax support checking
in xfs if the datadev and rtdev have different dax capabilities.
Signed-off-by: Darrick J. Wong <[email protected]>
---
drivers/dax/super.c | 9 +++++----
fs/ext2/super.c | 2 +-
fs/ext4/super.c | 2 +-
fs/xfs/xfs_ioctl.c | 3 ++-
fs/xfs/xfs_iops.c | 30 +++++++++++++++++++++++++-----
fs/xfs/xfs_super.c | 11 +++++++++--
include/linux/dax.h | 16 ++++++++++++----
7 files changed, 55 insertions(+), 18 deletions(-)
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index 3ec8046..c4db84f 100644
--- a/drivers/dax/super.c
+++ b/drivers/dax/super.c
@@ -72,8 +72,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
#endif
/**
- * __bdev_dax_supported() - Check if the device supports dax for filesystem
+ * bdev_dax_supported() - Check if the device supports dax for filesystem
* @sb: The superblock of the device
+ * @bdev: block device to check
* @blocksize: The block size of the device
*
* This is a library function for filesystems to check if the block device
@@ -81,9 +82,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
*
* Return: negative errno if unsupported, 0 if supported.
*/
-int __bdev_dax_supported(struct super_block *sb, int blocksize)
+int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
+ int blocksize)
{
- struct block_device *bdev = sb->s_bdev;
struct dax_device *dax_dev;
pgoff_t pgoff;
int err, id;
@@ -125,7 +126,7 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
return 0;
}
-EXPORT_SYMBOL_GPL(__bdev_dax_supported);
+EXPORT_SYMBOL_GPL(bdev_dax_supported);
#endif
enum dax_device_flags {
diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 7646818..6556993 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -958,7 +958,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
if (sbi->s_mount_opt & EXT2_MOUNT_DAX) {
- err = bdev_dax_supported(sb, blocksize);
+ err = sb_dax_supported(sb, blocksize);
if (err)
goto failed_mount;
}
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 7c46693..804a2d6 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -3712,7 +3712,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
" that may contain inline data");
goto failed_mount;
}
- err = bdev_dax_supported(sb, blocksize);
+ err = sb_dax_supported(sb, blocksize);
if (err)
goto failed_mount;
}
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 89fb1eb..277355f 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -1103,7 +1103,8 @@ xfs_ioctl_setattr_dax_invalidate(
if (fa->fsx_xflags & FS_XFLAG_DAX) {
if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
return -EINVAL;
- if (bdev_dax_supported(sb, sb->s_blocksize) < 0)
+ if (bdev_dax_supported(sb, xfs_find_bdev_for_inode(VFS_I(ip)),
+ sb->s_blocksize) < 0)
return -EINVAL;
}
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 56475fc..66cd61c 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -1182,6 +1182,30 @@ static const struct inode_operations xfs_inline_symlink_inode_operations = {
.update_time = xfs_vn_update_time,
};
+/* Figure out if this file actually supports DAX. */
+static bool
+xfs_inode_supports_dax(
+ struct xfs_inode *ip)
+{
+ struct xfs_mount *mp = ip->i_mount;
+
+ /* Only supported on non-reflinked files. */
+ if (!S_ISREG(VFS_I(ip)->i_mode) || xfs_is_reflink_inode(ip))
+ return false;
+
+ /* DAX mount option or DAX iflag must be set. */
+ if (!(mp->m_flags & XFS_MOUNT_DAX) &&
+ !(ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
+ return false;
+
+ /* Block size must match page size */
+ if (mp->m_sb.sb_blocksize != PAGE_SIZE)
+ return false;
+
+ /* Device has to support DAX too. */
+ return xfs_find_daxdev_for_inode(VFS_I(ip)) != NULL;
+}
+
STATIC void
xfs_diflags_to_iflags(
struct inode *inode,
@@ -1200,11 +1224,7 @@ xfs_diflags_to_iflags(
inode->i_flags |= S_SYNC;
if (flags & XFS_DIFLAG_NOATIME)
inode->i_flags |= S_NOATIME;
- if (S_ISREG(inode->i_mode) &&
- ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
- !xfs_is_reflink_inode(ip) &&
- (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
- ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
+ if (xfs_inode_supports_dax(ip))
inode->i_flags |= S_DAX;
}
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index 6f1b917..c115bc7 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -1692,11 +1692,18 @@ xfs_fs_fill_super(
sb->s_flags |= SB_I_VERSION;
if (mp->m_flags & XFS_MOUNT_DAX) {
+ int error2 = 0;
+
xfs_warn(mp,
"DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
- error = bdev_dax_supported(sb, sb->s_blocksize);
- if (error) {
+ error = bdev_dax_supported(sb, mp->m_ddev_targp->bt_bdev,
+ sb->s_blocksize);
+ if (mp->m_rtdev_targp)
+ error2 = bdev_dax_supported(sb,
+ mp->m_rtdev_targp->bt_bdev,
+ sb->s_blocksize);
+ if (error && error2) {
xfs_alert(mp,
"DAX unsupported by block device. Turning off DAX.");
mp->m_flags &= ~XFS_MOUNT_DAX;
diff --git a/include/linux/dax.h b/include/linux/dax.h
index 5258346..1107a98 100644
--- a/include/linux/dax.h
+++ b/include/linux/dax.h
@@ -40,10 +40,11 @@ static inline void put_dax(struct dax_device *dax_dev)
int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff);
#if IS_ENABLED(CONFIG_FS_DAX)
-int __bdev_dax_supported(struct super_block *sb, int blocksize);
-static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
+int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
+ int blocksize);
+static inline int sb_dax_supported(struct super_block *sb, int blocksize)
{
- return __bdev_dax_supported(sb, blocksize);
+ return bdev_dax_supported(sb, sb->s_bdev, blocksize);
}
static inline struct dax_device *fs_dax_get_by_host(const char *host)
@@ -58,7 +59,14 @@ static inline void fs_put_dax(struct dax_device *dax_dev)
struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev);
#else
-static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
+static inline int bdev_dax_supported(struct super_block *sb,
+ struct block_device *bdev,
+ int blocksize)
+{
+ return -EOPNOTSUPP;
+}
+
+static inline int sb_dax_supported(struct super_block *sb, int blocksize)
{
return -EOPNOTSUPP;
}
On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> When using realtime device (rtdev) with xfs where the data device is not
> DAX capable, two issues arise. One is when data device is not DAX but the
> realtime device is DAX capable, we currently disable DAX.
> After passing this check, we are also not marking the inode as DAX capable.
> This change will allow DAX enabled if the data device or the realtime
> device is DAX capable. S_DAX will be marked for the inode if the file is
> residing on a DAX capable device. This will prevent the case of rtdev is not
> DAX and data device is DAX to create realtime files.
I'm confused by this description. I'm not sure what is broken, nor
what you are trying to fix.
I think what you want to do is enable DAX on RT devices separately
to the data device and vice versa?
i.e. is this what you are trying to acheive?
datadev dax rtdev dax DAX enabled on
----------- --------- --------------
no no neither
yes no datadev
no yes rtdev
yes yes both
>
> Signed-off-by: Dave Jiang <[email protected]>
> Reported-by: Darrick Wong <[email protected]>
> ---
> fs/xfs/xfs_iops.c | 3 ++-
> fs/xfs/xfs_super.c | 9 ++++++++-
> 2 files changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index 56475fcd76f2..ab352c325301 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> !xfs_is_reflink_inode(ip) &&
> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
This does not discriminate between the rtdev or the data dev. This
needs to call xfs_find_bdev_for_inode() to get the right device
for the inode config.
Further, if we add or remove the RT flag to the inode at a later
point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
flag at that point in time.
Which brings me to the real problem here: dynamically changing the
S_DAX flag is racy, dangerous and broken. It's not clear that this
should be allowed at all as the inode may have already been mmap()d
by the time the ioctl is called to set/clear the rt file state.
IOWs, right now we cannot support mixed DAX mode filesystems because
the generic DAX code does not support dynamic changing of the DAX
flag on an inode and so checking the block device state here is
irrelevant....
> inode->i_flags |= S_DAX;
> }
>
> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> index e8a687232614..5ac478924dce 100644
> --- a/fs/xfs/xfs_super.c
> +++ b/fs/xfs/xfs_super.c
> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
> sb->s_flags |= SB_I_VERSION;
>
> if (mp->m_flags & XFS_MOUNT_DAX) {
> + bool rtdev_is_dax = false;
> +
> xfs_warn(mp,
> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>
> + if (mp->m_rtdev_targp->bt_daxdev)
> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> + sb->s_blocksize) == 0)
> + rtdev_is_dax = true;
.... as this code here needs to turn off DAX here if any device
in the filesystem doesn't support DAX....
FWIW, the logic in the code is terrible (not your fault, Dave).
The logic reads
if (NOT bdev_dax_supported(rtdev)) then
rtdev supports DAX
That also needs fixing - we're checking something that has a boolean
return state (yes or no) and so it should define them in a way that
makes the caller logic read cleanly....
Cheers,
Dave.
--
Dave Chinner
[email protected]
On 02/01/2018 04:28 PM, Darrick J. Wong wrote:
>> [PATCH 2 2/2] xfs: fix rt_dev usage for DAX
>
> "[PATCH v2 2/2]" to distinguish the version number from the patch number
> more explicitly.
>
> On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
>> When using realtime device (rtdev) with xfs where the data device is not
>> DAX capable, two issues arise. One is when data device is not DAX but the
>> realtime device is DAX capable, we currently disable DAX.
>> After passing this check, we are also not marking the inode as DAX capable.
>> This change will allow DAX enabled if the data device or the realtime
>> device is DAX capable. S_DAX will be marked for the inode if the file is
>> residing on a DAX capable device. This will prevent the case of rtdev is not
>> DAX and data device is DAX to create realtime files.
>>
>> Signed-off-by: Dave Jiang <[email protected]>
>> Reported-by: Darrick Wong <[email protected]>
>> ---
>> fs/xfs/xfs_iops.c | 3 ++-
>> fs/xfs/xfs_super.c | 9 ++++++++-
>> 2 files changed, 10 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
>> index 56475fcd76f2..ab352c325301 100644
>> --- a/fs/xfs/xfs_iops.c
>> +++ b/fs/xfs/xfs_iops.c
>> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
>> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
>> !xfs_is_reflink_inode(ip) &&
>> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
>> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
>> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
>> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
>
> inode->i_sb->s_bdev is the data device bdev, so if the inode is a
> realtime file, we're checking the wrong device for daxiness, I think.
>
> Maybe this whole ugly switch statement should get turned into a helper
> function?
>
> xfs_ioctl_setattr_dax_invalidate needs to pick the right bdev to check.
>
>> inode->i_flags |= S_DAX;
>> }
>>
>> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
>> index e8a687232614..5ac478924dce 100644
>> --- a/fs/xfs/xfs_super.c
>> +++ b/fs/xfs/xfs_super.c
>> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
>> sb->s_flags |= SB_I_VERSION;
>>
>> if (mp->m_flags & XFS_MOUNT_DAX) {
>> + bool rtdev_is_dax = false;
>> +
>> xfs_warn(mp,
>> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>>
>> + if (mp->m_rtdev_targp->bt_daxdev)
>> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
>> + sb->s_blocksize) == 0)
>> + rtdev_is_dax = true;
>> +
>> error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
>> - if (error) {
>> + if (error && !rtdev_is_dax) {
>> xfs_alert(mp,
>> "DAX unsupported by block device. Turning off DAX.");
>> mp->m_flags &= ~XFS_MOUNT_DAX;
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
>> the body of a message to [email protected]
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
> Does the following patch fix everything for you?
>
> (Note that we can't switch S_DAX on a running fs so you have to remount
> the whole fs after setting the dax flag...)
Yes this passes my tests. However it looks like Dave Chinner has
additional concerns with regards to changing the S_DAX flag dynamically?
>
> --D
>
> --------------------
>
> fs: allow per-device dax status checking for filesystems
>
> Refactor __bdev_dax_supported into a sb_dax_supported helper for
> single-bdev filesystems and a regular bdev_dax_supported that takes a
> bdev parameter. This enables multi-device filesystems like xfs to check
> that a dax device can work for the particular filesystem. Once that's
> in place, actually fix all the parts of XFS where we need to be able to
> distinguish between datadev and rtdev.
>
> This patch fixes the problem where we screw up the dax support checking
> in xfs if the datadev and rtdev have different dax capabilities.
>
> Signed-off-by: Darrick J. Wong <[email protected]>
> ---
> drivers/dax/super.c | 9 +++++----
> fs/ext2/super.c | 2 +-
> fs/ext4/super.c | 2 +-
> fs/xfs/xfs_ioctl.c | 3 ++-
> fs/xfs/xfs_iops.c | 30 +++++++++++++++++++++++++-----
> fs/xfs/xfs_super.c | 11 +++++++++--
> include/linux/dax.h | 16 ++++++++++++----
> 7 files changed, 55 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/dax/super.c b/drivers/dax/super.c
> index 3ec8046..c4db84f 100644
> --- a/drivers/dax/super.c
> +++ b/drivers/dax/super.c
> @@ -72,8 +72,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> #endif
>
> /**
> - * __bdev_dax_supported() - Check if the device supports dax for filesystem
> + * bdev_dax_supported() - Check if the device supports dax for filesystem
> * @sb: The superblock of the device
> + * @bdev: block device to check
> * @blocksize: The block size of the device
> *
> * This is a library function for filesystems to check if the block device
> @@ -81,9 +82,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> *
> * Return: negative errno if unsupported, 0 if supported.
> */
> -int __bdev_dax_supported(struct super_block *sb, int blocksize)
> +int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
> + int blocksize)
> {
> - struct block_device *bdev = sb->s_bdev;
> struct dax_device *dax_dev;
> pgoff_t pgoff;
> int err, id;
> @@ -125,7 +126,7 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
>
> return 0;
> }
> -EXPORT_SYMBOL_GPL(__bdev_dax_supported);
> +EXPORT_SYMBOL_GPL(bdev_dax_supported);
> #endif
>
> enum dax_device_flags {
> diff --git a/fs/ext2/super.c b/fs/ext2/super.c
> index 7646818..6556993 100644
> --- a/fs/ext2/super.c
> +++ b/fs/ext2/super.c
> @@ -958,7 +958,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
> blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
>
> if (sbi->s_mount_opt & EXT2_MOUNT_DAX) {
> - err = bdev_dax_supported(sb, blocksize);
> + err = sb_dax_supported(sb, blocksize);
> if (err)
> goto failed_mount;
> }
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 7c46693..804a2d6 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -3712,7 +3712,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> " that may contain inline data");
> goto failed_mount;
> }
> - err = bdev_dax_supported(sb, blocksize);
> + err = sb_dax_supported(sb, blocksize);
> if (err)
> goto failed_mount;
> }
> diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
> index 89fb1eb..277355f 100644
> --- a/fs/xfs/xfs_ioctl.c
> +++ b/fs/xfs/xfs_ioctl.c
> @@ -1103,7 +1103,8 @@ xfs_ioctl_setattr_dax_invalidate(
> if (fa->fsx_xflags & FS_XFLAG_DAX) {
> if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
> return -EINVAL;
> - if (bdev_dax_supported(sb, sb->s_blocksize) < 0)
> + if (bdev_dax_supported(sb, xfs_find_bdev_for_inode(VFS_I(ip)),
> + sb->s_blocksize) < 0)
> return -EINVAL;
> }
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index 56475fc..66cd61c 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -1182,6 +1182,30 @@ static const struct inode_operations xfs_inline_symlink_inode_operations = {
> .update_time = xfs_vn_update_time,
> };
>
> +/* Figure out if this file actually supports DAX. */
> +static bool
> +xfs_inode_supports_dax(
> + struct xfs_inode *ip)
> +{
> + struct xfs_mount *mp = ip->i_mount;
> +
> + /* Only supported on non-reflinked files. */
> + if (!S_ISREG(VFS_I(ip)->i_mode) || xfs_is_reflink_inode(ip))
> + return false;
> +
> + /* DAX mount option or DAX iflag must be set. */
> + if (!(mp->m_flags & XFS_MOUNT_DAX) &&
> + !(ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> + return false;
> +
> + /* Block size must match page size */
> + if (mp->m_sb.sb_blocksize != PAGE_SIZE)
> + return false;
> +
> + /* Device has to support DAX too. */
> + return xfs_find_daxdev_for_inode(VFS_I(ip)) != NULL;
> +}
> +
> STATIC void
> xfs_diflags_to_iflags(
> struct inode *inode,
> @@ -1200,11 +1224,7 @@ xfs_diflags_to_iflags(
> inode->i_flags |= S_SYNC;
> if (flags & XFS_DIFLAG_NOATIME)
> inode->i_flags |= S_NOATIME;
> - if (S_ISREG(inode->i_mode) &&
> - ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> - !xfs_is_reflink_inode(ip) &&
> - (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> + if (xfs_inode_supports_dax(ip))
> inode->i_flags |= S_DAX;
> }
>
> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> index 6f1b917..c115bc7 100644
> --- a/fs/xfs/xfs_super.c
> +++ b/fs/xfs/xfs_super.c
> @@ -1692,11 +1692,18 @@ xfs_fs_fill_super(
> sb->s_flags |= SB_I_VERSION;
>
> if (mp->m_flags & XFS_MOUNT_DAX) {
> + int error2 = 0;
> +
> xfs_warn(mp,
> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>
> - error = bdev_dax_supported(sb, sb->s_blocksize);
> - if (error) {
> + error = bdev_dax_supported(sb, mp->m_ddev_targp->bt_bdev,
> + sb->s_blocksize);
> + if (mp->m_rtdev_targp)
> + error2 = bdev_dax_supported(sb,
> + mp->m_rtdev_targp->bt_bdev,
> + sb->s_blocksize);
> + if (error && error2) {
> xfs_alert(mp,
> "DAX unsupported by block device. Turning off DAX.");
> mp->m_flags &= ~XFS_MOUNT_DAX;
> diff --git a/include/linux/dax.h b/include/linux/dax.h
> index 5258346..1107a98 100644
> --- a/include/linux/dax.h
> +++ b/include/linux/dax.h
> @@ -40,10 +40,11 @@ static inline void put_dax(struct dax_device *dax_dev)
>
> int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff);
> #if IS_ENABLED(CONFIG_FS_DAX)
> -int __bdev_dax_supported(struct super_block *sb, int blocksize);
> -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> +int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
> + int blocksize);
> +static inline int sb_dax_supported(struct super_block *sb, int blocksize)
> {
> - return __bdev_dax_supported(sb, blocksize);
> + return bdev_dax_supported(sb, sb->s_bdev, blocksize);
> }
>
> static inline struct dax_device *fs_dax_get_by_host(const char *host)
> @@ -58,7 +59,14 @@ static inline void fs_put_dax(struct dax_device *dax_dev)
>
> struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev);
> #else
> -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> +static inline int bdev_dax_supported(struct super_block *sb,
> + struct block_device *bdev,
> + int blocksize)
> +{
> + return -EOPNOTSUPP;
> +}
> +
> +static inline int sb_dax_supported(struct super_block *sb, int blocksize)
> {
> return -EOPNOTSUPP;
> }
>
On 02/01/2018 04:44 PM, Dave Chinner wrote:
> On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
>> When using realtime device (rtdev) with xfs where the data device is not
>> DAX capable, two issues arise. One is when data device is not DAX but the
>> realtime device is DAX capable, we currently disable DAX.
>> After passing this check, we are also not marking the inode as DAX capable.
>> This change will allow DAX enabled if the data device or the realtime
>> device is DAX capable. S_DAX will be marked for the inode if the file is
>> residing on a DAX capable device. This will prevent the case of rtdev is not
>> DAX and data device is DAX to create realtime files.
>
> I'm confused by this description. I'm not sure what is broken, nor
> what you are trying to fix.
>
> I think what you want to do is enable DAX on RT devices separately
> to the data device and vice versa?
>
> i.e. is this what you are trying to acheive?
>
> datadev dax rtdev dax DAX enabled on
> ----------- --------- --------------
> no no neither
> yes no datadev
> no yes rtdev
> yes yes both
^ Yes that's pretty much what I was trying to say. I probably should've
just provided the table above. Looks like Darrick supplied a much
cleaner patch. Although I don't think it addresses the concerns you have
with regards to dynamically changing the S_DAX flag.
>
>
>>
>> Signed-off-by: Dave Jiang <[email protected]>
>> Reported-by: Darrick Wong <[email protected]>
>> ---
>> fs/xfs/xfs_iops.c | 3 ++-
>> fs/xfs/xfs_super.c | 9 ++++++++-
>> 2 files changed, 10 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
>> index 56475fcd76f2..ab352c325301 100644
>> --- a/fs/xfs/xfs_iops.c
>> +++ b/fs/xfs/xfs_iops.c
>> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
>> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
>> !xfs_is_reflink_inode(ip) &&
>> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
>> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
>> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
>> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
>
> This does not discriminate between the rtdev or the data dev. This
> needs to call xfs_find_bdev_for_inode() to get the right device
> for the inode config.
>
> Further, if we add or remove the RT flag to the inode at a later
> point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
> flag at that point in time.
>
> Which brings me to the real problem here: dynamically changing the
> S_DAX flag is racy, dangerous and broken. It's not clear that this
> should be allowed at all as the inode may have already been mmap()d
> by the time the ioctl is called to set/clear the rt file state.
>
> IOWs, right now we cannot support mixed DAX mode filesystems because
> the generic DAX code does not support dynamic changing of the DAX
> flag on an inode and so checking the block device state here is
> irrelevant....
>
>> inode->i_flags |= S_DAX;
>> }
>>
>> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
>> index e8a687232614..5ac478924dce 100644
>> --- a/fs/xfs/xfs_super.c
>> +++ b/fs/xfs/xfs_super.c
>> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
>> sb->s_flags |= SB_I_VERSION;
>>
>> if (mp->m_flags & XFS_MOUNT_DAX) {
>> + bool rtdev_is_dax = false;
>> +
>> xfs_warn(mp,
>> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>>
>> + if (mp->m_rtdev_targp->bt_daxdev)
>> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
>> + sb->s_blocksize) == 0)
>> + rtdev_is_dax = true;
>
> .... as this code here needs to turn off DAX here if any device
> in the filesystem doesn't support DAX....
>
>
> FWIW, the logic in the code is terrible (not your fault, Dave).
> The logic reads
>
> if (NOT bdev_dax_supported(rtdev)) then
> rtdev supports DAX
>
> That also needs fixing - we're checking something that has a boolean
> return state (yes or no) and so it should define them in a way that
> makes the caller logic read cleanly....
>
> Cheers,
>
> Dave.
>
On Thu, Feb 01, 2018 at 05:08:36PM -0700, Dave Jiang wrote:
>
> On 02/01/2018 04:28 PM, Darrick J. Wong wrote:
> >> [PATCH 2 2/2] xfs: fix rt_dev usage for DAX
> >
> > "[PATCH v2 2/2]" to distinguish the version number from the patch number
> > more explicitly.
> >
> > On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> >> When using realtime device (rtdev) with xfs where the data device is not
> >> DAX capable, two issues arise. One is when data device is not DAX but the
> >> realtime device is DAX capable, we currently disable DAX.
> >> After passing this check, we are also not marking the inode as DAX capable.
> >> This change will allow DAX enabled if the data device or the realtime
> >> device is DAX capable. S_DAX will be marked for the inode if the file is
> >> residing on a DAX capable device. This will prevent the case of rtdev is not
> >> DAX and data device is DAX to create realtime files.
> >>
> >> Signed-off-by: Dave Jiang <[email protected]>
> >> Reported-by: Darrick Wong <[email protected]>
> >> ---
> >> fs/xfs/xfs_iops.c | 3 ++-
> >> fs/xfs/xfs_super.c | 9 ++++++++-
> >> 2 files changed, 10 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> >> index 56475fcd76f2..ab352c325301 100644
> >> --- a/fs/xfs/xfs_iops.c
> >> +++ b/fs/xfs/xfs_iops.c
> >> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> >> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> >> !xfs_is_reflink_inode(ip) &&
> >> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> >> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> >> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> >> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
> >
> > inode->i_sb->s_bdev is the data device bdev, so if the inode is a
> > realtime file, we're checking the wrong device for daxiness, I think.
> >
> > Maybe this whole ugly switch statement should get turned into a helper
> > function?
> >
> > xfs_ioctl_setattr_dax_invalidate needs to pick the right bdev to check.
> >
> >> inode->i_flags |= S_DAX;
> >> }
> >>
> >> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> >> index e8a687232614..5ac478924dce 100644
> >> --- a/fs/xfs/xfs_super.c
> >> +++ b/fs/xfs/xfs_super.c
> >> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
> >> sb->s_flags |= SB_I_VERSION;
> >>
> >> if (mp->m_flags & XFS_MOUNT_DAX) {
> >> + bool rtdev_is_dax = false;
> >> +
> >> xfs_warn(mp,
> >> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
> >>
> >> + if (mp->m_rtdev_targp->bt_daxdev)
> >> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> >> + sb->s_blocksize) == 0)
> >> + rtdev_is_dax = true;
> >> +
> >> error = bdev_dax_supported(sb->s_bdev, sb->s_blocksize);
> >> - if (error) {
> >> + if (error && !rtdev_is_dax) {
> >> xfs_alert(mp,
> >> "DAX unsupported by block device. Turning off DAX.");
> >> mp->m_flags &= ~XFS_MOUNT_DAX;
> >>
> >> --
> >> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> >> the body of a message to [email protected]
> >> More majordomo info at http://vger.kernel.org/majordomo-info.html
> >
> > Does the following patch fix everything for you?
> >
> > (Note that we can't switch S_DAX on a running fs so you have to remount
> > the whole fs after setting the dax flag...)
>
> Yes this passes my tests. However it looks like Dave Chinner has
> additional concerns with regards to changing the S_DAX flag dynamically?
The patch doesn't even touch /that/ part, other than updating the
bdev_dax_supported function call site. Dynamically changing S_DAX has
been disabled since 742d84290739 ("xfs: disable per-inode DAX flag") but
I was going to let the dax/pmem/mm developers sort that one out. In the
meantime we could at least probe the devices correctly.
This is turning into a series that refactors the functions; changes the
return value into the boolean that we actually care about; and then
fixes the xfs problems.
--D
>
>
> >
> > --D
> >
> > --------------------
> >
> > fs: allow per-device dax status checking for filesystems
> >
> > Refactor __bdev_dax_supported into a sb_dax_supported helper for
> > single-bdev filesystems and a regular bdev_dax_supported that takes a
> > bdev parameter. This enables multi-device filesystems like xfs to check
> > that a dax device can work for the particular filesystem. Once that's
> > in place, actually fix all the parts of XFS where we need to be able to
> > distinguish between datadev and rtdev.
> >
> > This patch fixes the problem where we screw up the dax support checking
> > in xfs if the datadev and rtdev have different dax capabilities.
> >
> > Signed-off-by: Darrick J. Wong <[email protected]>
> > ---
> > drivers/dax/super.c | 9 +++++----
> > fs/ext2/super.c | 2 +-
> > fs/ext4/super.c | 2 +-
> > fs/xfs/xfs_ioctl.c | 3 ++-
> > fs/xfs/xfs_iops.c | 30 +++++++++++++++++++++++++-----
> > fs/xfs/xfs_super.c | 11 +++++++++--
> > include/linux/dax.h | 16 ++++++++++++----
> > 7 files changed, 55 insertions(+), 18 deletions(-)
> >
> > diff --git a/drivers/dax/super.c b/drivers/dax/super.c
> > index 3ec8046..c4db84f 100644
> > --- a/drivers/dax/super.c
> > +++ b/drivers/dax/super.c
> > @@ -72,8 +72,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> > #endif
> >
> > /**
> > - * __bdev_dax_supported() - Check if the device supports dax for filesystem
> > + * bdev_dax_supported() - Check if the device supports dax for filesystem
> > * @sb: The superblock of the device
> > + * @bdev: block device to check
> > * @blocksize: The block size of the device
> > *
> > * This is a library function for filesystems to check if the block device
> > @@ -81,9 +82,9 @@ EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
> > *
> > * Return: negative errno if unsupported, 0 if supported.
> > */
> > -int __bdev_dax_supported(struct super_block *sb, int blocksize)
> > +int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
> > + int blocksize)
> > {
> > - struct block_device *bdev = sb->s_bdev;
> > struct dax_device *dax_dev;
> > pgoff_t pgoff;
> > int err, id;
> > @@ -125,7 +126,7 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
> >
> > return 0;
> > }
> > -EXPORT_SYMBOL_GPL(__bdev_dax_supported);
> > +EXPORT_SYMBOL_GPL(bdev_dax_supported);
> > #endif
> >
> > enum dax_device_flags {
> > diff --git a/fs/ext2/super.c b/fs/ext2/super.c
> > index 7646818..6556993 100644
> > --- a/fs/ext2/super.c
> > +++ b/fs/ext2/super.c
> > @@ -958,7 +958,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent)
> > blocksize = BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
> >
> > if (sbi->s_mount_opt & EXT2_MOUNT_DAX) {
> > - err = bdev_dax_supported(sb, blocksize);
> > + err = sb_dax_supported(sb, blocksize);
> > if (err)
> > goto failed_mount;
> > }
> > diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> > index 7c46693..804a2d6 100644
> > --- a/fs/ext4/super.c
> > +++ b/fs/ext4/super.c
> > @@ -3712,7 +3712,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> > " that may contain inline data");
> > goto failed_mount;
> > }
> > - err = bdev_dax_supported(sb, blocksize);
> > + err = sb_dax_supported(sb, blocksize);
> > if (err)
> > goto failed_mount;
> > }
> > diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
> > index 89fb1eb..277355f 100644
> > --- a/fs/xfs/xfs_ioctl.c
> > +++ b/fs/xfs/xfs_ioctl.c
> > @@ -1103,7 +1103,8 @@ xfs_ioctl_setattr_dax_invalidate(
> > if (fa->fsx_xflags & FS_XFLAG_DAX) {
> > if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)))
> > return -EINVAL;
> > - if (bdev_dax_supported(sb, sb->s_blocksize) < 0)
> > + if (bdev_dax_supported(sb, xfs_find_bdev_for_inode(VFS_I(ip)),
> > + sb->s_blocksize) < 0)
> > return -EINVAL;
> > }
> >
> > diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> > index 56475fc..66cd61c 100644
> > --- a/fs/xfs/xfs_iops.c
> > +++ b/fs/xfs/xfs_iops.c
> > @@ -1182,6 +1182,30 @@ static const struct inode_operations xfs_inline_symlink_inode_operations = {
> > .update_time = xfs_vn_update_time,
> > };
> >
> > +/* Figure out if this file actually supports DAX. */
> > +static bool
> > +xfs_inode_supports_dax(
> > + struct xfs_inode *ip)
> > +{
> > + struct xfs_mount *mp = ip->i_mount;
> > +
> > + /* Only supported on non-reflinked files. */
> > + if (!S_ISREG(VFS_I(ip)->i_mode) || xfs_is_reflink_inode(ip))
> > + return false;
> > +
> > + /* DAX mount option or DAX iflag must be set. */
> > + if (!(mp->m_flags & XFS_MOUNT_DAX) &&
> > + !(ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> > + return false;
> > +
> > + /* Block size must match page size */
> > + if (mp->m_sb.sb_blocksize != PAGE_SIZE)
> > + return false;
> > +
> > + /* Device has to support DAX too. */
> > + return xfs_find_daxdev_for_inode(VFS_I(ip)) != NULL;
> > +}
> > +
> > STATIC void
> > xfs_diflags_to_iflags(
> > struct inode *inode,
> > @@ -1200,11 +1224,7 @@ xfs_diflags_to_iflags(
> > inode->i_flags |= S_SYNC;
> > if (flags & XFS_DIFLAG_NOATIME)
> > inode->i_flags |= S_NOATIME;
> > - if (S_ISREG(inode->i_mode) &&
> > - ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> > - !xfs_is_reflink_inode(ip) &&
> > - (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> > - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> > + if (xfs_inode_supports_dax(ip))
> > inode->i_flags |= S_DAX;
> > }
> >
> > diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> > index 6f1b917..c115bc7 100644
> > --- a/fs/xfs/xfs_super.c
> > +++ b/fs/xfs/xfs_super.c
> > @@ -1692,11 +1692,18 @@ xfs_fs_fill_super(
> > sb->s_flags |= SB_I_VERSION;
> >
> > if (mp->m_flags & XFS_MOUNT_DAX) {
> > + int error2 = 0;
> > +
> > xfs_warn(mp,
> > "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
> >
> > - error = bdev_dax_supported(sb, sb->s_blocksize);
> > - if (error) {
> > + error = bdev_dax_supported(sb, mp->m_ddev_targp->bt_bdev,
> > + sb->s_blocksize);
> > + if (mp->m_rtdev_targp)
> > + error2 = bdev_dax_supported(sb,
> > + mp->m_rtdev_targp->bt_bdev,
> > + sb->s_blocksize);
> > + if (error && error2) {
> > xfs_alert(mp,
> > "DAX unsupported by block device. Turning off DAX.");
> > mp->m_flags &= ~XFS_MOUNT_DAX;
> > diff --git a/include/linux/dax.h b/include/linux/dax.h
> > index 5258346..1107a98 100644
> > --- a/include/linux/dax.h
> > +++ b/include/linux/dax.h
> > @@ -40,10 +40,11 @@ static inline void put_dax(struct dax_device *dax_dev)
> >
> > int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff);
> > #if IS_ENABLED(CONFIG_FS_DAX)
> > -int __bdev_dax_supported(struct super_block *sb, int blocksize);
> > -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> > +int bdev_dax_supported(struct super_block *sb, struct block_device *bdev,
> > + int blocksize);
> > +static inline int sb_dax_supported(struct super_block *sb, int blocksize)
> > {
> > - return __bdev_dax_supported(sb, blocksize);
> > + return bdev_dax_supported(sb, sb->s_bdev, blocksize);
> > }
> >
> > static inline struct dax_device *fs_dax_get_by_host(const char *host)
> > @@ -58,7 +59,14 @@ static inline void fs_put_dax(struct dax_device *dax_dev)
> >
> > struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev);
> > #else
> > -static inline int bdev_dax_supported(struct super_block *sb, int blocksize)
> > +static inline int bdev_dax_supported(struct super_block *sb,
> > + struct block_device *bdev,
> > + int blocksize)
> > +{
> > + return -EOPNOTSUPP;
> > +}
> > +
> > +static inline int sb_dax_supported(struct super_block *sb, int blocksize)
> > {
> > return -EOPNOTSUPP;
> > }
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Feb 02, 2018 at 10:44:13AM +1100, Dave Chinner wrote:
> On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> > When using realtime device (rtdev) with xfs where the data device is not
> > DAX capable, two issues arise. One is when data device is not DAX but the
> > realtime device is DAX capable, we currently disable DAX.
> > After passing this check, we are also not marking the inode as DAX capable.
> > This change will allow DAX enabled if the data device or the realtime
> > device is DAX capable. S_DAX will be marked for the inode if the file is
> > residing on a DAX capable device. This will prevent the case of rtdev is not
> > DAX and data device is DAX to create realtime files.
>
> I'm confused by this description. I'm not sure what is broken, nor
> what you are trying to fix.
>
> I think what you want to do is enable DAX on RT devices separately
> to the data device and vice versa?
>
> i.e. is this what you are trying to acheive?
>
> datadev dax rtdev dax DAX enabled on
> ----------- --------- --------------
> no no neither
> yes no datadev
> no yes rtdev
> yes yes both
>
>
> >
> > Signed-off-by: Dave Jiang <[email protected]>
> > Reported-by: Darrick Wong <[email protected]>
> > ---
> > fs/xfs/xfs_iops.c | 3 ++-
> > fs/xfs/xfs_super.c | 9 ++++++++-
> > 2 files changed, 10 insertions(+), 2 deletions(-)
> >
> > diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> > index 56475fcd76f2..ab352c325301 100644
> > --- a/fs/xfs/xfs_iops.c
> > +++ b/fs/xfs/xfs_iops.c
> > @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> > ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> > !xfs_is_reflink_inode(ip) &&
> > (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> > - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> > + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> > + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
>
> This does not discriminate between the rtdev or the data dev. This
> needs to call xfs_find_bdev_for_inode() to get the right device
> for the inode config.
>
> Further, if we add or remove the RT flag to the inode at a later
> point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
> flag at that point in time.
Ah, right, I'd missed that subtlety in my earlier replies. Ok, add
another patch to this series to reevaluate S_DAX when we change the RT
flag.
> Which brings me to the real problem here: dynamically changing the
> S_DAX flag is racy, dangerous and broken. It's not clear that this
> should be allowed at all as the inode may have already been mmap()d
> by the time the ioctl is called to set/clear the rt file state.
Agreed that this is a mess. Either this needs to get fixed in the dax
code, or we need to decide that we're not going to support reconfiguring
the dax flag at all, except possibly for empty files (similar to how we
restrict changes to the rt flag).
> IOWs, right now we cannot support mixed DAX mode filesystems because
> the generic DAX code does not support dynamic changing of the DAX
> flag on an inode and so checking the block device state here is
> irrelevant....
>
> > inode->i_flags |= S_DAX;
> > }
> >
> > diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> > index e8a687232614..5ac478924dce 100644
> > --- a/fs/xfs/xfs_super.c
> > +++ b/fs/xfs/xfs_super.c
> > @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
> > sb->s_flags |= SB_I_VERSION;
> >
> > if (mp->m_flags & XFS_MOUNT_DAX) {
> > + bool rtdev_is_dax = false;
> > +
> > xfs_warn(mp,
> > "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
> >
> > + if (mp->m_rtdev_targp->bt_daxdev)
> > + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> > + sb->s_blocksize) == 0)
> > + rtdev_is_dax = true;
>
> .... as this code here needs to turn off DAX here if any device
> in the filesystem doesn't support DAX....
I think it'd be useful to be able to have a pmem rt device even if the
data device doesn't support it. Or rather, I have a few clients who
have expressed interest in this sort of configuration.
--D
>
>
> FWIW, the logic in the code is terrible (not your fault, Dave).
> The logic reads
>
> if (NOT bdev_dax_supported(rtdev)) then
> rtdev supports DAX
>
> That also needs fixing - we're checking something that has a boolean
> return state (yes or no) and so it should define them in a way that
> makes the caller logic read cleanly....
>
> Cheers,
>
> Dave.
> --
> Dave Chinner
> [email protected]
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Feb 01, 2018 at 05:13:40PM -0700, Dave Jiang wrote:
>
>
> On 02/01/2018 04:44 PM, Dave Chinner wrote:
> > On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> >> When using realtime device (rtdev) with xfs where the data device is not
> >> DAX capable, two issues arise. One is when data device is not DAX but the
> >> realtime device is DAX capable, we currently disable DAX.
> >> After passing this check, we are also not marking the inode as DAX capable.
> >> This change will allow DAX enabled if the data device or the realtime
> >> device is DAX capable. S_DAX will be marked for the inode if the file is
> >> residing on a DAX capable device. This will prevent the case of rtdev is not
> >> DAX and data device is DAX to create realtime files.
> >
> > I'm confused by this description. I'm not sure what is broken, nor
> > what you are trying to fix.
> >
> > I think what you want to do is enable DAX on RT devices separately
> > to the data device and vice versa?
> >
> > i.e. is this what you are trying to acheive?
> >
> > datadev dax rtdev dax DAX enabled on
> > ----------- --------- --------------
> > no no neither
> > yes no datadev
> > no yes rtdev
> > yes yes both
>
> ^ Yes that's pretty much what I was trying to say. I probably should've
> just provided the table above.
Ok, good to know I understood what you want to achieve. Now we've
just got to work out how to do it. :)
Cheers,
Dave.
--
Dave Chinner
[email protected]
On Thu, Feb 01, 2018 at 04:43:32PM -0800, Darrick J. Wong wrote:
> On Fri, Feb 02, 2018 at 10:44:13AM +1100, Dave Chinner wrote:
> > On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> > > Signed-off-by: Dave Jiang <[email protected]>
> > > Reported-by: Darrick Wong <[email protected]>
> > > ---
> > > fs/xfs/xfs_iops.c | 3 ++-
> > > fs/xfs/xfs_super.c | 9 ++++++++-
> > > 2 files changed, 10 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> > > index 56475fcd76f2..ab352c325301 100644
> > > --- a/fs/xfs/xfs_iops.c
> > > +++ b/fs/xfs/xfs_iops.c
> > > @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> > > ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> > > !xfs_is_reflink_inode(ip) &&
> > > (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> > > - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> > > + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> > > + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
> >
> > This does not discriminate between the rtdev or the data dev. This
> > needs to call xfs_find_bdev_for_inode() to get the right device
> > for the inode config.
> >
> > Further, if we add or remove the RT flag to the inode at a later
> > point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
> > flag at that point in time.
>
> Ah, right, I'd missed that subtlety in my earlier replies. Ok, add
> another patch to this series to reevaluate S_DAX when we change the RT
> flag.
>
> > Which brings me to the real problem here: dynamically changing the
> > S_DAX flag is racy, dangerous and broken. It's not clear that this
> > should be allowed at all as the inode may have already been mmap()d
> > by the time the ioctl is called to set/clear the rt file state.
>
> Agreed that this is a mess. Either this needs to get fixed in the dax
> code, or we need to decide that we're not going to support reconfiguring
> the dax flag at all, except possibly for empty files (similar to how we
> restrict changes to the rt flag).
Hmmm, the rt flag limitations are to do with whether extents have
been allocated or not. IIRC the issue with S_DAX is whether the file
has been mmap()d or not which we can't check for reliably even if
the file is empty...
.....
> > > if (mp->m_flags & XFS_MOUNT_DAX) {
> > > + bool rtdev_is_dax = false;
> > > +
> > > xfs_warn(mp,
> > > "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
> > >
> > > + if (mp->m_rtdev_targp->bt_daxdev)
> > > + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> > > + sb->s_blocksize) == 0)
> > > + rtdev_is_dax = true;
> >
> > .... as this code here needs to turn off DAX here if any device
> > in the filesystem doesn't support DAX....
>
> I think it'd be useful to be able to have a pmem rt device even if the
> data device doesn't support it. Or rather, I have a few clients who
> have expressed interest in this sort of configuration.
Understood.
I suspect all this means we can only support DAX on the rt device by
setting the RT flag on file create, at least until the dynamic S_DAX
issues are sorted out. And that also means it can't be cleared from
files that have it set. So perhaps all we need to do is disallow
changing the RT flag on regular files via the ioctl if XFS_MOUNT_DAX
is set?
Cheers,
Dave.
--
Dave Chinner
[email protected]
On 02/01/2018 05:43 PM, Darrick J. Wong wrote:
> On Fri, Feb 02, 2018 at 10:44:13AM +1100, Dave Chinner wrote:
>> On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
>>> When using realtime device (rtdev) with xfs where the data device is not
>>> DAX capable, two issues arise. One is when data device is not DAX but the
>>> realtime device is DAX capable, we currently disable DAX.
>>> After passing this check, we are also not marking the inode as DAX capable.
>>> This change will allow DAX enabled if the data device or the realtime
>>> device is DAX capable. S_DAX will be marked for the inode if the file is
>>> residing on a DAX capable device. This will prevent the case of rtdev is not
>>> DAX and data device is DAX to create realtime files.
>>
>> I'm confused by this description. I'm not sure what is broken, nor
>> what you are trying to fix.
>>
>> I think what you want to do is enable DAX on RT devices separately
>> to the data device and vice versa?
>>
>> i.e. is this what you are trying to acheive?
>>
>> datadev dax rtdev dax DAX enabled on
>> ----------- --------- --------------
>> no no neither
>> yes no datadev
>> no yes rtdev
>> yes yes both
>>
>>
>>>
>>> Signed-off-by: Dave Jiang <[email protected]>
>>> Reported-by: Darrick Wong <[email protected]>
>>> ---
>>> fs/xfs/xfs_iops.c | 3 ++-
>>> fs/xfs/xfs_super.c | 9 ++++++++-
>>> 2 files changed, 10 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
>>> index 56475fcd76f2..ab352c325301 100644
>>> --- a/fs/xfs/xfs_iops.c
>>> +++ b/fs/xfs/xfs_iops.c
>>> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
>>> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
>>> !xfs_is_reflink_inode(ip) &&
>>> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
>>> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
>>> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
>>> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
>>
>> This does not discriminate between the rtdev or the data dev. This
>> needs to call xfs_find_bdev_for_inode() to get the right device
>> for the inode config.
>>
>> Further, if we add or remove the RT flag to the inode at a later
>> point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
>> flag at that point in time.
>
> Ah, right, I'd missed that subtlety in my earlier replies. Ok, add
> another patch to this series to reevaluate S_DAX when we change the RT
> flag.
>
>> Which brings me to the real problem here: dynamically changing the
>> S_DAX flag is racy, dangerous and broken. It's not clear that this
>> should be allowed at all as the inode may have already been mmap()d
>> by the time the ioctl is called to set/clear the rt file state.
>
> Agreed that this is a mess. Either this needs to get fixed in the dax
> code, or we need to decide that we're not going to support reconfiguring
> the dax flag at all, except possibly for empty files (similar to how we
> restrict changes to the rt flag).
Does this mean we should add a check in xfs_ioctl_setattr_xflags() to
reject removing of realtime flag if S_DAX is set on the inode until the
dynamic change issue is sorted out?
>
>> IOWs, right now we cannot support mixed DAX mode filesystems because
>> the generic DAX code does not support dynamic changing of the DAX
>> flag on an inode and so checking the block device state here is
>> irrelevant....
>>
>>> inode->i_flags |= S_DAX;
>>> }
>>>
>>> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
>>> index e8a687232614..5ac478924dce 100644
>>> --- a/fs/xfs/xfs_super.c
>>> +++ b/fs/xfs/xfs_super.c
>>> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
>>> sb->s_flags |= SB_I_VERSION;
>>>
>>> if (mp->m_flags & XFS_MOUNT_DAX) {
>>> + bool rtdev_is_dax = false;
>>> +
>>> xfs_warn(mp,
>>> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
>>>
>>> + if (mp->m_rtdev_targp->bt_daxdev)
>>> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
>>> + sb->s_blocksize) == 0)
>>> + rtdev_is_dax = true;
>>
>> .... as this code here needs to turn off DAX here if any device
>> in the filesystem doesn't support DAX....
>
> I think it'd be useful to be able to have a pmem rt device even if the
> data device doesn't support it. Or rather, I have a few clients who
> have expressed interest in this sort of configuration.
>
> --D
>
>>
>>
>> FWIW, the logic in the code is terrible (not your fault, Dave).
>> The logic reads
>>
>> if (NOT bdev_dax_supported(rtdev)) then
>> rtdev supports DAX
>>
>> That also needs fixing - we're checking something that has a boolean
>> return state (yes or no) and so it should define them in a way that
>> makes the caller logic read cleanly....
>>
>> Cheers,
>>
>> Dave.
>> --
>> Dave Chinner
>> [email protected]
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
>> the body of a message to [email protected]
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Feb 06, 2018 at 03:32:00PM -0700, Dave Jiang wrote:
> On 02/01/2018 05:43 PM, Darrick J. Wong wrote:
> > On Fri, Feb 02, 2018 at 10:44:13AM +1100, Dave Chinner wrote:
> >> On Thu, Feb 01, 2018 at 01:33:05PM -0700, Dave Jiang wrote:
> >>> When using realtime device (rtdev) with xfs where the data device is not
> >>> DAX capable, two issues arise. One is when data device is not DAX but the
> >>> realtime device is DAX capable, we currently disable DAX.
> >>> After passing this check, we are also not marking the inode as DAX capable.
> >>> This change will allow DAX enabled if the data device or the realtime
> >>> device is DAX capable. S_DAX will be marked for the inode if the file is
> >>> residing on a DAX capable device. This will prevent the case of rtdev is not
> >>> DAX and data device is DAX to create realtime files.
> >>
> >> I'm confused by this description. I'm not sure what is broken, nor
> >> what you are trying to fix.
> >>
> >> I think what you want to do is enable DAX on RT devices separately
> >> to the data device and vice versa?
> >>
> >> i.e. is this what you are trying to acheive?
> >>
> >> datadev dax rtdev dax DAX enabled on
> >> ----------- --------- --------------
> >> no no neither
> >> yes no datadev
> >> no yes rtdev
> >> yes yes both
> >>
> >>
> >>>
> >>> Signed-off-by: Dave Jiang <[email protected]>
> >>> Reported-by: Darrick Wong <[email protected]>
> >>> ---
> >>> fs/xfs/xfs_iops.c | 3 ++-
> >>> fs/xfs/xfs_super.c | 9 ++++++++-
> >>> 2 files changed, 10 insertions(+), 2 deletions(-)
> >>>
> >>> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> >>> index 56475fcd76f2..ab352c325301 100644
> >>> --- a/fs/xfs/xfs_iops.c
> >>> +++ b/fs/xfs/xfs_iops.c
> >>> @@ -1204,7 +1204,8 @@ xfs_diflags_to_iflags(
> >>> ip->i_mount->m_sb.sb_blocksize == PAGE_SIZE &&
> >>> !xfs_is_reflink_inode(ip) &&
> >>> (ip->i_mount->m_flags & XFS_MOUNT_DAX ||
> >>> - ip->i_d.di_flags2 & XFS_DIFLAG2_DAX))
> >>> + ip->i_d.di_flags2 & XFS_DIFLAG2_DAX) &&
> >>> + blk_queue_dax(bdev_get_queue(inode->i_sb->s_bdev)))
> >>
> >> This does not discriminate between the rtdev or the data dev. This
> >> needs to call xfs_find_bdev_for_inode() to get the right device
> >> for the inode config.
> >>
> >> Further, if we add or remove the RT flag to the inode at a later
> >> point in time (e.g. via ioctl) we also need to re-evaluate the S_DAX
> >> flag at that point in time.
> >
> > Ah, right, I'd missed that subtlety in my earlier replies. Ok, add
> > another patch to this series to reevaluate S_DAX when we change the RT
> > flag.
> >
> >> Which brings me to the real problem here: dynamically changing the
> >> S_DAX flag is racy, dangerous and broken. It's not clear that this
> >> should be allowed at all as the inode may have already been mmap()d
> >> by the time the ioctl is called to set/clear the rt file state.
>
> > Agreed that this is a mess. Either this needs to get fixed in the dax
> > code, or we need to decide that we're not going to support reconfiguring
> > the dax flag at all, except possibly for empty files (similar to how we
> > restrict changes to the rt flag).
>
> Does this mean we should add a check in xfs_ioctl_setattr_xflags() to
> reject removing of realtime flag if S_DAX is set on the inode until the
> dynamic change issue is sorted out?
Well, I /was/ tiptoeing around this while trying to take care of the
other 4.16 stuff, but now that I'm done with merge window stuff I'll let
loose. :)
The last time I paid much attention to DAX was the thread "re-enable XFS
per-inode DAX"[1] last September. Motivating me to merge anything else
into DAX involves convincing me that we (mm, fs, dax developers) have
some kind of agreement about what we want the user-visible interfaces to
DAX to look like. Namely:
0. On what level do we allow users / administrators to control usage of
the dax paths? Can the hardware convey enough detail to the kernel that
the kernel can make a reasonable decision on its own whether buffered or
dax io make more sense? If so, can we please just have that? If not,
why?
1. If we want to let users override whatever decision the kernel makes,
how should we do this? One mount option that applies to everything,
like ext4? Inheritable inode flags, like xfs? Do we have one to force
it on even if the kernel doesn't want to? Do we have another to force
it off even if the kernel wants to? Do we even want to go down this
path? Can we get away with making the answer to Q0 "yes" and then see
if anyone actually complains about not having fine-grained control?
2. Under what conditions can we support dynamic changing of S_DAX on
inodes at runtime? Will this switching work at any time? Only for
files that are open but not mmap'd? Only for files that are empty?
3. The MAP_SYNC support that was merged into 4.15 -- is this sufficient
to allow this fsyncless clflush business that everyone seems to want?
4. Can someone please fix the XFS iomap_begin function to handle CoW
properly? I think it's a simple matter of allocate blocks, memcpy, and
remap, though I don't know how to do that. ;)
5. Do we test any of this stuff?
The thread from last September left off with promises to go define what
interface and behaviors we are providing to userspace, but afaict none
of that ever happened? If we don't resolve these questions before LSF
then I think what's needed is to lock everyone in a room to hash all
this out. :P
--D
PS: My personal inclination is {yes, get rid of all that until someone
complains, i think so but haven't tested it, ???, i sure hope so}.
[1] https://marc.info/?l=linux-xfs&m=150638135225793&w=2
>
> >
> >> IOWs, right now we cannot support mixed DAX mode filesystems because
> >> the generic DAX code does not support dynamic changing of the DAX
> >> flag on an inode and so checking the block device state here is
> >> irrelevant....
> >>
> >>> inode->i_flags |= S_DAX;
> >>> }
> >>>
> >>> diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> >>> index e8a687232614..5ac478924dce 100644
> >>> --- a/fs/xfs/xfs_super.c
> >>> +++ b/fs/xfs/xfs_super.c
> >>> @@ -1649,11 +1649,18 @@ xfs_fs_fill_super(
> >>> sb->s_flags |= SB_I_VERSION;
> >>>
> >>> if (mp->m_flags & XFS_MOUNT_DAX) {
> >>> + bool rtdev_is_dax = false;
> >>> +
> >>> xfs_warn(mp,
> >>> "DAX enabled. Warning: EXPERIMENTAL, use at your own risk");
> >>>
> >>> + if (mp->m_rtdev_targp->bt_daxdev)
> >>> + if (bdev_dax_supported(mp->m_rtdev_targp->bt_bdev,
> >>> + sb->s_blocksize) == 0)
> >>> + rtdev_is_dax = true;
> >>
> >> .... as this code here needs to turn off DAX here if any device
> >> in the filesystem doesn't support DAX....
> >
> > I think it'd be useful to be able to have a pmem rt device even if the
> > data device doesn't support it. Or rather, I have a few clients who
> > have expressed interest in this sort of configuration.
> >
> > --D
> >
> >>
> >>
> >> FWIW, the logic in the code is terrible (not your fault, Dave).
> >> The logic reads
> >>
> >> if (NOT bdev_dax_supported(rtdev)) then
> >> rtdev supports DAX
> >>
> >> That also needs fixing - we're checking something that has a boolean
> >> return state (yes or no) and so it should define them in a way that
> >> makes the caller logic read cleanly....
> >>
> >> Cheers,
> >>
> >> Dave.
> >> --
> >> Dave Chinner
> >> [email protected]
> >> --
> >> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> >> the body of a message to [email protected]
> >> More majordomo info at http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Feb 6, 2018 at 3:19 PM, Darrick J. Wong <[email protected]> wrote:
> On Tue, Feb 06, 2018 at 03:32:00PM -0700, Dave Jiang wrote:
[..]
> The last time I paid much attention to DAX was the thread "re-enable XFS
> per-inode DAX"[1] last September. Motivating me to merge anything else
> into DAX involves convincing me that we (mm, fs, dax developers) have
> some kind of agreement about what we want the user-visible interfaces to
> DAX to look like. Namely:
>
> 0. On what level do we allow users / administrators to control usage of
> the dax paths? Can the hardware convey enough detail to the kernel that
> the kernel can make a reasonable decision on its own whether buffered or
> dax io make more sense? If so, can we please just have that? If not,
> why?
>
> 1. If we want to let users override whatever decision the kernel makes,
> how should we do this? One mount option that applies to everything,
> like ext4? Inheritable inode flags, like xfs? Do we have one to force
> it on even if the kernel doesn't want to? Do we have another to force
> it off even if the kernel wants to? Do we even want to go down this
> path? Can we get away with making the answer to Q0 "yes" and then see
> if anyone actually complains about not having fine-grained control?
I think we will always have folks that want to force it on, i.e. the
MAP_SYNC user crowd. However, I think we might have some people what
will complain if they can't force it off. For example, I'm in the
process of killing off support for passing filesystem-dax mappings
through to guests because VFIO has the same "pin pages / fs-blocks
forever" problem as RDMA. Passing page cache through to a guest works
fine and it would be a shame if that silently stopped working in the
future. Given the page pinning constraint I'm not sure we can ever
support dynamically enabling DAX behind the user's back, at least not
until we kill off any "pin pages / fs-blocks forever" users, or
otherwise convert them to take a lease.
On Tue, Feb 06, 2018 at 03:19:15PM -0800, Darrick J. Wong wrote:
<>
> The last time I paid much attention to DAX was the thread "re-enable XFS
> per-inode DAX"[1] last September. Motivating me to merge anything else
> into DAX involves convincing me that we (mm, fs, dax developers) have
> some kind of agreement about what we want the user-visible interfaces to
> DAX to look like.
Yep, I agree that is the next step.
> Namely:
>
> 0. On what level do we allow users / administrators to control usage of
> the dax paths? Can the hardware convey enough detail to the kernel that
> the kernel can make a reasonable decision on its own whether buffered or
> dax io make more sense? If so, can we please just have that? If not,
> why?
Maybe eventually via the HMAT, but I don't think we have any systems today
that do a good job of this.
> 1. If we want to let users override whatever decision the kernel makes,
> how should we do this? One mount option that applies to everything,
> like ext4? Inheritable inode flags, like xfs? Do we have one to force
> it on even if the kernel doesn't want to? Do we have another to force
> it off even if the kernel wants to? Do we even want to go down this
> path? Can we get away with making the answer to Q0 "yes" and then see
> if anyone actually complains about not having fine-grained control?
I agree with Dan's assessment that even if we can make the kernel smart enough
to know when it's not a performance loss to use DAX (i.e. the persistent
memory you're using DAX on is just as fast as the page cache), users will
probably still want to retain the ability to force it on for use cases like
MAP_SYNC, and force it off for things like RDMA or VFIO, at least until the
page pinning work is complete.
Personally I'm still hopeful that we can have both the mount option and the
inheritable inode flags, and that we can figure out what we need to to get
S_DAX transitions happening again.
> 2. Under what conditions can we support dynamic changing of S_DAX on
> inodes at runtime? Will this switching work at any time? Only for
> files that are open but not mmap'd? Only for files that are empty?
>
> 3. The MAP_SYNC support that was merged into 4.15 -- is this sufficient
> to allow this fsyncless clflush business that everyone seems to want?
Yep, I think so. The next big battles are S_DAX transitions, per-inode DAX
support, and of course the page pinning / leases code that Dan & Christoph
have been talking about.
> 4. Can someone please fix the XFS iomap_begin function to handle CoW
> properly? I think it's a simple matter of allocate blocks, memcpy, and
> remap, though I don't know how to do that. ;)
>
> 5. Do we test any of this stuff?
Yes, I think in general we do a pretty good job of DAX test case coverage
between a combination of xfstests (which I have added to as I've fixed DAX
related bugs), nfit_test and the ndctl unit tests. hch has recently suggested
we start using blktests as well, though I don't think we've actually made any
new tests there yet. Suggestions on how we can get better test coverage are
welcome.
> The thread from last September left off with promises to go define what
> interface and behaviors we are providing to userspace, but afaict none
> of that ever happened? If we don't resolve these questions before LSF
> then I think what's needed is to lock everyone in a room to hash all
> this out. :P
Yep, that's accurate. I got pulled off onto other work and am just now
finding my way back. I think talking about it at LSF sounds great, but it's a
shame that hch won't be available. It'll be nice to finally meet dchinner,
though. :)
> --D
>
> PS: My personal inclination is {yes, get rid of all that until someone
> complains, i think so but haven't tested it, ???, i sure hope so}.
>
> [1] https://marc.info/?l=linux-xfs&m=150638135225793&w=2