[v11,06/07] xfs: support CoW in fsdax mode

Commit Message

Shiyang Ruan May 8, 2022, 2:36 p.m. UTC

In fsdax mode, WRITE and ZERO on a shared extent need CoW performed.
After that, new allocated extents needs to be remapped to the file.
So, add a CoW identification in ->iomap_begin(), and implement
->iomap_end() to do the remapping work.

Signed-off-by: Shiyang Ruan <[e-mail]>
Reviewed-by: Darrick J. Wong <[e-mail]>
---
 fs/xfs/xfs_file.c  |  7 ++-----
 fs/xfs/xfs_iomap.c | 46 +++++++++++++++++++++++++++++++++++++++++++++-
 fs/xfs/xfs_iomap.h |  3 +++
 3 files changed, 50 insertions(+), 6 deletions(-)

Comments

Christoph Hellwig May 10, 2022, 5:45 a.m. UTC | #1

> +#ifdef CONFIG_FS_DAX
> +int
> +xfs_dax_fault(
> +	struct vm_fault		*vmf,
> +	enum page_entry_size	pe_size,
> +	bool			write_fault,
> +	pfn_t			*pfn)
> +{
> +	return dax_iomap_fault(vmf, pe_size, pfn, NULL,
> +			(write_fault && !vmf->cow_page) ?
> +				&xfs_dax_write_iomap_ops :
> +				&xfs_read_iomap_ops);
> +}
> +#endif

Is there any reason this is in xfs_iomap.c and not xfs_file.c?

Otherwise the patch looks good:


Reviewed-by: Christoph Hellwig <[e-mail]>

Shiyang Ruan May 10, 2022, 10:06 a.m. UTC | #2

在 2022/5/10 13:45, Christoph Hellwig 写道:
>> +#ifdef CONFIG_FS_DAX
>> +int
>> +xfs_dax_fault(
>> +	struct vm_fault		*vmf,
>> +	enum page_entry_size	pe_size,
>> +	bool			write_fault,
>> +	pfn_t			*pfn)
>> +{
>> +	return dax_iomap_fault(vmf, pe_size, pfn, NULL,
>> +			(write_fault && !vmf->cow_page) ?
>> +				&xfs_dax_write_iomap_ops :
>> +				&xfs_read_iomap_ops);
>> +}
>> +#endif
> 
> Is there any reason this is in xfs_iomap.c and not xfs_file.c?

Yes, It's better to put it in xfs_file.c since it's the only caller.  I 
didn't notice it...


--
Thanks,
Ruan.

> 
> Otherwise the patch looks good:
> 
> 
> Reviewed-by: Christoph Hellwig <[e-mail]>

Patch

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index af954a5b71f8..5a4508b23b51 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -669,7 +669,7 @@  xfs_file_dax_write(
 	pos = iocb->ki_pos;
 
 	trace_xfs_file_dax_write(iocb, from);
-	ret = dax_iomap_rw(iocb, from, &xfs_direct_write_iomap_ops);
+	ret = dax_iomap_rw(iocb, from, &xfs_dax_write_iomap_ops);
 	if (ret > 0 && iocb->ki_pos > i_size_read(inode)) {
 		i_size_write(inode, iocb->ki_pos);
 		error = xfs_setfilesize(ip, pos, ret);
@@ -1285,10 +1285,7 @@  __xfs_filemap_fault(
 		pfn_t pfn;
 
 		xfs_ilock(XFS_I(inode), XFS_MMAPLOCK_SHARED);
-		ret = dax_iomap_fault(vmf, pe_size, &pfn, NULL,
-				(write_fault && !vmf->cow_page) ?
-				 &xfs_direct_write_iomap_ops :
-				 &xfs_read_iomap_ops);
+		ret = xfs_dax_fault(vmf, pe_size, write_fault, &pfn);
 		if (ret & VM_FAULT_NEEDDSYNC)
 			ret = dax_finish_sync_fault(vmf, pe_size, pfn);
 		xfs_iunlock(XFS_I(inode), XFS_MMAPLOCK_SHARED);
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 5a393259a3a3..e35842215d22 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -27,6 +27,7 @@ 
 #include "xfs_dquot_item.h"
 #include "xfs_dquot.h"
 #include "xfs_reflink.h"
+#include "linux/dax.h"
 
 #define XFS_ALLOC_ALIGN(mp, off) \
 	(((off) >> mp->m_allocsize_log) << mp->m_allocsize_log)
@@ -773,7 +774,8 @@  xfs_direct_write_iomap_begin(
 
 		/* may drop and re-acquire the ilock */
 		error = xfs_reflink_allocate_cow(ip, &imap, &cmap, &shared,
-				&lockmode, flags & IOMAP_DIRECT);
+				&lockmode,
+				(flags & IOMAP_DIRECT) || IS_DAX(inode));
 		if (error)
 			goto out_unlock;
 		if (shared)
@@ -867,6 +869,33 @@  const struct iomap_ops xfs_direct_write_iomap_ops = {
 	.iomap_begin		= xfs_direct_write_iomap_begin,
 };
 
+static int
+xfs_dax_write_iomap_end(
+	struct inode		*inode,
+	loff_t			pos,
+	loff_t			length,
+	ssize_t			written,
+	unsigned		flags,
+	struct iomap		*iomap)
+{
+	struct xfs_inode	*ip = XFS_I(inode);
+
+	if (!xfs_is_cow_inode(ip))
+		return 0;
+
+	if (!written) {
+		xfs_reflink_cancel_cow_range(ip, pos, length, true);
+		return 0;
+	}
+
+	return xfs_reflink_end_cow(ip, pos, written);
+}
+
+const struct iomap_ops xfs_dax_write_iomap_ops = {
+	.iomap_begin	= xfs_direct_write_iomap_begin,
+	.iomap_end	= xfs_dax_write_iomap_end,
+};
+
 static int
 xfs_buffered_write_iomap_begin(
 	struct inode		*inode,
@@ -1358,3 +1387,18 @@  xfs_truncate_page(
 	return iomap_truncate_page(inode, pos, did_zero,
 				   &xfs_buffered_write_iomap_ops);
 }
+
+#ifdef CONFIG_FS_DAX
+int
+xfs_dax_fault(
+	struct vm_fault		*vmf,
+	enum page_entry_size	pe_size,
+	bool			write_fault,
+	pfn_t			*pfn)
+{
+	return dax_iomap_fault(vmf, pe_size, pfn, NULL,
+			(write_fault && !vmf->cow_page) ?
+				&xfs_dax_write_iomap_ops :
+				&xfs_read_iomap_ops);
+}
+#endif
diff --git a/fs/xfs/xfs_iomap.h b/fs/xfs/xfs_iomap.h
index e88dc162c785..89dfa3bb099f 100644
--- a/fs/xfs/xfs_iomap.h
+++ b/fs/xfs/xfs_iomap.h
@@ -25,6 +25,8 @@  int xfs_bmbt_to_iomap(struct xfs_inode *ip, struct iomap *iomap,
 int xfs_zero_range(struct xfs_inode *ip, loff_t pos, loff_t len,
 		bool *did_zero);
 int xfs_truncate_page(struct xfs_inode *ip, loff_t pos, bool *did_zero);
+int xfs_dax_fault(struct vm_fault *vmf, enum page_entry_size pe_size,
+		bool write_fault, pfn_t *pfn);
 
 static inline xfs_filblks_t
 xfs_aligned_fsb_count(
@@ -51,5 +53,6 @@  extern const struct iomap_ops xfs_direct_write_iomap_ops;
 extern const struct iomap_ops xfs_read_iomap_ops;
 extern const struct iomap_ops xfs_seek_iomap_ops;
 extern const struct iomap_ops xfs_xattr_iomap_ops;
+extern const struct iomap_ops xfs_dax_write_iomap_ops;
 
 #endif /* __XFS_IOMAP_H__*/