[17/17] xfs: support for synchronous DAX faults

Commit Message

Jan Kara Oct. 19, 2017, 12:58 p.m. UTC

From: Christoph Hellwig <[e-mail]>

Return IOMAP_F_DIRTY from xfs_file_iomap_begin() when asked to prepare
blocks for writing and the inode is pinned, and has dirty fields other
than the timestamps.  In __xfs_filemap_fault() we then detect this case
and call dax_finish_sync_fault() to make sure all metadata is committed,
and to insert the page table entry.

Note that this will also dirty corresponding radix tree entry which is
what we want - fsync(2) will still provide data integrity guarantees for
applications not using userspace flushing. And applications using
userspace flushing can avoid calling fsync(2) and thus avoid the
performance overhead.

[JK: Added VM_SYNC flag handling]

Signed-off-by: Christoph Hellwig <[e-mail]>
Signed-off-by: Jan Kara <[e-mail]>
---
 fs/xfs/xfs_file.c  | 17 ++++++++++++++++-
 fs/xfs/xfs_iomap.c |  5 +++++
 2 files changed, 21 insertions(+), 1 deletion(-)

Comments

Christoph Hellwig Oct. 19, 2017, 1:17 p.m. UTC | #1

> +#define XFS_MAP_SUPPORTED (LEGACY_MAP_MASK | MAP_SYNC)

I'd kill this define.  Also is there any good reason that we have
to add LEGACY_MAP_MASK instead of assuming it's supported in the
core?

>  #endif
>  	.mmap		= xfs_file_mmap,
> +	.mmap_supported_flags	= XFS_MAP_SUPPORTED,
>  	.open		= xfs_file_open,
>  	.release	= xfs_file_release,
>  	.fsync		= xfs_file_fsync,

I usually either reformat all members to be aligned again, or if that's
too much churn (in this case it probably is) just use a single space
before the = to minimize the alignment differences.

Otherwise your changes look good to me.

Patch

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 7c6b8def6eed..9b8058774af3 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -44,6 +44,7 @@ 
 #include <linux/falloc.h>
 #include <linux/pagevec.h>
 #include <linux/backing-dev.h>
+#include <linux/mman.h>
 
 static const struct vm_operations_struct xfs_file_vm_ops;
 
@@ -1040,7 +1041,11 @@  __xfs_filemap_fault(
 
 	xfs_ilock(XFS_I(inode), XFS_MMAPLOCK_SHARED);
 	if (IS_DAX(inode)) {
-		ret = dax_iomap_fault(vmf, pe_size, NULL, &xfs_iomap_ops);
+		pfn_t pfn;
+
+		ret = dax_iomap_fault(vmf, pe_size, &pfn, &xfs_iomap_ops);
+		if (ret & VM_FAULT_NEEDDSYNC)
+			ret = dax_finish_sync_fault(vmf, pe_size, pfn);
 	} else {
 		if (write_fault)
 			ret = iomap_page_mkwrite(vmf, &xfs_iomap_ops);
@@ -1131,6 +1136,13 @@  xfs_file_mmap(
 	struct file	*filp,
 	struct vm_area_struct *vma)
 {
+	/*
+	 * We don't support synchronous mappings for non-DAX files. At least
+	 * until someone comes with a sensible use case.
+	 */
+	if (!IS_DAX(file_inode(filp)) && (vma->vm_flags & VM_SYNC))
+		return -EOPNOTSUPP;
+
 	file_accessed(filp);
 	vma->vm_ops = &xfs_file_vm_ops;
 	if (IS_DAX(file_inode(filp)))
@@ -1138,6 +1150,8 @@  xfs_file_mmap(
 	return 0;
 }
 
+#define XFS_MAP_SUPPORTED (LEGACY_MAP_MASK | MAP_SYNC)
+
 const struct file_operations xfs_file_operations = {
 	.llseek		= xfs_file_llseek,
 	.read_iter	= xfs_file_read_iter,
@@ -1149,6 +1163,7 @@  const struct file_operations xfs_file_operations = {
 	.compat_ioctl	= xfs_file_compat_ioctl,
 #endif
 	.mmap		= xfs_file_mmap,
+	.mmap_supported_flags	= XFS_MAP_SUPPORTED,
 	.open		= xfs_file_open,
 	.release	= xfs_file_release,
 	.fsync		= xfs_file_fsync,
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index f179bdf1644d..b43be199fbdf 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -33,6 +33,7 @@ 
 #include "xfs_error.h"
 #include "xfs_trans.h"
 #include "xfs_trans_space.h"
+#include "xfs_inode_item.h"
 #include "xfs_iomap.h"
 #include "xfs_trace.h"
 #include "xfs_icache.h"
@@ -1086,6 +1087,10 @@  xfs_file_iomap_begin(
 		trace_xfs_iomap_found(ip, offset, length, 0, &imap);
 	}
 
+	if ((flags & IOMAP_WRITE) && xfs_ipincount(ip) &&
+	    (ip->i_itemp->ili_fsync_fields & ~XFS_ILOG_TIMESTAMP))
+		iomap->flags |= IOMAP_F_DIRTY;
+
 	xfs_bmbt_to_iomap(ip, iomap, &imap);
 
 	if (shared)