linux-nvdimm.lists.01.org archive mirror
 help / color / mirror / Atom feed
From: Shiyang Ruan <ruansy.fnst@cn.fujitsu.com>
To: <linux-xfs@vger.kernel.org>, <linux-nvdimm@lists.01.org>,
	<darrick.wong@oracle.com>, <rgoldwyn@suse.de>,
	<hch@infradead.org>, <david@fromorbit.com>
Cc: linux-kernel@vger.kernel.org, gujx@cn.fujitsu.com, qi.fuli@fujitsu.com
Subject: [RFC PATCH v2 6/7] xfs: handle copy-on-write in fsdax write() path.
Date: Wed, 30 Oct 2019 12:13:57 +0800	[thread overview]
Message-ID: <20191030041358.14450-7-ruansy.fnst@cn.fujitsu.com> (raw)
In-Reply-To: <20191030041358.14450-1-ruansy.fnst@cn.fujitsu.com>

In fsdax mode, WRITE and ZERO on a shared extent need COW mechanism
performed.  After COW, new extents needs to be remapped to the file.

Signed-off-by: Shiyang Ruan <ruansy.fnst@cn.fujitsu.com>
---
 fs/xfs/xfs_bmap_util.c |  6 +++++-
 fs/xfs/xfs_file.c      | 10 +++++++---
 fs/xfs/xfs_iomap.c     |  3 ++-
 fs/xfs/xfs_iops.c      | 11 ++++++++---
 fs/xfs/xfs_reflink.c   |  2 ++
 5 files changed, 24 insertions(+), 8 deletions(-)

diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 99bf372ed551..1b7d0665d889 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1114,10 +1114,14 @@ xfs_free_file_space(
 	if (offset + len > XFS_ISIZE(ip))
 		len = XFS_ISIZE(ip) - offset;
 	error = iomap_zero_range(VFS_I(ip), offset, len, NULL,
-			&xfs_buffered_write_iomap_ops);
+		  IS_DAX(VFS_I(ip)) ?
+		  &xfs_direct_write_iomap_ops : &xfs_buffered_write_iomap_ops);
 	if (error)
 		return error;
 
+	if (xfs_is_reflink_inode(ip))
+		xfs_reflink_end_cow(ip, offset, len);
+
 	/*
 	 * If we zeroed right up to EOF and EOF straddles a page boundary we
 	 * must make sure that the post-EOF area is also zeroed because the
diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 24659667d5cb..48071f16a436 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -594,9 +594,13 @@ xfs_file_dax_write(
 
 	trace_xfs_file_dax_write(ip, count, pos);
 	ret = dax_iomap_rw(iocb, from, &xfs_direct_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);
+	if (ret > 0) {
+		if (iocb->ki_pos > i_size_read(inode)) {
+			i_size_write(inode, iocb->ki_pos);
+			error = xfs_setfilesize(ip, pos, ret);
+		}
+		if (xfs_is_cow_inode(ip))
+			xfs_reflink_end_cow(ip, pos, ret);
 	}
 out:
 	xfs_iunlock(ip, iolock);
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index e8fb500e1880..db8ee371252f 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -772,13 +772,14 @@ xfs_direct_write_iomap_begin(
 		goto out_unlock;
 
 	if (imap_needs_cow(ip, flags, &imap, nimaps)) {
+		bool need_convert = flags & IOMAP_DIRECT || IS_DAX(inode);
 		error = -EAGAIN;
 		if (flags & IOMAP_NOWAIT)
 			goto out_unlock;
 
 		/* may drop and re-acquire the ilock */
 		error = xfs_reflink_allocate_cow(ip, &imap, &cmap, &shared,
-				&lockmode, flags & IOMAP_DIRECT);
+				&lockmode, need_convert);
 		if (error)
 			goto out_unlock;
 		if (shared)
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 329a34af8e79..0052a97ccebf 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -834,6 +834,7 @@ xfs_setattr_size(
 	int			error;
 	uint			lock_flags = 0;
 	bool			did_zeroing = false;
+	const struct iomap_ops	*ops;
 
 	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
 	ASSERT(xfs_isilocked(ip, XFS_MMAPLOCK_EXCL));
@@ -880,13 +881,17 @@ xfs_setattr_size(
 	 * extension, or zeroing out the rest of the block on a downward
 	 * truncate.
 	 */
+	if (IS_DAX(inode))
+		ops = &xfs_direct_write_iomap_ops;
+	else
+		ops = &xfs_buffered_write_iomap_ops;
+
 	if (newsize > oldsize) {
 		trace_xfs_zero_eof(ip, oldsize, newsize - oldsize);
 		error = iomap_zero_range(inode, oldsize, newsize - oldsize,
-				&did_zeroing, &xfs_buffered_write_iomap_ops);
+				&did_zeroing, ops);
 	} else {
-		error = iomap_truncate_page(inode, newsize, &did_zeroing,
-				&xfs_buffered_write_iomap_ops);
+		error = iomap_truncate_page(inode, newsize, &did_zeroing, ops);
 	}
 
 	if (error)
diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c
index 2c8ad581d4db..e3620bc794a2 100644
--- a/fs/xfs/xfs_reflink.c
+++ b/fs/xfs/xfs_reflink.c
@@ -1270,6 +1270,8 @@ xfs_reflink_zero_posteof(
 
 	trace_xfs_zero_eof(ip, isize, pos - isize);
 	return iomap_zero_range(VFS_I(ip), isize, pos - isize, NULL,
+			IS_DAX(VFS_I(ip)) ?
+			&xfs_direct_write_iomap_ops :
 			&xfs_buffered_write_iomap_ops);
 }
 
-- 
2.23.0


_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

  parent reply	other threads:[~2019-10-30  4:14 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-30  4:13 [RFC PATCH v2 0/7] xfs: reflink & dedupe for fsdax (read/write path) Shiyang Ruan
2019-10-30  4:13 ` [RFC PATCH v2 1/7] dax: Introduce dax_copy_edges() for COW Shiyang Ruan
2019-10-30  4:13 ` [RFC PATCH v2 2/7] dax: copy data before write Shiyang Ruan
2019-10-30  4:13 ` [RFC PATCH v2 3/7] dax: replace mmap entry in case of CoW Shiyang Ruan
2019-10-30  4:13 ` [RFC PATCH v2 4/7] fs: dedup file range to use a compare function Shiyang Ruan
2019-10-30  4:13 ` [RFC PATCH v2 5/7] dax: memcpy before zeroing range Shiyang Ruan
2019-10-30  4:13 ` Shiyang Ruan [this message]
2019-10-30  4:13 ` [RFC PATCH v2 7/7] xfs: support dedupe for fsdax Shiyang Ruan
2019-10-30 11:48 ` [RFC PATCH v2 0/7] xfs: reflink & dedupe for fsdax (read/write path) Goldwyn Rodrigues
2019-10-31  4:54   ` Shiyang Ruan
2019-11-08  3:10 ` Shiyang Ruan
2019-11-08  3:30   ` Dan Williams
2019-11-14 20:24     ` Dave Chinner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20191030041358.14450-7-ruansy.fnst@cn.fujitsu.com \
    --to=ruansy.fnst@cn.fujitsu.com \
    --cc=darrick.wong@oracle.com \
    --cc=david@fromorbit.com \
    --cc=gujx@cn.fujitsu.com \
    --cc=hch@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nvdimm@lists.01.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=qi.fuli@fujitsu.com \
    --cc=rgoldwyn@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).