All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: david@fromorbit.com, darrick.wong@oracle.com
Cc: linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-btrfs@vger.kernel.org, ocfs2-devel@oss.oracle.com,
	sandeen@redhat.com
Subject: [PATCH 11/15] vfs: allow short clone and dedupe operations
Date: Thu, 04 Oct 2018 17:45:54 -0700	[thread overview]
Message-ID: <153870035471.29072.15690301164025555492.stgit@magnolia> (raw)
In-Reply-To: <153870027422.29072.7433543674436957232.stgit@magnolia>

From: Darrick J. Wong <darrick.wong@oracle.com>

Allow the clone and dedupe prep function to shorten the request if the
caller can handle it.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/read_write.c    |    8 ++------
 include/linux/fs.h |    2 +-
 mm/filemap.c       |   19 +++++++++++++++----
 3 files changed, 18 insertions(+), 11 deletions(-)


diff --git a/fs/read_write.c b/fs/read_write.c
index 4eaea52f70a8..292d68c2f47c 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1726,7 +1726,6 @@ int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
 {
 	struct inode *inode_in = file_inode(file_in);
 	struct inode *inode_out = file_inode(file_out);
-	uint64_t nlen;
 	loff_t isize;
 	bool same_inode = (inode_in == inode_out);
 	bool is_dedupe = (flags & CLONERANGE_DEDUPE);
@@ -1758,13 +1757,10 @@ int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
 	}
 
 	/* Check that we don't violate system file offset limits. */
-	nlen = *len;
-	ret = generic_clone_checks(file_in, pos_in, file_out, pos_out, &nlen,
-			is_dedupe);
+	ret = generic_clone_checks(file_in, pos_in, file_out, pos_out, len,
+			flags);
 	if (ret)
 		return ret;
-	if (nlen != *len)
-		return -EINVAL;
 
 	/* Wait for the completion of any pending IOs on both files */
 	inode_dio_wait(inode_in);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ae5685c31270..eb35363478e5 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2988,7 +2988,7 @@ extern int generic_file_readonly_mmap(struct file *, struct vm_area_struct *);
 extern ssize_t generic_write_checks(struct kiocb *, struct iov_iter *);
 extern int generic_clone_checks(struct file *file_in, loff_t pos_in,
 				struct file *file_out, loff_t pos_out,
-				uint64_t *count, bool is_dedupe);
+				uint64_t *count, unsigned int flags);
 extern ssize_t generic_file_read_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t __generic_file_write_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_write_iter(struct kiocb *, struct iov_iter *);
diff --git a/mm/filemap.c b/mm/filemap.c
index f74391721234..013451b8017f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2983,7 +2983,7 @@ EXPORT_SYMBOL(generic_write_checks);
  */
 int generic_clone_checks(struct file *file_in, loff_t pos_in,
 			 struct file *file_out, loff_t pos_out,
-			 uint64_t *req_count, bool is_dedupe)
+			 uint64_t *req_count, unsigned int flags)
 {
 	struct inode *inode_in = file_in->f_mapping->host;
 	struct inode *inode_out = file_out->f_mapping->host;
@@ -3005,7 +3005,7 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	size_out = i_size_read(inode_out);
 
 	/* Dedupe requires both ranges to be within EOF. */
-	if (is_dedupe &&
+	if ((flags & CLONERANGE_DEDUPE) &&
 	    (pos_in >= size_in || pos_in + count > size_in ||
 	     pos_out >= size_out || pos_out + count > size_out))
 		return -EINVAL;
@@ -3056,8 +3056,12 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	if (pos_in + count == size_in) {
 		bcount = ALIGN(size_in, bs) - pos_in;
 	} else {
-		if (!IS_ALIGNED(count, bs))
-			return -EINVAL;
+		if (!IS_ALIGNED(count, bs)) {
+			if (flags & CLONERANGE_SHORT)
+				count = ALIGN_DOWN(count, bs);
+			else
+				return -EINVAL;
+		}
 
 		bcount = count;
 	}
@@ -3068,6 +3072,13 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	    pos_out < pos_in + bcount)
 		return -EINVAL;
 
+	/*
+	 * We shortened the request but the caller can't deal with that, so
+	 * bounce the request back to userspace.
+	 */
+	if (*req_count != count && !(flags & CLONERANGE_SHORT))
+		return -EINVAL;
+
 	*req_count = count;
 	return 0;
 }


WARNING: multiple messages have this Message-ID
From: Darrick J. Wong <darrick.wong@oracle.com>
To: david@fromorbit.com, darrick.wong@oracle.com
Cc: linux-xfs@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-btrfs@vger.kernel.org, ocfs2-devel@oss.oracle.com,
	sandeen@redhat.com
Subject: [Ocfs2-devel] [PATCH 11/15] vfs: allow short clone and dedupe operations
Date: Thu, 04 Oct 2018 17:45:54 -0700	[thread overview]
Message-ID: <153870035471.29072.15690301164025555492.stgit@magnolia> (raw)
In-Reply-To: <153870027422.29072.7433543674436957232.stgit@magnolia>

From: Darrick J. Wong <darrick.wong@oracle.com>

Allow the clone and dedupe prep function to shorten the request if the
caller can handle it.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/read_write.c    |    8 ++------
 include/linux/fs.h |    2 +-
 mm/filemap.c       |   19 +++++++++++++++----
 3 files changed, 18 insertions(+), 11 deletions(-)


diff --git a/fs/read_write.c b/fs/read_write.c
index 4eaea52f70a8..292d68c2f47c 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1726,7 +1726,6 @@ int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
 {
 	struct inode *inode_in = file_inode(file_in);
 	struct inode *inode_out = file_inode(file_out);
-	uint64_t nlen;
 	loff_t isize;
 	bool same_inode = (inode_in == inode_out);
 	bool is_dedupe = (flags & CLONERANGE_DEDUPE);
@@ -1758,13 +1757,10 @@ int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
 	}
 
 	/* Check that we don't violate system file offset limits. */
-	nlen = *len;
-	ret = generic_clone_checks(file_in, pos_in, file_out, pos_out, &nlen,
-			is_dedupe);
+	ret = generic_clone_checks(file_in, pos_in, file_out, pos_out, len,
+			flags);
 	if (ret)
 		return ret;
-	if (nlen != *len)
-		return -EINVAL;
 
 	/* Wait for the completion of any pending IOs on both files */
 	inode_dio_wait(inode_in);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ae5685c31270..eb35363478e5 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2988,7 +2988,7 @@ extern int generic_file_readonly_mmap(struct file *, struct vm_area_struct *);
 extern ssize_t generic_write_checks(struct kiocb *, struct iov_iter *);
 extern int generic_clone_checks(struct file *file_in, loff_t pos_in,
 				struct file *file_out, loff_t pos_out,
-				uint64_t *count, bool is_dedupe);
+				uint64_t *count, unsigned int flags);
 extern ssize_t generic_file_read_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t __generic_file_write_iter(struct kiocb *, struct iov_iter *);
 extern ssize_t generic_file_write_iter(struct kiocb *, struct iov_iter *);
diff --git a/mm/filemap.c b/mm/filemap.c
index f74391721234..013451b8017f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2983,7 +2983,7 @@ EXPORT_SYMBOL(generic_write_checks);
  */
 int generic_clone_checks(struct file *file_in, loff_t pos_in,
 			 struct file *file_out, loff_t pos_out,
-			 uint64_t *req_count, bool is_dedupe)
+			 uint64_t *req_count, unsigned int flags)
 {
 	struct inode *inode_in = file_in->f_mapping->host;
 	struct inode *inode_out = file_out->f_mapping->host;
@@ -3005,7 +3005,7 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	size_out = i_size_read(inode_out);
 
 	/* Dedupe requires both ranges to be within EOF. */
-	if (is_dedupe &&
+	if ((flags & CLONERANGE_DEDUPE) &&
 	    (pos_in >= size_in || pos_in + count > size_in ||
 	     pos_out >= size_out || pos_out + count > size_out))
 		return -EINVAL;
@@ -3056,8 +3056,12 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	if (pos_in + count == size_in) {
 		bcount = ALIGN(size_in, bs) - pos_in;
 	} else {
-		if (!IS_ALIGNED(count, bs))
-			return -EINVAL;
+		if (!IS_ALIGNED(count, bs)) {
+			if (flags & CLONERANGE_SHORT)
+				count = ALIGN_DOWN(count, bs);
+			else
+				return -EINVAL;
+		}
 
 		bcount = count;
 	}
@@ -3068,6 +3072,13 @@ int generic_clone_checks(struct file *file_in, loff_t pos_in,
 	    pos_out < pos_in + bcount)
 		return -EINVAL;
 
+	/*
+	 * We shortened the request but the caller can't deal with that, so
+	 * bounce the request back to userspace.
+	 */
+	if (*req_count != count && !(flags & CLONERANGE_SHORT))
+		return -EINVAL;
+
 	*req_count = count;
 	return 0;
 }

  parent reply	other threads:[~2018-10-05  0:46 UTC|newest]

Thread overview: 82+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-10-05  0:44 [PATCH 00/15] fs: fixes for serious clone/dedupe problems Darrick J. Wong
2018-10-05  0:44 ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:44 ` [PATCH 01/15] xfs: add a per-xfs trace_printk macro Darrick J. Wong
2018-10-05  0:44   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:44 ` [PATCH 02/15] xfs: refactor clonerange preparation into a separate helper Darrick J. Wong
2018-10-05  0:44   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  5:28   ` Dave Chinner
2018-10-05  5:28     ` [Ocfs2-devel] " Dave Chinner
2018-10-05 17:06     ` Darrick J. Wong
2018-10-05 17:06       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-06 10:30     ` Christoph Hellwig
2018-10-06 10:30       ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  7:02   ` Dave Chinner
2018-10-05  7:02     ` [Ocfs2-devel] " Dave Chinner
2018-10-05  9:02     ` Dave Chinner
2018-10-05  9:02       ` [Ocfs2-devel] " Dave Chinner
2018-10-05 17:21       ` Darrick J. Wong
2018-10-05 17:21         ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05 23:42         ` Dave Chinner
2018-10-05 23:42           ` [Ocfs2-devel] " Dave Chinner
2018-10-05  0:44 ` [PATCH 03/15] xfs: zero posteof blocks when cloning above eof Darrick J. Wong
2018-10-05  0:44   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  5:28   ` Dave Chinner
2018-10-05  5:28     ` [Ocfs2-devel] " Dave Chinner
2018-10-06 10:34   ` Christoph Hellwig
2018-10-06 10:34     ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  0:45 ` [PATCH 04/15] xfs: update ctime and remove suid before cloning files Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  5:30   ` Dave Chinner
2018-10-05  5:30     ` [Ocfs2-devel] " Dave Chinner
2018-10-06 10:35   ` Christoph Hellwig
2018-10-06 10:35     ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  0:45 ` [PATCH 05/15] vfs: check file ranges " Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-06 10:38   ` Christoph Hellwig
2018-10-06 10:38     ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  0:45 ` [PATCH 06/15] vfs: strengthen checking of file range inputs to clone/dedupe range Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  6:10   ` Amir Goldstein
2018-10-05 17:36     ` Darrick J. Wong
2018-10-05 17:36       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:45 ` [PATCH 07/15] vfs: skip zero-length dedupe requests Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  8:39   ` Amir Goldstein
2018-10-06 10:39   ` Christoph Hellwig
2018-10-06 10:39     ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  0:45 ` [PATCH 08/15] vfs: change clone and dedupe range function pointers to return bytes completed Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  8:06   ` Amir Goldstein
2018-10-05 21:47     ` Darrick J. Wong
2018-10-05 21:47       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-06 10:41   ` Christoph Hellwig
2018-10-06 10:41     ` [Ocfs2-devel] " Christoph Hellwig
2018-10-08 18:59     ` Darrick J. Wong
2018-10-08 18:59       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:45 ` [PATCH 09/15] vfs: pass operation flags to {clone, dedupe}_file_range implementations Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  7:07   ` Amir Goldstein
2018-10-05 17:50     ` Darrick J. Wong
2018-10-05 17:50       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-06 10:44       ` Christoph Hellwig
2018-10-06 10:44         ` [Ocfs2-devel] " Christoph Hellwig
2018-10-05  0:45 ` [PATCH 10/15] vfs: make cloning to source file eof more explicit Darrick J. Wong
2018-10-05  0:45   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  6:47   ` Amir Goldstein
2018-10-05  0:45 ` Darrick J. Wong [this message]
2018-10-05  0:45   ` [Ocfs2-devel] [PATCH 11/15] vfs: allow short clone and dedupe operations Darrick J. Wong
2018-10-05  0:46 ` [PATCH 12/15] vfs: implement opportunistic short dedupe Darrick J. Wong
2018-10-05  0:46   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  6:40   ` Amir Goldstein
2018-10-05 17:42     ` Darrick J. Wong
2018-10-05 17:42       ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:46 ` [PATCH 13/15] ocfs2: truncate page cache for clone destination file before remapping Darrick J. Wong
2018-10-05  0:46   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:46 ` [PATCH 14/15] ocfs2: support partial clone range and dedupe range Darrick J. Wong
2018-10-05  0:46   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  0:46 ` [PATCH 15/15] xfs: support returning partial reflink results Darrick J. Wong
2018-10-05  0:46   ` [Ocfs2-devel] " Darrick J. Wong
2018-10-05  1:17 ` [PATCH 00/15] fs: fixes for serious clone/dedupe problems Dave Chinner
2018-10-05  1:17   ` [Ocfs2-devel] " Dave Chinner
2018-10-05  1:24   ` Darrick J. Wong
2018-10-05  1:24     ` [Ocfs2-devel] " Darrick J. Wong

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=153870035471.29072.15690301164025555492.stgit@magnolia \
    --to=darrick.wong@oracle.com \
    --cc=david@fromorbit.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=ocfs2-devel@oss.oracle.com \
    --cc=sandeen@redhat.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.