fiemap cleanups v4

fiemap cleanups v4

[Christoph Hellwig]

Hi all,

This series cleans up the fiemap support in ext4 and in general.

Ted or Al, can one of you pick this up?  It touches both ext4 and core
code, so either tree could work.


Changes since v3:
 - dropped the fixes that have been merged int mainline

Changes since v2:
 - commit message typo
 - doc updates
 - use d_inode in cifs
 - add a missing return statement in cifs
 - remove the filemap_write_and_wait call from ext4_ioctl_get_es_cache

Changes since v1:
 - rename fiemap_validate to fiemap_prep
 - lift FIEMAP_FLAG_SYNC handling to common code
 - add a new linux/fiemap.h header
 - remove __generic_block_fiemap
 - remove access_ok calls from fiemap and ext4

[PATCH 1/9] ext4: split _ext4_fiemap

[Christoph Hellwig]

The fiemap and EXT4_IOC_GET_ES_CACHE cases share almost no code, so split
them into entirely separate functions.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/extents.c | 72 +++++++++++++++++++++++------------------------
 1 file changed, 35 insertions(+), 37 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 2b4b94542e34d..d2a2a3ba5c44a 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4854,11 +4854,9 @@ static int ext4_fiemap_check_ranges(struct inode *inode, u64 start, u64 *len)
 	return 0;
 }
 
-static int _ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
-			__u64 start, __u64 len, bool from_es_cache)
+int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		u64 start, u64 len)
 {
-	ext4_lblk_t start_blk;
-	u32 ext4_fiemap_flags = FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR;
 	int error = 0;
 
 	if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) {
@@ -4868,10 +4866,7 @@ static int _ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
 	}
 
-	if (from_es_cache)
-		ext4_fiemap_flags &= FIEMAP_FLAG_XATTR;
-
-	if (fiemap_check_flags(fieinfo, ext4_fiemap_flags))
+	if (fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR))
 		return -EBADR;
 
 	/*
@@ -4885,40 +4880,20 @@ static int _ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 
 	if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
 		fieinfo->fi_flags &= ~FIEMAP_FLAG_XATTR;
-		error = iomap_fiemap(inode, fieinfo, start, len,
-				     &ext4_iomap_xattr_ops);
-	} else if (!from_es_cache) {
-		error = iomap_fiemap(inode, fieinfo, start, len,
-				     &ext4_iomap_report_ops);
-	} else {
-		ext4_lblk_t len_blks;
-		__u64 last_blk;
-
-		start_blk = start >> inode->i_sb->s_blocksize_bits;
-		last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
-		if (last_blk >= EXT_MAX_BLOCKS)
-			last_blk = EXT_MAX_BLOCKS-1;
-		len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
-
-		/*
-		 * Walk the extent tree gathering extent information
-		 * and pushing extents back to the user.
-		 */
-		error = ext4_fill_es_cache_info(inode, start_blk, len_blks,
-						fieinfo);
+		return iomap_fiemap(inode, fieinfo, start, len,
+				    &ext4_iomap_xattr_ops);
 	}
-	return error;
-}
 
-int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
-		__u64 start, __u64 len)
-{
-	return _ext4_fiemap(inode, fieinfo, start, len, false);
+	return iomap_fiemap(inode, fieinfo, start, len, &ext4_iomap_report_ops);
 }
 
 int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		      __u64 start, __u64 len)
 {
+	ext4_lblk_t start_blk, len_blks;
+	__u64 last_blk;
+	int error = 0;
+
 	if (ext4_has_inline_data(inode)) {
 		int has_inline;
 
@@ -4929,9 +4904,32 @@ int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			return 0;
 	}
 
-	return _ext4_fiemap(inode, fieinfo, start, len, true);
-}
+	if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) {
+		error = ext4_ext_precache(inode);
+		if (error)
+			return error;
+		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
+	}
+
+	if (fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC))
+		return -EBADR;
 
+	error = ext4_fiemap_check_ranges(inode, start, &len);
+	if (error)
+		return error;
+
+	start_blk = start >> inode->i_sb->s_blocksize_bits;
+	last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
+	if (last_blk >= EXT_MAX_BLOCKS)
+		last_blk = EXT_MAX_BLOCKS-1;
+	len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
+
+	/*
+	 * Walk the extent tree gathering extent information
+	 * and pushing extents back to the user.
+	 */
+	return ext4_fill_es_cache_info(inode, start_blk, len_blks, fieinfo);
+}
 
 /*
  * ext4_access_path:
-- 
2.26.2

[PATCH 2/9] ext4: remove the call to fiemap_check_flags in ext4_fiemap

[Christoph Hellwig]

iomap_fiemap already calls fiemap_check_flags first thing, so this
additional check is redundant.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/extents.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index d2a2a3ba5c44a..a41ae7c510170 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4866,9 +4866,6 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
 	}
 
-	if (fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR))
-		return -EBADR;
-
 	/*
 	 * For bitmap files the maximum size limit could be smaller than
 	 * s_maxbytes, so check len here manually instead of just relying on the
-- 
2.26.2

[PATCH 3/9] fs: mark __generic_block_fiemap static

[Christoph Hellwig]

There is no caller left outside of ioctl.c.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/ioctl.c         | 4 +---
 include/linux/fs.h | 4 ----
 2 files changed, 1 insertion(+), 7 deletions(-)

diff --git a/fs/ioctl.c b/fs/ioctl.c
index 5e80b40bc1b5c..8fe5131b1deea 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -307,8 +307,7 @@ static inline loff_t blk_to_logical(struct inode *inode, sector_t blk)
  * If you use this function directly, you need to do your own locking. Use
  * generic_block_fiemap if you want the locking done for you.
  */
-
-int __generic_block_fiemap(struct inode *inode,
+static int __generic_block_fiemap(struct inode *inode,
 			   struct fiemap_extent_info *fieinfo, loff_t start,
 			   loff_t len, get_block_t *get_block)
 {
@@ -453,7 +452,6 @@ int __generic_block_fiemap(struct inode *inode,
 
 	return ret;
 }
-EXPORT_SYMBOL(__generic_block_fiemap);
 
 /**
  * generic_block_fiemap - FIEMAP for block based inodes
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 45cc10cdf6ddd..69b7619eb83d0 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -3299,10 +3299,6 @@ static inline int vfs_fstat(int fd, struct kstat *stat)
 extern const char *vfs_get_link(struct dentry *, struct delayed_call *);
 extern int vfs_readlink(struct dentry *, char __user *, int);
 
-extern int __generic_block_fiemap(struct inode *inode,
-				  struct fiemap_extent_info *fieinfo,
-				  loff_t start, loff_t len,
-				  get_block_t *get_block);
 extern int generic_block_fiemap(struct inode *inode,
 				struct fiemap_extent_info *fieinfo, u64 start,
 				u64 len, get_block_t *get_block);
-- 
2.26.2

[PATCH 4/9] fs: move the fiemap definitions out of fs.h

[Christoph Hellwig]

No need to pull the fiemap definitions into almost every file in the
kernel build.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/bad_inode.c              |  1 +
 fs/btrfs/extent_io.h        |  1 +
 fs/cifs/inode.c             |  1 +
 fs/cifs/smb2ops.c           |  1 +
 fs/ext2/inode.c             |  1 +
 fs/ext4/ext4.h              |  1 +
 fs/f2fs/data.c              |  1 +
 fs/f2fs/inline.c            |  1 +
 fs/gfs2/inode.c             |  1 +
 fs/hpfs/file.c              |  1 +
 fs/ioctl.c                  |  1 +
 fs/iomap/fiemap.c           |  1 +
 fs/nilfs2/inode.c           |  1 +
 fs/overlayfs/inode.c        |  1 +
 fs/xfs/xfs_iops.c           |  1 +
 include/linux/fiemap.h      | 24 ++++++++++++++++++++++++
 include/linux/fs.h          | 19 +------------------
 include/uapi/linux/fiemap.h |  6 +++---
 18 files changed, 43 insertions(+), 21 deletions(-)
 create mode 100644 include/linux/fiemap.h

diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index 8035d2a445617..54f0ce4442720 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -15,6 +15,7 @@
 #include <linux/time.h>
 #include <linux/namei.h>
 #include <linux/poll.h>
+#include <linux/fiemap.h>
 
 static int bad_file_open(struct inode *inode, struct file *filp)
 {
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 2ed65bd0760ea..817698bc06693 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -5,6 +5,7 @@
 
 #include <linux/rbtree.h>
 #include <linux/refcount.h>
+#include <linux/fiemap.h>
 #include "ulist.h"
 
 /*
diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c
index 5d2965a237305..115d28e7c2197 100644
--- a/fs/cifs/inode.c
+++ b/fs/cifs/inode.c
@@ -25,6 +25,7 @@
 #include <linux/freezer.h>
 #include <linux/sched/signal.h>
 #include <linux/wait_bit.h>
+#include <linux/fiemap.h>
 
 #include <asm/div64.h>
 #include "cifsfs.h"
diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
index f829f4165d38c..09047f1ddfb66 100644
--- a/fs/cifs/smb2ops.c
+++ b/fs/cifs/smb2ops.c
@@ -12,6 +12,7 @@
 #include <linux/uuid.h>
 #include <linux/sort.h>
 #include <crypto/aead.h>
+#include <linux/fiemap.h>
 #include "cifsfs.h"
 #include "cifsglob.h"
 #include "smb2pdu.h"
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index c885cf7d724b4..0f12a0e8a8d97 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -36,6 +36,7 @@
 #include <linux/iomap.h>
 #include <linux/namei.h>
 #include <linux/uio.h>
+#include <linux/fiemap.h>
 #include "ext2.h"
 #include "acl.h"
 #include "xattr.h"
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index ad2dbf6e49245..06f97a3a943f6 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -36,6 +36,7 @@
 #include <crypto/hash.h>
 #include <linux/falloc.h>
 #include <linux/percpu-rwsem.h>
+#include <linux/fiemap.h>
 #ifdef __KERNEL__
 #include <linux/compat.h>
 #endif
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index cdf2f626bea7a..25abbbb65ba09 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -19,6 +19,7 @@
 #include <linux/uio.h>
 #include <linux/cleancache.h>
 #include <linux/sched/signal.h>
+#include <linux/fiemap.h>
 
 #include "f2fs.h"
 #include "node.h"
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index 4167e54081518..9686ffea177e7 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -8,6 +8,7 @@
 
 #include <linux/fs.h>
 #include <linux/f2fs_fs.h>
+#include <linux/fiemap.h>
 
 #include "f2fs.h"
 #include "node.h"
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 5acd3ce30759b..a1337bf31e49f 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -17,6 +17,7 @@
 #include <linux/crc32.h>
 #include <linux/iomap.h>
 #include <linux/security.h>
+#include <linux/fiemap.h>
 #include <linux/uaccess.h>
 
 #include "gfs2.h"
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index b36abf9cb345a..62959a8e43ad8 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -9,6 +9,7 @@
 
 #include "hpfs_fn.h"
 #include <linux/mpage.h>
+#include <linux/fiemap.h>
 
 #define BLOCKS(size) (((size) + 511) >> 9)
 
diff --git a/fs/ioctl.c b/fs/ioctl.c
index 8fe5131b1deea..3f300cc07dee4 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -18,6 +18,7 @@
 #include <linux/buffer_head.h>
 #include <linux/falloc.h>
 #include <linux/sched/signal.h>
+#include <linux/fiemap.h>
 
 #include "internal.h"
 
diff --git a/fs/iomap/fiemap.c b/fs/iomap/fiemap.c
index d55e8f491a5e5..0a807bbb2b4af 100644
--- a/fs/iomap/fiemap.c
+++ b/fs/iomap/fiemap.c
@@ -6,6 +6,7 @@
 #include <linux/compiler.h>
 #include <linux/fs.h>
 #include <linux/iomap.h>
+#include <linux/fiemap.h>
 
 struct fiemap_ctx {
 	struct fiemap_extent_info *fi;
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 671085512e0fd..6e1aca38931f3 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -14,6 +14,7 @@
 #include <linux/pagemap.h>
 #include <linux/writeback.h>
 #include <linux/uio.h>
+#include <linux/fiemap.h>
 #include "nilfs.h"
 #include "btnode.h"
 #include "segment.h"
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index 981f11ec51bc6..625e58da4e82a 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -10,6 +10,7 @@
 #include <linux/xattr.h>
 #include <linux/posix_acl.h>
 #include <linux/ratelimit.h>
+#include <linux/fiemap.h>
 #include "overlayfs.h"
 
 
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index f7a99b3bbcf7a..44c353998ac5c 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -25,6 +25,7 @@
 #include <linux/posix_acl.h>
 #include <linux/security.h>
 #include <linux/iversion.h>
+#include <linux/fiemap.h>
 
 /*
  * Directories have different lock order w.r.t. mmap_sem compared to regular
diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
new file mode 100644
index 0000000000000..240d4f7d9116a
--- /dev/null
+++ b/include/linux/fiemap.h
@@ -0,0 +1,24 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _LINUX_FIEMAP_H
+#define _LINUX_FIEMAP_H 1
+
+#include <uapi/linux/fiemap.h>
+#include <linux/fs.h>
+
+struct fiemap_extent_info {
+	unsigned int fi_flags;		/* Flags as passed from user */
+	unsigned int fi_extents_mapped;	/* Number of mapped extents */
+	unsigned int fi_extents_max;	/* Size of fiemap_extent array */
+	struct fiemap_extent __user *fi_extents_start; /* Start of
+							fiemap_extent array */
+};
+
+int fiemap_fill_next_extent(struct fiemap_extent_info *info, u64 logical,
+			    u64 phys, u64 len, u32 flags);
+int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
+
+int generic_block_fiemap(struct inode *inode,
+		struct fiemap_extent_info *fieinfo, u64 start, u64 len,
+		get_block_t *get_block);
+
+#endif /* _LINUX_FIEMAP_H 1 */
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 69b7619eb83d0..6bb0de9995012 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -24,7 +24,6 @@
 #include <linux/capability.h>
 #include <linux/semaphore.h>
 #include <linux/fcntl.h>
-#include <linux/fiemap.h>
 #include <linux/rculist_bl.h>
 #include <linux/atomic.h>
 #include <linux/shrinker.h>
@@ -48,6 +47,7 @@ struct backing_dev_info;
 struct bdi_writeback;
 struct bio;
 struct export_operations;
+struct fiemap_extent_info;
 struct hd_geometry;
 struct iovec;
 struct kiocb;
@@ -1745,19 +1745,6 @@ extern long compat_ptr_ioctl(struct file *file, unsigned int cmd,
 extern void inode_init_owner(struct inode *inode, const struct inode *dir,
 			umode_t mode);
 extern bool may_open_dev(const struct path *path);
-/*
- * VFS FS_IOC_FIEMAP helper definitions.
- */
-struct fiemap_extent_info {
-	unsigned int fi_flags;		/* Flags as passed from user */
-	unsigned int fi_extents_mapped;	/* Number of mapped extents */
-	unsigned int fi_extents_max;	/* Size of fiemap_extent array */
-	struct fiemap_extent __user *fi_extents_start; /* Start of
-							fiemap_extent array */
-};
-int fiemap_fill_next_extent(struct fiemap_extent_info *info, u64 logical,
-			    u64 phys, u64 len, u32 flags);
-int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
 
 /*
  * This is the "filldir" function type, used by readdir() to let
@@ -3299,10 +3286,6 @@ static inline int vfs_fstat(int fd, struct kstat *stat)
 extern const char *vfs_get_link(struct dentry *, struct delayed_call *);
 extern int vfs_readlink(struct dentry *, char __user *, int);
 
-extern int generic_block_fiemap(struct inode *inode,
-				struct fiemap_extent_info *fieinfo, u64 start,
-				u64 len, get_block_t *get_block);
-
 extern struct file_system_type *get_filesystem(struct file_system_type *fs);
 extern void put_filesystem(struct file_system_type *fs);
 extern struct file_system_type *get_fs_type(const char *name);
diff --git a/include/uapi/linux/fiemap.h b/include/uapi/linux/fiemap.h
index 8c0bc24d5d955..07c1cdcb715e8 100644
--- a/include/uapi/linux/fiemap.h
+++ b/include/uapi/linux/fiemap.h
@@ -9,8 +9,8 @@
  *          Andreas Dilger <adilger@sun.com>
  */
 
-#ifndef _LINUX_FIEMAP_H
-#define _LINUX_FIEMAP_H
+#ifndef _UAPI_LINUX_FIEMAP_H
+#define _UAPI_LINUX_FIEMAP_H
 
 #include <linux/types.h>
 
@@ -67,4 +67,4 @@ struct fiemap {
 #define FIEMAP_EXTENT_SHARED		0x00002000 /* Space shared with other
 						    * files. */
 
-#endif /* _LINUX_FIEMAP_H */
+#endif /* _UAPI_LINUX_FIEMAP_H */
-- 
2.26.2

[PATCH 5/9] iomap: fix the iomap_fiemap prototype

[Christoph Hellwig]

iomap_fiemap should take u64 start and len arguments, just like the
->fiemap prototype.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/iomap/fiemap.c     | 2 +-
 include/linux/iomap.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/iomap/fiemap.c b/fs/iomap/fiemap.c
index 0a807bbb2b4af..449705575acf9 100644
--- a/fs/iomap/fiemap.c
+++ b/fs/iomap/fiemap.c
@@ -66,7 +66,7 @@ iomap_fiemap_actor(struct inode *inode, loff_t pos, loff_t length, void *data,
 }
 
 int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi,
-		loff_t start, loff_t len, const struct iomap_ops *ops)
+		u64 start, u64 len, const struct iomap_ops *ops)
 {
 	struct fiemap_ctx ctx;
 	loff_t ret;
diff --git a/include/linux/iomap.h b/include/linux/iomap.h
index 8b09463dae0db..63db02528b702 100644
--- a/include/linux/iomap.h
+++ b/include/linux/iomap.h
@@ -178,7 +178,7 @@ int iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
 vm_fault_t iomap_page_mkwrite(struct vm_fault *vmf,
 			const struct iomap_ops *ops);
 int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
-		loff_t start, loff_t len, const struct iomap_ops *ops);
+		u64 start, u64 len, const struct iomap_ops *ops);
 loff_t iomap_seek_hole(struct inode *inode, loff_t offset,
 		const struct iomap_ops *ops);
 loff_t iomap_seek_data(struct inode *inode, loff_t offset,
-- 
2.26.2

[PATCH 6/9] fs: move fiemap range validation into the file systems instances

[Christoph Hellwig]

Replace fiemap_check_flags with a fiemap_prep helper that also takes the
inode and mapped range, and performs the sanity check and truncation
previously done in fiemap_check_range.  This way the validation is inside
the file system itself and thus properly works for the stacked overlayfs
case as well.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
---
 Documentation/filesystems/fiemap.txt | 12 +++---
 fs/btrfs/inode.c                     |  2 +-
 fs/cifs/smb2ops.c                    |  6 ++-
 fs/ext4/extents.c                    |  5 ++-
 fs/f2fs/data.c                       |  3 +-
 fs/ioctl.c                           | 63 +++++++++++-----------------
 fs/iomap/fiemap.c                    |  2 +-
 fs/nilfs2/inode.c                    |  2 +-
 fs/ocfs2/extent_map.c                |  3 +-
 include/linux/fiemap.h               |  3 +-
 10 files changed, 47 insertions(+), 54 deletions(-)

diff --git a/Documentation/filesystems/fiemap.txt b/Documentation/filesystems/fiemap.txt
index ac87e6fda842b..35c8571eccb6e 100644
--- a/Documentation/filesystems/fiemap.txt
+++ b/Documentation/filesystems/fiemap.txt
@@ -203,16 +203,18 @@ EINTR once fatal signal received.
 
 
 Flag checking should be done at the beginning of the ->fiemap callback via the
-fiemap_check_flags() helper:
+fiemap_prep() helper:
 
-int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
+int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		u64 start, u64 *len, u32 supported_flags);
 
 The struct fieinfo should be passed in as received from ioctl_fiemap(). The
 set of fiemap flags which the fs understands should be passed via fs_flags. If
-fiemap_check_flags finds invalid user flags, it will place the bad values in
+fiemap_prep finds invalid user flags, it will place the bad values in
 fieinfo->fi_flags and return -EBADR. If the file system gets -EBADR, from
-fiemap_check_flags(), it should immediately exit, returning that error back to
-ioctl_fiemap().
+fiemap_prep(), it should immediately exit, returning that error back to
+ioctl_fiemap().  Additionally the range is validate against the supported
+maximum file size.
 
 
 For each extent in the request range, the file system should call
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 320d1062068d3..1f1ec361089b3 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -8250,7 +8250,7 @@ static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 {
 	int	ret;
 
-	ret = fiemap_check_flags(fieinfo, BTRFS_FIEMAP_FLAGS);
+	ret = fiemap_prep(inode, fieinfo, start, &len, BTRFS_FIEMAP_FLAGS);
 	if (ret)
 		return ret;
 
diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
index 09047f1ddfb66..828e53e795c6d 100644
--- a/fs/cifs/smb2ops.c
+++ b/fs/cifs/smb2ops.c
@@ -3408,8 +3408,10 @@ static int smb3_fiemap(struct cifs_tcon *tcon,
 	int i, num, rc, flags, last_blob;
 	u64 next;
 
-	if (fiemap_check_flags(fei, FIEMAP_FLAG_SYNC))
-		return -EBADR;
+	rc = fiemap_prep(d_inode(cfile->dentry), fei, start, &len,
+			FIEMAP_FLAG_SYNC);
+	if (rc)
+		return rc;
 
 	xid = get_xid();
  again:
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index a41ae7c510170..41f73dea92cac 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4908,8 +4908,9 @@ int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
 	}
 
-	if (fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC))
-		return -EBADR;
+	error = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
+	if (error)
+		return error;
 
 	error = ext4_fiemap_check_ranges(inode, start, &len);
 	if (error)
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 25abbbb65ba09..03faafc591b17 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -1825,7 +1825,8 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			return ret;
 	}
 
-	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR);
+	ret = fiemap_prep(inode, fieinfo, start, &len,
+			FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR);
 	if (ret)
 		return ret;
 
diff --git a/fs/ioctl.c b/fs/ioctl.c
index 3f300cc07dee4..56bbf02209aef 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -149,61 +149,50 @@ int fiemap_fill_next_extent(struct fiemap_extent_info *fieinfo, u64 logical,
 EXPORT_SYMBOL(fiemap_fill_next_extent);
 
 /**
- * fiemap_check_flags - check validity of requested flags for fiemap
+ * fiemap_prep - check validity of requested flags for fiemap
+ * @inode:	Inode to operate on
  * @fieinfo:	Fiemap context passed into ->fiemap
- * @fs_flags:	Set of fiemap flags that the file system understands
+ * @start:	Start of the mapped range
+ * @len:	Length of the mapped range, can be truncated by this function.
+ * @supported_flags:	Set of fiemap flags that the file system understands
  *
- * Called from file system ->fiemap callback. This will compute the
- * intersection of valid fiemap flags and those that the fs supports. That
- * value is then compared against the user supplied flags. In case of bad user
- * flags, the invalid values will be written into the fieinfo structure, and
- * -EBADR is returned, which tells ioctl_fiemap() to return those values to
- * userspace. For this reason, a return code of -EBADR should be preserved.
+ * This function must be called from each ->fiemap instance to validate the
+ * fiemap request against the file system parameters.
  *
- * Returns 0 on success, -EBADR on bad flags.
+ * Returns 0 on success, or a negative error on failure.
  */
-int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags)
+int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		u64 start, u64 *len, u32 supported_flags)
 {
+	u64 maxbytes = inode->i_sb->s_maxbytes;
 	u32 incompat_flags;
 
-	incompat_flags = fieinfo->fi_flags & ~(FIEMAP_FLAGS_COMPAT & fs_flags);
-	if (incompat_flags) {
-		fieinfo->fi_flags = incompat_flags;
-		return -EBADR;
-	}
-	return 0;
-}
-EXPORT_SYMBOL(fiemap_check_flags);
-
-static int fiemap_check_ranges(struct super_block *sb,
-			       u64 start, u64 len, u64 *new_len)
-{
-	u64 maxbytes = (u64) sb->s_maxbytes;
-
-	*new_len = len;
-
-	if (len == 0)
+	if (*len == 0)
 		return -EINVAL;
-
 	if (start > maxbytes)
 		return -EFBIG;
 
 	/*
 	 * Shrink request scope to what the fs can actually handle.
 	 */
-	if (len > maxbytes || (maxbytes - len) < start)
-		*new_len = maxbytes - start;
+	if (*len > maxbytes || (maxbytes - *len) < start)
+		*len = maxbytes - start;
 
+	supported_flags &= FIEMAP_FLAGS_COMPAT;
+	incompat_flags = fieinfo->fi_flags & ~supported_flags;
+	if (incompat_flags) {
+		fieinfo->fi_flags = incompat_flags;
+		return -EBADR;
+	}
 	return 0;
 }
+EXPORT_SYMBOL(fiemap_prep);
 
 static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
 {
 	struct fiemap fiemap;
 	struct fiemap_extent_info fieinfo = { 0, };
 	struct inode *inode = file_inode(filp);
-	struct super_block *sb = inode->i_sb;
-	u64 len;
 	int error;
 
 	if (!inode->i_op->fiemap)
@@ -215,11 +204,6 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
 	if (fiemap.fm_extent_count > FIEMAP_MAX_EXTENTS)
 		return -EINVAL;
 
-	error = fiemap_check_ranges(sb, fiemap.fm_start, fiemap.fm_length,
-				    &len);
-	if (error)
-		return error;
-
 	fieinfo.fi_flags = fiemap.fm_flags;
 	fieinfo.fi_extents_max = fiemap.fm_extent_count;
 	fieinfo.fi_extents_start = ufiemap->fm_extents;
@@ -232,7 +216,8 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
 	if (fieinfo.fi_flags & FIEMAP_FLAG_SYNC)
 		filemap_write_and_wait(inode->i_mapping);
 
-	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start, len);
+	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start,
+			fiemap.fm_length);
 	fiemap.fm_flags = fieinfo.fi_flags;
 	fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;
 	if (copy_to_user(ufiemap, &fiemap, sizeof(fiemap)))
@@ -320,7 +305,7 @@ static int __generic_block_fiemap(struct inode *inode,
 	bool past_eof = false, whole_file = false;
 	int ret = 0;
 
-	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC);
+	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
 	if (ret)
 		return ret;
 
diff --git a/fs/iomap/fiemap.c b/fs/iomap/fiemap.c
index 449705575acf9..89dca4a97e4a2 100644
--- a/fs/iomap/fiemap.c
+++ b/fs/iomap/fiemap.c
@@ -75,7 +75,7 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi,
 	ctx.fi = fi;
 	ctx.prev.type = IOMAP_HOLE;
 
-	ret = fiemap_check_flags(fi, FIEMAP_FLAG_SYNC);
+	ret = fiemap_prep(inode, fi, start, &len, FIEMAP_FLAG_SYNC);
 	if (ret)
 		return ret;
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 6e1aca38931f3..052c2da11e4d7 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -1006,7 +1006,7 @@ int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	unsigned int blkbits = inode->i_blkbits;
 	int ret, n;
 
-	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC);
+	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
 	if (ret)
 		return ret;
 
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index e3e2d1b2af51a..3744179b73fa1 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -746,7 +746,8 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	struct buffer_head *di_bh = NULL;
 	struct ocfs2_extent_rec rec;
 
-	ret = fiemap_check_flags(fieinfo, OCFS2_FIEMAP_FLAGS);
+	ret = fiemap_prep(inode, fieinfo, map_start, &map_len,
+			OCFS2_FIEMAP_FLAGS);
 	if (ret)
 		return ret;
 
diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
index 240d4f7d9116a..4e624c4665837 100644
--- a/include/linux/fiemap.h
+++ b/include/linux/fiemap.h
@@ -13,9 +13,10 @@ struct fiemap_extent_info {
 							fiemap_extent array */
 };
 
+int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
+		u64 start, u64 *len, u32 supported_flags);
 int fiemap_fill_next_extent(struct fiemap_extent_info *info, u64 logical,
 			    u64 phys, u64 len, u32 flags);
-int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
 
 int generic_block_fiemap(struct inode *inode,
 		struct fiemap_extent_info *fieinfo, u64 start, u64 len,
-- 
2.26.2

[PATCH 7/9] fs: handle FIEMAP_FLAG_SYNC in fiemap_prep

[Christoph Hellwig]

By moving FIEMAP_FLAG_SYNC handling to fiemap_prep we ensure it is
handled once instead of duplicated, but can still be done under fs locks,
like xfs/iomap intended with its duplicate handling.  Also make sure the
error value of filemap_write_and_wait is propagated to user space.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/btrfs/inode.c      |  4 +---
 fs/cifs/smb2ops.c     |  3 +--
 fs/ext4/extents.c     |  2 +-
 fs/ext4/ioctl.c       |  3 ---
 fs/f2fs/data.c        |  3 +--
 fs/ioctl.c            | 10 ++++++----
 fs/iomap/fiemap.c     |  8 +-------
 fs/nilfs2/inode.c     |  2 +-
 fs/ocfs2/extent_map.c |  5 +----
 fs/overlayfs/inode.c  |  4 ----
 10 files changed, 13 insertions(+), 31 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 1f1ec361089b3..529ffa5e7b452 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -8243,14 +8243,12 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	return ret;
 }
 
-#define BTRFS_FIEMAP_FLAGS	(FIEMAP_FLAG_SYNC)
-
 static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len)
 {
 	int	ret;
 
-	ret = fiemap_prep(inode, fieinfo, start, &len, BTRFS_FIEMAP_FLAGS);
+	ret = fiemap_prep(inode, fieinfo, start, &len, 0);
 	if (ret)
 		return ret;
 
diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
index 828e53e795c6d..300ade2acc41e 100644
--- a/fs/cifs/smb2ops.c
+++ b/fs/cifs/smb2ops.c
@@ -3408,8 +3408,7 @@ static int smb3_fiemap(struct cifs_tcon *tcon,
 	int i, num, rc, flags, last_blob;
 	u64 next;
 
-	rc = fiemap_prep(d_inode(cfile->dentry), fei, start, &len,
-			FIEMAP_FLAG_SYNC);
+	rc = fiemap_prep(d_inode(cfile->dentry), fei, start, &len, 0);
 	if (rc)
 		return rc;
 
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 41f73dea92cac..93574e88f6543 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4908,7 +4908,7 @@ int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
 	}
 
-	error = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
+	error = fiemap_prep(inode, fieinfo, start, &len, 0);
 	if (error)
 		return error;
 
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index 0746532ba463d..f81acbbb1b12e 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -759,9 +759,6 @@ static int ext4_ioctl_get_es_cache(struct file *filp, unsigned long arg)
 		       fieinfo.fi_extents_max * sizeof(struct fiemap_extent)))
 		return -EFAULT;
 
-	if (fieinfo.fi_flags & FIEMAP_FLAG_SYNC)
-		filemap_write_and_wait(inode->i_mapping);
-
 	error = ext4_get_es_cache(inode, &fieinfo, fiemap.fm_start,
 			fiemap.fm_length);
 	fiemap.fm_flags = fieinfo.fi_flags;
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 03faafc591b17..9de7dc476ed16 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -1825,8 +1825,7 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 			return ret;
 	}
 
-	ret = fiemap_prep(inode, fieinfo, start, &len,
-			FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR);
+	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_XATTR);
 	if (ret)
 		return ret;
 
diff --git a/fs/ioctl.c b/fs/ioctl.c
index 56bbf02209aef..b16e962340db6 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -166,6 +166,7 @@ int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
 {
 	u64 maxbytes = inode->i_sb->s_maxbytes;
 	u32 incompat_flags;
+	int ret = 0;
 
 	if (*len == 0)
 		return -EINVAL;
@@ -178,13 +179,17 @@ int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	if (*len > maxbytes || (maxbytes - *len) < start)
 		*len = maxbytes - start;
 
+	supported_flags |= FIEMAP_FLAG_SYNC;
 	supported_flags &= FIEMAP_FLAGS_COMPAT;
 	incompat_flags = fieinfo->fi_flags & ~supported_flags;
 	if (incompat_flags) {
 		fieinfo->fi_flags = incompat_flags;
 		return -EBADR;
 	}
-	return 0;
+
+	if (fieinfo->fi_flags & FIEMAP_FLAG_SYNC)
+		ret = filemap_write_and_wait(inode->i_mapping);
+	return ret;
 }
 EXPORT_SYMBOL(fiemap_prep);
 
@@ -213,9 +218,6 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
 		       fieinfo.fi_extents_max * sizeof(struct fiemap_extent)))
 		return -EFAULT;
 
-	if (fieinfo.fi_flags & FIEMAP_FLAG_SYNC)
-		filemap_write_and_wait(inode->i_mapping);
-
 	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start,
 			fiemap.fm_length);
 	fiemap.fm_flags = fieinfo.fi_flags;
diff --git a/fs/iomap/fiemap.c b/fs/iomap/fiemap.c
index 89dca4a97e4a2..aab070df4a217 100644
--- a/fs/iomap/fiemap.c
+++ b/fs/iomap/fiemap.c
@@ -75,16 +75,10 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi,
 	ctx.fi = fi;
 	ctx.prev.type = IOMAP_HOLE;
 
-	ret = fiemap_prep(inode, fi, start, &len, FIEMAP_FLAG_SYNC);
+	ret = fiemap_prep(inode, fi, start, &len, 0);
 	if (ret)
 		return ret;
 
-	if (fi->fi_flags & FIEMAP_FLAG_SYNC) {
-		ret = filemap_write_and_wait(inode->i_mapping);
-		if (ret)
-			return ret;
-	}
-
 	while (len > 0) {
 		ret = iomap_apply(inode, start, len, IOMAP_REPORT, ops, &ctx,
 				iomap_fiemap_actor);
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 052c2da11e4d7..25b0d368ecdb2 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -1006,7 +1006,7 @@ int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	unsigned int blkbits = inode->i_blkbits;
 	int ret, n;
 
-	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
+	ret = fiemap_prep(inode, fieinfo, start, &len, 0);
 	if (ret)
 		return ret;
 
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 3744179b73fa1..a94852af5510d 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -733,8 +733,6 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,
 	return 0;
 }
 
-#define OCFS2_FIEMAP_FLAGS	(FIEMAP_FLAG_SYNC)
-
 int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		 u64 map_start, u64 map_len)
 {
@@ -746,8 +744,7 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	struct buffer_head *di_bh = NULL;
 	struct ocfs2_extent_rec rec;
 
-	ret = fiemap_prep(inode, fieinfo, map_start, &map_len,
-			OCFS2_FIEMAP_FLAGS);
+	ret = fiemap_prep(inode, fieinfo, map_start, &map_len, 0);
 	if (ret)
 		return ret;
 
diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c
index 625e58da4e82a..7af76b9004ebb 100644
--- a/fs/overlayfs/inode.c
+++ b/fs/overlayfs/inode.c
@@ -480,10 +480,6 @@ static int ovl_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		return -EOPNOTSUPP;
 
 	old_cred = ovl_override_creds(inode->i_sb);
-
-	if (fieinfo->fi_flags & FIEMAP_FLAG_SYNC)
-		filemap_write_and_wait(realinode->i_mapping);
-
 	err = realinode->i_op->fiemap(realinode, fieinfo, start, len);
 	revert_creds(old_cred);
 
-- 
2.26.2

[PATCH 8/9] fs: remove the access_ok() check in ioctl_fiemap

[Christoph Hellwig]

access_ok just checks we are fed a proper user pointer.  We also do that
in copy_to_user itself, so no need to do this early.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/ioctl.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/fs/ioctl.c b/fs/ioctl.c
index b16e962340db6..d69786d1dd911 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -213,13 +213,9 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
 	fieinfo.fi_extents_max = fiemap.fm_extent_count;
 	fieinfo.fi_extents_start = ufiemap->fm_extents;
 
-	if (fiemap.fm_extent_count != 0 &&
-	    !access_ok(fieinfo.fi_extents_start,
-		       fieinfo.fi_extents_max * sizeof(struct fiemap_extent)))
-		return -EFAULT;
-
 	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start,
 			fiemap.fm_length);
+
 	fiemap.fm_flags = fieinfo.fi_flags;
 	fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;
 	if (copy_to_user(ufiemap, &fiemap, sizeof(fiemap)))
-- 
2.26.2

[PATCH 9/9] ext4: remove the access_ok() check in ext4_ioctl_get_es_cache

[Christoph Hellwig]

access_ok just checks we are fed a proper user pointer.  We also do that
in copy_to_user itself, so no need to do this early.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/ioctl.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index f81acbbb1b12e..2162db0c747d2 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -754,11 +754,6 @@ static int ext4_ioctl_get_es_cache(struct file *filp, unsigned long arg)
 	fieinfo.fi_extents_max = fiemap.fm_extent_count;
 	fieinfo.fi_extents_start = ufiemap->fm_extents;
 
-	if (fiemap.fm_extent_count != 0 &&
-	    !access_ok(fieinfo.fi_extents_start,
-		       fieinfo.fi_extents_max * sizeof(struct fiemap_extent)))
-		return -EFAULT;
-
 	error = ext4_get_es_cache(inode, &fieinfo, fiemap.fm_start,
 			fiemap.fm_length);
 	fiemap.fm_flags = fieinfo.fi_flags;
-- 
2.26.2

Re: fiemap cleanups v4

[Al Viro]

On Sat, May 23, 2020 at 09:30:07AM +0200, Christoph Hellwig wrote:
> Hi all,
> 
> This series cleans up the fiemap support in ext4 and in general.
> 
> Ted or Al, can one of you pick this up?  It touches both ext4 and core
> code, so either tree could work.
> 
> 
> Changes since v3:
>  - dropped the fixes that have been merged int mainline
> 
> Changes since v2:
>  - commit message typo
>  - doc updates
>  - use d_inode in cifs
>  - add a missing return statement in cifs
>  - remove the filemap_write_and_wait call from ext4_ioctl_get_es_cache
> 
> Changes since v1:
>  - rename fiemap_validate to fiemap_prep
>  - lift FIEMAP_FLAG_SYNC handling to common code
>  - add a new linux/fiemap.h header
>  - remove __generic_block_fiemap
>  - remove access_ok calls from fiemap and ext4

Hmmm...  I can do an immutable shared branch, no problem.  What would
you prefer for a branchpoint for that one?

Re: fiemap cleanups v4

[Christoph Hellwig]

On Sat, May 23, 2020 at 04:52:16PM +0100, Al Viro wrote:
> Hmmm...  I can do an immutable shared branch, no problem.  What would
> you prefer for a branchpoint for that one?

This needs pretty much latest Linus' tree as it sits on top of the bug
fixes merged this week.  But then again I'm not sure we even need
a shared tree, as there shouldn't be any conflicts.  But looking at
linux-next it seems Ted has actually pulled in the whole series, so
I think we actually are done.

But while I've got you attention:  Can you take a look at my
"clean up kernel_{read,write} & friends v2" series?

Re: fiemap cleanups v4

[Theodore Y. Ts'o]

On Sat, May 23, 2020 at 04:52:16PM +0100, Al Viro wrote:
> On Sat, May 23, 2020 at 09:30:07AM +0200, Christoph Hellwig wrote:
> > Hi all,
> > 
> > This series cleans up the fiemap support in ext4 and in general.
> > 
> > Ted or Al, can one of you pick this up?  It touches both ext4 and core
> > code, so either tree could work.
> > 
> > 
> > Changes since v3:
> >  - dropped the fixes that have been merged int mainline
> > 
> > Changes since v2:
> >  - commit message typo
> >  - doc updates
> >  - use d_inode in cifs
> >  - add a missing return statement in cifs
> >  - remove the filemap_write_and_wait call from ext4_ioctl_get_es_cache
> > 
> > Changes since v1:
> >  - rename fiemap_validate to fiemap_prep
> >  - lift FIEMAP_FLAG_SYNC handling to common code
> >  - add a new linux/fiemap.h header
> >  - remove __generic_block_fiemap
> >  - remove access_ok calls from fiemap and ext4
> 
> Hmmm...  I can do an immutable shared branch, no problem.  What would
> you prefer for a branchpoint for that one?

I thought we had already agreed to run these patches through the ext4
git tree, since most of the changes affect the ext4 tree (and there
aren't any other iomap fiemap changes pending as far as I know).

The v3 versions of these patches have been part of the ext4 dev tree
since May 19th.  Since the ext4 dev tree is rewinding, I can easily
update it fiemap-fixes patch to be on top of the first two patches
which Linus has already accepted, and then merge it into the ext4 dev
branch.

							- Ted

Re: fiemap cleanups v4

[Christoph Hellwig]

On Sun, May 24, 2020 at 03:17:13PM -0400, Theodore Y. Ts'o wrote:
> > Hmmm...  I can do an immutable shared branch, no problem.  What would
> > you prefer for a branchpoint for that one?
> 
> I thought we had already agreed to run these patches through the ext4
> git tree, since most of the changes affect the ext4 tree (and there
> aren't any other iomap fiemap changes pending as far as I know).
> 
> The v3 versions of these patches have been part of the ext4 dev tree
> since May 19th.  Since the ext4 dev tree is rewinding, I can easily
> update it fiemap-fixes patch to be on top of the first two patches
> which Linus has already accepted, and then merge it into the ext4 dev
> branch.

I think we're all fine and set, unless Al really needs a shared
branch.  Based on what is in linux-next and out on the lists I don't
see any real need, though.

Re: [PATCH 6/9] fs: move fiemap range validation into the file systems instances

[Darrick J. Wong]

On Sat, May 23, 2020 at 09:30:13AM +0200, Christoph Hellwig wrote:
> Replace fiemap_check_flags with a fiemap_prep helper that also takes the
> inode and mapped range, and performs the sanity check and truncation
> previously done in fiemap_check_range.  This way the validation is inside
> the file system itself and thus properly works for the stacked overlayfs
> case as well.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> Reviewed-by: Amir Goldstein <amir73il@gmail.com>

Still looks ok,
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>

--D

> ---
>  Documentation/filesystems/fiemap.txt | 12 +++---
>  fs/btrfs/inode.c                     |  2 +-
>  fs/cifs/smb2ops.c                    |  6 ++-
>  fs/ext4/extents.c                    |  5 ++-
>  fs/f2fs/data.c                       |  3 +-
>  fs/ioctl.c                           | 63 +++++++++++-----------------
>  fs/iomap/fiemap.c                    |  2 +-
>  fs/nilfs2/inode.c                    |  2 +-
>  fs/ocfs2/extent_map.c                |  3 +-
>  include/linux/fiemap.h               |  3 +-
>  10 files changed, 47 insertions(+), 54 deletions(-)
> 
> diff --git a/Documentation/filesystems/fiemap.txt b/Documentation/filesystems/fiemap.txt
> index ac87e6fda842b..35c8571eccb6e 100644
> --- a/Documentation/filesystems/fiemap.txt
> +++ b/Documentation/filesystems/fiemap.txt
> @@ -203,16 +203,18 @@ EINTR once fatal signal received.
>  
>  
>  Flag checking should be done at the beginning of the ->fiemap callback via the
> -fiemap_check_flags() helper:
> +fiemap_prep() helper:
>  
> -int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
> +int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
> +		u64 start, u64 *len, u32 supported_flags);
>  
>  The struct fieinfo should be passed in as received from ioctl_fiemap(). The
>  set of fiemap flags which the fs understands should be passed via fs_flags. If
> -fiemap_check_flags finds invalid user flags, it will place the bad values in
> +fiemap_prep finds invalid user flags, it will place the bad values in
>  fieinfo->fi_flags and return -EBADR. If the file system gets -EBADR, from
> -fiemap_check_flags(), it should immediately exit, returning that error back to
> -ioctl_fiemap().
> +fiemap_prep(), it should immediately exit, returning that error back to
> +ioctl_fiemap().  Additionally the range is validate against the supported
> +maximum file size.
>  
>  
>  For each extent in the request range, the file system should call
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index 320d1062068d3..1f1ec361089b3 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -8250,7 +8250,7 @@ static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
>  {
>  	int	ret;
>  
> -	ret = fiemap_check_flags(fieinfo, BTRFS_FIEMAP_FLAGS);
> +	ret = fiemap_prep(inode, fieinfo, start, &len, BTRFS_FIEMAP_FLAGS);
>  	if (ret)
>  		return ret;
>  
> diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c
> index 09047f1ddfb66..828e53e795c6d 100644
> --- a/fs/cifs/smb2ops.c
> +++ b/fs/cifs/smb2ops.c
> @@ -3408,8 +3408,10 @@ static int smb3_fiemap(struct cifs_tcon *tcon,
>  	int i, num, rc, flags, last_blob;
>  	u64 next;
>  
> -	if (fiemap_check_flags(fei, FIEMAP_FLAG_SYNC))
> -		return -EBADR;
> +	rc = fiemap_prep(d_inode(cfile->dentry), fei, start, &len,
> +			FIEMAP_FLAG_SYNC);
> +	if (rc)
> +		return rc;
>  
>  	xid = get_xid();
>   again:
> diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> index a41ae7c510170..41f73dea92cac 100644
> --- a/fs/ext4/extents.c
> +++ b/fs/ext4/extents.c
> @@ -4908,8 +4908,9 @@ int ext4_get_es_cache(struct inode *inode, struct fiemap_extent_info *fieinfo,
>  		fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
>  	}
>  
> -	if (fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC))
> -		return -EBADR;
> +	error = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
> +	if (error)
> +		return error;
>  
>  	error = ext4_fiemap_check_ranges(inode, start, &len);
>  	if (error)
> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
> index 25abbbb65ba09..03faafc591b17 100644
> --- a/fs/f2fs/data.c
> +++ b/fs/f2fs/data.c
> @@ -1825,7 +1825,8 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
>  			return ret;
>  	}
>  
> -	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR);
> +	ret = fiemap_prep(inode, fieinfo, start, &len,
> +			FIEMAP_FLAG_SYNC | FIEMAP_FLAG_XATTR);
>  	if (ret)
>  		return ret;
>  
> diff --git a/fs/ioctl.c b/fs/ioctl.c
> index 3f300cc07dee4..56bbf02209aef 100644
> --- a/fs/ioctl.c
> +++ b/fs/ioctl.c
> @@ -149,61 +149,50 @@ int fiemap_fill_next_extent(struct fiemap_extent_info *fieinfo, u64 logical,
>  EXPORT_SYMBOL(fiemap_fill_next_extent);
>  
>  /**
> - * fiemap_check_flags - check validity of requested flags for fiemap
> + * fiemap_prep - check validity of requested flags for fiemap
> + * @inode:	Inode to operate on
>   * @fieinfo:	Fiemap context passed into ->fiemap
> - * @fs_flags:	Set of fiemap flags that the file system understands
> + * @start:	Start of the mapped range
> + * @len:	Length of the mapped range, can be truncated by this function.
> + * @supported_flags:	Set of fiemap flags that the file system understands
>   *
> - * Called from file system ->fiemap callback. This will compute the
> - * intersection of valid fiemap flags and those that the fs supports. That
> - * value is then compared against the user supplied flags. In case of bad user
> - * flags, the invalid values will be written into the fieinfo structure, and
> - * -EBADR is returned, which tells ioctl_fiemap() to return those values to
> - * userspace. For this reason, a return code of -EBADR should be preserved.
> + * This function must be called from each ->fiemap instance to validate the
> + * fiemap request against the file system parameters.
>   *
> - * Returns 0 on success, -EBADR on bad flags.
> + * Returns 0 on success, or a negative error on failure.
>   */
> -int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags)
> +int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
> +		u64 start, u64 *len, u32 supported_flags)
>  {
> +	u64 maxbytes = inode->i_sb->s_maxbytes;
>  	u32 incompat_flags;
>  
> -	incompat_flags = fieinfo->fi_flags & ~(FIEMAP_FLAGS_COMPAT & fs_flags);
> -	if (incompat_flags) {
> -		fieinfo->fi_flags = incompat_flags;
> -		return -EBADR;
> -	}
> -	return 0;
> -}
> -EXPORT_SYMBOL(fiemap_check_flags);
> -
> -static int fiemap_check_ranges(struct super_block *sb,
> -			       u64 start, u64 len, u64 *new_len)
> -{
> -	u64 maxbytes = (u64) sb->s_maxbytes;
> -
> -	*new_len = len;
> -
> -	if (len == 0)
> +	if (*len == 0)
>  		return -EINVAL;
> -
>  	if (start > maxbytes)
>  		return -EFBIG;
>  
>  	/*
>  	 * Shrink request scope to what the fs can actually handle.
>  	 */
> -	if (len > maxbytes || (maxbytes - len) < start)
> -		*new_len = maxbytes - start;
> +	if (*len > maxbytes || (maxbytes - *len) < start)
> +		*len = maxbytes - start;
>  
> +	supported_flags &= FIEMAP_FLAGS_COMPAT;
> +	incompat_flags = fieinfo->fi_flags & ~supported_flags;
> +	if (incompat_flags) {
> +		fieinfo->fi_flags = incompat_flags;
> +		return -EBADR;
> +	}
>  	return 0;
>  }
> +EXPORT_SYMBOL(fiemap_prep);
>  
>  static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
>  {
>  	struct fiemap fiemap;
>  	struct fiemap_extent_info fieinfo = { 0, };
>  	struct inode *inode = file_inode(filp);
> -	struct super_block *sb = inode->i_sb;
> -	u64 len;
>  	int error;
>  
>  	if (!inode->i_op->fiemap)
> @@ -215,11 +204,6 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
>  	if (fiemap.fm_extent_count > FIEMAP_MAX_EXTENTS)
>  		return -EINVAL;
>  
> -	error = fiemap_check_ranges(sb, fiemap.fm_start, fiemap.fm_length,
> -				    &len);
> -	if (error)
> -		return error;
> -
>  	fieinfo.fi_flags = fiemap.fm_flags;
>  	fieinfo.fi_extents_max = fiemap.fm_extent_count;
>  	fieinfo.fi_extents_start = ufiemap->fm_extents;
> @@ -232,7 +216,8 @@ static int ioctl_fiemap(struct file *filp, struct fiemap __user *ufiemap)
>  	if (fieinfo.fi_flags & FIEMAP_FLAG_SYNC)
>  		filemap_write_and_wait(inode->i_mapping);
>  
> -	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start, len);
> +	error = inode->i_op->fiemap(inode, &fieinfo, fiemap.fm_start,
> +			fiemap.fm_length);
>  	fiemap.fm_flags = fieinfo.fi_flags;
>  	fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;
>  	if (copy_to_user(ufiemap, &fiemap, sizeof(fiemap)))
> @@ -320,7 +305,7 @@ static int __generic_block_fiemap(struct inode *inode,
>  	bool past_eof = false, whole_file = false;
>  	int ret = 0;
>  
> -	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC);
> +	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
>  	if (ret)
>  		return ret;
>  
> diff --git a/fs/iomap/fiemap.c b/fs/iomap/fiemap.c
> index 449705575acf9..89dca4a97e4a2 100644
> --- a/fs/iomap/fiemap.c
> +++ b/fs/iomap/fiemap.c
> @@ -75,7 +75,7 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi,
>  	ctx.fi = fi;
>  	ctx.prev.type = IOMAP_HOLE;
>  
> -	ret = fiemap_check_flags(fi, FIEMAP_FLAG_SYNC);
> +	ret = fiemap_prep(inode, fi, start, &len, FIEMAP_FLAG_SYNC);
>  	if (ret)
>  		return ret;
>  
> diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
> index 6e1aca38931f3..052c2da11e4d7 100644
> --- a/fs/nilfs2/inode.c
> +++ b/fs/nilfs2/inode.c
> @@ -1006,7 +1006,7 @@ int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
>  	unsigned int blkbits = inode->i_blkbits;
>  	int ret, n;
>  
> -	ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC);
> +	ret = fiemap_prep(inode, fieinfo, start, &len, FIEMAP_FLAG_SYNC);
>  	if (ret)
>  		return ret;
>  
> diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
> index e3e2d1b2af51a..3744179b73fa1 100644
> --- a/fs/ocfs2/extent_map.c
> +++ b/fs/ocfs2/extent_map.c
> @@ -746,7 +746,8 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
>  	struct buffer_head *di_bh = NULL;
>  	struct ocfs2_extent_rec rec;
>  
> -	ret = fiemap_check_flags(fieinfo, OCFS2_FIEMAP_FLAGS);
> +	ret = fiemap_prep(inode, fieinfo, map_start, &map_len,
> +			OCFS2_FIEMAP_FLAGS);
>  	if (ret)
>  		return ret;
>  
> diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
> index 240d4f7d9116a..4e624c4665837 100644
> --- a/include/linux/fiemap.h
> +++ b/include/linux/fiemap.h
> @@ -13,9 +13,10 @@ struct fiemap_extent_info {
>  							fiemap_extent array */
>  };
>  
> +int fiemap_prep(struct inode *inode, struct fiemap_extent_info *fieinfo,
> +		u64 start, u64 *len, u32 supported_flags);
>  int fiemap_fill_next_extent(struct fiemap_extent_info *info, u64 logical,
>  			    u64 phys, u64 len, u32 flags);
> -int fiemap_check_flags(struct fiemap_extent_info *fieinfo, u32 fs_flags);
>  
>  int generic_block_fiemap(struct inode *inode,
>  		struct fiemap_extent_info *fieinfo, u64 start, u64 len,
> -- 
> 2.26.2
>