ext4: support GETFSMAP ioctls

Support the GETFSMAP ioctls so that we can use the xfs free space
management tools to probe ext4 as well.  Note that this is a partial
implementation -- we only report fixed-location metadata and free space;
everything else is reported as "unknown".
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>

fs/ext4/Makefile

@@ -4,11 +4,11 @@
 
 obj-$(CONFIG_EXT4_FS) += ext4.o
 
-ext4-y	:= balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o page-io.o \
-		ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \
-		ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o \
-		mmp.o indirect.o extents_status.o xattr.o xattr_user.o \
-		xattr_trusted.o inline.o readpage.o sysfs.o
+ext4-y	:= balloc.o bitmap.o block_validity.o dir.o ext4_jbd2.o extents.o \
+		extents_status.o file.o fsmap.o fsync.o hash.o ialloc.o \
+		indirect.o inline.o inode.o ioctl.o mballoc.o migrate.o \
+		mmp.o move_extent.o namei.o page-io.o readpage.o resize.o \
+		super.o symlink.o sysfs.o xattr.o xattr_trusted.o xattr_user.o
 
 ext4-$(CONFIG_EXT4_FS_POSIX_ACL)	+= acl.o
 ext4-$(CONFIG_EXT4_FS_SECURITY)		+= xattr_security.o

fs/ext4/fsmap.h

+/*
+ * Copyright (C) 2017 Oracle.  All Rights Reserved.
+ *
+ * Author: Darrick J. Wong <darrick.wong@oracle.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+#ifndef __EXT4_FSMAP_H__
+#define	__EXT4_FSMAP_H__
+
+struct fsmap;
+
+/* internal fsmap representation */
+struct ext4_fsmap {
+	struct list_head	fmr_list;
+	dev_t		fmr_device;	/* device id */
+	uint32_t	fmr_flags;	/* mapping flags */
+	uint64_t	fmr_physical;	/* device offset of segment */
+	uint64_t	fmr_owner;	/* owner id */
+	uint64_t	fmr_length;	/* length of segment, blocks */
+};
+
+struct ext4_fsmap_head {
+	uint32_t	fmh_iflags;	/* control flags */
+	uint32_t	fmh_oflags;	/* output flags */
+	unsigned int	fmh_count;	/* # of entries in array incl. input */
+	unsigned int	fmh_entries;	/* # of entries filled in (output). */
+
+	struct ext4_fsmap fmh_keys[2];	/* low and high keys */
+};
+
+void ext4_fsmap_from_internal(struct super_block *sb, struct fsmap *dest,
+		struct ext4_fsmap *src);
+void ext4_fsmap_to_internal(struct super_block *sb, struct ext4_fsmap *dest,
+		struct fsmap *src);
+
+/* fsmap to userspace formatter - copy to user & advance pointer */
+typedef int (*ext4_fsmap_format_t)(struct ext4_fsmap *, void *);
+
+int ext4_getfsmap(struct super_block *sb, struct ext4_fsmap_head *head,
+		ext4_fsmap_format_t formatter, void *arg);
+
+#define EXT4_QUERY_RANGE_ABORT		1
+#define EXT4_QUERY_RANGE_CONTINUE	0
+
+/*	fmr_owner special values for FS_IOC_GETFSMAP; some share w/ XFS */
+#define EXT4_FMR_OWN_FREE	FMR_OWN_FREE      /* free space */
+#define EXT4_FMR_OWN_UNKNOWN	FMR_OWN_UNKNOWN   /* unknown owner */
+#define EXT4_FMR_OWN_FS		FMR_OWNER('X', 1) /* static fs metadata */
+#define EXT4_FMR_OWN_LOG	FMR_OWNER('X', 2) /* journalling log */
+#define EXT4_FMR_OWN_INODES	FMR_OWNER('X', 5) /* inodes */
+#define EXT4_FMR_OWN_GDT	FMR_OWNER('f', 1) /* group descriptors */
+#define EXT4_FMR_OWN_RESV_GDT	FMR_OWNER('f', 2) /* reserved gdt blocks */
+#define EXT4_FMR_OWN_BLKBM	FMR_OWNER('f', 3) /* inode bitmap */
+#define EXT4_FMR_OWN_INOBM	FMR_OWNER('f', 4) /* block bitmap */
+
+#endif /* __EXT4_FSMAP_H__ */

fs/ext4/ioctl.c

@@ -19,6 +19,9 @@
 #include <linux/delay.h>
 #include "ext4_jbd2.h"
 #include "ext4.h"
+#include <linux/fsmap.h>
+#include "fsmap.h"
+#include <trace/events/ext4.h>
 
 /**
  * Swap memory between @a and @b for @len bytes.
@@ -489,6 +492,90 @@ int ext4_shutdown(struct super_block *sb, unsigned long arg)
 	return 0;
 }
 
+struct getfsmap_info {
+	struct super_block	*gi_sb;
+	struct fsmap_head __user *gi_data;
+	unsigned int		gi_idx;
+	__u32			gi_last_flags;
+};
+
+static int ext4_getfsmap_format(struct ext4_fsmap *xfm, void *priv)
+{
+	struct getfsmap_info *info = priv;
+	struct fsmap fm;
+
+	trace_ext4_getfsmap_mapping(info->gi_sb, xfm);
+
+	info->gi_last_flags = xfm->fmr_flags;
+	ext4_fsmap_from_internal(info->gi_sb, &fm, xfm);
+	if (copy_to_user(&info->gi_data->fmh_recs[info->gi_idx++], &fm,
+			sizeof(struct fsmap)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int ext4_ioc_getfsmap(struct super_block *sb,
+			     struct fsmap_head __user *arg)
+{
+	struct getfsmap_info info = {0};
+	struct ext4_fsmap_head xhead = {0};
+	struct fsmap_head head;
+	bool aborted = false;
+	int error;
+
+	if (copy_from_user(&head, arg, sizeof(struct fsmap_head)))
+		return -EFAULT;
+	if (memchr_inv(head.fmh_reserved, 0, sizeof(head.fmh_reserved)) ||
+	    memchr_inv(head.fmh_keys[0].fmr_reserved, 0,
+		       sizeof(head.fmh_keys[0].fmr_reserved)) ||
+	    memchr_inv(head.fmh_keys[1].fmr_reserved, 0,
+		       sizeof(head.fmh_keys[1].fmr_reserved)))
+		return -EINVAL;
+	/*
+	 * ext4 doesn't report file extents at all, so the only valid
+	 * file offsets are the magic ones (all zeroes or all ones).
+	 */
+	if (head.fmh_keys[0].fmr_offset ||
+	    (head.fmh_keys[1].fmr_offset != 0 &&
+	     head.fmh_keys[1].fmr_offset != -1ULL))
+		return -EINVAL;
+
+	xhead.fmh_iflags = head.fmh_iflags;
+	xhead.fmh_count = head.fmh_count;
+	ext4_fsmap_to_internal(sb, &xhead.fmh_keys[0], &head.fmh_keys[0]);
+	ext4_fsmap_to_internal(sb, &xhead.fmh_keys[1], &head.fmh_keys[1]);
+
+	trace_ext4_getfsmap_low_key(sb, &xhead.fmh_keys[0]);
+	trace_ext4_getfsmap_high_key(sb, &xhead.fmh_keys[1]);
+
+	info.gi_sb = sb;
+	info.gi_data = arg;
+	error = ext4_getfsmap(sb, &xhead, ext4_getfsmap_format, &info);
+	if (error == EXT4_QUERY_RANGE_ABORT) {
+		error = 0;
+		aborted = true;
+	} else if (error)
+		return error;
+
+	/* If we didn't abort, set the "last" flag in the last fmx */
+	if (!aborted && info.gi_idx) {
+		info.gi_last_flags |= FMR_OF_LAST;
+		if (copy_to_user(&info.gi_data->fmh_recs[info.gi_idx - 1].fmr_flags,
+				 &info.gi_last_flags,
+				 sizeof(info.gi_last_flags)))
+			return -EFAULT;
+	}
+
+	/* copy back header */
+	head.fmh_entries = xhead.fmh_entries;
+	head.fmh_oflags = xhead.fmh_oflags;
+	if (copy_to_user(arg, &head, sizeof(struct fsmap_head)))
+		return -EFAULT;
+
+	return 0;
+}
+
 long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 {
 	struct inode *inode = file_inode(filp);
@@ -499,6 +586,8 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	ext4_debug("cmd = %u, arg = %lu\n", cmd, arg);
 
 	switch (cmd) {
+	case FS_IOC_GETFSMAP:
+		return ext4_ioc_getfsmap(sb, (void __user *)arg);
 	case EXT4_IOC_GETFLAGS:
 		ext4_get_inode_flags(ei);
 		flags = ei->i_flags & EXT4_FL_USER_VISIBLE;
@@ -1009,6 +1098,7 @@ long ext4_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case EXT4_IOC_GET_ENCRYPTION_PWSALT:
 	case EXT4_IOC_GET_ENCRYPTION_POLICY:
 	case EXT4_IOC_SHUTDOWN:
+	case FS_IOC_GETFSMAP:
 		break;
 	default:
 		return -ENOIOCTLCMD;

fs/ext4/mballoc.c

@@ -5277,3 +5277,52 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range)
 	range->len = EXT4_C2B(EXT4_SB(sb), trimmed) << sb->s_blocksize_bits;
 	return ret;
 }
+
+/* Iterate all the free extents in the group. */
+int
+ext4_mballoc_query_range(
+	struct super_block		*sb,
+	ext4_group_t			group,
+	ext4_grpblk_t			start,
+	ext4_grpblk_t			end,
+	ext4_mballoc_query_range_fn	formatter,
+	void				*priv)
+{
+	void				*bitmap;
+	ext4_grpblk_t			next;
+	struct ext4_buddy		e4b;
+	int				error;
+
+	error = ext4_mb_load_buddy(sb, group, &e4b);
+	if (error)
+		return error;
+	bitmap = e4b.bd_bitmap;
+
+	ext4_lock_group(sb, group);
+
+	start = (e4b.bd_info->bb_first_free > start) ?
+		e4b.bd_info->bb_first_free : start;
+	if (end >= EXT4_CLUSTERS_PER_GROUP(sb))
+		end = EXT4_CLUSTERS_PER_GROUP(sb) - 1;
+
+	while (start <= end) {
+		start = mb_find_next_zero_bit(bitmap, end + 1, start);
+		if (start > end)
+			break;
+		next = mb_find_next_bit(bitmap, end + 1, start);
+
+		ext4_unlock_group(sb, group);
+		error = formatter(sb, group, start, next - start, priv);
+		if (error)
+			goto out_unload;
+		ext4_lock_group(sb, group);
+
+		start = next + 1;
+	}
+
+	ext4_unlock_group(sb, group);
+out_unload:
+	ext4_mb_unload_buddy(&e4b);
+
+	return error;
+}

fs/ext4/mballoc.h

@@ -199,4 +199,21 @@ static inline ext4_fsblk_t ext4_grp_offs_to_block(struct super_block *sb,
 	return ext4_group_first_block_no(sb, fex->fe_group) +
 		(fex->fe_start << EXT4_SB(sb)->s_cluster_bits);
 }
+
+typedef int (*ext4_mballoc_query_range_fn)(
+	struct super_block		*sb,
+	ext4_group_t			agno,
+	ext4_grpblk_t			start,
+	ext4_grpblk_t			len,
+	void				*priv);
+
+int
+ext4_mballoc_query_range(
+	struct super_block		*sb,
+	ext4_group_t			agno,
+	ext4_grpblk_t			start,
+	ext4_grpblk_t			end,
+	ext4_mballoc_query_range_fn	formatter,
+	void				*priv);
+
 #endif

fs/ext4/super.c

@@ -49,6 +49,7 @@
 #include "xattr.h"
 #include "acl.h"
 #include "mballoc.h"
+#include "fsmap.h"
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/ext4.h>

include/trace/events/ext4.h

@@ -15,6 +15,7 @@ struct ext4_inode_info;
 struct mpage_da_data;
 struct ext4_map_blocks;
 struct extent_status;
+struct ext4_fsmap;
 
 #define EXT4_I(inode) (container_of(inode, struct ext4_inode_info, vfs_inode))
 
@@ -2529,6 +2530,79 @@ TRACE_EVENT(ext4_es_shrink,
 		  __entry->scan_time, __entry->nr_skipped, __entry->retried)
 );
 
+/* fsmap traces */
+DECLARE_EVENT_CLASS(ext4_fsmap_class,
+	TP_PROTO(struct super_block *sb, u32 keydev, u32 agno, u64 bno, u64 len,
+		 u64 owner),
+	TP_ARGS(sb, keydev, agno, bno, len, owner),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(dev_t, keydev)
+		__field(u32, agno)
+		__field(u64, bno)
+		__field(u64, len)
+		__field(u64, owner)
+	),
+	TP_fast_assign(
+		__entry->dev = sb->s_bdev->bd_dev;
+		__entry->keydev = new_decode_dev(keydev);
+		__entry->agno = agno;
+		__entry->bno = bno;
+		__entry->len = len;
+		__entry->owner = owner;
+	),
+	TP_printk("dev %d:%d keydev %d:%d agno %u bno %llu len %llu owner %lld\n",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  MAJOR(__entry->keydev), MINOR(__entry->keydev),
+		  __entry->agno,
+		  __entry->bno,
+		  __entry->len,
+		  __entry->owner)
+)
+#define DEFINE_FSMAP_EVENT(name) \
+DEFINE_EVENT(ext4_fsmap_class, name, \
+	TP_PROTO(struct super_block *sb, u32 keydev, u32 agno, u64 bno, u64 len, \
+		 u64 owner), \
+	TP_ARGS(sb, keydev, agno, bno, len, owner))
+DEFINE_FSMAP_EVENT(ext4_fsmap_low_key);
+DEFINE_FSMAP_EVENT(ext4_fsmap_high_key);
+DEFINE_FSMAP_EVENT(ext4_fsmap_mapping);
+
+DECLARE_EVENT_CLASS(ext4_getfsmap_class,
+	TP_PROTO(struct super_block *sb, struct ext4_fsmap *fsmap),
+	TP_ARGS(sb, fsmap),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(dev_t, keydev)
+		__field(u64, block)
+		__field(u64, len)
+		__field(u64, owner)
+		__field(u64, flags)
+	),
+	TP_fast_assign(
+		__entry->dev = sb->s_bdev->bd_dev;
+		__entry->keydev = new_decode_dev(fsmap->fmr_device);
+		__entry->block = fsmap->fmr_physical;
+		__entry->len = fsmap->fmr_length;
+		__entry->owner = fsmap->fmr_owner;
+		__entry->flags = fsmap->fmr_flags;
+	),
+	TP_printk("dev %d:%d keydev %d:%d block %llu len %llu owner %lld flags 0x%llx\n",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  MAJOR(__entry->keydev), MINOR(__entry->keydev),
+		  __entry->block,
+		  __entry->len,
+		  __entry->owner,
+		  __entry->flags)
+)
+#define DEFINE_GETFSMAP_EVENT(name) \
+DEFINE_EVENT(ext4_getfsmap_class, name, \
+	TP_PROTO(struct super_block *sb, struct ext4_fsmap *fsmap), \
+	TP_ARGS(sb, fsmap))
+DEFINE_GETFSMAP_EVENT(ext4_getfsmap_low_key);
+DEFINE_GETFSMAP_EVENT(ext4_getfsmap_high_key);
+DEFINE_GETFSMAP_EVENT(ext4_getfsmap_mapping);
+
 #endif /* _TRACE_EXT4_H */
 
 /* This part must be outside protection */