f2fs: support flexible inline xattr size
Now, in product, more and more features based on file encryption were
introduced, their demand of xattr space is increasing, however, inline
xattr has fixed-size of 200 bytes, once inline xattr space is full, new
increased xattr data would occupy additional xattr block which may bring
us more space usage and performance regression during persisting.
In order to resolve above issue, it's better to expand inline xattr size
flexibly according to user's requirement.
So this patch introduces new filesystem feature 'flexible inline xattr',
and new mount option 'inline_xattr_size=%u', once mkfs enables the
feature, we can use the option to make f2fs supporting flexible inline
xattr size.
To support this feature, we add extra attribute i_inline_xattr_size in
inode layout, indicating that how many space inline xattr borrows from
block address mapping space in inode layout, by this, we can easily
locate and store flexible-sized inline xattr data in inode.
Inode disk layout:
+----------------------+
| .i_mode |
| ... |
| .i_ext |
+----------------------+
| .i_extra_isize |
| .i_inline_xattr_size |-----------+
| ... | |
+----------------------+ |
| .i_addr | |
| - block address or | |
| - inline data | |
+----------------------+<---+ v
| inline xattr | +---inline xattr range
+----------------------+<---+
| .i_nid |
+----------------------+
| node_footer |
| (nid, ino, offset) |
+----------------------+
Note that, we have to cnosider backward compatibility which reserved
inline_data space, 200 bytes, all the time, reported by Sheng Yong.
Previous inline data or directory always reserved 200 bytes in inode layout,
even if inline_xattr is disabled. In order to keep inline_dentry's structure
for backward compatibility, we get the space back only from inline_data.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Reported-by: Sheng Yong <shengyong1@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index dabb43e..9b36791 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -94,6 +94,7 @@
#define F2FS_MOUNT_GRPQUOTA 0x00100000
#define F2FS_MOUNT_PRJQUOTA 0x00200000
#define F2FS_MOUNT_QUOTA 0x00400000
+#define F2FS_MOUNT_INLINE_XATTR_SIZE 0x00800000
#define clear_opt(sbi, option) ((sbi)->mount_opt.opt &= ~F2FS_MOUNT_##option)
#define set_opt(sbi, option) ((sbi)->mount_opt.opt |= F2FS_MOUNT_##option)
@@ -119,6 +120,7 @@
#define F2FS_FEATURE_EXTRA_ATTR 0x0008
#define F2FS_FEATURE_PRJQUOTA 0x0010
#define F2FS_FEATURE_INODE_CHKSUM 0x0020
+#define F2FS_FEATURE_FLEXIBLE_INLINE_XATTR 0x0040
#define F2FS_HAS_FEATURE(sb, mask) \
((F2FS_SB(sb)->raw_super->feature & cpu_to_le32(mask)) != 0)
@@ -395,11 +397,14 @@
/* for inline stuff */
#define DEF_INLINE_RESERVED_SIZE 1
+#define DEF_MIN_INLINE_SIZE 1
static inline int get_extra_isize(struct inode *inode);
-#define MAX_INLINE_DATA(inode) (sizeof(__le32) * \
- (CUR_ADDRS_PER_INODE(inode) - \
- DEF_INLINE_RESERVED_SIZE - \
- F2FS_INLINE_XATTR_ADDRS))
+static inline int get_inline_xattr_addrs(struct inode *inode);
+#define F2FS_INLINE_XATTR_ADDRS(inode) get_inline_xattr_addrs(inode)
+#define MAX_INLINE_DATA(inode) (sizeof(__le32) * \
+ (CUR_ADDRS_PER_INODE(inode) - \
+ F2FS_INLINE_XATTR_ADDRS(inode) - \
+ DEF_INLINE_RESERVED_SIZE))
/* for inline dir */
#define NR_INLINE_DENTRY(inode) (MAX_INLINE_DATA(inode) * BITS_PER_BYTE / \
@@ -610,6 +615,7 @@
int i_extra_isize; /* size of extra space located in i_addr */
kprojid_t i_projid; /* id for project quota */
+ int i_inline_xattr_size; /* inline xattr size */
};
static inline void get_extent_info(struct extent_info *ext,
@@ -1067,6 +1073,7 @@
loff_t max_file_blocks; /* max block index of file */
int active_logs; /* # of active logs */
int dir_level; /* directory level */
+ int inline_xattr_size; /* inline xattr size */
block_t user_block_count; /* # of user blocks */
block_t total_valid_block_count; /* # of valid blocks */
@@ -2184,25 +2191,20 @@
static inline unsigned int addrs_per_inode(struct inode *inode)
{
- if (f2fs_has_inline_xattr(inode))
- return CUR_ADDRS_PER_INODE(inode) - F2FS_INLINE_XATTR_ADDRS;
- return CUR_ADDRS_PER_INODE(inode);
+ return CUR_ADDRS_PER_INODE(inode) - F2FS_INLINE_XATTR_ADDRS(inode);
}
-static inline void *inline_xattr_addr(struct page *page)
+static inline void *inline_xattr_addr(struct inode *inode, struct page *page)
{
struct f2fs_inode *ri = F2FS_INODE(page);
return (void *)&(ri->i_addr[DEF_ADDRS_PER_INODE -
- F2FS_INLINE_XATTR_ADDRS]);
+ F2FS_INLINE_XATTR_ADDRS(inode)]);
}
static inline int inline_xattr_size(struct inode *inode)
{
- if (f2fs_has_inline_xattr(inode))
- return F2FS_INLINE_XATTR_ADDRS << 2;
- else
- return 0;
+ return get_inline_xattr_addrs(inode) * sizeof(__le32);
}
static inline int f2fs_has_inline_data(struct inode *inode)
@@ -2382,6 +2384,12 @@
return F2FS_I(inode)->i_extra_isize / sizeof(__le32);
}
+static inline int f2fs_sb_has_flexible_inline_xattr(struct super_block *sb);
+static inline int get_inline_xattr_addrs(struct inode *inode)
+{
+ return F2FS_I(inode)->i_inline_xattr_size;
+}
+
#define get_inode_mode(i) \
((is_inode_flag_set(i, FI_ACL_MODE)) ? \
(F2FS_I(i)->i_acl_mode) : ((i)->i_mode))
@@ -3051,6 +3059,11 @@
return F2FS_HAS_FEATURE(sb, F2FS_FEATURE_INODE_CHKSUM);
}
+static inline int f2fs_sb_has_flexible_inline_xattr(struct super_block *sb)
+{
+ return F2FS_HAS_FEATURE(sb, F2FS_FEATURE_FLEXIBLE_INLINE_XATTR);
+}
+
#ifdef CONFIG_BLK_DEV_ZONED
static inline int get_blkz_type(struct f2fs_sb_info *sbi,
struct block_device *bdev, block_t blkaddr)
diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
index 3617e7fc..9684d535 100644
--- a/fs/f2fs/inode.c
+++ b/fs/f2fs/inode.c
@@ -232,6 +232,23 @@
fi->i_extra_isize = f2fs_has_extra_attr(inode) ?
le16_to_cpu(ri->i_extra_isize) : 0;
+ if (f2fs_sb_has_flexible_inline_xattr(sbi->sb)) {
+ f2fs_bug_on(sbi, !f2fs_has_extra_attr(inode));
+ fi->i_inline_xattr_size = le16_to_cpu(ri->i_inline_xattr_size);
+ } else if (f2fs_has_inline_xattr(inode) ||
+ f2fs_has_inline_dentry(inode)) {
+ fi->i_inline_xattr_size = DEFAULT_INLINE_XATTR_ADDRS;
+ } else {
+
+ /*
+ * Previous inline data or directory always reserved 200 bytes
+ * in inode layout, even if inline_xattr is disabled. In order
+ * to keep inline_dentry's structure for backward compatibility,
+ * we get the space back only from inline_data.
+ */
+ fi->i_inline_xattr_size = 0;
+ }
+
/* check data exist */
if (f2fs_has_inline_data(inode) && !f2fs_exist_data(inode))
__recover_inline_status(inode, node_page);
@@ -384,6 +401,10 @@
if (f2fs_has_extra_attr(inode)) {
ri->i_extra_isize = cpu_to_le16(F2FS_I(inode)->i_extra_isize);
+ if (f2fs_sb_has_flexible_inline_xattr(F2FS_I_SB(inode)->sb))
+ ri->i_inline_xattr_size =
+ cpu_to_le16(F2FS_I(inode)->i_inline_xattr_size);
+
if (f2fs_sb_has_project_quota(F2FS_I_SB(inode)->sb) &&
F2FS_FITS_IN_INODE(ri, F2FS_I(inode)->i_extra_isize,
i_projid)) {
diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c
index df7880b..5deadc9 100644
--- a/fs/f2fs/namei.c
+++ b/fs/f2fs/namei.c
@@ -29,6 +29,7 @@
nid_t ino;
struct inode *inode;
bool nid_free = false;
+ int xattr_size = 0;
int err;
inode = new_inode(dir->i_sb);
@@ -86,11 +87,23 @@
if (test_opt(sbi, INLINE_XATTR))
set_inode_flag(inode, FI_INLINE_XATTR);
+
if (test_opt(sbi, INLINE_DATA) && f2fs_may_inline_data(inode))
set_inode_flag(inode, FI_INLINE_DATA);
if (f2fs_may_inline_dentry(inode))
set_inode_flag(inode, FI_INLINE_DENTRY);
+ if (f2fs_sb_has_flexible_inline_xattr(sbi->sb)) {
+ f2fs_bug_on(sbi, !f2fs_has_extra_attr(inode));
+ if (f2fs_has_inline_xattr(inode))
+ xattr_size = sbi->inline_xattr_size;
+ /* Otherwise, will be 0 */
+ } else if (f2fs_has_inline_xattr(inode) ||
+ f2fs_has_inline_dentry(inode)) {
+ xattr_size = DEFAULT_INLINE_XATTR_ADDRS;
+ }
+ F2FS_I(inode)->i_inline_xattr_size = xattr_size;
+
f2fs_init_extent_tree(inode, NULL);
stat_inc_inline_xattr(inode);
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index bdb7e92..f44d837 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -2191,8 +2191,8 @@
goto update_inode;
}
- dst_addr = inline_xattr_addr(ipage);
- src_addr = inline_xattr_addr(page);
+ dst_addr = inline_xattr_addr(inode, ipage);
+ src_addr = inline_xattr_addr(inode, page);
inline_size = inline_xattr_size(inode);
f2fs_wait_on_page_writeback(ipage, NODE, true);
@@ -2281,6 +2281,12 @@
dst->i_inline = src->i_inline & (F2FS_INLINE_XATTR | F2FS_EXTRA_ATTR);
if (dst->i_inline & F2FS_EXTRA_ATTR) {
dst->i_extra_isize = src->i_extra_isize;
+
+ if (f2fs_sb_has_flexible_inline_xattr(sbi->sb) &&
+ F2FS_FITS_IN_INODE(src, le16_to_cpu(src->i_extra_isize),
+ i_inline_xattr_size))
+ dst->i_inline_xattr_size = src->i_inline_xattr_size;
+
if (f2fs_sb_has_project_quota(sbi->sb) &&
F2FS_FITS_IN_INODE(src, le16_to_cpu(src->i_extra_isize),
i_projid))
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 272625a..53a7fd8 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -92,6 +92,7 @@
Opt_disable_ext_identify,
Opt_inline_xattr,
Opt_noinline_xattr,
+ Opt_inline_xattr_size,
Opt_inline_data,
Opt_inline_dentry,
Opt_noinline_dentry,
@@ -141,6 +142,7 @@
{Opt_disable_ext_identify, "disable_ext_identify"},
{Opt_inline_xattr, "inline_xattr"},
{Opt_noinline_xattr, "noinline_xattr"},
+ {Opt_inline_xattr_size, "inline_xattr_size=%u"},
{Opt_inline_data, "inline_data"},
{Opt_inline_dentry, "inline_dentry"},
{Opt_noinline_dentry, "noinline_dentry"},
@@ -383,6 +385,12 @@
case Opt_noinline_xattr:
clear_opt(sbi, INLINE_XATTR);
break;
+ case Opt_inline_xattr_size:
+ if (args->from && match_int(args, &arg))
+ return -EINVAL;
+ set_opt(sbi, INLINE_XATTR_SIZE);
+ sbi->inline_xattr_size = arg;
+ break;
#else
case Opt_user_xattr:
f2fs_msg(sb, KERN_INFO,
@@ -604,6 +612,24 @@
F2FS_IO_SIZE_KB(sbi));
return -EINVAL;
}
+
+ if (test_opt(sbi, INLINE_XATTR_SIZE)) {
+ if (!test_opt(sbi, INLINE_XATTR)) {
+ f2fs_msg(sb, KERN_ERR,
+ "inline_xattr_size option should be "
+ "set with inline_xattr option");
+ return -EINVAL;
+ }
+ if (!sbi->inline_xattr_size ||
+ sbi->inline_xattr_size >= DEF_ADDRS_PER_INODE -
+ F2FS_TOTAL_EXTRA_ATTR_SIZE -
+ DEF_INLINE_RESERVED_SIZE -
+ DEF_MIN_INLINE_SIZE) {
+ f2fs_msg(sb, KERN_ERR,
+ "inline xattr size is out of range");
+ return -EINVAL;
+ }
+ }
return 0;
}
@@ -1050,6 +1076,9 @@
seq_puts(seq, ",inline_xattr");
else
seq_puts(seq, ",noinline_xattr");
+ if (test_opt(sbi, INLINE_XATTR_SIZE))
+ seq_printf(seq, ",inline_xattr_size=%u",
+ sbi->inline_xattr_size);
#endif
#ifdef CONFIG_F2FS_FS_POSIX_ACL
if (test_opt(sbi, POSIX_ACL))
@@ -1112,6 +1141,7 @@
{
/* init some FS parameters */
sbi->active_logs = NR_CURSEG_TYPE;
+ sbi->inline_xattr_size = DEFAULT_INLINE_XATTR_ADDRS;
set_opt(sbi, BG_GC);
set_opt(sbi, INLINE_XATTR);
@@ -1670,7 +1700,7 @@
/*
* note: previously, result is equal to (DEF_ADDRS_PER_INODE -
- * F2FS_INLINE_XATTR_ADDRS), but now f2fs try to reserve more
+ * DEFAULT_INLINE_XATTR_ADDRS), but now f2fs try to reserve more
* space in inode.i_addr, it will be more safe to reassign
* result as zero.
*/
diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
index 48ebe61..e09e59c 100644
--- a/fs/f2fs/sysfs.c
+++ b/fs/f2fs/sysfs.c
@@ -107,6 +107,9 @@
if (f2fs_sb_has_inode_chksum(sb))
len += snprintf(buf + len, PAGE_SIZE - len, "%s%s",
len ? ", " : "", "inode_checksum");
+ if (f2fs_sb_has_flexible_inline_xattr(sb))
+ len += snprintf(buf + len, PAGE_SIZE - len, "%s%s",
+ len ? ", " : "", "flexible_inline_xattr");
len += snprintf(buf + len, PAGE_SIZE - len, "\n");
return len;
}
@@ -216,6 +219,7 @@
FEAT_EXTRA_ATTR,
FEAT_PROJECT_QUOTA,
FEAT_INODE_CHECKSUM,
+ FEAT_FLEXIBLE_INLINE_XATTR,
};
static ssize_t f2fs_feature_show(struct f2fs_attr *a,
@@ -228,6 +232,7 @@
case FEAT_EXTRA_ATTR:
case FEAT_PROJECT_QUOTA:
case FEAT_INODE_CHECKSUM:
+ case FEAT_FLEXIBLE_INLINE_XATTR:
return snprintf(buf, PAGE_SIZE, "supported\n");
}
return 0;
@@ -299,6 +304,7 @@
F2FS_FEATURE_RO_ATTR(extra_attr, FEAT_EXTRA_ATTR);
F2FS_FEATURE_RO_ATTR(project_quota, FEAT_PROJECT_QUOTA);
F2FS_FEATURE_RO_ATTR(inode_checksum, FEAT_INODE_CHECKSUM);
+F2FS_FEATURE_RO_ATTR(flexible_inline_xattr, FEAT_FLEXIBLE_INLINE_XATTR);
#define ATTR_LIST(name) (&f2fs_attr_##name.attr)
static struct attribute *f2fs_attrs[] = {
@@ -346,6 +352,7 @@
ATTR_LIST(extra_attr),
ATTR_LIST(project_quota),
ATTR_LIST(inode_checksum),
+ ATTR_LIST(flexible_inline_xattr),
NULL,
};
diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
index 9fbcb76..52b46f5 100644
--- a/fs/f2fs/xattr.c
+++ b/fs/f2fs/xattr.c
@@ -217,12 +217,12 @@
return entry;
}
-static struct f2fs_xattr_entry *__find_inline_xattr(void *base_addr,
- void **last_addr, int index,
- size_t len, const char *name)
+static struct f2fs_xattr_entry *__find_inline_xattr(struct inode *inode,
+ void *base_addr, void **last_addr, int index,
+ size_t len, const char *name)
{
struct f2fs_xattr_entry *entry;
- unsigned int inline_size = F2FS_INLINE_XATTR_ADDRS << 2;
+ unsigned int inline_size = inline_xattr_size(inode);
list_for_each_xattr(entry, base_addr) {
if ((void *)entry + sizeof(__u32) > base_addr + inline_size ||
@@ -250,13 +250,13 @@
void *inline_addr;
if (ipage) {
- inline_addr = inline_xattr_addr(ipage);
+ inline_addr = inline_xattr_addr(inode, ipage);
} else {
page = get_node_page(sbi, inode->i_ino);
if (IS_ERR(page))
return PTR_ERR(page);
- inline_addr = inline_xattr_addr(page);
+ inline_addr = inline_xattr_addr(inode, page);
}
memcpy(txattr_addr, inline_addr, inline_size);
f2fs_put_page(page, 1);
@@ -309,7 +309,7 @@
if (err)
goto out;
- *xe = __find_inline_xattr(txattr_addr, &last_addr,
+ *xe = __find_inline_xattr(inode, txattr_addr, &last_addr,
index, len, name);
if (*xe)
goto check;
@@ -404,7 +404,7 @@
void *inline_addr;
if (ipage) {
- inline_addr = inline_xattr_addr(ipage);
+ inline_addr = inline_xattr_addr(inode, ipage);
f2fs_wait_on_page_writeback(ipage, NODE, true);
set_page_dirty(ipage);
} else {
@@ -413,7 +413,7 @@
alloc_nid_failed(sbi, new_nid);
return PTR_ERR(page);
}
- inline_addr = inline_xattr_addr(page);
+ inline_addr = inline_xattr_addr(inode, page);
f2fs_wait_on_page_writeback(page, NODE, true);
}
memcpy(inline_addr, txattr_addr, inline_size);