Skip to contentSkip to conversation

f2fs: allow out-place-update for direct IO in LFS mode

PATCH Superseded Chao Yu 3 participants 6 messages Sep 16, 2018 – Sep 19, 2018 F2FS file system

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Chao Yu

author maintainer
#0 lore raw

PATCH 1/1 f2fs: allow out-place-update for direct IO in LFS mode

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0
maintainer
#1 lore raw
Quoted 6 lines · Chao Yu

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

I remember this is to serialize all the IOs in one buffered IO path to keep the IO sequence espeicially useful for SMR drivers. Could you confirm whether there is any IO reordering problem?

Quoted 201 lines

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0
#2 lore raw
Quoted 11 lines · Jaegeuk Kim
Quoted 6 lines · Chao Yu

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

I remember this is to serialize all the IOs in one buffered IO path to keep the IO sequence espeicially useful for SMR drivers. Could you confirm whether there is any IO reordering problem?

IMO, buffered IO will encounter reordering issue more easily, for example:

User write 2mb data via DIO, but according to policy on LFS mode, it will fallback to buffered IO.

Before dio triggers writeback, an async writeback from background is triggered, so all IO distribution can be:

page LBA range
async IO: 0 - 255 x + 0, x + 255
sync IO: 256 - 511 x + 256, x + 511

Then block IO scheduler can issue sync IO before async IO, result IO reordering.

Anyway, in DIO path we will suffer less IO reordering issue.

Thanks,

Quoted 205 lines
Quoted 202 lines

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0

.

maintainer
#3 lore raw
Quoted 21 lines · Chao Yu
Quoted 11 lines · Jaegeuk Kim
Quoted 6 lines · Chao Yu

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

I remember this is to serialize all the IOs in one buffered IO path to keep the IO sequence espeicially useful for SMR drivers. Could you confirm whether there is any IO reordering problem?

IMO, buffered IO will encounter reordering issue more easily, for example:

User write 2mb data via DIO, but according to policy on LFS mode, it will fallback to buffered IO.

Before dio triggers writeback, an async writeback from background is triggered, so all IO distribution can be:

Quoted 8 lines

page LBA range
async IO: 0 - 255 x + 0, x + 255
sync IO: 256 - 511 x + 256, x + 511

Then block IO scheduler can issue sync IO before async IO, result IO reordering.

Anyway, in DIO path we will suffer less IO reordering issue.

My concern was on mixed DIO and buffered IOs where we don't have a way to serialize all the IOs.

Quoted 208 lines

Thanks,

Quoted 205 lines
Quoted 202 lines

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0

.

#4 lore raw
Quoted 26 lines · Jaegeuk Kim
Quoted 25 lines · Chao Yu
Quoted 11 lines · Jaegeuk Kim
Quoted 6 lines · Chao Yu

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

I remember this is to serialize all the IOs in one buffered IO path to keep the IO sequence espeicially useful for SMR drivers. Could you confirm whether there is any IO reordering problem?

IMO, buffered IO will encounter reordering issue more easily, for example:

User write 2mb data via DIO, but according to policy on LFS mode, it will fallback to buffered IO.

Before dio triggers writeback, an async writeback from background is triggered, so all IO distribution can be:

page LBA range
async IO: 0 - 255 x + 0, x + 255
sync IO: 256 - 511 x + 256, x + 511

Quoted 7 lines
Quoted 4 lines

Then block IO scheduler can issue sync IO before async IO, result IO reordering.

Anyway, in DIO path we will suffer less IO reordering issue.

My concern was on mixed DIO and buffered IOs where we don't have a way to serialize all the IOs.

As I explained above, even we always use buffered IO, there is no such guarantee for serializing all async/sync IO in block layer, am I missing something?

If you really don't want to change the logic for SMR, how about let me just leave that case?

Thanks,

Quoted 212 lines
Quoted 209 lines

Thanks,

Quoted 205 lines
Quoted 202 lines

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0

.

.

maintainer
#5 lore raw
Quoted 35 lines · Chao Yu
Quoted 26 lines · Jaegeuk Kim
Quoted 25 lines · Chao Yu
Quoted 11 lines · Jaegeuk Kim
Quoted 6 lines · Chao Yu

From: Chao Yu <[email protected]>

Normally, DIO uses in-pllace-update, but in LFS mode, f2fs doesn't allow triggering any in-place-update writes, so we fallback direct write to buffered write, result in bad performance of large size write.

I remember this is to serialize all the IOs in one buffered IO path to keep the IO sequence espeicially useful for SMR drivers. Could you confirm whether there is any IO reordering problem?

IMO, buffered IO will encounter reordering issue more easily, for example:

User write 2mb data via DIO, but according to policy on LFS mode, it will fallback to buffered IO.

Before dio triggers writeback, an async writeback from background is triggered, so all IO distribution can be:

page LBA range
async IO: 0 - 255 x + 0, x + 255
sync IO: 256 - 511 x + 256, x + 511

Quoted 7 lines
Quoted 4 lines

Then block IO scheduler can issue sync IO before async IO, result IO reordering.

Anyway, in DIO path we will suffer less IO reordering issue.

My concern was on mixed DIO and buffered IOs where we don't have a way to serialize all the IOs.

Quoted 6 lines

As I explained above, even we always use buffered IO, there is no such guarantee for serializing all async/sync IO in block layer, am I missing something?

If you really don't want to change the logic for SMR, how about let me just leave that case?

Okay, that could be fine to me.
Thanks,

Quoted 215 lines

Thanks,

Quoted 212 lines
Quoted 209 lines

Thanks,

Quoted 205 lines
Quoted 202 lines

This patch adds to support triggering out-place-update for direct IO to enhance its performance.

Note that it needs to exclude direct read IO during direct write, since new data writing to new block address will no be valid until write finished.

storage: zram

time xfs_io -f -d /mnt/f2fs/file -c "pwrite 0 1073741824" -c "fsync"

Before:
real 0m13.061s
user 0m0.327s
sys 0m12.486s

After:
real 0m6.448s
user 0m0.228s
sys 0m6.212s

---
 fs/f2fs/data.c | 41 +++++++++++++++++++++++++++++++++--------
 fs/f2fs/f2fs.h | 39 +++++++++++++++++++++++++++++++++++----
 fs/f2fs/file.c |  3 ++-
 3 files changed, 70 insertions(+), 13 deletions(-)
fs/f2fs/data.c +33 -8
index 048f9cf1d589..9937dc2f384a 100644
@@ -894,7 +894,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type)
 
 	dn->data_blkaddr = datablock_addr(dn->inode,
 				dn->node_page, dn->ofs_in_node);
-	if (dn->data_blkaddr == NEW_ADDR)
+	if (dn->data_blkaddr != NULL_ADDR)
 		goto alloc;
 
 	if (unlikely((err = inc_valid_block_count(sbi, dn->inode, &count))))
@@ -950,7 +950,7 @@ int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from)
 
 	if (direct_io) {
 		map.m_seg_type = f2fs_rw_hint_to_seg_type(iocb->ki_hint);
-		flag = f2fs_force_buffered_io(inode, WRITE) ?
+		flag = f2fs_force_buffered_io(inode, iocb, from) ?
 					F2FS_GET_BLOCK_PRE_AIO :
 					F2FS_GET_BLOCK_PRE_DIO;
 		goto map_blocks;
@@ -1069,7 +1069,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
 		goto sync_out;
 	}
 
-	if (!is_valid_data_blkaddr(sbi, blkaddr)) {
+	if (is_valid_data_blkaddr(sbi, blkaddr)) {
+		/* use out-place-update for driect IO under LFS mode */
+		if (test_opt(sbi, LFS) && create &&
+				flag == F2FS_GET_BLOCK_DEFAULT) {
+			err = __allocate_data_block(&dn, map->m_seg_type);
+			if (!err)
+				set_inode_flag(inode, FI_APPEND_WRITE);
+		}
+	} else {
 		if (create) {
 			if (unlikely(f2fs_cp_error(sbi))) {
 				err = -EIO;
@@ -2491,36 +2499,53 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 	struct address_space *mapping = iocb->ki_filp->f_mapping;
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
 	size_t count = iov_iter_count(iter);
 	loff_t offset = iocb->ki_pos;
 	int rw = iov_iter_rw(iter);
 	int err;
 	enum rw_hint hint = iocb->ki_hint;
 	int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	bool lock_read;
 
 	err = check_direct_IO(inode, iter, offset);
 	if (err)
 		return err < 0 ? err : 0;
 
-	if (f2fs_force_buffered_io(inode, rw))
+	if (f2fs_force_buffered_io(inode, iocb, iter))
 		return 0;
 
+	lock_read = allow_outplace_dio(inode, iocb, iter);
+
 	trace_f2fs_direct_IO_enter(inode, offset, count, rw);
 
 	if (rw == WRITE && whint_mode == WHINT_MODE_OFF)
 		iocb->ki_hint = WRITE_LIFE_NOT_SET;
 
-	if (!down_read_trylock(&F2FS_I(inode)->i_gc_rwsem[rw])) {
-		if (iocb->ki_flags & IOCB_NOWAIT) {
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[rw])) {
+			iocb->ki_hint = hint;
+			err = -EAGAIN;
+			goto out;
+		}
+		if (lock_read && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[rw]);
 			iocb->ki_hint = hint;
 			err = -EAGAIN;
 			goto out;
 		}
-		down_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+	} else {
+		down_read(&fi->i_gc_rwsem[rw]);
+		if (lock_read)
+			down_read(&fi->i_gc_rwsem[READ]);
 	}
 
 	err = blockdev_direct_IO(iocb, inode, iter, get_data_block_dio);
-	up_read(&F2FS_I(inode)->i_gc_rwsem[rw]);
+
+	if (lock_read)
+		up_read(&fi->i_gc_rwsem[READ]);
+
+	up_read(&fi->i_gc_rwsem[rw]);
 
 	if (rw == WRITE) {
 		if (whint_mode == WHINT_MODE_OFF)
fs/f2fs/f2fs.h +35 -4
index 6f59ad180be7..a6dc9d581ee4 100644
@@ -8,6 +8,7 @@
 #ifndef _LINUX_F2FS_H
 #define _LINUX_F2FS_H
 
+#include <linux/uio.h>
 #include <linux/types.h>
 #include <linux/page-flags.h>
 #include <linux/buffer_head.h>
@@ -3461,11 +3462,41 @@ static inline bool f2fs_may_encrypt(struct inode *inode)
 #endif
 }
 
-static inline bool f2fs_force_buffered_io(struct inode *inode, int rw)
+static inline int block_unaligned_IO(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
 {
-	return (f2fs_post_read_required(inode) ||
-			(rw == WRITE && test_opt(F2FS_I_SB(inode), LFS)) ||
-			F2FS_I_SB(inode)->s_ndevs);
+	unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
+	unsigned int blocksize_mask = (1 << i_blkbits) - 1;
+	loff_t offset = iocb->ki_pos;
+	unsigned long align = offset | iov_iter_alignment(iter);
+
+	return align & blocksize_mask;
+}
+
+static inline int allow_outplace_dio(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	return (test_opt(sbi, LFS) && (rw == WRITE) &&
+				!block_unaligned_IO(inode, iocb, iter));
+}
+
+static inline bool f2fs_force_buffered_io(struct inode *inode,
+				struct kiocb *iocb, struct iov_iter *iter)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	int rw = iov_iter_rw(iter);
+
+	if (f2fs_post_read_required(inode))
+		return true;
+	if (sbi->s_ndevs)
+		return true;
+	if (test_opt(sbi, LFS) && (rw == WRITE) &&
+				block_unaligned_IO(inode, iocb, iter))
+		return true;
+	return false;
 }
 
 #ifdef CONFIG_F2FS_FAULT_INJECTION
fs/f2fs/file.c +2 -1
index 428e7398bd89..70491631e040 100644
@@ -3018,7 +3018,8 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 				if (!f2fs_overwrite_io(inode, iocb->ki_pos,
 						iov_iter_count(from)) ||
 					f2fs_has_inline_data(inode) ||
-					f2fs_force_buffered_io(inode, WRITE)) {
+					f2fs_force_buffered_io(inode,
+							iocb, from)) {
 						clear_inode_flag(inode,
 								FI_NO_PREALLOC);
 						inode_unlock(inode);
Signature · 1 line
2.18.0

.

.