[RFC PATCH 2/3] ocfs2: switch dio read path from buffer_head to iomap

This patch migrates the DIO read path from the legacy buffer_head
infrastructure to the modern and more efficient iomap framework.

One thing need to care: The generic iomap layer clears
`iocb->private` during its internal execution. To preserve the OCFS2
rw lock state across the DIO lifecycle, we leverage the top 2 bits
of `iocb->ki_flags` (`IOCB_OCFS2_RW_LOCK` and `IOCB_OCFS2_RW_LOCK_LEVEL`)
to safely carry the lock state into the `end_io` callback.

Signed-off-by: Heming Zhao <[email protected]>
---
 fs/ocfs2/Kconfig    |   1 +
 fs/ocfs2/aops.c     | 101 ++++++++++++++++++++++++++++++++++++++++++++
 fs/ocfs2/aops.h     |  24 +++++++++++
 fs/ocfs2/file.c     |  53 ++++++++++++++++++++---
 fs/ocfs2/ocfs2.h    |   3 ++
 fs/ocfs2/ocfs2_fs.h |   3 ++
 6 files changed, 179 insertions(+), 6 deletions(-)

diff --git a/fs/ocfs2/Kconfig b/fs/ocfs2/Kconfig
index 2514d36cbe01..bf1678a5eb01 100644
--- a/fs/ocfs2/Kconfig
+++ b/fs/ocfs2/Kconfig
@@ -7,6 +7,7 @@ config OCFS2_FS
 	select CRC32
 	select QUOTA
 	select QUOTA_TREE
+	select FS_IOMAP
 	select FS_POSIX_ACL
 	select LEGACY_DIRECT_IO
 	help
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 08df5e3b5196..59d445dd61f6 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -4,6 +4,7 @@
  */
 
 #include <linux/fs.h>
+#include <linux/iomap.h>
 #include <linux/slab.h>
 #include <linux/highmem.h>
 #include <linux/pagemap.h>
@@ -2556,6 +2557,106 @@ static ssize_t ocfs2_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 				    ocfs2_dio_end_io, 0);
 }
 
+static int ocfs2_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
+		unsigned int flags, struct iomap *iomap, struct iomap *srcmap)
+{
+	int ret;
+	struct ocfs2_map_block map;
+	struct ocfs2_inode_info *oi = OCFS2_I(inode);
+	u8 blkbits = inode->i_blkbits;
+
+	if ((offset >> blkbits) > OCFS2_MAX_LOGICAL_BLOCK)
+		return -EINVAL;
+
+	/*
+	 * Calculate the first and last logical blocks respectively.
+	 */
+	map.lblk = offset >> blkbits;
+	map.len = min_t(loff_t, (offset + length - 1) >> blkbits,
+			OCFS2_MAX_LOGICAL_BLOCK) - map.lblk + 1;
+	map.flags = 0;
+
+	if (flags & IOMAP_WRITE) {
+		/* todo */
+	} else {
+		down_read(&oi->ip_alloc_sem);
+		ret = ocfs2_map_blocks(inode, &map, 0);
+		up_read(&oi->ip_alloc_sem);
+	}
+
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Before returning to iomap, let's ensure the allocated mapping
+	 * covers the entire requested length for atomic writes.
+	 */
+	if (flags & IOMAP_ATOMIC) {
+		if (map.len < (length >> blkbits)) {
+			WARN_ON_ONCE(1);
+			return -EINVAL;
+		}
+	}
+
+	/* may use ocfs2_set_iomap() to replace all blow code */
+	iomap->bdev = inode->i_sb->s_bdev;
+	iomap->offset = (u64)map.lblk << blkbits;
+	iomap->length = (u64)map.len << blkbits;
+
+	if (map.pblk == 0) {
+		iomap->type = IOMAP_HOLE;
+		iomap->addr = IOMAP_NULL_ADDR;
+	} else {
+		if (map.flags & OCFS2_MAP_UNWRITTEN) {
+			iomap->type = IOMAP_UNWRITTEN;
+		} else if (map.flags & OCFS2_MAP_MAPPED) {
+			iomap->type = IOMAP_MAPPED;
+		} else {
+			WARN_ON_ONCE(1);
+			return -EIO;
+		}
+		iomap->addr = (sector_t)map.pblk << blkbits;
+	}
+
+	if (map.flags & OCFS2_MAP_NEW)
+		iomap->flags |= IOMAP_F_NEW;
+
+	return 0;
+}
+
+const struct iomap_ops ocfs2_iomap_ops = {
+	.iomap_begin	= ocfs2_iomap_begin,
+};
+
+static int ocfs2_dio_read_end_io(struct kiocb *iocb, ssize_t size, int error,
+		unsigned int flags)
+{
+	struct inode *inode = file_inode(iocb->ki_filp);
+	int level;
+
+	if (error) {
+		mlog_ratelimited(ML_ERROR, "Direct IO failed, bytes = %lld errno:%d",
+				 (long long)size, error);
+		goto bail;
+	}
+
+	BUG_ON(!ocfs2_iomap_iocb_is_rw_locked(iocb));
+
+	level = ocfs2_iomap_iocb_rw_locked_level(iocb);
+
+	/* sync dio unlock job done in .read_iter or .write_iter */
+	if (ocfs2_iomap_iocb_is_rw_locked(iocb))
+		ocfs2_rw_unlock(inode, level);
+
+	ocfs2_iocb_clear_rw_locked(iocb);
+	ocfs2_iomap_iocb_clear_rw_locked(iocb);
+bail:
+	return error;
+}
+
+const struct iomap_dio_ops ocfs2_dio_r_ops = {
+	.end_io = ocfs2_dio_read_end_io,
+};
 const struct address_space_operations ocfs2_aops = {
 	.dirty_folio		= block_dirty_folio,
 	.read_folio		= ocfs2_read_folio,
diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h
index 8dd6edd7c1a1..5966fa725135 100644
--- a/fs/ocfs2/aops.h
+++ b/fs/ocfs2/aops.h
@@ -74,4 +74,28 @@ enum ocfs2_iocb_lock_bits {
 #define ocfs2_iocb_rw_locked_level(iocb) \
 	test_bit(OCFS2_IOCB_RW_LOCK_LEVEL, (unsigned long *)&iocb->private)
 
+/*
+ * Since iomap clear iocb->private in some cases, we use iocb->ki_flags
+ * top 2-bit to store the rw lock state. Hope ki_flags never touch these 2-bit.
+ */
+#define IOCB_OCFS2_RW_LOCK		(1 << 30)
+#define IOCB_OCFS2_RW_LOCK_LEVEL	(1 << 31)
+static inline void ocfs2_iomap_iocb_set_rw_locked(struct kiocb *iocb, int level)
+{
+	iocb->ki_flags |= IOCB_OCFS2_RW_LOCK;
+	if (level)
+		iocb->ki_flags |= IOCB_OCFS2_RW_LOCK_LEVEL;
+	else
+		iocb->ki_flags &= ~IOCB_OCFS2_RW_LOCK_LEVEL;
+}
+
+#define ocfs2_iomap_iocb_is_rw_locked(iocb) \
+	(iocb->ki_flags & IOCB_OCFS2_RW_LOCK)
+#define ocfs2_iomap_iocb_init_rw_locked(iocb) \
+	(iocb->ki_flags &= ~(IOCB_OCFS2_RW_LOCK | IOCB_OCFS2_RW_LOCK_LEVEL))
+#define ocfs2_iomap_iocb_clear_rw_locked(iocb) \
+	(iocb->ki_flags &= ~IOCB_OCFS2_RW_LOCK)
+#define ocfs2_iomap_iocb_rw_locked_level(iocb) \
+	((iocb->ki_flags & IOCB_OCFS2_RW_LOCK_LEVEL) ? 1 : 0)
+
 #endif /* OCFS2_FILE_H */
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index d6e977ba6565..e879312ff86b 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -9,6 +9,7 @@
 
 #include <linux/capability.h>
 #include <linux/fs.h>
+#include <linux/iomap.h>
 #include <linux/types.h>
 #include <linux/slab.h>
 #include <linux/highmem.h>
@@ -2374,6 +2375,29 @@ static int ocfs2_prepare_inode_for_write(struct file *file,
 	return ret;
 }
 
+static bool ocfs2_should_use_dio(struct kiocb *iocb, struct iov_iter *iter,
+				struct inode *inode)
+{
+	unsigned int dio_align = bdev_logical_block_size(inode->i_sb->s_bdev);
+
+	/*
+	 * Fallback to buffered I/O if we see an inode without
+	 * extents.
+	 */
+	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL)
+		return false;
+
+	/*
+	 * Allow Direct I/O for any write requests aligned to the underlying
+	 * logical sector size (e.g., 512 bytes).
+	 */
+	if (IS_ALIGNED(iocb->ki_pos | iov_iter_alignment(iter), dio_align)) {
+		return true;
+	} else {
+		return false;
+	}
+}
+
 static ssize_t ocfs2_file_write_iter(struct kiocb *iocb,
 				    struct iov_iter *from)
 {
@@ -2548,7 +2572,6 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb,
 			filp->f_path.dentry->d_name.name,
 			to->nr_segs);	/* GRRRRR */
 
-
 	if (!inode) {
 		ret = -EINVAL;
 		mlog_errno(ret);
@@ -2558,26 +2581,37 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb,
 	if (!direct_io && nowait)
 		return -EOPNOTSUPP;
 
+	if (!iov_iter_count(to))
+		return 0; /* skip atime */
+
+	/*
+	 * the following two lines do the same thing. Once iomap completely
+	 * replaces buffer_head, we can remove ocfs2_iocb_init_rw_locked(),
+	 * ocfs2_iocb_clear_rw_locked() and ocfs2_iocb_rw_locked_level().
+	 */
 	ocfs2_iocb_init_rw_locked(iocb);
+	ocfs2_iomap_iocb_init_rw_locked(iocb);
 
 	/*
 	 * buffered reads protect themselves in ->read_folio().  O_DIRECT reads
 	 * need locks to protect pending reads from racing with truncate.
 	 */
 	if (direct_io) {
+		rw_level = 0;
 		if (nowait)
-			ret = ocfs2_try_rw_lock(inode, 0);
+			ret = ocfs2_try_rw_lock(inode, rw_level);
 		else
-			ret = ocfs2_rw_lock(inode, 0);
+			ret = ocfs2_rw_lock(inode, rw_level);
 
 		if (ret < 0) {
 			if (ret != -EAGAIN)
 				mlog_errno(ret);
 			goto bail;
 		}
-		rw_level = 0;
 		/* communicate with ocfs2_dio_end_io */
 		ocfs2_iocb_set_rw_locked(iocb, rw_level);
+		/* communicate with ocfs2_dio_read_end_io */
+		ocfs2_iomap_iocb_set_rw_locked(iocb, rw_level);
 	}
 
 	/*
@@ -2598,7 +2632,12 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb,
 	}
 	ocfs2_inode_unlock(inode, lock_level);
 
-	ret = generic_file_read_iter(iocb, to);
+	if (direct_io && ocfs2_should_use_dio(iocb, to, inode)) {
+		ret = iomap_dio_rw(iocb, to, &ocfs2_iomap_ops, &ocfs2_dio_r_ops, 0, NULL, 0);
+	} else {
+		iocb->ki_flags &= ~IOCB_DIRECT;
+		ret = generic_file_read_iter(iocb, to);
+	}
 	trace_generic_file_read_iter_ret(ret);
 
 	/* buffered aio wouldn't have proper lock coverage today */
@@ -2610,8 +2649,10 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb,
 	}
 
 bail:
-	if (rw_level != -1)
+	if (rw_level != -1) {
 		ocfs2_rw_unlock(inode, rw_level);
+		ocfs2_iomap_iocb_clear_rw_locked(iocb);
+	}
 
 	return ret;
 }
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 095f7ae5dded..9397947583fb 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -549,6 +549,9 @@ struct ocfs2_map_block {
 	unsigned int flags;
 };
 
+extern const struct iomap_ops ocfs2_iomap_ops;
+extern const struct iomap_dio_ops ocfs2_dio_r_ops;
+
 /* Flags used by ocfs2_map_blocks() */
 #define OCFS2_GET_BLOCKS_CREATE	(0x0001)
 
diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h
index c501eb3cdcda..000bd014acbd 100644
--- a/fs/ocfs2/ocfs2_fs.h
+++ b/fs/ocfs2/ocfs2_fs.h
@@ -314,6 +314,9 @@
  */
 #define OCFS2_CLUSTER_O2CB_GLOBAL_HEARTBEAT	(0x01)
 
+/* Max logical block we can support */
+#define OCFS2_MAX_LOGICAL_BLOCK	(0xFFFFFFFE)
+
 struct ocfs2_system_inode_info {
 	char	*si_name;
 	int	si_iflags;
-- 
2.54.0