1 files changed, 415 insertions, 94 deletions

diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 92ec2699bc85..3c98ef6af97d 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -24,6 +24,7 @@
 #include <linux/sched/signal.h>
 #include <linux/fileattr.h>
 #include <linux/fadvise.h>
+#include <linux/iomap.h>
 
 #include "f2fs.h"
 #include "node.h"
@@ -1232,7 +1233,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode,
 			if (ret)
 				return ret;
 
-			ret = f2fs_get_node_info(sbi, dn.nid, &ni);
+			ret = f2fs_get_node_info(sbi, dn.nid, &ni, false);
 			if (ret) {
 				f2fs_put_dnode(&dn);
 				return ret;
@@ -1687,6 +1688,7 @@ next_alloc:
 
 		map.m_seg_type = CURSEG_COLD_DATA_PINNED;
 		err = f2fs_map_blocks(inode, &map, 1, F2FS_GET_BLOCK_PRE_DIO);
+		file_dont_truncate(inode);
 
 		up_write(&sbi->pin_sem);
 
@@ -1748,7 +1750,11 @@ static long f2fs_fallocate(struct file *file, int mode,
 		(mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE)))
 		return -EOPNOTSUPP;
 
-	if (f2fs_compressed_file(inode) &&
+	/*
+	 * Pinned file should not support partial trucation since the block
+	 * can be used by applications.
+	 */
+	if ((f2fs_compressed_file(inode) || f2fs_is_pinned_file(inode)) &&
 		(mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE |
 			FALLOC_FL_ZERO_RANGE | FALLOC_FL_INSERT_RANGE)))
 		return -EOPNOTSUPP;
@@ -3143,17 +3149,17 @@ static int f2fs_ioc_set_pin_file(struct file *filp, unsigned long arg)
 
 	inode_lock(inode);
 
-	if (f2fs_should_update_outplace(inode, NULL)) {
-		ret = -EINVAL;
-		goto out;
-	}
-
 	if (!pin) {
 		clear_inode_flag(inode, FI_PIN_FILE);
 		f2fs_i_gc_failures_write(inode, 0);
 		goto done;
 	}
 
+	if (f2fs_should_update_outplace(inode, NULL)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
 	if (f2fs_pin_file_control(inode, false)) {
 		ret = -EAGAIN;
 		goto out;
@@ -4218,27 +4224,385 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	return __f2fs_ioctl(filp, cmd, arg);
 }
 
-static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter)
+/*
+ * Return %true if the given read or write request should use direct I/O, or
+ * %false if it should use buffered I/O.
+ */
+static bool f2fs_should_use_dio(struct inode *inode, struct kiocb *iocb,
+				struct iov_iter *iter)
+{
+	unsigned int align;
+
+	if (!(iocb->ki_flags & IOCB_DIRECT))
+		return false;
+
+	if (f2fs_force_buffered_io(inode, iocb, iter))
+		return false;
+
+	/*
+	 * Direct I/O not aligned to the disk's logical_block_size will be
+	 * attempted, but will fail with -EINVAL.
+	 *
+	 * f2fs additionally requires that direct I/O be aligned to the
+	 * filesystem block size, which is often a stricter requirement.
+	 * However, f2fs traditionally falls back to buffered I/O on requests
+	 * that are logical_block_size-aligned but not fs-block aligned.
+	 *
+	 * The below logic implements this behavior.
+	 */
+	align = iocb->ki_pos | iov_iter_alignment(iter);
+	if (!IS_ALIGNED(align, i_blocksize(inode)) &&
+	    IS_ALIGNED(align, bdev_logical_block_size(inode->i_sb->s_bdev)))
+		return false;
+
+	return true;
+}
+
+static int f2fs_dio_read_end_io(struct kiocb *iocb, ssize_t size, int error,
+				unsigned int flags)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp));
+
+	dec_page_count(sbi, F2FS_DIO_READ);
+	if (error)
+		return error;
+	f2fs_update_iostat(sbi, APP_DIRECT_READ_IO, size);
+	return 0;
+}
+
+static const struct iomap_dio_ops f2fs_iomap_dio_read_ops = {
+	.end_io = f2fs_dio_read_end_io,
+};
+
+static ssize_t f2fs_dio_read_iter(struct kiocb *iocb, struct iov_iter *to)
 {
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file_inode(file);
-	int ret;
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
+	const loff_t pos = iocb->ki_pos;
+	const size_t count = iov_iter_count(to);
+	struct iomap_dio *dio;
+	ssize_t ret;
+
+	if (count == 0)
+		return 0; /* skip atime update */
+
+	trace_f2fs_direct_IO_enter(inode, iocb, count, READ);
+
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		if (!down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			ret = -EAGAIN;
+			goto out;
+		}
+	} else {
+		down_read(&fi->i_gc_rwsem[READ]);
+	}
+
+	/*
+	 * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of
+	 * the higher-level function iomap_dio_rw() in order to ensure that the
+	 * F2FS_DIO_READ counter will be decremented correctly in all cases.
+	 */
+	inc_page_count(sbi, F2FS_DIO_READ);
+	dio = __iomap_dio_rw(iocb, to, &f2fs_iomap_ops,
+			     &f2fs_iomap_dio_read_ops, 0, 0);
+	if (IS_ERR_OR_NULL(dio)) {
+		ret = PTR_ERR_OR_ZERO(dio);
+		if (ret != -EIOCBQUEUED)
+			dec_page_count(sbi, F2FS_DIO_READ);
+	} else {
+		ret = iomap_dio_complete(dio);
+	}
+
+	up_read(&fi->i_gc_rwsem[READ]);
+
+	file_accessed(file);
+out:
+	trace_f2fs_direct_IO_exit(inode, pos, count, READ, ret);
+	return ret;
+}
+
+static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+	struct inode *inode = file_inode(iocb->ki_filp);
+	ssize_t ret;
 
 	if (!f2fs_is_compress_backend_ready(inode))
 		return -EOPNOTSUPP;
 
-	ret = generic_file_read_iter(iocb, iter);
+	if (f2fs_should_use_dio(inode, iocb, to))
+		return f2fs_dio_read_iter(iocb, to);
 
+	ret = filemap_read(iocb, to, 0);
 	if (ret > 0)
-		f2fs_update_iostat(F2FS_I_SB(inode), APP_READ_IO, ret);
+		f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_READ_IO, ret);
+	return ret;
+}
+
+static ssize_t f2fs_write_checks(struct kiocb *iocb, struct iov_iter *from)
+{
+	struct file *file = iocb->ki_filp;
+	struct inode *inode = file_inode(file);
+	ssize_t count;
+	int err;
+
+	if (IS_IMMUTABLE(inode))
+		return -EPERM;
+
+	if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED))
+		return -EPERM;
+
+	count = generic_write_checks(iocb, from);
+	if (count <= 0)
+		return count;
+
+	err = file_modified(file);
+	if (err)
+		return err;
+	return count;
+}
+
+/*
+ * Preallocate blocks for a write request, if it is possible and helpful to do
+ * so.  Returns a positive number if blocks may have been preallocated, 0 if no
+ * blocks were preallocated, or a negative errno value if something went
+ * seriously wrong.  Also sets FI_PREALLOCATED_ALL on the inode if *all* the
+ * requested blocks (not just some of them) have been allocated.
+ */
+static int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *iter,
+				   bool dio)
+{
+	struct inode *inode = file_inode(iocb->ki_filp);
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	const loff_t pos = iocb->ki_pos;
+	const size_t count = iov_iter_count(iter);
+	struct f2fs_map_blocks map = {};
+	int flag;
+	int ret;
+
+	/* If it will be an out-of-place direct write, don't bother. */
+	if (dio && f2fs_lfs_mode(sbi))
+		return 0;
+	/*
+	 * Don't preallocate holes aligned to DIO_SKIP_HOLES which turns into
+	 * buffered IO, if DIO meets any holes.
+	 */
+	if (dio && i_size_read(inode) &&
+		(F2FS_BYTES_TO_BLK(pos) < F2FS_BLK_ALIGN(i_size_read(inode))))
+		return 0;
+
+	/* No-wait I/O can't allocate blocks. */
+	if (iocb->ki_flags & IOCB_NOWAIT)
+		return 0;
+
+	/* If it will be a short write, don't bother. */
+	if (fault_in_iov_iter_readable(iter, count))
+		return 0;
+
+	if (f2fs_has_inline_data(inode)) {
+		/* If the data will fit inline, don't bother. */
+		if (pos + count <= MAX_INLINE_DATA(inode))
+			return 0;
+		ret = f2fs_convert_inline_inode(inode);
+		if (ret)
+			return ret;
+	}
+
+	/* Do not preallocate blocks that will be written partially in 4KB. */
+	map.m_lblk = F2FS_BLK_ALIGN(pos);
+	map.m_len = F2FS_BYTES_TO_BLK(pos + count);
+	if (map.m_len > map.m_lblk)
+		map.m_len -= map.m_lblk;
+	else
+		map.m_len = 0;
+	map.m_may_create = true;
+	if (dio) {
+		map.m_seg_type = f2fs_rw_hint_to_seg_type(inode->i_write_hint);
+		flag = F2FS_GET_BLOCK_PRE_DIO;
+	} else {
+		map.m_seg_type = NO_CHECK_TYPE;
+		flag = F2FS_GET_BLOCK_PRE_AIO;
+	}
+
+	ret = f2fs_map_blocks(inode, &map, 1, flag);
+	/* -ENOSPC|-EDQUOT are fine to report the number of allocated blocks. */
+	if (ret < 0 && !((ret == -ENOSPC || ret == -EDQUOT) && map.m_len > 0))
+		return ret;
+	if (ret == 0)
+		set_inode_flag(inode, FI_PREALLOCATED_ALL);
+	return map.m_len;
+}
+
+static ssize_t f2fs_buffered_write_iter(struct kiocb *iocb,
+					struct iov_iter *from)
+{
+	struct file *file = iocb->ki_filp;
+	struct inode *inode = file_inode(file);
+	ssize_t ret;
 
+	if (iocb->ki_flags & IOCB_NOWAIT)
+		return -EOPNOTSUPP;
+
+	current->backing_dev_info = inode_to_bdi(inode);
+	ret = generic_perform_write(file, from, iocb->ki_pos);
+	current->backing_dev_info = NULL;
+
+	if (ret > 0) {
+		iocb->ki_pos += ret;
+		f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_IO, ret);
+	}
 	return ret;
 }
 
-static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
+static int f2fs_dio_write_end_io(struct kiocb *iocb, ssize_t size, int error,
+				 unsigned int flags)
+{
+	struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp));
+
+	dec_page_count(sbi, F2FS_DIO_WRITE);
+	if (error)
+		return error;
+	f2fs_update_iostat(sbi, APP_DIRECT_IO, size);
+	return 0;
+}
+
+static const struct iomap_dio_ops f2fs_iomap_dio_write_ops = {
+	.end_io = f2fs_dio_write_end_io,
+};
+
+static ssize_t f2fs_dio_write_iter(struct kiocb *iocb, struct iov_iter *from,
+				   bool *may_need_sync)
 {
 	struct file *file = iocb->ki_filp;
 	struct inode *inode = file_inode(file);
+	struct f2fs_inode_info *fi = F2FS_I(inode);
+	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+	const bool do_opu = f2fs_lfs_mode(sbi);
+	const int whint_mode = F2FS_OPTION(sbi).whint_mode;
+	const loff_t pos = iocb->ki_pos;
+	const ssize_t count = iov_iter_count(from);
+	const enum rw_hint hint = iocb->ki_hint;
+	unsigned int dio_flags;
+	struct iomap_dio *dio;
+	ssize_t ret;
+
+	trace_f2fs_direct_IO_enter(inode, iocb, count, WRITE);
+
+	if (iocb->ki_flags & IOCB_NOWAIT) {
+		/* f2fs_convert_inline_inode() and block allocation can block */
+		if (f2fs_has_inline_data(inode) ||
+		    !f2fs_overwrite_io(inode, pos, count)) {
+			ret = -EAGAIN;
+			goto out;
+		}
+
+		if (!down_read_trylock(&fi->i_gc_rwsem[WRITE])) {
+			ret = -EAGAIN;
+			goto out;
+		}
+		if (do_opu && !down_read_trylock(&fi->i_gc_rwsem[READ])) {
+			up_read(&fi->i_gc_rwsem[WRITE]);
+			ret = -EAGAIN;
+			goto out;
+		}
+	} else {
+		ret = f2fs_convert_inline_inode(inode);
+		if (ret)
+			goto out;
+
+		down_read(&fi->i_gc_rwsem[WRITE]);
+		if (do_opu)
+			down_read(&fi->i_gc_rwsem[READ]);
+	}
+	if (whint_mode == WHINT_MODE_OFF)
+		iocb->ki_hint = WRITE_LIFE_NOT_SET;
+
+	/*
+	 * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of
+	 * the higher-level function iomap_dio_rw() in order to ensure that the
+	 * F2FS_DIO_WRITE counter will be decremented correctly in all cases.
+	 */
+	inc_page_count(sbi, F2FS_DIO_WRITE);
+	dio_flags = 0;
+	if (pos + count > inode->i_size)
+		dio_flags |= IOMAP_DIO_FORCE_WAIT;
+	dio = __iomap_dio_rw(iocb, from, &f2fs_iomap_ops,
+			     &f2fs_iomap_dio_write_ops, dio_flags, 0);
+	if (IS_ERR_OR_NULL(dio)) {
+		ret = PTR_ERR_OR_ZERO(dio);
+		if (ret == -ENOTBLK)
+			ret = 0;
+		if (ret != -EIOCBQUEUED)
+			dec_page_count(sbi, F2FS_DIO_WRITE);
+	} else {
+		ret = iomap_dio_complete(dio);
+	}
+
+	if (whint_mode == WHINT_MODE_OFF)
+		iocb->ki_hint = hint;
+	if (do_opu)
+		up_read(&fi->i_gc_rwsem[READ]);
+	up_read(&fi->i_gc_rwsem[WRITE]);
+
+	if (ret < 0)
+		goto out;
+	if (pos + ret > inode->i_size)
+		f2fs_i_size_write(inode, pos + ret);
+	if (!do_opu)
+		set_inode_flag(inode, FI_UPDATE_WRITE);
+
+	if (iov_iter_count(from)) {
+		ssize_t ret2;
+		loff_t bufio_start_pos = iocb->ki_pos;
+
+		/*
+		 * The direct write was partial, so we need to fall back to a
+		 * buffered write for the remainder.
+		 */
+
+		ret2 = f2fs_buffered_write_iter(iocb, from);
+		if (iov_iter_count(from))
+			f2fs_write_failed(inode, iocb->ki_pos);
+		if (ret2 < 0)
+			goto out;
+
+		/*
+		 * Ensure that the pagecache pages are written to disk and
+		 * invalidated to preserve the expected O_DIRECT semantics.
+		 */
+		if (ret2 > 0) {
+			loff_t bufio_end_pos = bufio_start_pos + ret2 - 1;
+
+			ret += ret2;
+
+			ret2 = filemap_write_and_wait_range(file->f_mapping,
+							    bufio_start_pos,
+							    bufio_end_pos);
+			if (ret2 < 0)
+				goto out;
+			invalidate_mapping_pages(file->f_mapping,
+						 bufio_start_pos >> PAGE_SHIFT,
+						 bufio_end_pos >> PAGE_SHIFT);
+		}
+	} else {
+		/* iomap_dio_rw() already handled the generic_write_sync(). */
+		*may_need_sync = false;
+	}
+out:
+	trace_f2fs_direct_IO_exit(inode, pos, count, WRITE, ret);
+	return ret;
+}
+
+static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
+{
+	struct inode *inode = file_inode(iocb->ki_filp);
+	const loff_t orig_pos = iocb->ki_pos;
+	const size_t orig_count = iov_iter_count(from);
+	loff_t target_size;
+	bool dio;
+	bool may_need_sync = true;
+	int preallocated;
 	ssize_t ret;
 
 	if (unlikely(f2fs_cp_error(F2FS_I_SB(inode)))) {
@@ -4260,91 +4624,42 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
 		inode_lock(inode);
 	}
 
-	if (unlikely(IS_IMMUTABLE(inode))) {
-		ret = -EPERM;
-		goto unlock;
-	}
-
-	if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) {
-		ret = -EPERM;
-		goto unlock;
-	}
-
-	ret = generic_write_checks(iocb, from);
-	if (ret > 0) {
-		bool preallocated = false;
-		size_t target_size = 0;
-		int err;
-
-		if (fault_in_iov_iter_readable(from, iov_iter_count(from)))
-			set_inode_flag(inode, FI_NO_PREALLOC);
-
-		if ((iocb->ki_flags & IOCB_NOWAIT)) {
-			if (!f2fs_overwrite_io(inode, iocb->ki_pos,
-						iov_iter_count(from)) ||
-				f2fs_has_inline_data(inode) ||
-				f2fs_force_buffered_io(inode, iocb, from)) {
-				clear_inode_flag(inode, FI_NO_PREALLOC);
-				inode_unlock(inode);
-				ret = -EAGAIN;
-				goto out;
-			}
-			goto write;
-		}
-
-		if (is_inode_flag_set(inode, FI_NO_PREALLOC))
-			goto write;
-
-		if (iocb->ki_flags & IOCB_DIRECT) {
-			/*
-			 * Convert inline data for Direct I/O before entering
-			 * f2fs_direct_IO().
-			 */
-			err = f2fs_convert_inline_inode(inode);
-			if (err)
-				goto out_err;
-			/*
-			 * If force_buffere_io() is true, we have to allocate
-			 * blocks all the time, since f2fs_direct_IO will fall
-			 * back to buffered IO.
-			 */
-			if (!f2fs_force_buffered_io(inode, iocb, from) &&
-					f2fs_lfs_mode(F2FS_I_SB(inode)))
-				goto write;
-		}
-		preallocated = true;
-		target_size = iocb->ki_pos + iov_iter_count(from);
+	ret = f2fs_write_checks(iocb, from);
+	if (ret <= 0)
+		goto out_unlock;
 
-		err = f2fs_preallocate_blocks(iocb, from);
-		if (err) {
-out_err:
-			clear_inode_flag(inode, FI_NO_PREALLOC);
-			inode_unlock(inode);
-			ret = err;
-			goto out;
-		}
-write:
-		ret = __generic_file_write_iter(iocb, from);
-		clear_inode_flag(inode, FI_NO_PREALLOC);
+	/* Determine whether we will do a direct write or a buffered write. */
+	dio = f2fs_should_use_dio(inode, iocb, from);
 
-		/* if we couldn't write data, we should deallocate blocks. */
-		if (preallocated && i_size_read(inode) < target_size) {
-			down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
-			filemap_invalidate_lock(inode->i_mapping);
-			f2fs_truncate(inode);
-			filemap_invalidate_unlock(inode->i_mapping);
-			up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
-		}
+	/* Possibly preallocate the blocks for the write. */
+	target_size = iocb->ki_pos + iov_iter_count(from);
+	preallocated = f2fs_preallocate_blocks(iocb, from, dio);
+	if (preallocated < 0)
+		ret = preallocated;
+	else
+		/* Do the actual write. */
+		ret = dio ?
+			f2fs_dio_write_iter(iocb, from, &may_need_sync):
+			f2fs_buffered_write_iter(iocb, from);
 
-		if (ret > 0)
-			f2fs_update_iostat(F2FS_I_SB(inode), APP_WRITE_IO, ret);
+	/* Don't leave any preallocated blocks around past i_size. */
+	if (preallocated && i_size_read(inode) < target_size) {
+		down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+		filemap_invalidate_lock(inode->i_mapping);
+		if (!f2fs_truncate(inode))
+			file_dont_truncate(inode);
+		filemap_invalidate_unlock(inode->i_mapping);
+		up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+	} else {
+		file_dont_truncate(inode);
 	}
-unlock:
+
+	clear_inode_flag(inode, FI_PREALLOCATED_ALL);
+out_unlock:
 	inode_unlock(inode);
 out:
-	trace_f2fs_file_write_iter(inode, iocb->ki_pos,
-					iov_iter_count(from), ret);
-	if (ret > 0)
+	trace_f2fs_file_write_iter(inode, orig_pos, orig_count, ret);
+	if (ret > 0 && may_need_sync)
 		ret = generic_write_sync(iocb, ret);
 	return ret;
 }
@@ -4352,12 +4667,12 @@ out:
 static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len,
 		int advice)
 {
-	struct inode *inode;
 	struct address_space *mapping;
 	struct backing_dev_info *bdi;
+	struct inode *inode = file_inode(filp);
+	int err;
 
 	if (advice == POSIX_FADV_SEQUENTIAL) {
-		inode = file_inode(filp);
 		if (S_ISFIFO(inode->i_mode))
 			return -ESPIPE;
 
@@ -4374,7 +4689,13 @@ static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len,
 		return 0;
 	}
 
-	return generic_fadvise(filp, offset, len, advice);
+	err = generic_fadvise(filp, offset, len, advice);
+	if (!err && advice == POSIX_FADV_DONTNEED &&
+		test_opt(F2FS_I_SB(inode), COMPRESS_CACHE) &&
+		f2fs_compressed_file(inode))
+		f2fs_invalidate_compress_pages(F2FS_I_SB(inode), inode->i_ino);
+
+	return err;
 }
 
 #ifdef CONFIG_COMPAT