On 9/9/19 4:48 AM, Matthew Bobrowski wrote:
This patch introduces a new direct IO read path that makes use of the
iomap infrastructure.
The new function ext4_dio_read_iter() is responsible for calling into
the iomap infrastructure via iomap_dio_rw(). If the inode in question
does not pass preliminary checks in ext4_dio_checks(), then we simply
fallback to buffered IO and take that path to fulfil the request. It's
imperative that we drop the IOCB_DIRECT flag from iocb->ki_flags in
order to prevent generic_file_read_iter() from trying to take the
direct IO code path again.
Signed-off-by: Matthew Bobrowski <mbobrowski@xxxxxxxxxxxxxx>
Looks good to me.
Reviewed-by: Ritesh Harjani <riteshh@xxxxxxxxxxxxx>
---
fs/ext4/file.c | 58 ++++++++++++++++++++++++++++++++++++++++++++++----
1 file changed, 54 insertions(+), 4 deletions(-)
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 70b0438dbc94..e52e3928dc25 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -34,6 +34,53 @@
#include "xattr.h"
#include "acl.h"
+static bool ext4_dio_checks(struct inode *inode)
+{
+#if IS_ENABLED(CONFIG_FS_ENCRYPTION)
+ if (IS_ENCRYPTED(inode))
+ return false;
+#endif
+ if (ext4_should_journal_data(inode))
+ return false;
+ if (ext4_has_inline_data(inode))
+ return false;
+ return true;
+}
+
+static ssize_t ext4_dio_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+ ssize_t ret;
+ struct inode *inode = file_inode(iocb->ki_filp);
+
+ /*
+ * Get exclusion from truncate and other inode operations.
+ */
+ if (!inode_trylock_shared(inode)) {
+ if (iocb->ki_flags & IOCB_NOWAIT)
+ return -EAGAIN;
+ inode_lock_shared(inode);
+ }
+
+ if (!ext4_dio_checks(inode)) {
+ inode_unlock_shared(inode);
+ /*
+ * Fallback to buffered IO if the operation being
+ * performed on the inode is not supported by direct
+ * IO. The IOCB_DIRECT flag flags needs to be cleared
+ * here to ensure that the direct IO code path within
+ * generic_file_read_iter() is not taken again.
+ */
+ iocb->ki_flags &= ~IOCB_DIRECT;
+ return generic_file_read_iter(iocb, to);
+ }
+
+ ret = iomap_dio_rw(iocb, to, &ext4_iomap_ops, NULL);
+ inode_unlock_shared(inode);
+
+ file_accessed(iocb->ki_filp);
+ return ret;
+}
+
#ifdef CONFIG_FS_DAX
static ssize_t ext4_dax_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
@@ -64,16 +111,19 @@ static ssize_t ext4_dax_read_iter(struct kiocb *iocb, struct iov_iter *to)
static ssize_t ext4_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
- if (unlikely(ext4_forced_shutdown(EXT4_SB(file_inode(iocb->ki_filp)->i_sb))))
+ struct inode *inode = file_inode(iocb->ki_filp);
+
+ if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
return -EIO;
if (!iov_iter_count(to))
return 0; /* skip atime */
-#ifdef CONFIG_FS_DAX
- if (IS_DAX(file_inode(iocb->ki_filp)))
+ if (IS_DAX(inode))
return ext4_dax_read_iter(iocb, to);
-#endif
+
+ if (iocb->ki_flags & IOCB_DIRECT)
+ return ext4_dio_read_iter(iocb, to);
return generic_file_read_iter(iocb, to);
}
