Re: [RFC] what to do with IOCB_DSYNC?

[Matthew Wilcox <willy@infradead.org>]

On Mon, May 23, 2022 at 12:42:33AM +0000, Al Viro wrote:
> At the moment it's carved up into 6 commits:
> 	btrfs_direct_write(): cleaner way to handle generic_write_sync() suppression

I happpen to need something similar for JFS, so I tried it out.
There's a mistake where dio->flags should have been dio_flags.
I also added a comment on the definition of DIO_NOSYNC.

(the btrfs patch didn't apply cleanly, so I just dropped it since I
wasn't going to be touching btrfs anyway)

diff --git a/fs/iomap/direct-io.c b/fs/iomap/direct-io.c
index 370c3241618a..6a361131080f 100644
--- a/fs/iomap/direct-io.c
+++ b/fs/iomap/direct-io.c
@@ -548,7 +548,8 @@ __iomap_dio_rw(struct kiocb *iocb, struct iov_iter *iter,
 		}
 
 		/* for data sync or sync, we need sync completion processing */
-		if (iocb->ki_flags & IOCB_DSYNC)
+		if (iocb->ki_flags & IOCB_DSYNC &&
+		    !(dio_flags & IOMAP_DIO_NOSYNC))
 			dio->flags |= IOMAP_DIO_NEED_SYNC;
 
 		/*
diff --git a/include/linux/iomap.h b/include/linux/iomap.h
index e552097c67e0..c8622d8f064e 100644
--- a/include/linux/iomap.h
+++ b/include/linux/iomap.h
@@ -353,6 +353,12 @@ struct iomap_dio_ops {
  */
 #define IOMAP_DIO_PARTIAL		(1 << 2)
 
+/*
+ * The caller will sync the write if needed; do not sync it within
+ * iomap_dio_rw.  Overrides IOMAP_DIO_FORCE_WAIT.
+ */
+#define IOMAP_DIO_NOSYNC		(1 << 3)
+
 ssize_t iomap_dio_rw(struct kiocb *iocb, struct iov_iter *iter,
 		const struct iomap_ops *ops, const struct iomap_dio_ops *dops,
 		unsigned int dio_flags, void *private, size_t done_before);