diff mbox series

[v4,26/46] btrfs: keep track of fscrypt info and orig_start for dio reads

Message ID 5e74c0395e4f58082b9446fb0105c0cb99e8338d.1701468306.git.josef@toxicpanda.com (mailing list archive)
State New, archived
Headers show
Series btrfs: add fscrypt support | expand

Commit Message

Josef Bacik Dec. 1, 2023, 10:11 p.m. UTC 
We keep track of this information in the ordered extent for writes, but
we need it for reads as well.  Add fscrypt_extent_info and orig_start to
the dio_data so we can populate this on reads.  This will be used later
when we attach the fscrypt context to the bios.

Signed-off-by: Josef Bacik <josef@toxicpanda.com>
---
 fs/btrfs/inode.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

Comments

Eric Biggers Dec. 5, 2023, 5:44 a.m. UTC | #1 
On Fri, Dec 01, 2023 at 05:11:23PM -0500, Josef Bacik wrote:
> We keep track of this information in the ordered extent for writes, but
> we need it for reads as well.  Add fscrypt_extent_info and orig_start to
> the dio_data so we can populate this on reads.  This will be used later
> when we attach the fscrypt context to the bios.
> 
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>

Does this mean that btrfs will support direct I/O on encrypted files from the
beginning?  Are you enforcing FS block alignment, as is required for encrypted
direct I/O?

- Eric
diff mbox series

Patch

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 9c704b1cfe05..dbdc01f25215 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -83,6 +83,8 @@  struct btrfs_dio_data {
 	ssize_t submitted;
 	struct extent_changeset *data_reserved;
 	struct btrfs_ordered_extent *ordered;
+	struct fscrypt_extent_info *fscrypt_info;
+	u64 orig_start;
 	bool data_space_reserved;
 	bool nocow_done;
 };
@@ -7767,6 +7769,10 @@  static int btrfs_dio_iomap_begin(struct inode *inode, loff_t start,
 							       release_len);
 		}
 	} else {
+		dio_data->fscrypt_info =
+			fscrypt_get_extent_info(em->fscrypt_info);
+		dio_data->orig_start = em->orig_start;
+
 		/*
 		 * We need to unlock only the end area that we aren't using.
 		 * The rest is going to be unlocked by the endio routine.
@@ -7848,6 +7854,11 @@  static int btrfs_dio_iomap_end(struct inode *inode, loff_t pos, loff_t length,
 		dio_data->ordered = NULL;
 	}
 
+	if (dio_data->fscrypt_info) {
+		fscrypt_put_extent_info(dio_data->fscrypt_info);
+		dio_data->fscrypt_info = NULL;
+	}
+
 	if (write)
 		extent_changeset_free(dio_data->data_reserved);
 	return ret;