Thread (37 messages) 37 messages, 3 authors, 2011-10-21

Re: [PATCH 12/28] ext4: Use i_generation in inode-related metadata checksums

From: Andreas Dilger <adilger.kernel@dilger.ca>
Date: 2011-10-13 00:06:43
Also in: linux-fsdevel, lkml

On 2011-10-12, at 3:28 PM, Darrick J. Wong wrote:
On Wed, Oct 12, 2011 at 12:52:30PM -0700, Andreas Dilger wrote:
quoted
On 2011-10-08, at 12:55 AM, Darrick J. Wong wrote:
quoted
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index f18bfe3..fdf0b1e 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -149,6 +149,10 @@ flags_out:
		if (!inode_owner_or_capable(inode))
			return -EPERM;

+		if (EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb,
+				EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
+			return -ENOTTY;
This should get an ext4_warning() in the non-checksum case to warn
users that this ioctl is deprecated and will be removed in the
future unless there is a good reason to keep it.
Ok.  Maybe put it in feature_removal_schedule.txt too?  Maybe not; the
ioctl is not being totally removed, it's merely unsupported for the
metadata_csum case.
I understand it is not being removed entirely, but since it doesn't
work at all with the checksum-enabled case, we may as well mark it
deprecated for the non-checksum case to allow us to catch any users
(however unlikely I think it is) earlier rather than later.

Cheers, Andreas




Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help