This ioctl tries to second guess direct I/O parameters which aren't
a filesystem drivers business and shouldn't be exposed as an ioctl
to start with.
Signed-off-by: Christoph Hellwig <[email protected]>
Index: linux-2.6/fs/xfs/linux-2.6/xfs_ioctl32.c
===================================================================
--- linux-2.6.orig/fs/xfs/linux-2.6/xfs_ioctl32.c 2005-11-18 10:29:41.000000000 +0100
+++ linux-2.6/fs/xfs/linux-2.6/xfs_ioctl32.c 2005-11-28 23:21:39.000000000 +0100
@@ -115,7 +115,6 @@
vnode_t *vp = LINVFS_GET_VP(inode);
switch (cmd) {
- case XFS_IOC_DIOINFO:
case XFS_IOC_FSGEOMETRY_V1:
case XFS_IOC_FSGEOMETRY:
case XFS_IOC_GETVERSION:
On Mon, Nov 28, 2005 at 11:25:01PM +0100, Christoph Hellwig wrote:
> This ioctl tries to second guess direct I/O parameters which aren't
> a filesystem drivers business and shouldn't be exposed as an ioctl
> to start with.
Unfortunately there are some applications that will now start
to see errors from this ioctl if we go this route - whereas
before they would've been "functional", now they will break.
So, I think we need a different solution here. Yes, I agree
its a stupid call to have on Linux, but here we are, and apps
ported straight from IRIX have been made to use this.
cheers.
--
Nathan