Message ID | 20150730044502.GA21118@canonical.com |
---|---|
State | New |
Headers | show |
Kind of a giant patch set for such a small problem. I'd like to see the usual battery of regression test results in the bug report.
On 07/30/2015 09:25 AM, Tim Gardner wrote: > Kind of a giant patch set for such a small problem. I'd like to see the > usual battery of regression test results in the bug report. > Agree, before I sent this request I tested the reproducer script in a VM with and without these patches. They do in fact fix what they claim. In addition I just ran the unionmount-testsuite before and after the patchset and no regressions were noticed. I'll also add this to the bug. --chris
On Wed, Jul 29, 2015 at 11:45:03PM -0500, Chris J Arges wrote: > BugLink: http://bugs.launchpad.net/bugs/1479468 > > The following changes since commit b146040071a57e0c91df9eeee7b592cd9216e3ff: > > UBUNTU: Ubuntu-3.19.0-26.27 (2015-07-28 18:25:05 +0100) > > are available in the git repository at: > > git://kernel.ubuntu.com/arges/ubuntu-vivid.git lp1479468 > > for you to fetch changes up to 69403dc50ba22c9057e2fac354d60bfadefeacec: > > fix a braino in ovl_d_select_inode() (2015-07-29 23:42:20 -0500) > > ---------------------------------------------------------------- > Al Viro (1): > fix a braino in ovl_d_select_inode() > > David Howells (3): > VFS: Add a fallthrough flag for marking virtual dentries > overlay: Call ovl_drop_write() earlier in ovl_dentry_open() > overlayfs: Make f_path always point to the overlay and f_inode to the underlay > > fs/dcache.c | 24 ++++++++++++++++++++++-- > fs/internal.h | 1 + > fs/open.c | 49 +++++++++++++++++++++++++----------------------- > fs/overlayfs/inode.c | 25 ++++++++++-------------- > fs/overlayfs/overlayfs.h | 1 + > fs/overlayfs/super.c | 1 + > include/linux/dcache.h | 11 +++++++++++ > include/linux/fs.h | 2 -- > 8 files changed, 72 insertions(+), 42 deletions(-) Ugg, well none of these in isolation look too bad. What sort of testing has been done with these applied? I am slightly concerned about the changes to vfs_open() as they obviously interact with most everything. If you can assure me this will be battered to death with a lot of filesystem tests I guess it is ok. Acked-by: Andy Whitcroft <apw@canonical.com> -apw
Applied to the master-next branch of Vivid.