[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <1480587303-25088-1-git-send-email-mszeredi@redhat.com>
Date: Thu, 1 Dec 2016 11:14:55 +0100
From: Miklos Szeredi <mszeredi@...hat.com>
To: Al Viro <viro@...iv.linux.org.uk>
Cc: linux-unionfs@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-fsdevel@...r.kernel.org
Subject: [PATCH 0/8] overlayfs: fix ro/rw fd data inconsistecies
Hi Al,
I'd like to hear your opinion on this patchset before queuing up for 4.10.
Here's the problem:
A file is opened for read-only, opened read-write (resulting in a copy up)
and modified. The data read back from the the read-only fd will be stale
in this case (the read-only file descriptor still refers to the lower,
unmodified file).
This patchset fixes issues related to this corner case.
The VFS impact is minimal and performance in the non-corner cases shouldn't
suffer.
Git tree is here:
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs.git #overlayfs-rorw
Pathset is against #overlayfs-next.
Thanks,
Miklos
---
Miklos Szeredi (8):
vfs: allow overlayfs to intercept file ops
vfs: export filp_clone_open()
mm: ovl: copy-up on MAP_SHARED
ovl: add infrastructure for intercepting file ops
ovl: intercept read_iter
ovl: intercept mmap
ovl: intercept fsync
Revert "ovl: Warn on copy up if a process has a R/O fd open to the
lower file"
fs/internal.h | 1 -
fs/open.c | 2 +-
fs/overlayfs/copy_up.c | 34 -------
fs/overlayfs/inode.c | 227 +++++++++++++++++++++++++++++++++++++++++++++++
fs/overlayfs/overlayfs.h | 2 +
fs/overlayfs/super.c | 1 +
include/linux/fs.h | 1 +
mm/util.c | 22 +++++
8 files changed, 254 insertions(+), 36 deletions(-)
--
2.5.5
Powered by blists - more mailing lists