kernel_optimize_test/fs
Eryu Guan 9ecd10b7a0 direct-io: fix direct write stale data exposure from concurrent buffered read
Currently direct writes inside i_size on a DIO_SKIP_HOLES filesystem are
not allowed to allocate blocks(get_more_blocks() sets 'create' to 0
before calling get_block() callback), if it's a sparse file, direct
writes fall back to buffered writes to avoid stale data exposure from
concurrent buffered read.  But there're two cases that can result in
stale data exposure are not correctly detected.

1. The detection for "writing inside i_size" is not sufficient,
   writes can be treated as "extending writes" wrongly.  For example,
   direct write 1FSB (file system block) to a 1FSB sparse file on
   ext2/3/4, starting from offset 0, in this case it's writing inside
   i_size, but 'create' is non-zero, because 'block_in_file' and
   '(i_size_read(inode) >> blkbits' are both zero.

2. Direct writes starting from or beyong i_size (not inside i_size)
   also could trigger block allocation and expose stale data.  For
   example, consider a sparse file with i_size of 2k, and a write to
   offset 2k or 3k into the file, with a filesystem block size of 4k.
   (Thanks to Jeff Moyer for pointing this case out in his review.)

The first problem can be demostrated by running ltp-aiodio test ADSP045
many times.  When testing on extN filesystems, I see test failures
occasionally, buffered read could read non-zero (stale) data.

ADSP045: dio_sparse -a 4k -w 4k -s 2k -n 1

dio_sparse    0  TINFO  :  Dirtying free blocks
dio_sparse    0  TINFO  :  Starting I/O tests
non zero buffer at buf[0] => 0xffffffaa,ffffffaa,ffffffaa,ffffffaa
non-zero read at offset 0
dio_sparse    0  TINFO  :  Killing childrens(s)
dio_sparse    1  TFAIL  :  dio_sparse.c:191: 1 children(s) exited abnormally

The second problem can also be reproduced easily by a hacked dio_sparse
program, which accepts an option to specify the write offset.

What we should really do is to disable block allocation for writes that
could result in filling holes inside i_size.

Link: http://lkml.kernel.org/r/1463156728-13357-1-git-send-email-guaneryu@gmail.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-05-27 14:49:37 -07:00
..
9p Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
adfs
affs Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
afs Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2016-05-17 16:26:30 -07:00
autofs4 dcache_{readdir,dir_lseek}() users: switch to ->iterate_shared 2016-05-02 19:49:32 -04:00
befs fs/befs/io.c:befs_bread(): remove unneeded initialization to NULL 2016-05-23 17:04:14 -07:00
bfs more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
btrfs Merge branch 'for-linus-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs 2016-05-21 10:49:22 -07:00
cachefiles mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
ceph Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client 2016-05-26 14:10:32 -07:00
cifs CIFS: Remove some obsolete comments 2016-05-19 21:56:34 -05:00
coda introduce a parallel variant of ->iterate() 2016-05-02 19:49:29 -04:00
configfs configfs_readdir(): make safe under shared lock 2016-05-09 11:41:13 -04:00
cramfs more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
crypto fscrypto/f2fs: allow fs-specific key prefix for fs encryption 2016-05-07 10:32:33 -07:00
debugfs Merge 4.6-rc4 into driver-core-next 2016-04-19 04:28:28 +09:00
devpts devpts: more pty driver interface cleanups 2016-04-26 15:47:32 -07:00
dlm mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
ecryptfs Merge branch 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-18 11:51:59 -07:00
efivarfs fs/efivarfs/inode.c: use generic UUID library 2016-05-20 17:58:30 -07:00
efs fs/efs/super.c: fix return value 2016-05-20 17:58:30 -07:00
exofs Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial 2016-05-17 17:05:30 -07:00
exportfs introduce a parallel variant of ->iterate() 2016-05-02 19:49:29 -04:00
ext2 DAX error handling for 4.7 2016-05-26 19:34:26 -07:00
ext4 DAX error handling for 4.7 2016-05-26 19:34:26 -07:00
f2fs Enhancement 2016-05-21 18:25:28 -07:00
fat Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
freevxfs more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
fscache mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
fuse Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
gfs2 We've got nine patches this time: 2016-05-20 15:11:26 -07:00
hfs Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
hfsplus Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
hostfs hostfs: switch to ->iterate_shared() 2016-05-12 19:49:30 -04:00
hpfs hpfs: switch to ->iterate_shared() 2016-05-12 19:47:13 -04:00
hugetlbfs mm, fs: remove remaining PAGE_CACHE_* and page_cache_{get,release} usage 2016-04-04 10:41:08 -07:00
isofs Merge branch 'ovl-fixes' into for-linus 2016-05-11 00:00:29 -04:00
jbd2 Fix a number of bugs, most notably a potential stale data exposure 2016-05-24 12:55:26 -07:00
jffs2 more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
jfs Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-18 10:08:45 -07:00
kernfs driver core update for 4.7-rc1 2016-05-20 21:26:15 -07:00
lockd
logfs logfs: no need to lock directory in lseek 2016-05-09 11:42:19 -04:00
minix simple local filesystems: switch to ->iterate_shared() 2016-05-02 19:49:32 -04:00
ncpfs mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
nfs NFS client updates for Linux 4.7 2016-05-26 10:33:33 -07:00
nfs_common
nfsd A very quiet cycle for nfsd, mainly just an RDMA update from Chuck Lever. 2016-05-24 14:39:20 -07:00
nilfs2 nilfs2: fix block comments 2016-05-23 17:04:14 -07:00
nls
notify fsnotify: avoid spurious EMFILE errors from inotify_init() 2016-05-19 19:12:14 -07:00
ntfs fs: simplify the generic_write_sync prototype 2016-05-01 19:58:39 -04:00
ocfs2 ocfs2: bump up o2cb network protocol version 2016-05-27 14:49:37 -07:00
omfs more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
openpromfs more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
orangefs orangefs: don't open-code inode_lock/inode_unlock 2016-05-02 19:47:23 -04:00
overlayfs Merge branch 'ovl-fixes' into for-linus 2016-05-17 02:17:59 -04:00
proc mm, proc: make clear_refs killable 2016-05-23 17:04:14 -07:00
pstore mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
qnx4 more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
qnx6 more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
quota fs/quota: use nla_put_u64_64bit() 2016-04-26 12:00:48 -04:00
ramfs tmpfs/ramfs: fix VM_MAYSHARE mappings for NOMMU 2016-05-20 17:58:30 -07:00
reiserfs lib/uuid.c: move generate_random_uuid() to uuid.c 2016-05-20 17:58:30 -07:00
romfs romfs, squashfs: switch to ->iterate_shared() 2016-05-09 11:41:15 -04:00
squashfs romfs, squashfs: switch to ->iterate_shared() 2016-05-09 11:41:15 -04:00
sysfs
sysv simple local filesystems: switch to ->iterate_shared() 2016-05-02 19:49:32 -04:00
tracefs
ubifs lib/uuid.c: move generate_random_uuid() to uuid.c 2016-05-20 17:58:30 -07:00
udf Merge branch 'work.preadv2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 15:05:23 -07:00
ufs simple local filesystems: switch to ->iterate_shared() 2016-05-02 19:49:32 -04:00
xfs DAX error handling for 4.7 2016-05-26 19:34:26 -07:00
aio.c aio: make aio_setup_ring killable 2016-05-23 17:04:14 -07:00
anon_inodes.c
attr.c
bad_inode.c ->getxattr(): pass dentry and inode as separate arguments 2016-04-11 00:48:00 -04:00
binfmt_aout.c mm, aout: handle vm_brk failures 2016-05-23 17:04:14 -07:00
binfmt_elf_fdpic.c Merge branch 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-18 11:51:59 -07:00
binfmt_elf.c mm, elf: handle vm_brk error 2016-05-23 17:04:14 -07:00
binfmt_em86.c
binfmt_flat.c
binfmt_misc.c
binfmt_script.c
block_dev.c DAX error handling for 4.7 2016-05-26 19:34:26 -07:00
buffer.c mm, page_alloc: avoid looking up the first zone in a zonelist twice 2016-05-19 19:12:14 -07:00
char_dev.c chrdev: emit a warning when we go below dynamic major range 2016-03-29 10:11:44 -07:00
compat_binfmt_elf.c
compat_ioctl.c
compat.c Fix a number of bugs, most notably a potential stale data exposure 2016-05-24 12:55:26 -07:00
coredump.c coredump: make coredump_wait wait for mmap_sem for write killable 2016-05-23 17:04:14 -07:00
dax.c Filesystem DAX locking for 4.7 2016-05-26 20:00:28 -07:00
dcache.c Merge branch 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-18 11:51:59 -07:00
dcookies.c
direct-io.c direct-io: fix direct write stale data exposure from concurrent buffered read 2016-05-27 14:49:37 -07:00
drop_caches.c
eventfd.c eventfd: document lockless access in eventfd_poll 2016-03-22 15:36:02 -07:00
eventpoll.c fs: poll/select/recvmmsg: use timespec64 for timeout events 2016-05-19 19:12:14 -07:00
exec.c exec: make exec path waiting for mmap_sem killable 2016-05-23 17:04:14 -07:00
fcntl.c
fhandle.c fs/coredump: prevent fsuid=0 dumps into user-controlled directories 2016-03-22 15:36:02 -07:00
file_table.c
file.c give readdir(2)/getdents(2)/etc. uniform exclusion with lseek() 2016-05-02 19:49:28 -04:00
filesystems.c
fs_pin.c
fs_struct.c
fs-writeback.c mm,writeback: don't use memory reserves for wb_start_writeback 2016-05-20 17:58:30 -07:00
inode.c parallel lookups: actual switch to rwsem 2016-05-02 19:49:28 -04:00
internal.h
ioctl.c
Kconfig dax: Make huge page handling depend of CONFIG_BROKEN 2016-05-19 15:13:17 -06:00
Kconfig.binfmt ELF/MIPS build fix 2016-05-23 17:04:14 -07:00
libfs.c more trivial ->iterate_shared conversions 2016-05-09 11:41:14 -04:00
locks.c
Makefile Merge tag 'ofs-pull-tag-1' of git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux 2016-03-26 12:59:04 -07:00
mbcache.c
mount.h
mpage.c mm, fs: remove remaining PAGE_CACHE_* and page_cache_{get,release} usage 2016-04-04 10:41:08 -07:00
namei.c xfs: update for 4.7-rc1 2016-05-26 10:13:40 -07:00
namespace.c
no-block.c
nsfs.c
open.c Merge branch 'work.const-path' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-17 14:41:03 -07:00
pipe.c mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
pnode.c propogate_mnt: Handle the first propogated copy being a slave 2016-05-05 09:54:45 -05:00
pnode.h
posix_acl.c xattr_handler: pass dentry and inode as separate arguments of ->get() 2016-04-10 20:48:24 -04:00
proc_namespace.c
read_write.c Merge branch 'work.iov_iter' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2016-05-18 11:46:23 -07:00
readdir.c Fix a number of bugs, most notably a potential stale data exposure 2016-05-24 12:55:26 -07:00
select.c fs: poll/select/recvmmsg: use timespec64 for timeout events 2016-05-19 19:12:14 -07:00
seq_file.c Make file credentials available to the seqfile interfaces 2016-04-14 12:56:09 -07:00
signalfd.c
splice.c Merge branch 'ovl-fixes' into for-linus 2016-05-11 00:00:29 -04:00
stack.c
stat.c
statfs.c
super.c Merge branch 'master' into for-next 2016-04-18 11:18:55 +02:00
sync.c mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros 2016-04-04 10:41:08 -07:00
timerfd.c
userfaultfd.c userfaultfd: don't pin the user memory in userfaultfd_file_create() 2016-05-20 17:58:30 -07:00
utimes.c
xattr.c make xattr_resolve_handlers() safe to use with NULL ->s_xattr 2016-05-25 17:34:41 -04:00