linux-stable.git/fs, branch v4.6.5

ovl: warn instead of error if d_type is not supported

2016-07-27T15:42:25+00:00

commit e7c0b5991dd1be7b6f6dc2b54a15a0f47b64b007 upstream.

overlay needs underlying fs to support d_type. Recently I put in a
patch in to detect this condition and started failing mount if
underlying fs did not support d_type.

But this breaks existing configurations over kernel upgrade. Those who
are running docker (partially broken configuration) with xfs not
supporting d_type, are surprised that after kernel upgrade docker does
not run anymore.

https://github.com/docker/docker/issues/22937#issuecomment-229881315

So instead of erroring out, detect broken configuration and warn
about it. This should allow existing docker setups to continue
working after kernel upgrade.

Signed-off-by: Vivek Goyal 
Signed-off-by: Miklos Szeredi 
Fixes: 45aebeaf4f67 ("ovl: Ensure upper filesystem supports d_type")
Signed-off-by: Greg Kroah-Hartman

ovl: Do d_type check only if work dir creation was successful

2016-07-27T15:42:25+00:00

commit 21765194cecf2e4514ad75244df459f188140a0f upstream.

d_type check requires successful creation of workdir as iterates
through work dir and expects work dir to be present in it. If that's
not the case, this check will always return d_type not supported even
if underlying filesystem might be supporting it.

So don't do this check if work dir creation failed in previous step.

Signed-off-by: Vivek Goyal 
Signed-off-by: Miklos Szeredi 
Signed-off-by: Greg Kroah-Hartman

File names with trailing period or space need special case conversion

2016-07-27T15:42:24+00:00

commit 45e8a2583d97ca758a55c608f78c4cef562644d1 upstream.

POSIX allows files with trailing spaces or a trailing period but
SMB3 does not, so convert these using the normal Services For Mac
mapping as we do for other reserved characters such as
	: < > | ? *
This is similar to what Macs do for the same problem over SMB3.

Signed-off-by: Steve French 
Acked-by: Pavel Shilovsky 
Signed-off-by: Greg Kroah-Hartman

cifs: dynamic allocation of ntlmssp blob

2016-07-27T15:42:24+00:00

commit b8da344b74c822e966c6d19d6b2321efe82c5d97 upstream.

In sess_auth_rawntlmssp_authenticate(), the ntlmssp blob is allocated
statically and its size is an "empirical" 5*sizeof(struct
_AUTHENTICATE_MESSAGE) (320B on x86_64). I don't know where this value
comes from or if it was ever appropriate, but it is currently
insufficient: the user and domain name in UTF16 could take 1kB by
themselves. Because of that, build_ntlmssp_auth_blob() might corrupt
memory (out-of-bounds write). The size of ntlmssp_blob in
SMB2_sess_setup() is too small too (sizeof(struct _NEGOTIATE_MESSAGE)
+ 500).

This patch allocates the blob dynamically in
build_ntlmssp_auth_blob().

Signed-off-by: Jerome Marchand 
Signed-off-by: Steve French 
Signed-off-by: Greg Kroah-Hartman

Fix reconnect to not defer smb3 session reconnect long after socket reconnect

2016-07-27T15:42:24+00:00

commit 4fcd1813e6404dd4420c7d12fb483f9320f0bf93 upstream.

Azure server blocks clients that open a socket and don't do anything on it.
In our reconnect scenarios, we can reconnect the tcp session and
detect the socket is available but we defer the negprot and SMB3 session
setup and tree connect reconnection until the next i/o is requested, but
this looks suspicous to some servers who expect SMB3 negprog and session
setup soon after a socket is created.

In the echo thread, reconnect SMB3 sessions and tree connections
that are disconnected.  A later patch will replay persistent (and
resilient) handle opens.

Signed-off-by: Steve French 
Acked-by: Pavel Shilovsky 
Signed-off-by: Greg Kroah-Hartman

ovl: verify upper dentry in ovl_remove_and_whiteout()

2016-07-27T15:42:24+00:00

commit cfc9fde0b07c3b44b570057c5f93dda59dca1c94 upstream.

The upper dentry may become stale before we call ovl_lock_rename_workdir.
For example, someone could (mistakenly or maliciously) manually unlink(2)
it directly from upperdir.

To ensure it is not stale, let's lookup it after ovl_lock_rename_workdir
and and check if it matches the upper dentry.

Essentially, it is the same problem and similar solution as in
commit 11f3710417d0 ("ovl: verify upper dentry before unlink and rename").

Signed-off-by: Maxim Patlasov 
Signed-off-by: Miklos Szeredi 
Signed-off-by: Greg Kroah-Hartman

ovl: handle ATTR_KILL*

2016-07-27T15:42:24+00:00

commit b99c2d913810e56682a538c9f2394d76fca808f8 upstream.

Before 4bacc9c9234c ("overlayfs: Make f_path...") file->f_path pointed to
the underlying file, hence suid/sgid removal on write worked fine.

After that patch file->f_path pointed to the overlay file, and the file
mode bits weren't copied to overlay_inode->i_mode.  So the suid/sgid
removal simply stopped working.

The fix is to copy the mode bits, but then ovl_setattr() needs to clear
ATTR_MODE to avoid the BUG() in notify_change().  So do this first, then in
the next patch copy the mode.

Reported-by: Eryu Guan 
Signed-off-by: Miklos Szeredi 
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Signed-off-by: Greg Kroah-Hartman

ovl: Copy up underlying inode's ->i_mode to overlay inode

2016-07-27T15:42:23+00:00

commit 07a2daab49c549a37b5b744cbebb6e3f445f12bc upstream.

Right now when a new overlay inode is created, we initialize overlay
inode's ->i_mode from underlying inode ->i_mode but we retain only
file type bits (S_IFMT) and discard permission bits.

This patch changes it and retains permission bits too. This should allow
overlay to do permission checks on overlay inode itself in task context.

[SzM] It also fixes clearing suid/sgid bits on write.

Signed-off-by: Vivek Goyal 
Reported-by: Eryu Guan 
Signed-off-by: Miklos Szeredi 
Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
Signed-off-by: Greg Kroah-Hartman

ovl: get_write_access() in truncate

2016-07-27T15:42:23+00:00

commit 03bea60409328de54e4ff7ec41672e12a9cb0908 upstream.

When truncating a file we should check write access on the underlying
inode.  And we should do so on the lower file as well (before copy-up) for
consistency.

Original patch and test case by Aihua Zhang.

 - - >o >o - - test.c - - >o >o - -
#include 
#include 
#include 

int main(int argc, char *argv[])
{
	int ret;

	ret = truncate(argv[0], 4096);
	if (ret != -1) {
		fprintf(stderr, "truncate(argv[0]) should have failed\n");
		return 1;
	}
	if (errno != ETXTBSY) {
		perror("truncate(argv[0])");
		return 1;
	}

	return 0;
}
 - - >o >o - - >o >o - - >o >o - -

Reported-by: Aihua Zhang 
Signed-off-by: Miklos Szeredi 
Signed-off-by: Greg Kroah-Hartman

ovl: fix dentry leak for default_permissions

2016-07-27T15:42:23+00:00

commit a4859d75944a726533ab86d24bb5ffd1b2b7d6cc upstream.

When using the 'default_permissions' mount option, ovl_permission() on
non-directories was missing a dput(alias), resulting in "BUG Dentry still
in use".

Signed-off-by: Miklos Szeredi 
Fixes: 8d3095f4ad47 ("ovl: default permissions")
Signed-off-by: Greg Kroah-Hartman