Correct kernel-doc comments pointed out by the
automated kernel test robot.
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

checkpatch complains about source files with filenames (e.g. in
these cases just below the SPDX header in comments at the top of
various files in fs/cifs). It also is helpful to change this now
so will be less confusing when the parent directory is renamed
e.g. from fs/cifs to fs/smb_client (or fs/smbfs)
Reviewed-by: NRonnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

for SMB1.
This removes the dependency to DES.
Signed-off-by: NRonnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

When remouting a DFS share, force a new DFS referral of the path and
if the currently cached targets do not match any of the new targets or
there was no cached targets, then mark it for reconnect.

For example:

    $ mount //dom/dfs/link /mnt -o username=foo,password=bar
    $ ls /mnt
    oldfile.txt

    change target share of 'link' in server settings

    $ mount /mnt -o remount,username=foo,password=bar
    $ ls /mnt
    newfile.txt
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Make sure that we do not share tcp sessions of dfs mounts when
mounting regular shares that connect to same server.  DFS connections
rely on a single instance of tcp in order to do failover properly in
cifs_reconnect().
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

We have a few ref counters srv_count, ses_count and
tc_count which we use for ref counting. Added a WARN_ON
during the decrement of each of these counters to make
sure that they don't go below their minimum values.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Although it is unlikely to be have ended up with a null
session pointer calling cifs_try_adding_channels in cifs_mount.
Coverity correctly notes that we are already checking for
it earlier (when we return from do_dfs_failover), so at
a minimum to clarify the code we should make sure we also
check for it when we exit the loop so we don't end up calling
cifs_try_adding_channels or mount_setup_tlink with a null
ses pointer.

Addresses-Coverity: 1505608 ("Derefernce after null check")
Reviewed-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

When there is no cached DFS referral of tcon->dfs_path, then reconnect
to same share.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: <stable@vger.kernel.org>
Signed-off-by: NSteve French <stfrench@microsoft.com>

We recently fixed DNS resolution of the server hostname during reconnect.
However, server IP address may change, even when the old one continues
to server (although sub-optimally).

We should schedule the next DNS resolution based on the TTL of
the DNS record used for the last resolution. This way, we resolve the
server hostname again when a DNS record expires.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Reviewed-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: <stable@vger.kernel.org> # v5.11+
Signed-off-by: NSteve French <stfrench@microsoft.com>

There were three places where we were not taking the spinlock
around updates to server->tcpStatus when it was being modified.
To be consistent (also removes Coverity warning) and to remove
possibility of race best to lock all places where it is updated.
Two of the three were in initialization of the field and can't
race - but added lock around the other.

Addresses-Coverity: 1399512 ("Data race condition")
Reviewed-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

In the other places where we update ses->status we protect the
updates via GlobalMid_Lock. So to be consistent add the same
locking around it in cifs_put_smb_ses where it was missing.

Addresses-Coverity: 1268904 ("Data race condition")
Signed-off-by: NSteve French <stfrench@microsoft.com>

We weren't checking if tcon is null before setting dfs path,
although we check for null tcon in an earlier assignment statement.

Addresses-Coverity: 1476411 ("Dereference after null check")
Signed-off-by: NSteve French <stfrench@microsoft.com>

Add SPDX license identifier and replace license boilerplate.
Corrects various checkpatch errors with the older format for
noting the LGPL license.
Signed-off-by: NSteve French <stfrench@microsoft.com>

It isn't enough to have unshared tcons because multiple DFS mounts can
connect to same target server and failover to different servers, so we
can't use a single tcp server for such cases.

For the simplest solution, use nosharesock option to achieve that.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Convert all dfs paths to dfs cache's local codepage (@cache_cp) and
avoid mixing them with different charsets.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

At every mount, keep all sessions alive that were used for chasing the
DFS referrals as long as the dfs mounts are active.

Use those sessions in DFS cache to refresh all active tcons as well as
cached entries.  They will be managed by a list of mount_group
structures that will be indexed by a randomly generated uuid at mount
time, so we can put all the sessions related to specific dfs mounts
and avoid leaking them.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

On session close, the IPC is closed and the server must release all
tcons of the session.  It doesn't matter if we send a ipc close or
not.

Besides, it will make the server to not close durable and resilient
files on session close, as specified in MS-SMB2 3.3.5.6 Receiving an
SMB2 LOGOFF Request.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

The commit 315db9a0 ("cifs: fix leak in cifs_smb3_do_mount() ctx")
revealed an existing bug when mounting shares that contain a prefix
path or DFS links.

cifs_setup_volume_info() requires the @devname to contain the full
path (UNC + prefix) to update the fs context with the new UNC and
prepath values, however we were passing only the UNC
path (old_ctx->UNC) in @device thus discarding any prefix paths.

Instead of concatenating both old_ctx->{UNC,prepath} and pass it in
@devname, just keep the dup'ed values of UNC and prepath in
cifs_sb->ctx after calling smb3_fs_context_dup(), and fix
smb3_parse_devname() to correctly parse and not leak the new UNC and
prefix paths.

Cc: <stable@vger.kernel.org> # v5.11+
Fixes: 315db9a0 ("cifs: fix leak in cifs_smb3_do_mount() ctx")
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Acked-by: NDavid Disseldorp <ddiss@suse.de>
Signed-off-by: NSteve French <stfrench@microsoft.com>

When the tcp connection is not ready to send requests,
we keep retrying echo with an interval of zero.

This seems unnecessary, and this fix changes the interval
between echoes to what is specified as echo_interval.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

We can detect server unresponsiveness only if echoes are enabled.
Echoes can be disabled under two scenarios:
1. The connection is low on credits, so we've disabled echoes/oplocks.
2. The connection has not seen any request till now (other than
negotiate/sess-setup), which is when we enable these two, based on
the credits available.

So this fix will check for dead connection, only when echo is enabled.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
CC: <stable@vger.kernel.org> # v5.8+
Signed-off-by: NSteve French <stfrench@microsoft.com>

strndup(s, strlen(s)) is a highly unidiomatic way to spell strdup(s);
it's *NOT* safer in any way, since strlen() is just as sensitive to
NUL-termination as strdup() is.

strndup() is for situations when you need a copy of a known-sized
substring, not a magic security juju to drive the bad spirits away.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NSteve French <stfrench@microsoft.com>

This commit doesn't change the logic of SWN.

Add dummy implementation of SWN functions when SWN is disabled instead
of using ifdef sections.

The dummy functions get optimized out, this leads to clearer code and
compile time type-checking regardless of config options with no
runtime penalty.

Leave the simple ifdefs section as-is.

A single bitfield (bool foo:1) on its own will use up one int. Move
tcon->use_witness out of ifdefs with the other tcon bitfields.
Signed-off-by: NAurelien Aptel <aaptel@suse.com>
Reviewed-by: NSamuel Cabrero <scabrero@suse.de>
Signed-off-by: NSteve French <stfrench@microsoft.com>

On cifs_reconnect, make sure that DNS resolution happens again.
It could be the cause of connection to go dead in the first place.

This also contains the fix for a build issue identified by Intel bot.
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Reviewed-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NPavel Shilovsky <pshilov@microsoft.com>
CC: <stable@vger.kernel.org> # 5.11+
Signed-off-by: NSteve French <stfrench@microsoft.com>

The MIDs are mostly printed as decimal, so let's make it consistent.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Reviewed-by: NRonnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

With multichannel, operations like the queries
from "ls -lR" can cause all credits to be used and
errors to be returned since max_credits was not
being set correctly on the secondary channels and
thus the client was requesting 0 credits incorrectly
in some cases (which can lead to not having
enough credits to perform any operation on that
channel).
Signed-off-by: NAurelien Aptel <aaptel@suse.com>
CC: <stable@vger.kernel.org> # v5.8+
Reviewed-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

The iterator, ITER_DISCARD, that can only be used in READ mode and
just discards any data copied to it, was added to allow a network
filesystem to discard any unwanted data sent by a server.
Convert cifs_discard_from_socket() to use this.
Signed-off-by: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Some servers seem to mistakenly report different values for
capabilities and share flags, so we can't always rely on those values
to decide whether the resolved target can handle any new DFS
referrals.

Add a new helper is_referral_server() to check if all resolved targets
can handle new DFS referrals by directly looking at the
GET_DFS_REFERRAL.ReferralHeaderFlags value as specified in MS-DFSC
2.2.4 RESP_GET_DFS_REFERRAL in addition to is_tcon_dfs().
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: stable@vger.kernel.org # 5.11
Signed-off-by: NSteve French <stfrench@microsoft.com>

Handle the case where a resolved target share is like
//server/users/dir, and the user "foo" has no read permission to
access the parent folder "users" but has access to the final path
component "dir".

is_path_remote() already implements that, so call it directly.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: stable@vger.kernel.org # 5.11
Signed-off-by: NSteve French <stfrench@microsoft.com>

In do_dfs_failover(), the mount_get_conns() function requires the full
fs context in order to get new connection to server, so clone the
original context and change it accordingly when retrying the DFS
targets in the referral.

If failover was successful, then update original context with the new
UNC, prefix path and ip address.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: stable@vger.kernel.org # 5.11
Signed-off-by: NSteve French <stfrench@microsoft.com>

Skip DFS resolving when mounting with 'nodfs' even if
CONFIG_CIFS_DFS_UPCALL is enabled.
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Cc: stable@vger.kernel.org # 5.11
Reviewed-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

The new optional mount parameter "acregmax" allows a different
timeout for file metadata ("acdirmax" now allows controlling timeout
for directory metadata).  Setting "actimeo" still works as before,
and changes timeout for both files and directories, but
specifying "acregmax" or "acdirmax" allows overriding the
default more granularly which can be a big performance benefit
on some workloads. "acregmax" is already used by NFS as a mount
parameter (albeit with a larger default and thus looser caching).
Suggested-by: NTom Talpey <tom@talpey.com>
Reviewed-By: NTom Talpey <tom@talpey.com>
Reviewed-by: NRonnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

nfs and cifs on Linux currently have a mount parameter "actimeo" to control
metadata (attribute) caching but cifs does not have additional mount
parameters to allow distinguishing between caching directory metadata
(e.g. needed to revalidate paths) and that for files.

Add new mount parameter "acdirmax" to allow caching metadata for
directories more loosely than file data.  NFS adjusts metadata
caching from acdirmin to acdirmax (and another two mount parms
for files) but to reduce complexity, it is safer to just introduce
the one mount parm to allow caching directories longer. The
defaults for acdirmax and actimeo (for cifs.ko) are conservative,
1 second (NFS defaults acdirmax to 60 seconds). For many workloads,
setting acdirmax to a higher value is safe and will improve
performance.  This patch leaves unchanged the default values
for caching metadata for files and directories but gives the
user more flexibility in adjusting them safely for their workload
via the new mount parm.
Signed-off-by: NSteve French <stfrench@microsoft.com>
Reviewed-by: NRonnie Sahlberg <lsahlber@redhat.com>
Reviewed-By: NTom Talpey <tom@talpey.com>

When server returns error STATUS_NETWORK_NAME_DELETED, TCON
must be marked for reconnect. So, subsequent IO does the tree
connect again.
Signed-off-by: NRohith Surabattula <rohiths@microsoft.com>
Reviewed-by: NPavel Shilovsky <pshilov@microsoft.com>
Reviewed-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

ses->serverName is not the server name, but the string form
of the ip address of the server.  Change the name to ip_addr
to avoid confusion (and fix the array length to match
maximum length of ipv6 address).
Reviewed-by: NShyam Prasad N <sprasad@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Introduced a new field conn_id in TCP_Server_Info structure.
This is a non-persistent unique identifier maintained by the client
for a connection to a file server. For this, a global counter named
tcpSesNextId is maintained. On allocating a new TCP_Server_Info,
this counter is incremented and assigned.

Changed the dynamic tracepoints related to reconnects and
crediting to be more informative (with conn_id printed).
Debugging a crediting issue helped me understand the
important things to print here.

Always call dynamic tracepoints outside the scope of spinlocks.
To do this, copy out the credits and in_flight fields of the
server struct before dropping the lock.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Reviewed-by: NPavel Shilovsky <pshilov@microsoft.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Typo: exiting --> existing
Signed-off-by: NSteve French <stfrench@microsoft.com>

While debugging another issue today, Steve and I noticed that if a
subdir for a file share is already mounted on the client, any new
mount of any other subdir (or the file share root) of the same share
results in sharing the cifs superblock, which e.g. can result in
incorrect device name.

While setting prefix path for the root of a cifs_sb,
CIFS_MOUNT_USE_PREFIX_PATH flag should also be set.
Without it, prepath is not even considered in some places,
and output of "mount" and various /proc/<>/*mount* related
options can be missing part of the device name.
Signed-off-by: NShyam Prasad N <sprasad@microsoft.com>
Reviewed-by: NAurelien Aptel <aaptel@suse.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>

This fixes a regression following dfs links that was introduced in the
patch series for the new mount api.
Signed-off-by: NRonnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

The new mount API requires additional changes to how DFS
is handled. Additional testing of DFS uncovered problems
with domain based DFS referrals (a follow on patch addresses
DFS links) which this patch addresses.
Signed-off-by: NRonnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: NPaulo Alcantara (SUSE) <pc@cjr.nz>
Signed-off-by: NSteve French <stfrench@microsoft.com>

Fix the follow warnings:

./fs/cifs/connect.c: WARNING: Comparison of 0/1 to bool variable
Reported-by: NAbaci Robot <abaci@linux.alibaba.com>
Signed-off-by: NJiapeng Zhong <abaci-bugfix@linux.alibaba.com>
Signed-off-by: NSteve French <stfrench@microsoft.com>