Add a zip file with all patches in the July 24th 2023 backport of all relevant cifs.ko patches to 5.15

Add full backport list of commits

cifs: update internal module version number for cifs.ko

From 2.43 to 2.44

Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix mid leak during reconnection after timeout threshold

When the number of responses with status of STATUS_IO_TIMEOUT
exceeds a specified threshold (NUM_STATUS_IO_TIMEOUT), we reconnect
the connection. But we do not return the mid, or the credits
returned for the mid, or reduce the number of in-flight requests.

This bug could result in the server->in_flight count to go bad,
and also cause a leak in the mids.

This change moves the check to a few lines below where the
response is decrypted, even of the response is read from the
transform header. This way, the code for returning the mids
can be reused.

Also, the cifs_reconnect was reconnecting just the transport
connection before. In case of multi-channel, this may not be
what we want to do after several timeouts. Changed that to
reconnect the session and the tree too.

Also renamed NUM_STATUS_IO_TIMEOUT to a more appropriate name
MAX_STATUS_IO_TIMEOUT.

Fixes: 8e670f77c4a5 ("Handle STATUS_IO_TIMEOUT gracefully")
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix missed ses refcounting

Use new cifs_smb_ses_inc_refcount() helper to get an active reference
of @ses and @ses->dfs_root_ses (if set).  This will prevent
@ses->dfs_root_ses of being put in the next call to cifs_put_smb_ses()
and thus potentially causing an use-after-free bug.

Fixes: 8e3554150d6c ("cifs: fix sharing of DFS connections")
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: Fix -Wstringop-overflow issues

pSMB->hdr.Protocol is an array of size 4 bytes, hence when the compiler
analyzes this line of code

parm_data = ((char *) &pSMB->hdr.Protocol) + offset;

it legitimately complains about the fact that offset points outside the
bounds of the array. Notice that the compiler gives priority to the object
as an array, rather than merely the address of one more byte in a structure
to wich offset should be added (which seems to be the actual intention of
the original implementation).

Fix this by explicitly instructing the compiler to treat the code as a
sequence of bytes in struct smb_com_transaction2_spi_req, and not as an
array accessed through pointer notation.

Notice that ((char *)pSMB) + sizeof(pSMB->hdr.smb_buf_length) points to
the same address as ((char *) &pSMB->hdr.Protocol), therefore this results
in no differences in binary output.

Fixes the following -Wstringop-overflow warnings when built s390
architecture with defconfig (GCC 13):
  CC [M]  fs/smb/client/cifssmb.o
In function 'cifs_init_ace',
    inlined from 'posix_acl_to_cifs' at fs/smb/client/cifssmb.c:3046:3,
    inlined from 'cifs_do_set_acl' at fs/smb/client/cifssmb.c:3191:15:
fs/smb/client/cifssmb.c:2987:31: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=]
 2987 |         cifs_ace->cifs_e_perm = local_ace->e_perm;
      |         ~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~
In file included from fs/smb/client/cifssmb.c:27:
fs/smb/client/cifspdu.h: In function 'cifs_do_set_acl':
fs/smb/client/cifspdu.h:384:14: note: at offset [7, 11] into destination object 'Protocol' of size 4
  384 |         __u8 Protocol[4];
      |              ^~~~~~~~
In function 'cifs_init_ace',
    inlined from 'posix_acl_to_cifs' at fs/smb/client/cifssmb.c:3046:3,
    inlined from 'cifs_do_set_acl' at fs/smb/client/cifssmb.c:3191:15:
fs/smb/client/cifssmb.c:2988:30: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=]
 2988 |         cifs_ace->cifs_e_tag =  local_ace->e_tag;
      |         ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~
fs/smb/client/cifspdu.h: In function 'cifs_do_set_acl':
fs/smb/client/cifspdu.h:384:14: note: at offset [6, 10] into destination object 'Protocol' of size 4
  384 |         __u8 Protocol[4];
      |              ^~~~~~~~

This helps with the ongoing efforts to globally enable
-Wstringop-overflow.

Link: https://github.com/KSPP/linux/issues/310
Fixes: dc1af4c4b472 ("cifs: implement set acl method")
Cc: stable@vger.kernel.org
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: if deferred close is disabled then close files immediately

If defer close timeout value is set to 0, then there is no
need to include files in the deferred close list and utilize
the delayed worker for closing. Instead, we can close them
immediately.

Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
---
 fs/smb/client/file.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/smb/client/file.c b/fs/smb/client/file.c
index 879bc8e6555c..fc5acc95cd13 100644
--- a/fs/smb/client/file.c
+++ b/fs/smb/client/file.c
@@ -1080,8 +1080,8 @@ int cifs_close(struct inode *inode, struct file *file)
  cfile = file->private_data;
  file->private_data = NULL;
  dclose = kmalloc(sizeof(struct cifs_deferred_close), GFP_KERNEL);
- if ((cinode->oplock == CIFS_CACHE_RHW_FLG) &&
-     cinode->lease_granted &&
+ if ((cifs_sb->ctx->closetimeo && cinode->oplock == CIFS_CACHE_RHW_FLG)
+     && cinode->lease_granted &&
      !test_bit(CIFS_INO_CLOSE_ON_LOCK, &cinode->flags) &&
      dclose) {
  if (test_and_clear_bit(CIFS_INO_MODIFIED_ATTR, &cinode->flags)) {

smb: client: remove redundant pointer 'server'

The pointer 'server' is assigned but never read, the pointer is
redundant and can be removed. Cleans up clang scan build warning:

fs/smb/client/dfs.c:217:3: warning: Value stored to 'server' is
never read [deadcode.DeadStores]

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix session state transition to avoid use-after-free issue

We switch session state to SES_EXITING without cifs_tcp_ses_lock now,
it may lead to potential use-after-free issue.

Consider the following execution processes:

Thread 1:
__cifs_put_smb_ses()
    spin_lock(&cifs_tcp_ses_lock)
    if (--ses->ses_count > 0)
        spin_unlock(&cifs_tcp_ses_lock)
        return
    spin_unlock(&cifs_tcp_ses_lock)
        ---> **GAP**
    spin_lock(&ses->ses_lock)
    if (ses->ses_status == SES_GOOD)
        ses->ses_status = SES_EXITING
    spin_unlock(&ses->ses_lock)

Thread 2:
cifs_find_smb_ses()
    spin_lock(&cifs_tcp_ses_lock)
    list_for_each_entry(ses, ...)
        spin_lock(&ses->ses_lock)
        if (ses->ses_status == SES_EXITING)
            spin_unlock(&ses->ses_lock)
            continue
        ...
        spin_unlock(&ses->ses_lock)
    if (ret)
        cifs_smb_ses_inc_refcount(ret)
    spin_unlock(&cifs_tcp_ses_lock)

If thread 1 is preempted in the gap and thread 2 start executing, thread 2
will get the session, and soon thread 1 will switch the session state to
SES_EXITING and start releasing it, even though thread 1 had increased the
session's refcount and still uses it.

So switch session state under cifs_tcp_ses_lock to eliminate this gap.

Signed-off-by: Winston Wen <wentao@uniontech.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: new dynamic tracepoint to track ses not found errors

It is perfectly valid to not find session not found errors
when a reconnect of a session happens when requests for the
same session are happening in parallel.

We had these log messages as VFS logs. My last change dumped
these logs as FYI logs.

This change just creates a new dynamic tracepoint to capture
events of this type, just in case it is useful while
debugging issues in the future.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: log session id when a matching ses is not found

We do not log the session id in crypt_setup when a matching
session is not found. Printing the session id helps debugging
here. This change does just that.

This change also changes this log to FYI, since it is normal to
see then during a reconnect. Doing the same for a similar log
in case of signed connections.

The plan is to have a tracepoint for this event, so that we will
be able to see this event if need be. That will be done as
another change.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: improve DFS mount check

Some servers may return error codes from REQ_GET_DFS_REFERRAL requests
that are unexpected by the client, so to make it easier, assume
non-DFS mounts when the client can't get the initial DFS referral of
@ctx->UNC in dfs_mount_share().

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix shared DFS root mounts with different prefixes

When having two DFS root mounts that are connected to same namespace,
same mount options but different prefix paths, we can't really use the
shared @server->origin_fullpath when chasing DFS links in them.

Move the origin_fullpath field to cifs_tcon structure so when having
shared DFS root mounts with different prefix paths, and we need to
chase any DFS links, dfs_get_automount_devname() will pick up the
correct full path out of the @tcon that will be used for the new
mount.

Before patch

  mount.cifs //dom/dfs/dir /mnt/1 -o ...
  mount.cifs //dom/dfs /mnt/2 -o ...
  # shared server, ses, tcon
  # server: origin_fullpath=//dom/dfs/dir

  # @server->origin_fullpath + '/dir/link1'
  $ ls /mnt/2/dir/link1
  ls: cannot open directory '/mnt/2/dir/link1': No such file or directory

After patch

  mount.cifs //dom/dfs/dir /mnt/1 -o ...
  mount.cifs //dom/dfs /mnt/2 -o ...
  # shared server & ses
  # tcon_1: origin_fullpath=//dom/dfs/dir
  # tcon_2: origin_fullpath=//dom/dfs

  # @tcon_2->origin_fullpath + '/dir/link1'
  $ ls /mnt/2/dir/link1
  dir0  dir1  dir10  dir3  dir5  dir6  dir7  dir9  target2_file.txt  tsub

Fixes: 8e3554150d6c ("cifs: fix sharing of DFS connections")
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix parsing of source mount option

Handle trailing and leading separators when parsing UNC and prefix
paths in smb3_parse_devname().  Then, store the sanitised paths in
smb3_fs_context::source.

This fixes the following cases

$ mount //srv/share// /mnt/1 -o ...
$ cat /mnt/1/d0/f0
cat: /mnt/1/d0/f0: Invalid argument

The -EINVAL was returned because the client sent SMB2_CREATE "\\d0\f0"
rather than SMB2_CREATE "\d0\f0".

$ mount //srv//share /mnt/1 -o ...
mount: Invalid argument

The -EINVAL was returned correctly although the client only realised
it after sending a couple of bad requests rather than bailing out
earlier when parsing mount options.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix broken file attrs with nodfs mounts

*_get_inode_info() functions expect -EREMOTE when query path info
calls find a DFS link, regardless whether !CONFIG_CIFS_DFS_UPCALL or
'nodfs' mount option.  Otherwise, those files will miss the fake DFS
file attributes.

Before patch

  $ mount.cifs //srv/dfs /mnt/1 -o ...,nodfs
  $ ls -l /mnt/1
  ls: cannot access '/mnt/1/link': Operation not supported
  total 0
  -rwxr-xr-x 1 root root 0 Jul 26  2022 dfstest2_file1.txt
  drwxr-xr-x 2 root root 0 Aug  8  2022 dir1
  d????????? ? ?    ?    ?            ? link

After patch

  $ mount.cifs //srv/dfs /mnt/1 -o ...,nodfs
  $ ls -l /mnt/1
  total 0
  -rwxr-xr-x 1 root root 0 Jul 26  2022 dfstest2_file1.txt
  drwxr-xr-x 2 root root 0 Aug  8  2022 dir1
  drwx--x--x 2 root root 0 Jun 26 20:29 link

Fixes: c877ce47e137 ("cifs: reduce roundtrips on create/qinfo requests")
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print client_guid in DebugData

Having the ClientGUID info makes it easier to debug
issues related to a client on a server that serves a
number of clients.

This change prints the ClientGUID in DebugData.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Acked-by: Tom Talpey <tom@talpey.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix session state check in smb2_find_smb_ses

Chech the session state and skip it if it's exiting.

Signed-off-by: Winston Wen <wentao@uniontech.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix session state check in reconnect to avoid
 use-after-free issue

Don't collect exiting session in smb2_reconnect_server(), because it
will be released soon.

Note that the exiting session will stay in server->smb_ses_list until
it complete the cifs_free_ipc() and logoff() and then delete itself
from the list.

Signed-off-by: Winston Wen <wentao@uniontech.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: do all necessary checks for credits within or before locking

All the server credits and in-flight info is protected by req_lock.
Once the req_lock is held, and we've determined that we have enough
credits to continue, this lock cannot be dropped till we've made the
changes to credits and in-flight count.

However, we used to drop the lock in order to avoid deadlock with
the recent srv_lock. This could cause the checks already made to be
invalidated.

Fixed it by moving the server status check to before locking req_lock.

Fixes: d7d7a66aacd6 ("cifs: avoid use of global locks for high contention data")
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: prevent use-after-free by freeing the cfile later

In smb2_compound_op we have a possible use-after-free
which can cause hard to debug problems later on.

This was revealed during stress testing with KASAN enabled
kernel. Fixing it by moving the cfile free call to
a few lines below, after the usage.

Fixes: 76894f3e2f71 ("cifs: improve symlink handling for smb2+")
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix warning in generic_ip_connect()

This fixes the following warning reported by kernel test robot

  fs/smb/client/connect.c:2974 generic_ip_connect() error: we
  previously assumed 'socket' could be null (see line 2962)

Link: https://lore.kernel.org/all/202306170124.CtQqzf0I-lkp@intel.com/
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix warning in CIFSFindNext()

This fixes the following warning reported by kernel test robot

  fs/smb/client/cifssmb.c:4216 CIFSFindNext() warn: missing error
  code? 'rc'

Link: https://lore.kernel.org/all/202306170124.CtQqzf0I-lkp@intel.com/
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix warning in CIFSFindFirst()

This fixes the following warning reported by kernel test robot

  fs/smb/client/cifssmb.c:4089 CIFSFindFirst() warn: missing error
  code? 'rc'

Link: https://lore.kernel.org/all/202306170124.CtQqzf0I-lkp@intel.com/
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: do not reserve too many oplock credits

There were cases reported where servers will sometimes return more
credits than requested on oplock break responses, which can lead to
most of the credits being allocated for oplock breaks (instead of
for normal operations like read and write) if number of SMB3 requests
in flight always stays above 0 (the oplock and echo credits are
rebalanced when in flight requests goes down to zero).

If oplock credits gets unexpectedly large (e.g. three is more than it
would ever be expected to be) and in flight requests are greater than
zero, then rebalance the oplock credits and regular credits (go
back to reserving just one oplock credit).

Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print more detail when invalidate_inode_mapping fails

We had seen cases where cifs_invalidate_mapping was logging:
   "Could not invalidate inode ..."
if invalidate_inode_pages2 fails but this message does not show what
the rc is.  Update the logged message to also log the return code.

Suggested-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix warning in cifs_smb3_do_mount()

This fixes the following warning reported by kernel test robot

  fs/smb/client/cifsfs.c:982 cifs_smb3_do_mount() warn: possible
  memory leak of 'cifs_sb'

Link: https://lore.kernel.org/all/202306170124.CtQqzf0I-lkp@intel.com/
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: client: fix warning in cifs_match_super()

Fix potential dereference of ERR_PTR @tlink as reported by kernel test
robot

  fs/smb/client/connect.c:2775 cifs_match_super() error: 'tlink'
  dereferencing possible ERR_PTR()

Link: https://lore.kernel.org/all/202306170124.CtQqzf0I-lkp@intel.com/
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print nosharesock value while dumping mount options

We print most other mount options for a mount when dumping
the mount entries. But miss out the nosharesock value.
This change will print that in addition to the other options.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: Do not send lease break acknowledgment if all
 file handles have been closed

In case if all existing file handles are deferred handles and if all of
them gets closed due to handle lease break then we dont need to send
lease break acknowledgment to server, because last handle close will be
considered as lease break ack.
After closing deferred handels, we check for openfile list of inode,
if its empty then we skip sending lease break ack.

Fixes: 59a556aebc43 ("SMB3: drop reference to cfile before sending oplock break")
Reviewed-by: Tom Talpey <tom@talpey.com>
Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: add a warning when the in-flight count goes negative

We've seen the in-flight count go into negative with some
internal stress testing in Microsoft.

Adding a WARN when this happens, in hope of understanding
why this happens when it happens.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix lease break oops in xfstest generic/098

umount can race with lease break so need to check if
tcon->ses->server is still valid to send the lease
break response.

Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Fixes: 59a556aebc43 ("SMB3: drop reference to cfile before sending oplock break")
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix max_credits implementation

The current implementation of max_credits on the client does
not work because the CreditRequest logic for several commands
does not take max_credits into account.

Still, we can end up asking the server for more credits, depending
on the number of credits in flight. For this, we need to
limit the credits while parsing the responses too.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix sockaddr comparison in iface_cmp

iface_cmp used to simply do a memcmp of the two
provided struct sockaddrs. The comparison needs to do more
based on the address family. Similar logic was already
present in cifs_match_ipaddr. Doing something similar now.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb/client: print "Unknown" instead of bogus link speed value

The virtio driver for Linux guests will not set a link speed to its
paravirtualized NICs.  This will be seen as -1 in the ethernet layer, and
when some servers (e.g. samba) fetches it, it's converted to an unsigned
value (and multiplied by 1000 * 1000), so in client side we end up with:

   1)      Speed: 4294967295000000 bps

in DebugData.

This patch introduces a helper that returns a speed string (in Mbps or
Gbps) if interface speed is valid (>= SPEED_10 and <= SPEED_800000), or
"Unknown" otherwise.

The reason to not change the value in iface->speed is because we don't
know the real speed of the HW backing the server NIC, so let's keep
considering these as the fastest NICs available.

Also print "Capabilities: None" when the interface doesn't support any.

Signed-off-by: Enzo Matsumiya <ematsumiya@suse.de>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print all credit counters in DebugData

Output of /proc/fs/cifs/DebugData shows only the per-connection
counter for the number of credits of regular type. i.e. the
credits reserved for echo and oplocks are not displayed.

There have been situations recently where having this info
would have been useful. This change prints the credit counters
of all three types: regular, echo, oplocks.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix status checks in cifs_tree_connect

The ordering of status checks at the beginning of
cifs_tree_connect is wrong. As a result, a tcon
which is good may stay marked as needing reconnect
infinitely.

Fixes: 2f0e4f034220 ("cifs: check only tcon status on tcon related functions")
Cc: stable@vger.kernel.org # 6.3
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: remove obsolete comment

Because do_gettimeofday has been removed and replaced by ktime_get_real_ts64,
So just remove the comment as it's not needed now.

Signed-off-by: 鑫华 <jixianghua@xfusion.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: address unused variable warning

Fix trivial unused variable warning (when SMB1 support disabled)

"ioctl.c:324:17: warning: variable 'caps' set but not used [-Wunused-but-set-variable]"

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202305250056.oZhsJmdD-lkp@intel.com/
Signed-off-by: Steve French <stfrench@microsoft.com>

smb: delete an unnecessary statement

We don't need to set the list iterators to NULL before a
list_for_each_entry() loop because they are assigned inside the
macro.

Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Mukesh Ojha <quic_mojha@quicinc.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: missing null check in SMB2_change_notify

If plen is null when passed in, we only checked for null
in one of the two places where it could be used. Although
plen is always valid (not null) for current callers of the
SMB2_change_notify function, this change makes it more consistent.

Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Closes: https://lore.kernel.org/all/202305251831.3V1gbbFs-lkp@intel.com/
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: mapchars mount option ignored

There are two ways that special characters (not allowed in some
other operating systems like Windows, but allowed in POSIX) have
been mapped in the past ("SFU" and "SFM" mappings) to allow them
to be stored in a range reserved for special chars. The default
for Linux has been to use "mapposix" (ie the SFM mapping) but
the conversion to the new mount API in the 5.11 kernel broke
the ability to override the default mapping of the reserved
characters (like '?' and '*' and '\') via "mapchars" mount option.

This patch fixes that - so can now mount with "mapchars"
mount option to override the default ("mapposix" ie SFM) mapping.

Reported-by: Tyler Spivey <tspivey8@gmail.com>
Fixes: 24e0a1eff9e2 ("cifs: switch to new mount api")
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: display debug information better for encryption

Fix /proc/fs/cifs/DebugData to use the same case for "encryption"
(ie "Encryption" with init capital letter was used in one place).
In addition, if gcm256 encryption (intead of gcm128) is used on
a connection to a server, note that in the DebugData as well.

It now displays (when gcm256 negotiated):
 Security type: RawNTLMSSP  SessionId: 0x86125800bc000b0d encrypted(gcm256)

Acked-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix smb1 mount regression

cifs.ko maps NT_STATUS_NOT_FOUND to -EIO when SMB1 servers couldn't
resolve referral paths.  Proceed to tree connect when we get -EIO from
dfs_get_referral() as well.

Reported-by: Kris Karas (Bug Reporting) <bugs-a21@moonlit-rail.com>
Tested-by: Woody Suwalski <terraluna977@gmail.com>
Fixes: 8e3554150d6c ("cifs: fix sharing of DFS connections")
Cc: stable@vger.kernel.org # v6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: drop reference to cfile before sending oplock break

In cifs_oplock_break function we drop reference to a cfile at
the end of function, due to which close command goes on wire
after lease break acknowledgment even if file is already closed
by application but we had deferred the handle close.
If other client with limited file shareaccess waiting on lease
break ack proceeds operation on that file as soon as first client
sends ack, then we may encounter status sharing violation error
because of open handle.
Solution is to put reference to cfile(send close on wire if last ref)
and then send oplock acknowledgment to server.

Fixes: 9e31678fb403 ("SMB3: fix lease break timeout when multiple deferred close handles for the same file.")
Cc: stable@kernel.org
Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: Close all deferred handles of inode in case of handle lease break

Oplock break may occur for different file handle than the deferred
handle. Check for inode deferred closes list, if it's not empty then
close all the deferred handles of inode because we should not cache
handles if we dont have handle lease.

Eg: If openfilelist has one deferred file handle and another open file
handle from app for a same file, then on a lease break we choose the
first handle in openfile list. The first handle in list can be deferred
handle or actual open file handle from app. In case if it is actual open
handle then today, we don't close deferred handles if we lose handle lease
on a file. Problem with this is, later if app decides to close the existing
open handle then we still be caching deferred handles until deferred close
timeout. Leaving open handle may result in sharing violation when windows
client tries to open a file with limited file share access.

So we should check for deferred list of inode and walk through the list of
deferred files in inode and close all deferred files.

Fixes: 9e31678fb403 ("SMB3: fix lease break timeout when multiple deferred close handles for the same file.")
Cc: stable@kernel.org
Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: release leases for deferred close handles when freezing

We should not be caching closed files when freeze is invoked on an fs
(so we can release resources more gracefully).

Fixes xfstests generic/068 generic/390 generic/491

Reviewed-by: David Howells <dhowells@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: fix problem remounting a share after shutdown

xfstests generic/392 showed a problem where even after a
shutdown call was made on a mount, we would still attempt
to use the (now inaccessible) superblock if another mount
was attempted for the same share.

Reported-by: David Howells <dhowells@redhat.com>
Reviewed-by: David Howells <dhowells@redhat.com>
Cc: <stable@vger.kernel.org>
Fixes: 087f757b0129 ("cifs: add shutdown support")
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: force unmount was failing to close deferred close files

In investigating a failure with xfstest generic/392 it
was noticed that mounts were reusing a superblock that should
already have been freed. This turned out to be related to
deferred close files keeping a reference count until the
closetimeo expired.

Currently the only way an fs knows that mount is beginning is
when force unmount is called, but when this, ie umount_begin(),
is called all deferred close files on the share (tree
connection) should be closed immediately (unless shared by
another mount) to avoid using excess resources on the server
and to avoid reusing a superblock which should already be freed.

In umount_begin, close all deferred close handles for that
share if this is the last mount using that share on this
client (ie send the SMB3 close request over the wire for those
that have been already closed by the app but that we have
kept a handle lease open for and have not sent closes to the
server for yet).

Reported-by: David Howells <dhowells@redhat.com>
Acked-by: Bharath SM <bharathsm@microsoft.com>
Cc: <stable@vger.kernel.org>
Fixes: 78c09634f7dc ("Cifs: Fix kernel oops caused by deferred close for files.")
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: improve parallel reads of large files

rasize (ra_pages) should be set higher than read size by default
to allow parallel reads when reading large files in order to
improve performance (otherwise there is much dead time on the
network when doing readahead of large files).  Default rasize
to twice readsize.

Signed-off-by: Steve French <stfrench@microsoft.com>

do not reuse connection if share marked as isolated

"SHAREFLAG_ISOLATED_TRANSPORT" indicates that we should not reuse the socket
for this share (for future mounts).  Mark the socket as server->nosharesock if
share flags returned include SHAREFLAG_ISOLATED_TRANSPORT.

See MS-SMB2 MS-SMB2 2.2.10 and 3.2.5.5

Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix pcchunk length type in smb2_copychunk_range

Change type of pcchunk->Length from u32 to u64 to match
smb2_copychunk_range arguments type. Fixes the problem where performing
server-side copy with CIFS_IOC_COPYCHUNK_FILE ioctl resulted in incomplete
copy of large files while returning -EINVAL.

Fixes: 9bf0c9cd4314 ("CIFS: Fix SMB2/SMB3 Copy offload support (refcopy) for large files")
Cc: <stable@vger.kernel.org>
Signed-off-by: Pawel Witek <pawel.ireneusz.witek@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix sharing of DFS connections

When matching DFS connections, we can't rely on the values set in
cifs_sb_info::prepath and cifs_tcon::tree_name as they might change
during DFS failover.  The DFS referrals related to a specific DFS tcon
are already matched earlier in match_server(), therefore we can safely
skip those checks altogether as the connection is guaranteed to be
unique for the DFS tcon.

Besides, when creating or finding an SMB session, make sure to also
refcount any DFS root session related to it (cifs_ses::dfs_root_ses),
so if a new DFS mount ends up reusing the connection from the old
mount while there was an umount(2) still in progress (e.g. umount(2)
-> cifs_umount() -> reconnect -> cifs_put_tcon()), the connection
could potentially be put right after the umount(2) finished.

Patch has minor update to include fix for unused variable issue
noted by the kernel test robot

Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/oe-kbuild-all/202305041040.j7W2xQSy-lkp@intel.com/
Cc: stable@vger.kernel.org # v6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: avoid potential races when handling multiple dfs tcons

Now that a DFS tcon manages its own list of DFS referrals and
sessions, there is no point in having a single worker to refresh
referrals of all DFS tcons.  Make it faster and less prone to race
conditions when having several mounts by queueing a worker per DFS
tcon that will take care of refreshing only the DFS referrals related
to it.

Cc: stable@vger.kernel.org # v6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: protect access of TCP_Server_Info::{origin,leaf}_fullpath

Protect access of TCP_Server_Info::{origin,leaf}_fullpath when
matching DFS connections, and get rid of
TCP_Server_Info::current_fullpath while we're at it.

Cc: stable@vger.kernel.org # v6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix potential race when tree connecting ipc

Protect access of TCP_Server_Info::hostname when building the ipc tree
name as it might get freed in cifsd thread and thus causing an
use-after-free bug in __tree_connect_dfs_target().  Also, while at it,
update status of IPC tcon on success and then avoid any extra tree
connects.

Cc: stable@vger.kernel.org # v6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix potential use-after-free bugs in TCP_Server_Info::hostname

TCP_Server_Info::hostname may be updated once or many times during
reconnect, so protect its access outside reconnect path as well and
then prevent any potential use-after-free bugs.

Cc: stable@vger.kernel.org
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print smb3_fs_context::source when mounting

Print full device name (UNC + optional prefix) from @old_ctx->source
when printing info about mount.

Before patch

  mount.cifs //srv/share/dir /mnt -o ...
  dmesg
  ...
  CIFS: Attempting to mount \\srv\share

After patch

  mount.cifs //srv/share/dir /mnt -o ...
  dmesg
  ...
  CIFS: Attempting to mount //srv/share/dir

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: protect session status check in smb2_reconnect()

Use @ses->ses_lock to protect access of @ses->ses_status.

Cc: stable@vger.kernel.org
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3.1.1: correct definition for app_instance_id create contexts

The name lengths were incorrect for two create contexts.
     SMB2_CREATE_APP_INSTANCE_ID
     SMB2_CREATE_APP_INSTANCE_VERSION

Update the definitions for these two to match the protocol specs.

Acked-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: update internal module version number for cifs.ko

From 2.42 to 2.43

Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: move some common open context structs to smbfs_common

create durable and create durable reconnect context and the maximal
access create context struct definitions can be put in common code in
smbfs_common

Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: make query_on_disk_id open context consistent and move to common code

cifs and ksmbd were using a slightly different version of the query_on_disk_id
struct which could be confusing. Use the ksmbd version of this struct, and
move it to common code.

Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3.1.1: add new tree connect ShareFlags

Also update these flag names in a few places to match the simpler
easier to understand names now used in the protocol documentation
(see MS-SMB2 section 2.2.10)

Acked-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: missing lock when updating session status

Coverity noted a place where we were not grabbing
the ses_lock when setting (and checking) ses_status.

Addresses-Coverity: 1536833 ("Data race condition (MISSING_LOCK)")
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: Close deferred file handles in case of handle lease break

We should not cache deferred file handles if we dont have
handle lease on a file. And we should immediately close all
deferred handles in case of handle lease break.

Fixes: 9e31678fb403 ("SMB3: fix lease break timeout when multiple deferred close handles for the same file.")
Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

SMB3: Add missing locks to protect deferred close file list

cifs_del_deferred_close function has a critical section which modifies
the deferred close file list. We must acquire deferred_lock before
calling cifs_del_deferred_close function.

Fixes: ca08d0eac020 ("cifs: Fix memory leak on the deferred close")
Signed-off-by: Bharath SM <bharathsm@microsoft.com>
Acked-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Acked-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Avoid a cast in add_lease_context()

We have the correctly-typed struct smb2_create_req * available in the
caller.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by Ralph Boehme <slow@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Simplify SMB2_open_init()

Reduce code duplication by calculating req->CreateContextsLength in
one place.

This is the last reference to "req" in the add_*_context functions,
remove that parameter.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Enzo Matsumiya <ematsumiya@suse.de>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Simplify SMB2_open_init()

Reduce code duplication by stitching together create contexts in one
place.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Enzo Matsumiya <ematsumiya@suse.de>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Simplify SMB2_open_init()

We can point to the create contexts in just one place, we don't have
to do this in every add_*_context routine.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Enzo Matsumiya <ematsumiya@suse.de>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: avoid dup prefix path in dfs_get_automount_devname()

@server->origin_fullpath already contains the tree name + optional
prefix, so avoid calling __build_path_from_dentry_optional_prefix() as
it might end up duplicating prefix path from @cifs_sb->prepath into
final full path.

Instead, generate DFS full path by simply merging
@server->origin_fullpath with dentry's path.

This fixes the following case

mount.cifs //root/dfs/dir /mnt/ -o ...
ls /mnt/link

where cifs_dfs_do_automount() will call smb3_parse_devname() with
@devname set to "//root/dfs/dir/link" instead of
"//root/dfs/dir/dir/link".

Fixes: 7ad54b98fc1f ("cifs: use origin fullpath for automounts")
Cc: <stable@vger.kernel.org> # 6.2+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix negotiate context parsing

smb311_decode_neg_context() doesn't properly check against SMB packet
boundaries prior to accessing individual negotiate context entries. This
is due to the length check omitting the eight byte smb2_neg_context
header, as well as incorrect decrementing of len_of_ctxts.

Fixes: 5100d8a3fe03 ("SMB311: Improve checking of negotiate security contexts")
Reported-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: David Disseldorp <ddiss@suse.de>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: double lock in cifs_reconnect_tcon()

This lock was supposed to be an unlock.

Fixes: 6cc041e90c17 ("cifs: avoid races in parallel reconnects in smb1")
Signed-off-by: Dan Carpenter <error27@gmail.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: sanitize paths in cifs_update_super_prepath.

After a server reboot, clients are failing to move files with ENOENT.
This is caused by DFS referrals containing multiple separators, which
the server move call doesn't recognize.

v1: Initial patch.
v2: Move prototype to header.

Link: https://bugzilla.redhat.com/show_bug.cgi?id=2182472
Fixes: a31080899d5f ("cifs: sanitize multiple delimiters in prepath")
Actually-Fixes: 24e0a1eff9e2 ("cifs: switch to new mount api")
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Thiago Rafael Becker <tbecker@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: get rid of dead check in smb2_reconnect()

The SMB2_IOCTL check in the switch statement will never be true as we
return earlier from smb2_reconnect() if @smb2_command == SMB2_IOCTL.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: prevent infinite recursion in CIFSGetDFSRefer()

We can't call smb_init() in CIFSGetDFSRefer() as cifs_reconnect_tcon()
may end up calling CIFSGetDFSRefer() again to get new DFS referrals
and thus causing an infinite recursion.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: avoid races in parallel reconnects in smb1

Prevent multiple threads of doing negotiate, session setup and tree
connect by holding @ses->session_mutex in cifs_reconnect_tcon() while
reconnecting session and tcon.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix DFS traversal oops without CONFIG_CIFS_DFS_UPCALL

When compiled with CONFIG_CIFS_DFS_UPCALL disabled, cifs_dfs_d_automount
is NULL. cifs.ko logic for mapping CIFS_FATTR_DFS_REFERRAL attributes to
S_AUTOMOUNT and corresponding dentry flags is retained regardless of
CONFIG_CIFS_DFS_UPCALL, leading to a NULL pointer dereference in
VFS follow_automount() when traversing a DFS referral link:
  BUG: kernel NULL pointer dereference, address: 0000000000000000
  ...
  Call Trace:
   <TASK>
   __traverse_mounts+0xb5/0x220
   ? cifs_revalidate_mapping+0x65/0xc0 [cifs]
   step_into+0x195/0x610
   ? lookup_fast+0xe2/0xf0
   path_lookupat+0x64/0x140
   filename_lookup+0xc2/0x140
   ? __create_object+0x299/0x380
   ? kmem_cache_alloc+0x119/0x220
   ? user_path_at_empty+0x31/0x50
   user_path_at_empty+0x31/0x50
   __x64_sys_chdir+0x2a/0xd0
   ? exit_to_user_mode_prepare+0xca/0x100
   do_syscall_64+0x42/0x90
   entry_SYSCALL_64_after_hwframe+0x72/0xdc

This fix adds an inline cifs_dfs_d_automount() {return -EREMOTE} handler
when CONFIG_CIFS_DFS_UPCALL is disabled. An alternative would be to
avoid flagging S_AUTOMOUNT, etc. without CONFIG_CIFS_DFS_UPCALL. This
approach was chosen as it provides more control over the error path.

Signed-off-by: David Disseldorp <ddiss@suse.de>
Cc: stable@vger.kernel.org
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: fix unusable share after force unmount failure

If user does forced unmount ("umount -f") while files are still open
on the share (as was seen in a Kubernetes example running on SMB3.1.1
mount) then we were marking the share as "TID_EXITING" in umount_begin()
which caused all subsequent operations (except write) to fail ... but
unfortunately when umount_begin() is called we do not know yet that
there are open files or active references on the share that would prevent
unmount from succeeding.  Kubernetes had example when they were doing
umount -f when files were open which caused the share to become
unusable until the files were closed (and the umount retried).

Fix this so that TID_EXITING is not set until we are about to send
the tree disconnect (not at the beginning of forced umounts in
umount_begin) so that if "umount -f" fails (due to open files or
references) the mount is still usable.

Cc: stable@vger.kernel.org
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix dentry lookups in directory handle cache

Get rid of any prefix paths in @path before lookup_positive_unlocked()
as it will call ->lookup() which already adds those prefix paths
through build_path_from_dentry().

This has caused a performance regression when mounting shares with a
prefix path where readdir(2) would end up retrying several times to
open bad directory names that contained duplicate prefix paths.

Fix this by skipping any prefix paths in @path before calling
lookup_positive_unlocked().

Fixes: e4029e072673 ("cifs: find and use the dentry for cached non-root directories also")
Cc: stable@vger.kernel.org # 6.1+
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

smb3: lower default deferred close timeout to address perf regression

Performance tests with large number of threads noted that the change
of the default closetimeo (deferred close timeout between when
close is done by application and when client has to send the close
to the server), to 5 seconds from 1 second, significantly degraded
perf in some cases like this (in the filebench example reported,
the stats show close requests on the wire taking twice as long,
and 50% regression in filebench perf). This is stil configurable
via mount parm closetimeo, but to be safe, decrease default back
to its previous value of 1 second.

Reported-by: Yin Fengwei <fengwei.yin@intel.com>
Reported-by: kernel test robot <yujie.liu@intel.com>
Link: https://lore.kernel.org/lkml/997614df-10d4-af53-9571-edec36b0e2f3@intel.com/
Fixes: 5efdd9122eff ("smb3: allow deferred close timeout to be configurable")
Cc: stable@vger.kernel.org # 6.0+
Tested-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix missing unload_nls() in smb2_reconnect()

Make sure to unload_nls() @nls_codepage if we no longer need it.

Fixes: bc962159e8e3 ("cifs: avoid race conditions with parallel reconnects")
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: avoid race conditions with parallel reconnects

When multiple processes/channels do reconnects in parallel
we used to return success immediately
negotiate/session-setup/tree-connect, causing race conditions
between processes that enter the function in parallel.
This caused several errors related to session not found to
show up during parallel reconnects.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: append path to open_enter trace event

We do not dump the file path for smb3_open_enter ftrace
calls, which is a severe handicap while debugging
using ftrace evens. This change adds that info.

Unfortunately, we're not updating the path in open params
in many places; which I had to do as a part of this change.
SMB2_open gets path in utf16 format, but it's easier of
path is supplied as char pointer in oparms.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: print session id while listing open files

In the output of /proc/fs/cifs/open_files, we only print
the tree id for the tcon of each open file. It becomes
difficult to know which tcon these files belong to with
just the tree id.

This change dumps ses id in addition to all other data today.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: dump pending mids for all channels in DebugData

Currently, we only dump the pending mid information only
on the primary channel in /proc/fs/cifs/DebugData.
If multichannel is active, we do not print the pending MID
list on secondary channels.

This change will dump the pending mids for all the channels
based on server->conn_id.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: empty interface list when server doesn't support query interfaces

When querying server interfaces returns -EOPNOTSUPP,
clear the list of interfaces. Assumption is that multichannel
would be disabled too.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: do not poll server interfaces too regularly

We have the server interface list hanging off the tcon
structure today for reasons unknown. So each tcon which is
connected to a file server can query them separately,
which is really unnecessary. To avoid this, in the query
function, we will check the time of last update of the
interface list, and avoid querying the server if it is
within a certain range.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: lock chan_lock outside match_session

Coverity had rightly indicated a possible deadlock
due to chan_lock being done inside match_session.
All callers of match_* functions should pick up the
necessary locks and call them.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Fixes: 724244cdb382 ("cifs: protect session channel fields with chan_lock")
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: check only tcon status on tcon related functions

We had a couple of checks for session in cifs_tree_connect
and cifs_mark_open_files_invalid, which were unnecessary.
And that was done with ses_lock. Changed that to tc_lock too.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: use DFS root session instead of tcon ses

Use DFS root session whenever possible to get new DFS referrals
otherwise we might end up with an IPC tcon (tcon->ses->tcon_ipc) that
doesn't respond to them.  It should be safe accessing
@ses->dfs_root_ses directly in cifs_inval_name_dfs_link_error() as it
has same lifetime as of @tcon.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: return DFS root session id in DebugData

Return the DFS root session id in /proc/fs/cifs/DebugData to make it
easier to track which IPC tcon was used to get new DFS referrals for a
specific connection, and aids in debugging.

A simple output of it would be

  Sessions:
  1) Address: 192.168.1.13 Uses: 1 Capability: 0x300067   Session Status: 1
  Security type: RawNTLMSSP  SessionId: 0xd80000000009
  User: 0 Cred User: 0
  DFS root session id: 0x128006c000035

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: fix use-after-free bug in refresh_cache_worker()

The UAF bug occurred because we were putting DFS root sessions in
cifs_umount() while DFS cache refresher was being executed.

Make DFS root sessions have same lifetime as DFS tcons so we can avoid
the use-after-free bug is DFS cache refresher and other places that
require IPCs to get new DFS referrals on.  Also, get rid of mount
group handling in DFS cache as we no longer need it.

This fixes below use-after-free bug catched by KASAN

[ 379.946955] BUG: KASAN: use-after-free in __refresh_tcon.isra.0+0x10b/0xc10 [cifs]
[ 379.947642] Read of size 8 at addr ffff888018f57030 by task kworker/u4:3/56
[ 379.948096]
[ 379.948208] CPU: 0 PID: 56 Comm: kworker/u4:3 Not tainted 6.2.0-rc7-lku #23
[ 379.948661] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS
rel-1.16.0-0-gd239552-rebuilt.opensuse.org 04/01/2014
[ 379.949368] Workqueue: cifs-dfscache refresh_cache_worker [cifs]
[ 379.949942] Call Trace:
[ 379.950113] <TASK>
[ 379.950260] dump_stack_lvl+0x50/0x67
[ 379.950510] print_report+0x16a/0x48e
[ 379.950759] ? __virt_addr_valid+0xd8/0x160
[ 379.951040] ? __phys_addr+0x41/0x80
[ 379.951285] kasan_report+0xdb/0x110
[ 379.951533] ? __refresh_tcon.isra.0+0x10b/0xc10 [cifs]
[ 379.952056] ? __refresh_tcon.isra.0+0x10b/0xc10 [cifs]
[ 379.952585] __refresh_tcon.isra.0+0x10b/0xc10 [cifs]
[ 379.953096] ? __pfx___refresh_tcon.isra.0+0x10/0x10 [cifs]
[ 379.953637] ? __pfx___mutex_lock+0x10/0x10
[ 379.953915] ? lock_release+0xb6/0x720
[ 379.954167] ? __pfx_lock_acquire+0x10/0x10
[ 379.954443] ? refresh_cache_worker+0x34e/0x6d0 [cifs]
[ 379.954960] ? __pfx_wb_workfn+0x10/0x10
[ 379.955239] refresh_cache_worker+0x4ad/0x6d0 [cifs]
[ 379.955755] ? __pfx_refresh_cache_worker+0x10/0x10 [cifs]
[ 379.956323] ? __pfx_lock_acquired+0x10/0x10
[ 379.956615] ? read_word_at_a_time+0xe/0x20
[ 379.956898] ? lockdep_hardirqs_on_prepare+0x12/0x220
[ 379.957235] process_one_work+0x535/0x990
[ 379.957509] ? __pfx_process_one_work+0x10/0x10
[ 379.957812] ? lock_acquired+0xb7/0x5f0
[ 379.958069] ? __list_add_valid+0x37/0xd0
[ 379.958341] ? __list_add_valid+0x37/0xd0
[ 379.958611] worker_thread+0x8e/0x630
[ 379.958861] ? __pfx_worker_thread+0x10/0x10
[ 379.959148] kthread+0x17d/0x1b0
[ 379.959369] ? __pfx_kthread+0x10/0x10
[ 379.959630] ret_from_fork+0x2c/0x50
[ 379.959879] </TASK>

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: set DFS root session in cifs_get_smb_ses()

Set the DFS root session pointer earlier when creating a new SMB
session to prevent racing with smb2_reconnect(), cifs_reconnect_tcon()
and DFS cache refresher.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: generate signkey for the channel that's reconnecting

Before my changes to how multichannel reconnects work, the
primary channel was always used to do a non-binding session
setup. With my changes, that is not the case anymore.
Missed this place where channel at index 0 was forcibly
updated with the signing key.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Fix smb2_set_path_size()

If cifs_get_writable_path() finds a writable file, smb2_compound_op()
must use that file's FID and not the COMPOUND_FID.

Cc: stable@vger.kernel.org
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: Move the in_send statistic to __smb_send_rqst()

When send SMB_COM_NT_CANCEL and RFC1002_SESSION_REQUEST, the
in_send statistic was lost.

Let's move the in_send statistic to the send function to avoid
this scenario.

Fixes: 7ee1af765dfa ("[CIFS]")
Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5@huawei.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: prevent data race in cifs_reconnect_tcon()

Make sure to get an up-to-date TCP_Server_Info::nr_targets value prior
to waiting the server to be reconnected in cifs_reconnect_tcon().  It
is set in cifs_tcp_ses_needs_reconnect() and protected by
TCP_Server_Info::srv_lock.

Create a new cifs_wait_for_server_reconnect() helper that can be used
by both SMB2+ and CIFS reconnect code.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: improve checking of DFS links over STATUS_OBJECT_NAME_INVALID

Do not map STATUS_OBJECT_NAME_INVALID to -EREMOTE under non-DFS
shares, or 'nodfs' mounts or CONFIG_CIFS_DFS_UPCALL=n builds.
Otherwise, in the slow path, get a referral to figure out whether it
is an actual DFS link.

This could be simply reproduced under a non-DFS share by running the
following

  $ mount.cifs //srv/share /mnt -o ...
  $ cat /mnt/$(printf '\U110000')
  cat: '/mnt/'$'\364\220\200\200': Object is remote

Fixes: c877ce47e137 ("cifs: reduce roundtrips on create/qinfo requests")
CC: stable@vger.kernel.org # 6.2
Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>

cifs: reuse cifs_match_ipaddr for comparison of dstaddr too

We have two pieces of code that does pretty much the same
comparison. This change reuses cifs_match_ipaddr within
match_address.

Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>