| Age | Commit message (Collapse) | Author |
|---|
|
Pull smb client fixes from Steve French:
- Multichannel reconnect channel selection fix
- Fix for smbdirect (RDMA) disconnect bug
- Fix for incorrect username length check
- Fix memory leak in mount parm processing
* tag 'v6.18-rc5-smb-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
smb: client: let smbd_disconnect_rdma_connection() turn CREATED into DISCONNECTED
smb: fix invalid username check in smb3_fs_context_parse_param()
cifs: client: fix memory leak in smb3_fs_context_parse_param
smb: client: fix cifs_pick_channel when channel needs reconnect
|
|
DISCONNECTED
When smbd_disconnect_rdma_connection() turns SMBDIRECT_SOCKET_CREATED
into SMBDIRECT_SOCKET_ERROR, we'll have the situation that
smbd_disconnect_rdma_work() will set SMBDIRECT_SOCKET_DISCONNECTING
and call rdma_disconnect(), which likely fails as we never reached
the RDMA_CM_EVENT_ESTABLISHED. it means that
wait_event(sc->status_wait, sc->status == SMBDIRECT_SOCKET_DISCONNECTED)
in smbd_destroy() will hang forever in SMBDIRECT_SOCKET_DISCONNECTING
never reaching SMBDIRECT_SOCKET_DISCONNECTED.
So we directly go from SMBDIRECT_SOCKET_CREATED to
SMBDIRECT_SOCKET_DISCONNECTED.
Fixes: ffbfc73e84eb ("smb: client: let smbd_disconnect_rdma_connection() set SMBDIRECT_SOCKET_ERROR...")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Since the maximum return value of strnlen(..., CIFS_MAX_USERNAME_LEN)
is CIFS_MAX_USERNAME_LEN, length check in smb3_fs_context_parse_param()
is always FALSE and invalid.
Fix the comparison in if statement.
Signed-off-by: Yiqi Sun <sunyiqixm@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
DISCONNECTED
When smb_direct_disconnect_rdma_connection() turns SMBDIRECT_SOCKET_CREATED
into SMBDIRECT_SOCKET_ERROR, we'll have the situation that
smb_direct_disconnect_rdma_work() will set SMBDIRECT_SOCKET_DISCONNECTING
and call rdma_disconnect(), which likely fails as we never reached
the RDMA_CM_EVENT_ESTABLISHED. it means that
wait_event(sc->status_wait, sc->status == SMBDIRECT_SOCKET_DISCONNECTED)
in free_transport() will hang forever in SMBDIRECT_SOCKET_DISCONNECTING
never reaching SMBDIRECT_SOCKET_DISCONNECTED.
So we directly go from SMBDIRECT_SOCKET_CREATED to
SMBDIRECT_SOCKET_DISCONNECTED.
Fixes: b3fd52a0d85c ("smb: server: let smb_direct_disconnect_rdma_connection() set SMBDIRECT_SOCKET_ERROR...")
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
When the per-IP connection limit is exceeded in ksmbd_kthread_fn(),
the code sets ret = -EAGAIN and continues the accept loop without
closing the just-accepted socket. That leaks one socket per rejected
attempt from a single IP and enables a trivial remote DoS.
Release client_sk before continuing.
This bug was found with ZeroPath.
Cc: stable@vger.kernel.org
Signed-off-by: Joshua Rogers <linux@joshua.hu>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
smb_direct_prepare_negotiation() posts a recv and then, if
smb_direct_accept_client() fails, calls put_recvmsg() on the same
buffer. That unmaps and recycles a buffer that is still posted on
the QP., which can lead to device DMA into unmapped or reused memory.
Track whether the recv was posted and only return it if it was never
posted. If accept fails after a post, leave it for teardown to drain
and complete safely.
Signed-off-by: Joshua Rogers <linux@joshua.hu>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
The user calls fsconfig twice, but when the program exits, free() only
frees ctx->source for the second fsconfig, not the first.
Regarding fc->source, there is no code in the fs context related to its
memory reclamation.
To fix this memory leak, release the source memory corresponding to ctx
or fc before each parsing.
syzbot reported:
BUG: memory leak
unreferenced object 0xffff888128afa360 (size 96):
backtrace (crc 79c9c7ba):
kstrdup+0x3c/0x80 mm/util.c:84
smb3_fs_context_parse_param+0x229b/0x36c0 fs/smb/client/fs_context.c:1444
BUG: memory leak
unreferenced object 0xffff888112c7d900 (size 96):
backtrace (crc 79c9c7ba):
smb3_fs_context_fullpath+0x70/0x1b0 fs/smb/client/fs_context.c:629
smb3_fs_context_parse_param+0x2266/0x36c0 fs/smb/client/fs_context.c:1438
Reported-by: syzbot+72afd4c236e6bc3f4bac@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=72afd4c236e6bc3f4bac
Cc: stable@vger.kernel.org
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
cifs_pick_channel iterates candidate channels using cur. The
reconnect-state test mistakenly used a different variable.
This checked the wrong slot and would cause us to skip a healthy channel
and to dispatch on one that needs reconnect, occasionally failing
operations when a channel was down.
Fix by replacing for the correct variable.
Fixes: fc43a8ac396d ("cifs: cifs_pick_channel should try selecting active channels")
Cc: stable@vger.kernel.org
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Henrique Carvalho <henrique.carvalho@suse.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Pull smb client fixes from Steve French:
- Fix change notify packet validation check
- Refcount fix (e.g. rename error paths)
- Fix potential UAF due to missing locks on directory lease refcount
* tag 'v6.18rc4-SMB-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
smb: client: validate change notify buffer before copy
smb: client: fix refcount leak in smb2_set_path_attr
smb: client: fix potential UAF in smb2_close_cached_fid()
|
|
SMB2_change_notify called smb2_validate_iov() but ignored the return
code, then kmemdup()ed using server provided OutputBufferOffset/Length.
Check the return of smb2_validate_iov() and bail out on error.
Discovered with help from the ZeroPath security tooling.
Signed-off-by: Joshua Rogers <linux@joshua.hu>
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Cc: stable@vger.kernel.org
Fixes: e3e9463414f61 ("smb3: improve SMB3 change notification support")
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Fix refcount leak in `smb2_set_path_attr` when path conversion fails.
Function `cifs_get_writable_path` returns `cfile` with its reference
counter `cfile->count` increased on success. Function `smb2_compound_op`
would decrease the reference counter for `cfile`, as stated in its
comment. By calling `smb2_rename_path`, the reference counter of `cfile`
would leak if `cifs_convert_path_to_utf16` fails in `smb2_set_path_attr`.
Fixes: 8de9e86c67ba ("cifs: create a helper to find a writeable handle by path name")
Acked-by: Henrique Carvalho <henrique.carvalho@suse.com>
Signed-off-by: Shuhao Fu <sfual@cse.ust.hk>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
find_or_create_cached_dir() could grab a new reference after kref_put()
had seen the refcount drop to zero but before cfid_list_lock is acquired
in smb2_close_cached_fid(), leading to use-after-free.
Switch to kref_put_lock() so cfid_release() is called with
cfid_list_lock held, closing that gap.
Fixes: ebe98f1447bb ("cifs: enable caching of directories for which a lease is held")
Cc: stable@vger.kernel.org
Reported-by: Jay Shin <jaeshin@redhat.com>
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Signed-off-by: Henrique Carvalho <henrique.carvalho@suse.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Current ksmbd_rdma_capable_netdev fails to mark certain RDMA-capable
inerfaces such as IPoIB as RDMA capable after reverting GUID matching code
due to layer violation.
This patch check the ARPHRD_INFINIBAND type safely identifies an IPoIB
interface without introducing a layer violation, ensuring RDMA
functionality is correctly enabled for these interfaces.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
If user set bridge interface as actual RDMA-capable NICs are lower devices,
ksmbd can not detect as RDMA capable. This patch can detect the RDMA
capable lower devices from bridge master or VLAN. With this change, ksmbd
can accept both TCP and RDMA connections through the same bridge IP
address, allowing mixed transport operation without requiring separate
interfaces.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Pull smb client fixes from Steve French:
- fix potential UAF in statfs
- DFS fix for expired referrals
- fix minor modinfo typo
- small improvement to reconnect for smbdirect
* tag '6.18-rc3-smb-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
smb: client: call smbd_destroy() in the same splace as kernel_sock_shutdown()/sock_release()
smb: client: handle lack of IPC in dfs_cache_refresh()
smb: client: fix potential cfid UAF in smb2_query_info_compound
cifs: fix typo in enable_gcm_256 module parameter
|
|
kernel_sock_shutdown()/sock_release()
With commit b0432201a11b ("smb: client: let destroy_mr_list() keep
smbdirect_mr_io memory if registered") the changes from commit
214bab448476 ("cifs: Call MID callback before destroying transport") and
commit 1d2a4f57cebd ("cifs:smbd When reconnecting to server, call
smbd_destroy() after all MIDs have been called") are no longer needed.
And it's better to use the same logic flow, so that
the chance of smbdirect related problems is smaller.
Fixes: 214bab448476 ("cifs: Call MID callback before destroying transport")
Fixes: 1d2a4f57cebd ("cifs:smbd When reconnecting to server, call smbd_destroy() after all MIDs have been called")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
In very rare cases, DFS mounts could end up with SMB sessions without
any IPC connections. These mounts are only possible when having
unexpired cached DFS referrals, hence not requiring any IPC
connections during the mount process.
Try to establish those missing IPC connections when refreshing DFS
referrals. If the server is still rejecting it, then simply ignore
and leave expired cached DFS referral for any potential DFS failovers.
Reported-by: Jay Shin <jaeshin@redhat.com>
Signed-off-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Cc: David Howells <dhowells@redhat.com>
Cc: linux-cifs@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
When smb2_query_info_compound() retries, a previously allocated cfid may
have been freed in the first attempt.
Because cfid wasn't reset on replay, later cleanup could act on a stale
pointer, leading to a potential use-after-free.
Reinitialize cfid to NULL under the replay label.
Example trace (trimmed):
refcount_t: underflow; use-after-free.
WARNING: CPU: 1 PID: 11224 at ../lib/refcount.c:28 refcount_warn_saturate+0x9c/0x110
[...]
RIP: 0010:refcount_warn_saturate+0x9c/0x110
[...]
Call Trace:
<TASK>
smb2_query_info_compound+0x29c/0x5c0 [cifs f90b72658819bd21c94769b6a652029a07a7172f]
? step_into+0x10d/0x690
? __legitimize_path+0x28/0x60
smb2_queryfs+0x6a/0xf0 [cifs f90b72658819bd21c94769b6a652029a07a7172f]
smb311_queryfs+0x12d/0x140 [cifs f90b72658819bd21c94769b6a652029a07a7172f]
? kmem_cache_alloc+0x18a/0x340
? getname_flags+0x46/0x1e0
cifs_statfs+0x9f/0x2b0 [cifs f90b72658819bd21c94769b6a652029a07a7172f]
statfs_by_dentry+0x67/0x90
vfs_statfs+0x16/0xd0
user_statfs+0x54/0xa0
__do_sys_statfs+0x20/0x50
do_syscall_64+0x58/0x80
Cc: stable@kernel.org
Fixes: 4f1fffa237692 ("cifs: commands that are retried should have replay flag set")
Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.com>
Acked-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Enzo Matsumiya <ematsumiya@suse.de>
Signed-off-by: Henrique Carvalho <henrique.carvalho@suse.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
smb_direct_disconnect_rdma_work()
All handlers triggered by ib_drain_qp() should already see the
broken connection.
smb_direct_cm_handler() is called under a mutex of the rdma_cm,
we should make sure ib_drain_qp() and all rdma layer logic completes
and unlocks the mutex.
It means free_transport() will also already see the connection
as SMBDIRECT_SOCKET_DISCONNECTED, so we need to call
crdma_[un]lock_handler(sc->rdma.cm_id) around
ib_drain_qp(), rdma_destroy_qp(), ib_free_cq() and ib_dealloc_pd().
Otherwise we free resources while the ib_drain_qp() within
smb_direct_cm_handler() is still running.
We have to unlock before rdma_destroy_id() as it locks again.
Fixes: 141fa9824c0f ("ksmbd: call ib_drain_qp when disconnected")
Fixes: 4c564f03e23b ("smb: server: make use of common smbdirect_socket")
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
We now activate sc->recv_io.posted.refill_work and sc->idle.immediate_work
only after a successful negotiation, before sending the negotiation
response.
It means the queue_work(sc->workqueue, &sc->recv_io.posted.refill_work)
in put_recvmsg() of the negotiate request, is a no-op now.
It also means our explicit smb_direct_post_recv_credits() will
have queue_work(sc->workqueue, &sc->idle.immediate_work) as no-op.
This should make sure we don't have races and post any immediate
data_transfer message that tries to grant credits to the peer,
before we send the negotiation response, as that will grant
the initial credits to the peer.
Fixes: 0626e6641f6b ("cifsd: add server handler for central processing and tranport layers")
Fixes: 1cde0a74a7a8 ("smb: server: don't use delayed_work for post_recv_credits_work")
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
handle_response() dereferences the payload as a 4-byte handle without
verifying that the declared payload size is at least 4 bytes. A malformed
or truncated message from ksmbd.mountd can lead to a 4-byte read past the
declared payload size. Validate the size before dereferencing.
This is a minimal fix to guard the initial handle read.
Fixes: 0626e6641f6b ("cifsd: add server handler for central processing and tranport layers")
Cc: stable@vger.kernel.org
Reported-by: Qianchang Zhao <pioooooooooip@gmail.com>
Signed-off-by: Qianchang Zhao <pioooooooooip@gmail.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Fix typo in description of enable_gcm_256 module parameter
Suggested-by: Thomas Spear <speeddymon@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Pull smb server fixes from Steve French:
"smbdirect (RDMA) fixes in order avoid potential submission queue
overflows:
- free transport teardown fix
- credit related fixes (five server related, one client related)"
* tag 'v6.18-rc2-smb-server-fixes' of git://git.samba.org/ksmbd:
smb: server: let free_transport() wait for SMBDIRECT_SOCKET_DISCONNECTED
smb: client: make use of smbdirect_socket.send_io.lcredits.*
smb: server: make use of smbdirect_socket.send_io.lcredits.*
smb: server: simplify sibling_list handling in smb_direct_flush_send_list/send_done
smb: server: smb_direct_disconnect_rdma_connection() already wakes all waiters on error
smb: smbdirect: introduce smbdirect_socket.send_io.lcredits.*
smb: server: allocate enough space for RW WRs and ib_drain_qp()
|
|
We should wait for the rdma_cm to become SMBDIRECT_SOCKET_DISCONNECTED!
At least on the client side (with similar code)
wait_event_interruptible() often returns with -ERESTARTSYS instead of
waiting for SMBDIRECT_SOCKET_DISCONNECTED.
We should use wait_event() here too, which makes the code be identical
in client and server, which will help when moving to common functions.
Fixes: b31606097de8 ("smb: server: move smb_direct_disconnect_rdma_work() into free_transport()")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Make cifs #include cifsglob.h in advance of #including trace.h so that the
structures defined in cifsglob.h can be accessed directly by the cifs
tracepoints rather than the callers having to manually pass in the bits and
pieces.
This should allow the tracepoints to be made more efficient to use as well
as easier to read in the code.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Paulo Alcantara <pc@manguebit.org>
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
As the SMB1 and SMB2/3 calc_signature functions are called from separate
sign and verify paths, just call them directly rather than using a function
pointer. The SMB3 calc_signature then jumps to the SMB2 variant if
necessary.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Enzo Matsumiya <ematsumiya@suse.de>
cc: Paulo Alcantara <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: Tom Talpey <tom@talpey.com>
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
There is no need to force a lookup by unhashing the moved dentry after
successfully renaming the file on server. The file metadata will be
re-fetched from server, if necessary, in the next call to
->d_revalidate() anyways.
Signed-off-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Reviewed-by: David Howells <dhowells@redhat.com>
Cc: stable@vger.kernel.org
Cc: linux-cifs@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Fix TCP_Server_Info::credits to be signed, just as echo_credits and
oplock_credits are. This also fixes what ought to get at least a
compilation warning if not an outright error in *get_credits_field() as a
pointer to the unsigned server->credits field is passed back as a pointer
to a signed int.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: linux-cifs@vger.kernel.org
Cc: stable@vger.kernel.org
Acked-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Acked-by: Pavel Shilovskiy <pshilovskiy@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
This makes the logic to prevent on overflow of
the send submission queue with ib_post_send() easier.
As we first get a local credit and then a remote credit
before we mark us as pending.
For now we'll keep the logic around smbdirect_socket.send_io.pending.*,
but that will likely change or be removed completely.
The server will get a similar logic soon, so
we'll be able to share the send code in future.
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
This introduces logic to prevent on overflow of
the send submission queue with ib_post_send() easier.
As we first get a local credit and then a remote credit
before we mark us as pending.
From reading the git history of the linux smbdirect
implementations in client and server) it was seen
that a peer granted more credits than we requested.
I guess that only happened because of bugs in our
implementation which was active as client and server.
I guess Windows won't do that.
So the local credits make sure we only use the amount
of credits we asked for.
Fixes: 0626e6641f6b ("cifsd: add server handler for central processing and tranport layers")
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
smb_direct_flush_send_list/send_done
We have a list handling that is much easier to understand:
1. Before smb_direct_flush_send_list() is called all
struct smbdirect_send_io messages are part of
send_ctx->msg_list
2. Before smb_direct_flush_send_list() calls
smb_direct_post_send() we remove the last
element in send_ctx->msg_list and move all
others into last->sibling_list. As only
last has IB_SEND_SIGNALED and gets a completion
vis send_done().
3. send_done() has an easy way to free all others
in sendmsg->sibling_list (if there are any).
And use list_for_each_entry_safe() instead of
a complex custom logic.
This will help us to share send_done() in common
code soon, as it will work fine for the client too,
where last->sibling_list is currently always an empty list.
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
waiters on error
There's no need to care about pending or credit counters when we
already disconnecting.
And all related wait_event conditions already check for broken
connections too.
This will simplify the code and makes the following changes simpler.
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
This will be used to implement a logic in order to make sure
we don't overflow the send submission queue for ib_post_send().
We will initialize the local credits with the
fixed sp->send_credit_target value, which matches
the reserved slots in the submission queue for ib_post_send().
We will be a local credit first and then wait for a remote credit,
if we managed to get both we are allowed to post an
IB_WR_SEND[_WITH_INV]. The local credit is given back to
the pool when we get the local ib_post_send() completion,
while remote credits are granted by the peer.
From reading the git history of the linux smbdirect
implementations in client and server) it was seen
that a peer granted more credits than we requested.
I guess that only happened because of bugs in our
implementation which was active as client and server.
I guess Windows won't do that.
So the local credits make sure we only use the amount
of credits we asked for.
The client already has some logic for this based on
smbdirect_socket.send_io.pending.count, but that
counts in the order direction and makes it complex it
share common logic for various credits classes.
That logic will be replaced soon.
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Make use of rdma_rw_mr_factor() to calculate the number of rw
credits and the number of pages per RDMA RW operation.
We get the same numbers for iWarp connections, tested
with siw.ko and irdma.ko (in iWarp mode).
siw:
CIFS: max_qp_rd_atom=128, max_fast_reg_page_list_len = 256
CIFS: max_sgl_rd=0, max_sge_rd=1
CIFS: responder_resources=32 max_frmr_depth=256 mr_io.type=0
CIFS: max_send_wr 384, device reporting max_cqe 3276800 max_qp_wr 32768
ksmbd: max_fast_reg_page_list_len = 256, max_sgl_rd=0, max_sge_rd=1
ksmbd: device reporting max_cqe 3276800 max_qp_wr 32768
ksmbd: Old sc->rw_io.credits: max = 9, num_pages = 256
ksmbd: New sc->rw_io.credits: max = 9, num_pages = 256, maxpages=2048
ksmbd: Info: rdma_send_wr 27 + max_send_wr 256 = 283
irdma (in iWarp mode):
CIFS: max_qp_rd_atom=127, max_fast_reg_page_list_len = 262144
CIFS: max_sgl_rd=0, max_sge_rd=13
CIFS: responder_resources=32 max_frmr_depth=2048 mr_io.type=0
CIFS: max_send_wr 384, device reporting max_cqe 1048574 max_qp_wr 4063
ksmbd: max_fast_reg_page_list_len = 262144, max_sgl_rd=0, max_sge_rd=13
ksmbd: device reporting max_cqe 1048574 max_qp_wr 4063
ksmbd: Old sc->rw_io.credits: max = 9, num_pages = 256
ksmbd: New sc->rw_io.credits: max = 9, num_pages = 256, maxpages=2048
ksmbd: rdma_send_wr 27 + max_send_wr 256 = 283
This means that we get the different correct numbers for ROCE,
tested with rdma_rxe.ko and irdma.ko (in RoCEv2 mode).
rxe:
CIFS: max_qp_rd_atom=128, max_fast_reg_page_list_len = 512
CIFS: max_sgl_rd=0, max_sge_rd=32
CIFS: responder_resources=32 max_frmr_depth=512 mr_io.type=0
CIFS: max_send_wr 384, device reporting max_cqe 32767 max_qp_wr 1048576
ksmbd: max_fast_reg_page_list_len = 512, max_sgl_rd=0, max_sge_rd=32
ksmbd: device reporting max_cqe 32767 max_qp_wr 1048576
ksmbd: Old sc->rw_io.credits: max = 9, num_pages = 256
ksmbd: New sc->rw_io.credits: max = 65, num_pages = 32, maxpages=2048
ksmbd: rdma_send_wr 65 + max_send_wr 256 = 321
irdma (in RoCEv2 mode):
CIFS: max_qp_rd_atom=127, max_fast_reg_page_list_len = 262144,
CIFS: max_sgl_rd=0, max_sge_rd=13
CIFS: responder_resources=32 max_frmr_depth=2048 mr_io.type=0
CIFS: max_send_wr 384, device reporting max_cqe 1048574 max_qp_wr 4063
ksmbd: max_fast_reg_page_list_len = 262144, max_sgl_rd=0, max_sge_rd=13
ksmbd: device reporting max_cqe 1048574 max_qp_wr 4063
ksmbd: Old sc->rw_io.credits: max = 9, num_pages = 256,
ksmbd: New sc->rw_io.credits: max = 159, num_pages = 13, maxpages=2048
ksmbd: rdma_send_wr 159 + max_send_wr 256 = 415
And rely on rdma_rw_init_qp() to setup ib_mr_pool_init() for
RW MRs. ib_mr_pool_destroy() will be called by rdma_rw_cleanup_mrs().
It seems the code was implemented before the rdma_rw_* layer
was fully established in the kernel.
While there also add additional space for ib_drain_qp().
This should make sure ib_post_send() will never fail
because the submission queue is full.
Fixes: ddbdc861e37c ("ksmbd: smbd: introduce read/write credits for RDMA read/write")
Fixes: 4c564f03e23b ("smb: server: make use of common smbdirect_socket")
Fixes: 177368b99243 ("smb: server: make use of common smbdirect_socket_parameters")
Fixes: 95475d8886bd ("smb: server: make use smbdirect_socket.rw_io.credits")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Add missing smb3_rw_credits tracepoints to cifs_readv_callback() (for SMB1)
to match those of SMB2/3.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Steve French <sfrench@samba.org>
cc: Paulo Alcantara <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: Tom Talpey <tom@talpey.com>
cc: linux-cifs@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
The IB_WR_REG_MR and IB_WR_LOCAL_INV operations for smbdirect_mr_io
structures should never fail because the submission or completion queues
are too small. So we allocate more send_wr depending on the (local) max
number of MRs.
While there also add additional space for ib_drain_qp().
This should make sure ib_post_send() will never fail
because the submission queue is full.
Fixes: f198186aa9bb ("CIFS: SMBD: Establish SMB Direct connection")
Fixes: cc55f65dd352 ("smb: client: make use of common smbdirect_socket_parameters")
Cc: stable@vger.kernel.org
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Pull smb client fixes from Steve French:
"smb client fixes, security and smbdirect improvements, and some minor cleanup:
- Important OOB DFS fix
- Fix various potential tcon refcount leaks
- smbdirect (RDMA) fixes (following up from test event a few weeks
ago):
- Fixes to improve and simplify handling of memory lifetime of
smbdirect_mr_io structures, when a connection gets disconnected
- Make sure we really wait to reach SMBDIRECT_SOCKET_DISCONNECTED
before destroying resources
- Make sure the send/recv submission/completion queues are large
enough to avoid ib_post_send() from failing under pressure
- convert cifs.ko to use the recommended crypto libraries (instead of
crypto_shash), this also can improve performance
- Three small cleanup patches"
* tag '6.18-rc1-smb-client-fixes' of git://git.samba.org/sfrench/cifs-2.6: (24 commits)
smb: client: Consolidate cmac(aes) shash allocation
smb: client: Remove obsolete crypto_shash allocations
smb: client: Use HMAC-MD5 library for NTLMv2
smb: client: Use MD5 library for SMB1 signature calculation
smb: client: Use MD5 library for M-F symlink hashing
smb: client: Use HMAC-SHA256 library for SMB2 signature calculation
smb: client: Use HMAC-SHA256 library for key generation
smb: client: Use SHA-512 library for SMB3.1.1 preauth hash
cifs: parse_dfs_referrals: prevent oob on malformed input
smb: client: Fix refcount leak for cifs_sb_tlink
smb: client: let smbd_destroy() wait for SMBDIRECT_SOCKET_DISCONNECTED
smb: move some duplicate definitions to common/cifsglob.h
smb: client: let destroy_mr_list() keep smbdirect_mr_io memory if registered
smb: client: let destroy_mr_list() call ib_dereg_mr() before ib_dma_unmap_sg()
smb: client: call ib_dma_unmap_sg if mr->sgt.nents is not 0
smb: client: improve logic in smbd_deregister_mr()
smb: client: improve logic in smbd_register_mr()
smb: client: improve logic in allocate_mr_list()
smb: client: let destroy_mr_list() remove locked from the list
smb: client: let destroy_mr_list() call list_del(&mr->list)
...
|
|
Now that smb3_crypto_shash_allocate() and smb311_crypto_shash_allocate()
are identical and only allocate "cmac(aes)", delete the latter and
replace the call to it with the former.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Now that the SMB client accesses MD5, HMAC-MD5, HMAC-SHA256, and SHA-512
only via the library API and not via crypto_shash, allocating
crypto_shash objects for these algorithms is no longer necessary.
Remove all these allocations, their corresponding kconfig selections,
and their corresponding module soft dependencies.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
For the HMAC-MD5 computations in NTLMv2, use the HMAC-MD5 library
instead of a "hmac(md5)" crypto_shash. This is simpler and faster.
With the library there's no need to allocate memory, no need to handle
errors, and the HMAC-MD5 code is accessed directly without inefficient
indirect calls and other unnecessary API overhead.
To preserve the existing behavior of NTLMv2 support being disabled when
the kernel is booted with "fips=1", make setup_ntlmv2_rsp() check
fips_enabled itself. Previously it relied on the error from
cifs_alloc_hash("hmac(md5)", &hmacmd5).
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Convert cifs_calc_signature() to use the MD5 library instead of a "md5"
crypto_shash. This is simpler and faster. With the library there's no
need to allocate memory, no need to handle errors, and the MD5 code is
accessed directly without inefficient indirect calls and other
unnecessary API overhead.
To preserve the existing behavior of MD5 signature support being
disabled when the kernel is booted with "fips=1", make
cifs_calc_signature() check fips_enabled itself. Previously it relied
on the error from cifs_alloc_hash("md5", &server->secmech.md5).
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Convert parse_mf_symlink() and format_mf_symlink() to use the MD5
library instead of a "md5" crypto_shash. This is simpler and faster.
With the library there's no need to allocate memory, no need to handle
errors, and the MD5 code is accessed directly without inefficient
indirect calls and other unnecessary API overhead.
This also fixes an issue where these functions did not work on kernels
booted in FIPS mode. The use of MD5 here is for data integrity rather
than a security purpose, so it can use a non-FIPS-approved algorithm.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Convert smb2_calc_signature() to use the HMAC-SHA256 library instead of
a "hmac(sha256)" crypto_shash. This is simpler and faster. With the
library there's no need to allocate memory, no need to handle errors,
and the HMAC-SHA256 code is accessed directly without inefficient
indirect calls and other unnecessary API overhead.
To make this possible, make __cifs_calc_signature() support both the
HMAC-SHA256 library and crypto_shash. (crypto_shash is still needed for
HMAC-MD5 and AES-CMAC. A later commit will switch HMAC-MD5 from shash
to the library. I'd like to eventually do the same for AES-CMAC, but it
doesn't have a library API yet. So for now, shash is still needed.)
Also remove the unnecessary 'sigptr' variable.
For now smb3_crypto_shash_allocate() still allocates a "hmac(sha256)"
crypto_shash. It will be removed in a later commit.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Convert generate_key() to use the HMAC-SHA256 library instead of a
"hmac(sha256)" crypto_shash. This is simpler and faster. With the
library there's no need to allocate memory, no need to handle errors,
and the HMAC-SHA256 code is accessed directly without inefficient
indirect calls and other unnecessary API overhead.
Also remove the unnecessary 'hashptr' variable.
For now smb3_crypto_shash_allocate() still allocates a "hmac(sha256)"
crypto_shash. It will be removed in a later commit.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Convert smb311_update_preauth_hash() to use the SHA-512 library instead
of a "sha512" crypto_shash. This is simpler and faster. With the
library there's no need to allocate memory, no need to handle errors,
and the SHA-512 code is accessed directly without inefficient indirect
calls and other unnecessary API overhead.
Remove the call to smb311_crypto_shash_allocate() from
smb311_update_preauth_hash(), since it appears to have been needed only
to allocate the "sha512" crypto_shash. (It also had the side effect of
allocating the "cmac(aes)" crypto_shash, but that's also done in
generate_key() which is where the AES-CMAC key is initialized.)
For now the "sha512" crypto_shash is still being allocated elsewhere.
It will be removed in a later commit.
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Malicious SMB server can send invalid reply to FSCTL_DFS_GET_REFERRALS
- reply smaller than sizeof(struct get_dfs_referral_rsp)
- reply with number of referrals smaller than NumberOfReferrals in the
header
Processing of such replies will cause oob.
Return -EINVAL error on such replies to prevent oob-s.
Signed-off-by: Eugene Korenevsky <ekorenevsky@aliyun.com>
Cc: stable@vger.kernel.org
Suggested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
Fix three refcount inconsistency issues related to `cifs_sb_tlink`.
Comments for `cifs_sb_tlink` state that `cifs_put_tlink()` needs to be
called after successful calls to `cifs_sb_tlink()`. Three calls fail to
update refcount accordingly, leading to possible resource leaks.
Fixes: 8ceb98437946 ("CIFS: Move rename to ops struct")
Fixes: 2f1afe25997f ("cifs: Use smb 2 - 3 and cifsacl mount options getacl functions")
Fixes: 366ed846df60 ("cifs: Use smb 2 - 3 and cifsacl mount options setacl function")
Cc: stable@vger.kernel.org
Signed-off-by: Shuhao Fu <sfual@cse.ust.hk>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
We should wait for the rdma_cm to become SMBDIRECT_SOCKET_DISCONNECTED,
it turns out that (at least running some xfstests e.g. cifs/001)
often triggers the case where wait_event_interruptible() returns
with -ERESTARTSYS instead of waiting for SMBDIRECT_SOCKET_DISCONNECTED
to be reached.
Or we are already in SMBDIRECT_SOCKET_DISCONNECTING and never wait
for SMBDIRECT_SOCKET_DISCONNECTED.
Fixes: 050b8c374019 ("smbd: Make upper layer decide when to destroy the transport")
Fixes: e8b3bfe9bc65 ("cifs: smbd: Don't destroy transport on RDMA disconnect")
Fixes: b0aa92a229ab ("smb: client: make sure smbd_disconnect_rdma_work() doesn't run after smbd_destroy() took over")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
In order to maintain the code more easily, move duplicate definitions to
new common header file.
Co-developed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: ZhangGuoDong <zhangguodong@kylinos.cn>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
If a smbdirect_mr_io structure if still visible to callers of
smbd_register_mr() we can't free the related memory when the
connection is disconnected! Otherwise smbd_deregister_mr()
will crash.
Now we use a mutex and refcounting in order to keep the
memory around if the connection is disconnected.
It means smbd_deregister_mr() can be called at any later time to free
the memory, which is no longer referenced by nor referencing the
connection.
It also means smbd_destroy() no longer needs to wait for
mr_io.used.count to become 0.
Fixes: 050b8c374019 ("smbd: Make upper layer decide when to destroy the transport")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
