It is not necessary to deregister a memory registration after it has been
successfully invalidated.
Signed-off-by: Long Li <longli@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
When issuing SMB1 read/write, pass the page offset to transport.
Signed-off-by: Long Li <longli@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
In some error conditions, resp_buftype can be passed uninitialised to
free_rsp_buf(), potentially resulting in a spurious debug message.
If resp_buftype randomly had the value 1 (CIFS_SMALL_BUFFER) then this
would log a debug message.
The rsp pointer is initialised to NULL so there is no other side-effect.
Detected by CoverityScan, CID 1438585 ("Uninitialized scalar variable")
Detected by CoverityScan, CID 1438667 ("Uninitialized scalar variable")
Detected by CoverityScan, CID 1438764 ("Uninitialized scalar variable")
Signed-off-by: Garry McNulty <garrmcnu@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Be able to log a ftrace message on success and/or failure of
sending a lease break response to the server.
Example output:
TASK-PID CPU# |||| TIMESTAMP FUNCTION
| | | |||| | |
kworker/1:1-5681 [001] .... 11123.530457: smb3_lease_done: sid=0x291e3e0f tid=0x8ba43071 lease_key=0x1852ca0d3ecd9b55847750a86716fde lease_state=0x0
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
In network file system it is fairly easy for server and client
atime vs. mtime to get confused (and atime updated less frequently)
which we noticed broke some apps which expect atime >= mtime
Also ignore relatime mount option (rather than error on it) since
relatime is basically what some network server fs are doing
(relatime).
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Modern servers often support 8MB as maximum i/o size, and we see some
performance benefits (my testing showed 1 to 13% on write paths,
and 1 to 3% on read paths for increasing the default to 4MB). If server
doesn't support larger i/o size, during negotiate protocol it is already
set correctly to the server's maximum if lower than 4MB.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
As we reset credits later in the reconnect path, useful
to have optional (cifsFYI) debug message.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
tcon->Flags is only used by SMB1 code and changing it is not permanent
(you lose the setting on tcon reconnect).
* Move the setting to superblock flags (per mount-points).
* Make automount callback exit early when flag present
* Make dfs resolving happening in mount syscall exit early if flag present
Signed-off-by: Aurelien Aptel <aaptel@suse.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Acked-by: Pavel Shilovsky <pshilov@microsoft.com>
Each time we reconnect to the same server, bump an instance
counter (and display in /proc/fs/cifs/DebugData) to make it
easier to debug.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
Previously reserved dpen response field changed in smb3
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
/proc/fs/cifs/Stats when CONFIG_CIFS_STATS2 is enabled logs 'slow'
responses, but depending on the server you are debugging a
one second timeout may be too fast, so allow setting it to
a larger number of seconds via new module parameter
/sys/module/cifs/parameters/slow_rsp_threshold
or via modprobe:
slow_rsp_threshold:Amount of time (in seconds) to wait before
logging that a response is delayed.
Default: 1 (if set to 0 disables msg). (uint)
Recommended values are 0 (disabled) to 32767 (9 hours) with
the default remaining as 1 second.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
note smb3 (and common more modern servers) in the module description
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
For a network file system we generally prefer large i/o, but
if the server returns invalid file system block/sector sizes
in cifs (vers=1.0) QFSInfo then set block size to a default
of a reasonable minimum (4K).
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Currently, "echo 0 > /proc/fs/cifs/Stats" resets all of the stats
except the session and share reconnect counts. Fix it to
reset those as well.
CC: Stable <stable@vger.kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
When "backup intent" is requested on the mount (e.g. backupuid or
backupgid mount options), the corresponding flag was missing from
some of the new compounding operations as well (now that
open_query_close is gone).
Related to kernel bugzilla #200953
Reported-and-tested-by: <whh@rubrik.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
So we don't overflow the io vector arrays accidentally
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Get rid of smb2_open_op_close() as all operations are now migrated
to smb2_compound_op().
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
We never pass is_falloc==true here anyway and if we ever need to support
is_falloc in the future, SMB2_set_eof is such a trivial wrapper around
send_set_info() that we can/should just create a differently named wrapper
for that new functionality.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Cuts number of network roundtrips significantly for some common syscalls
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
This changes SMB2_OP_SET_EOF to use compounding in some situations.
This is part of the path based API to truncate a file.
Most of the time this will however not be invoked for SMB2 since
cifs_set_file_size() will as far as I can tell almost always just
open the file synchronously and switch to the handle based truncate
code path, thus bypassing the compounding we add here.
Rewriting cifs_set_file_size() and make that whole pile of code more
compounding friendly, and also easier to read and understand, is a
different project though and not for this patch.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
This and previous patches drop the number of roundtrips we need for rmdir()
from 6 to 2.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
so that we can use these later for compounded set-info calls.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
This,and previous patches, drops the number of roundtrips from five to two
for unlink()
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
This with the previous patch changes mkdir() from needing 6 roundtrips to
just 3.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
This turns most open/query-info/close patterns in cifs.ko
to become compounds.
This changes stat from using 3 roundtrips to just a single one.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
When processing the mids for compounds we would only add credits based on
the last successful mid in the compound which would leak credits and
eventually triggering a re-connect.
Fix this by splitting the mid processing part into two loops instead of one
where the first loop just waits for all mids and then counts how many
credits we were granted for the whole compound.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Add tracepoint to catch potential cases where a pending operation overlapping a
reconnect could fail and incorrectly refund its credits causing the client
to think it has more credits available than the server thinks it does.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Fixes gcc '-Wunused-but-set-variable' warning:
fs/cifs/ioctl.c: In function 'cifs_ioctl':
fs/cifs/ioctl.c:164:23: warning:
variable 'cifs_sb' set but not used [-Wunused-but-set-variable]
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Use kmemdup rather than duplicating its implementation
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Some servers (e.g. Azure) return "STATUS_NOT_IMPLEMENTED" rather
than "STATUS_INVALID_DEVICE_REQUEST" on query network interface
info at mount. This shouldn't cause us to log a warning message
automatically. Don't log this unless noisier cifsFYI is enabled.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Fixes problem (discovered by Aurelien) introduced by recent commit:
commit b24df3e30c
("cifs: update receive_encrypted_standard to handle compounded responses")
which broke the ability to respond to some lease breaks
(lease breaks being ignored is a problem since can block
server response for duration of the lease break timeout).
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
For compounded PDUs we whould only wake the waiting thread for the
very last PDU of the compound.
We do this so that we are guaranteed that the demultiplex_thread will
not process or access any of those MIDs any more once the send/recv
thread starts processing.
Else there is a race where at the end of the send/recv processing we
will try to delete all the mids of the compound. If the multiplex
thread still has other mids to process at this point for this compound
this can lead to an oops.
Needed to fix recent commit:
commit 730928c8f4
("cifs: update smb2_queryfs() to use compounding")
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
cifs_delete_mid() is called once we are finished handling a mid and we
expect no more work done on this mid.
Needed to fix recent commit:
commit 730928c8f4
("cifs: update smb2_queryfs() to use compounding")
Add a warning if someone tries to dequeue a mid that has already been
flagged to be deleted.
Also change list_del() to list_del_init() so that if we have similar bugs
resurface in the future we will not oops.
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
When mounting a Windows share that is the root of a drive (eg. C$)
the server does not return . and .. directory entries. This results in
the smb2 code path erroneously skipping the 2 first entries.
Pseudo-code of the readdir() code path:
cifs_readdir(struct file, struct dir_context)
initiate_cifs_search <-- if no reponse cached yet
server->ops->query_dir_first
dir_emit_dots
dir_emit <-- adds "." and ".." if we're at pos=0
find_cifs_entry
initiate_cifs_search <-- if pos < start of current response
(restart search)
server->ops->query_dir_next <-- if pos > end of current response
(fetch next search res)
for(...) <-- loops over cur response entries
starting at pos
cifs_filldir <-- skip . and .., emit entry
cifs_fill_dirent
dir_emit
pos++
A) dir_emit_dots() always adds . & ..
and sets the current dir pos to 2 (0 and 1 are done).
Therefore we always want the index_to_find to be 2 regardless of if
the response has . and ..
B) smb1 code initializes index_of_last_entry with a +2 offset
in cifssmb.c CIFSFindFirst():
psrch_inf->index_of_last_entry = 2 /* skip . and .. */ +
psrch_inf->entries_in_buffer;
Later in find_cifs_entry() we want to find the next dir entry at pos=2
as a result of (A)
first_entry_in_buffer = cfile->srch_inf.index_of_last_entry -
cfile->srch_inf.entries_in_buffer;
This var is the dir pos that the first entry in the buffer will
have therefore it must be 2 in the first call.
If we don't offset index_of_last_entry by 2 (like in (B)),
first_entry_in_buffer=0 but we were instructed to get pos=2 so this
code in find_cifs_entry() skips the 2 first which is ok for non-root
shares, as it skips . and .. from the response but is not ok for root
shares where the 2 first are actual files
pos_in_buf = index_to_find - first_entry_in_buffer;
// pos_in_buf=2
// we skip 2 first response entries :(
for (i = 0; (i < (pos_in_buf)) && (cur_ent != NULL); i++) {
/* go entry by entry figuring out which is first */
cur_ent = nxt_dir_entry(cur_ent, end_of_smb,
cfile->srch_inf.info_level);
}
C) cifs_filldir() skips . and .. so we can safely ignore them for now.
Sample program:
int main(int argc, char **argv)
{
const char *path = argc >= 2 ? argv[1] : ".";
DIR *dh;
struct dirent *de;
printf("listing path <%s>\n", path);
dh = opendir(path);
if (!dh) {
printf("opendir error %d\n", errno);
return 1;
}
while (1) {
de = readdir(dh);
if (!de) {
if (errno) {
printf("readdir error %d\n", errno);
return 1;
}
printf("end of listing\n");
break;
}
printf("off=%lu <%s>\n", de->d_off, de->d_name);
}
return 0;
}
Before the fix with SMB1 on root shares:
<.> off=1
<..> off=2
<$Recycle.Bin> off=3
<bootmgr> off=4
and on non-root shares:
<.> off=1
<..> off=4 <-- after adding .., the offsets jumps to +2 because
<2536> off=5 we skipped . and .. from response buffer (C)
<411> off=6 but still incremented pos
<file> off=7
<fsx> off=8
Therefore the fix for smb2 is to mimic smb1 behaviour and offset the
index_of_last_entry by 2.
Test results comparing smb1 and smb2 before/after the fix on root
share, non-root shares and on large directories (ie. multi-response
dir listing):
PRE FIX
=======
pre-1-root VS pre-2-root:
ERR pre-2-root is missing [bootmgr, $Recycle.Bin]
pre-1-nonroot VS pre-2-nonroot:
OK~ same files, same order, different offsets
pre-1-nonroot-large VS pre-2-nonroot-large:
OK~ same files, same order, different offsets
POST FIX
========
post-1-root VS post-2-root:
OK same files, same order, same offsets
post-1-nonroot VS post-2-nonroot:
OK same files, same order, same offsets
post-1-nonroot-large VS post-2-nonroot-large:
OK same files, same order, same offsets
REGRESSION?
===========
pre-1-root VS post-1-root:
OK same files, same order, same offsets
pre-1-nonroot VS post-1-nonroot:
OK same files, same order, same offsets
BugLink: https://bugzilla.samba.org/show_bug.cgi?id=13107
Signed-off-by: Aurelien Aptel <aaptel@suse.com>
Signed-off-by: Paulo Alcantara <palcantara@suse.deR>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Stable <stable@vger.kernel.org>
We need to verify that the "data_offset" is within bounds.
Reported-by: Dr Silvio Cesare of InfoSect <silvio.cesare@gmail.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
The "le32_to_cpu(rsp->OutputOffset) + *plen" addition can overflow and
wrap around to a smaller value which looks like it would lead to an
information leak.
Fixes: 4a72dafa19 ("SMB2 FSCTL and IOCTL worker function")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
CC: Stable <stable@vger.kernel.org>
The problem is that "entryptr + next_offset" and "entryptr + len + size"
can wrap. I ended up changing the type of "entryptr" because it makes
the math easier when we don't have to do so much casting.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
CC: Stable <stable@vger.kernel.org>
The "old_entry + le32_to_cpu(pDirInfo->NextEntryOffset)" can wrap
around so I have added a check for integer overflow.
Reported-by: Dr Silvio Cesare of InfoSect <silvio.cesare@gmail.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Stable <stable@vger.kernel.org>
This got lost in commit 0fdfef9aa7,
which removed CONFIG_CIFS_SMB311.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Fixes: 0fdfef9aa7 ("smb3: simplify code by removing CONFIG_CIFS_SMB311")
CC: Stable <stable@vger.kernel.org>
CC: linux-cifs@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
A powerpc build of cifs with gcc v8.2.0 produces this warning:
fs/cifs/cifssmb.c: In function ‘CIFSSMBNegotiate’:
fs/cifs/cifssmb.c:605:3: warning: ‘strncpy’ writing 16 bytes into a region of size 1 overflows the destination [-Wstringop-overflow=]
strncpy(pSMB->DialectsArray+count, protocols[i].name, 16);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Since we are already doing a strlen() on the source, change the strncpy
to a memcpy().
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Steve French <stfrench@microsoft.com>
This patch is required allows access to a Microsoft fileserver failover
cluster behind a 1:1 NAT firewall.
The change also provides stronger context for authentication and share
connection (see MS-SMB2 3.3.5.7 and MS-SRVS 3.1.6.8) as noted by
Tom Talpey, and addresses comments about the buffer size for the UNC
made by Aurélien Aptel.
Signed-off-by: Thomas Werschlein <thomas.werschlein@geo.uzh.ch>
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Tom Talpey <ttalpey@microsoft.com>
Reviewed-by: Aurelien Aptel <aaptel@suse.com>
CC: Stable <stable@vger.kernel.org>
Although servers will typically ignore unsupported features,
we should advertise the support for directory leases (as
Windows e.g. does) in the negotiate protocol capabilities we
pass to the server, and should check for the server capability
(CAP_DIRECTORY_LEASING) before sending a lease request for an
open of a directory. This will prevent us from accidentally
sending directory leases to SMB2.1 or SMB2 server for example.
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Stable <stable@vger.kernel.org>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
I ran into some cases where server was returning the wrong length
on frames but I couldn't easily match them to the command in the
network trace (or server logs) since I need the command and/or
multiplex id to find the offending SMB2/SMB3 command. Add these
two fields to the log message. In the case of padding too much
it may not be a problem in all cases but might have correlated
to a network disconnect case in some problems we have been
looking at. In the case of frame too short is even more important.
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Ronnie Sahlberg <lsahlber@redhat.com>
When "backup intent" is requested on the mount (e.g. backupuid or
backupgid mount options), the corresponding flag needs to be set
on opens of directories (and files) but was missing in some
places causing access denied trying to enumerate and backup
servers.
Fixes kernel bugzilla #200953https://bugzilla.kernel.org/show_bug.cgi?id=200953
Reported-and-tested-by: <whh@rubrik.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Stable <stable@vger.kernel.org>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
When a Mac client saves an item containing a backslash to a file server
the backslash is represented in the CIFS/SMB protocol as as U+F026.
Before this change, listing a directory containing an item with a
backslash in its name will return that item with the backslash
represented with a true backslash character (U+005C) because
convert_sfm_character mapped U+F026 to U+005C when interpretting the
CIFS/SMB protocol response. However, attempting to open or stat the
path using a true backslash will result in an error because
convert_to_sfm_char does not map U+005C back to U+F026 causing the
CIFS/SMB request to be made with the backslash represented as U+005C.
This change simply prevents the U+F026 to U+005C conversion from
happenning. This is analogous to how the code does not do any
translation of UNI_SLASH (U+F000).
Signed-off-by: Jon Kuhn <jkuhn@barracuda.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
The kmalloc was not being checked - if it fails issue a warning
and return -ENOMEM to the caller.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Fixes: b8da344b74 ("cifs: dynamic allocation of ntlmssp blob")
Signed-off-by: Steve French <stfrench@microsoft.com>
Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com>
cc: Stable <stable@vger.kernel.org>`
Some SMB2/3 servers, Win2016 but possibly others too, adds padding
not only between PDUs in a compound but also to the final PDU.
This padding extends the PDU to a multiple of 8 bytes.
Check if the unexpected length looks like this might be the case
and avoid triggering the log messages for :
"SMB2 server sent bad RFC1001 len %d not %d\n"
Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
rdma.git merge resolution for the 4.19 merge window
Conflicts:
drivers/infiniband/core/rdma_core.c
- Use the rdma code and revise with the new spelling for
atomic_fetch_add_unless
drivers/nvme/host/rdma.c
- Replace max_sge with max_send_sge in new blk code
drivers/nvme/target/rdma.c
- Use the blk code and revise to use NULL for ib_post_recv when
appropriate
- Replace max_sge with max_recv_sge in new blk code
net/rds/ib_send.c
- Use the net code and revise to use NULL for ib_post_recv when
appropriate
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
