linux-stable.git/net/sunrpc, branch v4.1.26

sunrpc: fix stripping of padded MIC tokens

2016-06-06T23:12:33+00:00

[ Upstream commit c0cb8bf3a8e4bd82e640862cdd8891400405cb89 ]

The length of the GSS MIC token need not be a multiple of four bytes.
It is then padded by XDR to a multiple of 4 B, but unwrap_integ_data()
would previously only trim mic.len + 4 B. The remaining up to three
bytes would then trigger a check in nfs4svc_decode_compoundargs(),
leading to a "garbage args" error and mount failure:

nfs4svc_decode_compoundargs: compound not properly padded!
nfsd: failed to decode arguments!

This would prevent older clients using the pre-RFC 4121 MIC format
(37-byte MIC including a 9-byte OID) from mounting exports from v3.9+
servers using krb5i.

The trimming was introduced by commit 4c190e2f913f ("sunrpc: trim off
trailing checksum before returning decrypted or integrity authenticated
buffer").

Fixes: 4c190e2f913f "unrpc: trim off trailing checksum..."
Signed-off-by: Tomáš Trnka 
Cc: stable@vger.kernel.org
Acked-by: Jeff Layton 
Signed-off-by: J. Bruce Fields 
Signed-off-by: Sasha Levin

sunrpc/cache: fix off-by-one in qword_get()

2016-03-07T21:30:48+00:00

[ Upstream commit b7052cd7bcf3c1478796e93e3dff2b44c9e82943 ]

The qword_get() function NUL-terminates its output buffer.  If the input
string is in hex format \xXXXX... and the same length as the output
buffer, there is an off-by-one:

  int qword_get(char **bpp, char *dest, int bufsize)
  {
      ...
      while (len < bufsize) {
          ...
          *dest++ = (h << 4) | l;
          len++;
      }
      ...
      *dest = '\0';
      return len;
  }

This patch ensures the NUL terminator doesn't fall outside the output
buffer.

Signed-off-by: Stefan Hajnoczi 
Cc: stable@vger.kernel.org
Signed-off-by: J. Bruce Fields 
Signed-off-by: Sasha Levin

svcrdma: handle rdma read with a non-zero initial page offset

2015-10-27T00:51:59+00:00

commit c91aed9896946721bb30705ea2904edb3725dd61 upstream.

The server rdma_read_chunk_lcl() and rdma_read_chunk_frmr() functions
were not taking into account the initial page_offset when determining
the rdma read length.  This resulted in a read who's starting address
and length exceeded the base/bounds of the frmr.

The server gets an async error from the rdma device and kills the
connection, and the client then reconnects and resends.  This repeats
indefinitely, and the application hangs.

Most work loads don't tickle this bug apparently, but one test hit it
every time: building the linux kernel on a 16 core node with 'make -j
16 O=/mnt/0' where /mnt/0 is a ramdisk mounted via NFSRDMA.

This bug seems to only be tripped with devices having small fastreg page
list depths.  I didn't see it with mlx4, for instance.

Fixes: 0bf4828983df ('svcrdma: refactor marshalling logic')
Signed-off-by: Steve Wise 
Tested-by: Chuck Lever 
Signed-off-by: J. Bruce Fields 
Signed-off-by: Greg Kroah-Hartman

svcrdma: Fix send_reply() scatter/gather set-up

2015-10-22T21:43:16+00:00

commit 9d11b51ce7c150a69e761e30518f294fc73d55ff upstream.

The Linux NFS server returns garbage in the data payload of inline
NFS/RDMA READ replies. These are READs of under 1000 bytes or so
where the client has not provided either a reply chunk or a write
list.

The NFS server delivers the data payload for an NFS READ reply to
the transport in an xdr_buf page list. If the NFS client did not
provide a reply chunk or a write list, send_reply() is supposed to
set up a separate sge for the page containing the READ data, and
another sge for XDR padding if needed, then post all of the sges via
a single SEND Work Request.

The problem is send_reply() does not advance through the xdr_buf
when setting up scatter/gather entries for SEND WR. It always calls
dma_map_xdr with xdr_off set to zero. When there's more than one
sge, dma_map_xdr() sets up the SEND sge's so they all point to the
xdr_buf's head.

The current Linux NFS/RDMA client always provides a reply chunk or
a write list when performing an NFS READ over RDMA. Therefore, it
does not exercise this particular case. The Linux server has never
had to use more than one extra sge for building RPC/RDMA replies
with a Linux client.

However, an NFS/RDMA client _is_ allowed to send small NFS READs
without setting up a write list or reply chunk. The NFS READ reply
fits entirely within the inline reply buffer in this case. This is
perhaps a more efficient way of performing NFS READs that the Linux
NFS/RDMA client may some day adopt.

Fixes: b432e6b3d9c1 ('svcrdma: Change DMA mapping logic to . . .')
BugLink: https://bugzilla.linux-nfs.org/show_bug.cgi?id=285
Signed-off-by: Chuck Lever 
Signed-off-by: J. Bruce Fields 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: Lock the transport layer on shutdown

2015-09-29T17:26:11+00:00

commit 79234c3db6842a3de03817211d891e0c2878f756 upstream.

Avoid all races with the connect/disconnect handlers by taking the
transport lock.

Reported-by:"Suzuki K. Poulose" 
Acked-by: Jeff Layton 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: Ensure that we wait for connections to complete before retrying

2015-09-29T17:26:10+00:00

commit 0fdea1e8a2853f79d39b8555cc9de16a7e0ab26f upstream.

Commit 718ba5b87343, moved the responsibility for unlocking the socket to
xs_tcp_setup_socket, meaning that the socket will be unlocked before we
know that it has finished trying to connect. The following patch is based on
an initial patch by Russell King to ensure that we delay clearing the
XPRT_CONNECTING flag until we either know that we failed to initiate
a connection attempt, or the connection attempt itself failed.

Fixes: 718ba5b87343 ("SUNRPC: Add helpers to prevent socket create from racing")
Reported-by: Russell King 
Reported-by: Russell King 
Tested-by: Russell King 
Tested-by: Benjamin Coddington 
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: xs_reset_transport must mark the connection as disconnected

2015-09-29T17:26:10+00:00

commit 0c78789e3a030615c6650fde89546cadf40ec2cc upstream.

In case the reconnection attempt fails.

Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: Fix a thinko in xs_connect()

2015-09-29T17:26:10+00:00

commit 99b1a4c32ad22024ac6198a4337aaec5ea23168f upstream.

It is rather pointless to test the value of transport->inet after
calling xs_reset_transport(), since it will always be zero, and
so we will never see any exponential back off behaviour.
Also don't force early connections for SOFTCONN tasks. If the server
disconnects us, we should respect the exponential backoff.

Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

SUNRPC: Fix a memory leak in the backchannel code

2015-08-03T16:29:17+00:00

commit 88de6af24f2b48b06c514d3c3d0a8f22fafe30bd upstream.

req->rq_private_buf isn't initialised when xprt_setup_backchannel calls
xprt_free_allocation.

Fixes: fb7a0b9addbdb ("nfs41: New backchannel helper routines")
Signed-off-by: Trond Myklebust 
Signed-off-by: Greg Kroah-Hartman

svcrpc: fix potential GSSX_ACCEPT_SEC_CONTEXT decoding failures

2015-05-04T16:02:40+00:00

In an environment where the KDC is running Active Directory, the
exported composite name field returned in the context could be large
enough to span a page boundary.  Attaching a scratch buffer to the
decoding xdr_stream helps deal with those cases.

The case where we saw this was actually due to behavior that's been
fixed in newer gss-proxy versions, but we're fixing it here too.

Signed-off-by: Scott Mayhew 
Cc: stable@vger.kernel.org
Reviewed-by: Simo Sorce 
Signed-off-by: J. Bruce Fields