Merge branch 'strparser'

Tom Herbert says:

====================
strp: Stream parser for messages

This patch set introduces a utility for parsing application layer
protocol messages in a TCP stream. This is a generalization of the
mechanism implemented of Kernel Connection Multiplexor.

This patch set adapts KCM to use the strparser. We expect that kTLS
can use this mechanism also. RDS would probably be another candidate
to use a common stream parsing mechanism.

The API includes a context structure, a set of callbacks, utility
functions, and a data ready function. The callbacks include
a parse_msg function that is called to perform parsing (e.g.
BPF parsing in case of KCM), and a rcv_msg function that is called
when a full message has been completed.

For strparser we specify the return codes from the parser to allow
the backend to indicate that control of the socket should be
transferred back to userspace to handle some exceptions in the
stream: The return values are:

      >0 : indicates length of successfully parsed message
       0  : indicates more data must be received to parse the message
       -ESTRPIPE : current message should not be processed by the
          kernel, return control of the socket to userspace which
          can proceed to read the messages itself
       other < 0 : Error is parsing, give control back to userspace
          assuming that synchronization is lost and the stream
          is unrecoverable (application expected to close TCP socket)

There is one issue I haven't been able to fully resolve. If parse_msg
returns ESTRPIPE (wants control back to userspace) the parser may
already have consumed some bytes of the message. There is no way to
put bytes back into the TCP receive queue and tcp_read_sock does not
allow an easy way to peek messages. In lieu of a better solution, we
return ENODATA on the socket to indicate that the data stream is
unrecoverable (application needs to close socket). This condition
should only happen if an application layer message header is split
across two skbuffs and parsing just the first skbuff wasn't sufficient
to determine the that transfer to userspace is needed.

This patch set contains:

  - strparser implementation
  - changes to kcm to use strparser
  - strparser.txt documentation

v2:
  - Add copyright notice to C files
  - Remove GPL module license from strparser.c
  - Add report of rxpause

v3:
  - Restore GPL module license
  - Use EXPORT_SYMBOL_GPL

v4:
  - Removed unused function, changed another to be static as suggested
    by davem
  - Rewoked data_ready to be called from upper layer, no longer requires
    taking over socket data_ready callback as suggested by Lance Chao

Tested:
  - Ran a KCM thrash test for 24 hours. No behavioral or performance
    differences observed.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 136 insertions, 0 deletions

diff --git a/Documentation/networking/strparser.txt b/Documentation/networking/strparser.txt
new file mode 100644
index 000000000000..a0bf573dfa61
--- /dev/null
+++ b/Documentation/networking/strparser.txt
@@ -0,0 +1,136 @@
+Stream Parser
+-------------
+
+The stream parser (strparser) is a utility that parses messages of an
+application layer protocol running over a TCP connection. The stream
+parser works in conjunction with an upper layer in the kernel to provide
+kernel support for application layer messages. For instance, Kernel
+Connection Multiplexor (KCM) uses the Stream Parser to parse messages
+using a BPF program.
+
+Interface
+---------
+
+The API includes a context structure, a set of callbacks, utility
+functions, and a data_ready function. The callbacks include
+a parse_msg function that is called to perform parsing (e.g.
+BPF parsing in case of KCM), and a rcv_msg function that is called
+when a full message has been completed.
+
+A stream parser can be instantiated for a TCP connection. This is done
+by:
+
+strp_init(struct strparser *strp, struct sock *csk,
+	  struct strp_callbacks *cb)
+
+strp is a struct of type strparser that is allocated by the upper layer.
+csk is the TCP socket associated with the stream parser. Callbacks are
+called by the stream parser.
+
+Callbacks
+---------
+
+There are four callbacks:
+
+int (*parse_msg)(struct strparser *strp, struct sk_buff *skb);
+
+    parse_msg is called to determine the length of the next message
+    in the stream. The upper layer must implement this function. It
+    should parse the sk_buff as containing the headers for the
+    next application layer messages in the stream.
+
+    The skb->cb in the input skb is a struct strp_rx_msg. Only
+    the offset field is relevant in parse_msg and gives the offset
+    where the message starts in the skb.
+
+    The return values of this function are:
+
+    >0 : indicates length of successfully parsed message
+    0  : indicates more data must be received to parse the message
+    -ESTRPIPE : current message should not be processed by the
+          kernel, return control of the socket to userspace which
+          can proceed to read the messages itself
+    other < 0 : Error is parsing, give control back to userspace
+          assuming that synchronization is lost and the stream
+          is unrecoverable (application expected to close TCP socket)
+
+    In the case that an error is returned (return value is less than
+    zero) the stream parser will set the error on TCP socket and wake
+    it up. If parse_msg returned -ESTRPIPE and the stream parser had
+    previously read some bytes for the current message, then the error
+    set on the attached socket is ENODATA since the stream is
+    unrecoverable in that case.
+
+void (*rcv_msg)(struct strparser *strp, struct sk_buff *skb);
+
+    rcv_msg is called when a full message has been received and
+    is queued. The callee must consume the sk_buff; it can
+    call strp_pause to prevent any further messages from being
+    received in rcv_msg (see strp_pause below). This callback
+    must be set.
+
+    The skb->cb in the input skb is a struct strp_rx_msg. This
+    struct contains two fields: offset and full_len. Offset is
+    where the message starts in the skb, and full_len is the
+    the length of the message. skb->len - offset may be greater
+    then full_len since strparser does not trim the skb.
+
+int (*read_sock_done)(struct strparser *strp, int err);
+
+     read_sock_done is called when the stream parser is done reading
+     the TCP socket. The stream parser may read multiple messages
+     in a loop and this function allows cleanup to occur when existing
+     the loop. If the callback is not set (NULL in strp_init) a
+     default function is used.
+
+void (*abort_parser)(struct strparser *strp, int err);
+
+     This function is called when stream parser encounters an error
+     in parsing. The default function stops the stream parser for the
+     TCP socket and sets the error in the socket. The default function
+     can be changed by setting the callback to non-NULL in strp_init.
+
+Functions
+---------
+
+The upper layer calls strp_tcp_data_ready when data is ready on the lower
+socket for strparser to process. This should be called from a data_ready
+callback that is set on the socket.
+
+strp_stop is called to completely stop stream parser operations. This
+is called internally when the stream parser encounters an error, and
+it is called from the upper layer when unattaching a TCP socket.
+
+strp_done is called to unattach the stream parser from the TCP socket.
+This must be called after the stream processor has be stopped.
+
+strp_check_rcv is called to check for new messages on the socket. This
+is normally called at initialization of the a stream parser instance
+of after strp_unpause.
+
+Statistics
+----------
+
+Various counters are kept for each stream parser for a TCP socket.
+These are in the strp_stats structure. strp_aggr_stats is a convenience
+structure for accumulating statistics for multiple stream parser
+instances. save_strp_stats and aggregate_strp_stats are helper functions
+to save and aggregate statistics.
+
+Message assembly limits
+-----------------------
+
+The stream parser provide mechanisms to limit the resources consumed by
+message assembly.
+
+A timer is set when assembly starts for a new message. The message
+timeout is taken from rcvtime for the associated TCP socket. If the
+timer fires before assembly completes the stream parser is aborted
+and the ETIMEDOUT error is set on the TCP socket.
+
+Message length is limited to the receive buffer size of the associated
+TCP socket. If the length returned by parse_msg is greater than
+the socket buffer size then the stream parser is aborted with
+EMSGSIZE error set on the TCP socket. Note that this makes the
+maximum size of receive skbuffs for a socket with a stream parser
+to be 2*sk_rcvbuf of the TCP socket.