Ignore:
Timestamp:
Oct 10, 2020, 4:56:11 PM (4 years ago)
Author:
alain
Message:

Introduce support for both TCP in the dev_nic.h & dev_nic.c files.

File:
1 edited

Legend:

Unmodified
Added
Removed

  • trunk/kernel/devices/dev_nic.h

    r657 r663  
    5656 * In this case, the channel index is defined by an hash function computed from the remote
    5757 * IP address and port. This index is computed by the hardware for an RX packet, and is
    58  * computed by the kernel for a TX packet, using a specific driver function. TODO ...
     58 * computed by the kernel for a TX packet, using a specific driver function.
    5959 * The 2*N chdevs, and the associated server threads implementing the protocols stack,
    6060 * are distributed in 2*N different clusters.
    … …  
    6363 *
    6464 * On the user side, ALMOS-MKH implements the POSIX socket API.
    65  * The kernel functions implementing the socket related syscalls are :
    66  * - dev_nic_socket()    : create a local socket registered in process fd_array[].
    67  * - dev_nic_bind()      : attach a local IP address and port to a local socket.
    68  * - dev_nic_listen()    : local server makes a passive open.
    69  * - dev_nic_connect()   : local client makes an active open to a remote server.
    70  * - dev_nic_accept()    : local server accept a new remote client.
    71  * - dev_nic_send()      : send data on a connected socket.
    72  * - dev_nic_recv()      : receive data on a connected socket.
    73  * - dev_nic_sendto()    : send a packet to a remote (IP address/port).
    74  * - dev_nic_recvfrom()  : receive a paket from a remote (IP address/port).
    75  * - dev_nic_close()     : close a socket
    76  *
    77  * 3) TX stream
    78  *
    79  * The internal API between the client threads and the TX server thread defines
    80  * the 3 following commands:
    81  *   . SOCKET_TX_CONNECT : request to execute the 3 steps TCP connection handshake.
    82  *   . SOCKET_TX_SEND    : send data to a remote socket (UDP or TCP).
    83  *   . SOCKET_TX_CLOSE   : request to execute the 3 steps TCP close handshake.
    84  *
    85  * - These 3 commands are blocking for the client thread that registers the command in the
    86  *   socket descriptor, blocks on the BLOCKED_IO condition, and deschedules.
    87  * - The TX server thread is acting as a multiplexer. It scans the list of attached sockets,
    88  *   to handle all valid commands: one UDP packet or TCP segment per iteration.
    89  *   It uses the user buffer defined by the client thread, and attached to socket descriptor,
    90  *   as a retransmission buffer. It blocks and deschedules on the BLOCKED_CLIENT condition,
    91  *   when there is no more active TX command registered in any socket. It is re-activated
    92  *   by the first client thread registering a new TX command in the socket descriptor.
    93  *   It unblocks a client thread only when a command is fully completed. It signals errors
    94  *   to the client thread using the tx_error field in socket descriptor.
    95  *
    96  * 4) RX stream
    97  *
    98  * The communication between the RX server thread and the client threads expecting data
    99  * is done through receive buffers (one private buffer per socket) that are handled
    100  * as single-writer / single reader-FIFOs, called rx_buf.
    101  * - The RX server thread is acting as a demultiplexor: it handle one TCP segment or UDP
    102  *   packet per iteration, and register the data in the rx_buf of the socket matching
    103  *   the packet. It simply discard all packets that does not match a registered socket.
    104  *   When a client thread is registered in the socket descriptor, the RX server thread
    105  *   unblocks this client thread as soon as there is data available in rx_buf.
    106  *   It blocks and deschedules on the BLOCKED_ISR condition when there is no more packets
    107  *   in the NIC_RX queue. It is unblocked by the hardware ISR.
    108  * - The client thread simply access the rx_buf attached to socket descriptor, and consumes
    109  *   the available data when the rx_buf is non empty. It blocks on the BLOCKED_IO condition,
    110  *   and deschedules when the rx_buf is empty.
    111  *
    112  * 5) R2T queue
    113  *
    114  * To implement the TCP "3 steps handshake" protocol, the RX server thread can directly
    115  * request the associated TX server thread to send control packets in  the TX stream,
    116  * using a dedicate R2T (RX to TX) FIFO stored in the socket descriptor.
    117  *
    118  * 6) NIC driver API
    119  *
    120  * The generic NIC device "driver" API defines the following commands to the NIC driver:
    121  * - READABLE : returns true if at least one RX paquet is available in RX queue.
    122  * - WRITABLE : returns true if at least one empty slot is available in TX queue.
    123  * - READ     : consume one packet from the RX queue.
    124  * - WRITE    : produce one packet to the TX queue.
     65 * The following kernel functions implement the socket related syscalls :
     66 * - socket_build()     : create a local socket registered in process fd_array[].
     67 * - socket_bind()      : attach a local IP address and port to a local socket.
     68 * - socket_listen()    : local server makes a passive open.
     69 * - socket_connect()   : local client makes an active open to a remote server.
     70 * - socket_accept()    : local server accept a new remote client.
     71 * - socket_send()      : send data on a connected socket.
     72 * - socket_recv()      : receive data on a connected socket.
     73 * - socket_sendto()    : send a packet to a remote (IP address/port).
     74 * - socket_recvfrom()  : receive a paket from a remote (IP address/port).
     75 *
     76 * 3) NIC TX and NIC_RX server threads
     77 *
     78 * The dev_nic_tx_server() & dev_nic_rx_server() functions defined below execute
     79 * the user commands stored in the sockets to implement the [ETH / IP / TCP or UDP]
     80 * protocols stack, as defined in the <ksocket.c> and <ksocket.h> files.
     81 *
     82 * 4) NIC driver API
     83 *
     84 * The generic NIC device "driver" API defines the following commands, used by the
     85 * NIC_TX and NIC_RX server threads, running in the cluster containing the relevant chdev,
     86 * to access the NIC_TX and NIC_RX packets queues:
     87 *
     88 * - READ     : consume one packet from the NIC_RX queue.
     89 * - WRITE    : produce one packet to the NIC_TX queue.
     90 *
    12591 * All RX or TX paquets are sent or received in standard 2 Kbytes kernel buffers,
    12692 * that are dynamically allocated by the protocols stack.
    127  *
    12893 * The actual TX an RX queues structures depends on the hardware NIC implementation,
    12994 * and are defined in the HAL specific driver code.
     95 *
     96 * Moreover, the generic NIC device "driver" API defines the following commands,
     97 * used directly by a client thread running in any cluster, to access the NIC device
     98 * configuration or status registers:
     99 *
     100 * - GET_KEY      : get channel index from remote IP address and port
     101 * - SET_RUN      : activate/desactivate one channel
     102 * - GET_INSTRU   : get one instrumentation counter value
     103 * - CLEAR_INSTRU : reset all instrumentation counters
    130104 *
    131105 * WARNING: the WTI mailboxes used by the driver ro receive events from the hardware
    … …  
    140114struct chdev_s;
    141115
    142 /******************************************************************************************
    143  *   Various constants used by the Protocols stack
    144  *****************************************************************************************/
    145 
    146 #define SRC_MAC_54             0x54
    147 #define SRC_MAC_32             0x32
    148 #define SRC_MAC_10             0x10
    149 #define DST_MAC_54             0x54
    150 #define DST_MAC_32             0x32
    151 #define DST_MAC_10             0x10
     116/*****************************************************************************************
     117 *   Various constants used by the protocols stack
     118 ****************************************************************************************/
     119
     120#define SRC_MAC_5              0x66         // This is a temporary short-cut for debug
     121#define SRC_MAC_4              0x55
     122#define SRC_MAC_3              0x44
     123#define SRC_MAC_2              0x33
     124#define SRC_MAC_1              0x22
     125#define SRC_MAC_0              0x11
     126
     127#define DST_MAC_5              0x66         // This is a temporary short-cut for debug
     128#define DST_MAC_4              0x55
     129#define DST_MAC_3              0x44
     130#define DST_MAC_2              0x33
     131#define DST_MAC_1              0x22
     132#define DST_MAC_0              0x11
    152133
    153134#define TCP_HEAD_LEN           20
    … …  
    159140#define PROTOCOL_TCP           0x06
    160141
    161 #define TCP_ISS                0x10000
    162 
    163 #define PAYLOAD_MAX_LEN        1500         // max payload for and UDP packet or a TCP segment
     142#define TCP_ISS_CLIENT         0x10000      // initial sequence number for TCP client
     143#define TCP_ISS_SERVER         0x20000      // initial sequence number for TCP server
     144#define TCP_MAX_WINDOW         0xFFFFF      // initial TCP send window
     145
     146#define PAYLOAD_MAX_LEN        1500         // max length for an UDP packet / TCP segment
    164147
    165148#define TCP_FLAG_FIN           0x01
    … …  
    172155#define NIC_RX_BUF_SIZE        0x100000     // 1 Mbytes
    173156#define NIC_R2T_QUEUE_SIZE     0x64         // smallest KCM size
    174 #define NIC_CRQ_QUEUE_SIZE     0x8          // 8 * sizeof(sockaddr_t) = smallest KCM size
    175 #define NIC_PKT_MAX_SIZE       1500         // for Ethernet
    176 #define NIC_KERNEL_BUF_SIZE    2000         // for on ETH/IP/TCP packet
     157#define NIC_CRQ_QUEUE_SIZE     0x8          // actual size is 8 * sizeof(sockaddr_t)
     158#define NIC_KERNEL_BUF_SIZE    0x800        // 2 Kbytes for one ETH/IP/TCP packet
    177159
    178160/*****************************************************************************************
    179  * This defines the extension for the generic NIC device.
    180  * The actual queue descriptor depends on the implementation.
    181  *
    182  * WARNING : for all NIC_TX and NIC_RX chdevs, the xlist rooted in in the chdev
    183  *           ("wait_root" and "wait_lock" fields) is actually a list of sockets.
     161 * This structure defines the specific chdev extension for NIC device:
     162 * - queue : local pointer on the memory mapped queue of TX or RX packets, used
     163 *   by the NIC driver to move packets to/from the NIC hardware. The actual descriptor
     164 *   depends on the NIC implementation.
     165 * - root  : root of an xlist of sockets that are in the LISTEN state, waiting one or
     166 *   several TCP connection requests from remote processes. It is only used by the
     167 *   NIC_RX server thread attached to a NIC_RX chdev.
     168 * - lock  : lock protecting concurrent access to the litening sockets list.
    184169 ****************************************************************************************/
    185170
    186171typedef struct nic_extend_s
    187172{
    188     void         * queue;       /*! local pointer on NIC queue descriptor (RX or TX)    */
     173    void            * queue;    /*! pointer on NIC packets queue descriptor (RX or TX)  */
     174    xlist_entry_t     root;     /*! root of listening sockets list (only used in RX[0]) */
     175    remote_busylock_t lock;     /*! lock protecting this list  (only used in RX[0]      */
    189176}
    190177nic_extend_t;
    … …  
    204191/****************************************************************************************
    205192 * This defines the (implementation independant) commands to access the NIC hardware.
    206  * These commands are registered by the NIC_TX and NIC_RX server threads in the
    207  * server thread descriptor, to be used by the NIC driver.
    208  * The buffer is always a 2K bytes kernel buffer, containing an Ethernet packet.
     193 * There is two types of commands
     194 * - The first 2 commands are used by the NIC_TX and NIC_RX server threads, and stored
     195 *   in the server thread descriptor, to access the NIC_RX & NIC_TX packet queues.
     196 *   The buffer is always a 2K bytes kernel buffer, containing an Ethernet packet.
     197 * - The next 4 synchronous commands are used by the client th, and stored in the
     198 *   client thread descriptor, to directly access the NIC registers.
    209199 ****************************************************************************************/
    210200
    211201typedef enum nic_cmd_e
    212202{
    213     NIC_CMD_WRITABLE  = 10,   /*! test TX queue not full (for a given packet length)    */
    214     NIC_CMD_WRITE     = 11,   /*! put one (given length) packet to TX queue             */
    215     NIC_CMD_READABLE  = 12,   /*! test RX queue not empty (for any packet length)       */
    216     NIC_CMD_READ      = 13,   /*! get one (any length) packet from RX queue             */
     203    NIC_CMD_WRITE        = 10,  /*! put one (given length) packet to TX queue           */
     204    NIC_CMD_READ         = 11,  /*! get one (any length) packet from RX queue           */
     205
     206    NIC_CMD_GET_KEY      = 20,  /*! return channel index from IP address and port       */
     207    NIC_CMD_SET_RUN      = 21,  /*! enable/disable one NIC channel                      */
     208    NIC_CMD_GET_INSTRU   = 22,  /*! return one intrumentation register value            */
     209    NIC_CMD_CLEAR_INSTRU = 23,  /*! reset all instrumentation registers                 */
    217210}
    218211nic_cmd_t;
    … …  
    222215    xptr_t      dev_xp;       /*! extended pointer on NIC chdev descriptor              */
    223216    nic_cmd_t   type;         /*! command type                                          */
    224     uint8_t   * buffer;       /*! local pointer on buffer (kernel or user space)        */ 
     217    uint8_t   * buffer;       /*! local pointer on kernel buffer                        */ 
    225218    uint32_t    length;       /*! number of bytes in buffer                             */
    226219    uint32_t    status;       /*! return value (depends on command type)                */
    … …  
    228221}
    229222nic_command_t;
    230 
    231 /*****************************************************************************************
    232  * This structure defines a socket descriptor. In order to parallelize the transfers,
    233  * the set of all registered sockets is split in several subsets.
    234  * The number of subsets is the number of  NIC channels.
    235  * The distribution key is computed from the (remote_addr/remote_port) couple.
    236  * This computation is done by the NIC hardware for RX packets,
    237  * and by the dev_nic_connect() function for the TX packets.
    238  *
    239  * A socket is attached to the NIC_TX[channel] & NIC_RX[channel] chdevs.
    240  * Each socket descriptor allows the TX and TX server threads to access various buffers:
    241  * - the user "send" buffer contains the data to be send by the TX server thread.
    242  * - the kernel "receive" buffer contains the data received by the RX server thread.
    243  * - the kernel "r2t" buffer allows the RX server thread to make direct requests
    244  *   to the associated TX server (to implement the TCP 3 steps handshake).
    245  *
    246  * The synchronisation mechanism between the clients threads and the servers threads
    247  * is different for TX and RX transfers:
    248  *
    249  * 1) For a TX transfer, it can exist only one client thread for a given socket,
    250  *    the transfer is always initiated by the local process, and all TX commands
    251  *    (CONNECT/SEND/CLOSE) are blocking for the client thread. The user buffer is
    252  *    used by TCP to handle retransmissions when required.in case of re
    253  *    The client thread registers the command in the thread descriptor, registers itself
    254  *    in the socket descriptor, unblocks the TX server thread from the BLOCKED_CLIENT
    255  *    condition, blocks itself on the BLOCKED_IO condition, and deschedules.
    256  *    When the command is completed, the TX server thread unblocks the client thread.
    257  *    The TX server blocks itself on the BLOCKED_CLIENT condition, when there is no
    258  *    pending commands and the R2T queue is empty. It is unblocked when a client
    259  *    register a new command, or when the TX server thread register a mew request
    260  *    in the R2T queue.
    261  *    The tx_valid flip-flop is SET by the client thread to signal a valid command.
    262  *    It is RESET by the server thread when the command is completed: For a SEND,
    263  *    all bytes have been sent (UDP) or acknowledged (TCP).
    264  *
    265  * 2) For an RX transfer, it can exist only one client thread for a given socket,
    266  *    but the transfer is initiated by the remote process, and the  RECV command
    267  *    is not really blocking: the data can arrive before the local RECV command is
    268  *    executed, and the server thread does not wait to receive all requested data
    269  *    to deliver data to client thread. Therefore each socket contains a receive
    270  *    buffer (rx_buf) handled as a single-writer/single-reader fifo.
    271  *    The client thread consumes data from the rx_buf when possible. It blocks on the
    272  *    BLOCKED_IO condition and deschedules when the rx_buf is empty.
    273  *    It is unblocked by the RX server thread when new data is available in the rx_buf.
    274  *    The RX server blocks itself on the BLOCKED_ISR condition When the NIC_RX packets
    275  *    queue is empty. It is unblocked by the hardware when new packets are available.
    276  *
    277  * Note : the socket domains and types are defined in the "shared_socket.h" file.
    278  ****************************************************************************************/
    279 
    280 /******************************************************************************************
    281  * This function returns a printable string for a given NIC command <type>.
    282  ******************************************************************************************
    283  * @ type    : NIC command type
    284  *****************************************************************************************/
    285 char * nic_cmd_str( uint32_t type );
    286    
    287 /******************************************************************************************
    288  * This function returns a printable string for a given socket <state>.
    289  ******************************************************************************************
    290  * @ state    : socket state
    291  *****************************************************************************************/
    292 char * socket_state_str( uint32_t state );
    293223
    294224/******************************************************************************************
    … …  
    299229 * device and the specific data structures when required.
    300230 * It creates the associated server thread and allocates a WTI from local ICU.
    301  * For a TX_NIC chedv, it allocates and initializes the R2T waiting queue used by the
    302  * NIC_RX[channel] server to send direct requests to the NIC_TX[channel] server.
     231 * For a TX_NIC chedv, it allocates and initializes the R2T queue used by the
     232 * NIC_RX[channel] server to send direct requests to the NIC_TX[channel] server,
     233 * and the CRQ queue used to register connection requests.
    303234 * It must de executed by a local thread.
     235 * For NIC_TX and NIC_RX chdevs, the "wait_root" field is actually a list of sockets.
    304236 ******************************************************************************************
    305237 * @ chdev    : local pointer on NIC chdev descriptor.
    … …  
    307239void dev_nic_init( struct chdev_s * chdev );
    308240
    309 
    310 /* functions implementing the socket API */
    311 
    312 /****************************************************************************************
    313  * This function implements the socket() syscall.
    314  * This function allocates and intializes in the calling thread cluster:
    315  * - a new socket descriptor, defined by the <domain> and <type> arguments,
    316  * - a new file descriptor, associated to this socket,
    317  * It registers the file descriptor in the reference process fd_array[], set
    318  * the socket state to IDLE, and returns the <fdid> value.
    319  ****************************************************************************************
    320  * @ domain  : [in] socket protocol family (AF_UNIX / AF_INET)
    321  * @ type    : [in] socket type (SOCK_DGRAM / SOCK_STREAM).
    322  * @ return a file descriptor <fdid> if success / return -1 if failure.
    323 ***************************************************************************************/
    324 int dev_nic_socket( uint32_t   domain,
    325                     uint32_t   type );
    326 
    327 /****************************************************************************************
    328  * This function implements the bind() syscall.
    329  * It initializes the  "local_addr" and "local_port" fields in the socket
    330  * descriptor identified by the <fdid> argument and set the socket state to BOUND.
    331  * It can be called by a thread running in any cluster.
    332  ****************************************************************************************
    333  * @ fdid      : [in] file descriptor identifying the socket.
    334  * @ addr      : [in] local IP address.
    335  * @ port      : [in] local port.
    336  * @ return 0 if success / return -1 if failure.
    337  ***************************************************************************************/
    338 int dev_nic_bind( uint32_t fdid,
    339                   uint32_t addr,
    340                   uint16_t port );
    341 
    342 /****************************************************************************************
    343  * This function implements the listen() syscall().
    344  * It is called by a (local) server process to specify the max size of the queue
    345  * registering the (remote) client process connections, and set the socket identified
    346  * by the <fdid> argument to LISTEN state. It applies only to sockets of type TCP.
    347  * It can be called by a thread running in any cluster.
    348  * TODO handle the <max_pending> argument...
    349  ****************************************************************************************
    350  * @ fdid        : [in] file descriptor identifying the local server socket.
    351  * @ max_pending : [in] max number of accepted remote client connections.
    352  ***************************************************************************************/
    353 int dev_nic_listen( uint32_t fdid,
    354                     uint32_t max_pending );
    355 
    356 /****************************************************************************************
    357  * This function implements the connect() syscall.
    358  * It is used by a (local) client process to connect a local socket identified by
    359  * the <fdid> argument, to a remote socket identified by the <remote_addr> and
    360  * <remote_port> arguments. It can be used for both  UDP and TCP sockets.
    361  * It computes the nic_channel index from <remote_addr> and <remote_port> values,
    362  * and initializes "remote_addr","remote_port", "nic_channel" in local socket.
    363  * It registers the socket in the two lists of clients rooted in the NIC_RX[channel]
    364  * and NIC_TX[channel] chdevs.  It can be called by a thread running in any cluster.
    365  * WARNING : the clients are the socket descriptors, and NOT the threads descriptors.
    366  ****************************************************************************************
    367  * Implementation Note:
    368  * - For a TCP socket, it updates the "remote_addr", "remote_port", "nic_channel" fields
    369  *   in the socket descriptor defined by the <fdid> argument, and register this socket,
    370  *   in the lists of sockets attached to the NIC_TX and NIC_RX chdevs.
    371  *   Then, it registers a CONNECT command in the "nic_cmd" field ot the client thread
    372  *   descriptor to request the NIC_TX server thread to execute the 3 steps handshake,
    373  *   and updates the "tx_client" field in the socket descriptor. It unblocks the NIC_TX
    374  *   server thread, blocks on the THREAD_BLOCKED_IO condition and deschedules.
    375  * - For an UDP socket, it simply updates "remote_addr", "remote_port", "nic_channel"
    376  *   in the socket descriptor defined by the <fdid> argument, and register this socket,
    377  *   in the lists of sockets attached to the NIC_TX and NIC_RX chdevs.
    378  *   Then, it set the socket state to CONNECT, without unblocking the NIC_TX server
    379  *   thread, and without blocking itself.
    380  * TODO : the nic_channel index computation must be done by a driver specific function.
    381  ****************************************************************************************
    382  * @ fdid          : [in] file descriptor identifying the socket.
    383  * @ remote_addr   : [in] remote IP address.
    384  * @ remote_port   : [in] remote port.
    385  * @ return 0 if success / return -1 if failure.
    386  ***************************************************************************************/
    387 int dev_nic_connect( uint32_t  fdid,
    388                      uint32_t  remote_addr,
    389                      uint16_t  remote_port );
    390 
    391 /****************************************************************************************
    392  * This function implements the accept() syscall().
    393  * It is executed by a server process, waiting for one (or several) client process(es)
    394  * requesting a connection on a socket identified by the <fdid> argument.
    395  * This socket was previouly created with socket(), bound to a local address with bind(),
    396  * and is listening for connections after a listen().
    397  * This function extracts the first connection request on the CRQQ queue of pending
    398  * requests, creates a new socket with the same properties as the existing socket,
    399  * and allocates a new file descriptor for this new socket.
    400  * If no pending connections are present on the queue, it blocks the caller until a
    401  * connection is present.
    402  * The new socket cannot accept more connections, but the original socket remains open.
    403  * It returns the new socket <fdid>, and register in the <address> an <port> arguments
    404  * the remote client IP address & port. It applies only to sockets of type SOCK_STREAM.
    405  ****************************************************************************************
    406  * @ fdid         : [in] file descriptor identifying the listening socket.
    407  * @ address      : [out] server IP address.
    408  * @ port         : [out] server port address length in bytes.
    409  * @ return the new socket <fdid> if success / return -1 if failure
    410  ***************************************************************************************/
    411 int dev_nic_accept( uint32_t   fdid,
    412                     uint32_t * address,
    413                     uint16_t * port );
    414 
    415 /****************************************************************************************
    416  * This blocking function implements the send() syscall.
    417  * It is used to send data stored in the user buffer, identified the <u_buf> and <length>
    418  * arguments, to a connected (TCP or UDP) socket, identified by the <fdid> argument.
    419  * The work is actually done by the NIC_TX server thread, and the synchronisation
    420  * between the client and the server threads uses the "rx_valid" set/reset flip-flop:
    421  * The client thread registers itself in the socket descriptor, registers in the queue
    422  * rooted in the NIC_TX[index] chdev, set "rx_valid", unblocks the server thread, and
    423  * finally blocks on THREAD_BLOCKED_IO, and deschedules.
    424  * When the TX server thread completes the command (all data has been sent for an UDP
    425  * socket, or acknowledeged for a TCP socket), the server thread reset "rx_valid" and
    426  * unblocks the client thread.
    427  * This function can be called by a thread running in any cluster.
    428  * WARNING : This implementation does not support several concurent SEND/SENDTO commands
    429  * on the same socket, as only one TX thread can register in a given socket.
    430  ****************************************************************************************
    431  * @ fdid      : [in] file descriptor identifying the socket.
    432  * @ u_buf     : [in] pointer on buffer containing packet in user space.
    433  * @ length    : [in] packet size in bytes.
    434  * @ return number of sent bytes if success / return -1 if failure.
    435  ***************************************************************************************/
    436 int dev_nic_send( uint32_t    fdid,
    437                   uint8_t   * u_buf,
    438                   uint32_t    length );
    439 
    440 /****************************************************************************************
    441  * This blocking function implements the sendto() syscall.
    442  * It registers the <remote_addr> and <remote_port> arguments in the local socket
    443  * descriptor, and does the same thing as the dev_nic_send() function above,
    444  * but can be called  on an unconnected UDP socket.
    445  ****************************************************************************************
    446  * @ fdid        : [in] file descriptor identifying the socket.
    447  * @ u_buf       : [in] pointer on buffer containing packet in user space.
    448  * @ length      : [in] packet size in bytes.
    449  * @ remote_addr : [in] destination IP address.
    450  * @ remote_port : [in] destination port.
    451  * @ return number of sent bytes if success / return -1 if failure.
    452  ***************************************************************************************/
    453 int dev_nic_sendto( uint32_t    fdid,
    454                     uint8_t   * u_buf,
    455                     uint32_t    length,
    456                     uint32_t    remote_addr,
    457                     uint32_t    remote_port );
    458 
    459 /****************************************************************************************
    460  * This blocking function implements the recv() syscall.
    461  * It is used to receive data that has been stored by the NIC_RX server thread in the
    462  * rx_buf of a connected (TCP or UDP) socket, identified by the <fdid> argument.
    463  * The synchronisation between the client and the server threads uses the "rx_valid"
    464  * set/reset flip-flop: If "rx_valid" is set, the client simply moves the available
    465  * data from the "rx_buf" to the user buffer identified by the <u_buf> and <length>
    466  * arguments, and reset the "rx_valid" flip_flop. If "rx_valid" is not set, the client
    467  * thread register itself in the socket descriptor, registers in the clients queue rooted
    468  * in the NIC_RX[index] chdev, and finally blocks on THREAD_BLOCKED_IO, and deschedules.
    469  * The client thread is re-activated by the RX server, that set the "rx_valid" flip-flop
    470  * as soon as data is available in the "rcv_buf" (can be less than the user buffer size).
    471  * This  function can be called by a thread running in any cluster.
    472  * WARNING : This implementation does not support several concurent RECV/RECVFROM
    473  * commands on the same socket, as only one RX thread can register in a given socket.
    474  ****************************************************************************************
    475  * @ fdid      : [in] file descriptor identifying the socket.
    476  * @ u_buf     : [in] pointer on buffer in user space.
    477  * @ length    : [in] buffer size in bytes.
    478  * @ return number of received bytes if success / return -1 if failure.
    479  ***************************************************************************************/
    480 int dev_nic_recv( uint32_t    fdid,
    481                   uint8_t   * u_buf,
    482                   uint32_t    length );
    483 
    484 /****************************************************************************************
    485  * This blocking function implements the recvfrom() syscall.
    486  * It registers the <remote_addr> and <remote_port> arguments in the local socket
    487  * descriptor, and does the same thing as the dev_nic_recv() function above,
    488  * but can be called  on an unconnected UDP socket.
    489  ****************************************************************************************
    490  * @ fdid        : [in] file descriptor identifying the socket.
    491  * @ u_buf       : [in] pointer on buffer containing packet in user space.
    492  * @ length      : [in] packet size in bytes.
    493  * @ remote_addr : [in] destination IP address.
    494  * @ remote_port : [in] destination port.
    495  * @ return number of received bytes if success / return -1 if failure.
    496  ***************************************************************************************/
    497 int dev_nic_recvfrom( uint32_t    fdid,
    498                       uint8_t   * u_buf,
    499                       uint32_t    length,
    500                       uint32_t    remote_addr,
    501                       uint32_t    remote_port );
    502 
    503 
    504 /*  Instrumentation functions */
    505 
     241/* Functions directly called by a client thread in any cluster */
     242 
     243/******************************************************************************************
     244 * This function compute a channel index in range [0,nic_channels[ from the remote IP
     245 * address <addr> and <port>, by calling the relevant driver command.
     246 ******************************************************************************************
     247 * @ addr     : [in] IP address.
     248 * @ port     : [in] TCP/UDP port.
     249 * @ return the selected channel index
     250 *****************************************************************************************/
     251uint32_t dev_nic_get_key( uint32_t addr,
     252                          uint16_t port );
     253
     254/******************************************************************************************
     255 * This function activate / de-activate a NIC channel DMA engine identified by the
     256 * <channel> argument, as defined by the <run> argument.
     257 ******************************************************************************************
     258 * @ channel  : [in] NIC channel index.
     259 * @ run      : [in] activate if non-zero / desactivate if zero.
     260 * @ return 0 if success / return -1 if error.
     261 *****************************************************************************************/
     262error_t dev_nic_set_run( uint32_t channel,
     263                         uint32_t run );
    506264
    507265/******************************************************************************************
    508266 * This instrumentation function displays on the TXT0 kernel terminal the content
    509267 * of the instrumentation registers contained in the NIC device.
    510  *****************************************************************************************/
    511 void dev_nic_print_stats( void );
     268 ******************************************************************************************
     269 * @ return 0 if success / return -1 if error.
     270 *****************************************************************************************/
     271error_t dev_nic_get_instru( void );
    512272
    513273/******************************************************************************************
    514274 * This instrumentation function reset all instrumentation registers contained
    515275 * in the NIC device.
    516  *****************************************************************************************/
    517 void dev_nic_clear_stats( void );
    518 
     276 ******************************************************************************************
     277 * @ return 0 if success / return -1 if error.
     278 *****************************************************************************************/
     279error_t dev_nic_clear_instru( void );
    519280
    520281/* Functions executed by the TX and RX server threads */
    … …  
    532293 * When the TX client threads queue is empty, it blocks on THREAD_BLOCKED_CLIENT
    533294 * condition and deschedules. It is re-activated by a client thread registering a command.
     295 * When the NIC_TX packet queue is full, it blocks on the THREAD_BLOCKED_ISR condition
     296 * and deschedules. It is reactivated by the NIC_TX DMA engine.
    534297 ******************************************************************************************
    535298 * Implementation note:
    … …  
    556319 * requests to the TX server, using the R2T queue attached to the involved socket.
    557320 * It blocks on the THREAD_BLOCKED_ISR condition and deschedules when the NIC_RX_QUEUE
    558  * is empty. It is re-activated by the NIC_RX_ISR, when the queue becomes non empty.
     321 * is empty, and is re-activated by the NIC_RX_ISR, when the queue becomes non empty.
    559322 ******************************************************************************************
    560323 * Implementation note:
    … …  
    570333void dev_nic_rx_server( struct chdev_s * chdev );
    571334
     335/******************************************************************************************
     336 * This function displays all the fields of an ETH/IP/TCP segment or ETH/IP/UDP packet.
     337 ******************************************************************************************
     338 * @ is_tx   : [in] sent packet if true / received packet if false.
     339 * @ pid     : [in] process identifier.
     340 * @ trdid   : [in] thread identifier.
     341 * @ cycle   : [in] date (number of cycles).
     342 * @ buf     : [in] local pointer on kernel buffer containing the packet.
     343 *****************************************************************************************/
     344void dev_nic_packet_display( bool_t    is_tx,
     345                             pid_t     pid,
     346                             trdid_t   trdid,
     347                             uint32_t  cycle,
     348                             uint8_t * buf );
     349
    572350#endif  /* _DEV_NIC_H */
Note: See TracChangeset for help on using the changeset viewer.