ipcpd: Adds a shim over IEEE 802.2 over IEEE 802.3

This adds a shim over LLC over Ethernet. It uses the raw socket API to
send messages directly over an interface.

3 files changed, 1212 insertions, 0 deletions

diff --git a/src/ipcpd/shim-eth-llc/CMakeLists.txt b/src/ipcpd/shim-eth-llc/CMakeLists.txt
new file mode 100644
index 00000000..14441113
--- /dev/null
+++ b/src/ipcpd/shim-eth-llc/CMakeLists.txt
@@ -0,0 +1,37 @@
+get_filename_component(CURRENT_SOURCE_PARENT_DIR
+  ${CMAKE_CURRENT_SOURCE_DIR} DIRECTORY)
+get_filename_component(CURRENT_BINARY_PARENT_DIR
+  ${CMAKE_CURRENT_BINARY_DIR} DIRECTORY)
+
+include_directories(${CMAKE_CURRENT_SOURCE_DIR})
+include_directories(${CMAKE_CURRENT_BINARY_DIR})
+
+include_directories(${CURRENT_SOURCE_PARENT_DIR})
+include_directories(${CURRENT_BINARY_PARENT_DIR})
+
+include_directories(${CMAKE_SOURCE_DIR}/include)
+include_directories(${CMAKE_BINARY_DIR}/include)
+
+find_package(ProtobufC REQUIRED)
+include_directories(${PROTOBUF_INCLUDE_DIRS})
+protobuf_generate_c(SHIM_ETH_LLC_PROTO_SRCS SHIM_ETH_LLC_PROTO_HDRS
+  shim_eth_llc_messages.proto)
+
+SET(IPCP_SHIM_ETH_LLC_TARGET ipcpd-shim-eth-llc
+  CACHE STRING "IPCP_SHIM_ETH_LLC_TARGET")
+
+set(SHIM_ETH_LLC_SOURCES
+  # Add source files here
+  ${CMAKE_CURRENT_SOURCE_DIR}/main.c)
+
+add_executable (ipcpd-shim-eth-llc ${SHIM_ETH_LLC_SOURCES} ${IPCP_SOURCES}
+ ${SHIM_ETH_LLC_PROTO_SRCS})
+target_link_libraries (ipcpd-shim-eth-llc LINK_PUBLIC ouroboros
+ ${PROTOBUF_C_LIBRARY})
+
+include(MacroAddCompileFlags)
+if (CMAKE_BUILD_TYPE MATCHES Debug)
+  MACRO_ADD_COMPILE_FLAGS(ipcpd-shim-eth-llc -DCONFIG_OUROBOROS_DEBUG)
+endif (CMAKE_BUILD_TYPE MATCHES Debug)
+
+install(TARGETS ipcpd-shim-eth-llc RUNTIME DESTINATION bin)
diff --git a/src/ipcpd/shim-eth-llc/main.c b/src/ipcpd/shim-eth-llc/main.c
new file mode 100644
index 00000000..73d6288e
--- /dev/null
+++ b/src/ipcpd/shim-eth-llc/main.c
@@ -0,0 +1,1160 @@
+/*
+ * Ouroboros - Copyright (C) 2016
+ *
+ * Shim IPC process over Ethernet with LLC
+ *
+ *    Sander Vrijders <sander.vrijders@intec.ugent.be>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <ouroboros/config.h>
+
+#define _DEFAULT_SOURCE
+
+#include "ipcp.h"
+#include "flow.h"
+#include <ouroboros/shm_du_map.h>
+#include <ouroboros/shm_ap_rbuff.h>
+#include <ouroboros/list.h>
+#include <ouroboros/utils.h>
+#include <ouroboros/ipcp.h>
+#include <ouroboros/dif_config.h>
+#include <ouroboros/sockets.h>
+#include <ouroboros/bitmap.h>
+#include <ouroboros/flow.h>
+#include <ouroboros/dev.h>
+#include <ouroboros/rw_lock.h>
+
+#define OUROBOROS_PREFIX "ipcpd/shim-eth-llc"
+
+#include <ouroboros/logs.h>
+
+#include <net/if.h>
+#include <signal.h>
+#include <stdlib.h>
+#include <pthread.h>
+#include <fcntl.h>
+#include <unistd.h>
+#include <errno.h>
+#include <string.h>
+#include <sys/socket.h>
+#include <sys/types.h>
+#include <sys/ioctl.h>
+#include <netinet/in.h>
+#include <linux/if_packet.h>
+#include <linux/if_ether.h>
+
+#include "shim_eth_llc_messages.pb-c.h"
+
+typedef ShimEthLlcMsg shim_eth_llc_msg_t;
+
+#define THIS_TYPE IPCP_SHIM_ETH_LLC
+#define MGMT_SAP 0x01
+#define SHIM_ETH_LLC_MAX_SDU_SIZE 1500
+#define MAC_SIZE 6
+#define MAX_SAPS 64
+
+/* global for trapping signal */
+int irmd_pid;
+
+struct ipcp * _ipcp;
+
+#define shim_data(type) ((struct eth_llc_ipcp_data *) type->data)
+
+#define ipcp_flow(index) ((struct flow *) &(shim_data(_ipcp)->flows[index]))
+
+struct eth_llc_flow {
+        struct flow flow;
+        uint8_t     sap;
+        uint8_t     r_sap;
+        uint8_t     r_addr[MAC_SIZE];
+};
+
+struct eth_llc_ipcp_data {
+        /* Keep ipcp_data first for polymorphism. */
+        struct ipcp_data      ipcp_data;
+
+        struct sockaddr_ll    device;
+        int                   s_fd;
+
+        struct bmp *          indices;
+        struct bmp *          saps;
+
+        struct shm_du_map *   dum;
+        struct shm_ap_rbuff * rb;
+
+        struct eth_llc_flow   flows[AP_MAX_FLOWS];
+        rw_lock_t             flows_lock;
+
+        pthread_t             mainloop;
+        pthread_t             sdu_writer;
+        pthread_t             sdu_reader;
+};
+
+struct eth_llc_ipcp_data * eth_llc_ipcp_data_create()
+{
+        struct eth_llc_ipcp_data * eth_llc_data;
+        enum ipcp_type             ipcp_type;
+
+        eth_llc_data = malloc(sizeof(*eth_llc_data));
+        if (eth_llc_data == NULL) {
+                LOG_ERR("Failed to allocate.");
+                return NULL;
+        }
+
+        ipcp_type = THIS_TYPE;
+        if (ipcp_data_init((struct ipcp_data *) eth_llc_data,
+                           ipcp_type) == NULL) {
+                free(eth_llc_data);
+                return NULL;
+        }
+
+        eth_llc_data->dum = shm_du_map_open();
+        if (eth_llc_data->dum == NULL) {
+                free(eth_llc_data);
+                return NULL;
+        }
+
+        eth_llc_data->rb = shm_ap_rbuff_create();
+        if (eth_llc_data->rb == NULL) {
+                shm_du_map_close(eth_llc_data->dum);
+                free(eth_llc_data);
+                return NULL;
+        }
+
+        eth_llc_data->indices = bmp_create(AP_MAX_FLOWS, 0);
+        if (eth_llc_data->indices == NULL) {
+                shm_ap_rbuff_destroy(eth_llc_data->rb);
+                shm_du_map_close(eth_llc_data->dum);
+                free(eth_llc_data);
+                return NULL;
+        }
+
+        eth_llc_data->saps = bmp_create(MAX_SAPS, 2);
+        if (eth_llc_data->indices == NULL) {
+                bmp_destroy(eth_llc_data->indices);
+                shm_ap_rbuff_destroy(eth_llc_data->rb);
+                shm_du_map_close(eth_llc_data->dum);
+                free(eth_llc_data);
+                return NULL;
+        }
+
+        rw_lock_init(&eth_llc_data->flows_lock);
+
+        return eth_llc_data;
+}
+
+void eth_llc_ipcp_data_destroy()
+{
+        int i = 0;
+
+        if (_ipcp == NULL)
+                return;
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+
+        if (_ipcp->state != IPCP_SHUTDOWN)
+                LOG_WARN("Cleaning up while not in shutdown.");
+
+        if (shim_data(_ipcp)->dum != NULL)
+                shm_du_map_close(shim_data(_ipcp)->dum);
+        if (shim_data(_ipcp)->rb != NULL)
+                shm_ap_rbuff_destroy(shim_data(_ipcp)->rb);
+        if (shim_data(_ipcp)->indices != NULL)
+                bmp_destroy(shim_data(_ipcp)->indices);
+        if (shim_data(_ipcp)->saps != NULL)
+                bmp_destroy(shim_data(_ipcp)->saps);
+
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        for (i = 0; i < AP_MAX_FLOWS; i ++)
+                if (ipcp_flow(i)->rb != NULL)
+                        shm_ap_rbuff_close(ipcp_flow(i)->rb);
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        free(_ipcp->data);
+}
+
+/* only call this under flows_lock */
+static int port_id_to_index(int port_id)
+{
+        int i;
+
+        for (i = 0; i < AP_MAX_FLOWS; ++i) {
+                if (ipcp_flow(i)->port_id == port_id
+                    && ipcp_flow(i)->state != FLOW_NULL)
+                        return i;
+        }
+
+        return -1;
+}
+
+/* only call this under flows_lock */
+static int addr_and_saps_to_index(uint8_t r_addr[MAC_SIZE],
+                                  uint8_t r_sap,
+                                  uint8_t sap)
+{
+        int i = 0;
+
+        for (i = 0; i < AP_MAX_FLOWS; i++) {
+                if (shim_data(_ipcp)->flows[i].r_sap == r_sap &&
+                    shim_data(_ipcp)->flows[i].sap == sap &&
+                    !memcmp(shim_data(_ipcp)->flows[i].r_addr,
+                            r_addr,
+                            MAC_SIZE)) {
+                        return i;
+                }
+        }
+
+        return -1;
+}
+
+/* only call this under flows_lock */
+static int sap_to_index(uint8_t sap)
+{
+        int i = 0;
+
+        for (i = 0; i < AP_MAX_FLOWS; i++) {
+                if (shim_data(_ipcp)->flows[i].sap == sap) {
+                        return i;
+                }
+        }
+
+        return -1;
+}
+
+/* only call this under flows_lock */
+static void destroy_ipcp_flow(int index)
+{
+        ipcp_flow(index)->port_id = -1;
+        if (ipcp_flow(index)->rb != NULL)
+                shm_ap_rbuff_close(ipcp_flow(index)->rb);
+        ipcp_flow(index)->rb = NULL;
+        ipcp_flow(index)->state = FLOW_NULL;
+        bmp_release(shim_data(_ipcp)->indices, index);
+        bmp_release(shim_data(_ipcp)->saps,
+                    shim_data(_ipcp)->flows[index].sap);
+}
+
+static uint8_t reverse_bits(uint8_t b)
+{
+        b = (b & 0xF0) >> 4 | (b & 0x0F) << 4;
+        b = (b & 0xCC) >> 2 | (b & 0x33) << 2;
+        b = (b & 0xAA) >> 1 | (b & 0x55) << 1;
+
+        return b;
+}
+
+static int eth_llc_ipcp_send_frame(uint8_t   dst_addr[MAC_SIZE],
+                                   uint8_t   dsap,
+                                   uint8_t   ssap,
+                                   uint8_t * payload,
+                                   size_t    len)
+{
+        uint8_t frame[SHIM_ETH_LLC_MAX_SDU_SIZE];
+        int frame_len = 0;
+        struct sockaddr_ll device;
+        uint8_t cf = 0x03;
+        int fd;
+        uint16_t length = 0;
+
+        if (payload == NULL) {
+                LOG_ERR("Payload was NULL.");
+                return -1;
+        }
+
+        length = htons(len);
+
+        memcpy(frame, dst_addr, MAC_SIZE * sizeof(uint8_t));
+        frame_len += MAC_SIZE;
+        memcpy(frame + frame_len,
+               shim_data(_ipcp)->device.sll_addr,
+               MAC_SIZE * sizeof(uint8_t));
+        frame_len += MAC_SIZE;
+        memcpy(frame + frame_len, &length, 2 * sizeof(uint8_t));
+        frame_len += 2 * sizeof(uint8_t);
+        memcpy(frame + frame_len, &dsap, sizeof(uint8_t));
+        frame_len += sizeof(uint8_t);
+        memcpy(frame + frame_len, &ssap, sizeof(uint8_t));
+        frame_len += sizeof(uint8_t);
+        memcpy(frame + frame_len, &cf, sizeof(uint8_t));
+        frame_len += sizeof(uint8_t);
+        memcpy(frame + frame_len, payload, len);
+        frame_len += len;
+
+        rw_lock_rdlock(&_ipcp->state_lock);
+        device = (shim_data(_ipcp))->device;
+        fd = (shim_data(_ipcp))->s_fd;
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        if (sendto(fd, frame, frame_len, 0,
+                   (struct sockaddr *) &device, sizeof(device)) <= 0) {
+                LOG_ERR("Failed to send message");
+                return -1;
+        }
+
+        return 0;
+}
+
+static int eth_llc_ipcp_send_mgmt_frame(shim_eth_llc_msg_t * msg,
+                                        uint8_t              dst_addr[MAC_SIZE])
+{
+        size_t    len;
+        uint8_t * buf;
+
+        len = shim_eth_llc_msg__get_packed_size(msg);
+        if (len == 0) {
+                return -1;
+        }
+
+        buf = malloc(len);
+        if (buf == NULL) {
+                return -1;
+        }
+
+        shim_eth_llc_msg__pack(msg, buf);
+
+        if (eth_llc_ipcp_send_frame(dst_addr, reverse_bits(MGMT_SAP),
+                                    reverse_bits(MGMT_SAP), buf, len)) {
+                LOG_ERR("Failed to send mgmt frame");
+                return -1;
+        }
+
+        free(buf);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_port_alloc(uint8_t dst_addr[MAC_SIZE],
+                                   uint8_t ssap,
+                                   char *  dst_name,
+                                   char *  src_ap_name,
+                                   char *  src_ae_name)
+{
+        shim_eth_llc_msg_t msg = SHIM_ETH_LLC_MSG__INIT;
+
+        msg.code        = SHIM_ETH_LLC_MSG_CODE__FLOW_REQ;
+        msg.ssap        = ssap;
+        msg.dst_name    = dst_name;
+        msg.src_ap_name = src_ap_name;
+        msg.src_ae_name = src_ae_name;
+
+        return eth_llc_ipcp_send_mgmt_frame(&msg, dst_addr);
+}
+
+static int eth_llc_ipcp_port_alloc_resp(uint8_t dst_addr[MAC_SIZE],
+                                        uint8_t ssap,
+                                        uint8_t dsap,
+                                        int     response)
+{
+        shim_eth_llc_msg_t msg = SHIM_ETH_LLC_MSG__INIT;
+
+        msg.code         = SHIM_ETH_LLC_MSG_CODE__FLOW_REPLY;
+        msg.ssap         = ssap;
+        msg.has_dsap     = true;
+        msg.dsap         = dsap;
+        msg.has_response = true;
+        msg.response     = response;
+
+        return eth_llc_ipcp_send_mgmt_frame(&msg, dst_addr);
+}
+
+static int eth_llc_ipcp_port_dealloc(uint8_t dst_addr[MAC_SIZE],
+                                     uint8_t ssap)
+{
+        shim_eth_llc_msg_t msg = SHIM_ETH_LLC_MSG__INIT;
+
+        msg.code = SHIM_ETH_LLC_MSG_CODE__FLOW_DEALLOC;
+        msg.ssap = ssap;
+
+        return eth_llc_ipcp_send_mgmt_frame(&msg, dst_addr);
+}
+
+static int eth_llc_ipcp_port_req(uint8_t r_sap,
+                                 uint8_t r_addr[MAC_SIZE],
+                                 char *  dst_name,
+                                 char *  src_ap_name,
+                                 char *  src_ae_name)
+{
+        int port_id;
+        ssize_t index = 0;
+        int i;
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        index = bmp_allocate(shim_data(_ipcp)->indices);
+        if (index < 0) {
+                rw_lock_unlock(&_ipcp->state_lock);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                LOG_ERR("Out of free indices");
+                return -1;
+        }
+
+        /* reply to IRM */
+        port_id = ipcp_flow_req_arr(getpid(),
+                                    dst_name,
+                                    src_ap_name,
+                                    src_ae_name);
+
+        if (port_id < 0) {
+                bmp_release(shim_data(_ipcp)->indices, index);
+                rw_lock_unlock(&_ipcp->state_lock);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                LOG_ERR("Could not get port id from IRMd");
+                return -1;
+        }
+
+        ipcp_flow(index)->port_id = port_id;
+        ipcp_flow(index)->rb      = NULL;
+        ipcp_flow(index)->state   = FLOW_PENDING;
+        shim_data(_ipcp)->flows[index].r_sap = r_sap;
+        for (i = 0; i < MAC_SIZE; i++) {
+                shim_data(_ipcp)->flows[index].r_addr[i] = r_addr[i];
+        }
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        LOG_DBG("New flow request, port_id %d, remote SAP %d.", port_id, r_sap);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_port_alloc_reply(uint8_t ssap,
+                                         uint8_t r_addr[MAC_SIZE],
+                                         int     dsap,
+                                         int     response)
+{
+        int index = -1;
+        int ret = 0;
+        int port_id = -1;
+        int i;
+
+        rw_lock_rdlock(&_ipcp->state_lock);
+        rw_lock_rdlock(&shim_data(_ipcp)->flows_lock);
+
+        index = sap_to_index(ssap);
+        if (index < 0) {
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_ERR("No flow found with that SAP");
+                return -1; /* -EFLOWNOTFOUND */
+        }
+
+        if (ipcp_flow(index)->state != FLOW_PENDING) {
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return -1; /* -EFLOWNOTPENDING */
+        }
+
+        port_id = ipcp_flow(index)->port_id;
+
+        if (response) {
+                destroy_ipcp_flow(index);
+        } else {
+                ipcp_flow(index)->state = FLOW_ALLOCATED;
+                shim_data(_ipcp)->flows[index].r_sap = dsap;
+                for (i = 0; i < MAC_SIZE; i++) {
+                        shim_data(_ipcp)->flows[index].r_addr[i] = r_addr[i];
+                }
+        }
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        LOG_DBG("Flow reply, port_id %d, remote SAP %d.", port_id, dsap);
+
+        if ((ret = ipcp_flow_alloc_reply(getpid(),
+                                         port_id,
+                                         response)) < 0) {
+                return -1; /* -EPIPE */
+        }
+
+        return ret;
+
+}
+
+static int eth_llc_ipcp_flow_dealloc_req(uint8_t ssap,
+                                         uint8_t r_addr[MAC_SIZE])
+{
+        int port_id = -1;
+        int i = 0;
+
+        rw_lock_rdlock(&_ipcp->state_lock);
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        i = sap_to_index(ssap);
+        if (i < 0) {
+                LOG_ERR("No flow found for remote deallocation request");
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return 0;
+        }
+
+        port_id = ipcp_flow(i)->port_id;
+        destroy_ipcp_flow(i);
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        ipcp_flow_dealloc(0, port_id);
+
+        LOG_DBG("Flow with port_id %d deallocated.", port_id);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_mgmt_frame(uint8_t * buf,
+                                   size_t    len,
+                                   uint8_t   r_addr[MAC_SIZE])
+{
+        shim_eth_llc_msg_t * msg = NULL;
+
+        msg = shim_eth_llc_msg__unpack(NULL, len, buf);
+        if (msg == NULL) {
+                return -1;
+        }
+
+        switch (msg->code) {
+        case SHIM_ETH_LLC_MSG_CODE__FLOW_REQ:
+                if (ipcp_data_is_in_registry(_ipcp->data,
+                                             msg->dst_name)) {
+                        eth_llc_ipcp_port_req(msg->ssap,
+                                              r_addr,
+                                              msg->dst_name,
+                                              msg->src_ap_name,
+                                              msg->src_ae_name);
+                }
+                break;
+        case SHIM_ETH_LLC_MSG_CODE__FLOW_REPLY:
+                eth_llc_ipcp_port_alloc_reply(msg->ssap,
+                                              r_addr,
+                                              msg->dsap,
+                                              msg->response);
+                break;
+        case SHIM_ETH_LLC_MSG_CODE__FLOW_DEALLOC:
+                eth_llc_ipcp_flow_dealloc_req(msg->ssap,
+                                              r_addr);
+                break;
+        default:
+                LOG_ERR("Unknown message received %d.",
+                        msg->code);
+                shim_eth_llc_msg__free_unpacked(msg, NULL);
+                return -1;
+        }
+
+        shim_eth_llc_msg__free_unpacked(msg, NULL);
+        return 0;
+}
+
+static void * eth_llc_ipcp_sdu_reader(void * o)
+{
+        uint8_t buf[SHIM_ETH_LLC_MAX_SDU_SIZE];
+        ssize_t index;
+        struct rb_entry e;
+        uint8_t src_mac[MAC_SIZE];
+        uint8_t dst_mac[MAC_SIZE];
+        uint8_t br_addr[MAC_SIZE];
+        int frame_len = 0;
+        uint8_t ssap = 0;
+        uint8_t dsap = 0;
+        int i = 0;
+        int j = 0;
+
+        memset(br_addr, 0xff, MAC_SIZE * sizeof(uint8_t));
+
+        while (true) {
+                rw_lock_rdlock(&_ipcp->state_lock);
+
+                if (_ipcp->state != IPCP_ENROLLED) {
+                        rw_lock_unlock(&_ipcp->state_lock);
+                        return (void *) 1; /* -ENOTENROLLED */
+                }
+
+                rw_lock_unlock(&_ipcp->state_lock);
+
+                if (recv(shim_data(_ipcp)->s_fd, buf,
+                         SHIM_ETH_LLC_MAX_SDU_SIZE, 0) < 0) {
+                        LOG_ERR("Failed to recv frame");
+                        continue;
+                }
+
+                for (i = 0; i < MAC_SIZE; i++) {
+                        dst_mac[i] = buf[i];
+                }
+
+                if (memcmp(shim_data(_ipcp)->device.sll_addr,
+                           dst_mac,
+                           MAC_SIZE) &&
+                    memcmp(br_addr, dst_mac, MAC_SIZE)) {
+                        LOG_DBG("Not a unicast or broadcast frame");
+                        continue;
+                }
+
+                for (; i < 2 * MAC_SIZE; i++) {
+                        src_mac[i - MAC_SIZE] = buf[i];
+                }
+
+                frame_len = ((buf[i]) << 8) + buf[i + 1];
+                i += 2;
+
+                dsap = reverse_bits(buf[i++]);
+                ssap = reverse_bits(buf[i++]);
+                i++;
+
+                if (ssap == MGMT_SAP &&
+                    dsap == MGMT_SAP) {
+                        eth_llc_ipcp_mgmt_frame((uint8_t *) (buf + i),
+                                                frame_len, src_mac);
+                } else {
+                        rw_lock_rdlock(&_ipcp->state_lock);
+                        rw_lock_rdlock(&shim_data(_ipcp)->flows_lock);
+
+                        j = addr_and_saps_to_index(src_mac, ssap, dsap);
+                        if (j < 0) {
+                                LOG_DBG("Received data for unknown flow");
+                                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                                rw_lock_unlock(&_ipcp->state_lock);
+                                continue;
+                        }
+
+                        while ((index =
+                                shm_create_du_buff(shim_data(_ipcp)->dum,
+                                                   frame_len, 0,
+                                                   (uint8_t *) (buf + i),
+                                                   frame_len)) < 0)
+                                ;
+
+                        e.index = index;
+                        e.port_id = ipcp_flow(j)->port_id;
+
+                        while (shm_ap_rbuff_write(ipcp_flow(j)->rb, &e) < 0)
+                                ;
+
+                        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                        rw_lock_unlock(&_ipcp->state_lock);
+                }
+        }
+
+        return (void *) 0;
+}
+
+static void * eth_llc_ipcp_sdu_writer(void * o)
+{
+        while (true) {
+                struct rb_entry * e;
+                int i;
+                int len = 0;
+                uint8_t * buf;
+                uint8_t ssap;
+                uint8_t dsap;
+
+                e = shm_ap_rbuff_read(shim_data(_ipcp)->rb);
+                if (e == NULL) {
+                        continue;
+                }
+
+                rw_lock_rdlock(&_ipcp->state_lock);
+
+                if (_ipcp->state != IPCP_ENROLLED) {
+                        rw_lock_unlock(&_ipcp->state_lock);
+                        return (void *) 1; /* -ENOTENROLLED */
+                }
+
+                len = shm_du_map_read_sdu((uint8_t **) &buf,
+                                          shim_data(_ipcp)->dum,
+                                          e->index);
+                if (len <= 0) {
+                        rw_lock_unlock(&_ipcp->state_lock);
+                        free(e);
+                        continue;
+                }
+
+                rw_lock_rdlock(&shim_data(_ipcp)->flows_lock);
+
+                i = port_id_to_index(e->port_id);
+                if (i < 0) {
+                        free(e);
+                        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                        rw_lock_unlock(&_ipcp->state_lock);
+                        continue;
+                }
+
+                ssap = reverse_bits(shim_data(_ipcp)->flows[i].sap);
+                dsap = reverse_bits(shim_data(_ipcp)->flows[i].r_sap);
+
+                if (eth_llc_ipcp_send_frame(shim_data(_ipcp)->flows[i].r_addr,
+                                            dsap, ssap, buf, len)) {
+                        LOG_ERR("Failed to send SDU");
+                }
+
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+
+                if (shim_data(_ipcp)->dum != NULL)
+                        shm_release_du_buff(shim_data(_ipcp)->dum, e->index);
+
+                rw_lock_unlock(&_ipcp->state_lock);
+        }
+
+        return (void *) 1;
+}
+
+void ipcp_sig_handler(int sig, siginfo_t * info, void * c)
+{
+        sigset_t  sigset;
+        sigemptyset(&sigset);
+        sigaddset(&sigset, SIGINT);
+
+        switch(sig) {
+        case SIGINT:
+        case SIGTERM:
+        case SIGHUP:
+                if (info->si_pid == irmd_pid) {
+                        bool clean_threads = false;
+                        LOG_DBG("Terminating by order of %d. Bye.",
+                                info->si_pid);
+
+                        rw_lock_wrlock(&_ipcp->state_lock);
+
+                        if (_ipcp->state == IPCP_ENROLLED)
+                                clean_threads = true;
+
+                        _ipcp->state = IPCP_SHUTDOWN;
+
+                        rw_lock_unlock(&_ipcp->state_lock);
+
+                        if (clean_threads) {
+                                pthread_cancel(shim_data(_ipcp)->sdu_reader);
+                                pthread_cancel(shim_data(_ipcp)->sdu_writer);
+
+                                pthread_join(shim_data(_ipcp)->sdu_writer,
+                                             NULL);
+                                pthread_join(shim_data(_ipcp)->sdu_reader,
+                                             NULL);
+                        }
+
+                        pthread_cancel(shim_data(_ipcp)->mainloop);
+
+                }
+        default:
+                return;
+        }
+}
+
+static int eth_llc_ipcp_bootstrap(struct dif_config * conf)
+{
+        int fd = -1;
+        struct ifreq ifr;
+        int index;
+
+        if (conf == NULL)
+                return -1; /* -EINVAL */
+
+        if (conf->type != THIS_TYPE) {
+                LOG_ERR("Config doesn't match IPCP type.");
+                return -1;
+        }
+
+        if (conf->if_name == NULL) {
+                LOG_ERR("Interface name is NULL");
+                return -1;
+        }
+
+        memset(&ifr, 0, sizeof(ifr));
+
+        fd = socket(AF_UNIX, SOCK_STREAM, 0);
+        if (fd < 0) {
+                LOG_ERR("Failed to open socket");
+                return -1;
+        }
+
+        memcpy(ifr.ifr_name, conf->if_name, strlen(conf->if_name));
+
+        if (ioctl(fd, SIOCGIFHWADDR, &ifr)) {
+                close(fd);
+                LOG_ERR("Failed to ioctl: %s", strerror(errno));
+                return -1;
+        }
+
+        close(fd);
+
+        index = if_nametoindex(conf->if_name);
+        if (index == 0) {
+                LOG_ERR("Failed to retrieve interface index");
+                return -1;
+        }
+
+        fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_802_2));
+        if (fd < 0) {
+                LOG_ERR("Failed to create socket: %s", strerror(errno));
+                return -1;
+        }
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+
+        if (_ipcp->state != IPCP_INIT) {
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_ERR("IPCP in wrong state.");
+                close(fd);
+                return -1;
+        }
+
+        shim_data(_ipcp)->s_fd = fd;
+
+        memset(&(shim_data(_ipcp)->device), 0,
+               sizeof(shim_data(_ipcp)->device));
+        shim_data(_ipcp)->device.sll_ifindex = index;
+        shim_data(_ipcp)->device.sll_family = AF_PACKET;
+        memcpy(shim_data(_ipcp)->device.sll_addr,
+               ifr.ifr_hwaddr.sa_data,
+               MAC_SIZE * sizeof (uint8_t));
+        shim_data(_ipcp)->device.sll_halen = MAC_SIZE;
+        shim_data(_ipcp)->device.sll_protocol = htons(ETH_P_802_3);
+
+        _ipcp->state = IPCP_ENROLLED;
+
+        pthread_create(&shim_data(_ipcp)->sdu_reader,
+                       NULL,
+                       eth_llc_ipcp_sdu_reader,
+                       NULL);
+
+        pthread_create(&shim_data(_ipcp)->sdu_writer,
+                       NULL,
+                       eth_llc_ipcp_sdu_writer,
+                       NULL);
+
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        LOG_DBG("Bootstrapped shim IPCP over Ethernet with LLC with pid %d.",
+                getpid());
+
+        return 0;
+}
+
+static int eth_llc_ipcp_name_reg(char * name)
+{
+        rw_lock_rdlock(&_ipcp->state_lock);
+
+        if (_ipcp->state != IPCP_ENROLLED) {
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_DBGF("Won't register with non-enrolled IPCP.");
+                return -1; /* -ENOTENROLLED */
+        }
+
+        if (ipcp_data_add_reg_entry(_ipcp->data, name)) {
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_ERR("Failed to add %s to local registry.", name);
+                return -1;
+        }
+
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        LOG_DBG("Registered %s.", name);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_name_unreg(char * name)
+{
+        rw_lock_rdlock(&_ipcp->state_lock);
+
+        ipcp_data_del_reg_entry(_ipcp->data, name);
+
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_flow_alloc(pid_t         n_pid,
+                                   int           port_id,
+                                   char *        dst_name,
+                                   char *        src_ap_name,
+                                   char *        src_ae_name,
+                                   enum qos_cube qos)
+{
+        struct shm_ap_rbuff * rb;
+        uint8_t ssap = 0;
+        uint8_t r_addr[MAC_SIZE];
+        int index = 0;
+
+        LOG_INFO("Allocating flow from %s to %s.", src_ap_name, dst_name);
+
+        if (dst_name == NULL || src_ap_name == NULL || src_ae_name == NULL)
+                return -1;
+
+        if (qos != QOS_CUBE_BE)
+                LOG_DBGF("QoS requested. Ethernet LLC can't do that. For now.");
+
+        rb = shm_ap_rbuff_open(n_pid);
+        if (rb == NULL)
+                return -1; /* -ENORBUFF */
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+
+        if (_ipcp->state != IPCP_ENROLLED) {
+                shm_ap_rbuff_close(rb);
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_DBGF("Won't allocate flow with non-enrolled IPCP.");
+                return -1; /* -ENOTENROLLED */
+        }
+
+        index = bmp_allocate(shim_data(_ipcp)->indices);
+        if (index < 0) {
+                shm_ap_rbuff_close(rb);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return -1;
+        }
+
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        ssap = bmp_allocate(shim_data(_ipcp)->saps);
+        if (ssap < 0) {
+                shm_ap_rbuff_close(rb);
+                bmp_release(shim_data(_ipcp)->indices, index);
+                rw_lock_unlock(&_ipcp->state_lock);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                return -1;
+        }
+
+        ipcp_flow(index)->port_id = port_id;
+        ipcp_flow(index)->state = FLOW_PENDING;
+        ipcp_flow(index)->rb = rb;
+        shim_data(_ipcp)->flows[index].sap = ssap;
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        memset(r_addr, 0xff, MAC_SIZE * sizeof(uint8_t));
+
+        if (eth_llc_ipcp_port_alloc(r_addr, ssap,
+                                    dst_name,
+                                    src_ap_name,
+                                    src_ae_name) < 0) {
+                LOG_DBGF("Port alloc returned -1.");
+                rw_lock_wrlock(&_ipcp->state_lock);
+                rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+                destroy_ipcp_flow(index);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return -1;
+        }
+
+        LOG_DBG("Pending flow with port_id %d on SAP %d.",
+                port_id, ssap);
+
+        return index;
+}
+
+static int eth_llc_ipcp_flow_alloc_resp(pid_t n_pid,
+                                        int   port_id,
+                                        int   response)
+{
+        struct shm_ap_rbuff * rb;
+        int index = -1;
+        uint8_t ssap = 0;
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        index = port_id_to_index(port_id);
+        if (index < 0) {
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_DBGF("Could not find flow with port_id %d.", port_id);
+                return -1;
+        }
+
+        if (ipcp_flow(index)->state != FLOW_PENDING) {
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                LOG_DBGF("Flow was not pending.");
+                return -1;
+        }
+
+        rb = shm_ap_rbuff_open(n_pid);
+        if (rb == NULL) {
+                LOG_ERR("Could not open N + 1 ringbuffer.");
+                ipcp_flow(index)->state = FLOW_NULL;
+                ipcp_flow(index)->port_id = -1;
+                bmp_release(shim_data(_ipcp)->indices, index);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return -1;
+        }
+
+        ssap = bmp_allocate(shim_data(_ipcp)->saps);
+        if (ssap < 0) {
+                ipcp_flow(index)->state = FLOW_NULL;
+                ipcp_flow(index)->port_id = -1;
+                shm_ap_rbuff_close(ipcp_flow(index)->rb);
+                bmp_release(shim_data(_ipcp)->indices, index);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return -1;
+        }
+
+        ipcp_flow(index)->state = FLOW_ALLOCATED;
+        ipcp_flow(index)->rb = rb;
+        shim_data(_ipcp)->flows[index].sap = ssap;
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        if (eth_llc_ipcp_port_alloc_resp(shim_data(_ipcp)->flows[index].r_addr,
+                                         shim_data(_ipcp)->flows[index].r_sap,
+                                         ssap,
+                                         response) < 0) {
+                rw_lock_rdlock(&_ipcp->state_lock);
+                rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+                destroy_ipcp_flow(index);
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+
+                LOG_DBGF("Could not send response.");
+                return -1;
+        }
+
+        LOG_DBG("Accepted flow, port_id %d, SAP %d.", port_id, ssap);
+
+        return 0;
+}
+
+static int eth_llc_ipcp_flow_dealloc(int port_id)
+{
+        int index = -1;
+        uint8_t sap;
+        uint8_t addr[MAC_SIZE];
+        int i;
+
+        rw_lock_rdlock(&_ipcp->state_lock);
+        rw_lock_wrlock(&shim_data(_ipcp)->flows_lock);
+
+        index = port_id_to_index(port_id);
+        if (index < 0) {
+                rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+                rw_lock_unlock(&_ipcp->state_lock);
+                return 0;
+        }
+
+        sap = shim_data(_ipcp)->flows[index].r_sap;
+        for (i = 0; i < MAC_SIZE; i++) {
+                addr[i] = shim_data(_ipcp)->flows[index].r_addr[i];
+        }
+
+        destroy_ipcp_flow(index);
+
+        rw_lock_unlock(&shim_data(_ipcp)->flows_lock);
+
+        if (eth_llc_ipcp_port_dealloc(addr, sap) < 0)
+                LOG_DBGF("Could not notify remote.");
+
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        LOG_DBG("Flow with port_id %d deallocated.", port_id);
+
+        return 0;
+}
+
+static struct ipcp_ops eth_llc_ops = {
+        .ipcp_bootstrap       = eth_llc_ipcp_bootstrap,
+        .ipcp_enroll          = NULL,                       /* shim */
+        .ipcp_reg             = NULL,                       /* shim */
+        .ipcp_unreg           = NULL,                       /* shim */
+        .ipcp_name_reg        = eth_llc_ipcp_name_reg,
+        .ipcp_name_unreg      = eth_llc_ipcp_name_unreg,
+        .ipcp_flow_alloc      = eth_llc_ipcp_flow_alloc,
+        .ipcp_flow_alloc_resp = eth_llc_ipcp_flow_alloc_resp,
+        .ipcp_flow_dealloc    = eth_llc_ipcp_flow_dealloc
+};
+
+int main(int argc, char * argv[])
+{
+        /* argument 1: pid of irmd ? */
+        /* argument 2: ap name */
+        struct sigaction sig_act;
+        sigset_t  sigset;
+        int i = 0;
+
+        sigemptyset(&sigset);
+        sigaddset(&sigset, SIGINT);
+        sigaddset(&sigset, SIGQUIT);
+        sigaddset(&sigset, SIGHUP);
+        sigaddset(&sigset, SIGPIPE);
+
+        if (ipcp_arg_check(argc, argv)) {
+                LOG_ERR("Wrong arguments.");
+                exit(1);
+        }
+
+        /* store the process id of the irmd */
+        irmd_pid = atoi(argv[1]);
+
+        /* init sig_act */
+        memset(&sig_act, 0, sizeof(sig_act));
+
+        /* install signal traps */
+        sig_act.sa_sigaction = &ipcp_sig_handler;
+        sig_act.sa_flags     = SA_SIGINFO;
+
+        sigaction(SIGINT,  &sig_act, NULL);
+        sigaction(SIGTERM, &sig_act, NULL);
+        sigaction(SIGHUP,  &sig_act, NULL);
+        sigaction(SIGPIPE, &sig_act, NULL);
+
+        _ipcp = ipcp_instance_create();
+        if (_ipcp == NULL) {
+                LOG_ERR("Failed to create instance");
+                exit(1);
+        }
+
+        _ipcp->data = (struct ipcp_data *) eth_llc_ipcp_data_create();
+        if (_ipcp->data == NULL) {
+                LOG_ERR("Failed to create instance data.");
+                free(_ipcp);
+                exit(1);
+        }
+
+        for (i = 0; i < AP_MAX_FLOWS; i++) {
+                ipcp_flow(i)->rb = NULL;
+                ipcp_flow(i)->port_id = -1;
+                ipcp_flow(i)->state = FLOW_NULL;
+        }
+
+        _ipcp->ops = &eth_llc_ops;
+        _ipcp->state = IPCP_INIT;
+
+        rw_lock_wrlock(&_ipcp->state_lock);
+
+        pthread_sigmask(SIG_BLOCK, &sigset, NULL);
+
+        pthread_create(&shim_data(_ipcp)->mainloop, NULL,
+                       ipcp_main_loop, _ipcp);
+
+        pthread_sigmask(SIG_UNBLOCK, &sigset, NULL);
+
+        rw_lock_unlock(&_ipcp->state_lock);
+
+        pthread_join(shim_data(_ipcp)->mainloop, NULL);
+
+        eth_llc_ipcp_data_destroy();
+
+        free(_ipcp);
+
+        exit(0);
+}
diff --git a/src/ipcpd/shim-eth-llc/shim_eth_llc_messages.proto b/src/ipcpd/shim-eth-llc/shim_eth_llc_messages.proto
new file mode 100644
index 00000000..3ea1af1f
--- /dev/null
+++ b/src/ipcpd/shim-eth-llc/shim_eth_llc_messages.proto
@@ -0,0 +1,15 @@
+enum shim_eth_llc_msg_code {
+        FLOW_REQ     = 1;
+        FLOW_REPLY   = 2;
+        FLOW_DEALLOC = 3;
+};
+
+message shim_eth_llc_msg {
+        required shim_eth_llc_msg_code code  = 1;
+        optional string dst_name             = 2;
+        optional string src_ap_name          = 3;
+        optional string src_ae_name          = 4;
+        required sint32 ssap                 = 5;
+        optional sint32 dsap                 = 6;
+        optional sint32 response             = 7;
+};