diff options
author | dimitri staessens <dimitri.staessens@intec.ugent.be> | 2016-10-31 18:35:17 +0100 |
---|---|---|
committer | dimitri staessens <dimitri.staessens@intec.ugent.be> | 2016-10-31 18:43:04 +0100 |
commit | 86b6d0ca1b871e629a95c9cf0ddceeeb7e1bd345 (patch) | |
tree | 02f048b2b148ad9426194967cbc332a93742ac52 | |
parent | add7d5f1e59c661b431ba86ba80d62badc14e762 (diff) | |
download | ouroboros-86b6d0ca1b871e629a95c9cf0ddceeeb7e1bd345.tar.gz ouroboros-86b6d0ca1b871e629a95c9cf0ddceeeb7e1bd345.zip |
lib, ipcpd: Improve flow allocation
All calls for opening rbuffs are now concentrated on the dev side.
This allows some simplifications in the np1 calls. The ipcp_fini call
will not destroy the mutex associated with the state, since the final
state needs to be checked before shutting down an IPCP.
-rw-r--r-- | include/ouroboros/np1_flow.h | 3 | ||||
-rw-r--r-- | src/ipcpd/ipcp.c | 16 | ||||
-rw-r--r-- | src/ipcpd/local/main.c | 20 | ||||
-rw-r--r-- | src/ipcpd/shim-eth-llc/main.c | 30 | ||||
-rw-r--r-- | src/ipcpd/shim-udp/main.c | 58 | ||||
-rw-r--r-- | src/lib/dev.c | 185 |
6 files changed, 169 insertions, 143 deletions
diff --git a/include/ouroboros/np1_flow.h b/include/ouroboros/np1_flow.h index 5b063213..4e7e6070 100644 --- a/include/ouroboros/np1_flow.h +++ b/include/ouroboros/np1_flow.h @@ -28,8 +28,7 @@ int np1_flow_alloc(pid_t n_api, int port_id); -int np1_flow_resp(pid_t n_api, - int port_id); +int np1_flow_resp(int port_id); int np1_flow_dealloc(int port_id); diff --git a/src/ipcpd/ipcp.c b/src/ipcpd/ipcp.c index b7eff3fb..4dff86f4 100644 --- a/src/ipcpd/ipcp.c +++ b/src/ipcpd/ipcp.c @@ -114,7 +114,6 @@ void ipcp_fini() ipcp_data_destroy(ipcpi.data); pthread_cond_destroy(&ipcpi.state_cond); - pthread_mutex_destroy(&ipcpi.state_mtx); pthread_rwlock_destroy(&ipcpi.state_lock); } @@ -349,7 +348,7 @@ void * ipcp_main_loop(void * o) } fd = np1_flow_alloc(msg->api, msg->port_id); if (fd < 0) { - LOG_ERR("Could not get fd for port_id. %d", + LOG_ERR("Failed allocating fd on port_id %d.", msg->port_id); ret_msg.has_result = true; ret_msg.result = -1; @@ -362,11 +361,6 @@ void * ipcp_main_loop(void * o) msg->dst_name, msg->src_ae_name, msg->qos_cube); - if (ret_msg.result < 0) { - LOG_DBG("Deallocate failed on port_id %d.", - msg->port_id); - flow_dealloc(fd); - } break; case IPCP_MSG_CODE__IPCP_FLOW_ALLOC_RESP: if (ipcpi.ops->ipcp_flow_alloc_resp == NULL) { @@ -375,10 +369,10 @@ void * ipcp_main_loop(void * o) } if (!msg->response) { - fd = np1_flow_resp(msg->api, msg->port_id); + fd = np1_flow_resp(msg->port_id); if (fd < 0) { - LOG_ERR("Could not get fd for port_id %d.", - msg->port_id); + LOG_WARN("Port_id %d is not known.", + msg->port_id); ret_msg.has_result = true; ret_msg.result = -1; break; @@ -397,7 +391,7 @@ void * ipcp_main_loop(void * o) fd = np1_flow_dealloc(msg->port_id); if (fd < 0) { - LOG_ERR("Could not deallocate port_id %d.", + LOG_WARN("Could not deallocate port_id %d.", msg->port_id); ret_msg.has_result = true; ret_msg.result = -1; diff --git a/src/ipcpd/local/main.c b/src/ipcpd/local/main.c index f95eeae4..f0c85084 100644 --- a/src/ipcpd/local/main.c +++ b/src/ipcpd/local/main.c @@ -88,14 +88,10 @@ static void * ipcp_local_sdu_loop(void * o) (void) o; - while (true) { + while (flow_event_wait(local_data.flows, local_data.fq, &timeout)) { int fd; ssize_t idx; - if (flow_event_wait(local_data.flows, local_data.fq, &timeout) - == -ETIMEDOUT) - continue; - pthread_rwlock_rdlock(&ipcpi.state_lock); if (ipcp_get_state() != IPCP_ENROLLED) { @@ -233,7 +229,8 @@ static int ipcp_local_flow_alloc(int fd, LOG_DBG("Allocating flow to %s on fd %d.", dst_name, fd); - assert(dst_name || src_ae_name); + assert(dst_name); + assert(src_ae_name); pthread_rwlock_rdlock(&ipcpi.state_lock); @@ -298,17 +295,24 @@ static int ipcp_local_flow_dealloc(int fd) ipcp_flow_fini(fd); pthread_rwlock_rdlock(&ipcpi.state_lock); + + if (ipcp_get_state() != IPCP_ENROLLED) { + pthread_rwlock_unlock(&ipcpi.state_lock); + LOG_DBG("Won't register with non-enrolled IPCP."); + return -1; /* -ENOTENROLLED */ + } + pthread_rwlock_wrlock(&local_data.lock); flow_set_del(local_data.flows, fd); local_data.in_out[fd] = -1; + flow_dealloc(fd); + pthread_rwlock_unlock(&local_data.lock); pthread_rwlock_unlock(&ipcpi.state_lock); - flow_dealloc(fd); - LOG_INFO("Flow with fd %d deallocated.", fd); return 0; diff --git a/src/ipcpd/shim-eth-llc/main.c b/src/ipcpd/shim-eth-llc/main.c index 0d4c3903..3f3c0e1e 100644 --- a/src/ipcpd/shim-eth-llc/main.c +++ b/src/ipcpd/shim-eth-llc/main.c @@ -624,18 +624,22 @@ static void * eth_llc_ipcp_sdu_writer(void * o) (void) o; - while (true) { - if (flow_event_wait(eth_llc_data.np1_flows, - eth_llc_data.fq, - &timeout) == -ETIMEDOUT) - continue; + while (flow_event_wait(eth_llc_data.np1_flows, + eth_llc_data.fq, + &timeout)) { + pthread_rwlock_rdlock(&ipcpi.state_lock); + + if (ipcp_get_state() != IPCP_ENROLLED) { + pthread_rwlock_unlock(&ipcpi.state_lock); + return (void *) -1; /* -ENOTENROLLED */ + } while ((fd = fqueue_next(eth_llc_data.fq)) >= 0) { if (ipcp_flow_read(fd, &sdb)) { LOG_ERR("Bad read from fd %d.", fd); continue; } - pthread_rwlock_rdlock(&ipcpi.state_lock); + pthread_rwlock_rdlock(ð_llc_data.flows_lock); ssap = reverse_bits(eth_llc_data.fd_to_ef[fd].sap); @@ -645,7 +649,6 @@ static void * eth_llc_ipcp_sdu_writer(void * o) MAC_SIZE); pthread_rwlock_unlock(ð_llc_data.flows_lock); - pthread_rwlock_unlock(&ipcpi.state_lock); eth_llc_ipcp_send_frame(r_addr, dsap, ssap, shm_du_buff_head(sdb), @@ -653,6 +656,8 @@ static void * eth_llc_ipcp_sdu_writer(void * o) - shm_du_buff_head(sdb)); ipcp_flow_del(sdb); } + + pthread_rwlock_unlock(&ipcpi.state_lock); } return (void *) 1; @@ -1045,6 +1050,13 @@ static int eth_llc_ipcp_flow_dealloc(int fd) ipcp_flow_fini(fd); pthread_rwlock_rdlock(&ipcpi.state_lock); + + if (ipcp_get_state() != IPCP_ENROLLED) { + pthread_rwlock_unlock(&ipcpi.state_lock); + LOG_DBG("Won't register with non-enrolled IPCP."); + return -1; /* -ENOTENROLLED */ + } + pthread_rwlock_wrlock(ð_llc_data.flows_lock); flow_set_del(eth_llc_data.np1_flows, fd); @@ -1058,11 +1070,11 @@ static int eth_llc_ipcp_flow_dealloc(int fd) eth_llc_data.ef_to_fd[sap] = -1; + flow_dealloc(fd); + pthread_rwlock_unlock(ð_llc_data.flows_lock); pthread_rwlock_unlock(&ipcpi.state_lock); - flow_dealloc(fd); - LOG_DBG("Flow with fd %d deallocated.", fd); return 0; diff --git a/src/ipcpd/shim-udp/main.c b/src/ipcpd/shim-udp/main.c index 503dbd0b..eff0bd94 100644 --- a/src/ipcpd/shim-udp/main.c +++ b/src/ipcpd/shim-udp/main.c @@ -452,7 +452,6 @@ static void * ipcp_udp_sdu_reader(void * o) continue; flags = fcntl(skfd, F_GETFL, 0); fcntl(skfd, F_SETFL, flags | O_NONBLOCK); - fd = udp_data.uf_to_fd[skfd]; n = sizeof(r_saddr); if ((n = recvfrom(skfd, &buf, @@ -462,7 +461,14 @@ static void * ipcp_udp_sdu_reader(void * o) (unsigned *) &n)) <= 0) continue; + pthread_rwlock_rdlock(&ipcpi.state_lock); + pthread_rwlock_rdlock(&udp_data.flows_lock); + + fd = udp_data.uf_to_fd[skfd]; flow_write(fd, buf, n); + + pthread_rwlock_unlock(&udp_data.flows_lock); + pthread_rwlock_unlock(&ipcpi.state_lock); } } @@ -477,11 +483,16 @@ static void * ipcp_udp_sdu_loop(void * o) (void) o; - while (true) { - if (flow_event_wait(udp_data.np1_flows, - udp_data.fq, - &timeout) == -ETIMEDOUT) - continue; + while (flow_event_wait(udp_data.np1_flows, udp_data.fq, &timeout)) { + pthread_rwlock_rdlock(&ipcpi.state_lock); + + if (ipcp_get_state() != IPCP_ENROLLED) { + pthread_rwlock_unlock(&ipcpi.state_lock); + return (void *) -1; /* -ENOTENROLLED */ + } + + + pthread_rwlock_rdlock(&udp_data.flows_lock); while ((fd = fqueue_next(udp_data.fq)) >= 0) { if (ipcp_flow_read(fd, &sdb)) { @@ -489,19 +500,18 @@ static void * ipcp_udp_sdu_loop(void * o) continue; } - pthread_rwlock_rdlock(&ipcpi.state_lock); - pthread_rwlock_rdlock(&udp_data.flows_lock); - if (send(udp_data.fd_to_uf[fd].skfd, shm_du_buff_head(sdb), shm_du_buff_tail(sdb) - shm_du_buff_head(sdb), 0) < 0) LOG_ERR("Failed to send SDU."); - pthread_rwlock_unlock(&udp_data.flows_lock); - pthread_rwlock_unlock(&ipcpi.state_lock); + ipcp_flow_del(sdb); + } - ipcp_flow_del(sdb); } + + pthread_rwlock_unlock(&udp_data.flows_lock); + pthread_rwlock_unlock(&ipcpi.state_lock); } return (void *) 1; @@ -948,8 +958,8 @@ static int ipcp_udp_flow_alloc(int fd, LOG_DBG("Allocating flow to %s.", dst_name); - if (dst_name == NULL || src_ae_name == NULL) - return -1; + assert(dst_name); + assert(src_ae_name); if (strlen(dst_name) > 255 || strlen(src_ae_name) > 255) { @@ -1101,32 +1111,36 @@ static int ipcp_udp_flow_dealloc(int fd) ipcp_flow_fini(fd); - flow_set_del(udp_data.np1_flows, fd); - pthread_rwlock_rdlock(&ipcpi.state_lock); + + if (ipcp_get_state() != IPCP_ENROLLED) { + pthread_rwlock_unlock(&ipcpi.state_lock); + LOG_DBG("Won't register with non-enrolled IPCP."); + return -1; /* -ENOTENROLLED */ + } + pthread_rwlock_wrlock(&udp_data.flows_lock); + flow_set_del(udp_data.np1_flows, fd); + skfd = udp_data.fd_to_uf[fd].skfd; udp_data.uf_to_fd[skfd] = -1; udp_data.fd_to_uf[fd].udp = -1; udp_data.fd_to_uf[fd].skfd = -1; + close(skfd); + pthread_rwlock_unlock(&udp_data.flows_lock); pthread_rwlock_rdlock(&udp_data.flows_lock); clr_fd(skfd); - pthread_rwlock_unlock(&udp_data.flows_lock); - pthread_rwlock_wrlock(&udp_data.flows_lock); - - close(skfd); + flow_dealloc(fd); pthread_rwlock_unlock(&udp_data.flows_lock); pthread_rwlock_unlock(&ipcpi.state_lock); - flow_dealloc(fd); - LOG_DBG("Flow with fd %d deallocated.", fd); return 0; diff --git a/src/lib/dev.c b/src/lib/dev.c index 4b97428e..fc8739a2 100644 --- a/src/lib/dev.c +++ b/src/lib/dev.c @@ -388,6 +388,14 @@ int flow_accept(char ** ae_name, struct qos_spec * qos) return -1; } + ai.flows[fd].tx_rb = shm_rbuff_open(recv_msg->api, recv_msg->port_id); + if (ai.flows[fd].tx_rb == NULL) { + reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + ai.flows[fd].set = shm_flow_set_open(recv_msg->api); if (ai.flows[fd].set == NULL) { reset_flow(fd); @@ -398,7 +406,6 @@ int flow_accept(char ** ae_name, struct qos_spec * qos) return -1; } - if (ae_name != NULL) { *ae_name = strdup(recv_msg->ae_name); if (*ae_name == NULL) { @@ -452,40 +459,34 @@ int flow_alloc_resp(int fd, int response) msg.port_id = ai.flows[fd].port_id; pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); msg.has_response = true; msg.response = response; recv_msg = send_recv_irm_msg(&msg); - if (recv_msg == NULL) { - pthread_rwlock_unlock(&ai.data_lock); + if (recv_msg == NULL) return -1; - } if (!recv_msg->has_result) { - pthread_rwlock_unlock(&ai.data_lock); irm_msg__free_unpacked(recv_msg, NULL); return -1; } ret = recv_msg->result; - pthread_rwlock_wrlock(&ai.flows_lock); + irm_msg__free_unpacked(recv_msg, NULL); + + if (response) { + pthread_rwlock_rdlock(&ai.data_lock); + pthread_rwlock_wrlock(&ai.flows_lock); - ai.flows[fd].tx_rb = shm_rbuff_open(ai.flows[fd].api, - ai.flows[fd].port_id); - if (ai.flows[fd].tx_rb == NULL) { reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); - return -1; } - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - - irm_msg__free_unpacked(recv_msg, NULL); - return ret; } @@ -535,9 +536,6 @@ int flow_alloc(char * dst_name, char * src_ae_name, struct qos_spec * qos) return -1; } - ai.flows[fd].port_id = recv_msg->port_id; - ai.flows[fd].oflags = FLOW_O_DEFAULT; - ai.flows[fd].api = recv_msg->api; ai.flows[fd].rx_rb = shm_rbuff_open(ai.api, recv_msg->port_id); if (ai.flows[fd].rx_rb == NULL) { reset_flow(fd); @@ -548,6 +546,25 @@ int flow_alloc(char * dst_name, char * src_ae_name, struct qos_spec * qos) return -1; } + ai.flows[fd].tx_rb = shm_rbuff_open(recv_msg->api, recv_msg->port_id); + if (ai.flows[fd].tx_rb == NULL) { + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + + ai.flows[fd].set = shm_flow_set_open(recv_msg->api); + if (ai.flows[fd].set == NULL) { + reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + + ai.flows[fd].port_id = recv_msg->port_id; + ai.flows[fd].oflags = FLOW_O_DEFAULT; + ai.flows[fd].api = recv_msg->api; + ai.ports[recv_msg->port_id].fd = fd; ai.ports[recv_msg->port_id].state = PORT_ID_ASSIGNED; @@ -582,22 +599,6 @@ int flow_alloc_res(int fd) msg.port_id = ai.flows[fd].port_id; - ai.flows[fd].tx_rb = shm_rbuff_open(ai.flows[fd].api, - ai.flows[fd].port_id); - if (ai.flows[fd].tx_rb == NULL) { - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } - - ai.flows[fd].set = shm_flow_set_open(ai.flows[fd].api); - if (ai.flows[fd].set == NULL) { - reset_flow(fd); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } - pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); @@ -764,10 +765,14 @@ ssize_t flow_write(int fd, void * buf, size_t count) } } else { /* blocking */ struct shm_rdrbuff * rdrb = ai.rdrb; - pid_t api = ai.flows[fd].api; + struct shm_rbuff * tx_rb = ai.flows[fd].tx_rb; + pid_t api = ai.flows[fd].api; + pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); + assert(tx_rb); + idx = shm_rdrbuff_write_b(rdrb, api, DU_BUFF_HEADSPACE, @@ -775,15 +780,13 @@ ssize_t flow_write(int fd, void * buf, size_t count) buf, count); - pthread_rwlock_rdlock(&ai.data_lock); - pthread_rwlock_rdlock(&ai.flows_lock); - - if (shm_rbuff_write(ai.flows[fd].tx_rb, idx) < 0) { - shm_rdrbuff_remove(ai.rdrb, idx); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); + if (shm_rbuff_write(tx_rb, idx) < 0) { + shm_rdrbuff_remove(rdrb, idx); return -ENOTALLOC; } + + pthread_rwlock_rdlock(&ai.data_lock); + pthread_rwlock_rdlock(&ai.flows_lock); } shm_flow_set_notify(ai.flows[fd].set, ai.flows[fd].port_id); @@ -993,7 +996,7 @@ int flow_event_wait(struct flow_set * set, { ssize_t ret; - if (set == NULL) + if (set == NULL || fq == NULL) return -EINVAL; if (fq->fqsize > 0) @@ -1038,6 +1041,22 @@ int np1_flow_alloc(pid_t n_api, int port_id) return -1; } + ai.flows[fd].tx_rb = shm_rbuff_open(n_api, port_id); + if (ai.flows[fd].tx_rb == NULL) { + reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + + ai.flows[fd].set = shm_flow_set_open(n_api); + if (ai.flows[fd].set == NULL) { + reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + ai.flows[fd].port_id = port_id; ai.flows[fd].oflags = FLOW_O_DEFAULT; ai.flows[fd].api = n_api; @@ -1066,8 +1085,7 @@ int np1_flow_dealloc(int port_id) return fd; } - -int np1_flow_resp(pid_t n_api, int port_id) +int np1_flow_resp(int port_id) { int fd; @@ -1077,28 +1095,6 @@ int np1_flow_resp(pid_t n_api, int port_id) pthread_rwlock_wrlock(&ai.flows_lock); fd = ai.ports[port_id].fd; - if (fd < 0) { - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return fd; - } - - ai.flows[fd].tx_rb = shm_rbuff_open(n_api, port_id); - if (ai.flows[fd].tx_rb == NULL) { - reset_flow(fd); - port_destroy(&ai.ports[port_id]); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } - - ai.flows[fd].set = shm_flow_set_open(n_api); - if (ai.flows[fd].set == NULL) { - reset_flow(fd); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); @@ -1166,21 +1162,47 @@ int ipcp_flow_req_arr(pid_t api, char * dst_name, char * src_ae_name) if (recv_msg == NULL) return -1; - if (!recv_msg->has_port_id) { + if (!recv_msg->has_port_id || !recv_msg->has_api) { irm_msg__free_unpacked(recv_msg, NULL); return -1; } + if (recv_msg->has_result && recv_msg->result) { + irm_msg__free_unpacked(recv_msg, NULL); + return -1; + } + port_id = recv_msg->port_id; - irm_msg__free_unpacked(recv_msg, NULL); - if (port_id < 0) + if (port_id < 0) { + irm_msg__free_unpacked(recv_msg, NULL); return -1; + } pthread_rwlock_rdlock(&ai.data_lock); pthread_rwlock_wrlock(&ai.flows_lock); ai.flows[fd].rx_rb = shm_rbuff_open(ai.api, port_id); if (ai.flows[fd].rx_rb == NULL) { + irm_msg__free_unpacked(recv_msg, NULL); + reset_flow(fd); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + + ai.flows[fd].tx_rb = shm_rbuff_open(recv_msg->api, port_id); + if (ai.flows[fd].tx_rb == NULL) { + irm_msg__free_unpacked(recv_msg, NULL); + reset_flow(fd); + port_destroy(&ai.ports[port_id]); + pthread_rwlock_unlock(&ai.flows_lock); + pthread_rwlock_unlock(&ai.data_lock); + return -1; + } + + ai.flows[fd].set = shm_flow_set_open(recv_msg->api); + if (ai.flows[fd].set == NULL) { + irm_msg__free_unpacked(recv_msg, NULL); reset_flow(fd); pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); @@ -1196,6 +1218,8 @@ int ipcp_flow_req_arr(pid_t api, char * dst_name, char * src_ae_name) pthread_rwlock_unlock(&ai.flows_lock); pthread_rwlock_unlock(&ai.data_lock); + irm_msg__free_unpacked(recv_msg, NULL); + return fd; } @@ -1228,27 +1252,6 @@ int ipcp_flow_alloc_reply(int fd, int response) ret = recv_msg->result; - pthread_rwlock_wrlock(&ai.flows_lock); - - ai.flows[fd].tx_rb = shm_rbuff_open(ai.flows[fd].api, - ai.flows[fd].port_id); - if (ai.flows[fd].tx_rb == NULL) { - reset_flow(fd); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } - - ai.flows[fd].set = shm_flow_set_open(ai.flows[fd].api); - if (ai.flows[fd].set == NULL) { - reset_flow(fd); - pthread_rwlock_unlock(&ai.flows_lock); - pthread_rwlock_unlock(&ai.data_lock); - return -1; - } - - pthread_rwlock_unlock(&ai.flows_lock); - irm_msg__free_unpacked(recv_msg, NULL); return ret; |