From 870e3fdfaee4991592cc29e90767abd0e9fba43b Mon Sep 17 00:00:00 2001 From: Dimitri Staessens Date: Mon, 28 Aug 2023 12:28:59 +0200 Subject: ipcpd: Don't release sdb before writing packet The flow allocator fa_alloc_resp would release the packet buffer (sdb) before writing if the response was a failure. Also sets the IPCP allocation timeout in nanoseconds as per the comment. Signed-off-by: Dimitri Staessens Signed-off-by: Sander Vrijders --- src/ipcpd/unicast/fa.c | 45 +++++++++++++++++++++++++-------------------- 1 file changed, 25 insertions(+), 20 deletions(-) diff --git a/src/ipcpd/unicast/fa.c b/src/ipcpd/unicast/fa.c index 2591a162..52ef9684 100644 --- a/src/ipcpd/unicast/fa.c +++ b/src/ipcpd/unicast/fa.c @@ -56,7 +56,7 @@ #define CLOCK_REALTIME_COARSE CLOCK_REALTIME #endif -#define TIMEOUT 10000 /* nanoseconds */ +#define TIMEOUT 10 * MILLION /* nanoseconds */ #define FLOW_REQ 0 #define FLOW_REPLY 1 @@ -475,7 +475,7 @@ static int fa_wait_irmd_alloc(uint8_t * dst, const void * data, size_t len) { - struct timespec ts = {0, TIMEOUT * 1000}; + struct timespec ts = {0, TIMEOUT}; struct timespec abstime; int fd; time_t mpl = IPCP_UNICAST_MPL; @@ -516,7 +516,7 @@ static int fa_wait_irmd_alloc(uint8_t * dst, static int fa_wait_irmd_alloc_resp(int fd) { - struct timespec ts = {0, TIMEOUT * 1000}; + struct timespec ts = {0, TIMEOUT}; struct timespec abstime; clock_gettime(PTHREAD_COND_CLOCK, &abstime); @@ -882,43 +882,48 @@ int fa_alloc_resp(int fd, flow = &fa.flows[fd]; if (fa_wait_irmd_alloc_resp(fd) < 0) - return -1; + goto fail_alloc_resp; if (ipcp_sdb_reserve(&sdb, sizeof(*msg) + len)) { - fa_flow_fini(flow); - return -1; + goto fail_reserve; } msg = (struct fa_msg *) shm_du_buff_head(sdb); memset(msg, 0, sizeof(*msg)); - pthread_rwlock_wrlock(&fa.flows_lock); - msg->code = FLOW_REPLY; - msg->r_eid = hton64(flow->r_eid); - msg->s_eid = hton64(flow->s_eid); msg->response = response; - if (len > 0) memcpy(msg + 1, data, len); + pthread_rwlock_rdlock(&fa.flows_lock); + + msg->r_eid = hton64(flow->r_eid); + msg->s_eid = hton64(flow->s_eid); + + pthread_rwlock_unlock(&fa.flows_lock); + + if (dt_write_packet(flow->r_addr, qc, fa.eid, sdb)) + goto fail_packet; + if (response < 0) { + pthread_rwlock_rdlock(&fa.flows_lock); fa_flow_fini(flow); - ipcp_sdb_release(sdb); + pthread_rwlock_unlock(&fa.flows_lock); } else { psched_add(fa.psched, fd); } - if (dt_write_packet(flow->r_addr, qc, fa.eid, sdb)) { - fa_flow_fini(flow); - pthread_rwlock_unlock(&fa.flows_lock); - ipcp_sdb_release(sdb); - return -1; - } + return 0; + fail_packet: + ipcp_sdb_release(sdb); + fail_reserve: + pthread_rwlock_wrlock(&fa.flows_lock); + fa_flow_fini(flow); pthread_rwlock_unlock(&fa.flows_lock); - - return 0; + fail_alloc_resp: + return -1; } int fa_dealloc(int fd) -- cgit v1.2.3