A possible data raise might mean some mutex lock is missing.
on running Valgrind Helgrind, output shows:
==62338== Possible data race during write of size 1 at 0x1FC8CB94 by thread #2==62338== Locks held: 1, at address 0x1FC8CB38==62338== at 0x4B129E5: libwsclient_handshake_thread (wsclient.c:792)==62338== by 0x484F30D: mythread_wrapper (hg_intercepts.c:406)==62338== by 0x1C467B19: start_thread (pthread_create.c:443)==62338== by 0x1C4EB8F3: clone (clone.S:100)==62338====62338== This conflicts with a previous read of size 1 by thread #1==62338== Locks held: none==62338== at 0x4B1157A: libwsclient_run (wsclient.c:50)
A possible data raise might mean some mutex lock is missing. on running Valgrind Helgrind, output shows:
==62338== Possible data race during write of size 1 at 0x1FC8CB94 by thread #2==62338== Locks held: 1, at address 0x1FC8CB38==62338== at 0x4B129E5: libwsclient_handshake_thread (wsclient.c:792)==62338== by 0x484F30D: mythread_wrapper (hg_intercepts.c:406)==62338== by 0x1C467B19: start_thread (pthread_create.c:443)==62338== by 0x1C4EB8F3: clone (clone.S:100)==62338====62338== This conflicts with a previous read of size 1 by thread #1==62338== Locks held: none==62338== at 0x4B1157A: libwsclient_run (wsclient.c:50)Fairly simple to resolve:
change:
if(pthread_create(&client->handshake_thread, NULL, libwsclient_handshake_thread, (void *)client)) {by:
ibwsclient_handshake_thread((void *)client);omitting the threading for the handshake solves it.