kgiusti commented on a change in pull request #760:
URL: https://github.com/apache/qpid-dispatch/pull/760#discussion_r438992885
##########
File path: src/posix/threading.c
##########
@@ -150,19 +150,38 @@ void sys_rwlock_unlock(sys_rwlock_t *lock)
struct sys_thread_t {
pthread_t thread;
+ void *(*f)(void *);
+ void *arg;
};
+static __thread sys_thread_t *_self;
+
+
+// bootstrap _self before calling main thread function
+//
+static void *_thread_init(void *arg)
+{
+ _self = (sys_thread_t *)arg;
+ return _self->f(_self->arg);
+}
+
+
sys_thread_t *sys_thread(void *(*run_function) (void *), void *arg)
{
sys_thread_t *thread = NEW(sys_thread_t);
- pthread_create(&(thread->thread), 0, run_function, arg);
+ thread->f = run_function;
+ thread->arg = arg;
+ pthread_create(&(thread->thread), 0, _thread_init, (void *)thread);
return thread;
}
-long sys_thread_id(sys_thread_t *thread) {
- return (long) thread->thread;
+
Review comment:
Yes and Yes!
There were no calls to sys_thread_id() in the code... so zero risk in
removing it.
Each thread is allocated a unique sys_thread_t pointer that exists in shared
memory so every thread's pointer will be globally unique. One of the new unit
tests verifies that is the case.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
