Realized yielding is actually faster by a lot
This commit is contained in:
parent
4dbaf216b3
commit
8642c075ea
@ -56,7 +56,7 @@ errval_t do_aos_urpc(
|
||||
// wait until the rpc call completes
|
||||
while(rpc->meta->call_in_progress) {
|
||||
// yield the thread because there might be useful stuff to do...
|
||||
// thread_yield();
|
||||
thread_yield();
|
||||
}
|
||||
|
||||
// memory barrier
|
||||
@ -180,9 +180,8 @@ int urpc_server(void *arg) {
|
||||
while (true) {
|
||||
// wait until an rpc call arrives
|
||||
while(urpc->meta->call_in_progress == false) {
|
||||
// sleep for a bit while there is nothing to do
|
||||
// barrelfish_usleep(100);
|
||||
// thread_yield();
|
||||
// yield the thread because there might be useful stuff to do...
|
||||
thread_yield();
|
||||
}
|
||||
|
||||
// memory barrier
|
||||
|
||||
Loading…
Reference in New Issue
Block a user