Browse Source

Only restart threads that are not stuck waiting on lp staged work, and add one mandatory queued work item for each restart.

nfactor-troky
Con Kolivas 14 years ago
parent
commit
ba4ba7f65c
  1. 10
      main.c

10
main.c

@ -1924,8 +1924,16 @@ static void *watchdog_thread(void *userdata)
for (i = 0; i < mining_threads; i++) { for (i = 0; i < mining_threads; i++) {
struct thr_info *thr = &thr_info[i]; struct thr_info *thr = &thr_info[i];
if (now.tv_sec - thr->last.tv_sec > 60) { /* Do not kill threads waiting on longpoll staged work */
if (now.tv_sec - thr->last.tv_sec > 60 && !lp_staged) {
applog(LOG_ERR, "Attempting to restart thread %d, idle for more than 60 seconds", i); applog(LOG_ERR, "Attempting to restart thread %d, idle for more than 60 seconds", i);
/* Create one mandatory work item */
inc_staged(1, true);
if (unlikely(!queue_request())) {
applog(LOG_ERR, "Failed to queue_request in watchdog_thread");
kill_work();
break;
}
reinit_thread(i); reinit_thread(i);
applog(LOG_WARNING, "Thread %d restarted", i); applog(LOG_WARNING, "Thread %d restarted", i);
} }

Loading…
Cancel
Save