[or-cvs] backport the hack to cull wedged cpuworkers
Roger Dingledine
arma at seul.org
Thu Apr 7 21:23:12 UTC 2005
Update of /home2/or/cvsroot/tor/src/or
In directory moria.mit.edu:/home2/arma/work/onion/0091/tor/src/or
Modified Files:
Tag: tor-0_0_9-patches
cpuworker.c
Log Message:
backport the hack to cull wedged cpuworkers
Index: cpuworker.c
===================================================================
RCS file: /home2/or/cvsroot/tor/src/or/cpuworker.c,v
retrieving revision 1.61.2.2
retrieving revision 1.61.2.3
diff -u -d -r1.61.2.2 -r1.61.2.3
--- cpuworker.c 28 Jan 2005 09:59:23 -0000 1.61.2.2
+++ cpuworker.c 7 Apr 2005 21:23:10 -0000 1.61.2.3
@@ -347,6 +347,34 @@
log_fn(LOG_WARN,"assign_to_cpuworker failed. Ignoring.");
}
+#define CPUWORKER_BUSY_TIMEOUT 100 /* seconds */
+
+/** We have a bug that I can't find. Sometimes, very rarely, cpuworkers
+ * get stuck in the 'busy' state, even though the cpuworker process
+ * thinks of itself as idle. I don't know why. But here's a workaround
+ * to kill any cpuworker that's been busy for more than 100 seconds. */
+static void
+cull_wedged_cpuworkers(void) {
+ connection_t **carray;
+ connection_t *conn;
+ int n_conns, i;
+ time_t now = time(NULL);
+
+ get_connection_array(&carray, &n_conns);
+ for (i = 0; i < n_conns; ++i) {
+ conn = carray[i];
+ if (!conn->marked_for_close &&
+ conn->type == CONN_TYPE_CPUWORKER &&
+ conn->state == CPUWORKER_STATE_BUSY_ONION &&
+ conn->timestamp_lastwritten + CPUWORKER_BUSY_TIMEOUT > now) {
+ log_fn(LOG_NOTICE,"Bug: closing wedged cpuworker. Can somebody find the bug?");
+ num_cpuworkers_busy--;
+ num_cpuworkers--;
+ connection_mark_for_close(conn);
+ }
+ }
+}
+
/** if cpuworker is defined, assert that he's idle, and use him. else,
* look for an idle cpuworker and use him. if none idle, queue task onto
* the pending onion list and return.
@@ -360,6 +388,9 @@
tor_assert(question_type == CPUWORKER_TASK_ONION);
+ cull_wedged_cpuworkers();
+ spawn_enough_cpuworkers();
+
if (question_type == CPUWORKER_TASK_ONION) {
circ = task;
More information about the tor-commits
mailing list