[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

[tor-commits] [tor/master] sched: Always initialize scheduler_last_run to now



commit 070064de899901daaf98aefc3335bc9f21acb5c6
Author: David Goulet <dgoulet@xxxxxxxxxxxxxx>
Date:   Fri Sep 29 14:07:55 2017 -0400

    sched: Always initialize scheduler_last_run to now
    
    Because our monotonic time interface doesn't play well with value set to 0,
    always initialize to now() the scheduler_last_run at init() of the KIST
    scheduler.
    
    Fixes #23696
    
    Signed-off-by: David Goulet <dgoulet@xxxxxxxxxxxxxx>
---
 changes/bug23696        |  3 +++
 src/or/scheduler_kist.c | 10 ++++++++++
 2 files changed, 13 insertions(+)

diff --git a/changes/bug23696 b/changes/bug23696
new file mode 100644
index 000000000..d447815c8
--- /dev/null
+++ b/changes/bug23696
@@ -0,0 +1,3 @@
+  o Minor bugfixes (scheduler, kist):
+    - Properly initialize the scheduler last run time counter so it is not 0
+      at the first tick. Fixes bug 23696; bugfix on 0.3.2.1-alpha.
diff --git a/src/or/scheduler_kist.c b/src/or/scheduler_kist.c
index 9e960cdef..5856cbea4 100644
--- a/src/or/scheduler_kist.c
+++ b/src/or/scheduler_kist.c
@@ -490,6 +490,16 @@ kist_scheduler_on_new_options(void)
 static void
 kist_scheduler_init(void)
 {
+  /* When initializing the scheduler, the last run could be 0 because it is
+   * declared static or a value in the past that was set when it was last
+   * used. In both cases, we want to initialize it to now so we don't risk
+   * using the value 0 which doesn't play well with our monotonic time
+   * interface.
+   *
+   * One side effect is that the first scheduler run will be at the next tick
+   * that is in now + 10 msec (KIST_SCHED_RUN_INTERVAL_DEFAULT) by default. */
+  monotime_get(&scheduler_last_run);
+
   kist_scheduler_on_new_options();
   IF_BUG_ONCE(sched_run_interval == 0) {
     log_warn(LD_SCHED, "We are initing the KIST scheduler and noticed the "



_______________________________________________
tor-commits mailing list
tor-commits@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-commits