[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
[tor-commits] [tor/release-0.2.2] Set timestamp_dirty on HS circuits as circuit_expire_building requires
commit 112d204fadab74426cc802e6d998a53f938a5248
Author: Robert Ransom <rransom.8774@xxxxxxxxx>
Date: Sun May 29 08:03:41 2011 -0700
Set timestamp_dirty on HS circuits as circuit_expire_building requires
Fixes part of #1297; bugfix on 48e0228f1e031a709c1deb149c7dfd187c3609cf,
when circuit_expire_building was changed to assume that timestamp_dirty
was set when a circuit changed purpose to _C_REND_READY. (It wasn't.)
---
changes/bug1297a | 8 ++++++++
src/or/or.h | 5 +++++
src/or/rendclient.c | 11 +++++++++++
3 files changed, 24 insertions(+), 0 deletions(-)
diff --git a/changes/bug1297a b/changes/bug1297a
new file mode 100644
index 0000000..459ef65
--- /dev/null
+++ b/changes/bug1297a
@@ -0,0 +1,8 @@
+ o Major bugfixes:
+ - Apply circuit timeouts to opened hidden-service-related circuits
+ based on the correct start time. Previously, we would apply the
+ circuit build timeout based on time since the circuit's
+ creation; it was supposed to be applied based on time since the
+ circuit entered its current state. Bugfix on 0.0.6; fixes part
+ of bug 1297.
+
diff --git a/src/or/or.h b/src/or/or.h
index 9cac5f3..6ccd957 100644
--- a/src/or/or.h
+++ b/src/or/or.h
@@ -2143,6 +2143,11 @@ typedef struct circuit_t {
* in time in order to indicate that a circuit shouldn't be used for new
* streams, but that it can stay alive as long as it has streams on it.
* That's a kludge we should fix.
+ *
+ * XXX023 The CBT code uses this field to record when HS-related
+ * circuits entered certain states. This usage probably won't
+ * interfere with this field's primary purpose, but we should
+ * document it more thoroughly to make sure of that.
*/
time_t timestamp_dirty;
diff --git a/src/or/rendclient.c b/src/or/rendclient.c
index 77e11c2..a542339 100644
--- a/src/or/rendclient.c
+++ b/src/or/rendclient.c
@@ -275,6 +275,10 @@ rend_client_send_introduction(origin_circuit_t *introcirc,
/* Now, we wait for an ACK or NAK on this circuit. */
introcirc->_base.purpose = CIRCUIT_PURPOSE_C_INTRODUCE_ACK_WAIT;
+ /* Set timestamp_dirty, because circuit_expire_building expects it
+ * to specify when a circuit entered the _C_INTRODUCE_ACK_WAIT
+ * state. */
+ introcirc->_base.timestamp_dirty = time(NULL);
return 0;
perm_err:
@@ -329,6 +333,10 @@ rend_client_introduction_acked(origin_circuit_t *circ,
circ->rend_data->onion_address, CIRCUIT_PURPOSE_C_REND_READY);
if (rendcirc) { /* remember the ack */
rendcirc->_base.purpose = CIRCUIT_PURPOSE_C_REND_READY_INTRO_ACKED;
+ /* Set timestamp_dirty, because circuit_expire_building expects
+ * it to specify when a circuit entered the
+ * _C_REND_READY_INTRO_ACKED state. */
+ rendcirc->_base.timestamp_dirty = time(NULL);
} else {
log_info(LD_REND,"...Found no rend circ. Dropping on the floor.");
}
@@ -674,6 +682,9 @@ rend_client_rendezvous_acked(origin_circuit_t *circ, const uint8_t *request,
log_info(LD_REND,"Got rendezvous ack. This circuit is now ready for "
"rendezvous.");
circ->_base.purpose = CIRCUIT_PURPOSE_C_REND_READY;
+ /* Set timestamp_dirty, because circuit_expire_building expects it
+ * to specify when a circuit entered the _C_REND_READY state. */
+ circ->_base.timestamp_dirty = time(NULL);
/* XXXX023 This is a pretty brute-force approach. It'd be better to
* attach only the connections that are waiting on this circuit, rather
* than trying to attach them all. See comments bug 743. */
_______________________________________________
tor-commits mailing list
tor-commits@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-commits