[tor-commits] [tor/release-0.2.2] Detect insanely large circuit build state; don't give its length to rand_int

arma at torproject.org arma at torproject.org
Mon Jun 20 01:04:13 UTC 2011


commit 42e4e156d95a1c28a666a5346d491c4ed71435dd
Author: Nick Mathewson <nickm at torproject.org>
Date:   Wed Jun 1 12:12:01 2011 -0400

    Detect insanely large circuit build state; don't give its length to rand_int
---
 changes/bug3306       |    4 ++++
 src/or/circuitbuild.c |   10 +++++++++-
 2 files changed, 13 insertions(+), 1 deletions(-)

diff --git a/changes/bug3306 b/changes/bug3306
index b1bb103..f868a24 100644
--- a/changes/bug3306
+++ b/changes/bug3306
@@ -3,3 +3,7 @@
       correctly. Previously, it accepted values up to UINT_MAX, but
       could return a negative number if given a value above INT_MAX+1.
       Found by George Kadianakis. Fixes bug 3306; bugfix on 0.2.2pre14.
+
+    - Avoid a segfault when reading a malformed circuit build state
+      with more than INT_MAX entries. Found by wanoskarnet. Bugfix on
+      0.2.2.4-alpha.
diff --git a/src/or/circuitbuild.c b/src/or/circuitbuild.c
index 3f08448..6d7e711 100644
--- a/src/or/circuitbuild.c
+++ b/src/or/circuitbuild.c
@@ -678,7 +678,15 @@ circuit_build_times_shuffle_and_store_array(circuit_build_times_t *cbt,
     log_notice(LD_CIRC, "The number of circuit times that this Tor version "
                "uses to calculate build times is less than the number stored "
                "in your state file. Decreasing the circuit time history from "
-               "%d to %d.", num_times, CBT_NCIRCUITS_TO_OBSERVE);
+               "%lu to %d.", (unsigned long)num_times,
+               CBT_NCIRCUITS_TO_OBSERVE);
+  }
+
+  if (n > INT_MAX-1) {
+    log_warn(LD_CIRC, "For some insane reasons, you had %lu circuit build "
+             "observations in your state file. That's far too many; probably "
+             "there's a bug here.", (unsigned long)n);
+    n = INT_MAX-1;
   }
 
   /* This code can only be run on a compact array */





More information about the tor-commits mailing list