In circuit_build_times_calculate_timeout() in circuitstats.c, avoid dividing
by zero in the pareto calculations.
If either the alpha or p parameters are 0, we would divide by zero, yielding
an infinite result; which would be clamped to INT32_MAX anyway. So rather
than dividing by zero, we just skip the offending calculation(s), and
use INT32_MAX for the result.
Division by zero traps under clang -fsanitize=undefined-trap -fsanitize-undefined-trap-on-error.
--- /dev/null
+ o Minor bugfixes:
+ - In circuit_build_times_calculate_timeout() in circuitstats.c, avoid
+ dividing by zero in the pareto calculations. This traps under
+ clang -fsanitize=undefined-trap -fsanitize-undefined-trap-on-error.
+ Fixes bug 13290.
tor_assert(1.0-quantile > 0);
tor_assert(cbt->Xm > 0);
- ret = cbt->Xm/pow(1.0-quantile,1.0/cbt->alpha);
+ /* If either alpha or p are 0, we would divide by zero, yielding an
+ * infinite (double) result; which would be clamped to INT32_MAX.
+ * Instead, initialise ret to INT32_MAX, and skip over these
+ * potentially illegal/trapping divides by zero.
+ */
+ ret = INT32_MAX;
+
+ if (cbt->alpha > 0) {
+ double p;
+ p = pow(1.0-quantile,1.0/cbt->alpha);
+ if (p > 0) {
+ ret = cbt->Xm/p;
+ }
+ }
+
if (ret > INT32_MAX) {
ret = INT32_MAX;
}