Latest push segfaults
Hi Nick, the commits you pushed this morning cause stem's integ tests to break the tor process with...
May 14 08:12:35.000 [notice] New control connection opened from 127.0.0.1.
May 14 08:12:36.000 [warn] Unrecognized service ID wgrobdz45bwja5bk on introduction circuit 4081277099.
May 14 08:12:36.000 [warn] Unrecognized service ID 7piyq5klvts2bl2v on introduction circuit 4179207724.
May 14 08:12:36.000 [warn] Unrecognized service ID wgrobdz45bwja5bk on introduction circuit 4045035595.
============================================================ T= 1431616356
Tor 0.2.7.1-alpha-dev (git-d05d21c89a21695b) died: Caught signal 11
tor(+0x126b5e)[0xb7725b5e]
tor(+0xa89db)[0xb76a79db]
tor(+0xa89db)[0xb76a79db]
tor(circuit_close_all_marked+0x56)[0xb76a7c16]
tor(+0x29574)[0xb7628574]
tor(+0x153bd1)[0xb7752bd1]
/usr/lib/libevent-2.0.so.5(event_base_loop+0x209)[0xb7549d39]
tor(do_main_loop+0x1b3)[0xb7629c63]
tor(tor_main+0x1f7d)[0xb762d64d]
tor(main+0x33)[0xb7625ef3]
/lib/i386-linux-gnu/libc.so.6(__libc_start_main+0xf3)[0xb71854d3]
tor(+0x26f3d)[0xb7625f3d]
This should be easily reproduceable (simply 'run_tests.py --integ' does the trick, though your spiffy new make target should do the trick too). I can narrow it down to a particular controller repro if needed.