Closed Bug 819039 Opened 13 years ago Closed 12 years ago

Intermittent crash in transport_unittests [@ nr_socket_recvfrom]

Categories

(Core :: WebRTC: Networking, defect, P5)

Product:
Core
Component:
WebRTC: Networking
Type:
defect

Tracking

()

Status:
RESOLVED FIXED
Tracking Flags:
Tracking Status
firefox20 - ---

People

(Reporter: emorley, Assigned: ekr)

References

Details

(4 keywords, Whiteboard: [WebRTC][blocking-webrtc+][qa-])

Crash Data

Linux x86-64 mozilla-inbound build on 2012-12-03 03:22:33 PST for push 3559bc4dfe1b slave: bld-linux64-ec2-635 https://tbpl.mozilla.org/php/getParsedLog.php?id=17553540&tree=Mozilla-Inbound { mozcrash INFO | PROCESS-CRASH | transport_unittests | application crashed (minidump found) mozcrash INFO | Crash dump filename: /tmp/tmpHqR5aV/50e958c8-d9ad-8212-137e1b9e-3ef789a5.dmp mozcrash INFO | Saved dump as /builds/slave/m-in-lnx64/minidumps/50e958c8-d9ad-8212-137e1b9e-3ef789a5.dmp cppunittests TEST-UNEXPECTED-FAIL | transport_unittests | test crashed cppunittests INFO | Running test runnable_utils_unittest To run this test set MOZ_WEBRTC_TESTS=1 in your environment Running tests... Setting up crash reporting [==========] Running 3 tests from 1 test case. [----------] Global test environment set-up. [----------] 3 tests from TimerTest [ RUN ] TimerTest.SimpleTimer Timer fired [ OK ] TimerTest.SimpleTimer (200 ms) [ RUN ] TimerTest.CancelTimer [ OK ] TimerTest.CancelTimer (2000 ms) [ RUN ] TimerTest.ScheduleTest Timer fired [ OK ] TimerTest.ScheduleTest (200 ms) [----------] 3 tests from TimerTest (2402 ms total) [----------] Global test environment tear-down [==========] 3 tests from 1 test case ran. (2404 ms total) [ PASSED ] 3 tests. Finished running tests. Running tests... Setting up crash reporting [==========] Running 2 tests from 1 test case. [----------] Global test environment set-up. [----------] 2 tests from SocketTransportServiceTest [ RUN ] SocketTransportServiceTest.SendEvent [ OK ] SocketTransportServiceTest.SendEvent (0 ms) [ RUN ] SocketTransportServiceTest.SendPacket [ OK ] SocketTransportServiceTest.SendPacket (0 ms) [----------] 2 tests from SocketTransportServiceTest (1 ms total) [----------] Global test environment tear-down [==========] 2 tests from 1 test case ran. (2 ms total) [ PASSED ] 2 tests. Finished running tests. Running tests... Setting up crash reporting [==========] Running 12 tests from 1 test case. [----------] Global test environment set-up. [----------] 12 tests from TransportTest [ RUN ] TransportTest.TestNoDtlsVerificationSettings [ OK ] TransportTest.TestNoDtlsVerificationSettings (282 ms) [ RUN ] TransportTest.TestConnect [ OK ] TransportTest.TestConnect (623 ms) [ RUN ] TransportTest.TestConnectAllowAll [ OK ] TransportTest.TestConnectAllowAll (613 ms) [ RUN ] TransportTest.TestConnectBadDigest [ OK ] TransportTest.TestConnectBadDigest (428 ms) [ RUN ] TransportTest.TestConnectTwoDigests [ OK ] TransportTest.TestConnectTwoDigests (357 ms) [ RUN ] TransportTest.TestConnectTwoDigestsFirstBad [ OK ] TransportTest.TestConnectTwoDigestsFirstBad (373 ms) [ RUN ] TransportTest.TestConnectTwoDigestsSecondBad [ OK ] TransportTest.TestConnectTwoDigestsSecondBad (466 ms) [ RUN ] TransportTest.TestConnectTwoDigestsBothBad [ OK ] TransportTest.TestConnectTwoDigestsBothBad (344 ms) [ RUN ] TransportTest.TestTransfer Received == 0 Received 1000 bytes [ OK ] TransportTest.TestTransfer (546 ms) [ RUN ] TransportTest.TestConnectLoseFirst Received == 0 Received 1000 bytes [ OK ] TransportTest.TestConnectLoseFirst (1583 ms) [ RUN ] TransportTest.TestConnectIce Got candidate candidate:0 1 UDP 2113667327 10.132.66.92 55957 typ host Got candidate candidate:1 1 UDP 1694302207 63.245.214.82 54265 typ srflx raddr 10.132.66.92 rport 55957 Got candidate candidate:0 1 UDP 2113667327 10.132.66.92 53069 typ host Got candidate candidate:1 1 UDP 1694302207 63.245.214.82 30882 typ srflx raddr 10.132.66.92 rport 53069 Now connected Now connected Operating system: Linux 0.0.0 Linux 2.6.32-220.23.1.el6.centos.plus.x86_64 #1 SMP Tue Jun 19 04:14:37 BST 2012 x86_64 CPU: amd64 family 6 model 26 stepping 5 8 CPUs Crash reason: SIGSEGV Crash address: 0x8 Thread 3 (crashed) 0 transport_unittests!nr_socket_recvfrom [nr_socket.c : 85 + 0x0] rbx = 0x00007f29efe69c8c r12 = 0x0000000000000000 r13 = 0x00007f29ee3fc810 r14 = 0x00007f29ee3fe980 r15 = 0x0000000000000010 rip = 0x0000000000429c56 rsp = 0x00007f29ee3fc7b0 rbp = 0x00007f29ee3fea30 Found by: given as instruction pointer in context 1 transport_unittests!nr_ice_socket_readable_cb [ice_socket.c : 63 + 0x20] rbx = 0x00007f29efe69c8c r12 = 0x0000000000000000 r13 = 0x00007f29ee3fc810 r14 = 0x00007f29ee3fe980 r15 = 0x0000000000000010 rip = 0x00000000004295de rsp = 0x00007f29ee3fc7c0 rbp = 0x00007f29ee3fea30 Found by: call frame info 2 transport_unittests!mozilla::NrSocket::OnSocketReady(PRFileDesc*, short) [nr_socket_prsock.cpp : 124 + 0x6] rbx = 0x00007f29ee4bdea0 r12 = 0x0000000000000001 r13 = 0x0000000000000003 r14 = 0x0000000000000000 r15 = 0x0000000000000010 rip = 0x000000000041c605 rsp = 0x00007f29ee3fea40 rbp = 0x00007f29ee3fea50 Found by: call frame info 3 libxul.so!nsSocketTransportService::DoPollIteration(bool) [nsSocketTransportService2.cpp : 807 + 0xb] rbx = 0x00007f29efecee20 r12 = 0x0000000000000000 r13 = 0x0000000000000003 r14 = 0x0000000000000000 r15 = 0x0000000000000010 rip = 0x00007f29f863752a rsp = 0x00007f29ee3fea60 rbp = 0x00007f29ee3feab0 Found by: call frame info 4 libxul.so!nsSocketTransportService::Run() [nsSocketTransportService2.cpp : 646 + 0x12] rbx = 0x00007f29efecee20 r12 = 0x00007f29efe67340 r13 = 0x00007f29ee3feacf r14 = 0x00007f29efecee68 r15 = 0x0000000000000000 rip = 0x00007f29f8637a29 rsp = 0x00007f29ee3feac0 rbp = 0x00007f29ee3feb00 Found by: call frame info 5 libxul.so!nsThread::ProcessNextEvent(bool, bool*) [nsThread.cpp : 627 + 0x5] rbx = 0x00007f29efe67340 r12 = 0x00007f29ee3fec01 r13 = 0x0000000000000000 r14 = 0x0000000000000000 r15 = 0x0000000000000003 rip = 0x00007f29f920ea49 rsp = 0x00007f29ee3feb10 rbp = 0x00007f29ee3febb0 Found by: call frame info 6 libxul.so!NS_ProcessNextEvent_P(nsIThread*, bool) [nsThreadUtils.cpp : 221 + 0xc] rbx = 0x00007f29efe67301 r12 = 0x00007f29ee3fec00 r13 = 0x00007f29efe67390 r14 = 0x0000000000000000 }
So we had this already as bug 814164 which was marked as fixed a couple of days ago. Here a frame is different but all other ones look the same. So bug 814164 is not fully fixed I assume.
Fixed for bug 814164 meant the test was disabled, so the root cause bug isn't fixed.
Will mark in-testsuite+ only cause this is already covered by an existing test that should be enabled when this is fixed.
Flags: in-testsuite+
Whiteboard: [WebRTC][blocking-webrtc+]
probably fixed by bug 792175; please retest (if it's enabled on alder (I think so), you can just push repeated build-only tries there of just linux64, or if repeatable locally just run the test in a loop)
Assignee: nobody → jsmith
Keywords: qawanted
Exactly the same stack I still see with attachment 692807 [details] (bug 822158). Just load the testcase and wait 5 seconds. Report: bp-7d1f299b-d0b2-45b3-9a9a-3290a2121217 0 XUL nr_socket_recvfrom nr_socket.c:85 1 XUL nr_ice_socket_readable_cb ice_socket.c:63 2 XUL mozilla::NrSocket::OnSocketReady nr_socket_prsock.cpp:190 3 XUL nsSocketTransportService::DoPollIteration nsSocketTransportService2.cpp:802 4 XUL nsSocketTransportService::Run nsSocketTransportService2.cpp:641 5 XUL _ZThn24_N24nsSocketTransportService3RunEv nsSocketTransportService2.cpp:685 6 XUL nsThread::ProcessNextEvent nsThread.cpp:627 7 XUL NS_ProcessNextEvent_P nsThreadUtils.cpp:237 8 XUL nsThread::ThreadFunc nsThread.cpp:265 9 libnspr4.dylib _pt_root ptthread.c:156
Assignee: jsmith → nobody
tracking-firefox20: --- → ?
Keywords: qawanted → reproducible, testcase
OS: Linux → All
Hardware: x86_64 → All
Flags: in-testsuite+ → in-testsuite?
whimboo: really? I didn't see that signature anywhere in that bug.
Assignee: nobody → ekr
The reporter ran this testcase on Win7 which could be the reason why he got another stack for his crash.
Don't see any reason to track this intermittent crash in testsuite, but will to look at uplift nomination any time one is available if it helps green the tree up.
tracking-firefox20: ? → -
Real bug, null-deref (sock or sock->vtbl). Likely fixed by bug 820102; we've seen none since that landed on ~12/21-22. If we don't see this again for a while we'll close it.
Priority: -- → P5
Status: NEW → RESOLVED
Closed: 12 years ago
Depends on: 820102
Resolution: --- → FIXED
Whiteboard: [WebRTC][blocking-webrtc+] → [WebRTC][blocking-webrtc+][qa-]
Flipping back to testsuite+ - there's already test coverage here. The renom isn't right.
Flags: in-testsuite? → in-testsuite+
You need to log in before you can comment on or make changes to this bug.