➜
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2020-03-16 21:59 by vstinner, last changed 2022-04-11 14:59 by admin. This issue is now closed.
| Messages (12) | |||
|---|---|---|---|
| msg364362 - (view) | Author: STINNER Victor (vstinner) *  |
Date: 2020-03-16 21:59 | |
AMD64 FreeBSD Shared 3.x: https://buildbot.python.org/all/#/builders/152/builds/409 ====================================================================== ERROR: testSendmsg (test.test_socket.SendmsgSCTPStreamTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 342, in _setUp self.__setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 2716, in setUp super().setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 2533, in setUp super().setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 646, in setUp conn, addr = self.serv.accept() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/socket.py", line 293, in accept fd, addr = self._accept() ConnectionAbortedError: [Errno 53] Software caused connection abort (...) ====================================================================== ERROR: testRecvmsgAfterClose (test.test_socket.RecvmsgSCTPStreamTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 342, in _setUp self.__setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 2533, in setUp super().setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 646, in setUp conn, addr = self.serv.accept() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/socket.py", line 293, in accept fd, addr = self._accept() ConnectionAbortedError: [Errno 53] Software caused connection abort (...) ====================================================================== ERROR: testRecvmsgIntoGenerator (test.test_socket.RecvmsgIntoSCTPStreamTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 342, in _setUp self.__setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 2533, in setUp super().setUp() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/test/test_socket.py", line 646, in setUp conn, addr = self.serv.accept() File "/usr/home/buildbot/python/3.x.koobs-freebsd-564d/build/Lib/socket.py", line 293, in accept fd, addr = self._accept() ConnectionAbortedError: [Errno 53] Software caused connection abort (...) |
|||
| msg364363 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-16 22:08 | |
It worked 10 days ago: https://buildbot.python.org/all/#/builders/152/builds/382 |
|||
| msg364367 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-16 22:27 | |
Same issue on AMD64 FreeBSD Shared 3.8: https://buildbot.python.org/all/#builders/18/builds/162 It looks more like a recent FreeBSD upgrade than a Python regression. 3.8 fail, build 162: os.uname: posix.uname_result(sysname='FreeBSD', nodename='130-CURRENT-amd64', release='13.0-CURRENT', version='FreeBSD 13.0-CURRENT #1 r358865: Wed Mar 11 21:35:42 UTC 2020 root@130-CURRENT-amd64:/usr/obj/usr/src/amd64.amd64/sys/GENERIC', machine='amd64') 3.8 ok, build 153 (12 days ago): os.uname: posix.uname_result(sysname='FreeBSD', nodename='130-CURRENT-amd64', release='13.0-CURRENT', version='FreeBSD 13.0-CURRENT #7 r356645: Sun Jan 12 14:57:12 AEDT 2020 root@130-CURRENT-amd64:/usr/obj/usr/src/amd64.amd64/sys/GENERIC-NODEBUG', machine='amd64') The kernel changed. |
|||
| msg364381 - (view) | Author: Kubilay Kocak (koobs)  |
Date: 2020-03-17 01:25 | |
Saw the following commit which *may* be related: "Log: sendfile() does currently not support SCTP sockets. Therefore, fail the call. " https://svnweb.freebsd.org/changeset/base/358965 I'm updating FreeBSD base on the BB worker to include that commit and will re-run the failing build |
|||
| msg364403 - (view) | Author: Kubilay Kocak (koobs) |
Date: 2020-03-17 09:44 | |
Update to BB complete. Rebuilding: https://buildbot.python.org/all/#/builders/18/builds/162 and https://buildbot.python.org/all/#/builders/152/builds/409 |
|||
| msg364406 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-17 10:43 | |
> https://buildbot.python.org/all/#/builders/152/builds/409 That's AMD64 FreeBSD Shared 3.x: there are still tons of SCTP failures. "Log: sendfile() does currently not support SCTP sockets. Therefore, fail the call." I don't understand why the test passed previously... was it simply skipped? Test on FreeBSD 12.1: --- vstinner@freebsd$ uname -a FreeBSD freebsd 12.1-RELEASE-p2 FreeBSD 12.1-RELEASE-p2 GENERIC amd64 vstinner@freebsd$ ./python -m test test_socket -v -m '*SCTP*' == CPython 3.9.0a4+ (heads/master:a45b695b9f, Mar 17 2020, 10:42:05) [Clang 8.0.1 (tags/RELEASE_801/final 366581)] == FreeBSD-12.1-RELEASE-p2-amd64-64bit-ELF little-endian == cwd: /usr/home/vstinner/python/master/build/test_python_1636 == CPU count: 8 == encodings: locale=UTF-8, FS=utf-8 0:00:00 load avg: 0.23 Run tests sequentially 0:00:00 load avg: 0.23 [1/1] test_socket testSendmsg (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgAfterClose (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgAncillaryGenerator (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgArray (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgBadArgs (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgBadCmsg (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgBadMultiCmsg (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgDataGenerator (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgDontWait (test.test_socket.SendmsgSCTPStreamTest) ... skipped 'MSG_DONTWAIT not known to work on this platform when sending' testSendmsgExcessCmsgReject (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgExplicitNoneAddr (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgGather (test.test_socket.SendmsgSCTPStreamTest) ... ok testSendmsgTimeout (test.test_socket.SendmsgSCTPStreamTest) ... ok testRecvmsg (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgAfterClose (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgBadArgs (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgEOF (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgExplicitDefaults (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgFromSendmsg (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgLongAncillaryBuf (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgOverflow (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgPeek (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgShortAncillaryBuf (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgShorter (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgTimeout (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsgTrunc (test.test_socket.RecvmsgSCTPStreamTest) ... ok testRecvmsg (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgAfterClose (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgEOF (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgExplicitDefaults (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgFromSendmsg (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgIntoArray (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgIntoBadArgs (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgIntoGenerator (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgIntoScatter (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgLongAncillaryBuf (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgOverflow (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgPeek (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgShortAncillaryBuf (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgShorter (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgTimeout (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok testRecvmsgTrunc (test.test_socket.RecvmsgIntoSCTPStreamTest) ... ok ---------------------------------------------------------------------- Ran 42 tests in 0.327s OK (skipped=1) == Tests result: SUCCESS == 1 test OK. Total duration: 746 ms Tests result: SUCCESS --- |
|||
| msg364408 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-17 10:54 | |
It seems like there is a race condition in some tests. Running the same test twice on the idle buildbot worker fails randomly: 130-CURRENT-amd64% ./python -m test test_socket -v -m test.test_socket.RecvmsgSCTPStreamTest.testRecvmsgAfterClose == CPython 3.9.0a4+ (heads/master:6ff79f6582, Mar 16 2020, 22:07:47) [Clang 10.0.0 (git@github.com:llvm/llvm-project.git llvmorg-10.0.0-rc3-1-gc290c == FreeBSD-13.0-CURRENT-amd64-64bit-ELF little-endian == cwd: /usr/home/haypo/python/master/build/test_python_31444 == CPU count: 2 == encodings: locale=UTF-8, FS=utf-8 (...) testRecvmsgAfterClose (test.test_socket.RecvmsgSCTPStreamTest) ... ok (...) 130-CURRENT-amd64% ./python -m test test_socket -v -m test.test_socket.RecvmsgSCTPStreamTest.testRecvmsgAfterClose (...) testRecvmsgAfterClose (test.test_socket.RecvmsgSCTPStreamTest) ... ERROR ====================================================================== ERROR: testRecvmsgAfterClose (test.test_socket.RecvmsgSCTPStreamTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/home/haypo/python/master/Lib/test/test_socket.py", line 342, in _setUp self.__setUp() File "/usr/home/haypo/python/master/Lib/test/test_socket.py", line 2533, in setUp super().setUp() File "/usr/home/haypo/python/master/Lib/test/test_socket.py", line 646, in setUp conn, addr = self.serv.accept() File "/usr/home/haypo/python/master/Lib/socket.py", line 293, in accept fd, addr = self._accept() ConnectionAbortedError: [Errno 53] Software caused connection abort (...) |
|||
| msg364579 - (view) | Author: Kubilay Kocak (koobs) |
Date: 2020-03-19 00:51 | |
I've been in touch with the FreeBSD SCTP maintainer and will continue to investigate that avenue. In the meantime, I'm re-running the last successful 3.x build (382 [1]) to see if that fails or not to isolate Cpython code changes as contributing causes [1] https://buildbot.python.org/all/#/builders/152/builds/382 |
|||
| msg365101 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-26 17:41 | |
Ok, we got the confirmation that it's a regression in the FreeBSD kernel in FreeBSD CURRENT. The following commit reverts a change which introduced the race condition: https://svnweb.freebsd.org/base?view=revision&revision=359302 I suggest to wait until this change lands on FreeBSD CURRENT buildbots (not workaround the bug in Python, since CURRENT is not released yet). |
|||
| msg365239 - (view) | Author: Kubilay Kocak (koobs) |
Date: 2020-03-29 01:49 | |
koobs-freebsd-564d (CURRENT) BB has been updated past https://svnweb.freebsd.org/base?view=revision&revision=359302 test_socket is no longer failing |
|||
| msg365256 - (view) | Author: STINNER Victor (vstinner) * |
Date: 2020-03-29 14:23 | |
Wow wow wow, that's huge. Python helps to detect FreeBSD kernel bugs :-) Thanks a lot Kubilay Kocak! |
|||
| msg365292 - (view) | Author: Kubilay Kocak (koobs) |
Date: 2020-03-30 02:56 | |
+1 to mutually beneficial relationships :) |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:59:28 | admin | set | github: 84163 |
| 2020-03-30 02:56:56 | koobs | set | messages: + msg365292 |
| 2020-03-29 14:23:53 | vstinner | set | messages: + msg365256 |
| 2020-03-29 01:49:38 | koobs | set | status: open -> closed resolution: fixed messages: + msg365239 stage: resolved |
| 2020-03-26 17:41:48 | vstinner | set | messages: + msg365101 |
| 2020-03-19 00:51:21 | koobs | set | messages: + msg364579 |
| 2020-03-17 10:54:17 | vstinner | set | messages: + msg364408 |
| 2020-03-17 10:43:57 | vstinner | set | messages: + msg364406 |
| 2020-03-17 09:44:33 | koobs | set | messages: + msg364403 |
| 2020-03-17 01:25:23 | koobs | set | messages: + msg364381 |
| 2020-03-16 22:27:06 | vstinner | set | messages: + msg364367 |
| 2020-03-16 22:08:18 | vstinner | set | messages: + msg364363 |
| 2020-03-16 21:59:26 | vstinner | create | |