hi,
I use the current Noad version and have experienced a couple of bugs. Here the first one:
If another noad is started before the previous ends, the earlier seems to segfault:
Sep 21 21:25:00 vdr noad[20105]: noad called with 'before' and online=1 and liverecording is no Sep 21 21:25:00 vdr noad[20105]: nothing to do yet Sep 21 21:25:52 vdr noad[15366]: noad aborted by signal Segmentation fault Sep 21 21:25:52 vdr noad[15366]: [bt] Execution path: Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x805b223] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x8057a47] Sep 21 21:25:52 vdr noad[15366]: [bt] [0xffffe420] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x804e4a3] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x80504f7] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x805081e] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x8051430] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x8052e36] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x8053182] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x805773d] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad [0x8058060] Sep 21 21:25:52 vdr noad[15366]: [bt] /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xf0) [0xb7bd0970] Sep 21 21:25:52 vdr noad[15366]: [bt] /usr/local/bin/noad(avcodec_register_all+0x41) [0x8049be1]
yours, Jouni
hi, Am Donnerstag, 22. September 2005 17:06 schrieb Jouni Karvo:
i've tried to reproduce this error without success. Maybe i can say more about this error if you translate the bt-addresses to function-names. This could be done with addr2line -e <your-noad-executable-here> -f -C <bt-address> The complete backtrace can be translated with grep '[bt.*noad.*[0x' <your-logfile-here> | sed -e 's/.*[//' -e s'/]//' | addr2line -e <your-noad-executable-here> -f -C
There should be more entries in the logfile about these two noad-instances. Please report all entries so i can see how these instances are started. What about other bugs you're talking about?
Regards theNoad
yours, Jouni
hi,
the Noad writes:
I am not sure if the presence of the other noad process is essential here. Here an example where the crash happens a lot later than the other noad stops (with the function names in the end)
Sep 21 14:15:57 vdr noad[19424]: noad arg[0]: /usr/local/bin/noad Sep 21 14:15:57 vdr noad[19424]: noad arg[1]: -c Sep 21 14:15:57 vdr noad[19424]: noad arg[2]: -o Sep 21 14:15:57 vdr noad[19424]: noad arg[3]: -j Sep 21 14:15:57 vdr noad[19424]: noad arg[4]: -a Sep 21 14:15:57 vdr noad[19424]: noad arg[5]: --nelonen Sep 21 14:15:57 vdr noad[19424]: noad arg[6]: - Sep 21 14:15:57 vdr noad[19424]: noad arg[7]: /video0/elokuvat/Elokuva:_Morsiamen_is<E4>/2005-09-17.17:55.40.02.rec Sep 21 14:15:57 vdr noad[19424]: noad args done Sep 21 14:15:57 vdr noad[19424]: Wednesday,21.09.2005 14:15:57 start noad-0.5.3 for /video0/elokuvat/Elokuva:_Morsiamen_is<E4>/2005-09-17.17:55.40.02.rec Sep 21 16:00:08 vdr vdr[1770]: ERROR (svdrp.c,364): Broken pipe Sep 21 18:00:08 vdr vdr[1770]: ERROR (svdrp.c,364): Broken pipe Sep 21 18:25:00 vdr noad[11702]: noad arg[0]: /usr/local/bin/noad Sep 21 18:25:00 vdr noad[11702]: noad arg[1]: -c Sep 21 18:25:00 vdr noad[11702]: noad arg[2]: -o Sep 21 18:25:00 vdr noad[11702]: noad arg[3]: -j Sep 21 18:25:00 vdr noad[11702]: noad arg[4]: -S Sep 21 18:25:00 vdr noad[11702]: noad arg[5]: --nelonen Sep 21 18:25:00 vdr noad[11702]: noad arg[6]: before Sep 21 18:25:00 vdr noad[11702]: noad arg[7]: /video0/jamie/Jamien_kouluruokaa/2005-09-21.18:25.40.21.rec Sep 21 18:25:00 vdr noad[11702]: noad args done Sep 21 18:25:00 vdr noad[11702]: noad called with 'before' and online=1 and liverecording is no Sep 21 18:25:00 vdr noad[11702]: nothing to do yet Sep 21 18:54:03 vdr noad[19424]: noad aborted by signal Segmentation fault Sep 21 18:54:03 vdr noad[19424]: [bt] Execution path: Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x805b223] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8057a47] Sep 21 18:54:03 vdr noad[19424]: [bt] [0xffffe420] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x804e4a3] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x80504f7] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8050a39] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8051430] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8052e36] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8053182] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x805773d] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad [0x8058060] Sep 21 18:54:03 vdr noad[19424]: [bt] /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xf0) [0xb7bd0970] Sep 21 18:54:03 vdr noad[19424]: [bt] /usr/local/bin/noad(avcodec_register_all+0x41) [0x8049be1]
dr2line -e `which noad` -f -C[0x' rap.txt | sed -e 's/.*[//' -e s'/]//' | ad show_stackframe(bool) /home/kex/installs/vdr/noad/noad-0.6.0/tools.cpp:1135 signal_handler /home/kex/installs/vdr/noad/noad-0.6.0/main.cpp:157 getPicInfo(pic_info*, int, int, bool, bool, bool) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:2345 checkMarkPair(cMark**, cMarks*, cFileName*) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:2726 pass2a(cMarks*, cFileName*) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:2789 MarkCleanup(cMarks*, cFileName*) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:1681 scanRecord(int, cMarks*) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:2063 doX11Scan(noadData*, char const*, int) /home/kex/installs/vdr/noad/noad-0.6.0/noad.cpp:2121 doNoad(bool, char const*) /home/kex/installs/vdr/noad/noad-0.6.0/main.cpp:80 main /home/kex/installs/vdr/noad/noad-0.6.0/main.cpp:577 _start ../sysdeps/i386/elf/start.S:105 kex@vdr:~$
The other problems I mentioned may come from faulty recordings - I have not yet upgraded to a VDR with a videorepacker, so the problems can come from there.
yours, Jouni
Am Dienstag, 27. September 2005 21:00 schrieb Jouni Karvo:
Looking into the sources i can't find any reason why noad should segfault here. Maybe more information about a noad run can help. Please add '-vvv' to your noad-call and let's see what we can see. As the logs are very big with -vvv, please put the resulting log into a zip-file and send it directly to me (theNoad@SoftHome.net). I've seen in the log that noad runs more than 4 hours for this recording??? Please give me also some informations about the machine noad runs on.
Hope the problem above isn't caused by a faulty record also.
Regards theNoad
-
Jeremy Hall -
Jouni Karvo -
the Noad