Re: Mkclean sigsev on a specific file
julroy67 <[email protected]>
| Newsgroups | gmane.comp.multimedia.matroska.user |
|---|---|
| Message-ID | <[email protected]> |
Sorry for the delay, but now, with this new release, it is finally fixed. Really, good work ! :) Thanks 2010/7/17 Steve Lhomme <[email protected]>: > I released new versions of both tools. They should now build more easily out > of the box on Linux, including x64 variants. > http://dl.matroska.org/downloads/mkvalidator/mkvalidator-0.1.11.tar.bz2 > http://dl.matroska.org/downloads/mkclean/mkclean-0.4.2.tar.bz2 > > On Sat, Jul 17, 2010 at 12:37 AM, Steve Lhomme <[email protected]> wrote: >> >> I already reported the problem to Handbrake. It's a low priority bug for >> them. >> I already fixed a bug that looks like this. Can you do another SVN >> checkout to make sure you're using the right value ? >> >> On Fri, Jul 16, 2010 at 8:38 PM, julroy67 <[email protected]> wrote: >>> >>> Nice, good to hear it, thanks. But now, shall I report this to the >>> handbrake devs ? And second one, I trie the svn version, and it seems >>> to work, but now it is a libc segfault : >>> http://julroy67.free.fr/DL/bt_libc.log :( >>> >>> But still thanks for the fix. >>> >>> 2010/7/16 Steve Lhomme <[email protected]>: >>> > Hi again, >>> > I've identified the problem. It seems that Handbrake is writing a CRC >>> > element at the wrong location (second element of SegmentInfo) and with >>> > the >>> > wrong size (12 instead of 4). That made mkclean behave inconsistently. >>> > I've >>> > identified other possible problems in mkclean and they are all fixed in >>> > SVN. >>> > I am not going to make a release today as I have other things I want to >>> > put >>> > in the next one (hopefully this week-end). >>> > If you don't want to wait for the next release, you can get the latest >>> > sources from SVN >>> > >>> > at https://matroska.svn.sourceforge.net/svnroot/matroska/trunk/foundation_src >>> > You will need to copy <root>/mkclean/default_config.h into >>> > <root>/config.h >>> > before running "coremake gcc_linux_x64" >>> > Steve >>> > On Fri, Jul 16, 2010 at 4:41 PM, julroy67 <[email protected]> wrote: >>> >> >>> >> Ok, here is the dump http://julroy67.free.fr/DL/TGBS_4925_dump_1M (« >>> >> dd if=Test.mkv of=TGBS_4925_dump_1M bs=1M count=1 » should be >>> >> correct). >>> >> >>> >> 2010/7/16 Steve Lhomme <[email protected]>: >>> >> > OK that all looks correct. Can you cut the first 1 or 2 first MB of >>> >> > the >>> >> > file >>> >> > and send them to me ? >>> >> > >>> >> > On Fri, Jul 16, 2010 at 4:30 PM, julroy67 <[email protected]> >>> >> > wrote: >>> >> >> >>> >> >> Yes, it plays fine (VLC and Mplayer at least). The mkvinfo simple >>> >> >> output : http://julroy67.free.fr/DL/TGBS_4925_mkvinfo.txt >>> >> >> >>> >> >> 2010/7/16 Steve Lhomme <[email protected]>: >>> >> >> > OK it seems your file has no (found) Tracks. Does it play >>> >> >> > correctly ? >>> >> >> > It's >>> >> >> > not normal for a file that big to have no Tracks (it's possible >>> >> >> > for >>> >> >> > files >>> >> >> > with only chapters or attachments). And theoretically the code >>> >> >> > should >>> >> >> > exit >>> >> >> > before the crash, as there is a check to see if there are Cluster >>> >> >> > but >>> >> >> > no >>> >> >> > matching Tracks. So I don't know how you are getting to that >>> >> >> > point. >>> >> >> > Can you send me the output of mkvinfo on that file ? You can use >>> >> >> > pastebin or >>> >> >> > that same web server you used. >>> >> >> > Thanks, >>> >> >> > Steve >>> >> >> > >>> >> >> > On Fri, Jul 16, 2010 at 3:50 PM, julroy67 <[email protected]> >>> >> >> > wrote: >>> >> >> >> >>> >> >> >> Here is the nice full output : >>> >> >> >> http://julroy67.free.fr/DL/mkclean_full_backtrace.log >>> >> >> >> >>> >> >> >> Sending the file can be made, but they all are over 1 Gb, the >>> >> >> >> smallest >>> >> >> >> is 1.2 Gb and would take a long time to upload if so. >>> >> >> >> >>> >> >> >> 2010/7/16 Steve Lhomme <[email protected]>: >>> >> >> >> > Can you get a bigger backtrace, up to the main() if possible ? >>> >> >> >> > Is >>> >> >> >> > it >>> >> >> >> > possible to get one of the files that crash (the smallest) ? >>> >> >> >> > >>> >> >> >> > On Fri, Jul 16, 2010 at 3:26 PM, julroy67 <[email protected]> >>> >> >> >> > wrote: >>> >> >> >> >> >>> >> >> >> >> So, better, than before. The first version, with a clean >>> >> >> >> >> build, >>> >> >> >> >> crashes like the arch package, same output. And now for the >>> >> >> >> >> debug >>> >> >> >> >> version, the real output (I have missed up something in the >>> >> >> >> >> other >>> >> >> >> >> version, that crashed at startup) : >>> >> >> >> >> >>> >> >> >> >> 15:23 julroy67@momoko-laptop ~/Apps/mkclean/mkclean-0.4.1% >>> >> >> >> >> gdb >>> >> >> >> >> release/gcc_linux_x64_debug/mkclean >>> >> >> >> >> GNU gdb (GDB) 7.1 >>> >> >> >> >> Copyright (C) 2010 Free Software Foundation, Inc. >>> >> >> >> >> License GPLv3+: GNU GPL version 3 or later >>> >> >> >> >> <http://gnu.org/licenses/gpl.html> >>> >> >> >> >> This is free software: you are free to change and >>> >> >> >> >> redistribute >>> >> >> >> >> it. >>> >> >> >> >> There is NO WARRANTY, to the extent permitted by law. Type >>> >> >> >> >> "show >>> >> >> >> >> copying" >>> >> >> >> >> and "show warranty" for details. >>> >> >> >> >> This GDB was configured as "x86_64-unknown-linux-gnu". >>> >> >> >> >> For bug reporting instructions, please see: >>> >> >> >> >> <http://www.gnu.org/software/gdb/bugs/>... >>> >> >> >> >> Reading symbols from >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> /home/julroy67/Apps/mkclean/mkclean-0.4.1/release/gcc_linux_x64_debug/mkclean...done. >>> >> >> >> >> (gdb) run --optimize --remux Test.mkv Test.mkv.2 >>> >> >> >> >> Starting program: >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> >> >> /home/julroy67/Apps/mkclean/mkclean-0.4.1/release/gcc_linux_x64_debug/mkclean >>> >> >> >> >> --optimize --remux Test.mkv Test.mkv.2 >>> >> >> >> >> [Thread debugging using libthread_db enabled] >>> >> >> >> >> Progress 1/3: 100% >>> >> >> >> >> >>> >> >> >> >> Program received signal SIGSEGV, Segmentation fault. >>> >> >> >> >> 0x000000000042b4af in EBML_MasterFindFirstElt (Element=0x0, >>> >> >> >> >> Context=0x656ce0, bCreateIfNull=0, SetDefault=0) at >>> >> >> >> >> ebmlmaster.c:46 >>> >> >> >> >> 46 for >>> >> >> >> >> (i=EBML_MasterChildren(Element);i;i=EBML_MasterNext(i)) >>> >> >> >> >> >>> >> >> >> >> This one is the good, I think. >>> >> >> >> >> >>> >> >> >> >> 2010/7/16 Steve Lhomme <[email protected]>: >>> >> >> >> >> > Before starting to debug you should try to build it first >>> >> >> >> >> > like >>> >> >> >> >> > I >>> >> >> >> >> > just >>> >> >> >> >> > described you. Using the available source package >>> >> >> >> >> > >>> >> >> >> >> > >>> >> >> >> >> > >>> >> >> >> >> > here: http://dl.matroska.org/downloads/mkclean/mkclean-0.4.1.tar.bz2 >>> >> >> >> >> > I don't know about archlinux and how you got that package >>> >> >> >> >> > and >>> >> >> >> >> > who >>> >> >> >> >> > made >>> >> >> >> >> > it. >>> >> >> >> >> > So I don't know what it contains and if it's correct. >>> >> >> >> >> > mkclean >>> >> >> >> >> > doesn't >>> >> >> >> >> > use >>> >> >> >> >> > the usual autoconf/automake system found on Linux, so maybe >>> >> >> >> >> > it >>> >> >> >> >> > was >>> >> >> >> >> > packaged >>> >> >> >> >> > incorrectly. >>> >> >> >> >> > If you manage to build it like I said and it still fails, >>> >> >> >> >> > you >>> >> >> >> >> > can >>> >> >> >> >> > try >>> >> >> >> >> > to >>> >> >> >> >> > rebuild it with debug informations like this: >>> >> >> >> >> > ./coremake distclean >>> >> >> >> >> > ./coremake gcc_linux_x64_debug >>> >> >> >> >> > >>> >> >> >> >> > Steve >>> >> >> >> >> > On Fri, Jul 16, 2010 at 3:05 PM, julroy67 >>> >> >> >> >> > <[email protected]> >>> >> >> >> >> > wrote: >>> >> >> >> >> >> >>> >> >> >> >> >> Mhh I don't know if I'm correctly generating the debug >>> >> >> >> >> >> version >>> >> >> >> >> >> of >>> >> >> >> >> >> mkclean, could you say me step by step how to it. To be >>> >> >> >> >> >> clear, >>> >> >> >> >> >> I'm >>> >> >> >> >> >> using archlinux, and archlinux AUR package of mkclean, it >>> >> >> >> >> >> works >>> >> >> >> >> >> for >>> >> >> >> >> >> most of the file, but for at least 2 of them, it crashes >>> >> >> >> >> >> as >>> >> >> >> >> >> this >>> >> >> >> >> >> : >>> >> >> >> >> >> >>> >> >> >> >> >> 14:59 julroy67@momoko-laptop ~% mkclean --optimize --remux >>> >> >> >> >> >> Test.mkv >>> >> >> >> >> >> Test.mkv.2 >>> >> >> >> >> >> Progress 1/3: 100% >>> >> >> >> >> >> zsh: segmentation fault mkclean --optimize --remux >>> >> >> >> >> >> Test.mkv >>> >> >> >> >> >> Test.mkv.2 >>> >> >> >> >> >> >>> >> >> >> >> >> And it seems that the debug binary that I've generated >>> >> >> >> >> >> crashes >>> >> >> >> >> >> always, >>> >> >> >> >> >> so I'm doing wrong, but still, mkclean has a error >>> >> >> >> >> >> somewhere >>> >> >> >> >> >> ^^' >>> >> >> >> >> >> So >>> >> >> >> >> >> to >>> >> >> >> >> >> be sure to do correctly, I think, a step by step debugging >>> >> >> >> >> >> guide >>> >> >> >> >> >> would >>> >> >> >> >> >> help. Sorry for the hassle. >>> >> >> >> >> >> >>> >> >> >> >> >> 2010/7/16 Steve Lhomme <[email protected]>: >>> >> >> >> >> >> > Hi. >>> >> >> >> >> >> > Given the nature of the crash, it looks like it should >>> >> >> >> >> >> > always >>> >> >> >> >> >> > crash >>> >> >> >> >> >> > as >>> >> >> >> >> >> > that >>> >> >> >> >> >> > error is on the init. It may have to do with linux x64 >>> >> >> >> >> >> > rather >>> >> >> >> >> >> > than >>> >> >> >> >> >> > 32 >>> >> >> >> >> >> > bits. >>> >> >> >> >> >> > I'll give it a try on my side. >>> >> >> >> >> >> > Or did you compile it with a 32 bits tool chain ? The >>> >> >> >> >> >> > regular >>> >> >> >> >> >> > bootstrap.sh >>> >> >> >> >> >> > and subsequent "make" command are building for 32 bits. >>> >> >> >> >> >> > In >>> >> >> >> >> >> > 0.4.2 >>> >> >> >> >> >> > that >>> >> >> >> >> >> > should >>> >> >> >> >> >> > be released soon I changed that and there is now a >>> >> >> >> >> >> > "configure" >>> >> >> >> >> >> > script. >>> >> >> >> >> >> > It >>> >> >> >> >> >> > will build the makefiles for 64 bits architecture on 64 >>> >> >> >> >> >> > bits >>> >> >> >> >> >> > ones... >>> >> >> >> >> >> > Meanwhile you can make sure or that by running that in >>> >> >> >> >> >> > the >>> >> >> >> >> >> > root >>> >> >> >> >> >> > folder >>> >> >> >> >> >> > of >>> >> >> >> >> >> > the sources: >>> >> >> >> >> >> > ./coremake distclean >>> >> >> >> >> >> > ./coremake gcc_linux_x64 >>> >> >> >> >> >> > Then you run "make -C mkclean" and it should work. >>> >> >> >> >> >> > >>> >> >> >> >> >> > On Fri, Jul 16, 2010 at 2:41 PM, julroy67 >>> >> >> >> >> >> > <[email protected]> >>> >> >> >> >> >> > wrote: >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> Hello, >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> I'm using mkclean to clean up my mkv files that are for >>> >> >> >> >> >> >> the >>> >> >> >> >> >> >> most >>> >> >> >> >> >> >> not >>> >> >> >> >> >> >> valid as mkvalidate says, but on certain files, mkclean >>> >> >> >> >> >> >> crashes. >>> >> >> >> >> >> >> I've >>> >> >> >> >> >> >> managed to get the backtrace from gdb. But I'm not sure >>> >> >> >> >> >> >> where >>> >> >> >> >> >> >> I >>> >> >> >> >> >> >> have >>> >> >> >> >> >> >> to report the issue. >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> For the backtrace here's it : >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> 14:30 julroy67@momoko-laptop >>> >> >> >> >> >> >> ~/Apps/mkclean/src/mkclean-0.4.1% >>> >> >> >> >> >> >> gdb >>> >> >> >> >> >> >> release/gcc_linux_x64/mkclean >>> >> >> >> >> >> >> GNU gdb (GDB) 7.1 >>> >> >> >> >> >> >> Copyright (C) 2010 Free Software Foundation, Inc. >>> >> >> >> >> >> >> License GPLv3+: GNU GPL version 3 or later >>> >> >> >> >> >> >> <http://gnu.org/licenses/gpl.html> >>> >> >> >> >> >> >> This is free software: you are free to change and >>> >> >> >> >> >> >> redistribute >>> >> >> >> >> >> >> it. >>> >> >> >> >> >> >> There is NO WARRANTY, to the extent permitted by law. >>> >> >> >> >> >> >> Type >>> >> >> >> >> >> >> "show >>> >> >> >> >> >> >> copying" >>> >> >> >> >> >> >> and "show warranty" for details. >>> >> >> >> >> >> >> This GDB was configured as "x86_64-unknown-linux-gnu". >>> >> >> >> >> >> >> For bug reporting instructions, please see: >>> >> >> >> >> >> >> <http://www.gnu.org/software/gdb/bugs/>... >>> >> >> >> >> >> >> Reading symbols from >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> /home/julroy67/Apps/mkclean/src/mkclean-0.4.1/release/gcc_linux_x64/mkclean...done. >>> >> >> >> >> >> >> (gdb) run --optimize --remux Test.mkv Test.mkv.2 >>> >> >> >> >> >> >> Starting program: >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> /home/julroy67/Apps/mkclean/src/mkclean-0.4.1/release/gcc_linux_x64/mkclean >>> >> >> >> >> >> >> --optimize --remux Test.mkv Test.mkv.2 >>> >> >> >> >> >> >> [Thread debugging using libthread_db enabled] >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> Program received signal SIGSEGV, Segmentation fault. >>> >> >> >> >> >> >> 0x000000000040180e in ProjectSettings >>> >> >> >> >> >> >> (p=0x7fffffffdc10) at >>> >> >> >> >> >> >> mkclean_stdafx.c:99 >>> >> >> >> >> >> >> 99 >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> Node_SetData((node*)p,NODECONTEXT_PROJECT_NAME,TYPE_STRING,PROJECT_NAME); >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> or see here : >>> >> >> >> >> >> >> http://julroy67.free.fr/DL/mkclean_crash_backtrace.log >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> So anyone can say what I've to do ? >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> Thanks. >>> >> >> >> >> >> >> _______________________________________________ >>> >> >> >> >> >> >> Matroska-users mailing list >>> >> >> >> >> >> >> [email protected] >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> >> >> >> Read Matroska-Users on GMane: >>> >> >> >> >> >> >> >>> >> >> >> >> >> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> >> >> > >>> >> >> >> >> >> > >>> >> >> >> >> >> > _______________________________________________ >>> >> >> >> >> >> > Matroska-users mailing list >>> >> >> >> >> >> > [email protected] >>> >> >> >> >> >> > >>> >> >> >> >> >> > >>> >> >> >> >> >> > >>> >> >> >> >> >> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> >> >> > Read Matroska-Users on GMane: >>> >> >> >> >> >> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> >> >> > >>> >> >> >> >> >> > >>> >> >> >> >> >> _______________________________________________ >>> >> >> >> >> >> Matroska-users mailing list >>> >> >> >> >> >> [email protected] >>> >> >> >> >> >> >>> >> >> >> >> >> >>> >> >> >> >> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> >> >> Read Matroska-Users on GMane: >>> >> >> >> >> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> >> > >>> >> >> >> >> > >>> >> >> >> >> > _______________________________________________ >>> >> >> >> >> > Matroska-users mailing list >>> >> >> >> >> > [email protected] >>> >> >> >> >> > >>> >> >> >> >> > >>> >> >> >> >> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> >> > Read Matroska-Users on GMane: >>> >> >> >> >> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> >> > >>> >> >> >> >> > >>> >> >> >> >> _______________________________________________ >>> >> >> >> >> Matroska-users mailing list >>> >> >> >> >> [email protected] >>> >> >> >> >> >>> >> >> >> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> >> Read Matroska-Users on GMane: >>> >> >> >> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> > >>> >> >> >> > >>> >> >> >> > _______________________________________________ >>> >> >> >> > Matroska-users mailing list >>> >> >> >> > [email protected] >>> >> >> >> > >>> >> >> >> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> > Read Matroska-Users on GMane: >>> >> >> >> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> >> > >>> >> >> >> > >>> >> >> >> _______________________________________________ >>> >> >> >> Matroska-users mailing list >>> >> >> >> [email protected] >>> >> >> >> >>> >> >> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> >> Read Matroska-Users on GMane: >>> >> >> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> > >>> >> >> > >>> >> >> > _______________________________________________ >>> >> >> > Matroska-users mailing list >>> >> >> > [email protected] >>> >> >> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> > Read Matroska-Users on GMane: >>> >> >> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> >> > >>> >> >> > >>> >> >> _______________________________________________ >>> >> >> Matroska-users mailing list >>> >> >> [email protected] >>> >> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> >> Read Matroska-Users on GMane: >>> >> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> > >>> >> > >>> >> > _______________________________________________ >>> >> > Matroska-users mailing list >>> >> > [email protected] >>> >> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> > Read Matroska-Users on GMane: >>> >> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> >> > >>> >> > >>> >> _______________________________________________ >>> >> Matroska-users mailing list >>> >> [email protected] >>> >> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> >> Read Matroska-Users on GMane: >>> >> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> > >>> > >>> > _______________________________________________ >>> > Matroska-users mailing list >>> > [email protected] >>> > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> > Read Matroska-Users on GMane: >>> > http://dir.gmane.org/gmane.comp.multimedia.matroska.user >>> > >>> > >>> _______________________________________________ >>> Matroska-users mailing list >>> [email protected] >>> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users >>> Read Matroska-Users on GMane: >>> http://dir.gmane.org/gmane.comp.multimedia.matroska.user >> > > > _______________________________________________ > Matroska-users mailing list > [email protected] > http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users > Read Matroska-Users on GMane: > http://dir.gmane.org/gmane.comp.multimedia.matroska.user > > _______________________________________________ Matroska-users mailing list [email protected] http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-users Read Matroska-Users on GMane: http://dir.gmane.org/gmane.comp.multimedia.matroska.user