On Thu, 2013-02-21 at 13:28 +0100, Xavier Fustero wrote: > Hiya, > > sorry to send many emails. But I realized there were more options in > valgrind so I executed it again using: > ah, you already got the interesting part.
> --leak-check=full --show-reachable=yes -v > > I have attached the output. > > Thanks a log, > Xavi > > > > > On 21 February 2013 13:18, Xavier Fustero <[email protected]> wrote: > > > Sorry to hassle today with my bug... I added another dump adding the > > --leak-check=full option. > > > > root@rsyslog:/var/tmp# /opt/valgrind/bin/valgrind --leak-check=full > > /usr/sbin/rsyslogd -n > > ==31957== Memcheck, a memory error detector > > ==31957== Copyright (C) 2002-2012, and GNU GPL'd, by Julian Seward et al. > > ==31957== Using Valgrind-3.8.1 and LibVEX; rerun with -h for copyright info > > ==31957== Command: /usr/sbin/rsyslogd -n > > ==31957== > > rsyslogd: Could no open output pipe '/dev/xconsole': No such file or > > directory [try http://www.rsyslog.com/e/2039 ] > > ==31957== Thread 2: > > ==31957== Jump to the invalid address stated on the next line > > ==31957== at 0x0: ??? > > ==31957== by 0x6ECA6E1: MsgSetInputName (in > > /usr/lib/rsyslog/imuxsock.so) would it be possible to compile with debug support and maybe -O0. This somehow looks a bit like things got mixed up, but I am missing the line number (which debug build will bring us, at least if the symbols are installed on that system). You may also use --enable-valgrind with rsyslog configure, which just prevents unloading of modules, so it has no effect on the actual code. Rainer > > ==31957== by 0x6EC1EE4: ??? (in /usr/lib/rsyslog/imuxsock.so) > > ==31957== by 0x6EC2BD2: ??? (in /usr/lib/rsyslog/imuxsock.so) > > ==31957== by 0x450F2E: ??? (in /usr/sbin/rsyslogd) > > ==31957== by 0x504E9C9: start_thread (pthread_create.c:300) > > ==31957== by 0x640021C: clone (clone.S:112) > > ==31957== Address 0x0 is not stack'd, malloc'd or (recently) free'd > > ==31957== > > ==31957== > > ==31957== Process terminating with default action of signal 11 (SIGSEGV) > > ==31957== Bad permissions for mapped region at address 0x0 > > ==31957== at 0x0: ??? > > ==31957== by 0x6ECA6E1: MsgSetInputName (in > > /usr/lib/rsyslog/imuxsock.so) > > ==31957== by 0x6EC1EE4: ??? (in /usr/lib/rsyslog/imuxsock.so) > > ==31957== by 0x6EC2BD2: ??? (in /usr/lib/rsyslog/imuxsock.so) > > ==31957== by 0x450F2E: ??? (in /usr/sbin/rsyslogd) > > ==31957== by 0x504E9C9: start_thread (pthread_create.c:300) > > ==31957== by 0x640021C: clone (clone.S:112) > > ==31957== > > ==31957== HEAP SUMMARY: > > ==31957== in use at exit: 201,203 bytes in 1,569 blocks > > ==31957== total heap usage: 2,228 allocs, 659 frees, 289,130 bytes > > allocated > > ==31957== > > ==31957== Thread 1: > > ==31957== 288 bytes in 1 blocks are possibly lost in loss record 1,001 of > > 1,053 > > ==31957== at 0x4C27B49: calloc (vg_replace_malloc.c:593) > > ==31957== by 0x4012495: _dl_allocate_tls (dl-tls.c:300) > > ==31957== by 0x504F728: pthread_create@@GLIBC_2.2.5 > > (allocatestack.c:561) > > ==31957== by 0x441642: wtpAdviseMaxWorkers (in /usr/sbin/rsyslogd) > > ==31957== by 0x446E62: qqueueEnqObj (in /usr/sbin/rsyslogd) > > ==31957== by 0x40EF7B: submitMsg (in /usr/sbin/rsyslogd) > > ==31957== by 0x40F15C: logmsgInternal (in /usr/sbin/rsyslogd) > > ==31957== by 0x410398: realMain (in /usr/sbin/rsyslogd) > > ==31957== by 0x6335C4C: (below main) (libc-start.c:226) > > ==31957== > > ==31957== 576 bytes in 2 blocks are possibly lost in loss record 1,019 of > > 1,053 > > ==31957== at 0x4C27B49: calloc (vg_replace_malloc.c:593) > > ==31957== by 0x4012495: _dl_allocate_tls (dl-tls.c:300) > > ==31957== by 0x504F728: pthread_create@@GLIBC_2.2.5 > > (allocatestack.c:561) > > ==31957== by 0x450ACD: thrdCreate (in /usr/sbin/rsyslogd) > > ==31957== by 0x427A98: activate (in /usr/sbin/rsyslogd) > > ==31957== by 0x41032B: realMain (in /usr/sbin/rsyslogd) > > ==31957== by 0x6335C4C: (below main) (libc-start.c:226) > > ==31957== > > ==31957== LEAK SUMMARY: > > ==31957== definitely lost: 0 bytes in 0 blocks > > ==31957== indirectly lost: 0 bytes in 0 blocks > > ==31957== possibly lost: 864 bytes in 3 blocks > > ==31957== still reachable: 200,339 bytes in 1,566 blocks > > ==31957== suppressed: 0 bytes in 0 blocks > > ==31957== Reachable blocks (those to which a pointer was found) are not > > shown. > > ==31957== To see them, rerun with: --leak-check=full --show-reachable=yes > > ==31957== > > ==31957== For counts of detected and suppressed errors, rerun with: -v > > ==31957== ERROR SUMMARY: 3 errors from 3 contexts (suppressed: 13 from 7) > > ==31957== could not unlink /tmp/vgdb-pipe-from-vgdb-to-31957-by-root-on-??? > > ==31957== could not unlink /tmp/vgdb-pipe-to-vgdb-from-31957-by-root-on-??? > > ==31957== could not unlink > > /tmp/vgdb-pipe-shared-mem-vgdb-31957-by-root-on-??? > > Killed > > > > > > Thanks, > > Xavi > > > > > > > > > > On 21 February 2013 12:41, Xavier Fustero <[email protected]> wrote: > > > >> Hi Rainer, > >> > >> I launched the rsyslog using valgrind: /opt/valgrind/bin/valgrind > >> /usr/sbin/rsyslogd -n > >> and then send a message using logger. This is what it dumps: > >> > >> ==31904== Memcheck, a memory error detector > >> ==31904== Copyright (C) 2002-2012, and GNU GPL'd, by Julian Seward et al. > >> ==31904== Using Valgrind-3.8.1 and LibVEX; rerun with -h for copyright > >> info > >> ==31904== Command: /usr/sbin/rsyslogd -n > >> ==31904== > >> rsyslogd: Could no open output pipe '/dev/xconsole': No such file or > >> directory [try http://www.rsyslog.com/e/2039 ] > >> ==31904== Thread 2: > >> ==31904== Jump to the invalid address stated on the next line > >> ==31904== at 0x0: ??? > >> ==31904== by 0x6ECA6E1: MsgSetInputName (in > >> /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x6EC1EE4: ??? (in /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x6EC2BD2: ??? (in /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x450F2E: ??? (in /usr/sbin/rsyslogd) > >> ==31904== by 0x504E9C9: start_thread (pthread_create.c:300) > >> ==31904== by 0x640021C: clone (clone.S:112) > >> ==31904== Address 0x0 is not stack'd, malloc'd or (recently) free'd > >> ==31904== > >> ==31904== > >> ==31904== Process terminating with default action of signal 11 (SIGSEGV) > >> ==31904== Bad permissions for mapped region at address 0x0 > >> ==31904== at 0x0: ??? > >> ==31904== by 0x6ECA6E1: MsgSetInputName (in > >> /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x6EC1EE4: ??? (in /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x6EC2BD2: ??? (in /usr/lib/rsyslog/imuxsock.so) > >> ==31904== by 0x450F2E: ??? (in /usr/sbin/rsyslogd) > >> ==31904== by 0x504E9C9: start_thread (pthread_create.c:300) > >> ==31904== by 0x640021C: clone (clone.S:112) > >> ==31904== > >> ==31904== HEAP SUMMARY: > >> ==31904== in use at exit: 201,203 bytes in 1,569 blocks > >> ==31904== total heap usage: 2,228 allocs, 659 frees, 289,130 bytes > >> allocated > >> ==31904== > >> ==31904== LEAK SUMMARY: > >> ==31904== definitely lost: 0 bytes in 0 blocks > >> ==31904== indirectly lost: 0 bytes in 0 blocks > >> ==31904== possibly lost: 864 bytes in 3 blocks > >> ==31904== still reachable: 200,339 bytes in 1,566 blocks > >> ==31904== suppressed: 0 bytes in 0 blocks > >> ==31904== Rerun with --leak-check=full to see details of leaked memory > >> ==31904== > >> ==31904== For counts of detected and suppressed errors, rerun with: -v > >> ==31904== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 13 from 7) > >> ==31904== could not unlink > >> /tmp/vgdb-pipe-from-vgdb-to-31904-by-root-on-??? > >> ==31904== could not unlink > >> /tmp/vgdb-pipe-to-vgdb-from-31904-by-root-on-??? > >> ==31904== could not unlink > >> /tmp/vgdb-pipe-shared-mem-vgdb-31904-by-root-on-??? > >> Killed > >> > >> > >> Thanks, > >> Xavi > >> > >> > >> > >> > >> On 21 February 2013 11:18, Rainer Gerhards <[email protected]>wrote: > >> > >>> On Thu, 2013-02-21 at 11:14 +0100, Xavier Fustero wrote: > >>> > Hi, > >>> > > >>> > I would like to ask if there is any Ubuntu expert here who could help > >>> me on > >>> > creating the rsyslog package to have rsyslog 7.2.5 running in my Ubuntu > >>> > 10.04. > >>> > > >>> > I have created several packages: > >>> > > >>> > json-c_0.9-1_amd64.deb > >>> > libee_0.4.1-1_amd64.deb > >>> > libestr_0.1.4-1_amd64.deb > >>> > liblognorm_0.3.5-1_amd64.deb > >>> > librelp_1.0.1-1_amd64.deb > >>> > rsyslog_7.2.5-1_amd64.deb > >>> > > >>> > Then I have 2 scenarios where one is working and the other no. > >>> > > >>> > > >>> > - *Scenario1:* > >>> > > >>> > > >>> > I have installed all above packages except rsyslog_7.2.5-1_amd64.deb. > >>> Then > >>> > I have installed any extra required package, downloaded the rsyslog and > >>> > compiled it in this box: > >>> > > >>> > apt-get install uuid-dev > >>> > > >>> > cd rsyslog-7.2.5 > >>> > > >>> > ./configure --libdir=/usr/lib --includedir=/usr/include > >>> > --enable-gssapi-krb5 --enable-kmsg --enable-debug --enable-memcheck > >>> > --enable-diagtools --enable-usertools --enable-mysql > >>> --enable-elasticsearch > >>> > --enable-gnutls --enable-mmnormalize --enable-mmjsonparse --enable-relp > >>> > --enable-imfile --enable-imptcp --enable-impstats --enable-pmlastmsg > >>> > > >>> > This seems to be working for me. If I launch the rsyslog by hand > >>> > /usr/local/sbin/rsyslogd the PrivDrop also seems to be working fine. It > >>> > takes a while for the root process to disappear but it does. > >>> > > >>> > > >>> > > >>> > - *Scenario 2* > >>> > > >>> > > >>> > This is the desired scenario as I am building server through templates > >>> and > >>> > I would like just to install packages at boot time. > >>> > I have installed all above packages. I don't know why but my rsyslog > >>> > package produces a segmentation fault every time I send a message. I > >>> have > >>> > created it according the above configuration parameters. > >>> > > >>> > But when I launch it in debug mode I get the last line when I get a > >>> message: > >>> > > >>> > ... > >>> > 6250.465837291:7fbc779ce700: regular consumer finished, iret=4, szlog > >>> > 0 sz phys 0 > >>> > 6250.465841734:7fbc779ce700: main Q:Reg/w0: worker IDLE, waiting for > >>> work. > >>> > 6253.511342997:7fbc799d2700: Message from UNIX socket: #3 > >>> > Segmentation fault > >>> > > >>> can you run this under valgrind? It hopefully points to the actual > >>> proiblem location. Make sure symbols are not stripped during > >>> compile/link. > >>> > >>> Rainer > >>> > I might miss something during the package creation because I am not > >>> able to > >>> > make a solid package for the rsyslog. > >>> > Does anyone have any hint on how to fix this? > >>> > > >>> > Thanks a log, > >>> > Xavi > >>> > _______________________________________________ > >>> > rsyslog mailing list > >>> > http://lists.adiscon.net/mailman/listinfo/rsyslog > >>> > http://www.rsyslog.com/professional-services/ > >>> > What's up with rsyslog? Follow https://twitter.com/rgerhards > >>> > NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a > >>> myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if > >>> you DON'T LIKE THAT. > >>> > >>> _______________________________________________ > >>> rsyslog mailing list > >>> http://lists.adiscon.net/mailman/listinfo/rsyslog > >>> http://www.rsyslog.com/professional-services/ > >>> What's up with rsyslog? Follow https://twitter.com/rgerhards > >>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad > >>> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you > >>> DON'T LIKE THAT. > >>> > >> > >> > > > _______________________________________________ > rsyslog mailing list > http://lists.adiscon.net/mailman/listinfo/rsyslog > http://www.rsyslog.com/professional-services/ > What's up with rsyslog? Follow https://twitter.com/rgerhards > NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of > sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T > LIKE THAT. _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
