Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sorry about late reply, got a bit busy with other things. On 11/07/2014 11:43 PM, Reco wrote: Hi. I'd try adding some memory (or swapspace) to this VM to satisfy mremap's wish to remap to exactly 8G virtual memory. Reco Tried this, set the VMs memory to 8gb and still get the segfault, the strace and core dumb look pretty identical to before. Also still not sure why I don't have the debug symbol things for the ldap lib, tried a few more times to install it to no avail. - -Joris -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUYpWUAAoJEORnMHMHY2Fro8wQALSgyakp9fXsP063F51s3av4 3kc698UvCW8Qy1TY/CfsstOhiOPaLUjtJIKNwzc7YaFi8NhzsZcJzTeikFGWSpKg TfxDlEv9kJQWQiKxKY4RLzW6ZXKqPJjFoN1ob/rY3hdy/OoNUD1rGApzallwptZ/ pAx0my0g7edZQwD5sbtKfo3unSTWD8j3XDJF0N0BwxqsT6HncpIAtbiNQ+PJpfm4 gQRnJYd/ZQ8aGb9CpxqrtAAbf4dAqVBYmRa4jl+2crRKFL9DX+se83OimDQzJk6C DlEPIxPk5vRNtKbqYfYLc8G4YBPYVY1q3OoBlizvEQvgLkpamoWp07smg4TjRqWo tsp6b2ORkpmCxDeH8W6Lt+uhrkszQm3g6WWKPJbbMQpoaytg3rLdI0e9n20axzgo RAa0XyD26xHWdAK0IjnzQjrjllplkbx/seRo3d0b0NudNZKy/yQIJvdw19UA2GoE c7d+AYgBcA3ulY+LUc4hNnfyNqtuzALJWiQ/kO0V9plOkGd/1Smn546iBBxxCC7q P3LWqP72HfX4VA1i+/+7TQL3i9/HegnvA5h+HVmRSU38bYYiKKdKKZh+DC1unBaW dTov6Q04N9iEo4bZdBla+UhDkV3awpwfJ1v4s8JO6Eircly0btIrrqnq2euC7dTL yUpIjET51WlRtOD+oUla =0htc -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/54629597.4060...@linux.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/08/2014 03:43 PM, Don Armstrong wrote: No, just don't send the whole core dump itself (the file called core) to the BTS (or this mailing list). It almost certainly has your machine ldap password in it. that makes sense, wont do that, thanks. You should at least have the symbols for libnss-ldap... can you check the output of nm -a for that library to make sure it actually has the debugging symbols? Where do i find the object file for libnss-ldap? I tried running nm -a on pretty much every file I could find with libnss-ldap in the name and it said it didn't recognize the file type. -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUX87bAAoJEORnMHMHY2Frkv4QAI8i1wfJfKh3LCuxeDDvnmmU 0WgtVSdJTgV+AfuFAq7Ud5/FHeHDAh2gxP3g3x6A4VMAKL7om9Zqu2uO+OkjJuXk m0Fq9BULBA6iaPvutfZ6YlvSG6OsMwS/U80W1++4B3GtQBGW2lOboj3z6r4vP9R3 Dba8hzvDJA5oiapgNagh30jIIl/Vs8Q4Ak/cs4aVssZMaM8LVIVFf7OZT3YikfSq 8K2mksTfscGMdlhZxVNeN5lgIq/17/noZ1ieYAQRvXcRTGGsKhXLVeOVqsuyHUFz uvx2+y0LURCKJmPtsSnKVIQAzBqS3KOro3QvrIALMTIyzi3iCKHjxlGnIhXVpje8 vMJNAezeQU1xluxBm/TrYBoo4Xrakc5uDYIzsiZGoOpperxeL5qh8RwntnR+7pTa zfv5sKOg4H0hWO4pizuqxS476rQiNjMXpyJzwSZX304lDp/DjoEgl+PFzN+X1CRj nZBRR26rPH/46GBZwTDyxs1MWh1DpKrNQAsladGQe/RgLHZmHqVgAknrfvwwuGay FU8DMCcam4UsjeBti96RyY4VNysaiTIXZw/jTqBld9j6ixNqH7pX2Jxau/gvl2YB Whpad+z4TVDw+A4CTY7jBiAuIMcX6UX7gF1XUp4knSNvM6OsjIWgVmPwxFr0u3pf TjqIWdT7sXkvhEGI6y5Z =3N+u -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545fcede.6000...@gmail.com
Re: useradd segmentation fault
On Sun, 09 Nov 2014, Joris Bolsens wrote: Where do i find the object file for libnss-ldap? I tried running nm -a on pretty much every file I could find with libnss-ldap in the name and it said it didn't recognize the file type. It should be symlinked from /lib64/libnss_ldap.so.2 or /lib/x86_64-linux-gnu/ and nm -Da or similar should do it. -- Don Armstrong http://www.donarmstrong.com There is no such thing as social gambling. Either you are there to cut the other bloke's heart out and eat it--or you're a sucker. If you don't like this choice--don't gamble. -- Robert Heinlein _Time Enough For Love_ p250 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141109205516.gf29...@teltox.donarmstrong.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/09/2014 12:55 PM, Don Armstrong wrote: On Sun, 09 Nov 2014, Joris Bolsens wrote: Where do i find the object file for libnss-ldap? I tried running nm -a on pretty much every file I could find with libnss-ldap in the name and it said it didn't recognize the file type. It should be symlinked from /lib64/libnss_ldap.so.2 or /lib/x86_64-linux-gnu/ and nm -Da or similar should do it. Thanks, found it. here is the output, # nm -Da libnss_ldap.so.2 34d8 t .init A EXPORTED w _ITM_deregisterTMCloneTable w _ITM_registerTMCloneTable w _Jv_RegisterClasses U __assert_fail U __ctype_b_loc w __cxa_finalize U __dn_expand U __errno_location U __fxstat w __gmon_start__ w __pthread_mutex_lock w __pthread_mutex_unlock U __register_atfork U __res_init U __res_search U __res_state U __snprintf_chk U __stack_chk_fail U __stpcpy_chk U __strdup U __syslog_chk U __xstat b010 T _nss_ldap_endaliasent d380 T _nss_ldap_endautomntent cdc0 T _nss_ldap_endetherent 9810 T _nss_ldap_endgrent a450 T _nss_ldap_endhostent a860 T _nss_ldap_endnetent 9af0 T _nss_ldap_endnetgrent aac0 T _nss_ldap_endprotoent 8210 T _nss_ldap_endpwent 9f80 T _nss_ldap_endrpcent b480 T _nss_ldap_endservent ae90 T _nss_ldap_endspent afa0 T _nss_ldap_getaliasbyname_r b040 T _nss_ldap_getaliasent_r d3b0 T _nss_ldap_getautomntbyname_r d250 T _nss_ldap_getautomntent_r cdf0 T _nss_ldap_getetherent_r 9840 T _nss_ldap_getgrent_r 9780 T _nss_ldap_getgrgid_r 9710 T _nss_ldap_getgrnam_r a350 T _nss_ldap_gethostbyaddr_r a270 T _nss_ldap_gethostbyname2_r a320 T _nss_ldap_gethostbyname_r a480 T _nss_ldap_gethostent_r ccb0 T _nss_ldap_gethostton_r a680 T _nss_ldap_getnetbyaddr_r a5f0 T _nss_ldap_getnetbyname_r a890 T _nss_ldap_getnetent_r 9c10 T _nss_ldap_getnetgrent_r cd00 T _nss_ldap_getntohost_r aa00 T _nss_ldap_getprotobyname_r aa50 T _nss_ldap_getprotobynumber_r aaf0 T _nss_ldap_getprotoent_r 8240 T _nss_ldap_getpwent_r 8150 T _nss_ldap_getpwnam_r 81a0 T _nss_ldap_getpwuid_r 9ec0 T _nss_ldap_getrpcbyname_r 9f10 T _nss_ldap_getrpcbynumber_r 9fb0 T _nss_ldap_getrpcent_r b350 T _nss_ldap_getservbyname_r b3d0 T _nss_ldap_getservbyport_r b4b0 T _nss_ldap_getservent_r aec0 T _nss_ldap_getspent_r ae20 T _nss_ldap_getspnam_r 96e0 T _nss_ldap_initgroups 9400 T _nss_ldap_initgroups_dyn aff0 T _nss_ldap_setaliasent d1e0 T _nss_ldap_setautomntent cda0 T _nss_ldap_setetherent 97f0 T _nss_ldap_setgrent a430 T _nss_ldap_sethostent a840 T _nss_ldap_setnetent 9b40 T _nss_ldap_setnetgrent aaa0 T _nss_ldap_setprotoent 81f0 T _nss_ldap_setpwent 9f60 T _nss_ldap_setrpcent b460 T _nss_ldap_setservent ae70 T _nss_ldap_setspent U access U ber_bvfree U ber_free U ber_pvt_opt_on U ber_set_option U calloc U close U dup U dup2 U ether_aton U ether_ntoa U fclose U fcntl U fgets U fileno U fopen U free U geteuid U getpeername U getpid U getsockname U gss_krb5_ccache_name U inet_addr U inet_makeaddr U inet_network U inet_ntoa U ldap_abandon U ldap_control_free U ldap_controls_free U ldap_count_entries U ldap_count_values U ldap_create_page_control U ldap_err2string U ldap_explode_dn U ldap_explode_rdn U ldap_first_attribute U ldap_first_entry U ldap_get_dn U ldap_get_option U ldap_get_values
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/07/2014 11:43 PM, Reco wrote: So, chances are - you've found a bug. Awesome, where/how should I report it? and memory info on top: KiB Mem: 2061128 total, 263680 used, 1797448 free,75180 buffers KiB Swap: 2094076 total,0 used, 2094076 free, 118000 cached I'd try adding some memory (or swapspace) to this VM to satisfy mremap's wish to remap to exactly 8G virtual memory. I'll give that a go and reply back, machine is in use right now so cant mess with memory, but will try asap. -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUXdwcAAoJEORnMHMHY2FrjEkP+QHkFC7KwuCMZoKpOGLAIqiv o1b7TsWisfunlBux8s7la9ipsbWZJ9lGy2rfSPuEGv/u5XGjiIODSRnXMD+H/DHq TQv7EQsdNPqEakcCL8XRTmFEkHQTy5Vb6Jjxt/bgusAcbYa/doItk0aVp56yqHY2 nTa6G8dr6/QkPDfGz4Gbby6xRdlMeHvEbtjix7vCA4fFW9bntE7O8/2SLR7RuCY2 6xM0zDNEfvznGPb5FFB4EVRBsJRzMXpcPDD0AHmIttMMSk1ZrXYbQEabxHDNQr8h 1R6IUsYUGRV+nVuPorYxeGzmNv2MJFRgSFwBGBS9+JHahoXf0L7fSyeqqVuRfIqy qaY6sUHv+b4drf0djKtY/69HWYmRF9st5b2ZgTdRj5nTp9PtqVZEchAjhNM9lF5V 8EReauywy6xa6fqSZMnz8Yf1FhzpTqGv0I50gmHF4p9oPcHJ7Dr06DBTcSsGPio0 I54bfrXxZ9NEu+NtdSbkLCUnBwzLrtQoyJexkzj+eon/MST2uUZxX4/i2OLOPZ61 1GXtTok5JdR/G8y0GphsqM6VJJqUWQrHRvMNpVkyD7ddldWmUHAiq+WovSP5o11w vGSEw3KCheIn6d0gP2vw9/RW7n+ZzFTwKPBiKdvzrBtlRhUNxm48d7mECuij7aGh MdlZg++FFBbvHZ3KCzNE =Sn9q -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545ddc21.5030...@linux.com
Re: useradd segmentation fault
Hi. On Sat, 08 Nov 2014 01:02:25 -0800 Joris Bolsens jo...@linux.com wrote: So, chances are - you've found a bug. Awesome, where/how should I report it? Install reportbug package. Invoke 'reportbug passwd'. The rest of the process will be explained to you by the reportbug. Reportbug requires sending specially-crafted e-mail to sub...@bugs.debian.org, so you may want to produce such e-mail with reportbug, and send it by hand to that address. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108122101.8583d128c28636c13b972...@gmail.com
Re: useradd segmentation fault
On Fri, 07 Nov 2014, Joris Bolsens wrote: Ran into a bit of a strange error with useradd today, whenever i run useradd I get a Segmentation fault. Before filing a bug, please 1) Enable coredumps 2) Get a coredump 3) Backtrace it 4) Install any missing -dbg packages which provide symbols for the bits which are segfaulting If you are still missing the appropriate symbols for useradd, please rebuild it with DEB_BUILD_OPTS=nostrip, and try again. -- Don Armstrong http://www.donarmstrong.com With one simple pill we cured unhappiness and art -- a softer world #437 http://www.asofterworld.com/index.php?id=437 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108173937.gs29...@teltox.donarmstrong.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sorry, pretty new to linux still. put a few questions in-line. On 11/08/2014 09:39 AM, Don Armstrong wrote: Before filing a bug, please 1) Enable coredumps Did some googling, I have to recompile the kernel for this? the one thing i did find on how to do it without was for the fedora kernel. 2) Get a coredump 3) Backtrace it Not sure how to do this, or what I would be looking for here. -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUXmXyAAoJEORnMHMHY2FrK68P/Azq1Bu+NHmB1rK8c3DAJIhW rNHdbA+S+6J1dcgYVCLB+SSldUR+14z6N5hUzgm86LjyCeKaYrpcTF+wxJJ3gvvn CP6Bbk/V7sAK56/4OwVYBu+DYWTVJnm6dOGVO03+vUFWXPjkQUmIopabq6IV6zNb hjyTCOEUBpUUrfDoowiny/U5GuBG6vit77ukaAhrlAjTk0nDZM2seHG+Ueh78gRf s8XUeSG9KAoaB5dq3i1VzvKENcImiRzzkNY2WNSZwX0oodlUGD9+i6iVwka8o9cX kjDIagVAc6sZQBeOShc+YSgQ7kwwROpz8uc6IqPuC4ZMH8zpZe8xjsatLIEDBVNJ pdEXjG8EBs4ZI11pi4/NyHsMdm31pySxPcikY1BJw7qyJYhBXXw+hBveFCbKGqD7 KDAyDUwXov+Kv81uvb0phLgYv2PuSuiGhQkGlq/6HWPsXlrHbAShSIFwg7NUZbbm wEMJksILjaP8pe9oX9O8/vozJQ1LUd/yOPkio9PnEyHDUl6CYjVEsBWvEAcAEpRq OCpL+GW5vT2eoZ+Ut16WxSgboQcPhOxpxviRheM4S9AMJp4m7aF45q624gdQj/Et QzLnbJEgc23Ccnh08WJIfr0e59jFeABhkMuXBDbey75M2q8RNhqU7cpnMec4p+47 47U1Wdgjgdcf0Xcw/T1e =r/PK -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545e65f8.6060...@gmail.com
Re: useradd segmentation fault
On Sat, 08 Nov 2014, Joris Bolsens wrote: Sorry, pretty new to linux still. put a few questions in-line. No worries. On 11/08/2014 09:39 AM, Don Armstrong wrote: Before filing a bug, please 1) Enable coredumps Did some googling, I have to recompile the kernel for this? the one thing i did find on how to do it without was for the fedora kernel. Nah. All you have to do is something like the following: ulimit -c unlimited; useradd foo; gdb /usr/sbin/useradd core; bt full; or similar. You'll also want to install libc6-dbg gdb; and similar. This will give you an idea of where the code is segfaulting, and whether it's in a library which useradd is calling, or something else entirely. If the backtrace doesn't look at all useful, something like: apt-get build-dep passwd; DEB_BUILD_OPTS=nostrip,noopt apt-get source -b passwd; should in theory give you a passwd .deb file with debugging symbols and no optimization which you can then install to give you a more useful backtrace. You also might start by just reinstalling passwd in case there was some corruption somewhere to make sure you can replicate the segfault. -- Don Armstrong http://www.donarmstrong.com I finally developed a computer with feelings. It just doesn't have feelings for me. -- a softer world #633 http://www.asofterworld.com/index.php?id=633 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108191639.gv29...@teltox.donarmstrong.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/08/2014 11:16 AM, Don Armstrong wrote: ulimit -c unlimited; useradd foo; gdb /usr/sbin/useradd core; bt full; or similar. You'll also want to install libc6-dbg gdb; and similar. This will give you an idea of where the code is segfaulting, and whether it's in a library which useradd is calling, or something else entirely. If the backtrace doesn't look at all useful, something like: apt-get build-dep passwd; DEB_BUILD_OPTS=nostrip,noopt apt-get source -b passwd; should in theory give you a passwd .deb file with debugging symbols and no optimization which you can then install to give you a more useful backtrace. You also might start by just reinstalling passwd in case there was some corruption somewhere to make sure you can replicate the segfault. Awesome, ill give all this a go and also maybe try settings its ram to 8gb and update with what happens. -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUXm2zAAoJEORnMHMHY2FrZgAQAJWST7FWrcHKkkpCEAg7NSHQ dHaE6ASyDuIY3iMtyGSrVRogHC9C0rooL0tEaUekynJcG5MH96XaWZoDAIZ0APpb fBfFigvRZF+LLFsU4N4CG9GvPsSX5MIFGP9zqme51j22IUCqVUedS25D4Rd4qLeV qa0GvA4xStk/d3pCKsVOIt8tC302PV5OcFuvfzrVBPKRRxUweXAP4wiQupb9UzVr iXdqlc1UtFaOUhvu6KGC0iF7VzLBn6hZnRPxUBQmzgphCkPGpRSxv4zdVG/s13VR Zapi2kOxtTqarpBtxG95Krp+E1lzlO70xOrOqjbUhGnBAzc1d4jW8PP8WhDmSzH7 MYmPpmLF7cWlIBwvsf0TOO+NL7Xk0yvl2p9bf6GnDsw5sstu/B7k3ISWgtybsvCD 5Ys4OzteYzM2ekfmKFmGuT8nZG03fjOdM8Wtn+XFg75/Vbdbp58ZECGdzYTTcFeB +AKkb7PHznu99I32djg7ICWQRGdFhe4EWB3W/Qo4u2AYcZgEn2jn+8Lz+agytTLh Zcf+2jvZoabT8VhCqYhuBi6peDLsZsV2k4GBR+/CFevYGD48s26b1+cAJgB0/dWp OJwAW5nSXTpIlXS21axZkyTkEF+mzPhBuNL6WJBOq5i1J/9fTFGQt5/L2wv/Olyj YD7hsadQ8qN5xBzO0GQd =iKYV -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545e6dba.5000...@linux.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I can't be 100% certain because I'm not sure what any of this means,
but it would seem I messed up somewhere as it shows `No Symbol Table
info available` for libnss_ldap.
This bit also seems possibly problematic
`warning: Can't read pathname for load map: Input/output error.1`
# gdb /usr/sbin/useradd core
GNU gdb (GDB) 7.4.1-debian
Copyright (C) 2012 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-linux-gnu.
For bug reporting instructions, please see:
http://www.gnu.org/software/gdb/bugs/...
Reading symbols from /usr/sbin/useradd...(no debugging symbols
found)...done.
[New LWP 9929]
warning: Can't read pathname for load map: Input/output error.
[Thread debugging using libthread_db enabled]
Using host libthread_db library /lib/x86_64-linux-gnu/libthread_db.so.1.
Core was generated by `useradd foo'.
Program terminated with signal 11, Segmentation fault.
#0 0x7f880fad5f0b in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
(gdb) bt full
#0 0x7f880fad5f0b in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#1 0x7f880fad3ef1 in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#2 0x7f880fad4a34 in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#3 0x7f880fad4bde in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#4 0x7f880fad6276 in _nss_ldap_getpwent_r () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#5 0x7f88108f1840 in __nss_getent_r
(getent_func_name=0x7f8810952156 getpwent_r,
setent_func_name=optimized out,
lookup_fct=0x7f88108f1ad0 *__GI___nss_passwd_lookup2,
nip=0x7f8810b8cd60, startp=optimized out, last_nip=optimized out,
stayopen_tmp=0x0, res=0, resbuf=0x7f8810b8cca0, buffer=0x0,
buflen=8589934592, result=0x7fffbff02008, h_errnop=0x0)
at getnssent_r.c:171
fct = {f = 0x7f880fad6240 _nss_ldap_getpwent_r, ptr =
0x7f880fad6240}
no_more = optimized out
status = 277813968
#6 0x7f88108af0a6 in __getpwent_r (resbuf=optimized out,
buffer=optimized out, buflen=optimized out, result=0x0)
at ../nss/getXXent_r.c:162
status = 0
save = 0
#7 0x7f88108f1463 in __nss_getent (func=0x7f88108af010
__getpwent_r, resbuf=0x7f8810b8cca0, buffer=0x7f8810b8ae18,
buflen=optimized out, buffer_size=0x7f8810b8ccd0, h_errnop=0x0)
at getnssent.c:38
result = optimized out
#8 0x7f88108aece2 in getpwent () at ../nss/getXXent.c:84
buffer_size = 8589934592
resbuf = {l = {pw_name = 0x7f860d33d012 Address
0x7f860d33d012 out of bounds, pw_passwd = 0x0, pw_uid = 628342823,
pw_gid = 3, pw_gecos = 0x1637b49 ¸\020\210\177,
pw_dir = 0x1637b50 0{c\001, pw_shell = 0x1637b5d },
ptr = 0x7f860d33d012}
save = 0
#9 0x00408815 in ?? ()
No symbol table info available.
#10 0x004043e5 in ?? ()
No symbol table info available.
#11 0x7f8810822ead in __libc_start_main (main=optimized out,
argc=optimized out, ubp_av=optimized out,
init=optimized out, fini=optimized out, rtld_fini=optimized
out, stack_end=0x7fffbff023f8) at libc-start.c:244
result = optimized out
unwind_buf = {cancel_jmp_buf = {{jmp_buf = {0,
4098621885896654353, 4218724, 140736413574144, 0, 0,
- -4098762729259336175,
-4039416709317984751}, mask_was_saved = 0}}, priv =
{pad = {0x0, 0x0, 0x40ff50, 0x7fffbff02408}, data = {
prev = 0x0, cleanup = 0x0, canceltype = 4259664}}}
not_first_call = optimized out
#12 0x00405f8d in ?? ()
No symbol table info available.
#13 0x7fffbff023f8 in ?? ()
No symbol table info available.
#14 0x001c in ?? ()
No symbol table info available.
#15 0x0002 in ?? ()
No symbol table info available.
#16 0x7fffbff02e90 in ?? ()
No symbol table info available.
#17 0x7fffbff02e98 in ?? ()
No symbol table info available.
#18 0x in ?? ()
No symbol table info available.
(gdb)
#0 0x7f880fad5f0b in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#1 0x7f880fad3ef1 in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#2 0x7f880fad4a34 in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#3 0x7f880fad4bde in ?? () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#4 0x7f880fad6276 in _nss_ldap_getpwent_r () from
/lib/x86_64-linux-gnu/libnss_ldap.so.2
No symbol table info available.
#5 0x7f88108f1840 in __nss_getent_r
(getent_func_name=0x7f8810952156 getpwent_r,
Re: useradd segmentation fault
On Sat, 08 Nov 2014, Joris Bolsens wrote: I can't be 100% certain because I'm not sure what any of this means, but it would seem I messed up somewhere as it shows `No Symbol Table info available` for libnss_ldap. Awesome. It looks like it's segfaulting in libnss_ldap, which you didn't build symbols for (and unfortunately, they're not currently available). apt-get build-dep libnss-ldap; DEB_BUILD_OPTS=nostrip,noopt apt-get source -b libnss-ldap; will give you those symbols. Then you can rerun /usr/sbin/useradd foo; and re-analyze the coredump. Please double check the bt full; output for any obvious password strings too before posting to this mailing list. [I should have told you to do that too, but I don't see any below, so you should be OK.] And also don't send the core file to the BTS either, unless you're going to change your ldap machine passwords. -- Don Armstrong http://www.donarmstrong.com The game of science is, in principle, without end. He who decides one day that scientific statements do not call for any further test, and that they can be regarded as finally verified, retires from the game. -- Sir Karl Popper _The Logic of Scientific Discovery_ §11 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108201650.gx29...@teltox.donarmstrong.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/08/2014 12:16 PM, Don Armstrong wrote: Please double check the bt full; output for any obvious password strings too before posting to this mailing list. [I should have told you to do that too, but I don't see any below, so you should be OK.] Yea i did a quick read through to make sure no identifying information was present before sending :p And also don't send the core file to the BTS either, unless you're going to change your ldap machine passwords. Not sure what you mean by this, don't post the full output of the command? -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUXp6hAAoJEORnMHMHY2Frg54QAKhSRkfZuH+SNXXb4daaDJ8E USRZirhICEnvtzeDu6Bz4BsCG6hX0Zv6PXufQvXQcqHodvbfWBBbOPjE2CDHlnhI oKewZJdYcWTym9HYxwBN5Y/FG+gyNygiVsSBFja37P+LkpH6I52BdNSyIhTDkukX 1EgKbWntn+AVCQJoxpuZ0rNUbdO7YTsYkgo+CBA+1vg46H68AqJ8DGQ5KrYDWu/g i0N5bS7j7SyZnO4L5jolWNeBXDW6vh/JIYa8uNc24jj8ndg4LmjJokH56jDFsA6C mS4j69sWjaHcm84BuIi9mgzcB6feiWFP7M5qkj6LrzEUutjE+Tur85pPR0ZfDy4U d5CgzVkrqAOh0qCPWQt6onozpoVjPXvT1V+KrwihSoNGkM2hebzNDOH+SljaYRO7 qO5FhWqeGj3nsNXe99yDeWAYzGXNToajTlOl9Z2DYYUF5ps3EAsqVFspMCskMrxZ o7ItPu81IATdtNknGQ0yxubXCiGiw0cxbuEugdNaF6VLKoNAnPmA68laJOrty5T1 H4A9gsq6kBaAFUZLaU501jzmR3Z0quYgzV9A6b81jZNphr2f1GXtXzdai6jsJ6yk yMDM6J+Ce7/oZXfYnMKDRrK7Bt6NrXd79NNSzucpF3YmavKg9bwTwKfKzQMHlp0R Trv80S1QtZJkjaZJY2Eu =sk4N -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545e9ea4.3050...@gmail.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 11/08/2014 12:16 PM, Don Armstrong wrote: apt-get build-dep libnss-ldap; DEB_BUILD_OPTS=nostrip,noopt apt-get source -b libnss-ldap; will give you those symbols. Then you can rerun /usr/sbin/useradd foo; and re-analyze the coredump. The coredump looks identical to before i built/installed the new libnss-ldap -BEGIN PGP SIGNATURE- Version: GnuPG v1 iQIcBAEBAgAGBQJUXqEhAAoJEORnMHMHY2FrDa0P/1OLykX1o9050yn1Lg/W6W3g cTO6qQNF6big1tikRyjvJ4vubo1OYcuZyxVZxHL7SGMxX7S7ndzRpZWPmEgyg+nG wj0GxZoVN8tJTREEaa5foIw4nSO9glPjM61gdCUKdHSnyBFy3Nq+c3asLhb8UGBY o6OPshKbFL8C/bbK1AJvMVoi3tu7LmmPJLHfMefXun8Z4ENa4uO/xfrLL8dK2Wie KzOAG3FpoKdTOAJ5uQdjM9pu8joTZq3/zWYfOyvQoImKZkdFb2whjFHDK1hu88Bf UHxsf7pgMIyqt7qWEN9ExjEstiwTSqTJXkP2I731q/UAaHpJQUjAdqwpZVtwptd3 U/YTd8LVKExZBG/jY5VUasDZ++D7g08jzI61+B27+syFGFGh7mUZMvGj9LHzitqb N5ckjk1ItlMPJr2YoJ8Jv8cTYrNw9KD7YuT/bKjd9zr89MIUcieX9JjVZndvWZQB Nmeo/vE3KqnScGou6yxuS1YNgAZoGYSPHQg83MlgVRrrGFqaQOaBifwIgp12yJYz p66stYt8eiZTJcgQ5VMOpgeo2zoiJpbi0S+mtv8RK2zVXeVDpMpTFqWd31Xo0S0m cIKTzei0SJJ6NaNVXUa85jfaMupE0anANeL5ofg1i1CcMYhuJSIrHVJ9YYf4agL4 rMtBcSUWdeKd9xLZE9D1 =5Ry2 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545ea123.9050...@gmail.com
Re: useradd segmentation fault
On Sat, 08 Nov 2014, Joris Bolsens wrote: Not sure what you mean by this, don't post the full output of the command? No, just don't send the whole core dump itself (the file called core) to the BTS (or this mailing list). It almost certainly has your machine ldap password in it. On Sat, 08 Nov 2014, Joris Bolsens wrote: The coredump looks identical to before i built/installed the new libnss-ldap You should at least have the symbols for libnss-ldap... can you check the output of nm -a for that library to make sure it actually has the debugging symbols? -- Don Armstrong http://www.donarmstrong.com Of course Pacman didn't influence us as kids. If it did, we'd be running around in darkened rooms, popping pills and listening to repetitive music. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108234354.gz29...@teltox.donarmstrong.com
Re: useradd segmentation fault
it seems that using the adduser works fine, not sure why i didn't think to try that before. However I am still curious why this occurred. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545d84dc.7050...@linux.com
Re: useradd segmentation fault
On 11/07/2014 09:04 PM, Joris Bolsens wrote:
Ran into a bit of a strange error with useradd today,
whenever i run useradd I get a Segmentation fault.
I tried running fsck as i read it might be due to corrupt filesystem, but that
didn't report any problems.
I reinstalled the passwd package, also to no avail.
I ran an strace and it seems it cannot allocate memory, which is strange as it
is running as a vm on a machine with 32gb of ram,
and is provisioned to use whatever it needs and has 2gb reserved to it, not to
mention that top shows only a fraction of that being in use.
I recently added ldap integration if that matters at all,
although none of the other VMs seem to be having issues with this
(granted the others are running CentOs, this being the ldap server)
Output of strace:
rt_sigaction(SIGPIPE, {SIG_DFL, [], SA_RESTORER, 0x7f259c6461e0}, NULL, 8) = 0
No idea, but why are you using realtime (rt_*) system calls?
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/545d9b70.1060...@ix.netcom.com
Re: useradd segmentation fault
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
No idea, that's just what i got when I ran the command, I left out a
bit at the top that did the ldap lookups, that all looked find and
didn't want to include that as 1) dont think it's relevant and 2)
contains some sensitive data (usernames and such), do you think that
could have something to do with it?
On 11/07/2014 08:26 PM, Marty wrote:
On 11/07/2014 09:04 PM, Joris Bolsens wrote:
Ran into a bit of a strange error with useradd today, whenever i
run useradd I get a Segmentation fault.
I tried running fsck as i read it might be due to corrupt
filesystem, but that didn't report any problems. I reinstalled
the passwd package, also to no avail.
I ran an strace and it seems it cannot allocate memory, which is
strange as it is running as a vm on a machine with 32gb of ram,
and is provisioned to use whatever it needs and has 2gb reserved
to it, not to mention that top shows only a fraction of that
being in use.
I recently added ldap integration if that matters at all,
although none of the other VMs seem to be having issues with
this (granted the others are running CentOs, this being the ldap
server)
Output of strace:
rt_sigaction(SIGPIPE, {SIG_DFL, [], SA_RESTORER,
0x7f259c6461e0}, NULL, 8) = 0
No idea, but why are you using realtime (rt_*) system calls?
-BEGIN PGP SIGNATURE-
Version: GnuPG v1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=9URa
-END PGP SIGNATURE-
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/545d9c2b.3060...@linux.com
Re: useradd segmentation fault
On 11/07/2014 11:29 PM, Joris Bolsens wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 No idea, that's just what i got when I ran the command, I left out a bit at the top that did the ldap lookups, that all looked find and didn't want to include that as 1) dont think it's relevant and 2) contains some sensitive data (usernames and such), do you think that could have something to do with it? Well I think yes if that's what you are doing, Linux realtime support is still experimental -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/545da752.3040...@ix.netcom.com
Re: useradd segmentation fault
Hi. On Fri, 07 Nov 2014 18:04:25 -0800 Joris Bolsens jo...@linux.com wrote: skip mremap(0x7f239914d000, 4294971392, 8589938688, MREMAP_MAYMOVE) = -1 EFAULT (Bad address) It goes downhill from here. Apparently useradd does not checks a return status of mremap, and, according to the mremap(2), EFAULT is defined as: EFAULT Segmentation fault. Some address in the range old_address to old_address+old_size is an invalid virtual memory address for this process. You can also get EFAULT even if there exist mappings that cover the whole address space requested, but those mappings are of different types. So, chances are - you've found a bug. and memory info on top: KiB Mem: 2061128 total, 263680 used, 1797448 free,75180 buffers KiB Swap: 2094076 total,0 used, 2094076 free, 118000 cached I'd try adding some memory (or swapspace) to this VM to satisfy mremap's wish to remap to exactly 8G virtual memory. and another interesting not, groupadd works fine, but groupdel also gives a segmentation fault. # groupadd testgrp # groupdel testgrp Segmentation fault And groupdel fails at the same output of strace on groupdel: mremap(0x7f7f71f1c000, 4294971392, 8589938688, MREMAP_MAYMOVE) = -1 EFAULT (Bad address) Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20141108104313.fa2a07ed7b979d9b0d829...@gmail.com
