Kenneth Marshall
2007-Dec-17 19:57 UTC
[Samba] problem with samba 3.0.28/Solaris 8/smbclient
Dear Samba users,
I am trying to update our local samba packages to 3.0.28.
They are built against heimdal-1.0.1 and openldap-2.3.38.
The Redhat Enterprise Linux 3 and 4 packages are working
fine so far in my limited testing. The problem with heimdal
and "net ads join..." has been fixed on all 3 platforms.
On the Solaris 8 server, the "net ads join..." works
correctly and the machine registers with the domain. The
problem is that the smbclient segfaults on the ls command.
A put command worked correctly. Here is what a truss
returned:
21056: ioctl(1, TCGETA, 0xFFBECFA4) = 0
21056: write(1, " s m b : \ k \ k t m \".., 14) = 14
21056: read(0, 0xFFBED3E3, 1) (sleeping...)
21056: signotifywait() (sleeping...)
21056: door_return(0x00000000, 0, 0x00000000, 0) (sleeping...)
21056: lwp_cond_wait(0xFF3834E8, 0xFF3834F8, 0xFF37CD80) (sleeping...)
21056: read(0, " l", 1) = 1
21056: write(1, " l", 1) = 1
21056: read(0, " s", 1) = 1
21056: write(1, " s", 1) = 1
21056: read(0, "\r", 1) = 1
21056: write(1, "\n", 1) = 1
21056: ioctl(0, TCSETSW, 0xFF345C28) = 0
21056: sigaction(SIGINT, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGTERM, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGQUIT, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGALRM, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGTSTP, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGTTOU, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGTTIN, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: sigaction(SIGWINCH, 0xFFBED4C0, 0xFFBED5C0) = 0
21056: time() = 1197919173
21056: write(6, "\0\0\0 bFF S M B 2\0\0\0".., 102) = 102
21056: poll(0xFFBEAF18, 1, 20000) = 1
21056: read(6, "\0\003AC", 4) = 4
21056: poll(0xFFBEAF88, 1, 20000) = 1
21056: read(6, "FF S M B 2\0\0\0\08801C8".., 940) = 940
21056: open("/usr/share/lib/zoneinfo/US/Central", O_RDONLY) = 8
21056: read(8, " T Z i f\0\0\0\0\0\0\0\0".., 8192) = 1279
21056: close(8) = 0
21056: stat64("/usr/site/samba-3.0.28/lib/C.msg", 0xFFBEA948) Err#2
ENOENT
21056: Incurred fault #6, FLTBOUNDS %pc = 0xFE833218
21056: siginfo: SIGSEGV SEGV_MAPERR addr=0x40498000
21056: Received signal #11, SIGSEGV [default]
21056: siginfo: SIGSEGV SEGV_MAPERR addr=0x40498000
21056: *** process killed ***
Here is a gdb backtrace:
(gdb) run -U 'AD\ktm' //storage.rice.edu/home -D k/ktm
Starting program: /build/samba-3.0.28/sunos5/bin/smbclient -U 'AD\ktm'
//storage.rice.edu/home -D k/ktm
[New LWP 1]
[New LWP 2]
[New LWP 3]
params.c:OpenConfFile() - Unable to open configuration file
"/usr/site/samba-3.0.28/lib/smb.conf":
No such file or directory
[New LWP 4]
[New LWP 5]
/build/samba-3.0.28/sunos5/bin/smbclient: Can't load
/usr/site/samba-3.0.28/lib/smb.conf - run testparm to debug it
Password:
Domain=[AD] OS=[BlueArc Titan 4.3.990q] Server=[BlueArc Titan 4.3.990q]
smb: \k\ktm\> ls
Program received signal SIGSEGV, Segmentation fault.
0xfe833218 in strlen () from /usr/lib/libc.so.1
(gdb) bt
#0 0xfe833218 in strlen () from /usr/lib/libc.so.1
#1 0xfedcadcc in wait_for_process () from
/usr/site/samba-3.0.28/ext/libroken.so.18
#2 0xfedcadcc in wait_for_process () from
/usr/site/samba-3.0.28/ext/libroken.so.18
Previous frame identical to this frame (corrupt stack?)
(gdb)
I am building a debug version of the heimdal 1.0.1 libraries to
try and get some more information, but I wanted to post a preliminary
message to see if this is a know issue or not and whether there are
some other debugging measures that could be tried.
Cheers,
Ken
Kenneth Marshall
2007-Dec-18 13:15 UTC
[Samba] problem with samba 3.0.28/Solaris 8/smbclient
On Mon, Dec 17, 2007 at 01:26:51PM -0600, Kenneth Marshall wrote:> Dear Samba users, > > I am trying to update our local samba packages to 3.0.28. > They are built against heimdal-1.0.1 and openldap-2.3.38. > The Redhat Enterprise Linux 3 and 4 packages are working > fine so far in my limited testing. The problem with heimdal > and "net ads join..." has been fixed on all 3 platforms. > On the Solaris 8 server, the "net ads join..." works > correctly and the machine registers with the domain. The > problem is that the smbclient segfaults on the ls command. > A put command worked correctly. Here is what a truss > returned: > > 21056: ioctl(1, TCGETA, 0xFFBECFA4) = 0 > 21056: write(1, " s m b : \ k \ k t m \".., 14) = 14 > 21056: read(0, 0xFFBED3E3, 1) (sleeping...) > 21056: signotifywait() (sleeping...) > 21056: door_return(0x00000000, 0, 0x00000000, 0) (sleeping...) > 21056: lwp_cond_wait(0xFF3834E8, 0xFF3834F8, 0xFF37CD80) (sleeping...) > 21056: read(0, " l", 1) = 1 > 21056: write(1, " l", 1) = 1 > 21056: read(0, " s", 1) = 1 > 21056: write(1, " s", 1) = 1 > 21056: read(0, "\r", 1) = 1 > 21056: write(1, "\n", 1) = 1 > 21056: ioctl(0, TCSETSW, 0xFF345C28) = 0 > 21056: sigaction(SIGINT, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGTERM, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGQUIT, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGALRM, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGTSTP, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGTTOU, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGTTIN, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: sigaction(SIGWINCH, 0xFFBED4C0, 0xFFBED5C0) = 0 > 21056: time() = 1197919173 > 21056: write(6, "\0\0\0 bFF S M B 2\0\0\0".., 102) = 102 > 21056: poll(0xFFBEAF18, 1, 20000) = 1 > 21056: read(6, "\0\003AC", 4) = 4 > 21056: poll(0xFFBEAF88, 1, 20000) = 1 > 21056: read(6, "FF S M B 2\0\0\0\08801C8".., 940) = 940 > 21056: open("/usr/share/lib/zoneinfo/US/Central", O_RDONLY) = 8 > 21056: read(8, " T Z i f\0\0\0\0\0\0\0\0".., 8192) = 1279 > 21056: close(8) = 0 > 21056: stat64("/usr/site/samba-3.0.28/lib/C.msg", 0xFFBEA948) Err#2 ENOENT > 21056: Incurred fault #6, FLTBOUNDS %pc = 0xFE833218 > 21056: siginfo: SIGSEGV SEGV_MAPERR addr=0x40498000 > 21056: Received signal #11, SIGSEGV [default] > 21056: siginfo: SIGSEGV SEGV_MAPERR addr=0x40498000 > 21056: *** process killed *** > > Here is a gdb backtrace: > > (gdb) run -U 'AD\ktm' //storage.rice.edu/home -D k/ktm > Starting program: /build/samba-3.0.28/sunos5/bin/smbclient -U 'AD\ktm' //storage.rice.edu/home -D k/ktm > [New LWP 1] > [New LWP 2] > [New LWP 3] > params.c:OpenConfFile() - Unable to open configuration file "/usr/site/samba-3.0.28/lib/smb.conf": > No such file or directory > [New LWP 4] > [New LWP 5] > /build/samba-3.0.28/sunos5/bin/smbclient: Can't load /usr/site/samba-3.0.28/lib/smb.conf - run testparm to debug it > Password: > Domain=[AD] OS=[BlueArc Titan 4.3.990q] Server=[BlueArc Titan 4.3.990q] > smb: \k\ktm\> ls > > Program received signal SIGSEGV, Segmentation fault. > 0xfe833218 in strlen () from /usr/lib/libc.so.1 > (gdb) bt > #0 0xfe833218 in strlen () from /usr/lib/libc.so.1 > #1 0xfedcadcc in wait_for_process () from /usr/site/samba-3.0.28/ext/libroken.so.18 > #2 0xfedcadcc in wait_for_process () from /usr/site/samba-3.0.28/ext/libroken.so.18 > Previous frame identical to this frame (corrupt stack?) > (gdb) > > I am building a debug version of the heimdal 1.0.1 libraries to > try and get some more information, but I wanted to post a preliminary > message to see if this is a know issue or not and whether there are > some other debugging measures that could be tried. > > Cheers, > Ken > -- > To unsubscribe from this list go to the following URL and read the > instructions: https://lists.samba.org/mailman/listinfo/samba >Dear Samba users, As promised here s the backtrace from the debug version of the Heimdal 1.0.1 libraries: Starting program: /build/samba-3.0.28/sunos5/bin/smbclient -U 'ADRICE\ktm' //storage.rice.edu/home -D k/ktm [New LWP 1] [New LWP 2] [New LWP 3] params.c:OpenConfFile() - Unable to open configuration file "/usr/site/samba-3.0.28/lib/smb.conf": No such file or directory [New LWP 4] [New LWP 5] /build/samba-3.0.28/sunos5/bin/smbclient: Can't load /usr/site/samba-3.0.28/lib/smb.conf - run testparm to debug it Password: Domain=[ADRICE] OS=[BlueArc Titan 4.3.990q] Server=[BlueArc Titan 4.3.990q] smb: \k\ktm\> ls Program received signal SIGSEGV, Segmentation fault. 0xfe733218 in strlen () from /usr/lib/libc.so.1 (gdb) bt #0 0xfe733218 in strlen () from /usr/lib/libc.so.1 #1 0xfef50b50 in append_string (state=0xffbeb130, arg=0x40498000 <Address 0x40498000 out of bounds>, width=0, prec=-1, flags=0) at snprintf.c:276 #2 0xfef513cc in xyzprintf (state=0xffbeb130, char_format=0x1e7318 " %-30s%7.7s %8.0f %s", ap=0xffbeb33c) at snprintf.c:446 #3 0xfef5213c in vasnprintf (ret=0xffbeb248, max_sz=0, format=0x1e7318 " %-30s%7.7s %8.0f %s", args=0xffbeb330) at snprintf.c:661 #4 0xfef52074 in vasprintf (ret=0xffbeb248, format=0x1e7318 " %-30s%7.7s %8.0f %s", args=0xffbeb330) at snprintf.c:638 #5 0x0013afb0 in d_vfprintf (f=0xfe7c0294, format=0x1e7318 " %-30s%7.7s %8.0f %s", ap=0xffbeb330) at lib/dprintf.c:47 #6 0x0013b250 in d_printf (format=0x1e7318 " %-30s%7.7s %8.0f %s") at lib/dprintf.c:102 #7 0x00031918 in display_finfo (finfo=0xffbec0d0) at client/client.c:388 #8 0x00032e98 in do_list_helper (mntpoint=0x252db4 "", f=0xffbec0d0, mask=0xffbed254 "\\k\\ktm\\*", state=0x0) at client/client.c:604 #9 0x000c5d0c in cli_list_new (cli=0x2531c0, Mask=0xffbed254 "\\k\\ktm\\*", attribute=22, fn=0x32b24 <do_list_helper>, state=0x0) at libsmb/clilist.c:355 #10 0x000c69d0 in cli_list (cli=0x2531c0, Mask=0xffbed254 "\\k\\ktm\\*", attribute=22, fn=0x32b24 <do_list_helper>, state=0x0) at libsmb/clilist.c:523 #11 0x0003334c in do_list (mask=0xffbedad6 "\\k\\ktm\\*", attribute=22, fn=0x317c8 <display_finfo>, rec=0, dirs=1) at client/client.c:677 #12 0x00033750 in cmd_dir () at client/client.c:725 #13 0x0003fe8c in process_stdin () at client/client.c:3694 #14 0x0004003c in process (base_directory=0xffbef758 "k\\ktm") at client/client.c:3728 #15 0x00041a74 in main (argc=6, argv=0xffbefbf4) at client/client.c:4132 (gdb) Does this point to a problem in the Samba code or is it a problem with the Heimdal library? Does anyone have any suggestions for further debugging? Cheers, Ken