netatalk / Bugs / #498 netatalk 3.0.2 "INTERNAL ERROR: Signal 11" on opening file in share

Ralph Böhme - 2013-02-27

http://netatalk.sourceforge.net/wiki/index.php/Developer_Infos#Debugging_process_crashes

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Ali J. - 2013-02-27
  
  Ralph: should I install the source version and try again? or do you just need the debug output?
  
  EDIT: building with debug flags right now
  
  Last edit: Ali J. 2013-02-27
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ralph Böhme - 2013-02-27

... "We need a stack-backtrace (SBT) from a corefile with debugging symbols." ...

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
- Ali J. - 2013-03-17
  
  ok - working on it - compiled netatalk with the debug flags, upped the ulimits:
  
  root@ubuntu:~# ulimit -a
  core file size (blocks, -c) unlimited
  data seg size (kbytes, -d) unlimited
  scheduling priority (-e) 0
  file size (blocks, -f) unlimited
  pending signals (-i) 7799
  max locked memory (kbytes, -l) 64
  max memory size (kbytes, -m) unlimited
  open files (-n) 1024
  pipe size (512 bytes, -p) 8
  POSIX message queues (bytes, -q) 819200
  real-time priority (-r) 0
  stack size (kbytes, -s) 8192
  cpu time (seconds, -t) unlimited
  max user processes (-u) 7799
  virtual memory (kbytes, -v) unlimited
  file locks (-x) unlimited
  
  but I'm not getting crash dumps in /var/crash
  
  attaching with gdb with the pid is tough since the process always crashes and doesn't stay long enough for me to attach, attaching to afpd directly doesn't give me a stack:
  
  root@ubuntu:~# gdb /usr/local/sbin/afpd
  GNU gdb (Ubuntu/Linaro 7.4-2012.04-0ubuntu2.1) 7.4-2012.04
  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://bugs.launchpad.net/gdb-linaro/...
  Reading symbols from /usr/local/sbin/afpd...done.
  (gdb) bt full
  No stack.
  (gdb) q
  
  Last edit: Ali J. 2013-03-17
  
  If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ali J. - 2013-03-17

additonal infos to reproduce this:

ubuntu 12.04
netatalk 3.0.2
using symlinks
and a file located in:

/home/username/folder/symlink(->somewhere/else)/file

just opening the file freezes the opening program in os x, the program can only be force exited

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ali J. - 2013-03-17

nvm I think I got them crashdumps, additional info: on this machine I additionally get a moving error:

Mar 17 15:17:49.797604 afpd[3444] {auth.c:242} (N:AFPDaemon): AFP3.3 Login by admin
Mar 17 15:17:49.798168 afpd[3444] {auth.c:591} (N:AFPDaemon): afp_disconnect: trying primary reconnect
Mar 17 15:17:49.798258 afpd[2488] {server_child.c:268} (N:Default): Reconnect: no child[3429]
Mar 17 15:17:54.798542 afpd[3444] {auth.c:624} (E:AFPDaemon): afp_disconnect: primary reconnect failed
Mar 17 15:17:54.814567 afpd[3444] {unix.c:97} (N:Default): run_cmd("mv"): status: 1
Mar 17 15:17:54.814627 afpd[3444] {desktop.c:218} (E:AFPDaemon): moving .AppleDesktop from "/home/protonet/dashboard/shared/files/system_users/admin/.AppleDesktop" to "/usr/local/var/netatalk/CNID//admin's home//.AppleDesktop" failed
Mar 17 15:17:54.816307 afpd[3444] {fault.c:123} (S:Default): ===============================================================
Mar 17 15:17:54.816345 afpd[3444] {fault.c:124} (S:Default): INTERNAL ERROR: Signal 11 in pid 3444 (3.0.2)
Mar 17 15:17:54.816353 afpd[3444] {fault.c:125} (S:Default): ===============================================================
Mar 17 15:17:54.817113 afpd[3444] {fault.c:96} (S:Default): PANIC: internal error
Mar 17 15:17:54.817128 afpd[3444] {fault.c:97} (S:Default): BACKTRACE: 11 stack frames:
Mar 17 15:17:54.817136 afpd[3444] {fault.c:103} (S:Default): #0 /usr/local/lib/libatalk.so.3(netatalk_panic+0x26) [0x7f7d5ba6a9e0]
Mar 17 15:17:54.817143 afpd[3444] {fault.c:103} (S:Default): #1 /usr/local/lib/libatalk.so.3(+0x42bfa) [0x7f7d5ba6abfa]
Mar 17 15:17:54.817150 afpd[3444] {fault.c:103} (S:Default): #2 /usr/local/lib/libatalk.so.3(+0x42c4e) [0x7f7d5ba6ac4e]
Mar 17 15:17:54.817158 afpd[3444] {fault.c:103} (S:Default): #3 /lib/x86_64-linux-gnu/libc.so.6(+0x364a0) [0x7f7d5ab934a0]
Mar 17 15:17:54.817165 afpd[3444] {fault.c:103} (S:Default): #4 /usr/local/sbin/afpd(cname+0x859) [0x41af08]
Mar 17 15:17:54.817172 afpd[3444] {fault.c:103} (S:Default): #5 /usr/local/sbin/afpd(afp_getfildirparams+0x12a) [0x428235]
Mar 17 15:17:54.817179 afpd[3444] {fault.c:103} (S:Default): #6 /usr/local/sbin/afpd(afp_over_dsi+0x741) [0x40c205]
Mar 17 15:17:54.817186 afpd[3444] {fault.c:103} (S:Default): #7 /usr/local/sbin/afpd() [0x42f338]
Mar 17 15:17:54.817193 afpd[3444] {fault.c:103} (S:Default): #8 /usr/local/sbin/afpd(main+0xb33) [0x42eec4]

core.3444

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ali J. - 2013-03-17

ALso, I've got a lot more coredumps ;)

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

[New LWP 3444]

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 `/usr/local/sbin/afpd -d -F /usr/local/etc/afp.conf'.
Program terminated with signal 6, Aborted.
#0  0x00007f7d5ab93425 in raise () from /lib/x86_64-linux-gnu/libc.so.6
(gdb) bt full
#0  0x00007f7d5ab93425 in raise () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#1  0x00007f7d5ab96b8b in abort () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#2  0x00007f7d5ba6ac39 in fault_report (sig=11) at fault.c:139
        counter = 1
#3  0x00007f7d5ba6ac4e in sig_fault (sig=11) at fault.c:147
No locals.
#4  <signal handler called>
No symbol table info available.
#5  0x000000000041af08 in cname (vol=0xbc9ef0, dir=0xbc9470, 
    cpath=0x7fffaecf59b0) at directory.c:1257
        path = "enable2start_foerderung_excel.xlsx", '\000' <repeats 4062 times>
        ret = {m_type = 3, 
          m_name = 0x64b8e0 "enable2start_foerderung_excel.xlsx", 
          u_name = 0x66181c "enable2start_foerderung_excel.xlsx", id = 0, 
          d_dir = 0x0, st_valid = 1, st_errno = 0, st = {st_dev = 2049, 
            st_ino = 1322925, st_nlink = 1, st_mode = 33204, st_uid = 8000, 
            st_gid = 1000, __pad0 = 0, st_rdev = 0, st_size = 31115, 
            st_blksize = 4096, st_blocks = 72, st_atim = {tv_sec = 1363529071, 
              tv_nsec = 891734436}, st_mtim = {tv_sec = 1319454817, 
              tv_nsec = 0}, st_ctim = {tv_sec = 1363527869, 
---Type <return> to continue, or q <return> to quit---
              tv_nsec = 509496000}, __unused = {0, 0, 0}}}
        cdir = 0xff9
        data = 0x7f7d5be6a042 ""
        p = 0x64b8ff "lsx"
        len = 0
        hint = 134217987
        len16 = 7936
        size = 7
        toUTF8 = 0
#6  0x0000000000428235 in afp_getfildirparams (obj=0x65e4c0, 
    ibuf=0x7f7d5be6a042 "", ibuflen=50, rbuf=0xbb0fa0 "", rbuflen=0xbb2fa0)
    at filedir.c:76
        st = 0x7f7d5ba54aea
        vol = 0xbc9ef0
        dir = 0xbc9470
        did = 369098752
        ret = 32767
        buflen = 12257456
        fbitmap = 59711
        dbitmap = 41791
        vid = 256
        s_path = 0x7fffaecf5a30
#7  0x000000000040c205 in afp_over_dsi (obj=0x65e4c0) at afp_dsi.c:626
---Type <return> to continue, or q <return> to quit---
        dsi = 0xbb08b0
        rc_idx = 11
        err = 0
        cmd = 2
        function = 34 '"'
        flag = 1
#8  0x000000000042f338 in dsi_start (obj=0x65e4c0, dsi=0xbb08b0, 
    server_children=0xbac0a0) at main.c:509
        child = 0x0
#9  0x000000000042eec4 in main (ac=4, av=0x7fffaecf5d68) at main.c:429
        i = 0
        sv = {__sigaction_handler = {sa_handler = 0x42dee1 <afp_goaway>, 
            sa_sigaction = 0x42dee1 <afp_goaway>}, sa_mask = {__val = {90625, 
              0 <repeats 15 times>}}, sa_flags = 268435456, sa_restorer = 0}
        sigs = {__val = {74241, 0 <repeats 15 times>}}
        ret = 1
        child = 0xbb4d50
        recon_ipc_fd = 0
        pid = 0
        saveerrno = 10

Last edit: Ralph Böhme 2013-03-27

Ralph Böhme - 2013-03-27

The crash happens when dereferencing curdir which was invalidated and set to NULL in cname_mtouname().
cname_mtouname() ends up calling dirlookup() when demangling mangled (name#CNID) names. There the curdir is invalidated and freed because the dircache detected a dev/inode cache difference and evicted the object from the cache. Unfortunately that was our "curdir".

Possible fixes:
- use ostat() in the dircache, this avoids the cache eviction in the first place
- in cname_mtouname() check whether curdir got NULL and reassign

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ralph Böhme - 2013-03-27

Proposed fix in [77d1bd102602aad0e6e6a38ac102a4735b9a58b7].

Related

Commit: <Commit _id='77d1bd102602aad0e6e6a38ac102a4735b9a58b7' tree_id='1f205e4d81ef758638b671b66eb7af375e6cf266' committed=I{'date': datetime.datetime(2013, 3, 26, 10, 43, 1), 'email': 'sloowfranklin@gmail.com', 'name': 'Ralph Boehme'} authored=I{'date': datetime.datetime(2013, 3, 26, 5, 58, 44), 'email': 'sloowfranklin@gmail.com', 'name': 'Ralph Boehme'} message='Use ostat in the dircache\n\nFixes a possible crash in cname() where cname_mtouname calls\ndirlookup() where the curdir is freed because the dircache\ndetected a dev/inode cache difference and evicted the object\nfrom the cache.\n\nFixes bug #498.\n' parent_ids=I['2fe9518a30bb4b5c0d568deb172dd32e998f5864'] child_ids=I[] repo_ids=I[ObjectId('503b705bfd48f843173d5fd3')]>

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Ralph Böhme - 2013-07-18

status: open --> closed

Group: -->
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

netatalk 3.0.2 "INTERNAL ERROR: Signal 11" on opening file in share

Group

Searches

Help

#498 netatalk 3.0.2 "INTERNAL ERROR: Signal 11" on opening file in share

Discussion

Related