dirscan segfault

[Anon Anon]

Hi all,

sorry if this question is in wrong forum, maybe some mod can move the thread to correct forum then?

 

I use dirscan v. 3.11 and have two active profiles to check if two different directories exist but dirscan segfaults over and over again (since nimbus trying to restart the probe):

[1727122.381022] dirscan[12919]: segfault at 0 ip 00007f7fa9b8a86f sp 00007f7fa9a6be38 error 4 in libc-2.13.so[7f7fa9a71000+180000]

[1727242.916167] dirscan[12932]: segfault at 0 ip 00007f7f5849686f sp 00007f7f58377e38 error 4 in libc-2.13.so[7f7f5837d000+180000]

[1727469.597456] dirscan[12971]: segfault at 0 ip 00007fce5906186f sp 00007fce58f42e38 error 4 in libc-2.13.so[7fce58f48000+180000]

GDB didn't reveal so much about the problem:

<GDB generic header>

Reading symbols from /opt/nimsoft/probes/system/dirscan/dirscan...(no debugging symbols found)...done.

[New LWP 12911]

[New LWP 12881]

 

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 `./dirscan'.

Program terminated with signal 11, Segmentation fault.

#0  0x00007eff9927f86f in ?? () from /lib/x86_64-linux-gnu/libc.so.6

(gdb) bt

#0  0x00007eff9927f86f in ?? () from /lib/x86_64-linux-gnu/libc.so.6

#1  0x000000000042c49f in ?? ()

#2  0x000000000041fc2d in ?? ()

#3  0x0000000000411cbf in ?? ()

#4  0x0000000000415b04 in ?? ()

#5  0x000000000041d43c in ?? ()

#6  0x00007eff996fab50 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0

#7  0x00007eff99240a7d in clone () from /lib/x86_64-linux-gnu/libc.so.6

#8  0x0000000000000000 in ?? ()

 

When i run the dirscan binary with strace, I'll see the following:

mmap(NULL, 8392704, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK, -1, 0) = 0x7f06297f8000

mprotect(0x7f06297f8000, 4096, PROT_NONE) = 0

clone(child_stack=0x7f0629ff7ff0, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0x7f0629ff89d0, tls=0x7f0629ff8700, child_tidptr=0x7f0629ff89d0) = 14152

 

 

And strace for pid 14152 (before segfault):

write(3, "Mar 31 12:07:09:235 [13966445114"..., 75) = 75

write(3, "Mar 31 12:07:09:235 [13966445114"..., 75) = 75

write(3, "Mar 31 12:07:09:235 [13966445114"..., 75) = 75

time([1427796429])                      = 1427796429

stat("/etc/localtime", {st_mode=S_IFREG|0644, st_size=1892, ...}) = 0

write(3, "Mar 31 12:07:09:235 [13966445114"..., 123) = 123

open("/opt/wordpress_install/uploads/cache", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 5

getdents(5, /* 3 entries */, 32768)     = 80

--- SIGSEGV (Segmentation fault) @ 0 (0) ---

 

Any clues what can cause this bug?

BR

//Robert

 

 

[becph01]

I ran into this before; mine was a sigsev 11 I believe.

 

I would start a fresh configuration (rename your .cfg and bring it back up) and configure this via Admin Console.

 

If you try running this through Valgrind and attach to the processes - I think you'll see a memory violation. I believe the path you're using is around 64 characters? From what I recall - it was running out of bounds somewhere and will eventually hit some memory it shouldn't.

 

From what I could gather when I tried to dissemble the code - there are two different profiles one for pre-Admin Console(C/C++) and one for post-Admin console (Java). The code path in Java/connected to Admin Console does not seem to suffer from this.

 

Regards,

Philip B