Linux Forums - Linux Help,Advice & support community:LinuxSolved.com
Linux in General => Linux Development & Programming => Topic started by: vababu on March 12, 2007, 09:47:25 AM
-
Hi,
We have an application on RHEL4 U4 which uses JNI calls.
During endurance, it fails after about 10 laks of message syaing the following
*** glibc detected *** double free or corruption (!prev): 0x094019e0 ***
bt of the core shows the following
(gdb) bt
#0 0x0021d7a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
#1 0x002857a5 in raise () from /lib/tls/libc.so.6
#2 0x00287209 in abort () from /lib/tls/libc.so.6
#3 0x002b971a in __libc_message () from /lib/tls/libc.so.6
#4 0x002bffbf in _int_free () from /lib/tls/libc.so.6
#5 0x002c033a in free () from /lib/tls/libc.so.6
#6 0xb1f56283 in Java_com_hp_opencall_map_mapJNI_delete_1map_1message_1t (jenv=0x809c398,
jcls=0xb256caec, jarg1=2562072288) at /root/map-build/map_wrap.c:1662
#7 0xb291aa88 in ?? ()
#8 0x0809c398 in ?? ()
#9 0xb256caec in ?? ()
#10 0x98b61ee0 in ?? ()
#11 0x00000000 in ?? ()
We tried to troubleshoot this issue by ensuring the build options, linker options etc. We have revisited the code for malloc/free but could not solve it
Kernel : 2.6.9-42.ELsmp
glibc : glibc-2.3.4-2.25
gcc : 3.4.6
We have been trying with valgrind but surprisingly the following options could not generating any log:
algrind --tool=memcheck --log-file=valgrind.output --num-callers=20 --leak-check=yes --trace-children=yes java xxxxxxxxxxxxxxxxx
Any known issues ? with these versions of glibc/gcc.When browsed many ppl have posted the same issue but no proper resolution.Any help would be appreciated
Thanks inadvance
babu
-
valgrind is great, and even better if you start a project using it, and dilligently fix errors it reports.
It isn't clear from your post, but since you mention JNI, I'm going to assume you are calling some C or C++ code from Java? If you provide a succinct example of the class(es) or functions which break you may get more help.
-chuck