Bugzilla – Bug 49
tracejob not working : memory error
Last modified: 2010-07-01 14:51:19 MDT
You need to log in before you can comment on or make changes to this bug.
Hello, I have installed torque 2.5.0 on a small cluster - head ubuntu (kernel 2.6.31-17-generic) - nodes debian (kernel 2.6.32-trunk-686) and I have successfully submitted some very simple test jobs. However, when I run tracejob, I get the following error: $ tracejob 18 *** glibc detected *** tracejob: malloc(): memory corruption: 0x093ed0c8 *** ======= Backtrace: ========= /lib/tls/i686/cmov/libc.so.6[0xe74ff1] /lib/tls/i686/cmov/libc.so.6[0xe77be3] /lib/tls/i686/cmov/libc.so.6(__libc_malloc+0x58)[0xe79898] /lib/tls/i686/cmov/libc.so.6(popen+0x21)[0xe67651] tracejob[0x804906f] tracejob[0x8049ad6] /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe6)[0xe20b56] tracejob[0x8048be1] ======= Memory map: ======== 0014e000-00174000 r-xp 00000000 08:06 596937 /usr/local/lib/libtorque.so.2.0.0 00174000-00175000 r--p 00025000 08:06 596937 /usr/local/lib/libtorque.so.2.0.0 00175000-00176000 rw-p 00026000 08:06 596937 /usr/local/lib/libtorque.so.2.0.0 00176000-00213000 rw-p 00000000 00:00 0 002e7000-00303000 r-xp 00000000 08:06 331747 /lib/libgcc_s.so.1 00303000-00304000 r--p 0001b000 08:06 331747 /lib/libgcc_s.so.1 00304000-00305000 rw-p 0001c000 08:06 331747 /lib/libgcc_s.so.1 00699000-006b4000 r-xp 00000000 08:06 328207 /lib/ld-2.10.1.so 006b4000-006b5000 r--p 0001a000 08:06 328207 /lib/ld-2.10.1.so 006b5000-006b6000 rw-p 0001b000 08:06 328207 /lib/ld-2.10.1.so 00bbb000-00bbc000 r-xp 00000000 00:00 0 [vdso] 00e0a000-00f48000 r-xp 00000000 08:06 344674 /lib/tls/i686/cmov/libc-2.10.1.so 00f48000-00f49000 ---p 0013e000 08:06 344674 /lib/tls/i686/cmov/libc-2.10.1.so 00f49000-00f4b000 r--p 0013e000 08:06 344674 /lib/tls/i686/cmov/libc-2.10.1.so 00f4b000-00f4c000 rw-p 00140000 08:06 344674 /lib/tls/i686/cmov/libc-2.10.1.so 00f4c000-00f4f000 rw-p 00000000 00:00 0 08048000-0804b000 r-xp 00000000 08:06 564595 /usr/local/bin/tracejob 0804b000-0804c000 r--p 00002000 08:06 564595 /usr/local/bin/tracejob 0804c000-0804d000 rw-p 00003000 08:06 564595 /usr/local/bin/tracejob 093ed000-0940e000 rw-p 00000000 00:00 0 [heap] b7600000-b7621000 rw-p 00000000 00:00 0 b7621000-b7700000 ---p 00000000 00:00 0 b7748000-b774a000 rw-p 00000000 00:00 0 b7761000-b7763000 rw-p 00000000 00:00 0 bf861000-bf876000 rw-p 00000000 00:00 0 [stack] Is it me or tracejob ?
(In reply to comment #0) I have been unable to get the 2.5.0 version of tracejob to fail on Ubuntu 2.6.28-18-generic and Ubuntu 2.6.31-20-generic. Are you seeing the failure on the Ubuntu or Debian machine?
Vincent, does compiling this with -g give any extra debugging info in the stacktrace ? Have you tried running the command under valgrind or gdb (again compiled with -g) ?
Hi Chris and Al, sorry for answering late, I wanted to check my reported bug beforehand but I no-longer have torque 2.5.0 ready. I'll try to rebuild it by the end of the week to let you know.
Michael, please look at the current trunk before commenting further.