|
From: Rich <rd...@lu...> - 2003-08-04 15:50:26
|
When I run my program directly it does not segfault. It immediately segfaults with valgrind-20030725 with libc 2.3.2. The same tests for libc 2.0 with valgrind 1.9.6 do not segfault. Any ideas on what's changed in glibc or valgrind? I've tried this with and without the option "--run-libc-freeres=no." valgrind --run-libc-freeres=no --leak-check=yes ./durendal ==24134== Memcheck, a.k.a. Valgrind, a memory error detector for x86-linux. ==24134== Copyright (C) 2002-2003, and GNU GPL'd, by Julian Seward. ==24134== Using valgrind-20030725, a program supervision framework for x86-linux. ==24134== Copyright (C) 2000-2003, and GNU GPL'd, by Julian Seward. ==24134== Estimated CPU clock rate is 2397 MHz ==24134== For more details, rerun with: -v ==24134== ==24134== Warning: segment-override prefix encountered, but thread has no LDT ==24134== Warning: segment access: virtual addr 8 exceeds segment limit of 0 ==24134== Invalid read of size 4 ==24134== at 0x4023912C: __pthread_alt_lock (in /lib/i686/libpthread-0.10.so) ==24134== by 0x40235D76: __GI___pthread_mutex_lock (in /lib/i686/libpthread-0.10.so) ==24134== by 0x4202AE54: __cxa_atexit_internal (in /lib/i686/libc-2.3.2.so) ==24134== by 0x4201589F: __libc_start_main (in /lib/i686/libc-2.3.2.so) ==24134== Address 0x8 is not stack'd, malloc'd or free'd Any ideas would be appreciated. Thanks, Rich |