hi all, i'm using thread_specific_ptr (boost 1.33.1, sorry) in a plugin architecture, where the tls identifies a singleton object of type Context (initialized using DCLP). boost::thread_specific_ptr<Context> Context::instance_; i also have a set of shared libraries and a main executable linked against a static library defining the Context class. this all works fine so long as the libraries are loaded and the threads created in that order, however, if a library is loaded after thread is started, then the thread_specific_ptr in that thread appears to be reset to null, without calling cleanup(). if i use the debugger to inspect the singleton ptr before and after loading the second module, it is, before: $1 = {boost::noncopyable_::noncopyable = {<No data fields>}, m_tss = {boost::noncopyable_::noncopyable = {<No data fields>}, m_slot = 8}} ..and after: $2 = {boost::noncopyable_::noncopyable = {<No data fields>}, m_tss = {boost::noncopyable_::noncopyable = {<No data fields>}, m_slot = 1}} is this correct behaviour for thread_specific_ptr? is there a way i can obtain a handle for the first Context object before it is destroyed? here's some info about my system. $ cat /proc/cpuinfo ... AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ ... $ uname -a Linux lumpen 2.6.19 #2 SMP Fri Jan 12 17:58:33 NZDT 2007 x86_64 x86_64 x86_64 GNU/Linux $ gcc -v Using built-in specs. Target: x86_64-redhat-linux Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --enable-shared --enable-threads=posix --enable-checking=release --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-libgcj-multifile --enable-languages=c,c++,objc,obj-c++,java,fortran,ada --enable-java-awt=gtk --disable-dssi --with-java-home=/usr/lib/jvm/java-1.4.2-gcj-1.4.2.0/jre --with-cpu=generic --host=x86_64-redhat-linux Thread model: posix gcc version 4.1.1 20060525 (Red Hat 4.1.1-1) any help would be appreciated. cheers, Jono