Hi, Sorry if this gets sent twice...
I grabbed the 1.5.0 source and built libunwind myself on a linux SUSE system. On this linux machine there already exists libunwind in /usr/lib64 but I want to build it myself. To build I’m running autogen.sh, then configure and finally make. When I run my test program which gets callstacks once a millisecond, I see that the libunwind I built takes 4 times longer than the one installed on the system. I’ve also tried adding -O2 -m64 to CFLAGS and CXXFLAGS but it doesn’t make a difference. Is there anything obvious in the above steps that would indicate why there is this huge difference in performance with the version I am building? For example, do I have to pass any options to autogeh.sh, configure, or make to perform an optimized production build? Thanks in advance, Matt
