I've been using DFHack with the LFR mod successfully on my home machine for some time now. As some of you may remember, I built this using an incredibly baroque procedure involving a 32-bit chrooted old version of Debian and compiled gcc 4.5 from source code.
After doing all that, someone told me "Oh, you can just remove the libstdc++.so.6 file that comes with DF."
Well, today I decided to try that approach on a different machine (at work). This is another Debian 7 64-bit system. This time, instead of doing the chrooted build as I did at home, I tried moving the DF-supplied libstdc++.so.6 library out of the way, and then building everything with the gcc 4.7 toolchain provided in Debian 7.
I am using the DFHack source code from
https://github.com/DFHack/dfhack with the one-line patch that is required to make it compile (adding #include <stdexception> in plugins/eventful.cpp).
After a few minor false starts, I was able to get this to build, and start up. I generated a world, and embarked, and was able to play about a month before the game locked up. Totally CPU-bound. I had to kill it. I tried reloading the save, but I had weird and inconsistent results. Sometimes it would segfault. Sometimes it would lock up before even displaying the embark. Once it got far enough to display the embark, and then segfaulted.
I started a second worldgen, thinking that maybe it was some weird glitch with just that one world. The second worldgen completed, and I embarked... and got the same lock-up. I played around with various things, including "./dfhack -g" but didn't really get anywhere.
So, then I decided to copy the entire hack/ subdirectory from my home computer (containing the DFHack stuff that I built with gcc 4.5) to the work computer. I moved the gcc-4.7-compiled hack/ subdirectory aside, and dropped the gcc-4.5-compiled hack/ subdirectory in its place. Then I moved the DF-supplied libstdc++.so.6 library back into place. I did not change anything else at all. Then I reran ./dfhack and loaded the second fortress (autosaved upon embark). This time I was able to play all the way through the spring, with no lock-ups or segfaults.