http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58437
--- Comment #10 from Jeffrey M. Birnbaum <jmbnyc at gmail dot com> --- Tammy, Something must have been in the air because your timestamp on the bug submission means that right at the time you were reporting the bug I was actively looking into why I was seeing horrible sort performance. I am working on an algo that requires a post sort of something 1B entries so I wanted to see worst case for std::sort on 500M and was stunned to see how poor it was. My CPU was Haswell so wondered if there was an issue so I tried another bug (that happened to have gcc 4.4.7) and realized it appeared to be a compiler or std lib regression. Anyhow, given the length of time that this has been out there it is weird that we reported the same bug within hours of each other. Best, /JMB