i thought this discussion was on 9fans, but i don't see it any more. this is a recent bug report in 9front
http://code.google.com/p/plan9front/source/detail?r=f80b7ef22cd2352d3823513024d21d3ea14f4854 6a, 6c, 6l: fix copy propagation Without an explicit signal for a truncation, copy propagation will sometimes propagate a 32-bit truncation and end up overwriting uses of the original 64-bit value. This was independently discovered and fixed in Go. See: http://golang.org/issue/1315 https://codereview.appspot.com/6002043/ Thanks Charles Forsyth for tips and advice. i was in a panic for a bit that this was still hanging around, but after a bit of investigation, i see this was fixed by charles and pushed out to 9atom quite a while ago. i have a dim recollection of finding the bug. the symptoms were spectacular. i believe they bit the log2 calculations, and blew up malloc. there are more changes from charles in 9atom, including a pretty good complement of avx instructions. it might be worth a look. - erik
