2.6.0 segfaults

There has been a noticeable increase of segmentation faults after teams started using 2.6.0. So far this has only been reproduced on MacOS. A quick look at the core dumps indicates these are all related to boost coroutines. If this is an unknown issue we can share additional info once we get a better understanding of what exactly is going on. The current theory is state info used by boost is being corrupted or unmapped/freed before use.

Do you have the same problems with 2.7? I doubt that 2.6 will receive any fixes after 2.7 got released. Also 2.8 will be released in ~2 weeks. So perhaps even check if the problem exists on master and report an issue at Issues · NixOS/nix · GitHub

I was told that 2.7.0 was experiencing this issue as well, but I have not actually seen it reproduced first hand. If it is found on anything after 2.6.0 I’ll report it in issues, thanks!