[Nix-dev] Re: libgmp impurity

[Peter Simons]

The gmp library in stdenv-updates uses plain GNU config.guess now. This
should remedy overly aggressive choice of optimizer flags. 

My understanding is that there is no trivial solution to the underlying
problem, though. Configure scripts usually guess a lot, and that's
inherently impure because the resulting configuration depends on other
things than just the Nix expression that drove the build.

A simple heuristic to diagnose this particular problem, however, would
be scan the build log files for certain "evil" flags, i.e. --march=xxx
where xxx is something other than the expected architecture. Maybe Hydra
could issue QA warnings whenever it finds such a flag?

Take care,
Peter