Pierre-Marie Pédrot

Results 661 comments of Pierre-Marie Pédrot

Note that coq/coq#19126 has probably made this PR much less usefull in practice.

It actually helped on the compcert example.

I think that the bench basically doesn't measure anything coqchk-related. The current run is just noise. But on the one-liner from compcert this PR reduces the coqchk invocation from ~87s...

CI failure unrelated, let @coqbot merge now

The CoqHammer backwards compatible overlay has been merged, this is ready to go.