Dmitry Vyukov

> There is a local declaration of _go_fuzz_previous_location at the top of each function Ah, interesting. I missed that. I guess this can work then. The uses of the global...

> One other aspect that makes benchmarking this particular change more challenging is that it is harder to use coverage as a metric to compare before and after the change,...

> Three other related comments: I don't know what to say. Maybe. FWIW in libfuzzer we have lots of tunables for such things and then one can always experiment and...

-race is useful with GOMAXPROCS=1 as it uses abstract relations to detect races. So as the first step we could make -race supported in go-fuzz-build. Is it just passing the...

We could detect -race build and avoid setting GOMAXPROCS=1. This sounds reasonable because if one uses -race, they are interested in races, and if we don't detect races with GOMAXPROCS=1...

Wonder if it's go/ast bug and affects go tool cover too.

> Looks like we're more or less on the same page here, except that I propose to keep the signature simple and uniform, and use methods to get fuzz data....

But overall I agree with idea of adding a context argument as it's much more flexible and extensible. I always feel a bit nervous designing public stable APIs... Skip/Fail/Failf look...

For context, Ian's proposal of using F.Useful/Discard: https://github.com/golang/go/issues/19109#issuecomment-280119030 Skip looks better than Discard, because Skip is testing.T method. For Interesting/Useful we could also consider Priority name, esp if it accepts...

https://github.com/golang/go/issues/19109#issuecomment-280147350 also mentions: ``` Could there be some default convention say a _fuzz/xxx directory (where xxx corresponds with FuzzXxx) and a method on the *testing.F object to load a different...