pocmatos
2020-10-13 14:26:48

Has anyone had any luck with the kernel fuzzer? No luck for me: https://gitlab.flux.utah.edu/xsmith/xsmith/-/issues/53


samth
2020-10-13 14:30:14

those files both error for me, but the files don’t seem “invalid”


samth
2020-10-13 14:31:47

I don’t know if they’re supposed to run successfully in all cases


pocmatos
2020-10-13 14:48:32

They seem to generate evaluations that trigger contract errors. I would expect the fuzzer to generate always valid evaluations modulo possibly infinite loops or something of the sort. It should track variable types and not make mistakes like it currently does.


pocmatos
2020-10-13 14:48:40

But I am happy to be proven incorrect.



samth
2020-10-13 14:51:08

i think one of those list-type should be array-type


pocmatos
2020-10-13 15:13:25

No idea to be honest - haven’t looked enough at this racr stuff in order to understand this.