replicate fuzz

[sanket1729]

303439353462626261626262626262626261626262626262626230346262
626261626262626262626262626262626262626262626262626262626262
626262623735626262626262626262626262626262626262626262626262
626262626262626262626262623834626262626262426262626262623637
35343030326262356262

Unable to replicate the crash from https://github.com/rust-bitcoin/rust-miniscript/runs/3041068605 locally. Creating a upstream to retest the same vector on CI

[sanket1729]

Okay, I have investigated this for a couple of hours. Giving up on this, any ideas appreciated. Maybe it was some random bug in one of the dependencies?

[sanket1729]

Well, now it's passing again. But this is flaky and should be investigated

[apoelstra]

Closing this, we realized that the issue was caused by the rust-secp "fuzztarget" crypto being broken. See rust-bitcoin/rust-secp256k1#264 where I fixed it and rust-bitcoin/rust-secp256k1#282 where Matt re-broke it to improve fuzzability.

I put fixing this on my list of potential intern jobs, it's not trivial to get internally consistent crypto which can still be broken by a fuzzer and which has similar parser behavior to the real libsecp.