this is a really cool read, well put together and very thorough, but near the end the author betrays a startling amount of naïveté: > Considering how simple th

this is a really cool read, well put together and very thorough, but near the end the author betrays a startling amount of naïveté: > Considering how simple the PoC is, I hope it's clear that in order to be able to exploit a vulnerability, the planets really need to align. The vuln needs to exist, you need to somehow figure out how to craft the payload (hard if you're not on the system, and have access to source code, compile listings, and a debugging session), and get the payload on the system. Then, when you have arbitrary code execution, you need to know how the system is set up to do interesting stuff. Open sockets, create files, edit files, delete files, what have you. This is not a good conclusion to come to. The overall, high level conclusion is that the hardware that Z/OS runs on still overlays code and data and has some rudimentary memory protections, but is not memory safe by design and can still be exploited via direct code injection or return to libc / ROP. Relying on the secrecy of the system (its configuration, code, and hardware) is not a strong guarantee as motivated attackers can defeat this. Or to put another way: for reliable x86 exploitation, the planets also need to align, but attackers make them align regularly and have for 20 years...