Re: Observations from a bytecode validation feasibility experiment
Luiz Henrique de Figueiredo <[email protected]>
| Newsgroups | gmane.comp.lang.lua.general |
|---|---|
| Message-ID | <CABt16qngkwZAwk1R3Br3FDUa22nhZ=kAw4Un_apKw62A7qmKPA@mail.gmail.com> |
> My main conclusion is that it's > gotten a lot harder, and a complete verifier would likely be as big as > (if not larger than) the luaV_execute core loop. A complete verifier is a hopeless task (halting problem etc.). We dropped even the simple integrity test in Lua 4.0, way back in 2000. > Many 5.1/5.2 validators relied on luaP_opmodes to handle most opcodes > generically, with a switch/case only covering the more complex cases. We did that for bytecode listing in luac.c, but it was not clear. Since Lua 5.4 luac.c uses a complete switch; it's much easier to maintain even if there is much repetition. > It's clear that bytecode validation isn't a priority for the community, It's an impossible task. Our take is that if you don't trust bytecode, don't allow Lua to load it. --lhf -- You received this message because you are subscribed to the Google Groups "lua-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/lua-l/CABt16qngkwZAwk1R3Br3FDUa22nhZ%3DkAw4Un_apKw62A7qmKPA%40mail.gmail.com.