libwebp VS kani

The thing that concerns me most is looking at the fix it is very difficult to see why this fix is correct. It also appears as there is lots of code without explicit bounds checks. It makes me worried because while the logic may be safe this makes the logic very complex. I wonder what the cost would be to add an explicit, local bounds check at every array access. This would serve as a backup that is much easier to verify. I suspect the cost would be relatively small. Small enough that I personally would be happy to pay it.

https://github.com/webmproject/libwebp/commit/902bc919033134...

This is also a great reminded that fuzzing isn't a solution to memory unsafe languages and libraries. If anything the massive amount of bugs found via fuzzing should scare us as it is likely only scratching the surface of the vulnerabilities that still lie in the code, a couple too many branches away from being likely to be found by fuzzing.

There's a follow-up fix, according to Debian[0]: https://github.com/webmproject/libwebp/commit/95ea5226c87044...

[0]: https://security-tracker.debian.org/tracker/CVE-2023-4863

The breakage [0] was introduced by the creator [1] of the project. If you want to audit 1674 commits over the past 12 years, it'd be easier to just audit the full project.

[0] https://github.com/webmproject/libwebp/commit/21735e06f7c1cb...

[1] https://github.com/webmproject/libwebp/commit/c3f41cb47e5f32...

If you like the command line, then you can use ffmpeg and ImageMagick, or use libwebp directly

The webp parser code is open source. Which means that even if Google decides to hide/obscure the code for webp, they'd legally not be allowed to prevent you from using older versions of the webp parser library. The only thing they could do is patent it, and then companies in the US (which has software patents, unfortunately) would have to pay royalties to decode it anyway; but here comes the next point

This is also the backend for Kani - Amazon's formal verification tool for Rust.

https://github.com/model-checking/kani

Nice read! Rust has pushed, and will continue to push, the limits of practical, bare metal, memory safe languages. And it's interesting to think about what's next, maybe eventually there will be some form of practical theorem proving "for the masses". Lean 4 looks great and has potential, but it's still mostly a language for mathematicians. There has been some research on AI constructed proofs, which could be the best of both worlds because then the type checker can verify that the AI generated code/proof is indeed correct. Tools like Kani are also a step forward in program correctness.

Ease setup in Amazon Linux 2 by @adpaco-aws in #2833

Limit --exclude to workspace packages by @tautschnig in #2808

Here's a summary of what's new in version 0.38.0:

> those applications need the proof for correctness so that more dangerous code---say, what would need `unsafe` in Rust---can be safely added

There are actually already tools built for this very purpose in Rust (see Kani [1] for instance).

Formal verification has a serious scaling problem, so forming programs in such a way that there are a few performance-critical areas that use unsafe routines seems like the best route. I feel like Rust leans into this paradigm with `unsafe` blocks.

[1] - https://github.com/model-checking/kani

Enable concrete playback for failure of UB checks by @zhassan-aws in https://github.com/model-checking/kani/pull/2727

Change default solver to CaDiCaL by @celinval in https://github.com/model-checking/kani/pull/2557 By default, Kani will now run CBMC with CaDiCaL, since this solver has outperformed Minisat in most of our benchmarks. User's should still be able to select Minisat (or a different solver) either by using #[solver] harness attribute, or by passing --solver= command line option.

Add support for sysconf by feliperodri in #2557