How is this different from tools like Valgrind and Address Sanitizer?

Posted Oct 28, 2025 21:26 UTC (Tue) by cyperpunks (subscriber, #39406)
In reply to: How is this different from tools like Valgrind and Address Sanitizer? by excors
Parent article: Fil-C: A memory-safe C implementation

So it's kind of small virtual machine with garbage collection that happens to be compatible with C/C++ based source code?

to post comments

How is this different from tools like Valgrind and Address Sanitizer?

Posted Oct 29, 2025 4:57 UTC (Wed) by Cyberax (✭ supporter ✭, #52523) [Link] (2 responses)

Yes, it can be thought about like this. Fil-C limits the primitives that are accessible to the C code so that no combination of them can (in theory) lead to memory safety issues.

It's somewhat analogous to compiling C into WebAssembly and then JIT-compiling WebAssembly.

The amazing thing is that it preserves most of C/C++ semantics.

How is this different from tools like Valgrind and Address Sanitizer?

Posted Oct 29, 2025 5:53 UTC (Wed) by willmo (subscriber, #82093) [Link] (1 responses)

> The amazing thing is that it preserves most of C/C++ semantics.

But (at least WRT memory safety) only the semantics of the abstract machine described by the language standards, and not the additional semantics (aka undefined behavior) of the straightforward mappings to typical hardware that we’re all accustomed to.

Very cool idea. :-)

How is this different from tools like Valgrind and Address Sanitizer?

Posted Oct 30, 2025 15:00 UTC (Thu) by atanasi (guest, #136067) [Link]

Fil-C strives to support undefined behavior:

'In addition to memory safety, Fil-C's other goal is fanatical compatibility. This means that Fil-C's capability model also has to allow most (or ideally all) "safe" uses of C pointers; that is, idioms that are in widespread use and don't lead to exploitation unless they violate the above rules. This means even supporting uses of pointers that the C spec deems to be undefined behavior.'