Hashing algorithm and cross-compilation

help
1

The std documentation is clear that the default hashing algorithm is not specified, so I cannot rely on it being the same across releases. But can I rely on it being the same across targets for the same release? The case I have in mind is when generating a static hash table during the build, where the hash function must match across the two platforms in case of cross-compilation.

2

I think the answer is no. The default hasher is randomized. So even with the same binary on the same platform you will get different hashes in different runs.

1 Like
3

Oh, I didn't realize that. Thanks!

4

It’s possible to use a different, non-randomized hash function (Hasher implementation). However, this does not give you portability across platforms, because the behavior of Hash::hash_slice() and of the Hash implementations of built-in types is explicitly not guaranteed.

Portability

Due to differences in endianness and type sizes, data fed by Hash to a Hasher should not be considered portable across platforms. Additionally the data passed by most standard library types should not be considered stable between compiler versions.

— Hash in std::hash - Rust

So, you need to:

  • Convert your key type to bytes explicitly, not using Hash.
  • Feed those bytes to a deterministic hash function.
4 Likes
5

Rust doesn't have any standard randomization function in my understanding. So, I wrote my own, but now I can use just the standard hashing algorithm for the purpose. Thank you for the tip.