What's a good pattern for fixed bits?

[shepmaster]

I'm working with a chip that needs to read a byte of the form 0b0100AAAW, where the upper bits must always be 0100. What's the best route for me to ensure that those bits will always be set to that?

Right now, I'm setting them via the default as the docs say

Default values are used as-is, even if they affect bits that aren't defined within the bitfield.

However, someone could call new_with_raw_value(0) and set those upper bits to an invalid value.

I was originally looking for something vaguely of this shape:

#[bits(4..=7, fixed = 0b0100)]
_: (),

This would let me specify that a range of bits should always be a given value and wouldn't generate setters or getters.

[poseidon-rises]

Did you ever figure this out? I have a requirement and I'm looking for a library that supports this.

[shepmaster]

I did not find anything better than setting the default as mentioned in my original post.

[danlehmann]

Hi, you’re correct that bitbybit doesn’t provide a guarantee that certain bits are a certain value. Given that users can always translate to and from integers that didn’t seem like a very important requirement

Something like that could probably be built but it would be a very different type of library I’d imagine.