[GauravPawar101]

Closes #798

[GauravPawar101]

Miri is expected to fail:

• rkyv achieves zero-copy deserialization by referencing an ArchivedVec<T> struct directly within a raw byte buffer. Stacked Borrows Restriction: Miri retags the archived reference (&ArchivedVec<T>). Under Stacked Borrows rules, that reference's valid provenance range is restricted strictly to the byte range of the ArchivedVec struct header itself (offsets [0xc..0x10]).
• Out-of-Bounds Pointer Derivation: Inside ArchivedVec::as_slice(), rkyv calculates a relative pointer to access the vector items stored at offsets [0x0..0xc]. Because [0x0..0xc] falls outside the ArchivedVec struct reference boundary ([0xc..0x10]), Stacked Borrows treats deriving pointers to [0x0..0xc] from &self as illegal pointer retagging.

Been hitting this for almost an hour

[alejandro-vaz]

then tell miri to stfu like we do in tests/rayon.rs

[alejandro-vaz]

peak commit names

[GauravPawar101]

To not embaras ourselves we should use something like bors, to squash these like rust-lang.
Btw in my defense my keyboard has no g
How would that be, we are not big enough but would be fun tho

[alejandro-vaz]

hahahahhahha fair, no G

squash merging is the only type of merge allowed so don't worry

I'm not sure how you'd set up bors

I think we shouldn't. it'd add complexity and the ROI is basically zero

btw review #815