Segfault on more than 256 enumerations

[vertexclique]

Happens on both nightly and on stable:
cargo 1.66.0 (d65d197ad 2022-11-15)

............../nightly-x86_64-unknown-linux-gnu/lib/librustc_driver-ce965b63303a743e.so(+0x3259ca3)[0x7f876605bca3]
/lib/x86_64-linux-gnu/libc.so.6(+0x42520)[0x7f8762a6a520]
............../XXXX/target/debug/deps/libnum_enum_derive-9556be49b01582cc.so(+0x260044)[0x7f8751ff7044]

[GilShoshan94]

Are you using #[repr(u8)] on your enum ?

If so, it's normal that it fails since u8 give you 256 discriminant.
Try #[repr(u16)] instead.

About the segfault instead of a compilation error, I have no idea.

[illicitonion]

@vertexclique Thanks for the bug report - please could you include a sample code snippet that reproduces the problem? I've tried to reproduce a few different ways, and haven't been able to...

[vertexclique]

Smaller repr on the larger variants is always compiler error. It's not segfault. I unfortunately can't share the repro code. Only thing is there variants are 0 variadics so literally something like this:

enum Data {
A,
B,
C,
... 900 more
}

I wrote 256 because it staggers during compilation. I suspect something else might be going on that I don't know.

[illicitonion]

Can you share exactly what attributes you have on the enum? What's its repr? What trait(s) are you deriving?

[vertexclique]

#[derive(Debug, Clone, Copy, PartialOrd, Display, PartialEq, Eq, TryFromPrimitive)]
#[repr(u16)]
pub enum Data {
  A,
  B,
  C,
  ... 900 more
}

[illicitonion]

As far as I can tell, this is actually blowing up the stack of rustc (at least that's the only way I could repro this), and #89 should fix it by simplifying the generated code.