@memcpy: Don't check for aliasing if the type being copied has a bitSize of in 0

[JonathanHallstrom]

Should close #21655
Unable to verify myself due to issues with building zig locally

[Rexicon226]

I'm unsure if this is an issue that needs fixing. @memcpying requires memory to be involved, and ZSTs do not have any memory representation neither comptime nor runtime. Just my 2c.

[rohlem]

I'm unsure if this is an issue that needs fixing.

The issue is #21655 , where correct code for a generic data structure in status-quo would need special-casing.
Discussion might make more sense there; I think it makes perfect sense to instead special-case the compiler builtin though.

[Rexicon226]

I'm unsure if this is an issue that needs fixing.

The issue is #21655 , where correct code for a generic data structure in status-quo would need special casing. Discussion might make more sense there; I think it makes perfect sense to instead special-case the compiler builtin though.

Gotcha thanks, didn't see that.

[JonathanHallstrom]

Seem CI is still failing, maybe not emitting anything in the case of zero bit types will fix it?

[JonathanHallstrom]

This way of implementing it feels a little bit sketchy, maybe it should just be an early return?

[JonathanHallstrom]

i seem to have messed something up in git

[mlugg]

This isn't quite safe due to some edge cases with comptime-only types. I'm going to wait until myself and @alexrp finish our discussion with Andrew on @memcpy semantics before proposing a path forward here -- hold tight.

[alexrp]

@mlugg this was superseded by #22631, right?

[mlugg]

Oops, I forgot to implement this into #22631. I'll look into this shortly.

[mlugg]

Superseded by #22708.