dont handle bool transmute

[bend-n]

removes transmute(u8) -> bool suggestion due to ambiguity, leave it for clippy

elaboration on ambiguity in question:
transmute::<u8, bool>(x) will codegen to an assume(u8 < 2);
_ == 1 or _ != 0 or _ % 2 == 0 would remove that assumption
match _ { x @ (0 | 1) => x == 1, _ => std::hint::unreachable_unchecked() } is very verbose

@rustbot label L-unnecessary_transmutes

[rustbot]

r? @Nadrieril

rustbot has assigned @Nadrieril.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[rustbot]

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

[rustbot]

Unknown labels: L-unnecessary-transmute

[rustbot]

Unknown labels: L-unnecessary-transmutes

[oli-obk]

Is there context/discussions that led to this change?

[bend-n]

sort of

[Nadrieril]

I don't know how we decide on this sort of thing.

I personally expect that this suggestion is useful, I can see myself thinking in terms of bytes and forgetting that there's a safe way to do this conversion. At the same time if I'm really doing byte-conversion shenanigans, writing _ == 1 doesn't correctly express my intent.

I'd personally err on the beginner-friendlier side, I've seen colleagues reach for transmute way too happily.

[bend-n]

well this already exists in clippy, as transmute_num_to_bool

[Nadrieril]

ok, that convinces me. very unsure on what basis I should decide whether to merge tho; @oli-obk do you know the etiquette? also wdyt of this?

[asquared31415]

Personally i think that when you're trying to give the compiler more information to optimize with, you should expect to have to write a little more or silence a warning because you're effectively saying "no, i really mean to do this, compiler." I don't know whether this should be demoted to only being in clippy, as the suggested change produces something with identical semantics and no possibility of UB on misuse.

[oli-obk]

I'm really doing byte-conversion shenanigans, writing _ == 1 doesn't correctly express my intent.

The alternative is UB, so I don't see why we'd remove this lint. As asquared said, when you want this opt hint, use expect and state why you consider this sound