Override <RangeInclusive as Iterator>::try_(r)fold

[sinkuu]

cc #45222

• Here preexists a test for folding inclusive ranges.
• try_rfold for Range is left unoverrided since I didn't see performance gain for it.

Benchmarks:

 name orig ns/iter override ns/iter diff ns/iter diff % speedup bench_triples_inclusive_range 41,481 28,547 -12,934 -31.18% x 1.45 bench_triples_range 25,245 21,600 -3,645 -14.44% x 1.17

[rust-highfive]

r? @withoutboats

(rust_highfive has picked a reviewer for you, use r? to override)

[scottmcm]

Shame that this is needed; I thought that the point of #51622 is that it wouldn't be.

Edit: In fact, I'd added this override in #48012, and that PR removed it 😞

cc @kennytm: Do you know anything that might have changed to make LLVM need this again?

[kennytm]

@scottmcm No idea 😐

[Centril]

r? @scottmcm

@bors delegate=scottmcm

[bors]

✌️ @scottmcm can now approve this pull request

[scottmcm]

Cross-reference: #57378

[scottmcm]

r? @rkruppe

Can you take this one too, please, since you have the other RangeInclusive perf one? I think it's better that someone actually on libs handles this, and I'm biased here.

[hanna-kruppe]

Sorry, it doesn't seem like I'll be able to give this PR proper attention in the near future. Please assign someone else.

PS: not sure if that's what you are saying, but I'm not on T-libs or even "affiliated".

[Centril]

Ok... musical chairs... The luck has come to r? @Kimundi