[mir] or-patterns duplicate the same branch

[tesuji]

Consider this snippet:

#![feature(const_fn_transmute)]
#![feature(or_patterns)]

use std::mem::transmute;

pub const fn foo3(bytes: [u8; 16]) -> Option<[u8; 4]> {
    // big endian `u32`s
    let dwords: [u32; 4] = unsafe { transmute(bytes) };
    const FF: u32 = 0x0000_ffff_u32.to_be();
    if let [0, 0, 0 | FF, ip] = dwords {
        Some(unsafe { transmute(ip) })
    } else {
        None
    }
}

The or-pattern 0 | FF has this MIR:

    bb3: {
        switchInt(_2[2 of 4]) -> [0_u32: bb6, 4294901760_u32: bb7, otherwise: bb4]; // scope 1 at src/lib.rs:10:19: 10:20
    }

While bb6 and bb7 points to the same branch:

    bb6: {
        StorageLive(_4);                 // scope 1 at src/lib.rs:10:27: 10:29
        _4 = _2[3 of 4];                 // scope 1 at src/lib.rs:10:27: 10:29
        goto -> bb5;                     // scope 1 at src/lib.rs:10:5: 14:6
    }

    bb7: {
        StorageLive(_4);                 // scope 1 at src/lib.rs:10:27: 10:29
        _4 = _2[3 of 4];                 // scope 1 at src/lib.rs:10:27: 10:29
        goto -> bb5;                     // scope 1 at src/lib.rs:10:5: 14:6
    }

I would expect the or-pattern points the same branch.

[tesuji]

@rustbot modify labels: A-mir F-or_patterns A-patterns

[JulianKnodt]

Was this resolved by #75537? I believe that this would actually match on it, just because all the statements are the same

[tesuji]

I've just checked. The MIR isn't improved with #75537 .

[JulianKnodt]

Ah woops yea I can't count, this has 3 branches not 2, which is the only permitted number of branches being optimized currently.