std.meta.Tuple can't handle slice of type array

[iacore]

Zig Version

0.11.0-dev.3268+9d4d96ca9

Steps to Reproduce and Observed Behavior

Run zig test on the following file

const std = @import("std");

const Arr: [3]type = .{ i32, i32, i32 };
const T: type = std.meta.Tuple(Arr[0..2]);

test "simple test" {
    var foo: T = .{ 0, 0 };
    _ = foo;
}

Output:

❯ zig build test
zig test Debug native: error: the following command failed with 1 compilation errors:
/home/user/.local/share/zig/0.11.0-dev.3268+9d4d96ca9/files/zig test /home/user/computing/debug/zig-meta-tuple/src/main.zig --cache-dir /home/user/computing/debug/zig-meta-tuple/zig-cache --global-cache-dir /home/user/.cache/zig --name test --listen=- 
Build Summary: 0/3 steps succeeded; 1 failed (disable with -fno-summary)
test transitive failure
└─ run test transitive failure
   └─ zig test Debug native 1 errors
/home/user/.local/share/zig/0.11.0-dev.3268+9d4d96ca9/files/lib/std/meta.zig:1060:60: error: comptime dereference requires '[2]type' to have a well-defined layout, but it does not.
    return CreateUniqueTuple(types.len, types[0..types.len].*);
                                        ~~~~~~~~~~~~~~~~~~~^~
src/main.zig:4:31: note: called from here
const T: type = std.meta.Tuple(Arr[0..2]);
                ~~~~~~~~~~~~~~^~~~~~~~~~~

Expected Behavior

Test should pass

[iacore]

If you replace Arr[0..2] with Arr[0..3] it works as expected (with a correct error this time).

[iacore]

This seems like a regression. Zig 0.9.1 is able to handle this.

Proof: https://github.com/Rdirtbike/cZPeg/

[nektro]

not sure if this is related or its own issue

const std = @import("std");

const Arr: [3]type = .{ i32, i32, i32 };

test "simple test" {
    @compileLog(Arr[0..3]);
    @compileLog(Arr[0..2]);
    @compileLog(Arr[0..1]);
}

test.zig:6:5: error: found compile log statement
    @compileLog(Arr[0..3]);
    ^~~~~~~~~~~~~~~~~~~~~~

Compile Log Output:
@as(*const [3]type, .{ i32, i32, i32 })
@as(*const [2]type, .{ i32, i32, i32 })
@as(*const [1]type, .{ i32, i32, i32 })

[Daggerfall-is-the-best-TES-game]

on zig-windows-x86_64-0.14.0-dev.154+768b17755 This test passes, so I think this has been fixed

[iacore]

can confirm.

[nektro]

we should double check there's test coverage for this before closing