Compiler crashes and gives no output

[Didas-git]

Zig Version

0.14.0-dev.2909+f3c29dcb2

Steps to Reproduce and Observed Behavior

Create a new zig project using zig init and replace the main.zig file with the following:

const G = struct {
    const L = struct { [:0]const u8 };
    comptime list: []*L = &.{},

    pub fn get(comptime self: *G, pattern: [:0]const u8) void {
        comptime {
            var temp: L = .{pattern};
            self.list = self.list ++ @constCast(.{&temp});
        }
    }
};

pub fn main() !void {
    comptime {
        var t = G{};
        t.get("s");
    }
}

What is observed is that the compiler simply quits with code 1 and doesn't give any information on what the issue is.

We have also ran the code with a debug build and it does give us an error, im appending the output bellow.
Crash.log

Expected Behavior

I would expect to get an error message for why it fails to compile, the code is most likely not valid zig but as a new person in the zig ecosystem it's fairly hard to find out what's wrong when there is no output.

[Didas-git]

After exploring for a while, we were able to narrow down the issue to the following:

const T = struct { []const u8 };
pub fn main() !void {
    comptime {
        var t: T = .{"s"};
        @constCast(.{&t});
    }
}

We reached the conclusion that the issue seems to come from the lack of sema being passed to latter functions, as the inner value of the tuple is a pointer when the printPtr gets called it passes down the comptime_alloc pointer without a sema as per:

~/zig/src/Value.zig:4143:34: 0x11034951b in pointerDerivationAdvanced__anon_421417 (zig)
            const sema = opt_sema.?;
                                 ^
~/zig/src/print_value.zig:318:46: 0x1112904eb in printPtr__anon_1024821 (zig)
        try ptr_val.pointerDerivationAdvanced(arena.allocator(), pt, false, null);
                                             ^
~/zig/src/print_value.zig:167:25: 0x11102629b in print__anon_662789 (zig)
            try printPtr(val, .rvalue, writer, level, pt, opt_sema);
                        ^

[xdBronch]

doesnt reproduce anymore, fixed by #25458 i assume