tiny.simd.loadDup128

1pub fn loadDup128(comptime D: type, input: []const D.Lane) D.Vector {2    const block_lanes = @min(D.lane_count, 16 / @sizeOf(D.Lane));3    std.debug.assert(input.len >= block_lanes);4    var result: D.Vector = undefined;5    inline for (0..D.lane_count) |index| {6        result[index] = input[index % block_lanes];7    }8    return result;9}

1pub const loadDup128 = memory.loadDup128;