Function rowmajor_index 

pub const fn rowmajor_index(l: usize, token: usize, k_group: usize) -> usize

Offset of the same (token, k_group) scale in the layout per_token_group_quant_fp8 writes: row-major [M, K/128], K-group contiguous. The in-tree fp8_gemm_t_blockscaled indexes this one.