pub fn mla_q_rope_writeback(
gpu: &dyn GpuBackend,
kernel: KernelHandle,
q_rope_direct: DevicePtr,
q_absorbed_buf: DevicePtr,
nq: u32,
rope: u32,
kv_lora: u32,
mla_cache_dim: u32,
stream: u64,
) -> Result<()>Expand description
2026-09-25: Write the rotated [nq, rope] Q rope parts back into q_absorbed_buf at offset
kv_lora of each head’s mla_cache_dim row.