Hi Paolo and All, One question is about ACCESS_ONCE() in virtscsi_pick_vq(), looks it needn't since both reading and writing tgt->req_vq holds tgt->tgt_lock. Another one is about the comment in virtscsi_req_done(), which said smp_read_barrier_depends() is needed for avoiding out of order between reading req_vq and decreasing tgt->reqs. But if I understand correctly, in virtscsi_req_done(), req_vq is read from vscsi->req_vqs[vq->index - VIRTIO_SCSI_VQ_BASE], instead of tgt->req_vq, and the former won't change wrt. inc/dec tgt->reqs, so can the barrier be removed? Any comments about the above? Thanks, -- Ming Lei -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html