With the development of the upcoming sixth-generation networks (6G), reconfigurable intelligent surfaces (RISs) have gained significant attention due to its ability of reconfiguring wireless channels via smart reflections. However, it remains critical challenges for channel state information (CSI) acquisitions of double-RIS systems with ultra-heavy pilot overhead and multi-path interference. This paper proposes a novel channel generation method in double-RIS communication systems based on the utilization of conditional diffusion model (CDM). The CDM model is carefully designed and trained on synthetic channel data to capture the high-dimensional channel characteristics. It addresses the limitations of conventional deep learning-based CSI estimation approaches, such as insufficient model understanding capability and poor environmental adaptability. Then we provide a detailed analysis of the diffusion process for channel generation, elaborating on how CDM is able to acquire CSI efficiently. The simulation results demonstrate that the proposed CDM based channel acquisition method outperforms benchmarks in terms of normalized mean squared error, achieving an improvement of approximately 18%.