pub fn execute_triangular_solve_kernel<T>(
_device: &GpuDevice,
_kernel: &GpuKernelHandle,
n: usize,
indptr_buffer: &GpuBuffer<u32>,
indices_buffer: &GpuBuffer<u32>,
data_buffer: &GpuBuffer<T>,
b_buffer: &GpuBuffer<T>,
x_buffer: &GpuBuffer<T>,
_config: &GpuKernelConfig,
) -> Result<(), GpuError>