H100 no kernel image is available for execution on the device

[kehuanfeng]

I am trying to install from source with the following change,

--- a/cuda/Makefile
+++ b/cuda/Makefile
@@ -6,7 +6,8 @@ OPTIONS=-Iincludes \
                -gencode arch=compute_72,code=sm_72 \
                -gencode arch=compute_75,code=sm_75 \
                -gencode arch=compute_80,code=sm_80 \
-               -gencode arch=compute_86,code=sm_86
+               -gencode arch=compute_86,code=sm_86 \
+               -gencode arch=compute_90,code=sm_90

so that it can support hopper.
After compilation done, the kernel can't still be loaded

  File "/mnt/cpm_kernels/cpm_kernels/kernels/base.py", line 48, in __call__
    func = self._prepare_func()
  File "/mnt/cpm_kernels/cpm_kernels/kernels/base.py", line 40, in _prepare_func
    self._module.get_module(), self._func_name
  File "/mnt/cpm_kernels/cpm_kernels/kernels/base.py", line 24, in get_module
    self._module[curr_device] = cuda.cuModuleLoadData(self._code)
  File "/mnt/cpm_kernels/cpm_kernels/library/base.py", line 94, in wrapper
    return f(*args, **kwargs)
  File "/mnt/cpm_kernels/cpm_kernels/library/cuda.py", line 233, in cuModuleLoadData
    checkCUStatus(cuda.cuModuleLoadData(ctypes.byref(module), data))
  File "/mnt/cpm_kernels/cpm_kernels/library/cuda.py", line 216, in checkCUStatus
    raise RuntimeError("CUDA Error: %s" % cuGetErrorString(error))
RuntimeError: CUDA Error: no kernel image is available for execution on the device