[Requirement] 支持torch.uint64在npu上的相关计算

[daifucong233]

除了创建 Tensor 啥也干不了

>>> A = torch.randint(0, 100, (1,), dtype=torch.uint64, device="cpu")
>>> A = torch.randint(0, 100, (1,), dtype=torch.uint64, device="npu")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
RuntimeError: self scalar_type:UInt64is not surpported.
[ERROR] 2026-08-03-15:37:30 (PID:235311, Device:0, RankID:0) ERR01002 OPS invalid type

>>> A = torch.empty((1,), dtype=torch.uint64, device="npu")
>>> A = A + 1
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
RuntimeError: add_out_npu_nocheck:third_party/op-plugin/op_plugin/ops/opapi/AddKernelNpuOpApi.cpp:52 NPU function error: call aclnnAdds failed, error code is 161002
[ERROR] 2026-08-03-15:41:28 (PID:235311, Device:0, RankID:0) ERR00100 PTA call acl api failed.
[PID: 235311] 2026-08-03-15:41:28.117.010 AclNN_Parameter_Error(EZ1001): Tensor self not implemented for DT_UINT64, should be in dtype support list [DT_FLOAT,DT_INT32,DT_INT64,DT_FLOAT16,DT_INT16,DT_INT8,DT_UINT8,DT_DOUBLE,DT_BOOL,DT_BFLOAT16,DT_COMPLEX64,DT_COMPLEX128,].