implement torch.linalg.cholesky xpu backend

[jiqing-feng]

🚀 The feature, motivation and pitch

torch.linalg.cholesky is used by llm-compressor gptq. Now, it will show below log on xpu:

/home/sdp/jiqing/llm-compressor/src/llmcompressor/modifiers/quantization/gptq/gptq_quantize.py:160: UserWarning: Aten Op fallback from XPU
to CPU happends. This may have performance implications. If need debug the fallback ops please set environment variable `PYTORCH_DEBUG_XPU_
FALLBACK=1`  (Triggered internally at /pytorch/build/xpu/ATen/RegisterXPU_0.cpp:54523.)
  H = torch.linalg.cholesky(H)

Alternatives

No response

Additional context

No response

[CuiYifeng]

@jiqing-feng Could you provide the proportion of torch.linalg.cholesky in E2E time?

[daisyden]

Also impact pytorch upstream ut.

[chuanqi129]

@CuiYifeng can we check whether this op is fallback?