Fix NaN propagation for float16 min and max operators

[adamreeve]

Description

This makes min and max with NaN for either operand always return NaN for float16 data, matching the behaviour of float and double.

The behaviour for floats and doubles was previously fixed for the CPU provider in #21492 and the CUDA provider in #19984, but these PRs didn't fix the behaviour for float16 due to tests causing asan errors. The memory access violations with float16 data have now been fixed in #22135, so this PR is a follow up to make float16 min and max behave the same as float and double for both the CPU and CUDA providers now that we can add tests for this.

Motivation and Context

Relevant previous issues (not float16 specific):

• Incorrect NaN handling for Min and Max operators on CPU with a single element input #21455
• clip behaviour with NaN values is different between GPU and CPU onnx inference onnx/onnx#6003

[tianleiwu]

May use __hmin_nan(a, b) for _Min and __hmax_nan(a, b) for _Max, when a/b is half or bfloat16 (casted to __nv_bfloat16). See https://docs.nvidia.com/cuda/archive/12.1.0/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html#group__CUDA__MATH____HALF__COMPARISON_1g9752fee573b47368538ab19495ab9623

Difference is that those functions return canonical NaN, and more consistent with float/double (see line 366).

[tianleiwu]

How about logic like

        Eigen::half scalar_input = static_cast<Eigen::half>(per_iter_bh.ScalarInput0<MLFloat16>());
        scalar_input = std::isnan(static_cast<float>(scalar_input)) ? Eigen::NumTraits<Eigen::half>::quiet_NaN() : scalar_input;

        if (is_min) {
            output_vec_map = input_1_vec_map.isNaN().select(
                Eigen::NumTraits<Eigen::half>::quiet_NaN(),
                input_1_vec_map.template min<Eigen::PropagateNaN>(scalar_input)
            );
        } else {
            output_vec_map = input_1_vec_map.isNaN().select(
                Eigen::NumTraits<Eigen::half>::quiet_NaN(),
                input_1_vec_map.template max<Eigen::PropagateNaN>(scalar_input)
            );
        }

[tianleiwu]

How about logic like

output_vec_map = (input_0_vec_map.isNaN() || input_1_vec_map.isNaN()).select(
                Eigen::NumTraits<Eigen::half>::quiet_NaN(),
                input_0_vec_map.template min<Eigen::PropagateNaN>(input_1_vec_map)
            );