Question

如果我尝试在具有NDIVA GPU的服务器上运行OpenCL内核，则会遇到以下问题，在我的Mac上没有问题。这行代码似乎是问题所在，

float largest_0 = max(float (sin_i_angle), float (cos_i_angle));

这是错误消息。

  File "threed_dp.py", line 918, in gpu_calculate_segment_costs_orig
    bld = prg.build()
  File "/work/mrdrygal/.local/lib/python3.6/site-packages/pyopencl/__init__.py", line 510, in build
    options_bytes=options_bytes, source=self._source)
  File "/work/mrdrygal/.local/lib/python3.6/site-packages/pyopencl/__init__.py", line 554, in _build_and_catch_errors
    raise err
pyopencl._cl.RuntimeError: clBuildProgram failed: BUILD_PROGRAM_FAILURE - clBuildProgram failed: BUILD_PROGRAM_FAILURE - clBuildProgram failed: BUILD_PROGRAM_FAILURE

Build on <pyopencl.Device 'Tesla P100-PCIE-16GB' on 'NVIDIA CUDA' at 0x3767e50>:

<kernel>:82:33: error: expected expression
          float largest_0 = max(float (sin_i_angle), float (cos_i_angle));

Answer 1

float (sin_i_angle)

在C中不是有效的表达式。在C ++中有效（显式调用float()上的讲师），因此也许这就是Apple的OpenCL编译器允许它的原因。您应该将行更改为：

float largest_0 = max((float)sin_i_angle, (float)cos_i_angle);

在NVIDIA GPU上编译GPU内核的预期表达错误

1 个答案: