mseeger commented on issue #8338: master branch cannot build on centos 7 with cuda-8.0 URL: https://github.com/apache/incubator-mxnet/issues/8338#issuecomment-343281399 Sorry I have no idea what is going on here. The errors you report do not happen in any of the tests on any system we test on. The warnings reported above should be gone. The error is reported in some tensor_gpu-inl.cuh(75) in mshadow, and not in mshadow_op.h. It is also not clear to me what real piece of code is raising the error (.cuh is a header file, containing templates). The only function in mshadow_op.h that appears in the errors, is smooth_l1_*. The macros do not affect this function, it could well be rolled back (but it seems this does not fix the issue). Somebody who knows what the templates in tensor_gpu-inl.cuh(75) are doing (a mshadow developer) would be the right person to ask. The code in mshadow_op.h before my changes was broken in many ways, not just that the double type was not supported. I cannot work on your setup, because it is not covered by our tests.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services