jinhuang415 commented on a change in pull request #9810: remove MKL_EXPERIMENTAL and update make files for MKL-DNN URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r170617385
########## File path: docker_multiarch/arm.crosscompile.mk ########## @@ -115,13 +100,10 @@ USE_LAPACK_PATH = USE_INTEL_PATH = NONE # If use MKL only for BLAS, choose static link automatically to allow python wrapper -ifeq ($(USE_MKL2017), 0) +USE_STATIC_MKL = NONE ifeq ($(USE_BLAS), mkl) USE_STATIC_MKL = 1 Review comment: @ashokei thanks for the explanation, so can we use MKLML instead of full MKL if USE_MKLDNN=1 and USE_BLAS=mkl (otherwise use full MKL library if USE_BLAS=mkl)? Another way to change I can think is to set USE_STATIC_MKL to 0 in make/config.mk by checking above logic (so don't need to change mshadow repo since it is used by more than 1 projects), what do you think? Of course we can address this in a separate PR. ---------------------------------------------------------------- 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: [email protected] With regards, Apache Git Services
