Error: unknown mnemonic `pause' -- `pause' when using cmake to build Caffe on Tegra X1

Hi,

I am trying to build the caffe fp16 version on a Tegra X1 platform.

Following is the error that happens. I have tried using make all -j3, make all -j2, when building. I am installing it over ssh. Can anyone give me some pointers to what might be causing it?

[ 80%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/hdf5.cpp.o
[ 80%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/signal_handler.cpp.o
[ 81%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/db_lmdb.cpp.o
[ 81%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/io.cpp.o
[ 82%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/db.cpp.o
[ 82%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/gpu_memory.cpp.o
[ 82%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/db_leveldb.cpp.o
[ 84%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/insert_splits.cpp.o
[ 84%] Building CXX object src/caffe/CMakeFiles/caffe.dir/util/benchmark.cpp.o
/tmp/ccof9TBz.s: Assembler messages:
/tmp/ccof9TBz.s:1309: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:1742: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:2188: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:2951: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:3859: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:4063: Error: unknown mnemonic `pause' -- `pause'
/tmp/ccof9TBz.s:4948: Error: unknown mnemonic `pause' -- `pause'
src/caffe/CMakeFiles/caffe.dir/build.make:2232: recipe for target 'src/caffe/CMakeFiles/caffe.dir/util/gpu_memory.cpp.o' failed
make[2]: *** [src/caffe/CMakeFiles/caffe.dir/util/gpu_memory.cpp.o] Error 1
make[2]: *** Waiting for unfinished jobs....
CMakeFiles/Makefile2:218: recipe for target 'src/caffe/CMakeFiles/caffe.dir/all' failed
make[1]: *** [src/caffe/CMakeFiles/caffe.dir/all] Error 2
Makefile:117: recipe for target 'all' failed
make: *** [all] Error 2

Hello, tousif111:
I guess you are using 64bit, right?

in file “3rdparty/cub/cub/host/spinlock.cuh”
it only contains #ifndef arm

you can comment out the following ‘pause’ code. or change arm to aarch64

br
ChenJian

Hi ChenJian,

I have the same question and my GPU is P40 on a ppc64le system.
And I didn’t find “3rdparty/cub/cub/host/spinlock.cuh” in my caffe directory
which clone from NVIDIA official webpage on github.

Do you have any suggestion?
Thanks

BR
Mike

Hi DJmimi,

For issue on other GPU than Tegra, you may post it to another board:
https://devtalk.nvidia.com/default/board/55/gpu-accelerated-libraries/

Thanks

Hi,

CXX/LD -o .build_release/tools/convert_imageset.bin
CXX/LD -o .build_release/tools/extract_features.bin
CXX/LD -o .build_release/tools/compute_image_mean.bin
CXX/LD -o .build_release/tools/upgrade_net_proto_binary.bin
.build_release/lib/libcaffe-nv.so: undefined reference to cudnnSetConvolutionMathType' collect2: error: ld returned 1 exit status Makefile:657: recipe for target '.build_release/tools/convert_imageset.bin' failed make: *** [.build_release/tools/convert_imageset.bin] Error 1 make: *** Waiting for unfinished jobs.... .build_release/lib/libcaffe-nv.so: undefined reference to cudnnSetConvolutionMathType’
collect2: error: ld returned 1 exit status
Makefile:657: recipe for target ‘.build_release/tools/upgrade_net_proto_binary.bin’ failed
make: *** [.build_release/tools/upgrade_net_proto_binary.bin] Error 1
.build_release/lib/libcaffe-nv.so: undefined reference to cudnnSetConvolutionMathType' collect2: error: ld returned 1 exit status .build_release/lib/libcaffe-nv.so: undefined reference to cudnnSetConvolutionMathType’
collect2: error: ld returned 1 exit status
Makefile:657: recipe for target ‘.build_release/tools/compute_image_mean.bin’ failed
make: *** [.build_release/tools/compute_image_mean.bin] Error 1
Makefile:657: recipe for target ‘.build_release/tools/extract_features.bin’ failed
make: *** [.build_release/tools/extract_features.bin] Error 1

How can I solve this error?

Hi Pratosha,

It seems your issue got resolved from below link, if not, please file a new topic with clear description.

https://devtalk.nvidia.com/default/topic/1022658/gpu-accelerated-libraries/install-cudnn-from-deb/post/5237199/#5237199

Thanks