Caffe: Caffe installation error with CUDNN V6.0
CXX src/caffe/data_transformer.cpp
In file included from ./include/caffe/util/device_alternate.hpp:40:0,
from ./include/caffe/common.hpp:19,
from ./include/caffe/blob.hpp:8,
from ./include/caffe/data_transformer.hpp:6,
from src/caffe/data_transformer.cpp:8:
./include/caffe/util/cudnn.hpp: In function ‘const char* cudnnGetErrorString(cudnnStatus_t)’:
./include/caffe/util/cudnn.hpp:21:10: warning: enumeration value ‘CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING’ not handled in switch [-Wswitch]
switch (status) {
^
./include/caffe/util/cudnn.hpp: In function ‘void caffe::cudnn::setConvolutionDesc(cudnnConvolutionStruct**, cudnnTensorDescriptor_t, cudnnFilterDescriptor_t, int, int, int, int)’:
./include/caffe/util/cudnn.hpp:113:70: error: too few arguments to function ‘cudnnStatus_t cudnnSetConvolution2dDescriptor(cudnnConvolutionDescriptor_t, int, int, int, int, int, int, cudnnConvolutionMode_t, cudnnDataType_t)’
pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION));
^
./include/caffe/util/cudnn.hpp:15:28: note: in definition of macro ‘CUDNN_CHECK’
cudnnStatus_t status = condition; \
^
In file included from ./include/caffe/util/cudnn.hpp:5:0,
from ./include/caffe/util/device_alternate.hpp:40,
from ./include/caffe/common.hpp:19,
from ./include/caffe/blob.hpp:8,
from ./include/caffe/data_transformer.hpp:6,
from src/caffe/data_transformer.cpp:8:
/usr/local/cuda-8.0/include/cudnn.h:500:27: note: declared here
cudnnStatus_t CUDNNWINAPI cudnnSetConvolution2dDescriptor( cudnnConvolutionDescriptor_t convDesc,
^
Makefile:585: recipe for target '.build_release/src/caffe/data_transformer.o' failed
make: *** [.build_release/src/caffe/data_transformer.o] Error 1
Kindy help me if anyone has any solution !
soumenms2015
All 21 comments
Which version of caffeare you using?
deepali-c
on 24 Jul 2017
I have the same problem.. Do you solved it?
fafulad
on 30 Aug 2017
Same problem using cuDNN v7.0 (August 3, 2017) for CUDA 8.0 on Ubuntu16.04
Saif-03
on 31 Aug 2017
Anyone found a solution?
letatanu
on 9 Sep 2017
You can use other caffe cudnn.hpp to replace current one.
Using：
cp /export/anshan/caffe/caffe-master/include/caffe/util/cudnn.hpp ./include/caffe/util/
AnshanTJU
on 12 Sep 2017
i meet the similar problem:
scopeserver@scopephotos:~/RaidDisk/DeepLearning/mwang/models/appearance-flow/caffe$ make all
CXX src/caffe/layers/softmax_layer.cpp
In file included from ./include/caffe/util/device_alternate.hpp:40:0,
from ./include/caffe/common.hpp:19,
from ./include/caffe/blob.hpp:8,
from ./include/caffe/layers/softmax_layer.hpp:6,
from src/caffe/layers/softmax_layer.cpp:4:
./include/caffe/util/cudnn.hpp: In function ‘const char* cudnnGetErrorString(cudnnStatus_t)’:
./include/caffe/util/cudnn.hpp:21:10: warning: enumeration value ‘CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING’ not handled in switch [-Wswitch]
switch (status) {
^
./include/caffe/util/cudnn.hpp: In function ‘void caffe::cudnn::setConvolutionDesc(cudnnConvolutionStruct**, cudnnTensorDescriptor_t, cudnnFilterDescriptor_t, int, int, int, int)’:
./include/caffe/util/cudnn.hpp:113:70: error: too few arguments to function ‘cudnnStatus_t cudnnSetConvolution2dDescriptor(cudnnConvolutionDescriptor_t, int, int, int, int, int, int, cudnnConvolutionMode_t, cudnnDataType_t)’
pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION));
^
./include/caffe/util/cudnn.hpp:15:28: note: in definition of macro ‘CUDNN_CHECK’
cudnnStatus_t status = condition; \
^
In file included from ./include/caffe/util/cudnn.hpp:5:0,
from ./include/caffe/util/device_alternate.hpp:40,
from ./include/caffe/common.hpp:19,
from ./include/caffe/blob.hpp:8,
from ./include/caffe/layers/softmax_layer.hpp:6,
from src/caffe/layers/softmax_layer.cpp:4:
/usr/local/cuda/include/cudnn.h:500:27: note: declared here
cudnnStatus_t CUDNNWINAPI cudnnSetConvolution2dDescriptor( cudnnConvolutionDescriptor_t convDesc,
^
Makefile:583: recipe for target '/home/scopeserver/RaidDisk/DeepLearning/mwang/models/appearance-flow/caffe/.build_release/src/caffe/layers/softmax_layer.o' failed
make: *** [/home/scopeserver/RaidDisk/DeepLearning/mwang/models/appearance-flow/caffe/.build_release/src/caffe/layers/softmax_layer.o] Error 1
civilman628
on 18 Sep 2017
cuDNN 7.0 build fail.
civilman628
on 18 Sep 2017
I'm having the same problem with Ubuntu 16.04.3 LTS with CUDA 8.0 and CUDNN 7.0.2.38-1
$ make all
CXX src/caffe/util/cudnn.cpp
In file included from /usr/local/cuda/include/channel_descriptor.h:62:0,
from /usr/local/cuda/include/cuda_runtime.h:90,
from /usr/include/cudnn.h:64,
from ./include/caffe/util/cudnn.hpp:5,
from src/caffe/util/cudnn.cpp:2:
/usr/local/cuda/include/cuda_runtime_api.h:1628:101: error: use of enum ‘cudaDeviceP2PAttr’ without previous declaration
extern __host__ __cudart_builtin__ cudaError_t CUDARTAPI cudaDeviceGetP2PAttribute(int *value, enum cudaDeviceP2PAttr attr, int srcDevice, int dstDevice);
^
In file included from /usr/local/cuda/include/channel_descriptor.h:62:0,
from /usr/local/cuda/include/cuda_runtime.h:90,
from /usr/include/cudnn.h:64,
from ./include/caffe/util/cudnn.hpp:5,
from src/caffe/util/cudnn.cpp:2:
/usr/local/cuda/include/cuda_runtime_api.h:5382:92: error: use of enum ‘cudaMemoryAdvise’ without previous declaration
extern __host__ cudaError_t CUDARTAPI cudaMemAdvise(const void *devPtr, size_t count, enum cudaMemoryAdvise advice, int device);
^
/usr/local/cuda/include/cuda_runtime_api.h:5438:98: error: use of enum ‘cudaMemRangeAttribute’ without previous declaration
extern __host__ cudaError_t CUDARTAPI cudaMemRangeGetAttribute(void *data, size_t dataSize, enum cudaMemRangeAttribute attribute, const void *devPtr, size_t count);
^
/usr/local/cuda/include/cuda_runtime_api.h:5474:102: error: use of enum ‘cudaMemRangeAttribute’ without previous declaration
extern __host__ cudaError_t CUDARTAPI cudaMemRangeGetAttributes(void **data, size_t *dataSizes, enum cudaMemRangeAttribute *attributes, size_t numAttributes, const void *devPtr, size_t count);
^
In file included from src/caffe/util/cudnn.cpp:2:0:
./include/caffe/util/cudnn.hpp: In function ‘const char* cudnnGetErrorString(cudnnStatus_t)’:
./include/caffe/util/cudnn.hpp:21:10: warning: enumeration value ‘CUDNN_STATUS_RUNTIME_IN_PROGRESS’ not handled in switch [-Wswitch]
switch (status) {
^
./include/caffe/util/cudnn.hpp:21:10: warning: enumeration value ‘CUDNN_STATUS_RUNTIME_FP_OVERFLOW’ not handled in switch [-Wswitch]
Makefile:581: recipe for target '.build_release/src/caffe/util/cudnn.o' failed
make: *** [.build_release/src/caffe/util/cudnn.o] Error 1
storyseeds
on 29 Sep 2017
I could solve the error by downloading the correct version of CUDNN.
for me it was v5.1 . You can try older version of CUDNN if the prblem persists. But from the error i could say your CUDNN version should be < 6.0
imranlivt
on 7 Oct 2017
For fixing the error "too few arguments" edit the cudnn.hpp file to add a suffix "_v4" for the cribbing function. i.e example replace function "cudnnSetConvolution2dDescriptor" with "cudnnSetConvolution2dDescriptor_v4".
Caffes requires cudnn v5. Cudnn v6 retains the functionality of v5 by maintaining separate function which goes by the same name except that is has a prefix. Make sure to use corrrect prefix by reading the well documented cudnn.hpp file
shriyashchougule
on 8 Nov 2017
check your Make.config
if you use CUDA_DIR := /usr/local/cuda-8.0
then
CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \
-gencode arch=compute_30,code=sm_30 \
-gencode arch=compute_35,code=sm_35 \
-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_52,code=sm_52 \
-gencode arch=compute_60,code=sm_60 \
-gencode arch=compute_61,code=sm_61 \
if you use CUDA_DIR := /usr/local/cuda-7.5
then
CUDA_ARCH := -gencode arch=compute_20,code=sm_20 \
-gencode arch=compute_20,code=sm_21 \
-gencode arch=compute_30,code=sm_30 \
-gencode arch=compute_35,code=sm_35 \
-gencode arch=compute_50,code=sm_50 \
-gencode arch=compute_52,code=sm_52 \
rock417
on 26 Dec 2017
Please always check your CUDA version and make sure you download the matching version of cuDNN. The above answers seem to provide sufficient help, so I'm closing this issue. Feel free to reopen if the problem reoccurs though.
Noiredd
on 2 Mar 2018
does the same problem persists in the latest caffe as well ? Is it tested with latest cuDNN and CUDA ?
aisri
on 28 Mar 2018
I'm using cuda-9.1 and i get the same errors when running make all
Following this guide
Edit:
As the guide suggested, i installed caffe-rc5 which turns out was the issue.
I cloned the latest repo of caffe and the issue went away.
ArthurSav
on 6 Apr 2018
It seems additional parameter is added to cudnnSetConvolution2dDescriptor.
I edited like this
cudnnSetConvolution2dDescriptor(*conv,
pad_h, pad_w, stride_h, stride_w, 1, 1, CUDNN_CROSS_CORRELATION, CUDNN_DATA_FLOAT))
and it compiled well.
(https://devtalk.nvidia.com/default/topic/1002826/gpu-accelerated-libraries/question-about-cudnnsetconvolution2ddescriptor/)
ByeongHeonYOO
on 6 May 2018
Faced the same problem. It's happening due to cudnn.hpp (Location: include/caffe/util/cudnn.hpp) . Update cudnn.hpp file. It is not considering the current cuDNN versions.
Update the cudnn.hpp from github repo, BVLC/caffe/include/caffe/util/cudnn.hpp
Or, go to this link https://github.com/BVLC/caffe/blob/master/include/caffe/util/cudnn.hpp [Copy this and replace the old one]
mdamircoder
on 13 Jul 2018
@mdamirsohail13 solved my problem
xiaofanglegoc
on 26 Sep 2018
@mdamirsohail13 solved my problem,too,thank you!
longlinzhou
on 12 Nov 2018
Faced the same problem. It's happening due to cudnn.hpp (Location: include/caffe/util/cudnn.hpp) . Update cudnn.hpp file. It is not considering the current cuDNN versions.
Update the cudnn.hpp from github repo, BVLC/caffe/include/caffe/util/cudnn.hpp
Or, go to this link https://github.com/BVLC/caffe/blob/master/include/caffe/util/cudnn.hpp [Copy this and replace the old one]
It works, you are my hero!!! Thank you!
BluebirdStory
on 13 Nov 2018
Faced the same problem. It's happening due to cudnn.hpp (Location: include/caffe/util/cudnn.hpp) . Update cudnn.hpp file. It is not considering the current cuDNN versions.
Update the cudnn.hpp from github repo, BVLC/caffe/include/caffe/util/cudnn.hpp
Or, go to this link https://github.com/BVLC/caffe/blob/master/include/caffe/util/cudnn.hpp [Copy this and replace the old one]
works for me
mxazz123
on 27 May 2019
You can use other caffe cudnn.hpp to replace current one.
Using：
cp /export/anshan/caffe/caffe-master/include/caffe/util/cudnn.hpp ./include/caffe/util/
God bless your soul, stranger
mas2tg
on 19 Jun 2020
Related issues
kelvinxu
·
3Comments
serimp
·
3Comments
OpenHero
·
3Comments
iamhankai
·
3Comments
prathmeshrmadhu
·
3Comments
Most helpful comment
Faced the same problem. It's happening due to cudnn.hpp (Location: include/caffe/util/cudnn.hpp) . Update cudnn.hpp file. It is not considering the current cuDNN versions.
Update the cudnn.hpp from github repo, BVLC/caffe/include/caffe/util/cudnn.hpp
Or, go to this link https://github.com/BVLC/caffe/blob/master/include/caffe/util/cudnn.hpp [Copy this and replace the old one]