rk3399用clang编译

[corleonechensiyu]

pi@NanoPi-M4:/MNN/build$ cmake .. -DCMAKE_SYSTEM_NAME=Linux -DCMAKE_SYSTEM_VERSION=1 -DCMAKE_SYSTEM_PROCESSOR=aarch64 -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++
-- The C compiler identification is Clang 6.0.0
-- The CXX compiler identification is Clang 6.0.0
-- Check for working C compiler: /usr/bin/clang
-- Check for working C compiler: /usr/bin/clang -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/clang++
-- Check for working CXX compiler: /usr/bin/clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- The ASM compiler identification is Clang
-- Found assembler: /usr/bin/clang
-- >>>>>>>>>>>>>
-- MNN BUILD INFO:
-- System: Linux
-- Processor: aarch64
-- DEBUG: OFF
-- Metal: OFF
-- OpenCL: OFF
-- OpenGL: OFF
-- Vulkan: OFF
-- OpenMP: ON
-- Hideen: TRUE
-- [*] searching .c
-- [] searching .cc
-- [] searching .cpp
-- [] searching *.s .S
-- [] Checking OpenMP
-- Found OpenMP_C: -fopenmp=libomp (found version "3.1")
-- Found OpenMP_CXX: -fopenmp=libomp (found version "3.1")
-- Found OpenMP: TRUE (found version "3.1")
-- /home/pi/MNN/build/benchmark
-- Configuring done
-- Generating done
-- Build files have been written to: /home/pi/MNN/build
pi@NanoPi-M4:/MNN/build$ make -j4
Scanning dependencies of target MNN
[ 0%] Building C object CMakeFiles/MNN.dir/source/core/MNNMemoryUtils.c.o
[ 0%] Building CXX object CMakeFiles/MNN.dir/source/core/AutoTime.cpp.o
[ 1%] Building CXX object CMakeFiles/MNN.dir/source/core/Backend.cpp.o
[ 1%] Building CXX object CMakeFiles/MNN.dir/source/core/BackendFactory.cpp.o
[ 1%] Building CXX object CMakeFiles/MNN.dir/source/core/BufferAllocator.cpp.o
[ 2%] Building CXX object CMakeFiles/MNN.dir/source/core/Execution.cpp.o
[ 2%] Building CXX object CMakeFiles/MNN.dir/source/core/Interpreter.cpp.o
[ 3%] Building CXX object CMakeFiles/MNN.dir/source/core/Pipeline.cpp.o
[ 3%] Building CXX object CMakeFiles/MNN.dir/source/core/Schedule.cpp.o
[ 3%] Building CXX object CMakeFiles/MNN.dir/source/core/Session.cpp.o
[ 4%] Building CXX object CMakeFiles/MNN.dir/source/core/SizeComputer.cpp.o
[ 4%] Building CXX object CMakeFiles/MNN.dir/source/core/Tensor.cpp.o
[ 5%] Building CXX object CMakeFiles/MNN.dir/source/core/TensorUtils.cpp.o
[ 5%] Building CXX object CMakeFiles/MNN.dir/source/core/WrapExecution.cpp.o
[ 5%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUArgMax.cpp.o
[ 6%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUAsString.cpp.o
[ 6%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUBackend.cpp.o
[ 7%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUBatchMatMul.cpp.o
[ 7%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUBatchToSpaceND.cpp.o
[ 7%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUBinary.cpp.o
[ 8%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUCast.cpp.o
[ 8%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUConcat.cpp.o
[ 9%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUConst.cpp.o
[ 9%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUConvolution.cpp.o
[ 9%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUConvolutionDepthwise.cpp.o
[ 10%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUCrop.cpp.o
[ 10%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUCropAndResize.cpp.o
[ 11%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUDeconvolution.cpp.o
[ 11%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUDeconvolutionDepthwise.cpp.o
[ 11%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUDequantize.cpp.o
[ 12%] Building CXX object CMakeFiles/MNN.dir/source/backend/cpu/CPUDetectionOutput.cpp.o
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
CONV_SETUP_KERNELSIZE(4);
^
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:36:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int dst_depth_quad = UP_DIV(output->channel(), KB)...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:36:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int dst_depth_quad = UP_DIV(output->channel(), KB)...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:38:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int src_z_step = input->width() * input->height() * KB...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:38:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int src_z_step = input->width() * input->height() * KB...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:39:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int src_batch_step = input->stride(0)...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:39:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int src_batch_step = input->stride(0)...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:42:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int dst_batch_step = output->stride(0)...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:42:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int dst_batch_step = output->stride(0)...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:56:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int strideX_step = strideX * KB...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:56:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int strideX_step = strideX * KB...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:57:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_sy_step = KB * KB * kernel_width...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:57:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_sy_step = KB * KB * kernel_width...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:58:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_sz_step = KB * KB * kernel_width * kernel_height...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:58:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_sz_step = KB * KB * kernel_width * kernel_height...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: unknown type name
'__unused'
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:59:5: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_z_step = kernel_height * kernel_width * src_depth_qua...
^
/home/pi/MNN/include/MNNDefine.h:23:19: note: expanded from macro 'MNNUnused'
#define MNNUnused __unused
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: error: expected
unqualified-id
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:59:15: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
MNNUnused int weight_z_step = kernel_height * kernel_width * src_depth_qua...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:155:71: error: use of undeclared
identifier 'dst_depth_quad'; did you mean 'src_depth_quad'?
...= [this, width, height, count, tileCount, threadNumber, dst_depth_quad, kernelCo...
^~~~~~~~~~~~~~
src_depth_quad
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: note: 'src_depth_quad'
declared here
CONV_SETUP_KERNELSIZE(4);
^
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:37:9: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
int src_depth_quad = UP_DIV(input->channel(), KB)...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:156:53: error: use of undeclared
identifier 'src_z_step'
padX, padY, src_height, src_width, src_z_step, dilateX, dilateY...
^
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:168:39: error: use of undeclared
identifier 'dst_depth_quad'; did you mean 'src_depth_quad'?
for (int dz = 0; dz < dst_depth_quad; ++dz) {
^~~~~~~~~~~~~~
src_depth_quad
/home/pi/MNN/source/backend/cpu/CPUDeconvolution.cpp:148:5: note: 'src_depth_quad'
declared here
CONV_SETUP_KERNELSIZE(4);
^
/home/pi/MNN/source/backend/cpu/compute/ConvOpt.h:37:9: note: expanded from macro
'CONV_SETUP_KERNELSIZE'
int src_depth_quad = UP_DIV(input->channel(), KB)...
^
fatal error: too many errors emitted, stopping now [-ferror-limit=]
20 errors generated.
CMakeFiles/MNN.dir/build.make:710: recipe for target 'CMakeFiles/MNN.dir/source/backend/cpu/CPUDeconvolution.cpp.o' failed
make[2]: *** [CMakeFiles/MNN.dir/source/backend/cpu/CPUDeconvolution.cpp.o] Error 1
make[2]: *** Waiting for unfinished jobs....
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/MNN.dir/all' failed
make[1]: *** [CMakeFiles/MNN.dir/all] Error 2
Makefile:129: recipe for target 'all' failed
make: *** [all] Error 2

[corleonechensiyu]

使用aarch64-linux-gnu-gcc/g++ 没问题
（为啥不建一个qq群呀，学生不用钉钉。。。）

[deepbreath]

使用aarch64-linux-gnu-gcc/g++ 没问题
（为啥不建一个qq群呀，学生不用钉钉。。。）

阿里用腾讯QQ?

[li-qing]

unused在较低版本的clang中无法直接使用，后续会改成__attribute((unused)).

[liushuan]

为什么用的 交叉编译呢？ 本地编译是否可以呢？

[alohali]

使用aarch64-linux-gnu-gcc/g++ 没问题
（为啥不建一个qq群呀，学生不用钉钉。。。）

阿里用腾讯QQ?

做开源这就狭隘了
不行建个支付宝群= =！