原文地址:http://blog.sina.com.cn/s/blog_141f234870102w8is.html
1.visual studio 2013 error C2371: 'int8_t' : redefinition;
2. error C3861: 'getpid': identifier not found
- pid
= getpid(); ——>pid = _getpid();
3.
参考:ffmpeg编译总结
- usleep(is->audio_st
&& is->show_audio ? rdftspeed*1000 : 5000); - —>
Sleep (is->audio_st && is->show_audio ? rdftspeed*1 : 5); - usleep(300
* 1000); —> Sleep(300);
4. error C3861: 'snprintf': identifier not found
参考:http://blog.163.com/wanghuajie@126/blog/static/452312862009111114434838
5. error C3861: '__builtin_popcount': identifier not found
参考:http://blog.csdn.net/rappy/article/details/1788969
- template
<</span>typename Dtype> - unsigned
int __builtin_popcount(Dtype u) - {
-
u = (u & 0x55555555) + ((u >> 1) & 0x55555555); -
u = (u & 0x33333333) + ((u >> 2) & 0x33333333); -
u = (u & 0x0F0F0F0F) + ((u >> 4) & 0x0F0F0F0F); -
u = (u & 0x00FF00FF) + ((u >> 8) & 0x00FF00FF); -
u = (u & 0x0000FFFF) + ((u >> 16) & 0x0000FFFF); -
return u; - }//wishchin!!!
6.error : identifier "::caffe::kBNLL_THRESHOLD" is undefined in device code
参考:
- Dtype
expval = exp(min(in_data[index], Dtype(kBNLL_THRESHOLD))); - ——>Dtype
expval = exp(min(in_data[index], Dtype(50)));
7. error C2660: 'mkdir' : function does not take 2 arguments
参考:
- CHECK_EQ(mkdir(filename_->c_str(),
0744), 0) << "mkdir " << filename_<< "failed";
8.error C2784: '_Ty std::max(std::initializer_list<_Elem>,_Pr)' : could not de
9.error C4996: 'std::_Copy_impl': Function call with parameters that may be unsafe - this call relies on the caller to check that the passed values are correct. To disable this warning, use -D_SCL_SECURE_NO_WARNINGS.
XXX:不断出现的
10. error C3861: 'close': identifier not found
error C3861: 'open': identifier not found
在io.hpp文件里加入#include
11. Check failed: error == cudaSuccess (2 vs. 0)
train的batch_size设置过大,改小后可以. (个人经验不足所以一直在修改生成h5 file里的“chunksize”这个变量从64改到了1都还是会有 out of memory 的错误,而且也发现怎么改那个变量所消耗的内存数几乎不变。为此郁闷了一整天....后来才发现应该要改的是 train.prototxt里的 “batch_size", 一改以后马上内存就小了。在此记录一下。)
12.Check failed: error == cudaSuccess (38 vs. 0)
可能原因:solver的参数没有设置或错误
13.check failed:error==cudasuccess(11
可能原因:test的batch_size 和solver的test_iter设置有问题
14.cudnn.h中
inline const char* cudnnGetErrorString(cudnnStatus_t status)
inline const char* CUDNNWINAPI
15.error : identifier "cudnnTensor4dDescriptor_t" is undefined
cudnn版本不兼容问题,下载cudnn R1替换掉就可以。