运行多线程时双重释放或损坏 我在C++程序中遇到了一个运行时错误“双自由或损坏”,它调用了一个可靠的库ANN,并使用OpenMP来对A进行循环化。p> *** glibc detected *** /home/tim/test/debug/test: double free or corruption (!prev): 0x0000000002527260 ***
这是否意味着地址0x0000000002527260处的内存被多次释放 错误发生在函数classify_variable_k()内部的“_search_struct->annkSearch(queryPt,k_max,nnIdx,dists,_eps);”处,而函数classify_variable_k()又位于OpenMP for loop inside function tune_complexity()内部 请注意,当OpenMP有多个线程时会发生此错误,而在单线程情况下不会发生此错误。不知道为什么 下面是我的代码。如果还不够诊断,就告诉我。谢谢你的帮助运行多线程时双重释放或损坏 我在C++程序中遇到了一个运行时错误“双自由或损坏”,它调用了一个可靠的库ANN,并使用OpenMP来对A进行循环化。p> *** glibc detected *** /home/tim/test/debug/test: double free or corruption (!prev): 0x0000000002527260 ***,c++,multithreading,C++,Multithreading,这是否意味着地址0x0000000002527260处的内存被多次释放 错误发生在函数classify_variable_k()内部的“_search_struct->annkSearch(queryPt,k_max,nnIdx,dists,_eps);”处,而函数classify_variable_k()又位于OpenMP for loop inside function tune_complexity()内部 请注意,当OpenMP有多个线程时会发生此错误,而在单线程情况下不会发生此错误。不
void KNNClassifier::train(int nb_examples, int dim, double **features, int * labels) {
_nPts = nb_examples;
_labels = labels;
_dataPts = features;
setting_ANN(_dist_type,1);
delete _search_struct;
if(strcmp(_search_neighbors, "brutal") == 0) {
_search_struct = new ANNbruteForce(_dataPts, _nPts, dim);
}else if(strcmp(_search_neighbors, "kdtree") == 0) {
_search_struct = new ANNkd_tree(_dataPts, _nPts, dim);
}
}
void KNNClassifier::classify_various_k(int dim, double *feature, int label, int *ks, double * errors, int nb_ks, int k_max) {
ANNpoint queryPt = 0;
ANNidxArray nnIdx = 0;
ANNdistArray dists = 0;
queryPt = feature;
nnIdx = new ANNidx[k_max];
dists = new ANNdist[k_max];
if(strcmp(_search_neighbors, "brutal") == 0) {
_search_struct->annkSearch(queryPt, k_max, nnIdx, dists, _eps);
}else if(strcmp(_search_neighbors, "kdtree") == 0) {
_search_struct->annkSearch(queryPt, k_max, nnIdx, dists, _eps); // where error occurs
}
for (int j = 0; j < nb_ks; j++)
{
scalar_t result = 0.0;
for (int i = 0; i < ks[j]; i++) {
result+=_labels[ nnIdx[i] ];
}
if (result*label<0) errors[j]++;
}
delete [] nnIdx;
delete [] dists;
}
void KNNClassifier::tune_complexity(int nb_examples, int dim, double **features, int *labels, int fold, char *method, int nb_examples_test, double **features_test, int *labels_test) {
int nb_try = (_k_max - _k_min) / scalar_t(_k_step);
scalar_t *error_validation = new scalar_t [nb_try];
int *ks = new int [nb_try];
for(int i=0; i < nb_try; i ++){
ks[i] = _k_min + _k_step * i;
}
if (strcmp(method, "ct")==0)
{
train(nb_examples, dim, features, labels );// train once for all nb of nbs in ks
for(int i=0; i < nb_try; i ++){
if (ks[i] > nb_examples){nb_try=i; break;}
error_validation[i] = 0;
}
int i = 0;
#pragma omp parallel shared(nb_examples_test, error_validation,features_test, labels_test, nb_try, ks) private(i)
{
#pragma omp for schedule(dynamic) nowait
for (i=0; i < nb_examples_test; i++)
{
classify_various_k(dim, features_test[i], labels_test[i], ks, error_validation, nb_try, ks[nb_try - 1]); // where error occurs
}
}
for (i=0; i < nb_try; i++)
{
error_validation[i]/=nb_examples_test;
}
}
......
}
void knn分类器::train(int nb_示例,int dim,双**特征,int*标签){
_nPts=nb_示例;
_标签=标签;
_数据点=特征;
设置ANN(_dist_type,1);
删除搜索结构;
如果(strcmp(_search_neighbories,“野蛮”)==0{
_search_struct=newannbruteforce(_dataPts,_nPts,dim);
}else if(strcmp(_search_neights,“kdtree”)==0{
_search_struct=新的ANNkd_树(_dataPts,_nPts,dim);
}
}
void knn分类器::分类各种(int dim,double*特征,int标签,int*ks,double*错误,int nb_ks,int k_max){
ANNpoint queryPt=0;
annidx数组nnIdx=0;
ANNdistArray dists=0;
queryPt=特征;
nnIdx=新的ANNidx[k_max];
dists=新的ANNdist[k_max];
如果(strcmp(_search_neighbories,“野蛮”)==0{
_搜索结构->annkSearch(queryPt、k_max、nnIdx、dists、eps);
}else if(strcmp(_search_neights,“kdtree”)==0{
_search_struct->annkSearch(queryPt,k_max,nnIdx,dists,_eps);//发生错误的位置
}
对于(int j=0;j
更新: 谢谢!我现在正试图通过使用“#pragma omp critical”来纠正classify_variable_k()中写入相同内存问题的冲突:
void knn分类器::分类各种(int dim,double*特征,int标签,int*ks,double*错误,int nb\u ks,int k\u max){
ANNpoint queryPt=0;
annidx数组nnIdx=0;
ANNdistArray dists=0;
queryPt=feature;//for(int i=0;icontent[i];}
nnIdx=新的ANNidx[k_max];
dists=新的ANNdist[k_max];
if(strcmp(_search_neights,“野蛮”)==0{//search
_搜索结构->annkSearch(queryPt、k_max、nnIdx、dists、eps);
}else if(strcmp(_search_neights,“kdtree”)==0{
_搜索结构->annkSearch(queryPt、k_max、nnIdx、dists、eps);
}
对于(int j=0;j 如果(result*label您的train方法在分配新内存之前删除了_search\u struct。因此,第一次调用train时,它会被删除。在调用train之前,是否有代码分配它?您可能会尝试删除垃圾内存(尽管我们没有代码告诉您)。我不知道这是否是您的问题,但是:
void KNNClassifier::train(int nb_examples, int dim, double **features, int * labels) {
...
delete _search_struct;
if(strcmp(_search_neighbors, "brutal") == 0) {
_search_struct = new ANNbruteForce(_dataPts, _nPts, dim);
}else if(strcmp(_search_neighbors, "kdtree") == 0) {
_search_struct = new ANNkd_tree(_dataPts, _nPts, dim);
}
}
如果您不属于if
或else if
子句,会发生什么?您已删除\u search\u struct
并将其指向垃圾。之后应将其设置为NULL
如果这不是问题,您可以尝试更换:
delete p;
与:
(或者类似地,对于delete[]
站点)。(但是,这可能会对第一次调用knnlindicator::train
造成问题。)
另外,必须:您真的需要执行所有这些手动分配和解除分配吗?为什么您不至少使用std::vector
而不是new[]
/delete[]
(这几乎总是不好的)?好的,既然您已经声明它在单线程情况下工作正常方法无效。您需要执行以下操作:
- 查找并行访问的所有变量
- 尤其是
delete p;
assert(p != NULL);
delete p;
p = NULL;
shared(nb_examples_test, error_validation,features_test, labels_test, nb_try, ks)
for (int i = 0; i < ks[j]; i++) {
result+=_labels[ nnIdx[i] ];
}
if (result*label<0) errors[j]++;