Merge branch 'master' of https://git01.codeplex.com/cntk into linux-gcc

Conflicts: DataReader/LMSequenceReader/SequenceReader.cpp
raden · Jul 22, 2015 · fce1212 · fce1212
2 parents 75acf2b + c437f85
commit fce1212
Show file tree

Hide file tree

Showing 22 changed files with 188 additions and 163 deletions.
diff --git a/CNTK.sln b/CNTK.sln
@@ -165,7 +165,7 @@ Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "CheckInSuites", "CheckInSui
 		{E6F26F9A-FF64-4F0A-B749-CD309EE357EE} = {E6F26F9A-FF64-4F0A-B749-CD309EE357EE}
 	EndProjectSection
 EndProject
-Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "NDRMReader", "DataReader\NDRMReader\NDRMReader.vcxproj", "{CE429AA2-3778-4619-8FD1-49BA3B81197B}"
+Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "SparsePCReader", "DataReader\SparsePCReader\SparsePCReader.vcxproj", "{CE429AA2-3778-4619-8FD1-49BA3B81197B}"
 EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution

diff --git a/Common/CrossProcessMutex_linux.h b/Common/CrossProcessMutex_linux.h
@@ -62,7 +62,9 @@ class CrossProcessMutex
             setupTimeout(1);
             int r = fcntl(fd, wait ? F_SETLKW : F_SETLK, &m_lock);
             if (errno == EINTR) {
+                sleep(1);
                 // retrying in the case of signal or timeout
+                close(fd);
                 continue;
             }
             if (r != 0) {
@@ -80,6 +82,7 @@ class CrossProcessMutex
                 // we have a race with 'unlink' call in Release() 
                 // our lock is held to the previous instance of the file;
                 // this is not a problem, we just need to retry locking the new file
+                close(fd);
                 continue;
             }
             else {

diff --git a/Common/Include/DataReader.h b/Common/Include/DataReader.h
@@ -58,7 +58,7 @@ class DATAREADER_API IDataReader
     typedef unsigned int LabelIdType;
     unsigned m_seed;
     size_t   mBlgSize;  /// number of utterances per minibatch
-    bool     mDoRandomize; 
+    bool     mDoRandomize = true;
 
     virtual void Init(const ConfigParameters& /*config*/) = 0;
     virtual void Destroy() = 0;
@@ -147,9 +147,7 @@ class DataReader: public IDataReader<ElemType>, protected Plugin
     virtual void Destroy();
 
     /// number of utterances per minibatch, for data parallelsim
-    size_t mNbrUttPerMinibatch; 
-
-    bool mDoRandomize; 
+    size_t mNbrUttPerMinibatch;
 
 public:
     // DataReader Constructor

diff --git a/DataReader/LMSequenceReader/SequenceReader.cpp b/DataReader/LMSequenceReader/SequenceReader.cpp
@@ -1742,8 +1742,8 @@ bool BatchSequenceReader<ElemType>::EnsureDataAvailable(size_t /*mbStartSample*/
         mNumRead = m_parser.Parse(CACHE_BLOG_SIZE, &m_labelTemp, &m_featureTemp, &seqPos);
         if (mNumRead == 0) return false;
 
-        if (mDoRandomize) 
-            std::random_shuffle(m_parser.mSentenceIndex2SentenceInfo.begin(), m_parser.mSentenceIndex2SentenceInfo.end());
+        //if (mDoRandomize)
+        std::random_shuffle(m_parser.mSentenceIndex2SentenceInfo.begin(), m_parser.mSentenceIndex2SentenceInfo.end());
 
         m_readNextSampleLine += mNumRead;
         sLn = FindNextSentences(mNumRead);
@@ -1998,10 +1998,9 @@ bool BatchSequenceReader<ElemType>::DataEnd(EndDataType endDataType)
         ret = !EnsureDataAvailable(m_mbStartSample);
         break;
     case endDataSentence:  // for fast reader each minibatch is considered a "sentence", so always true
-            for (auto ptr = mToProcess.begin(); ptr != mToProcess.end(); ptr++)
-                mProcessed[*ptr] = true;
-
-        mSentenceEnd = true;
+        if (mSentenceEnd)
+        for (auto ptr = mToProcess.begin(); ptr != mToProcess.end(); ptr++)
+            mProcessed[*ptr] = true;
         ret = mSentenceEnd;
         break;
     }
@@ -2094,6 +2093,10 @@ void BatchSequenceReader<ElemType>::GetLabelOutput(std::map < std::wstring,
     else // GPU
     {
         RuntimeError("GetLabelOutput::should use CPU for labels ");
+    }  
+    if (curDevId != CPUDEVICE && readerMode != ReaderMode::Class)
+    {
+        labels->TransferFromDeviceToDevice(CPUDEVICE, curDevId, false, false, false);
     }
 }
 
@@ -2126,5 +2129,4 @@ int BatchSequenceReader<ElemType>::GetSentenceEndIdFromOutputLabel()
 
 template class BatchSequenceReader<double>; 
 template class BatchSequenceReader<float>;
-
-}}}
+}}}
diff --git a/DataReader/NDRMReader/Exports.cpp → DataReader/SparsePCReader/Exports.cpp b/DataReader/NDRMReader/Exports.cpp → DataReader/SparsePCReader/Exports.cpp
@@ -9,14 +9,14 @@
 #include "stdafx.h"
 #define DATAREADER_EXPORTS
 #include "DataReader.h"
-#include "NDRMReader.h"
+#include "SparsePCReader.h"
 
 namespace Microsoft { namespace MSR { namespace CNTK {
 
 template<class ElemType>
 void DATAREADER_API GetReader(IDataReader<ElemType>** preader)
 {
-    *preader = new NDRMReader<ElemType>();
+    *preader = new SparsePCReader<ElemType>();
 }
 
 extern "C" DATAREADER_API void GetReaderF(IDataReader<float>** preader)