MRRTF-154: MCH raw data decoder now catches exceptions

Author: aphecetche

Detectors/MUON/MCH/Raw/Decoder/include/MCHRawDecoder/DataDecoder.h

@@ -158,8 +158,12 @@ class DataDecoder
   /// Must be called before processing the TmeFrame buffer
   void setFirstOrbitInTF(uint32_t orbit);
 
-  /// Decode one TimeFrame buffer and fill the vector of digits
-  void decodeBuffer(gsl::span<const std::byte> buf);
+  /** Decode one TimeFrame buffer and fill the vector of digits.
+   *  @return true if decoding went ok, or false otherwise.
+   *  if false is returned, the decoding of the (rest of the) TF should be
+   *  abandonned simply.
+   */
+  bool decodeBuffer(gsl::span<const std::byte> buf);
 
   /// Functions to set and get the calibration offset for the SAMPA time computation
   void setSampaBcOffset(uint32_t offset) { mSampaTimeOffset = offset; }

Detectors/MUON/MCH/Raw/Decoder/include/MCHRawDecoder/ErrorCodes.h

@@ -22,20 +22,20 @@ namespace raw
 {
 
 enum ErrorCodes {
-  ErrorParity = 1,                    // 1
-  ErrorHammingCorrectable = 1 << 1,   // 2
-  ErrorHammingUncorrectable = 1 << 2, // 4
-  ErrorBadClusterSize = 1 << 3,       // 8
-  ErrorBadPacketType = 1 << 4,        // 16
-  ErrorBadHeartBeatPacket = 1 << 5,   // 32
-  ErrorBadIncompleteWord = 1 << 6,    // 64
-  ErrorTruncatedData = 1 << 7,        // 128
-  ErrorBadELinkID = 1 << 8,           // 256
-  ErrorBadLinkID = 1 << 9,            // 512
-  ErrorUnknownLinkID = 1 << 10,       // 1024
-  ErrorBadDigitTime = 1 << 11,        // 2048
-  ErrorInvalidDigitTime = 1 << 12     // 4096
-
+  ErrorParity = 1,                           // 1
+  ErrorHammingCorrectable = 1 << 1,          // 2
+  ErrorHammingUncorrectable = 1 << 2,        // 4
+  ErrorBadClusterSize = 1 << 3,              // 8
+  ErrorBadPacketType = 1 << 4,               // 16
+  ErrorBadHeartBeatPacket = 1 << 5,          // 32
+  ErrorBadIncompleteWord = 1 << 6,           // 64
+  ErrorTruncatedData = 1 << 7,               // 128
+  ErrorBadELinkID = 1 << 8,                  // 256
+  ErrorBadLinkID = 1 << 9,                   // 512
+  ErrorUnknownLinkID = 1 << 10,              // 1024
+  ErrorBadDigitTime = 1 << 11,               // 2048
+  ErrorInvalidDigitTime = 1 << 12,           // 4096
+  ErrorNonRecoverableDecodingError = 1 << 13 // 8192
 };
 
 uint32_t getErrorCodesSize();

Detectors/MUON/MCH/Raw/Decoder/src/DataDecoder.cxx

@@ -24,6 +24,7 @@
 #include "CommonConstants/LHCConstants.h"
 #include "DetectorsRaw/RDHUtils.h"
 #include "DetectorsRaw/HBFUtils.h"
+#include "MCHBase/DecoderError.h"
 #include "MCHMappingInterface/Segmentation.h"
 #include "Framework/Logger.h"
 #include "MCHRawDecoder/ErrorCodes.h"
@@ -264,7 +265,7 @@ void DataDecoder::setFirstOrbitInTF(uint32_t orbit)
 
 //_________________________________________________________________________________________________
 
-void DataDecoder::decodeBuffer(gsl::span<const std::byte> buf)
+bool DataDecoder::decodeBuffer(gsl::span<const std::byte> buf)
 {
   if (mDebug) {
     std::cout << "\n\n============================\nStart of new buffer\n";
@@ -289,7 +290,12 @@ void DataDecoder::decodeBuffer(gsl::span<const std::byte> buf)
     auto pageSize = o2::raw::RDHUtils::getOffsetToNext(rdh);
 
     gsl::span<const std::byte> page(reinterpret_cast<const std::byte*>(rdh), pageSize);
-    decodePage(page);
+    try {
+      decodePage(page);
+    } catch (const std::exception& e) {
+      mErrors.emplace_back(DecoderError(0, 0, 0, ErrorNonRecoverableDecodingError));
+      return false;
+    }
 
     pageStart += pageSize;
   }
@@ -300,6 +306,7 @@ void DataDecoder::decodeBuffer(gsl::span<const std::byte> buf)
     std::cout << "[decodeBuffer] mDigits size: " << mDigits.size() << std::endl;
     dumpDigits();
   }
+  return true;
 }
 
 //_________________________________________________________________________________________________

Detectors/MUON/MCH/Raw/Decoder/src/ErrorCodes.cxx

@@ -20,7 +20,7 @@ namespace raw
 
 uint32_t getErrorCodesSize()
 {
-  return 13;
+  return 14;
 }
 
 void append(const char* msg, std::string& to)
@@ -76,6 +76,9 @@ std::string errorCodeAsString(uint32_t ec)
   if (ec & ErrorInvalidDigitTime) {
     append("Invalid Digit Time", msg);
   }
+  if (ec & ErrorNonRecoverableDecodingError) {
+    append("Non Recoverable", msg);
+  }
   return msg;
 }
 

Detectors/MUON/MCH/Workflow/src/DataDecoderSpec.cxx

@@ -122,15 +122,20 @@ class DataDecoderTask
     // get the input buffer
     auto& inputs = pc.inputs();
     DPLRawParser parser(inputs, o2::framework::select(mInputSpec.c_str()));
-    for (auto it = parser.begin(), end = parser.end(); it != end; ++it) {
+    bool abort{false};
+    for (auto it = parser.begin(), end = parser.end(); it != end && abort == false; ++it) {
       auto const* raw = it.raw();
       if (!raw) {
         continue;
       }
       size_t payloadSize = it.size();
 
       gsl::span<const std::byte> buffer(reinterpret_cast<const std::byte*>(raw), sizeof(RDH) + payloadSize);
-      mDecoder->decodeBuffer(buffer);
+      bool ok = mDecoder->decodeBuffer(buffer);
+      if (!ok) {
+        LOG(alarm) << "critical decoding error : aborting this TF decoding\n";
+        abort = true;
+      }
     }
   }