added equipment option dropPagesWithError

Author: sy-c

doc/configurationParameters.md

@@ -103,6 +103,7 @@ The parameters related to 3rd-party libraries are described here for convenience
 | equipment-* | dataPagesLogPath | string |  | Path where to save a summary of each data pages generated by equipment. |
 | equipment-* | debugFirstPages | int | 0 | If set, print debug information for first (given number of) data pages readout. |
 | equipment-* | disableOutput | int | 0 | If non-zero, data generated by this equipment is discarded immediately and is not pushed to output fifo of readout thread. Used for testing. |
+| equipment-* | dropPagesWithError | int | 0 | If set, the pages with RDH errors are discarded (requires rdhCheckEnabled or rdhUseFirstInPage). |
 | equipment-* | enabled | int | 1 | Enable (value=1) or disable (value=0) the equipment. |
 | equipment-* | equipmentType | string |  | The type of equipment to be instanciated. One of: dummy, rorc, cruEmulator |
 | equipment-* | firstPageOffset | bytes | | Offset of the first page, in bytes from the beginning of the memory pool. If not set (recommended), will start at memoryPoolPageSize (one free page is kept before the first usable page for readout internal use). |

doc/releaseNotes.md

@@ -524,3 +524,7 @@ This file describes the main feature changes for each readout.exe released versi
 ## v2.17.0 - 01/03/2022
 - Updated configuration syntax: section names ending with `-*` can be used to define default parameters. They are applied to all section with similar names. Existing key-value pairs are not overwritten, but are defined according to defaults if they don't exist. For example, it is possible to define the TFperiod for all equipments by adding a section named `[equipment-*]` with `TFperiod=32`.
 - Updated readout to new bookkeeping API.
+
+## next version
+- Updated configuration parameters:
+  - added equipment-*.dropPagesWithError: if set, the pages with RDH errors are discarded (requires rdhCheckEnabled or rdhUseFirstInPage). This may be used if downstream software is not robust to RDH errors.

src/ReadoutEquipment.cxx

@@ -145,7 +145,9 @@ ReadoutEquipment::ReadoutEquipment(ConfigFile& cfg, std::string cfgEntryPoint, b
   cfg.getOptionalValue<int>(cfgEntryPoint + ".rdhCheckDetectorField", cfgRdhCheckDetectorField);  
   // configuration parameter: | equipment-* | rdhCheckTrigger | int | 0 | If set, the RDH trigger counters are checked for consistency. |
   cfg.getOptionalValue<int>(cfgEntryPoint + ".rdhCheckTrigger", cfgRdhCheckTrigger);
-  theLog.log(LogInfoDevel_(3002), "RDH settings: rdhCheckEnabled=%d rdhDumpEnabled=%d rdhDumpErrorEnabled=%d rdhDumpWarningEnabled=%d rdhUseFirstInPageEnabled=%d rdhCheckFirstOrbit=%d rdhCheckDetectorField=%d", cfgRdhCheckEnabled, cfgRdhDumpEnabled, cfgRdhDumpErrorEnabled, cfgRdhDumpWarningEnabled, cfgRdhUseFirstInPageEnabled, cfgRdhCheckFirstOrbit, cfgRdhCheckDetectorField);
+  // configuration parameter: | equipment-* | dropPagesWithError | int | 0 | If set, the pages with RDH errors are discarded (requires rdhCheckEnabled or rdhUseFirstInPage). |
+  cfg.getOptionalValue<int>(cfgEntryPoint + ".dropPagesWithError", cfgDropPagesWithError);
+  theLog.log(LogInfoDevel_(3002), "RDH settings: rdhCheckEnabled=%d rdhDumpEnabled=%d rdhDumpErrorEnabled=%d rdhDumpWarningEnabled=%d rdhUseFirstInPageEnabled=%d rdhCheckFirstOrbit=%d rdhCheckDetectorField=%d dropPagesWithError=%d", cfgRdhCheckEnabled, cfgRdhDumpEnabled, cfgRdhDumpErrorEnabled, cfgRdhDumpWarningEnabled, cfgRdhUseFirstInPageEnabled, cfgRdhCheckFirstOrbit, cfgRdhCheckDetectorField, cfgDropPagesWithError);
 
   // configuration parameter: | equipment-* | ctpMode | int | 0 | If set, the detector field (CTP run mask) is checked. Incoming data is discarded until a new bit is set, and discarded again after this bit is unset. Automatically implies rdhCheckDetectorField=1 and rdhCheckDetectorField=1. |
   cfg.getOptionalValue<int>(cfgEntryPoint + ".ctpMode", cfgCtpMode);
@@ -430,7 +432,14 @@ Thread::CallbackResult ReadoutEquipment::threadCallback(void* arg)
 
 	// handle RDH-formatted data
 	if (ptr->cfgRdhUseFirstInPageEnabled) {
-          ptr->processRdh(nextBlock);
+          if ((ptr->processRdh(nextBlock)) && ptr->cfgDropPagesWithError) {
+            // drop pages with error when configured to do so
+            ptr->statsRdhCheckPagesDropped++;
+            static InfoLogger::AutoMuteToken logPageErrorDrop(LogWarningSupport_(3235), 10, 60);
+            theLog.log(logPageErrorDrop, "Equipment %s : page with RDH error has been discarded (total: %llu)", ptr->name.c_str(), ptr->statsRdhCheckPagesDropped);
+            nextBlock = nullptr;
+            continue;
+          }
 	}
 
         // discard data immediately if configured to do so
@@ -593,6 +602,7 @@ void ReadoutEquipment::initCounters()
   statsRdhCheckOk = 0;
   statsRdhCheckErr = 0;
   statsRdhCheckStreamErr = 0;
+  statsRdhCheckPagesDropped = 0;
 
   statsNumberOfTimeframes = 0;
 
@@ -621,7 +631,8 @@ void ReadoutEquipment::initCounters()
 void ReadoutEquipment::finalCounters()
 {
   if (cfgRdhCheckEnabled) {
-    theLog.log(LogInfoDevel_(3003), "Equipment %s : %llu timeframes, RDH checks %llu ok, %llu errors, %llu stream inconsistencies", name.c_str(), statsNumberOfTimeframes, statsRdhCheckOk, statsRdhCheckErr, statsRdhCheckStreamErr);
+    theLog.log(LogInfoDevel_(3003), "Equipment %s : %llu timeframes, RDH checks %llu ok, %llu errors, %llu stream inconsistencies, %llu pages with error dropped",
+      name.c_str(), statsNumberOfTimeframes, statsRdhCheckOk, statsRdhCheckErr, statsRdhCheckStreamErr, statsRdhCheckPagesDropped);
   }
 };
 
@@ -1037,7 +1048,7 @@ int ReadoutEquipment::processRdh(DataBlockContainerReference& block)
     }
   }
 
-  return 0;
+  return isPageError;
 }
 
 void ReadoutEquipment::abortThread() {

src/ReadoutEquipment.h

@@ -203,6 +203,7 @@ class ReadoutEquipment
   int saveErrorPagesCount; // counter for number of pages dumped so far
   std::string cfgDataPagesLogPath; // path to write to disk a summary for each data page received
   FILE *fpDataPagesLog = nullptr; // handle to data page log file
+  int cfgDropPagesWithError = 0; // if set, pages with RDH errors are dropped by readout
 
  protected:
   // get timeframe from orbit
@@ -218,6 +219,7 @@ class ReadoutEquipment
   unsigned long long statsRdhCheckOk = 0;        // number of RDH structs which have passed check ok
   unsigned long long statsRdhCheckErr = 0;       // number of RDH structs which have not passed check
   unsigned long long statsRdhCheckStreamErr = 0; // number of inconsistencies in RDH stream (e.g. ids/timing compared to previous RDH)
+  unsigned long long statsRdhCheckPagesDropped = 0; // number of pages discarded because of RDH errors (when configured to do so)
 };
 
 std::unique_ptr<ReadoutEquipment> getReadoutEquipmentDummy(ConfigFile& cfg, std::string cfgEntryPoint);

src/readoutConfigEditor.tcl

@@ -79,6 +79,7 @@ set configurationParametersDescriptor {
 | equipment-* | dataPagesLogPath | string |  | Path where to save a summary of each data pages generated by equipment. |
 | equipment-* | debugFirstPages | int | 0 | If set, print debug information for first (given number of) data pages readout. |
 | equipment-* | disableOutput | int | 0 | If non-zero, data generated by this equipment is discarded immediately and is not pushed to output fifo of readout thread. Used for testing. |
+| equipment-* | dropPagesWithError | int | 0 | If set, the pages with RDH errors are discarded (requires rdhCheckEnabled or rdhUseFirstInPage). |
 | equipment-* | enabled | int | 1 | Enable (value=1) or disable (value=0) the equipment. |
 | equipment-* | equipmentType | string |  | The type of equipment to be instanciated. One of: dummy, rorc, cruEmulator |
 | equipment-* | firstPageOffset | bytes | | Offset of the first page, in bytes from the beginning of the memory pool. If not set (recommended), will start at memoryPoolPageSize (one free page is kept before the first usable page for readout internal use). |