GLAST / LAT > DAQ and FSW > FSW > Doxygen Index> QSE / V2-3-2 > qse / sun-gcc
#include <QSE_aem.h>
Public Attributes | |
| unsigned short int | rsvd [4] |
| unsigned short int | timeout |
| unsigned short int | hdr_parity |
| unsigned short int | pha_overrun |
| unsigned short int | pha_underrun |
| unsigned short int | present |
| unsigned short int | pha_parity |
| unsigned short int _QSE_aem_cableStatus::hdr_parity |
Bit list of the cables header parity bits. This may or may not be a fatal error, preventing the navigation of the event. The design decision was to account for the condition in this word, but not abort the unpacking procedure. The downside of this is that if the parity error corrupts the control information sufficiently, it is very likely that this will result in some other error. On the other hand if the parity error is the result of corrupting one of the accept or hit bits, the data is likely still parsable, and, in this case, comes down to a case of data quality.
Referenced by QDF_acdEvtConstruct(), QSE_aemRepack(), and QSE_aemUnpackDat().
| unsigned short int _QSE_aem_cableStatus::pha_overrun |
Bit list of the cables with unmatched ADCs. The term overrun refers to the fact that, for a given cable, there are more accept bits than there are PHAs, as in the control has overrun the supply. At one point in the AEM design, this was an allowed state. The idea was that, for deadtime control, one could limit the number PHA values, but still record the fact that a channel wished to emit a PHA value, but was inhibited by this limit. However, the hardware was not designed to this specification.
Referenced by QDF_acdEvtConstruct(), and QSE_aemUnpackDat().
| unsigned short int _QSE_aem_cableStatus::pha_parity |
Bit list of the cables with ADCs having parity errors. This is not a fatal error, rather falling into a question of data quality.
Referenced by QDF_acdEvtConstruct(), and QSE_aemUnpackDat().
| unsigned short int _QSE_aem_cableStatus::pha_underrun |
Cables with underrun phas. The term underrun refers to the fact that, for a given cable, the number of accepts bits is fewer than the number of PHAs actually emitted. This is an error condition and, while the face value interpretation may be correct, the more likely scenario is that AEM data stream is corrupted and the unpacker has lost synch with what it is actually seeing and what it is reading.
Referenced by QDF_acdEvtConstruct(), and QSE_aemUnpackDat().
| unsigned short int _QSE_aem_cableStatus::present |
Cables present in the input stream. This mask could be used in an integrity check, checking this mask against the anticipated set of cables. For example, in a system with all 12 free boards, this mask should be 0xfff. Note that this mask says nothing about whether the cables have data or not.
Referenced by acd_std_unpack(), QSE_aemRepack(), and QSE_aemUnpackDat().
| unsigned short int _QSE_aem_cableStatus::rsvd[4] |
Reserved for future use
Referenced by QDF_acdEvtConstruct().
| unsigned short int _QSE_aem_cableStatus::timeout |
Bit list of the cables with a timeout. A timeout is equivalent to a cable without a start bit. This is generally not considered to be a fatal error in the sense that this should cause no problem in navigating the event. The data has integrity, the issue is data quality, an issue the user should be concerned with, not the unpacker
Referenced by QDF_acdEvtConstruct(), QSE_aemRepack(), and QSE_aemUnpackDat().
1.5.8