Skip to content

Multiple fixes in Alpide decoding/error handling #8586

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
shahor02 merged 1 commit into from
Apr 15, 2022
Merged

Multiple fixes in Alpide decoding/error handling #8586

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
69 changes: 39 additions & 30 deletions Detectors/ITSMFT/common/reconstruction/include/ITSMFTReconstruction/AlpideCoder.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -150,8 +150,10 @@ class AlpideCoder
uint32_t expectInp = ExpectChipHeader | ExpectChipEmpty; // data must always start with chip header or chip empty flag

chipData.clear();
LOG(debug) << "NewEntry";
while (buffer.next(dataC)) {
//
LOGP(debug, "dataC: {:#x} expect {:#b}", int(dataC), int(expectInp));
// ---------- chip info ?
uint8_t dataCM = dataC & (~MaskChipID);
//
Expand All @@ -161,7 +163,7 @@ class AlpideCoder
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::TruncatedChipEmpty);
#endif
return unexpectedEOF("CHIP_EMPTY:Timestamp");
return unexpectedEOF("CHIP_EMPTY:Timestamp"); // abandon cable data
}
chipData.resetChipID();
expectInp = ExpectChipHeader | ExpectChipEmpty;
Expand All @@ -174,7 +176,7 @@ class AlpideCoder
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::TruncatedChipHeader);
#endif
return unexpectedEOF("CHIP_HEADER");
return unexpectedEOF("CHIP_HEADER"); // abandon cable data
}
expectInp = ExpectRegion; // now expect region info
continue;
Expand All @@ -188,7 +190,6 @@ class AlpideCoder
}

if ((expectInp & ExpectChipTrailer) && dataCM == CHIPTRAILER) { // chip trailer was expected
expectInp = ExpectChipHeader | ExpectChipEmpty;
chipData.setROFlags(dataC & MaskROFlags);
#ifdef ALPIDE_DECODING_STAT
uint8_t roErr = dataC & MaskROFlags;
Expand All @@ -205,10 +206,10 @@ class AlpideCoder
}

if (!chipData.getData().size() && !chipData.isErrorSet()) {
nRightCHits = 0;
colDPrev = 0xffff;
chipData.clear();
continue;
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::TrailerAfterHeader);
#endif
return unexpectedEOF("Trailer after header"); // abandon cable data
}
break;
}
Expand All @@ -222,9 +223,11 @@ class AlpideCoder
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::TruncatedRegion);
#endif
return unexpectedEOF("CHIPDATA");
return unexpectedEOF("CHIPDATA"); // abandon cable data
}
dataS |= dataC;
LOGP(debug, "dataC: {:#x} dataS: {:#x} expect {:#b} in ExpectData", int(dataC), int(dataS), int(expectInp));

// we are decoding the pixel addres, if this is a DATALONG, we will fetch the mask later
uint16_t dColID = (dataS & MaskEncoder) >> 10;
uint16_t pixID = dataS & MaskPixID;
Expand All @@ -237,10 +240,11 @@ class AlpideCoder
// if we start new double column, transfer the hits accumulated in the right column buffer of prev. double column
if (colD != colDPrev) {
if (colD < colDPrev && colDPrev != 0xffff) {
needSorting = true;
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::WrongDColOrder);
chipData.setError(ChipStat::WrongDColOrder); // abandon cable data
#endif
return unexpectedEOF("Wrong column order"); // abandon cable data
needSorting = true; // effectively disabled
}
colDPrev++;
for (int ihr = 0; ihr < nRightCHits; ihr++) {
Expand All @@ -259,11 +263,13 @@ class AlpideCoder
uint8_t hitsPattern = 0;
if (!buffer.next(hitsPattern)) {
chipData.setError(ChipStat::TruncatedLondData);
return unexpectedEOF("CHIP_DATA_LONG:Pattern");
return unexpectedEOF("CHIP_DATA_LONG:Pattern"); // abandon cable data
}
if (hitsPattern & (~MaskHitMap)) {
return unexpectedEOF("CHIP_DATA_LONG:Pattern");
chipData.setError(ChipStat::WrongDataLongPattern);
return unexpectedEOF("CHIP_DATA_LONG:Pattern"); // abandon cable data
}
LOGP(debug, "hitsPattern: {:#b} expect {:#b}", int(hitsPattern), int(expectInp));
}
expectInp = ExpectChipTrailer | ExpectData | ExpectRegion;
continue; // end of DATA(SHORT or LONG) processing
Expand All @@ -287,13 +293,14 @@ class AlpideCoder
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::TruncatedLondData);
#endif
return unexpectedEOF("CHIP_DATA_LONG:Pattern");
return unexpectedEOF("CHIP_DATA_LONG:Pattern"); // abandon cable data
}
LOGP(debug, "hitsPattern: {:#b} expect {:#b}", int(hitsPattern), int(expectInp));
if (hitsPattern & (~MaskHitMap)) {
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::WrongDataLongPattern);
#endif
return unexpectedEOF("CHIP_DATA_LONG:Pattern");
return unexpectedEOF("CHIP_DATA_LONG:Pattern"); // abandon cable data
}
for (int ip = 0; ip < HitMapSize; ip++) {
if (hitsPattern & (0x1 << ip)) {
Expand All @@ -302,7 +309,7 @@ class AlpideCoder
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::WrongRow);
#endif
return unexpectedEOF(fmt::format("Non-existing encoder {} decoded, DataLong was {:x}", pixID, dataS));
return unexpectedEOF(fmt::format("Non-existing encoder {} decoded, DataLong was {:x}", pixID, dataS)); // abandon cable data
}
rightC = ((rowE & 0x1) ? !(addr & 0x1) : (addr & 0x1)); // true for right column / lalse for left
// the real columnt is int colE = colD + rightC;
Expand All @@ -314,11 +321,15 @@ class AlpideCoder
}
}
}
} else if (ChipStat::getAPENonCritical(dataC) >= 0) { // check for recoverable APE, if on: continue with ExpectChipTrailer | ExpectData | ExpectRegion expectation
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::DecErrors(ChipStat::getAPENonCritical(dataC)));
#endif
} else {
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::NoDataFound);
#endif
return unexpectedEOF(fmt::format("Expected DataShort or DataLong mask, got {:x}", dataS));
return unexpectedEOF(fmt::format("Expected DataShort or DataLong mask, got {:x}", dataS)); // abandon cable data
}
expectInp = ExpectChipTrailer | ExpectData | ExpectRegion;
continue; // end of DATA(SHORT or LONG) processing
Expand All @@ -338,8 +349,11 @@ class AlpideCoder
}

if (!dataC) {
buffer.clear(); // 0 padding reached (end of the cable data), no point in continuing
break;
if (expectInp == (ExpectChipHeader | ExpectChipEmpty)) {
continue;
}
chipData.setError(ChipStat::TruncatedBuffer);
return unexpectedEOF("Abandon on 0-padding"); // abandon cable data
}

// in case of BUSY VIOLATION the Trailer may come directly after the Header
Expand All @@ -358,17 +372,11 @@ class AlpideCoder
chipData.setError(ChipStat::DecErrors(codeAPE));
#endif
if (fatalAPE) {
buffer.clear(); // no point in contiunuing with this cable data
} else { // skip eventual padding
while (buffer.next(dataC)) {
if (dataC) { // padding is over, make 1 step back in the buffer
auto currPtr = buffer.getPtr();
buffer.setPtr(--currPtr);
break;
}
}
return unexpectedEOF(fmt::format("APE error {:#02x} [expectation = {:#02x}]", int(dataC), int(expectInp))); // abandon cable data
} else {
LOGP(error, "Code should not have entered here, APE: {:#02x}, expectation: {:#02x}", codeAPE, int(expectInp));
return unexpectedEOF(fmt::format("APE error {:#02x} [expectation = {:#02x}]", int(dataC), int(expectInp))); // abandon cable data
}
return unexpectedEOF(fmt::format("APE error {:#02x} [expectation = 0x{:#02x}]", int(dataC), int(expectInp)));
}
#ifdef ALPIDE_DECODING_STAT
chipData.setError(ChipStat::UnknownWord);
Expand All @@ -381,10 +389,11 @@ class AlpideCoder
std::memcpy(chipData.getRawErrBuff().data(), curPtr - offsBack, offsBack + offsAfter);
chipData.setNBytesInRawBuff(offsBack + offsAfter);
#endif
return unexpectedEOF(fmt::format("Unknown word 0x{:x} [expectation = 0x{:x}]", int(dataC), int(expectInp))); // error
return unexpectedEOF(fmt::format("Unknown word 0x{:x} [expectation = 0x{:x}]", int(dataC), int(expectInp))); // abandon cable data
}

if (needSorting && chipData.getData().size()) { // d.columns were in a wrong order, need to sort the data
if (needSorting && chipData.getData().size()) { // d.columns were in a wrong order, need to sort the data, RS: effectively disabled
LOGP(error, "This code path should have been disabled");
auto& pixData = chipData.getData();
std::sort(pixData.begin(), pixData.end(),
[](PixelData& a, PixelData& b) { return a.getCol() < b.getCol() || (a.getCol() == b.getCol() && a.getRowDirect() < b.getRowDirect()); });
Expand Down
35 changes: 31 additions & 4 deletions Detectors/ITSMFT/common/reconstruction/include/ITSMFTReconstruction/DecodingStat.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,8 +57,14 @@ struct ChipStat {
APE_FSM_ERROR, // FSM error (FATAL, SEU error, reached an unknown state)
APE_OCCUPANCY_RATE_LIMIT, // pending detector events limit (FATAL)
APE_OCCUPANCY_RATE_LIMIT_2, // pending detector events limit in packager(FATAL)
APE_LANE_PROTOCOL_ERROR, // lane protocol error
APE_RESERVED_FC, // reserved FC
APE_ERROR_NON_CRITICAL_BYTE, // Error in non critical byte
APE_OOT_NON_CRITICAL, // OOT non-critical
WrongDColOrder, // DColumns non increasing
InterleavedChipData, // Chip data interleaved on the cable
TruncatedBuffer, // truncated buffer, 0 padding
TrailerAfterHeader, // trailer seen after header w/o FE of FD set
NErrorsDefined
};

Expand Down Expand Up @@ -86,8 +92,14 @@ struct ChipStat {
"APE_FSM_ERROR", // FSM error (FATAL, SEU error, reached an unknown state)
"APE_OCCUPANCY_RATE_LIMIT", // pending detector events limit (FATAL)
"APE_OCCUPANCY_RATE_LIMIT_2", // pending detector events limit in packager(FATAL)
"APE_LANE_PROTOCOL_ERROR", // lane protocol error
"APE_RESERVED_FC", // reserved
"APE_ERROR_IN_NON_CRITICAL_BYTE", // Error in non critical byte
"APE_OOT_NON_CRITICAL", // OOT non-critical
"DColumns non-increasing", // DColumns non increasing
"Chip data interleaved on the cable" // Chip data interleaved on the cable
"Chip data interleaved on the cable", // Chip data interleaved on the cable
"TruncatedBuffer", // truncated buffer, 0 padding
"TrailerAfterHeader" // trailer seen after header w/o FE of FD set
};

static constexpr std::array<uint32_t, NErrorsDefined> ErrActions = {
Expand All @@ -114,8 +126,14 @@ struct ChipStat {
ErrActPropagate | ErrActDump, // FSM error (FATAL, SEU error, reached an unknown state)
ErrActPropagate | ErrActDump, // pending detector events limit (FATAL)
ErrActPropagate | ErrActDump, // pending detector events limit in packager(FATAL)
ErrActPropagate | ErrActDump, // lane protocol error
ErrActPropagate | ErrActDump, // reserved FC
ErrActPropagate | ErrActDump, // Error in non critical byte
ErrActPropagate | ErrActDump, // OOT non-critical
ErrActPropagate | ErrActDump, // DColumns non increasing
ErrActPropagate | ErrActDump // Chip data interleaved on the cable
ErrActPropagate | ErrActDump, // Chip data interleaved on the cable
ErrActPropagate | ErrActDump, // Truncated buffer while something was expected
ErrActPropagate | ErrActDump // trailer seen after header w/o FE of FD set
};
uint16_t feeID = -1;
size_t nHits = 0;
Expand All @@ -128,14 +146,23 @@ struct ChipStat {
memset(errorCounts.data(), 0, sizeof(uint32_t) * errorCounts.size());
nHits = 0;
}

static int getAPENonCritical(uint8_t c)
{
if (c == 0xfd || c == 0xfe) {
return APE_STRIP_START + c - 0xf2;
}
return -1;
}

// return APE DecErrors code or -1 if not APE error, set fatal flag if needd
static int getAPECode(uint8_t c, bool& ft)
{
if (c < 0xf2 || c > 0xfa) {
if (c < 0xf2 || c > 0xfe) {
ft = false;
return -1;
}
ft = c >= 0xf4;
ft = c >= 0xf2 && c <= 0xfe;
return APE_STRIP_START + c - 0xf2;
}
uint32_t getNErrors() const;
Expand Down
12 changes: 7 additions & 5 deletions Detectors/ITSMFT/common/reconstruction/include/ITSMFTReconstruction/RUDecodeData.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,21 +99,23 @@ int RUDecodeData::decodeROF(const Mapping& mp)
}
#endif
ret = -1; // discard decoded data
nhits = 0;
}
#ifdef ALPIDE_DECODING_STAT
fillChipStatistics(icab, chipData);
#endif
if (ret >= 0 && chipData->getChipID() < Mapping::getNChips()) { // make sure there was no error | upd: why? if there was a non fatal error, we use chip data
if (nhits) {
doneChips[chipData->getChipID()] = true;
ntot += nhits;
}
if (nhits && chipData->getChipID() < Mapping::getNChips()) {
doneChips[chipData->getChipID()] = true;
ntot += nhits;
if (++nChipsFired < chipsData.size()) { // fetch next free chip
chipData = &chipsData[nChipsFired];
} else {
break; // last chip decoded
}
}
if (ret < 0) {
break; // negative code was returned by decoder: abandon cable data
}
}
}

Expand Down