detectors/slsDetectorPackage
24
1
Fork 0
mirror of https://github.com/slsdetectorgroup/slsDetectorPackage.git synced 2026-01-16 07:38:04 +01:00
Code Issues Actions Packages Releases Activity

ctb receiver fix: extract in 8 bit batches instead of 32 for less padding

Browse Source
This commit is contained in:
maliakal_d
2019-05-27 11:30:54 +02:00
parent e2ef043e04
commit 7fdab08256
1 changed files with 12 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
slsReceiverSoftware/src/DataProcessor.cpp
View File

@@ -515,18 +515,18 @@ void DataProcessor::RearrangeDbitData(char* buf) {
const int numSamples = (ctbDigitalDataBytes / sizeof(uint64_t)); const int numSamples = (ctbDigitalDataBytes / sizeof(uint64_t));
const int digOffset = FIFO_HEADER_NUMBYTES + sizeof(sls_receiver_header) + (*ctbAnalogDataBytes); const int digOffset = FIFO_HEADER_NUMBYTES + sizeof(sls_receiver_header) + (*ctbAnalogDataBytes);
// ceil as numResult32Bits could be decimal // ceil as numResult8Bits could be decimal
const int numResult32Bits = ceil((double)(numSamples * (*ctbDbitList).size()) / 32.00); const int numResult8Bits = ceil((double)(numSamples * (*ctbDbitList).size()) / 8.00);
uint32_t result[numResult32Bits]; uint8_t result[numResult8Bits];
memset(result, 0, numResult32Bits * sizeof(uint32_t)); memset(result, 0, numResult8Bits * sizeof(uint8_t));
uint8_t* dest = result;
uint32_t* dest = result;
auto* source = (uint64_t*)(buf + digOffset + (*ctbDbitOffset)); auto* source = (uint64_t*)(buf + digOffset + (*ctbDbitOffset));
// loop through digital bit enable vector // loop through digital bit enable vector
int bitoffset = 0; int bitoffset = 0;
for (auto bi : (*ctbDbitList)) { for (auto bi : (*ctbDbitList)) {
// where numbits * numsamples is not a multiple of 32 // where numbits * numsamples is not a multiple of 8
if (bitoffset != 0) { if (bitoffset != 0) {
bitoffset = 0; bitoffset = 0;
++dest; ++dest;
@@ -534,12 +534,12 @@ void DataProcessor::RearrangeDbitData(char* buf) {
// loop through the frame digital data // loop through the frame digital data
for (auto ptr = source; ptr < (source + numSamples);) { for (auto ptr = source; ptr < (source + numSamples);) {
// get selected bit from each 32 bit // get selected bit from each 8 bit
uint32_t bit = (*ptr++ >> bi) & 1; uint8_t bit = (*ptr++ >> bi) & 1;
*dest |= bit << bitoffset; *dest |= bit << bitoffset;
++bitoffset; ++bitoffset;
// extract destination in 32 bit batches // extract destination in 8 bit batches
if (bitoffset == 32) { if (bitoffset == 8) {
bitoffset = 0; bitoffset = 0;
++dest; ++dest;
} }
@@ -548,8 +548,8 @@ void DataProcessor::RearrangeDbitData(char* buf) {
// copy back to buf and update size // copy back to buf and update size
memcpy(buf + digOffset, result, numResult32Bits * sizeof(uint32_t)); memcpy(buf + digOffset, result, numResult8Bits * sizeof(uint8_t));
(*((uint32_t*)buf)) = numResult32Bits * sizeof(uint32_t); (*((uint32_t*)buf)) = numResult8Bits * sizeof(uint8_t);
} }
/** eiger specific */ /** eiger specific */