CDF_cal_ped.c File Reference

The CAL pedestal event processing and compaction routines. More...

#include <LSE/CDF_cal.h>
#include <LSE/CDF_cal_ped.h>
#include <CDF_cal_def.h>
#include <dprintf.h>
#include <EDS/EBF_dir.h>
#include <EDS/EBF_cid.h>
#include <EDS/EBF_cal.h>
#include <EDS/ECR_cal.h>
#include <LDT/BFP.h>
#include <LDT/APE.h>
#include <LDT/HUFF.h>
#include <LDT/HDE.h>
#include <PBI/FFS.ih>
#include <stdio.h>

Classes
struct	_HistLimits
	Min and Max indices of a histogram. More...
Defines
#define	RNGS_COMPOSE(_beg)
	Composes an array of 2 bit range values in the order specified. For example, specifying RNG_COMPOSE(0) means that one expects the four range readout to occur in the order range 0,1,2,3, while specifying RNG_COMPOSE(1) results in an expected order of range 1,2,3,0. This is used to predict the range when processing the layer pairs in the four range readout order.
#define	LYR_OFFSETS(_lyr, _rng)
	Computes the offsets to the specified layer and range when all the logs in a tower are struck.
#define	RNG_OFFSETS(_rng)
	Creates a comma separated list of the offsets to the 8 layers for the specified range when all the logs in a tower are struck.
Typedefs
typedef struct _HistLimits	HistLimits
	Typedef for struct _HistLimits.
Functions
static __inline unsigned int	cal_ped1_evaluate (const EBF_dir *dir, int *rngP)
	Evaluates the CAL data to see if it is in a form that the compression routines understand.
static __inline unsigned int	cal_ped1_evaluate_layer (const EBF_calLogData *logs, int cnt, int twr, int lyr, int rng)
	Checks that the column numbers are as expected and that all the ranges are 0.
static __inline int	cal_ped1_process (short int *cidx, unsigned short int *hist, const EBF_dir *dir, const ECR_cal *clb)
	Creates the histograms and the array of pedestal subtracted indices used to compress the CAL portion of the event that is specific to single-range readout.
static __inline const unsigned int *	cal_ped1_process_layer (short int *cidx, unsigned short int *hist, const ECR_calLayer *clyr, const unsigned int *logs, int cnt, int twr, int lyr)
	Accumulates the histograms used in compressing the CAL data for one layer.
static __inline unsigned int	cal_ped1_compress (CDF_cal *cdf, unsigned int *buf, unsigned int pos, const EBF_dir *dir, const ECR_cal *clb)
	Makes a compressed copy of the CAL data when it is single range, non-zero suppressed.
static __inline unsigned int	cal_ped4_evaluate (const EBF_dir *dir, int *rngP)
	Evaluates the 4-range CAL data to see if it is in a form that the compression routines understand.
static __inline unsigned int	cal_ped4_evaluate_layer (const EBF_calLogData *logs, int cnt, int twr, int lyr, int rng)
	Checks 4-range CAL data for column numbers are as expected all the ranges are in the same order.
static __inline unsigned int	cal_ped4_compress (CDF_cal *cdf, unsigned int *buf, unsigned int pos, const EBF_dir *dir, const ECR_cal *clb)
	Makes a compressed copy of the CAL data when it is four-range, non-zero suppressed.
static __inline int	cal_ped4_process (short int *cidx, unsigned short int *hist, const EBF_dir *dir, const ECR_cal *clb, int rng)
	Creates the histograms and the array of pedestal subtracted indices used to compress the CAL portion of the event that is specific to single-range readout.
static __inline unsigned int	cdf_cal_ped_copy (CDF_cal *cdf, unsigned int *buf, unsigned int pos, const EBF_dir *dir)
	Puts out a copy of the CAL data, prefacing it with a style header and tower map.
static unsigned int	cal_ped_hdr_encode (unsigned int *buf, unsigned int pos, unsigned int fmt, unsigned int map)
	Encodes the CAL pedestal header information. This consists of a formatting style (currently 0 = copy, 1 = compress) plus the map of towers that have a CAL contribution.
unsigned int	CDF_cal_ped_compress (CDF_cal *cdf, unsigned int *buf, unsigned int pos, const EBF_dir *dir, const ECR_cal *clb, const CDF_err *err)
	Makes a compressed copy of the CAL data if the data is in a form that this routine understands. If not, a straight copy of the data is made.

---

Detailed Description

The CAL pedestal event processing and compaction routines.

Author:

JJRussell - russell@slac.stanford.edu

   CVS $Id: CDF_cal_ped.c,v 1.6 2011/03/26 17:35:23 russell Exp $

The encoding for the CAL pedestal data is

Object Number of Bits Description Format 2 0 = copy, 1 = standard compress

Format 0

The format style is followed by 16 bits indicating which towers have CAL data, followed by a verbatim copy of the CAL data which begins on the next 32-bit boundary.

Format 1

The format style is followed by

1. A single bit indicating whether anything unusual is going on. If this is 0, then all logs of the LAT are present. The exception field is yet to be defined. This provides an escape hatch if needed

Each range is then encoded as

1. 13 bits indicating the offset to the 0 bin of the pedestal subtracted values. This value should be subtracted from each recovered pedestal subtracted value after adding on that logs pedestal, ie

                   ADC = encoded_value - this_offset + pedestal
   

2. 4 bits giving the number of bits needed to store the run lengths of the valid codes
3. A list of <nbits> values giving the run lengths of the valid codes. This is terminated by a 0. For example, if the frequency distribution was

1 2 0 3 4 5 0 0 7 6 5

The longest run length is 3. This takes 2 bits to hold so this distribution would be encoded 10b, 01b, 11b, 10b, 11b, 00b.

1. 5 bits giving the number of bits used to encode the code lengths - 1 (since all valid codes must be at least 1 in length, the 1 is subtracted off).
2. The list of the code lengths -1, each occupying the number of bits specified by the previous field
3. The Huffman codes for the logs.

---

Define Documentation

#define LYR_OFFSETS	(	_lyr,
		_rng		)

Value:

(((_lyr) >>   1) * EDA_CAL_K_LOGS_PER_TOWER +   \
                  ((_lyr) &  0x1) * EDA_CAL_K_LOGS_PER_LAYER +   \
                   (_rng)         * EDA_CAL_K_LOGS_PER_LAYER * 2)

Computes the offsets to the specified layer and range when all the logs in a tower are struck.

Returns:

The offset to the layer of the specified range

Parameters:

	_lyr	The layer number
	_rng	The range

#define RNG_OFFSETS

(

_rng

)

Value:

LYR_OFFSETS(0, _rng), \
        LYR_OFFSETS(1, _rng), \
        LYR_OFFSETS(2, _rng), \
        LYR_OFFSETS(3, _rng), \
        LYR_OFFSETS(4, _rng), \
        LYR_OFFSETS(5, _rng), \
        LYR_OFFSETS(6, _rng), \
        LYR_OFFSETS(7, _rng)

Creates a comma separated list of the offsets to the 8 layers for the specified range when all the logs in a tower are struck.

Returns:

The comma separated list, appropriate to use when defining a data statement of these offsets

Parameters:

_rng

The range to compute the offsets for

Referenced by cal_ped4_process().

#define RNGS_COMPOSE

(

_beg

)

Value:

( ((((_beg) + 3) & 3) << 14) | ((((_beg) + 3) & 3) << 12) | \
          ((((_beg) + 2) & 3) << 10) | ((((_beg) + 2) & 3) <<  8) | \
          ((((_beg) + 1) & 3) <<  6) | ((((_beg) + 1) & 3) <<  4) | \
          ((((_beg) + 0) & 3) <<  2) | ((((_beg) + 0) & 3) <<  0) )

Composes an array of 2 bit range values in the order specified. For example, specifying RNG_COMPOSE(0) means that one expects the four range readout to occur in the order range 0,1,2,3, while specifying RNG_COMPOSE(1) results in an expected order of range 1,2,3,0. This is used to predict the range when processing the layer pairs in the four range readout order.

Referenced by cal_ped4_evaluate().

---

Function Documentation

static __inline unsigned int cal_ped1_compress	(	CDF_cal *	cdf,
		unsigned int *	buf,
		unsigned int	pos,
		const EBF_dir *	dir,
		const ECR_cal *	clb
	)			[static]

Makes a compressed copy of the CAL data when it is single range, non-zero suppressed.

Returns:

The updated write bit position

Parameters:

	cdf	The CAL compression data formatting handle
	buf	The output buffer
	pos	The position of the next bit to fill
	dir	The directory describing the event to be formatted
	clb	The CAL pedestals and gains

References cal_ped1_process(), _CDF_cal_ped::cidx, _CDF_cal::com, _CDF_cal_com::hde, _CDF_cal_ped_statistics::hist, _CDF_cal::ped, and _CDF_cal_ped::stats.

Referenced by CDF_cal_ped_compress().

static __inline unsigned int cal_ped1_evaluate	(	const EBF_dir *	dir,
		int *	rngP
	)			[static]

Evaluates the CAL data to see if it is in a form that the compression routines understand.

Returns:

A bit mask of the towers that are not understandable

Parameters:

	dir	The directory of the event to be evaluated
	rngP	Returned as the range for the event.

References cal_ped1_evaluate_layer().

Referenced by CDF_cal_ped_compress().

static __inline unsigned int cal_ped1_evaluate_layer	(	const EBF_calLogData *	logs,
		int	cnt,
		int	twr,
		int	lyr,
		int	rng
	)			[static]

Checks that the column numbers are as expected and that all the ranges are 0.

Returns:

The tower number as a bit mask

Parameters:

	logs	The logs for this layer
	cnt	The number of logs for this layer (likely 12)
	twr	The tower number for this layer
	lyr	The layer number for this layer (for debugging only)
	rng	The 'standard' range. Anything not matching this is considered an exception.

Referenced by cal_ped1_evaluate().

static __inline int cal_ped1_process	(	short int *	cidx,
		unsigned short int *	hist,
		const EBF_dir *	dir,
		const ECR_cal *	clb
	)			[static]

Creates the histograms and the array of pedestal subtracted indices used to compress the CAL portion of the event that is specific to single-range readout.

Returns:

The number of adcs

Parameters:

	cidx	The pedestal subtracted ADCs
	hist	The histogram to fill
	dir	The directory for this event
	clb	The calibration constants

References cal_ped1_process_layer().

Referenced by cal_ped1_compress().

static __inline const unsigned int * cal_ped1_process_layer	(	short int *	cidx,
		unsigned short int *	hist,
		const ECR_calLayer *	clyr,
		const unsigned int *	logs,
		int	cnt,
		int	twr,
		int	lyr
	)			[static]

Accumulates the histograms used in compressing the CAL data for one layer.

Returns:

Pointer to the next CAL log data

Parameters:

	cidx	The pedestal subtracted ADCs
	hist	The histogram to fill
	clyr	The calibration constants for this layer (holds the pedestals)
	logs	The array of CAL logs for this layer
	cnt	The number of logs in this layer
	twr	The tower number of this layer, debugging only
	lyr	The layer number of this layer, debugging only

Referenced by cal_ped1_process(), and cal_ped4_process().

static __inline unsigned int cal_ped4_compress	(	CDF_cal *	cdf,
		unsigned int *	buf,
		unsigned int	pos,
		const EBF_dir *	dir,
		const ECR_cal *	clb
	)			[static]

Makes a compressed copy of the CAL data when it is four-range, non-zero suppressed.

Returns:

The updated write bit position

Parameters:

	cdf	The CAL compression data formatting handle
	buf	The output buffer
	pos	The position of the next bit to fill
	dir	The directory describing the event to be formatted
	clb	The CAL pedestals and gains

References cal_ped4_process(), _CDF_cal_ped::cidx, _CDF_cal::com, _CDF_cal_com::hde, _CDF_cal_ped_statistics::hist, _CDF_cal::ped, and _CDF_cal_ped::stats.

Referenced by CDF_cal_ped_compress().

static __inline unsigned int cal_ped4_evaluate	(	const EBF_dir *	dir,
		int *	rngP
	)			[static]

Evaluates the 4-range CAL data to see if it is in a form that the compression routines understand.

Returns:

A bit mask of the towers that are not understandable

Parameters:

	dir	The directory of the event to be evaluated
	rngP	Returned as the range for the event.

References cal_ped4_evaluate_layer(), and RNGS_COMPOSE.

Referenced by CDF_cal_ped_compress().

static __inline unsigned int cal_ped4_evaluate_layer	(	const EBF_calLogData *	logs,
		int	cnt,
		int	twr,
		int	lyr,
		int	rng
	)			[static]

Checks 4-range CAL data for

• column numbers are as expected
• all the ranges are in the same order.

Returns:

If in error, tower number as a bit mask, else 0

Parameters:

	logs	The logs for this layer
	cnt	The number of logs for this layer (likely 12)
	twr	The tower number for this layer
	lyr	The layer number for this layer (for debugging only)
	rng	The expected range

Referenced by cal_ped4_evaluate().

static __inline int cal_ped4_process	(	short int *	cidx,
		unsigned short int *	hist,
		const EBF_dir *	dir,
		const ECR_cal *	clb,
		int	rng
	)			[static]

Creates the histograms and the array of pedestal subtracted indices used to compress the CAL portion of the event that is specific to single-range readout.

Returns:

The number of adcs

Parameters:

	cidx	The pedestal subtracted ADCs
	hist	The histogram to fill
	dir	The directory for this event
	clb	The calibration constants
	rng	The range to process

const unsigned short int off_buf[8];

References cal_ped1_process_layer(), and RNG_OFFSETS.

Referenced by cal_ped4_compress().

static unsigned int cal_ped_hdr_encode	(	unsigned int *	buf,
		unsigned int	pos,
		unsigned int	fmt,
		unsigned int	map
	)			[static]

Encodes the CAL pedestal header information. This consists of a formatting style (currently 0 = copy, 1 = compress) plus the map of towers that have a CAL contribution.

Returns:

The updated write bit position

Parameters:

	buf	The output buffer
	pos	The position of the next bit to fill
	fmt	The format style
	map	The map of towers with CAL data

Referenced by cdf_cal_ped_copy().

unsigned int CDF_cal_ped_compress	(	CDF_cal *	cdf,
		unsigned int *	buf,
		unsigned int	pos,
		const EBF_dir *	dir,
		const ECR_cal *	clb,
		const CDF_err *	err
	)

Makes a compressed copy of the CAL data if the data is in a form that this routine understands. If not, a straight copy of the data is made.

Returns:

The updated write bit position

Parameters:

	cdf	The CAL compression data formatting handle
	buf	The output buffer
	pos	The position of the next bit to fill
	dir	The directory describing the event to be formatted
	clb	The CAL pedestals and gains
	err	The error context block

This routine is merely a dispatch routine, vectoring control to either single range or four range compression.

References cal_ped1_compress(), cal_ped1_evaluate(), cal_ped4_compress(), cal_ped4_evaluate(), and cdf_cal_ped_copy().

static __inline unsigned int cdf_cal_ped_copy	(	CDF_cal *	cdf,
		unsigned int *	buf,
		unsigned int	pos,
		const EBF_dir *	dir
	)			[static]

Puts out a copy of the CAL data, prefacing it with a style header and tower map.

Returns:

The updated write bit position

Parameters:

	cdf	The CAL compression data formatting handle
	buf	The output buffer
	pos	The position of the next bit to fill
	dir	The directory describing the event to be formatted

This routine is called when the CAL data does not meet the expectations of the fancier compression encoding. This is either due to a variation that has not be specifically handled or due to an error in the data itself.

References cal_ped_hdr_encode(), and CDF_cal_copy().

Referenced by CDF_cal_ped_compress().