- last edited 07.10.02 -

11 Parameters of DVM-system

Base set of parameters, controlling Run-Time System operating, is provided with DVM-system and located in dvm_sys/par subdirectory. File names reflect purpose of the parameters, placed in the files:

File name extensions reflects a way of applying given parameter subset:

The correcting parameter sets are the files with the names:

A user chooses the base set, more suitable for his purposes by specifying corresponding extension in environment variable dvmpar. He can also have own parameter sets (in any his working directory and with any file names and extensions), correcting the base set. The user specifies in these files only the parameters (contained in any of files with .rel or .deb extensions), that he wishes to correct. These directories and file names must be listed in environment variable usrpar of the dvm-command startup file (see section 7).

All parameters of the base set have the commentaries. Full lists of parameters of dynamic control and the user trace are presented in sections 11.1 and 11.2. Parameters of Run-Time System information messages output and main parameters, controlling system trace (as more frequently used when debugging a program) are considered in sections 11.3 and 11.4.

11.1 The parameters of dynamic control

The parameters of dynamic control, listed below, are set in files with name usrdebug.

• EnableDynControl - flag of enabled dynamic control.
• DebugOptions.ErrorToScreen - if this parameter is not equal to 0 all dynamic control diagnostic messages will be outputted to the screen.
• DebugOptions.ErrorFile - the name of error diagnostic file. Error messages, found by dynamic control, are written to this file if parameter DebugOptions.ErrorToScreen is equal to 0. In the case of file opening failure the diagnostics will be outputted to the screen.
• DebugOptions.AppendErrorFile - the mode of a diagnostic file usage. If this parameter is equal to 0, the file will be updated every time when the program is started.
• DebugOptions.MaxErrors - the maximum count of errors after that the dynamic control will be disabled.
• DebugOptions.PrintStatistic - if flag is not equal to 0, then after program termination the statistics of dealing with variable table is printed. The statistics is written into the file with name, specified by DebugOptions.ErrorFile parameter or to the screen dependently on value of DebugOptions.ErrorToScreen parameter.
• DebugOptions.CheckVarInitialization - if flag is equal to 0, then checking scalar variables initialization is disabled.
• DebugOptions.CheckVarReadOnly - if flag is equal to 0, then checking usage of read-only variables is disabled.
• DebugOptions.CheckReductionAccess - if flag is equal to 0, checking usage of reduction variables is disabled.
• DebugOptions.CheckDisArrInitialization - if flag is equal to 0, checking distributed array elements initialization is disabled.
• DebugOptions.CheckDisArrLimits - if flag is equal to 0, checking violation of distributed array bounds is disabled.
• DebugOptions.CheckDisArrEdgeExchange - if flag is equal to 0, checking exchange of shadow edges of distributed arrays is disabled.
• DebugOptions.CheckDisArrSequentialAccess - if flag is equal to 0, checking usage of distributed array elements in the sequential program part is disabled.
• DebugOptions.CheckDisArrLocalElm - if flag is equal to 0, checking membership of an element to the distributed array local part is disabled.
• DebugOptions.CheckDataDependence - if flag is equal to 0, checking undeclared loop data dependence is disabled.
• DebugOptions.CheckRemoteBufferAccess - if flag is equal to 0, checking usage of remote access buffers is disabled.
• HashMethod - the parameter defines algorithm of the hash-value calculation. If it is equal to 0, StandartHashCalc algorithm is used; if it is equal to 1, OffsetHashCalc algorithm is used.
• HashOffsetValue - offset value for the OffsetHashCalc algorithm. The values from 0 up to 16 are allowed.
• DebugOptions.HashIndexSize - size of hash-table index for the variable table.
• DebugOptions.HashTableSize - size of hash-table arrays.
• DebugOptions.VarTableSize - size of variable tables.
• DynDebugMemoryLimit – size of additional memory (in bytes); dynamic debugger stops its work if all the memory is exhausted. But the program execution is not stopped. If the value is 0, the memory size is not limitted.
• DynDebugExecutionTimeLimit – time of dynamic debugger operating (in seconds). When the time is over, dynamic debugger stops. But the program execution is not stopped. If the value is 0, the time is not limitted.

11.2 Parameters of trace accumulation and comparison

The parameters, listed below, are set in files with the name usrdebug.

• EnableTrace - flag of enabled execution tracing.
• DelUsrTrace – flag to delete "old" (kept from previous startup) files containing calculation trace and detected errors. Deleting is performed only in the case if EnableDynControl and EnableTrace parameters are equal to zero.
• TraceOptions.TraceMode - specifying trace mode:

• TraceOptions.TraceLevel - trace accumulation level. It is relevant only in modes of trace accumulation and trace configuration file generation:

• ManualReductCalc - reduction computation mode. If flag is equal to 0, then the calculation mode is usual, and emulation of computation of every iteration of the loop at its own processor is performed in other case.
• TraceOptions.WriteEmptyIter - if flag is equal to 0 then the records, corresponding to iteration beginning, are excluded from trace, if inside the iteration there are no the variables accesses or nested loops.
• TraceOptions.TracePath - path for the trace files.
• TraceOptions.TraceFile - the name of the reference trace file. This file is looked for in the TraceOptions.TracePath directory. It is used only in the trace comparison mode.
• TraceOptions.Ext - extension of trace accumulated files. The name of a trace file is formed as <processor number>.<TraceOptions.Ext>. The trace file will be created in the TraceOptions.TracePath directory. It is used only in the trace accumulation mode.
• TraceOptions.FileLoopInfo - trace configuration file name. This file is searched in the TraceOptions.TracePath directory.
• TraceOptions.ErrorToScreen - if flag is not equal to 0 then all diagnostic messages will be outputted to the screen.
• TraceOptions.ErrorFile - the name of error diagnostic file. The file is used if parameter TraceOptions.ErrorToScreen is equal to 0. In the case of file opening failure, the error messages will be outputted to the screen.
• TraceOptions.AppendErrorFile - the mode of a diagnostic file usage. If the flag is equal to 0, the file will be cleared every time, when the program is started.
• TraceOptions.MaxErrors - the maximal count of errors after that the trace accumulation or comparison will be disabled.
• TraceOptions.PrintStatistic - output statistics after program termination. The statistics is outputted to the file TraseOptions.ErrorFile or to the screen dependently on value of TraceOptions.ErrorToScreen parameter.
• TraceOptions.Exp - accuracy of a float number comparison in the trace comparison mode.
• TraceOptions.WrtHeaderProc - the number of a processor that creates trace configuration file. Since the common structure of the program is the same for all the processors, then trace configuration file will be identical for any process. The only distinction is the trace size, because these values are calculated only for the current process. The summary trace size can be obtained by launching the program on a single processor.
• TraceOptions.SpaceIndent - the indent size used in the trace files
• TraceOptions.TableTraceSize - size of arrays for tracing.
• TraceOptions.HashIterIndex,
  TraceOptions.HashIterSize - hash-table parameters, used for searching iterations.
• TraceOptions.ReductHashIndexSize,
  TraceOptions.ReductHashTableSize,
  TraceOptions.ReductVarTableSize - the parameters of the variable-table, used for keeping reduction variables.
• TraceOptions.SaveThroughExec - if the parameter is not equal to 0, the trace will be written into the file in the process of program execution (there is no accumulating in the memory). Otherwise the trace will be written to the file after program termination.
• DynDebugMemoryLimit – size of additional memory (in bytes); dynamic debugger stops its work if all the memory is exhausted. But the program execution is not stopped. If the value is 0, the memory size is not limitted.
• DynDebugExecutionTimeLimit – time of dynamic debugger operating (in seconds). When the time is over, dynamic debugger stops. But the program execution is not stopped. If the value is 0, the time is not limitted.

11.3 Parameters of standard data streams redirection and Run-Time System informational messages output control

The listed below parameters are set in files with syspar, statist, systrace, trcevent, sysdebug and usrdebug names. A parameter is set in files with name syspar by default.

• StdOutToFile – flag to redirect standard output stream stdout into the file (files).
• StdOutFileName –parameter defining redirection mode and a name (names) of a file (files) for redirection of stdout stream. The parameter has the form:
        [<regim>][<path>[.<extension>]] .
  If StdOutToFile parameter is zero, stdout stream is redirected to the file (files) with the name (names)
        [<path>][<processor number>].<extension> .
  If <extension> option is not specified, the extension of the file (files) is 'sto'.
  The <regim> option defines a way to redirect stdout stream and can be "*" or "+". In the first case the redirection is performed at all the processors to its own file for every processor with the name above. In the second case the redirection is also performed at all the processors but to the common for all the processors file with the name
        <path>.<extension> .
  At that if <path> option (together with <extension> option) is not specified, the redirection is performed to file with "stdout" name (without extension).
  If <regim> option is not specified, the redirection of stdout stream is performed only at the processor with zero number to the file <path>.<extension> (or to "stdout" file, if <path> option is not specified).
• StdErrToFile – flag to redirect standard error output stream stderr to the file (files).
• StdErrFileName – parameter defining redirection mode and a name (names) of a file (files) for redirection of stderr stream. The parameter has the form:
        [<regim>][<path>[.<extension>]] .
  If StdErrToFile parameter is zero, stderr stream is redirected to the file (files) with the name (names)
        [<path>][<processor number>].<extension> .
  The <regim> option defines a way to redirect stderr stream and can be "*" or "+". In the first case the redirection is performed at all the processors to its own file for every proceccor with the name above. In the second case the redirection is also performed at all the processors but into the file common for all the processors file with the name
        <path>.<extension> .
  At that if <path> option (together with <extension> option) is not specified, the redirection is performed to the file with "stderr" name (without extension).
  If <regim> option is not specified, the redirection of stderr stream is performed only at the processor with zero number to the file <path>.<extension> (or to "stderr" file, if <path> option is not specified).
• DelStdStream – flag to delete "old" (kept from previous startup) files containing output streams.
• SysInfoPrint - common flag to enable output of Run-Time System informational messages.
• SysInfoStdOut - flag to output informational messages to stdout stream.
• SysInfoStdErr - flag to output informational messages to stderr stream.
• SysInfoFile - flag to output informational messages into a file.
• SysInfoFileName - file name for informational messages.
• SysInfoSepFile - flag to accumulate informational messages in separate for each processor temporary file (if SysInfoSepFile=1) with the name <internal processor number>.out. Before program termination all temporary files will be joined into single target file and then deleted. The target file name is specified by SysInfoFileName parameter. If SysInfoSepFile=0 each processor outputs informational messages directly to the file, specified by SysInfoFileName parameter.
• FatInfoNoOpen - flag to terminate the work in the case of opening failure of the file for the informational messages. If the flag is equal to zero, Run-Time System outputs warning message to the error stream and continues the program execution without output of informational messages into the file.
• DelSysInfo – flag to delete "old" (kept from previous startup) file containing informational messages.
• VersStartPrint - flag to output Run-Time System version number before starting the program.
• VersFullStartPrint - flag to output Run-Time System version number and version numbers of all parameter files before starting the program.
• VersFinishPrint - flag to output Run-Time System version number after the program termination.
• VersFullFinishPrint - flag to output Run-Time System version number and version numbers of all parameter files after the program termination.
• SysParPrint - flag to output initialization parameters of Run-Time System before starting the program.
• ProcListPrint - flag to output processor matrix before starting the program. Internal and external numbers and identifier are outputted for each processor.
• WeightListPrint - flag to output coordinate processor weights before starting the program.
• ParamRunPrint - flag to output information about initial parameters and starting modes of Run-Time System before starting the program (standard or non-standard startup, base message passing system type and so on).
• EndProgMemoryPrint – flag to output information about non-freed memory when terminating work.
• EndProgObjectPrint – flag to output information about undeleted DVM-objects when terminating work.
• EndProgCheckSumPrint – flag to output information about state of checked memory areas when terminating work.
• InputParPrint - flag to output messages about beginning input of each parameter file.
• EndReadParPrint - flag to output messages about completion of the input of each parameter file.
• TimeExpendPrint - mode to output time spent by the program. If the parameter is equal to zero, time consumption information is not outputted. The parameter, equal to 1, specifies brief output mode of time consumption information. The parameter with value, more than 1 (2,3,4,5) specifies detailed mode. The parameter is set in files with name statist.
• CallCountPrint - flag to output statistics of Run-Time System functions calls, acceptable for the user program (for each function the count of function calls is reported). The parameter is set in the files with name statist. If CallCountPrint value is equal to one the information about functions, invoked at least one time, is output only and if the value is equal to 2, the information about all functions is output.
• TraceClosePrint - flag to output messages about completion of uploading each buffer with accumulated system trace into a file (when tracing in memory buffer). The parameter is set in the files with name systrace.
• DbgInfoPrint - flag to output informational messages of dynamic control subsystem and execution trace. The parameter is set in files with name usrdebug.

11.4 Parameters controlling system tracing

Controlling trace mode and trace event set is implemented by the parameters, kept in files systrace.* and trcevent.*. Let consider main from them. A parameter is kept in files systrace.* by default.

11.4.1 Enabling and disabling tracing

• Is_DVM_TRACE - non-zero value of the parameter is common flag of enabled tracing.
• Is_ALL_TRACE - flag to trace all functions of Run-Time System (except of functions invoking DVM-debugger and information accumulation system).
• Is_DEB_TRACE - flag to trace functions invoking DVM-debugger.
• Is_STAT_TRACE - flag to trace function invoking information accumulation system.
• TraceProcList - list of internal numbers of processors with enabled tracing. When flag of common trace is on, tracing is performed only on the processors from the list. At the main processor the trace information is accumulated always independently of presence of its number (zero) in TraceProcList list.
• IsTraceProcList - flag of enabled TraceProcList list. If the parameter has zero value trace is accumulated at all the processors.
• IsEvent[<event number>] - the parameter value, equal to1 or 2, is the flag of enabled tracing of the event with the number <event number>. The value 2 specifies trace of given event in extended mode independently from common mode of extended trace. Zero value of the parameter disables tracing of specified event. The parameter is set in trcevent.* files. Usually (i.e. in base sets of parameters) tracing of all events is enabled and tracing of the events with 0, 1, 2 and 3 numbers is enabled in verbose mode.
• DisableTraceEvents - list of enabled for tracing event numbers (via comma). DisableTraceEvents list disables tracing of listed events, that was enabled by IsEvent parameter. The flag of DisableTraceEvents list end is -1. Tracing of the 0, 1, 2 and 3 events can't be disabled.
• IsDisableTraceEvents - flag of enabled DisableTraceEvents list. If the parameter is zero disabling of event tracing is not performed.
• BlockTrace - flag of postponed trace enabling. When common tracing is enabled the accumulation is started only by special instruction from the function void tron(void). Enabled tracing can be disabled by the function void troff(void) (and then can be again enabled and so on).

11.4.2 Specifying opened (enabled) trace streams

• TracePath - directory name for trace files (as when tracing in the buffer, as when tracing directly in the file). If directory name is not empty, the last symbol of the name must be "\" or "/".
• ScreenTrace - if the parameter is not equal to zero, the trace is outputted to information message stream (flag to output trace to the "screen"). Trace, outputted to informational message stream, doesn't contain informational messages and messages about errors of Run-Time System.
• BufferTrace - flag to accumulate trace in the memory buffer.
• TraceBufLength - trace buffer size in bytes.
• TraceBufferExt - trace file extension when tracing in the buffer.
• FileTrace - flag to accumulate trace directly in the files.
• FatTraceNoOpen - flag to terminate work in the case of opening failure of the trace file for direct trace accumulation. If the parameter is equal to zero, warning is outputted in the case of the file opening failure.
• TraceFileExt - trace file extension when tracing directly into the files.

11.4.3 Specifying trace modes

• DelSysTrace - flag to delete "old" trace files (remained from previous startup) when initializing Run-Time System.
• FullTrace - common flag of extended trace mode.
• FullTraceEvents - list of event numbers (via comma), traced in verbose mode (independently from common flag of verbose mode, specified by FullTrace parameter). The flag of FullTraceEvents list end is -1. FullTraceEvents list changes trace mode of listed events, but doesn't enable their tracing, disabled before.
• IsFullTraceEvents - flag of enabled FullTraceEvents list. If the parameter is zero trace mode of the listed in FullTraceEvents list events is not changed.
• MaxTraceLevel - common maximal traced enclosure level of function calls. Zero value of the parameter means tracing the functions, directly called by the user program.
• MaxEventLevel[<event number>] - maximal traced enclosure level of function calls for the event with the number <event number>. This parameter has priority under the parameter MaxTraceLevel. The parameter is set in trcevent.* files.
• FullBufferStop - flag to stop tracing if buffer is overflowed. If the parameter is equal to zero, then last information is written instead of most old information if the buffer is overflowed.
• BufferTraceUnLoad - flag to upload trace, accumulated in trace buffers, into the files before program termination. If flag is equal to 0, Run-Time System does not unload trace, accumulated in trace buffers, into the files, entrusting this work to the dump-program, started if it is required.
• LowDumpLevel - flag to upload trace, accumulated in the buffers, into the files, using low level input/output functions. Uploading trace buffers by high-level input/output functions is performed more slowly.
• SetTraceBuf - flag to trace directly into the files with buffering, performed by operation system. Trace mode with buffering is set at once after opening trace stream by the function setbuf.
• TraceFlush - flag to upload trace information from operation system buffer into the file after each event (when tracing directly into the files). The mode, specified by this parameter, is active only in the case of tracing with buffering (the parameter SetTraceBuf is not equal to 0).

11.4.4 Controlling form of accumulated information

• KeyWordName - flag to output "NAME=" before name of each event.
• PreUnderLine - flag to underline heading of each event (event heading is a trace line, containing event name, time, file name and line number).
• PreUnderLining - underlining string of heading of each event.
• PostUnderLine - flag to underline end line of output information of each event.
• PostUnderLining - underlining string of output information for each event.
• BufferTraceShift - coefficient of indentation of output information when tracing in buffer. The number of blanks, equal to product of this parameter by the current enclosure level of traced function calls will be inserted prior each output line.
• FileTraceShift - coefficient of indentation of output information when tracing directly into the files.
• CurrentTimeTrace - flag to output current system time after the time, passed from last trace event.
• TimePrecision - a number of digits after dot when printing time, passed from the previous trace event, or when printing current system time.

11.4.5 Controlling internal self-checking Run-Time System mechanisms, functioning during trace accumulation

In the process of trace information accumulation Run-Time System can check certain memory areas. In the case of such area modification it terminates the program and outputs corresponding error message. Such areas are:

• left (low) and right (high) control parts of each dynamically allocated memory block;
• memory area, explicitly specified by their initial and final addresses in parameter files systrace.*;
• memory area, containing code of the user program and Run-Time System (code memory).

The control parts of dynamically allocated memory blocks are created by special instruction in parameter files sysdebug.* and are filled by the code, also specified in these files. Their contents can be checked each time, when memory is allocated or freed, and also in the process of tracing.

Explicitly specified memory area is checked by comparison its reference checksum with the current one. The reference checksum is calculated when Run-Time System is initialized.

The initial and final addresses of code memory are specified by the program, starting Run-Time System by means of writing these addresses to the Run-Time System executable module. The same program calculates and writes to Run-Time System executable module reference checksum of code memory. During its operating (in particular, when trace events occurred) Run-Time System checks code memory, comparing its current checksum with reference one.

Controlling mechanisms of memory area checking is implemented by the following parameters of systrace.* files.

• EveryEventCheckBound - flag to check all dynamically allocated memory blocks every time when traced event occurred.
• EveryTraceCheckBound - flag to check all dynamically allocated memory blocks each time when trace output function is called. The checking includes the checking specified by parameter EveryEventCheckBound, because it is also performed for non-traced (disabled) events.
• dvm_StartAddr - initial address of checked memory area when tracing.
• dvm_FinalAddr - final address of checked memory area when tracing.
• EveryEventCheckMem - flag to check the memory area, specified by parameters dvm_StartAddr and dvm_FinalAddr for each traced event.
• EveryTraceCheckMem - flag to check the memory area, specified by parameters dvm_StartAddr and dvm_FinalAddr each time when trace output functions are called. The checking includes the checking, specified by parameter EveryEventCheckMem, because it is also performed for disabled trace events.
• EveryEventCheckCodeMem - flag to check memory area with program code for each traced event.
• EveryTraceCheckCodeMem - flag to check memory area with program code each time when trace output functions are called. The checking includes the checking, specified by parameter EveryEventCheckCodeMem.

11.4.6 Controlling output of additional information when tracing some Run-Time System functions in extended mode

• distr_Trace – flag to output additional information when tracing the function distr_. There are outputted abstract machine representation map, ranges of local values of indexes for each dimension of abstract machine representation, flag of replication along all dimensions of the processor system (Repl) and flag of mapping some element (the elements may be different) of abstract machine representation (Every) on each processor.
• align_Trace - flag to output additional information when tracing the function align_. There are reported distributed array map and ranges of local values of indexes for each its dimension.
• dacopy_Trace - flag to output additional information when tracing functions arrcpy_ and aarrcp_. There are reported distributed array maps, ranges of local values of indexes for each dimension, information about the execution branches of the copy program and other information.
• mappl_Trace – mode of additional information output during tracing mappl_ function:

• dopl_Trace – flag to output additional information when tracing the function dopl_. The next set of ranges of local values of indexes for each dimension of the parallel loop is output.
• RedVarTrace – flag to output addresses, types and values of all variables of reduction group when executing functions insred_, saverv_, saverd_ and strtrd_, and also at the beginning and at the end of execution of the function waitrd_. If RedVarTrace is positive, the information about first min(RedVarTrace, RedVarNumber) elements of reduction variable-array is reported.
• OutIndexTrace – flag to output result values of indexes for each dimension of distributed array when executing internal Run-Time System function GetIndexArray.
• diter_Trace – flag to output index array when the function diter_ of interconnection with DVM-debugger is traced.
• drmbuf_Trace – flag to output index array when function drmbuf_ of interconnection with DVM-debugger is traced.
• LoadWeightTrace - flag to trace the arrays with coordinate weights of processor loadinig when invoking setelw_ function.
• WeightArrayTrace - flag to trace the array with loading coordinate weights when exiting gettar_ function.
• AcrossTrace - flag to trace functioning ACROSS schemes.

12 Diagnostics messages of dynamic debugger

The common format of error messages of dynamic debugger:

(<processor number>)<context> File: <file>, Line: <line> (<count> times) <error message>

where:

12.1 Dynamic control

12.2 Trace accumulation and comparison

13 Structure of trace configuration file

EL: <construct number>
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EL: <construct number>

14 Execution trace structure

When executions are traced, accumulated trace information consists of two parts:

• trace header;
• trace body (may be omitted).

The header exists in trace even if the trace accumulation is disabled for whole program. Its structure looks like the structure of trace configuration loop, but without calculated values of trace size for whole program and for loops:

MODE = <accumulation trace level for whole program>,

EL: <construct number>
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EL: <construct number>

Trace body is absent, when trace accumulation is disabled for whole program. Otherwise trace body consists of a lot of records of the following types:

• Reading variable.
  RD: [<variable type>] <variable name> = <value>; {<file name>, <line number>}
• Referring to variable (before expression computation).
  BW: [<variable type>] <variable name>; {<file name>, <line number>}
• Result of assigning value to variable.
  AW: [<variable type>] <variable name> = <value>; {<file name>, <line number>}
• Reading reduction variable. A record in trace file is the following.
  RV_RD: [<variable type>] <variable name> = <value>; {<file name>, <line number>}
• Referring to reduction variable (before expression computation).
  RV_BW: [<variable type>] <variable name>; {<file name>, <line number>}
• Result of assigning value to reduction variable.
  RV_AW: [<variable type>] <variable name> = <value>; {<file name>, <line number>}
• Result of reduction computation.
  RV: [<variable type>] <value>; {<file name>, <line number>}
• Skipping a group of statements when referring an element of distributed array in a sequential branch of the program.
  SKP: {<file name>, <line number>}
• Parallel loop beginning.
  PL: <loop number> (<parent construct number or empty>) [<loop rank>] = <trace level: FULL, MODIFY, MINIMAL, NONE>, (<traced iteration range (can be absent)>); {<file name>, <line number>}
• Sequential loop beginning.
  SL: <loop number> (<parent construct number or empty>) [<loop rank>] = <trace level: FULL, MODIFY, MINIMAL, NONE>, (<traced iteration range (can be absent)>); {<file name>, <line number>}
• Task region beginning.
  TR: <region number> (<parent construct number or empty>) [region rank>] = <trace level: FULL, MODIFY, MINIMAL, NONE>, (<traced task range>); {<file name>, <line number>}
• Beginning of iteration (put in trace file only when the most nested loop iteration is executed) or parallel task.
  IT: <absolute iteration index (calculated from all values of all iteration variables) or task number>, (<iteration variable value>,<iteration variable value>,…).
• End of parallel loop or task region execution.
  EL: <construct number>; {<file name>, <line number>}

15 Error messages of Run-Time System

The error messages, occurring in process of Run-Time System operating, can be divided (by error level) on warnings and error messages, breaking the program execution. The errors, breaking the program execution, can be usual (a user errors most likely) and fatal, requiring, possibly, a participation of program builder.

The error messages have one of the following forms:

The error class is defined by functional and module structure of Run-Time System.

The file name and line number (as user program, as Run-Time System), where the error occurred are reported before the error message:

USRFILE = <file name of user program>; USRLINE = <line number>;
SYSFILE = <file name of Run-Time System>; SYSLINE = <line number>;

The errors may be divided on two groups according to their occurrence way. The first group is the errors, occurring on all the processors at once. Input/output processor outputs such error messages. The second group is the errors that can occur not on all the processors. Each processor outputs such error messages itself, adding to the message its internal and external numbers. When working on one processor, the error messages are output without its numbers.

Run-Time System provides possibility to output all error messages of the first group by all the processors and all errors of the second group only by input/output processor. The first possibility is implemented if the parameter MultiProcErrReg is set equal to 2 (parameter files with name syspar). To output error messages of the second group only by input/output processor this parameter must be equal to zero (this possibility is used only in experimental purposes). Standard value of parameter MultiProcErrReg is equal to 1.

All the error messages, divided by subjects and ordered by class numbers and error numbers in the class are described in document "DVM-program Run-Time System. Detailed design".

The numbers of the messages, classified by possible causes, are listed below. The numbers of "fatal error" messages are not listed. The errors, which can't occur in the programs, produced by C-DVM and F-DVM compilers (in the case of correct functioning of the compilers), but can occur only when programming in terms of Run-Time System functions (however, in the case of memory corruption by user program any error are possible in any program) are also not considered.

15.1 Start and completion errors

These errors are indicated by the messages with 01x.xxx and 02x.xxx numbers.

The following messages belong to this type:

• messages about opening failure of input and output files (parameter files, files of informational message output, stdout and stderr redirection files, trace files); the reasons of such messages are inaccessibility or absence of the files (directories) or wrong access rights of the file creation;
• messages about lack of a correspondence between versions of the system and parameter files;
• messages about syntax errors in parameter files (wrong keywords, index values, wrong or not specified parameter value);
• messages about lack of memory for system buffers;
• messages about trace file-writing failure.

Moreover on the program termination the messages with 022.002 and 022.003 numbers can appear. These messages indicate wrong checksum of code and data memory, i.e. modifying the memory, which can't be modified ("memory corruption").

During the program execution the messages with 230.000, 230.001, 230.002, 230.003, 230.004, 250.000, 250.001 numbers can also indicate "memory corruption".

15.2 Errors of the type "... is not a DVM object"

"The object, passed to the function, is not DVM-object".

Possible reason is that the object is used before creation/initialization or after deletion.

The message numbers are:

030.000, 030.010, 030.020, 032.000, 032.010, 032.015, 032.020, 032.030, 032.050, 032.070, 032.080, 032.085, 036.000, 036.010, 036.060, 036.066, 037.000, 037.010, 037.020, 038.000, 038.010, 038.012, 038.014, 038.032, 038.038, 038.040, 038.050, 044.000, 044.040, 046.011, 046.014, 046.050, 046.054, 046.090, 060.010, 060.065, 060.075, 060.115, 065.000, 065.003, 065.020, 069.020, 070.000, 070.002, 070.010, 070.016, 070.030, 070.040, 070.050, 070.060, 070.070, 070.080, 080.000, 080.020, 080.040, 080.050, 080.060, 080.070, 080.080, 080.150, 080.160, 080.200, 080.250, 120.006, 130.065, 132.010, 132.030, 132.040, 132.050, 144.010, 144.030, 144.040, 144.050, 219.000.

15.3 Errors of the type "[the object] is not a/the …"

"The object, passed to the function, is not the object of required type".

As a rule the correspondence between the object type and executed operation is checked by C-DVM and F-DVM compilers. This error is caused more likely by the object [descriptor] "corruption".

The message numbers are:

030.001, 030.011, 030.021, 032.001, 032.011, 032.016, 032.021, 032.031, 032.051, 032.071, 032.081, 032.086, 036.001, 036.011, 036.061, 036.067, 037.001, 037.011, 037.021, 038.001, 038.011, 038.013, 038.015, 038.033, 038.039, 038.041, 038.051, 040.005, 040.008, 040.010, 044.001, 044.002, 044.041, 044.042, 044.060, 046.000, 046.010, 046.012, 046.015, 046.051, 046.052, 046.055, 046.091, 060.011, 060.012, 060.013, 060.050, 060.051, 060.060, 060.061, 060.066, 060.070, 060.071, 060.076, 060.080, 060.081, 060.090, 060.091, 060.100, 060.116, 060.200, 065.001, 065.004, 065.021, 069.021, 070.001, 070.003, 070.008, 070.014, 070.031, 070.041, 070.051, 070.061, 070.071, 070.081, 080.001, 080.002, 080.021, 080.022, 080.041, 080.051, 080.061, 080.071, 080.081, 080.151, 080.161, 080.201, 080.202, 080.251, 080.252, 090.000, 090.001, 090.010, 090.011, 090.020, 090.030, 090.040, 090.050, 090.051, 090.060, 092.000, 092.001, 092.010, 092.011, 092.020, 094.000, 096.000, 110.000, 110.003, 110.009, 110.012, 120.000, 120.001, 120.002, 120.003, 120.005, 130.001, 130.004, 130.005, 130.031, 130.035, 130.061, 130.066, 131.000, 131.001, 131.020, 131.022, 131.040, 131.042, 132.011, 132.014, 132.016, 132.031, 132.041, 132.051, 133.001, 133.004, 133.005, 140.001, 140.020, 140.021, 140.030, 141.001, 141.030, 143.000, 143.001, 143.020, 143.022, 143.040, 143.042, 144.011, 144.014, 144.016, 144.031, 144.041, 144.051, 219.001, 219.010, 219.011, 219.020, 219.021.

15.4 Errors of the type "… is not a subsystem of the current/parental PS"

"Processor system, passed to the function, is not current/parental processor system or its direct or indirect subsystem".

The messages indicate a violation of subsystem hierarchy. For example, the subtask attempts to use/change the object, belonging to higher hierarchy level.

The message numbers are:

032.012, 032.013, 032.032, 032.052, 032.082, 032.083, 036.012, 036.013, 036.068, 036.069, 038.025, 038.026, 038.042, 038.043, 044.003, 044.044, 044.045, 046.017, 046.058, 046.059, 070.013, 070.027, 080.006, 080.025, 080.047, 080.057, 080.067, 080.157, 080.167, 080.206, 080.255, 090.004, 090.005, 090.014, 090.015, 090.023, 090.024, 092.004, 092.005, 092.014, 092.015, 092.023, 092.024, 094.006, 094.007, 096.006, 096.007, 101.001, 101.011, 110.002, 110.005, 110.011, 110.014, 130.003, 130.033, 130.037, 130.063, 130.067, 131.004, 131.005, 132.032, 132.033, 133.003, 140.003, 140.033, 141.003, 141.033, 143.004, 143.005, 144.032, 144.033, 219.002.

15.5 Errors of the type "…was not created by the current subtask" and "…was not started by the current subtask"

"The object passed to the function was not created in the current subtask" and "Asynchronous operation, passed to the function, was not started by the current subtask".

The messages indicate a violation of subsystem hierarchy. For example, the subtask attempts to use/change the object, belonging to other subtask. Or the subtask attempts to wait for asynchronous operation (implemented by the pair of START- WAIT statements), started by other subtask.

The message numbers are:

060.067, 060.077, 060.117, 065.006, 069.022, 070.005, 070.006, 070.032, 070.042, 070.052, 070.062, 070.072, 070.082, 080.003, 080.023, 080.056, 080.066, 080.073, 080.082, 080.156, 080.166, 080.203, 080.253, 120.008, 131.023, 131.043, 132.012, 132.017, 132.042, 132.052, 143.023, 143.043, 144.012, 144.017, 144.042, 144.052.

15.6 Errors of the type "…has not been aligned/mapped" and "…has already been aligned/mapped"; "…does not exist" and "…already exists"

Possible reason of such errors is wrong order of the array and other DVM-object creation/usage. For example, the array A, aligned with the array B, can be created (by malloc statement) only if the array B already exists (was created). Similarly, parallel loop can be aligned with the array, only if the array exists. Or vice versa repeated creation of the object is performed. Note that repeated creation of the array (malloc) in sequential program may be not detected, and (repeated) creation of the template or shadow edge group is invisible for sequential program.

The message numbers are:

032.017, 032.088, 036.002, 036.003, 036.014, 036.051, 038.002, 038.020, 038.021, 038.024, 038.031, 040.000, 040.006, 044.005, 044.006, 044.047, 046.001, 046.019, 046.020, 046.061, 060.014, 060.015, 060.052, 060.062, 060.072, 060.082, 060.101, 060.201, 065.007, 065.009, 065.022, 070.009, 070.015, 080.005, 080.024, 080.205, 080.254, 090.002, 090.003, 090.012, 090.013, 090.021, 090.022, 090.041, 090.031, 090.052, 090.053, 090.061, 092.002, 092.003, 092.012, 092.013, 092.021, 092.022, 094.004, 094.005, 096.004, 096.005, 101.000, 101.010, 110.001, 110.004, 110.010, 110.013, 130.000, 130.002, 130.006, 130.030, 130.032, 130.036, 130.060, 130.062, 133.000, 133.002, 133.006, 140.000, 140.002, 140.022, 140.032, 141.000, 141.002, 141.032, 219.012, 219.022.

15.7 Errors of the type "… has already been started/inserted …", "… has not been started/completed" and "… the reduction group is empty"

Such errors are a result of wrong sequence of the actions with DVM-objects: reduction operation and shadow edge groups, remote access buffers and so on. For example, new start of the operation before the completion of started one, start of the operation with empty group, waiting for completion of not started operation, including variables (arrays) in the group during execution of operation with this group, deleting object during the operation execution and so on.

The message numbers are:

060.120, 060.140, 069.023, 070.004, 070.033, 070.034, 070.044, 070.053, 070.054, 070.063, 070.073, 070.083, 080.007, 080.008, 080.009, 080.010, 080.011, 080.026, 080.027, 080.028, 080.029, 080.030, 080.042, 080.043, 080.044, 080.045, 080.046, 080.052, 080.053, 080.054, 080.062, 080.063, 080.064, 080.072, 080.083, 080.152, 080.153, 080.154, 080.162, 080.163, 080.164, 080.207, 080.208, 080.209, 080.210, 080.211, 080.256, 080.257, 080.258, 080.259, 080.260, 131.003, 131.024, 131.044, 131.045, 132.013, 132.019, 132.034, 132.043, 132.053, 143.003, 143.024, 143.044, 143.045, 144.013, 144.019, 144.034, 144.043, 144.053.

15.8 Index and value errors

The errors of this type occurs in the case of a boundary violation of DVM-object dimension (for example, when aligning array or describing parallel loop), indexing of not-existing dimension, wrong ratio between indexes (for example, initial index is more than last one), wrong DVM-object rank and so on. Incorrect values can be passed to Run-Time System, for example, in GENBLOCK vector or as shadow edge widths (or as result of "memory clearing"). Run-Time System messages contain the expression (<equality or inequality>) - incorrect ratio. Description of the values, contained in each message, can be found in document "DVM-program Run-Time System. Detailed design" (rtsDDe.doc file).

The message numbers are:

030.022, 030.023, 032.019, 032.022, 032.035, 032.036, 032.037, 032.038, 032.039, 032.055, 032.056, 032.057, 032.090, 032.091, 032.092, 036.020. 036.021, 036.022, 036.023, 036.024, 036.025, 036.026, 037.002, 037.003, 037.012, 037.022, 038.016, 038.017, 038.044, 038.045, 040.001, 040.002, 040.003, 044.019, 044.020, 044.021, 044.022, 044.023, 044.024, 044.025, 044.026, 044.027, 044.028, 046.013, 046.016, 046.030, 046.031, 046.056, 046.057, 060.000, 060.001, 060.019, 060.020, 060.021, 060.022, 060.023, 060.024, 060.025, 060.026, 060.027, 060.028, 060.029, 060.030, 069.000, 069.001, 069.002, 069.003, 069.024, 080.014, 080.015, 080.031, 080.032, 080.033, 080.034, 080.035, 080.214, 080.215, 080.216, 080.220, 080.225, 080.226, 080.227, 080.230, 080.235, 080.236, 080.237, 080.261, 080.264, 080.265, 080.266, 080.270, 080.275, 080.276, 080.277, 080.280, 080.285, 080.286, 080.287, 080.290, 080.291, 094.001, 094.002, 094.003, 096.001, 096.002, 096.003, 110.006, 110.015, 120.004, 120.200, 130.010, 130.011, 130.012, 130.013, 130.014, 130.015, 130.016, 130.017, 130.018, 130.019, 130.040, 130.041, 130.042, 130.043, 130.044, 130.045, 130.046, 130.047, 130.070, 140.010, 140.011, 140.023, 140.031, 140.036, 140.037, 140.038, 140.041, 140.042, 140.043, 140.044, 141.010, 141.011, 141.031, 141.036, 141.037, 141.038, 219.015, 219.016, 219.017, 219.025, 219.026, 219.027.

15.9 Other semantic errors

The array or template redistribution and remapping, which is impossible or was not allowed when the array or template was created.

The message numbers are:

030.002, 036.050, 036.065, 038.030, 038.037, 044.043, 046.053, 065.005, 065.010.

Attempt to execute parallel loop, in which the order of iterations is changed using ACROSS scheme (scheme of execution of the loops with data dependence between iterations).

The message numbers are: 060.110, 060.111.

Attempt to extend a set of local iterations of parallel loop with changed order of iteration execution.

The message numbers are: 060.210, 060.211.

Including reduction variable simultaneously in two different groups.

The message number is 070.007.

The distribution of arrays in SHADOW_START or SHADOW_WAIT sub-directive is differ from the loop distribution.

The message number is 060.040.

Incorrectly distributed objects in non-regular access (for example, a number of distributed dimensions of remote array is not equal to 1).

The message numbers are:

140.004, 140.034, 140.035, 140.040, 140.050, 140.051, 141.004, 141.034, 141.035.

15.10 Memory allocation and the number of objects errors

The errors of this type occur in the case of lack of the memory because of the errors in the sizes of created arrays. Allocation memory errors can also occur because of the memory "corruption" by the user program or its "looping".

The message numbers are:

200.000, 200.001, 200.002, 200.003, 200.004, 200.005, 200.006, 200.007, 200.008.

The following messages are output when too many objects (distributed arrays, remote access buffers) were created. In C-DVM these errors can be caused by imbalance of malloc and free statements for distributed arrays or by the usage of local in the function (or in the block) "static" distributed arrays (the arrays with constant dimensions).

The message numbers are:

040.004, 040.007, 130.029, 140.060.

15.11 Errors of low level message passing

Message passing errors are the result of incorrect interaction of Run-Time System with message passing system.

The following reasons are possible:

• hardware errors;
• errors in message passing system or its incorrect installation;
• wrong compilation of Run-Time System;
• "corruption" of internal tables of Run-Time System by user program.

It is recommended to developer or its representative.

The message numbers are:

210.000, 210.001, 210.002, 210.003, 210.004, 210.005, 212.000, 212.001, 212.002, 212.003, 213.000, 213.001, 213.002, 214.000, 214.001, 219.003.

16 Structure of system trace file

The system trace file has the following structure:

• the event of function call has the name call_<function_name> ,
• the event of return from function has the name ret_<function_name> .

The trace of each event cab be enabled or disabled by its number (see section 6 and 11.4).

Structure of accumulated trace information is defined by trace mode. In brief mode the following is outputted:

• event name;
• time, passed from previous event;
• name of user program file and line number in the file, defining event point in the program.

In detailed trace mode saving information is extended and depends on event type. If the event is function call, then its call parameters are outputted. When returning function its result parameters are output. For each event the detailed trace mode can be specified independently of common mode.

When tracing accumulated information also contains all informational messages and messages about Run-Time System errors.

To make representation of function calls belonging to different nested levels clear, desirable indentation may be specified in parameters, controlling system trace (see section 11.4).