Compiler Options

The following topics describe each of the Open PL/I Compiler options:

Attention: Unless otherwise indicated, PL/I Compiler options have no default setting and should be considered disabled unless explicitly specified to the Compiler.

Note: When compiling from the command line, enclose parameter values containing special characters (including spaces) in double quotes (" "). Optionally, you may also use double quotes to enclose parameter values that do not contain any special characters. Do not enclose parameter values in single quotes.

-#
Provides information on how the Compiler and linker would be invoked without actually invoking them.
-adt
Generates a file containing the data structures processed by the PL/I compiler.
-agginit
Performs initialization of aggregates compatible with 7.x versions of Open PL/I (scalar arrays only).
-ascii
Specifies ASCII character encoding for all character data.
-bifprec
Sets the precision of a built-in function's result to 15 or 31.
-bigendian
Specifies that all FIXED BINARY, CHARACTER VARYING, GRAPHIC VARYING and WIDECHAR VARYING items are implicitly declared with the BIGENDIAN data attribute, unless the NATIVE attribute is explicitly applied to a data item.
-bitsltr
Stores bit strings left-to-right within each byte.
-bss
Uses the .bss section for uninitialized static variables.
-c
Suppresses linking and produces a relocatable object file.
-cics
Invokes the CICS preprocessor for EXEC CICS support.
-ctf
Compiles code enabled for use of the Consolidated Tracing Facility (CTF Trace).
-currency_symbol
Allows an alternative currency symbol for picture specification.
-dc[8]
Tells the compiler and/or preprocessors to create a data collection (.dcf) file to be consumed by the IDE(s) and Debugger for enhanced functionality, such as where a variable is referenced, its size, etc.
-dc_append[8]
Instructs the invoked program not to overwrite existing data collection data, but instead to add to it. Generally used to affect stacked preprocessors or a preprocessor and compiler.
-dcfout[8]
Enables the generation of a non-standard .dcf filename and optional location.
-deb
Produces debugging information for CodeWatch, Micro Focus's source-level debugger.
-decimal
Sets the conditions under which a FIXEDOVERFLOW is raised for a FIXED DECIMAL overflow.
-default
Applies the CONNECTED attribute to array parameter descriptors by default.
-default_binary
Causes the BINARY attribute to be assumed if the FIXED attribute is specified without BINARY or DECIMAL.
-default_decimal
Causes the DECIMAL attribute to be assumed if the FIXED attribute is specified without BINARY or DECIMAL.
-defext / -nodefext
Determines whether initialized values for STATIC EXTERNAL variables are defined.
-dli
Invokes the DL/I preprocessor for EXEC DLI support.
-dummy
When passing arguments, causes a dummy argument to be created when the aligned and unaligned attributes differ between the argument and parameter descriptor.
-E
Causes any specified preprocessors to be run on the source file, and saves their output, but stops without running the Compiler.
-ebcdic
Specifies EBCDIC character encoding for all character data.
-err_on_stdout
Displays or prints compiler error messages to STDOUT instead of STDERR.
-exp
Produces an expanded listing of the generated code in the form of assembly language statements in the listing file.
-fb7asfb15
Treats Fixed Binary (7) as Fixed Binary (15).
-fbmaxp / -nofbmaxp
Determines whether to set the maximum Fixed Binary precision to a precise value or to use standard PL/I conversion rules to determine precision.
-fdasfb / -nofdasfb
On UNIX platforms only, determines how Float Decimal data is internally coded.
-fdcheck / -nofdcheck (deprecated)
Determines whether to check for a possible loss of digits when assigning or converting a Fixed Decimal.
-fdmaxp
Specifies the maximum FIXED DECIMAL precision.
-flag
Specifies the minimum diagnostic level of messages.
-fmaxp / -nofmaxp
Determines whether to use floating-point precision for trigonomic built-in functions or to use standard PL/I conversion rules to determine precision.
-fpcobol
Tells the Compiler to generate code compatible with COBOL handling of floating-point parameters.
-full_list
Tells the macro preprocessor to omit %NOLIST directives normally placed before code included by the %INCLUDE statement, thereby allowing the included code to be visible in the compilation listing.
-graphic
Enables support for the Graphic data type.
-help
Prints out a list of available compiler options to the console or STDOUT output, and to EXIT.
-hidemsg
Hides one or more diagnostic messages per compilation.
-highbif
Allows changing the value returned by the HIGH() built-in function.
-ims (deprecated)
Use -systemims instead.
-incl
Invokes the Include preprocessor to process selected or default include statements and produce a deck that is then treated as a single source file by another preprocessor or the Open PL/I Compiler.
-initcall
Causes the program being compiled to fetch and execute a call to the named routine before executing any procedural code.
-initcheck / -noinitcheck
Determines whether diagnostic warnings are reported for possibly uninitialized automatic variables at first use within the given procedure.
-insource / -noinsource
Determines enablement of source listings.
-ipath
Directs the Compiler to search specified directories for files to be included in the source program text by use of the %INCLUDE statement.
-ipubr
Causes bits for unsigned integers created using INT or POSINT and bit strings returned with UNSPEC to be inversed.
-isuffix
Appends a specified string to the filename specified in %INCLUDE statements, excluding those in which the quoted filename form of the %INCLUDE statement is used.
-l
Produces a compiler listing written to a specified or implied output file.
-largedata
On UNIX/Linux platforms only, selects all options recommended for programs requiring large amounts of storage for static variables, and in 64-bit environments, also disables the 2GB limit for initialized static storage.
-laxbased / -nolaxbased
Determines storage checking of BASED variables.
-laxcomment / -nolaxcomment
Determines whether comments containing a semicolon (;) are diagnosed.
-laxdefined / -nolaxdefined
Determines enablement of type and storage checking for DEFINED variables.
-laxif / -nolaxif
Determines whether the compiler allows or diagnoses IF, WHILE, UNTIL, and WHEN clauses that are not of type BIT(1), and whether to allow or flag assignments of the form A=B=C .
-laxinit / -nolaxinit
Determines enablement of dependency ordering of automatic initializers in block prologue.
-laxother / -nolaxother
Determines whether to check for a missing OTHERWISE clause in SELECT statements.
-laxtype
Provides backwards compatibility to declarations that are not compatible with mainframe syntax.
-longint
Changes the default precision of Fixed Binary variables from (15) to (31).
-lowercase (deprecated)
Converts all uppercase names of internal and external variables and constants to lowercase.
-macro / -nomacro
Determines invocation or suppression of the mfpp Open PL/I Macro Preprocessor.
-map
On UNIX platforms only, produces a storage allocation map at the end of the listing file.
-map
On Windows platforms only, produces a storage allocation map at the end of the listing file and/or generates a linker map for a specified filename.
-margini
For the listings produced by the -insource and -source options, specifies a character to place in the column immediately preceding the left margin, and also in the column immediately after the right margin, and optionally delimits the left and right margins in the source listing.
-margins
Sets the margins within which characters of the source file are interpreted as source code and outside which all characters are ignored by the Compiler.
-maxmsg
Specifies the maximum number of diagnostics of a specified severity or higher reported during compilation.
-medium_model / -model_medium
For the medium memory model, generates code to support programs requiring large amounts of storage for static variables.
-multiclose / -nomulticlose
Specifies whether multiple closure of groups is enabled.
-mvs (deprecated)
Use -systemmvs instead.
-names
Adds extra-lingual characters to identifiers via the compiler options, similar in nature to %PROCESS NAMES().
-nest
Causes the nesting of DO ... END, PROC ... END, and so on to be shown in the listing file.
-nodescriptors
Prevents the compiler from passing descriptors to called subroutines and functions. Equivalent in functionally to OPTIONS(NODESCRIPTOR).
-nodebuginfo
Tells the Compiler to not break the line for %FILE and %LINE information.
-nogotocheck
Disables runtime validity checking when attempting a goto on a subscripted label.
-noincludes
Tells the Compiler and/or macro preprocessor not to include the contents of %INCLUDE files in the compiled source.
-nolaxdcl
Tells the Compiler to not allow compiler declarations for undeclared variables.
-nolaxtype
Do not allow laxed TYPE/LIKE usage.
-noobj
Compiles for syntax and semantic checking only. Does not produce an object file.
-nonnative (deprecated)
Use -bigendian instead.
-range / -norange (deprecated)
This option is deprecated and will be removed in a future release. See the -prefix option.
-not_symbol
Allows an alternative NOT symbol.
-nounixdebug
Avoids generating the additional object file data needed to report source file and line number when a fatal run-time error occurs, or for use by the ONLOC built-in function.
-nounrefexternals
Prevents external entry symbols that are not referenced in the program from being generated in the object file.
-O
Invokes the Compiler optimization phase.
-o
Names the object file using a specified file name and passes it to ldpli.
-opt / -noopt
Specifies whether to invoke the optimization phase of the Compiler.
-optcics
Allows the CICS options to be passed to the CICS ECM.
-optdli
Allows the DL/I options to be passed to the DL/I (IMS) ECM.
-optexec
Lets you specify options for the EXEC preprocessor.
-options
Adds information about translated %PROCESS directives and command-line option processing to the compiler listing.
-optsql
Allows the SQL options to be passed to the SQL preprocessor.
-or_symbol
Allows an alternative OR symbol.
-paramcount
Enables the use of the LPIPARAMCOUNT built-in function, which is Open PL/I-specific.
-pass
Connects to a remote DB2 database.
-pic
Generates position-independent code, suitable for linking into a shared library.
-pl1g
Causes the Compiler to reinstate the behavior of releases before version 5.0 for certain operations.
-plitdli
Uses a specified IMS linkage when calling a PLITDLI entry point name rather than a user-supplied PLITDLI.
-pmod
Changes the output of the MOD built-in function to always return a positive result, even with a negative divisor.
-pp
Specifies the file name for the output file of the Open PL/I Macro Preprocessor.
-ppcics
Specifies the file name for the output file of the CICS preprocessor.
-ppdli
Specifies the file name for the output file of the DL/I preprocessor.
-pporder
Selects which preprocessors are to be invoked and in what order.
-ppsql
Specifies the file name for the output file of the SQL preprocessor.
-prefix
Allows attachment of one or more condition prefixes to a program without the need to edit the source code.
-proginfo
Causes the compiler to insert a data structure that allows the CBL_GET_PROGRAM_INFO function to identify and report PL/I programs when walking the program stack.
-proto
Provides more extensive prototype checking on arguments vs. parameters in entry declarations.
-rc
Specifies the maximum error code level that results in a zero return code from mfplx.
-report
Specifies the file name for the output of the Open PL/I Source Checker.
-roundtozero[1]
Changes rounding mode of floating point arithmetic to round towards zero.
-setnull
Designates a null pointer value to be returned by the NULL built-in function.
-setstack
Sets each byte of the storage to a default or specified value.
-shortint
Tells the Compiler to set the default FIXED BINARY precision to 15.
-so
On UNIX platforms only, builds shared callable procedure objects (.so) based on your operating system.
-source / -nosource
Determines enablement of source listings.
-sql
Invokes the SQL preprocessor for EXEC SQL support.
-stardesc
Tells the Compiler to produce argument descriptors for calls only when array bounds or string lengths in the parameter declaration are represented by the asterisk (*) symbol.
-stat
Prints Compiler statistics to the standard output as each Compiler phase completes.
-stbout
Generates a .stb file in the same directory as the output object file.
-strip
Tells ldpli to not generate debugging information.
-structure_fillers
Tells the Compiler to produce warnings for structures containing fillers.
-systemcics
Specifies that this OPTIONS(MAIN) program executes under the control of CICS.
-systemims
Specifies that this OPTIONS(MAIN) program executes under the control of IMS.
-systemmvs
Specifies that this OPTIONS(MAIN) program executes under the control of MVS.
-unaligned
A compile-time switch that causes byte instructions to be generated for all based references, which might point anywhere in memory.
-unrefexternals
Causes external entry symbols to be generated in the object file, even if they are not referenced in the program.
-v
Prints information to stdout relating to the compilation and link-step phases.
-variant
Passes a variant string to the Open PL/I Macro Preprocessor.
-vax
Adapts programs originally written with Digital Equipment Corporations's PL/I Compiler for the VAX platform.
-warn / -nowarn (deprecated)
Use -flag instead.
-window
Sets the value for the w parameter used in date-related built-in functions.
-xmlattr
Causes the XMLCHAR built-in function to enclose XML attributes in either single quotes or double quotes.
-xmlcase
Causes the XMLCHAR built-in function to generate tags in its XML output to be all uppercase or in the case used in the declaration.
-xref / -noxref
Specifies whether to include cross-reference information in the listing file.
-zalign
Specifies z/OS aggregate mapping and alignment rules.
-zchar
Specifies that the default alignment attribute for various data types is unaligned.
-zfloat
Changes the Float Binary single precision value to 21, and changes the Float Binary double maximum to 53.
-zp1
Causes all structures to be mapped as if the UNALIGNED attribute is specified at the level 1 structure level.

Parent topic: Compiling a Program

Related concepts

Debugging PL/I Programs Under CICS

Related reference

mfexecpp Utility

Related information

External Names

Open PL/I Compiler Diagnostic Messages

Compilation Statistics

Structure