This section provides compatibility information for those who:

• Port Compaq Fortran 77 and Compaq Fortran applications from OpenVMS systems to Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha Systems
• Design Compaq Fortran applications to run on multiple platforms, including OpenVMS and Compaq Tru64 UNIX or Linux Alpha systems

If your primary concern is the design and development of Compaq Fortran applications for only Compaq Tru64 UNIX (or other U*X) systems, consider skipping this section.

This section discusses the following topics:

• Compaq Fortran 77 extensions for OpenVMS systems that are not supported by this version of Compaq Fortran 77 or Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha Systems ( Section A.4.1)
• Porting Compaq Fortran data files from an OpenVMS system to a Compaq Tru64 UNIX or Linux Alpha system ( Section A.4.2)
• Nonnative VAX floating-point representations, provided for those converting unformatted OpenVMS floating-point data ( Section A.4.3)

Some extensions to the FORTRAN-77 standard provided by Compaq Fortran 77 and Compaq Fortran for OpenVMS Systems are specific to the OpenVMS operating system, VAX architecture, or certain products on OpenVMS systems. Such extensions are not included in Compaq Fortran 77 or Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha Systems.

For information on language compatibility between Compaq Fortran and Compaq Fortran 77 without regard to operating system or architecture differences, see Section A.2.

Compaq Fortran 77 and Compaq Fortran products for OpenVMS systems include:

• Compaq Fortran Version 7.2 for OpenVMS Alpha Systems
• Compaq Fortran 77 Version 7.2 for OpenVMS Alpha Systems
• Compaq Fortran 77 Version 6.6 for OpenVMS VAX Systems (previously called VAX FORTRAN)

Unless otherwise noted, the following list describes the Compaq Fortran 77 extensions in Compaq Fortran 77 and/or Compaq Fortran for OpenVMS systems that are not supported by Compaq Fortran Version 5.3 for Compaq Tru64 UNIX Systems:

• DICTIONARY statement
  The DICTIONARY and related support for the CDD/Repository (common data dictionary) product are not provided by Compaq Fortran or Compaq Fortran 77 for Compaq Tru64 UNIX Systems.
• Support for indexed sequential files
  I/O statement specifiers for indexed file (keyed access) record I/O using OpenVMS OPEN and INQUIRE statement specifiers are not provided by Compaq Fortran or Compaq Fortran 77 for Compaq Tru64 UNIX or Linux Systems, as follows:

  ACCESS= ' KEYED '	EXTENDSIZE
  INITIALSIZE	KEY
  NOSPANBLOCKS	ORGANIZATION= ' INDEXED '
  SHARED

• FORSYSDEF symbol definitions for OpenVMS systems
  The parameter definitions of run-time messages found in FORSYSDEF.TLB library module FORIOSDEF on OpenVMS systems are provided in the file /usr/include/foriosdef.f (see Section 8.2.2) on Compaq Tru64 UNIX Systems. On Compaq Tru64 UNIX and Linux Alpha systems, Compaq Fortran and Compaq Fortran 77 provides jacket routines to simplify calling system calls and library routines (see Chapter 12).
• The INCLUDE statement option of including text from text libraries.
  On Compaq Tru64 UNIX and Linux Alpha systems, OpenVMS text libraries are not supported.
• The %DESCR built-in function (for OpenVMS character descriptors).
  On Compaq Tru64 UNIX and Linux Alpha systems, character data is passed by address and hidden length. For information about calling or being called by procedures written in other languages, see Chapter 11.
• Run-time default I/O units spelled as FOR0nn.dat, SYS$INPUT, and so on
  In Compaq Fortran and Compaq Fortran 77 on Compaq Tru64 UNIX and Linux Alpha systems, these are environment variables FORTn, stdin , stdout , and so forth (see Section 7.5.1).
• VAX floating-point formats and related selection of the floating-point format in memory
  Only IEEE floating-point formats are supported in memory on Compaq Tru64 UNIX (Alpha) systems. (Compaq proprietary VAX floating-point formats are not supported in memory.) You can request conversion of unformatted files containing VAX floating-point formats into the appropriate IEEE memory format during record I/O (see Chapter 10).
  On OpenVMS VAX systems, you specify the floating-point format to be used in memory with either the option [NO]G_FLOATING in the OPTIONS statement or the qualifier /[NO]G_FLOATING on the FORTRAN command line.
  On OpenVMS Alpha systems, you specify the floating-point format to be used in memory using the /FLOAT qualifier on the FORTRAN command line.
• Stream record format differences
  With Compaq Fortran 77 and Compaq Fortran for OpenVMS systems, the Stream record type is delimited by CR-LF character sequence (carriage control and line feed characters). In Compaq Fortran for Compaq Tru64 UNIX systems, the Stream record type uses no delimiters.
  For more information on compatible record types, see Section A.4.2.
• Other differences related to the OpenVMS operating system and the Compaq Tru64 UNIX and Linux operating systems
  When parsing file specifications for the OPEN, INQUIRE, and INCLUDE statements, keep in mind that file names are case-sensitive on Compaq Tru64 UNIX and Linux systems and that OpenVMS file specification syntax differs from pathname syntax.
  For the INCLUDE statement, the network node names (terminated by "::"), logical names (usually terminated by ":"), and other OpenVMS file specification components are not recognized. Instead, the INCLUDE statement should specify a pathname, possibly with an absolute directory path.
• The OpenVMS operating system provides various system services (SYS$ prefix) and run-time library routines (LIB$, SMG$, and other prefixes) that are not supported on Compaq Tru64 UNIX and Linux systems. Compaq Tru64 UNIX systems support system calls and library routines with similar functions (but different names).
  To make programs more portable to other operating systems, wherever possible you should use standard-conforming Compaq Fortran intrinsic routines in place of routines specific to a particular operating system.
  For more information on specifying files, see Section 7.5.1.3.

The following language and VAX architecture features are associated only with Compaq Fortran 77 on OpenVMS VAX Systems (previously called VAX FORTRAN) and are not supported by Compaq Fortran Version 5.3 for Compaq Tru64 UNIX and Compaq Fortran Version 1.0 for Linux Alpha systems:

• Directed decomposition features and CPAR$ directives for parallel processing
  CPAR$ directives are treated as comments (ignored). Parallel processing capabilities (appropriate f90 options, OpenMP, Compaq Fortran parallel, and HPF data mapping directives) are provided by Compaq Fortran.
• OPTIONS statement options /BLAS, /NOBLAS, /CHECK=ALIGNMENT, /CHECK=NOALIGNMENT, /CHECK=ASSERTION, /CHECK=NOASSERTION, /G_FLOAT, and /NOG_FLOAT
  You can specify some of these options by using the corresponding f90 command-line options. The OPTIONS statement is treated as a comment (ignored).
• CDEC$ performance directives ASSERT and NOVECTOR are treated as comments (ignored).
• The REAL*16 floating-point data type
  On VAX systems, REAL*16 data is in H_float format. On Alpha systems, REAL*16 data is in the native IEEE style X_float; so are both halves (real and imaginary) of COMPLEX*32 data.
• The following subroutines for PDP-11 compatibility:

  ASSIGN CLOSE ERRSET

  ERRTST FDBSET IRAD50

  RAD50 R50ASC USEREX

• Radix-50 constants and character set
• The BLAS routines
  Similar basic linear algebra routines are provided in the Compaq Extended Mathematical Library (CXML) product (see Section 5.1.1).

The following language and VAX architecture features are interpretation differences between Compaq Fortran and Compaq Fortran 77 on Alpha systems and Compaq Fortran 77 on OpenVMS VAX Systems (previously called VAX FORTRAN):

• Random number generator (RAN)
  The RAN function (one argument) generates a different pattern of numbers in Compaq Fortran than in Compaq Fortran 77 on OpenVMS VAX Systems for the same random seed. Compaq Fortran and Compaq Fortran 77 use the same random seed. (The RAN and RANDU functions are provided for Compaq Fortran 77 on OpenVMS VAX Systems compatibility. See Compaq Fortran Language Reference Manual.)
• Hollerith constants in formatted I/O statements
  Compaq Fortran 77 on OpenVMS VAX Systems and Compaq Fortran behave differently if either of the following occurs:
  • Two different I/O statements refer to the same CHARACTER PARAMETER constant as their format specifier. For example:

    CHARACTER*(*) FMT2 PARAMETER (FMT2='(10Habcdefghij)') READ (5, FMT2) WRITE (6, FMT2)

  • Two different I/O statements use the identical character constant as their format specifier. For example:

    READ (5, '(10Habcdefghij)') WRITE (6, '(10Habcdefghij)')

  
  In Compaq Fortran 77 for OpenVMS VAX Systems, the parameter value obtained by the READ statement is modified. The parameter value modified by the READ statement is used as the output of the WRITE statement (FMT2 is ignored). However, in Compaq Fortran, the parameter value is not modified (the parameter value read by the READ statement has no effect on the parameter value written by the WRITE statement.)

For More Information:

• On language compatibility information about Compaq Fortran for Compaq Tru64 UNIX systems and Compaq Fortran 77, see Section A.2.2.
• On language interpretation differences between Compaq Fortran for Compaq Tru64 UNIX systems and Compaq Fortran 77, see Section A.2.3.
• About the Compaq Fortran language, see the Compaq Fortran Language Reference Manual.

When porting data between systems running the Compaq Tru64 UNIX and Linux Alpha operating systems and systems running the OpenVMS operating system, the file formats and the floating-point representations may differ.

The file and record formats of Compaq Fortran 77 on Compaq Tru64 UNIX systems are compatible with Compaq Fortran on Compaq Tru64 UNIX and Linux Alpha systems; they share the same language run-time I/O environment (see Chapter 7).

OpenVMS Fortran¹ files containing only character, integer, or logical data do not need field-by-field conversion, but the record types must match. The segmented record type is the same on OpenVMS Fortran systems and Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha systems. Certain other record types, such as variable-length records, differ between OpenVMS systems and Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha systems.

Table A-2 summarizes the OpenVMS Fortran record types and their equivalent record types in Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha systems.

Table A-2 Equivalent Record Types for OpenVMS Fortran and Compaq Fortran on Compaq Tru64 UNIX or Linux Alpha Systems

OpenVMS Fortran Record Type	Compaq Tru64 UNIX Fortran Record Type	Comments
Fixed-length	None	Equivalent (must be copied correctly) if you use sequential access and you specify the -vms option when compiling the Compaq Fortran file. Otherwise, convert the file to a different record type.
Variable-length	None	Not equivalent. Convert the file to a different record type.
Segmented	Segmented	Equivalent (must be copied correctly). Segmented data files can contain formatted or unformatted data.
Stream	None	Not equivalent. Convert the file to a different record type.
Stream_CR	Stream_CR	Equivalent (must be copied correctly).
Stream_LF	Stream_LF	Equivalent (must be copied correctly).

A.4.2.1 Matching Record Types

To match record types, there are several options:

• For the Segmented, Stream_CR, and Stream_LF record types, you do not need to convert the files.
• For fixed-length records where you will only use sequential access, use the -vms option when compiling the Compaq Fortran program that will access the OpenVMS Fortran files. For fixed-length records where you will use direct access, convert the files to a different record format.
• For incompatible record types, convert the files by writing a OpenVMS Fortran or C conversion program or by using the ANALYZE/RMS/FDL and CONVERT/FDL (or EXCHANGE/FDL) commands for an appropriate file. For instance, convert the OpenVMS Fortran file to the segmented record type.

Equivalent record types must be copied carefully to preserve control information and record characteristics. For example:

• Do not use the ASCII transfer mode for binary files.
• Segmented files must be copied in a manner that preserves record length information.

To transfer (copy) the files, choose one of the following methods:

• From an NFS mounted disk, use the cp command (see cp(1)).
• Perform a DECnet copy from a Compaq Tru64 UNIX or Linux Alpha system running the appropriate optional network software using dcp (see dcp(8)). Use the dcp -i option when you want to preserve record format information.
• Perform a copy from a Compaq Tru64 UNIX or Linux Alpha system with rcp , possibly by using an intermediate node running the appropriate optional network software when using a version of the OpenVMS operating system that does not support a compatible network protocol (optional product).
• Use ftp from a Compaq Tru64 UNIX or Linux Alpha system to copy a file between a Compaq Tru64 UNIX or Linux Alpha system and an OpenVMS system. Use the binary or ascii command to set the mode before you copy (get or put) the file. For example, use the ftp binary command before copying an unformatted file (such as the segmented record type).
• Perform a DECnet copy from an OpenVMS system with the EXCHANGE command with the /NETWORK and /TRANSFER=BLOCK qualifiers with a Compaq Tru64 UNIX system. To convert the file to Stream_LF format during the copy operation, use /TRANSFER=(BLOCK,RECORD_SEPARATOR=LF) instead of /TRANSFER=BLOCK, or specify the /FDL qualifier to the EXCHANGE command to specify the record type.

In addition to using the correct record type and carefully transferring the files, the data inside unformatted records may need to be converted. OpenVMS Fortran data files that contain VAX binary floating-point data must be converted before they can be accessed by a Compaq Fortran program. There are several methods:

• On an OpenVMS system, a Fortran program can convert files containing unformatted data to files containing formatted data. Once the files contain formatted data, they can be read by the appropriate Compaq Fortran programs. However, converting unformatted data to formatted data may result in a loss of accuracy for unformatted floating-point data.
• On an OpenVMS VAX system, a Compaq Fortran 77 program can read and write files containing unformatted data by using the Compaq Fortran conversion capabilities described in DEC Fortran User Manual for OpenVMS VAX Systems.
• On an OpenVMS Alpha system, a Fortran program can read and write files containing unformatted data by using the Compaq Fortran conversion capabilities described in the Compaq Fortran User Manual for OpenVMS Alpha Systems.
  A Compaq Fortran 77 for OpenVMS Alpha Systems program can also use the CVT$CONVERT_FLOAT routine to convert individual floating-point fields.
• On a Compaq Tru64 UNIX or Linux Alpha system, a Compaq Fortran program can read and write files containing unformatted data using the Compaq Fortran conversion capabilities described in Section 10.2. A program using the Compaq Fortran conversion capabilities can also convert such data to other formats.

If you need to convert unformatted floating-point data, keep in mind that Compaq Fortran 77 for OpenVMS VAX programs (VAX hardware) store the following:

• REAL*4 or COMPLEX*8 data in VAX F_float format
• REAL*8 or COMPLEX*16 data in either VAX D_float or G_float format
• REAL*16 data in VAX H_float format

In contrast, Compaq Fortran programs running on the Compaq Tru64 UNIX or Linux Alpha operating system on Alpha hardware store the following:

• REAL*4 or COMPLEX*8 data in IEEE S_float format
• REAL*8 or COMPLEX*16 data in IEEE T_float format
• REAL*16 data or COMPLEX*32 data in native (IEEE style) X_float format

Compaq Fortran 77 and Compaq Fortran for OpenVMS Alpha programs store floating-point data in the format specified by the /FLOAT qualifier:

• REAL*4 or COMPLEX*8 data in VAX F_float or IEEE S_float format
• REAL*8 or COMPLEX*16 data in VAX D_float, VAX G_float, or IEEE T_float format
• REAL*16 data or COMPLEX*32 data in native (IEEE style) X_float format

For information on Compaq Fortran data types, see Chapter 9.

For More Information:

• On Compaq Fortran I/O, see Chapter 7.
• About the Compaq Fortran language, see the Compaq Fortran Language Reference Manual.

This section provides information about VAX floating-point data formats. You can convert unformatted files from OpenVMS systems by using the methods described in Chapter 10.

On OpenVMS VAX systems, single-precision data (such as REAL*4) is stored in VAX F_float format and double-precision data (such as REAL*8) data can be stored in either VAX D_float or VAX G_float formats, depending on whether the /G_FLOATING qualifier was specified on the FORTRAN command line (see the DEC Fortran User Manual for OpenVMS VAX Systems).

On OpenVMS Alpha systems, you can specify the floating-point format in memory by using the /FLOAT qualifier (see the DEC Fortran User Manual for OpenVMS AXP Systems). Single-precision data on OpenVMS Alpha systems is stored in either VAX F_float or IEEE S_float formats; double-precision data can be stored in VAX D_float, VAX G_float, or IEEE T_float formats.

REAL*16 (extended precision) data is always stored in IEEE style X_float format on Alpha systems.

With VAX floating-point data types, the binary radix point is to the left of the most-significant bit.

A.4.3.1 VAX F_float REAL (KIND=4) or REAL*4

Intrinsic REAL (KIND=4) or REAL*4 F_float data occupies four contiguous bytes. Bits are labeled from the right, 0 through 31, as shown in Figure A-1.

Figure A-1 VAX F_float REAL (KIND =4) or REAL*4 Representation

The form of REAL (KIND=4) or REAL*4 F_float data is sign magnitude, where:

• Bit 15 is the sign bit (0 for positive numbers, 1 for negative numbers).
• Bits 14:7 are a binary exponent in excess 128 notation (binary exponents from --127 to 127 are represented by binary 1 to 255).
• Bits 6:0 and 31:16 are a normalized 24-bit fraction with the redundant most significant fraction bit not represented.

When converting unformatted F_float data from an OpenVMS system, the approximate range is 0.293873588E--38 to 1.7014117E38. The precision is approximately one part in 2**23, typically seven decimal digits.

A.4.3.2 VAX G_float REAL (KIND=8) or REAL*8

Intrinsic REAL (KIND=8) or REAL*8 (same as DOUBLE PRECISION) G_float data occupies eight contiguous bytes. The bits are labeled from the right, 0 through 63, as shown in Figure A-2.

Figure A-2 VAX G_float REAL (KIND =8) or REAL*8 Representation

The form of REAL (KIND=8) or REAL*8 G_float data is sign magnitude, where:

• Bit 15 is the sign bit (0 for positive numbers, 1 for negative numbers).
• Bits 14:4 are a binary exponent in excess 1024 notation (binary exponents from --1023 to 1023 are represented by the binary 1 to 2047).
• Bits 3:0 and 63:16 are a normalized 53-bit fraction with the redundant most significant fraction bit not represented.

When converting unformatted G_float data from an OpenVMS system, the approximate range is 0.5562684646268004D--308 to 0.89884656743115785407D308. The precision of G_float data is approximately one part in 2**52, typically 15 decimal digits.

A.4.3.3 VAX D_float REAL (KIND=8) or REAL*8

Intrinsic REAL (KIND=8) or REAL*8 (same as DOUBLE PRECISION) D_float data occupies eight contiguous bytes. Bits are labeled from the right, 0 through 63, as shown in Figure A-3.

Figure A-3 VAX D_float REAL (KIND =8) or REAL*8 Representation

The form of REAL (KIND=8) or REAL*8 D_float data is identical to an F_float real number, except for an additional 32 low-significance fraction bits. The exponent conventions and approximate range of values are the similar to those for F_float.

When converting unformatted D_float data from an OpenVMS system, the approximate range is 0.2938735877055719D--38 to 1.70141183460469229D38. The precision is approximately one part in 2**55, typically 16 decimal digits.

A.4.3.4 VAX F_float COMPLEX (KIND=4) or COMPLEX*8

Intrinsic COMPLEX (KIND=4) or COMPLEX*8 (single-precision COMPLEX) data in VAX F_float format occupies eight contiguous bytes containing a pair of REAL*4 values. The low-order four bytes contain REAL*4 data that represents the real part of the complex number. The high-order four bytes contain REAL (KIND=4) or REAL*4 data that represents the imaginary part of the complex number. Figure A-4 shows a COMPLEX*8 number in F_float format.

Figure A-4 VAX F_float COMPLEX (KIND =4) or COMPLEX*8 Representation

The limits for REAL (KIND=4) or REAL*4 apply to the two separate real and imaginary parts of a COMPLEX (KIND=4) or COMPLEX*8 number. Like REAL (KIND=4) or REAL*4 numbers, the sign bit representation is 0 (zero) for positive numbers and 1 for negative numbers.

A.4.3.5 VAX G_float and D_float COMPLEX (KIND=8) or COMPLEX*16

Intrinsic COMPLEX (KIND=8) or COMPLEX*16 (same as DOUBLE COMPLEX) data occupies 16 contiguous bytes containing a pair of REAL*8 or REAL (KIND=8) values. COMPLEX (KIND=8) or COMPLEX*16 data from an OpenVMS system is in one of the following REAL*8 or REAL (KIND=8) formats:

• VAX G_float format
• VAX D_float format

The low-order eight bytes contain REAL (KIND=8) or REAL*8 data that represents the real part of the complex data. The high-order eight bytes contain REAL (KIND=8) or REAL*8 data that represents the imaginary part of the complex data, as shown in Figure A-5 (for G_float) and Figure A-6 (for D_float).

Figure A-5 VAX G_float COMPLEX (KIND =8) or COMPLEX*16 Representation

Figure A-6 VAX D_float COMPLEX (KIND =8) or COMPLEX*16 Representation

The limits for REAL (KIND=8) or REAL*8 apply to the two separate real and imaginary parts of a COMPLEX (KIND=8) or COMPLEX*16 number. Like REAL (KIND=8) or REAL*8 numbers, the sign bit representation is 0 (zero) for positive numbers and 1 for negative numbers.

A.4.3.6 VAX H_float Representation

The REAL (KIND=16) or REAL*16 VAX H_float data format is used only on OpenVMS VAX systems. On Alpha systems, REAL (KIND=16) extended precision data is always stored in Alpha X_float format.

With VAX floating-point data types, the binary radix point is to the left of the most-significant bit.

As shown in Figure A-7, REAL*16 H_float data is 16 contiguous bytes starting on an arbitrary byte boundary. The bits are labeled from the right, 0 through 127.

Figure A-7 VAX H_float REAL*16 Representation (VAX Systems)

The form of an H_float REAL*16 data is sign magnitude with bit 15 the sign bit, bits 14:0 an excess 16384 binary exponent, and bits 127:16 a normalized 113-bit fraction with the redundant most significant fraction bit not represented.

The value of H_float data is in the approximate range 0.84*10**--4932 through 0.59*10**4932. The precision of H_float data is approximately one part in 2**112 or typically 33 decimal digits.

For More Information:

• On converting unformatted data files, see Chapter 10.
• On native floating-point ranges, see Table 9-1.