/osprey/crayf90/fe90/cf95.msgs

$
$  Copyright (C) 2006. QLogic Corporation. All Rights Reserved.
$
$  Copyright 2003, 2004, 2005, 2006 PathScale, Inc.  All Rights Reserved.
$
$  Copyright (C) 2000, 2001 Silicon Graphics, Inc.  All Rights Reserved.
$
$  This program is free software; you can redistribute it and/or modify it
$  under the terms of version 2 of the GNU General Public License as
$  published by the Free Software Foundation.
$
$  This program is distributed in the hope that it would be useful, but
$  WITHOUT ANY WARRANTY; without even the implied warranty of
$  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  
$
$  Further, this software is distributed without any warranty that it is
$  free of the rightful claim of any third person regarding infringement 
$  or the like.  Any license provided herein, whether implied or 
$  otherwise, applies only to this software file.  Patent licenses, if 
$  any, provided herein do not apply to combinations of this program with 
$  other software, or any other product whatsoever.  
$
$  You should have received a copy of the GNU General Public License along
$  with this program; if not, write the Free Software Foundation, Inc., 59
$  Temple Place - Suite 330, Boston MA 02111-1307, USA.
$
$  Contact information:  Silicon Graphics, Inc., 1600 Amphitheatre Pky,
$  Mountain View, CA 94043, or:
$
$  http://www.sgi.com
$
$  For further information regarding this notice, see:
$
$  http://oss.sgi.com/projects/GenInfo/NoticeExplan
$
$
$ |**************************************************************************|
$ |**                                                                      **|
$ |**                     MESSAGE AND EXPLANATION FILE                     **|
$ |**                                                                      **|
$ |**************************************************************************|
$ \**************************************************************************/
$
$
$ MESSAGE GROUPINGS:
$   0000 - 1999 : Front-end messages.  There is no particular order.
$                 The largest front-end message number is recorded in
$                 MAX_FE_MSG in messages.m.  If the front-end message number
$                 range changes, MAX_FE_MSG must also be updated.
$   2000 - 2099 : arith.a (folder) messages.  The largest folder message 
$                 number is recorded in MAX_FOLDER_MSG in messages.m.  If the 
$                 folder message number range changes, MAX_FOLDER_MSG must 
$                 also be updated.
$   2100 - 2499 : f90 command.
$   2500 - 4999 : Currently unused.
$   5000 - 5999 : Reserved for lint (will not be used by compilers).
$   6000 - 7999 : PDGCS messages.
$   8000 - 9999 : Compiler back-end (CCG, MPPCG, CMCS) messages.
$
$ Since message texts do not reflect the message level, a comment defining 
$ the level should precede each message entry.
$ --------------------------------------------------------------------------
$
$$ DO NOT REMOVE THIS LINE, psm USES IT TO FIND THE FIRST MESSAGE
$
$ Internal : 0001
$msg  0001 Message number %d, exceeds allowed range.
$nexp 0001
Internal : Message number %d, exceeds allowed range.
.PP
The message number exceeds the allowed maximum size.  It is either larger
than the maximum allowed message number for the front end or is less than 1.
.PP
This message should never be generated.  There is something wrong with the
message catalog or installation of the compiler.  Please notify your product
support organization with this error message number and any supporting
information.  This message does not indicate a problem with your code.
You may be able to change your code, so that the compiler does not try to
issue this message.
.ME
$
$ Internal : 0002
$msg  0002 Message level %d, is out of range.
$nexp 0002
Internal : Message level %d, is out of range.
.PP
The message level exceeds the allowed range.  It is not in the valid range
of message levels for the message system.
.PP
This message should never be generated.  There is something wrong with the
message catalog or installation of the compiler.  Please notify your product
support organization with this error message number and any supporting
information.  This message does not indicate a problem with your code.
You may be able to change your code, so that the compiler does not try to
issue this message.
.ME
$
$ Internal : 0003
$msg  0003 Message system failed trying to issue message %d
$nexp 0003
Internal : Message system failed trying to issue message %d
.PP
The message system was unable to issue the message indicated by the message
number included in this message.  This internal error can arise for a number
of possible reasons, including:

  * The message number is not in the message system messages file.

    It is probable that either the message number does not exist in the
    message catalog or the message catalog has been corrupted.

  * The message catalog is out of date or does not match the release of
    the compiler that tried to issue the message. 

  * The \*CNLSPATH\fR environment variable is not set correctly.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that the compiler does not try to issue this
message.
.ME
$
$ Error : 0004
$msg  0004 The MODULE PROCEDURE statement is only allowed in a generic interface block.
$nexp 0004
Error : The MODULE PROCEDURE statement is only allowed in a generic interface block.
.PP
A \*CMODULE PROCEDURE\fR statement can
only be specified in an interface block that has a \fIgeneric_spec\fR.
The compiler found
a \*CMODULE PROCEDURE\fR statement in an interface block that does not have a generic
specification.
.ME
$
$ Error : 0005
$msg  0005 The %s statement is out of order.
$nexp 0005
Error : The %s statement is out of order.
.PP
The current statement violates the Fortran standard statement ordering rules.
See the \fICFortran Language Reference Manual\fR, publication SR-3902, for details.
.ME
$
$ Error : 0006
$msg  0006 A keyword or identifier must follow the label or construct name.
$nexp 0006
Error : A keyword or identifier must follow the label or construct name.
.PP
The compiler has found a label or construct name.  A keyword or identifier 
must follow the label or construct name.  The following examples may clarify this.
.CS
10          ! This is an illegal statement
10 CONTINUE ! This is legal; a keyword follows 10
.CE
.ME
$
$ Error : 0007
$msg  0007 A construct name is not allowed on a %s statement.
$nexp 0007
Error : A construct name is not allowed on a %s statement.
.PP
A construct name can only appear on an \*CIF-THEN\fR statement, a \*CDO\fR statement, or
a \*CSELECT CASE\fR statement.  This error is generated when a construct name is
found on a statement other than those listed above.
.ME
$
$ Error : 0008
$msg  0008 The %s statement must precede all component definitions in derived type definition "%s".
$nexp 0008
Error : The %s statement must precede all component definitions in derived type definition "%s".
.PP
This \*CPRIVATE\fR or \*CSEQUENCE\fR statement follows one or more component definitions
in the  derived type definition.  The \*CPRIVATE\fR and \*CSEQUENCE\fR statements
must precede any component definitions.
.ME
$
$ Internal : 0009
$msg  0009 No forward reference entries exist for label %s even though it is undefined.
$nexp 0009
Internal : No forward reference entries exist for label %s even though it is undefined.
.PP
If an attribute entry exists for the label then it must be referenced in
the scoping unit.  If it was referenced but was never defined, there must be at
least one forward reference entry to represent the unresolved forward reference.
However, the forward reference chain for the label is empty.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0010
$msg  0010 Dummy argument "%s" is already defined in this argument list.
$nexp 0010
Error : Dummy argument "%s" is already defined in this argument list.
.PP
The name of a dummy argument appears more than once in the argument list on a
\*CFUNCTION\fR, \*CSUBROUTINE\fR, \*CENTRY\fR or statement function definition statement. 
.ME
$
$ Log_Warning : 0011
$msg  0011 The line size option is ignored because free source form is in effect.
$nexp 0011
Log_Warning : The line size option is ignored because free source form is in effect.
.PP
The line size option is ignored when the source form is set to \*Cfree\fR, because
the line may contain up to 132 characters.  With fixed source form the line
size option is valid, because the Fortran standard limits the line to 72
characters.  The option allows the line length to be extended to 80 characters.
.ME
$
$ Internal : 0012
$msg  0012 %s underflow.
$nexp 0012
Internal : %s underflow.
.PP
This message is generated from \*CPOP_SRC\fR macro in \*Csrc_input.m\fR.
The source stack has underflowed.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Ansi : 0013
$msg  0013 The Fortran standard requires multiple entry point function results "%s" and "%s" to both be default intrinsic types.
$nexp 0013
Ansi : The Fortran standard requires multiple entry point function results "%s" and "%s" to both be default intrinsic types.
.PP
A multiple entry point function (a function with one or more \*CENTRY\fR 
statements) can have entry points which return results of differing 
declared types only if all entry points return results which are scalar
and of type default real, default logical, double precision real, default
complex, or default integer.  If the entry point has the \*CPOINTER\fR attribute,
all entry points must be the same type and be pointers. The compiler allows
the entry points with the same declared types to have different kind parameters.
.ME
$
$ Limit : 0014
$msg  0014 Insufficient memory is available for compilation to continue.
$nexp 0014
Limit : Insufficient memory is available for compilation to continue.
.PP
The compiler ran out of memory during compilation of the program.  Check
with your product support organization and request more memory for the job.
Remember that program compilation may be only one of several processes
executing in the job space.
.ME
$
$ Error : 0015
$msg  0015 The %s statement is not allowed in a block data program unit.
$nexp 0015
Error : The %s statement is not allowed in a block data program unit.
.PP
Derived type definitions, type declaration statement and
the following statements are allowed in a block data program unit:
.CS
	COMMON
	DATA
	DIMENSION
	END BLOCK DATA
	EQUIVALENCE
	IMPLICIT
	IMPLICIT NONE
	INTRINSIC
	PARAMETER
	POINTER
	SAVE
	TARGET
	TASK COMMON
	USE
.CE
.PP
\fRThe \*CSEQUENCE\fR statement must be specified in a derived type definition
to be in a block data program unit.
.ME
$
$ Error : 0016
$msg  0016 The %s statement is not allowed in a main program unit.
$nexp 0016
Error : The %s statement is not allowed in a main program unit.
.PP
The following statements are not allowed in a main program unit:
.CS
	ENTRY
	INTENT
	OPTIONAL
	PRIVATE
	PUBLIC
	RETURN
.CE
.PP
These statements must be inside a construct or contained scoping unit to be in a main program unit:
.BL
\*CSEQUENCE\fR, \*CEND TYPE \fRin a derived type definition
.BL
\*CMODULE PROCEDURE \fRin a generic interface block
.BL
\*CELSE\fR, \*CELSE IF\fR, \*CEND IF\fR in an \*CIF\fR construct 
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct 
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct 
.BL
\*CCYCLE\fR, \*CEXIT\fR, \*CEND DO\fR in a \*CDO\fR construct  
.BL
\*CFUNCTION\fR, \*CSUBROUTINE\fR in an interface block
.BL
\*CEND FUNCTION\fR in an interface body 
.BL
\*CEND SUBROUTINE\fR in an interface body 
.BL
\*CEND INTERFACE \fR in an interface block
.ME
$
$ Error : 0017
$msg  0017 Cannot open "%s" file.  It is needed for debugging or tracing.
$nexp 0017
Error : Cannot open "%s" file.  It is needed for debugging or tracing.
.PP
The compiler cannot open the debug file or the trace file.
.ME
$
$ Error : 0018
$msg  0018 The %s statement must not follow a MODULE PROCEDURE statement in a generic interface block.
$nexp 0018
Error : The %s statement must not follow a MODULE PROCEDURE statement in a generic interface block.
.PP
The syntax for an interface block is:
.CS
        interface_stmt
          [interface_body]...
          [module_procedure_stmt]...
        end_interface_stmt
.CE
.ME
$
$ Error : 0019
$msg  0019 The %s statement is not allowed in a module.
$nexp 0019
Error : The %s statement is not allowed in a module.
.PP
Derived-type definitions, type declaration statements and the
following statements are allowed in a module (before the \*CEND MODULE\fR or
the \*CCONTAINS\fR statement):
.CS
	ALLOCATABLE
	COMMON
	DATA
	DIMENSION
	EQUIVALENCE
	EXTERNAL
	IMPLICIT
	IMPLICIT NONE
	INTERFACE
	INTRINSIC
	NAMELIST
	PARAMETER
	POINTER
	PUBLIC
	PRIVATE
	SAVE
	TARGET
	USE
.CE
.PP
\fRThese statements must be in an interface block or contained scoping unit to be in a module program unit.
.BL
\*CMODULE PROCEDURE\fR in a generic interface block
.BL
\*CSEQUENCE\fR, \*CEND TYPE\fR in a derived-type definition
.BL
\*CFUNCTION\fR, \*CSUBROUTINE\fR, \*CEND FUNCTION\fR, \*CEND SUBROUTINE\fR, \*CEND INTERFACE\fR
in an interface block
.ME
$
$ Ansi : 0020
$msg  0020 The use of DOUBLE COMPLEX is an extension to the Fortran standard.
$nexp 0020
Ansi : The use of DOUBLE COMPLEX is an extension to the Fortran standard.
.PP
The \*CDOUBLE COMPLEX\fR statement is not part of the Fortran standard.
.ME
$
$ Error : 0021
$msg  0021 The type attributes for entry points "%s" and "%s" conflict.
$nexp 0021
Error : The type attributes for entry points "%s" and "%s" conflict.
.PP
If the result of any entry point of a multiple entry function (a function with
one or more \*CENTRY\fR statements) is of type character, or a derived type 
then all entry points in that function must return a result that is of the same
type.  If the function result is of type character, all entry points must
return a result with the same declared character length.  If the result is of
a derived type, then all entries must be the same derived type.
.ME
$
$ Ansi : 0022
$msg  0022 The type of function result "%s" must be a default intrinsic type if it does not match the type of other entry points.
$nexp 0022
Ansi : The type of function result "%s" must be a default intrinsic type if it does not match the type of other entry points.
.PP
A multiple entry point function (a function with one or more \*CENTRY\fR 
statements) may have entry points which return results of differing
declared types only if all entry points return  results which are scalar
and of type default real, default logical, double precision real, default
complex, or default integer.  If the entry point has the \*CPOINTER\fR attribute,
all entry points must be of the same type and have the \*CPOINTER\fR attribute.
.PP
The compiler allows the mix of default and non-default real, logical, double 
precision, complex and integer.
.ME
$
$ Error : 0023
$msg  0023 Label %s is not defined in this scoping unit.
$nexp 0023
Error : Label %s is not defined in this scoping unit.
.PP
When a statement label is referenced, the label must be the statement label of a
branch target statement that appears in the same scoping unit as the label 
reference.  For example, if the following \*CGO TO\fR statement appears in a subprogram,
label 10 must be defined in the same subprogram.
.CS
     GO TO 10
.CE
.ME
$
$ Error : 0024
$msg  0024 The character constant "%s", for the I/O control information list item %s, is not allowed.
$nexp 0024
Error : The character constant "%s", for the I/O control information list item %s, is not allowed.
.PP
Many I/O control information list items allow only a limited set of valid
character constants.  A character constant was encountered that is not
valid for this control list item.  Following are the I/O control items and
their valid character constants:
.CS
OPEN I/O control                         
 list item                   Character constants
 STATUS    OLD, NEW, SCRATCH, REPLACE, UNKNOWN
 ACCESS    SEQUENTIAL, DIRECT
 FORM      FORMATTED, UNFORMATTED
 BLANK     NULL, ZERO
 POSITION  ASIS, REWIND, APPEND
 ACTION    READ, WRITE, READWRITE
 DELIM     APOSTROPHE, QUOTE, NONE
 PAD       YES, NO
.CE
.PP
For the \*CCLOSE\fR I/O control list item \*CSTATUS\fR, the 
allowed character constants are \*CKEEP\fR and \*CDELETE\fR.
.PP
For the \*CREAD/WRITE\fR I/O control list item \*CADVANCE\fR, 
the allowed character constants are \*CYES\fR and \*CNO\fR.
.ME
$
$ Error : 0025
$msg  0025 This statement is not allowed in a derived type definition.
$nexp 0025
Error : This statement is not allowed in a derived type definition.
.PP
The only statements allowed in a derived type definition are \*CPRIVATE\fR, \*CSEQUENCE\fR
and the component definition statement.  \*CPRIVATE\fR is only allowed if the
derived type definition is in a module program unit.
.ME
$
$ Error : 0026
$msg  0026 The %s statement is not allowed in an interface block; expected SUBROUTINE, FUNCTION, MODULE PROCEDURE or END statement.
$nexp 0026
Error : The %s statement is not allowed in an interface block; expected SUBROUTINE, FUNCTION, MODULE PROCEDURE or END statement.
.PP
The only statements allowed in an interface block are \*CFUNCTION\fR,
\*CSUBROUTINE\fR, \*CMODULE PROCEDURE\fR, and \*CEND INTERFACE\fR.
\*CFUNCTION\fR and \*CSUBROUTINE\fR start an interface body.
\*CEND INTERFACE\fR ends the interface block.
\*CMODULE PROCEDURE\fR can be specified if the interface block is a generic
interface.  A generic interface contains a list of module procedures that can
be referenced with the generic name.
.ME
$
$ Error : 0027
$msg  0027 Function entry "%s" has result name "%s".  It must be used when declaring the result to be an array.
$nexp 0027
Error : Function entry "%s" has result name "%s".  It must be used when declaring the result to be an array.
.PP
If a function has a result name, the result name must be used when declaring
the result to be an array.
.ME
$
$ Error : 0028
$msg  0028 A trailing ")" is missing in the program string.
$nexp 0028
Error : A trailing ")" is missing in the program string.
.PP
The parentheses groups do not match in the optional parenthesized list that can appear on 
the \*CPROGRAM\fR statement.
.ME
$
$ Error : 0029
$msg  0029 The maximum number of unnamed block data program units is 26 and has been exceeded.
$nexp 0029
Error : The maximum number of unnamed block data program units is 26 and has been exceeded.
.PP
The Fortran standard limits unnamed block data program units to 1 in an
executable program.  However, the compiler extends this limit to 26.  The
compiler detected more than 26 unnamed block data program units.
.ME
$
$ Ansi : 0030
$msg  0030 The Fortran standard does not allow more than one unnamed BLOCK DATA program unit.
$nexp 0030
Ansi : The Fortran standard does not allow more than one unnamed BLOCK DATA program unit.
.PP
The Fortran standard does not allow more than one unnamed block data 
program unit.
.ME
$
$ Ansi : 0031
$msg  0031 Arguments to a PROGRAM statement are an extension to the Fortran standard.
$nexp 0031
Ansi : Arguments to a PROGRAM statement are an extension to the Fortran standard.
.PP
A \*CPROGRAM\fR statement contains an argument list enclosed in parentheses.  The
Fortran standard does not provide for actual arguments to a main program.
.ME
$
$ Error : 0032
$msg  0032 The type specification "CHARACTER*(*)" is not allowed in an IMPLICIT statement.
$nexp 0032
Error : The type specification "CHARACTER*(*)" is not allowed in an IMPLICIT statement.
.PP
Assumed-length character type declarations are not allowed in \*CIMPLICIT\fR statements.
.ME
$
$ Error : 0033
$msg  0033 If the component is being typed as the derived type being defined, it must have the POINTER attribute.
$nexp 0033
Error : If the component is being typed as the derived type being defined, it must have the POINTER attribute.
.PP
The Fortran standard requires that a component of a derived type may not be
declared to be of the derived type, unless it is a pointer.  For example:
.CS
      TYPE BB
        TYPE(BB) :: A
      END TYPE
.CE
.PP
Component \*CA\fR is illegal, because it is defined as the type being declared (\*CBB\fR).
.CS
      TYPE ZZ
        TYPE(ZZ), POINTER :: A
      END TYPE 
.CE
.PP
Component \*CA\fR is legal here, because it is a pointer to a structure of the type being declared (\*CZZ\fR).
.ME
$
$ Internal : 0034
$msg  0034 The name or attribute index is not valid.  It is less than 0 or greater than the table index.
$nexp 0034
Internal : The name or attribute index is not valid.  It is less than 0 or greater than the table index.
.PP
The local name table index or attribute table index is not valid.  It is
either less than zero or greater than the last table index entered.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0035
$msg  0035 The character length of "%s" has exceeded the maximum length of %d.
$nexp 0035
Error : The character length of "%s" has exceeded the maximum length of %d.
.PP
The maximum character length was exceeded.
.ME
$
$ Error : 0036
$msg  0036 Function entry "%s" has result name "%s", which must be used when declaring the result to be a pointer.
$nexp 0036
Error : Function entry "%s" has result name "%s", which must be used when declaring the result to be a pointer.
.PP
If a function has a result name, the result name must be used when declaring
the result to be a pointer.
.ME
$
$ Ansi : 0037
$msg  0037 Input lines greater than 72 characters long are an extension to the Fortran standard.
$nexp 0037
Ansi : Input lines greater than 72 characters long are an extension to the Fortran standard.
.PP
The compiler detected an input line greater than 72 characters.  The Fortran
standard does not provide for input lines greater than 72 characters long
in fixed source form.
.ME
$
$ Internal : 0038
$msg  0038 This character is not valid.  Expected keyword DO following failed parse of type spec DOUBLE.
$nexp 0038
Internal : This character is not valid.  Expected keyword DO following failed parse of type spec DOUBLE.
.PP
This error is generated when \*CDO\fR is not encountered as expected.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0039
$msg  0039 Namelist group name "%s" was referenced in an I/O statement.  It must not be specified in this NAMELIST statement.
$nexp 0039
Error : Namelist group name "%s" was referenced in an I/O statement.  It must not be specified in this NAMELIST statement.
.PP
Namelist statements may be mixed in with executable statements.  However, all
definitions of a specific namelist group must occur before the first reference to the
namelist group by an I/O statement.
.ME
$
$ Error : 0040
$msg  0040 If this main program has no PROGRAM statement specified, then no name can be specified on the END PROGRAM statement.
$nexp 0040
Error : If this main program has no PROGRAM statement specified, then no name can be specified on the END PROGRAM statement.
.PP
The compiler detected an \*CEND PROGRAM\fR statement followed by a name for 
a program unit.  This program unit has no \*CPROGRAM\fR statement
and therefore no name.
.ME
$
$ Error : 0041
$msg  0041 The %s statement must only be specified once for derived type definition "%s".
$nexp 0041
Error : The %s statement must only be specified once for derived type definition "%s".
.PP
The \*CPRIVATE\fR and \*CSEQUENCE\fR statements must only be specified once in a derived type definition.
.ME
$
$ Internal : 0042
$msg  0042 Attribute entry %d has a variant problem with field %s.
$nexp 0042
Internal : Attribute entry %d has a variant problem with field %s.
.PP
The compiler detected a variant problem.  A field is being accessed in the
attribute table, but the wrong variant is active.  The field may actually be
in the attribute auxiliary table, but it is checking the attribute table
variant.  In general whether the field actually resides in the attribute
table or the attribute auxiliary table should make no difference in resolving
the problem.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0043
$msg  0043 Only one %s statement is allowed in each %s construct.
$nexp 0043
Error : Only one %s statement is allowed in each %s construct.
.PP
Each \*CIF\fR construct must only have one \*CELSE\fR block.  Each \*CWHERE\fR construct must only
have one \*CELSEWHERE\fR block.  The compiler detected more than one \*CELSE\fR
or \*CELSE WHERE\fR statement in the \*CIF\fR or \*CWHERE\fR construct.
.ME
$
$ Error : 0044
$msg  0044 The entry point name of the host, "%s", must not be redefined in an interface body.
$nexp 0044
Error : The entry point name of the host, "%s", must not be redefined in an interface body.
.PP
Within an external subprogram one of the entry point names was specified
in an explicit interface in a contained subprogram.  Entry point names
already have explicit interfaces in the host, so they may not have another
interface specified in a contained procedure.  The following example illustrates an
illegal redefinition.
.CS
   FUNCTION ABC()
     CONTAINS
       SUBROUTINE SAM()
         INTERFACE
            FUNCTION ABC() !  Illegal re-definition
            END FUNCTION   !  of ABC's interface
         END INTERFACE
       END SUBROUTINE
   END FUNCTION
.CE
.ME
$
$ Error : 0045
$msg  0045 Component "%s" is private type "%s", therefore derived type "%s" must also be private or have private components.
$nexp 0045
Error : Component "%s" is private type "%s", therefore derived type "%s" must also be private or have private components.
.PP
If a component of a derived type is declared to be a private type, either
the derived type definition must contain the \*CPRIVATE\fR statement or the
derived type must be private.
.ME
$
$ Ansi : 0046
$msg  0046 The TASK COMMON statement is an extension to the Fortran standard.
$nexp 0046
Ansi : The TASK COMMON statement is an extension to the Fortran standard.
.PP
The \*CTASK COMMON\fR statement is an extension to the Fortran standard.
.ME
$
$ Error : 0047
$msg  0047 A '*' must only be used with the UNIT or FMT I/O control information list items.
$nexp 0047
Error : A '*' must only be used with the UNIT or FMT I/O control information list items.
.PP
An asterisk (*) was encountered in an I/O control information list in a position 
or with a keyword other than \*CUNIT\fR or \*CFMT\fR.
.ME
$
$ Error : 0048
$msg  0048 The DO variable must not be defined while it is active.
$nexp 0048
Error : The DO variable must not be defined while it is active.
.PP
The name of an active \*CDO\fR variable appeared in a statement that would change the
\*CDO\fR variable's value.  The \*CDO\fR variable can be used but not redefined within the
range of its loop.
.ME
$
$ Log_Error : 0049
$msg  0049 Cannot open source file "%s".
$nexp 0049
Log_Error : Cannot open source file "%s".
.PP
The compiler cannot open the source file.
.ME
$
$ Internal : 0050
$msg  0050 Attempt to read past end of file.
$nexp 0050
Internal : Attempt to read past end of file.
.PP
This is generated from \*Cfixed_get_line\fR in \*Csrc_input.c\fR.
The compiler has attempted to read past EOF.
The source input file may be corrupted.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0051
$msg  0051 A compiler directive must not be followed by a Fortran continuation line.
$nexp 0051
Error : A compiler directive must not be followed by a Fortran continuation line.
.PP
A Fortran continuation source line was encountered following
a compiler directive. Compiler directives may be continued,
but they require the directive sentinal at the beginning
of the line. Compiler directives may not be embedded within
a continued Fortran source line.
.ME
$
$ Ansi : 0052
$msg  0052 More than %d continuation lines is an extension to the Fortran standard in %s source form.
$nexp 0052
Ansi : More than %d continuation lines is an extension to the Fortran standard in %s source form.
.PP
The Fortran standard only allows for 19 continuation lines in fixed
source form and 39 continuation lines in free form.
.ME
$
$ Error : 0053
$msg  0053 The first line of an include file must not be a continuation line.
$nexp 0053
Error : The first line of an include file must not be a continuation line.
.PP
The first line of an include file must not be a continuation line.
.ME
$
$ Error : 0054
$msg  0054 The source line following an INCLUDE line must not be a continuation line.
$nexp 0054
Error : The source line following an INCLUDE line must not be a continuation line.
.PP
The line following an \*CINCLUDE\fR line must not be a continuation line.
.ME
$
$ Warning : 0055
$msg  0055 The file line does not end with a newline.
$nexp 0055
Warning : The file line does not end with a newline.
.PP
The compiler expects all lines in the source file to end with a newline
character.  This line does not end with a newline character.
.ME
$
$ Error : 0056
$msg  0056 A continuation line must not contain a label.
$nexp 0056
Error : A continuation line must not contain a label.
.PP
It is illegal to define a label on a continuation line.
.ME
$
$ Error : 0057
$msg  0057 The file name length exceeds the maximum of %d characters.
$nexp 0057
Error : The file name length exceeds the maximum of %d characters.
.PP
The length of the name of the file exceeds the maximum length.
Shorten the name of the file.
.ME
$
$ Error : 0058
$msg  0058 The INCLUDE file name is missing.
$nexp 0058
Error : The INCLUDE file name is missing.
.PP
An \*CINCLUDE\fR line does not follow the expected form.  In the following
syntax, \fIfile\fR is the name of the file to be included:
.CS
      INCLUDE 'file'
.CE
.PP
The file name must be enclosed in apostrophes or quotation marks.  Only an embedded comment can
follow the closing delimiter.
.ME
$
$ Error : 0059
$msg  0059 There is a missing delimiter on an INCLUDE file name.
$nexp 0059
Error : There is a missing delimiter on an INCLUDE file name.
.PP
An \*CINCLUDE\fR line does not follow the expected form.  In the following
syntax, \*Vfile\fR is the name of the file to be included:
.CS
      INCLUDE 'file'
.CE
.PP
The file name must be enclosed in apostrophes or quotation marks.  Only embedded comments can
follow the closing delimiter.
.ME
$
$ Error : 0060
$msg  0060 The source following the INCLUDE file name is not a comment.
$nexp 0060
Error : The source following the INCLUDE file name is not a comment.
.PP
An \*CINCLUDE\fR line does not follow the expected form.  In the following
syntax, \*Vfile\fR is the name of the file to be included:
.CS
      INCLUDE 'file'
.CE
.PP
The file name must be enclosed in apostrophes or quotation marks.  Only an embedded comment can
follow the closing delimiter.
.ME
$
$ Error : 0061
$msg  0061 The INCLUDE line must not contain a statement label.
$nexp 0061
Error : The INCLUDE line must not contain a statement label.
.PP
An \*CINCLUDE\fR line does not follow the expected form.  In the following
syntax, \*Vfile\fR is the name of the file to be included:
.CS
      INCLUDE 'file'
.CE
.PP
The file name must be enclosed in apostrophes or quotation marks.  Only an embedded comment can
follow the closing delimiter.
.ME
$
$ Error : 0062
$msg  0062 Operator "%s" is not intrinsic.  All operations in a DATA statement expression must be intrinsic.
$nexp 0062
Error : Operator "%s" is not intrinsic.  All operations in a DATA statement expression must be intrinsic.
.PP
The \*CDATA\fR statement has the general form:
.CS
  data_stmt_object_list /data_stmt_value_list/
.CE
.PP
where a \*Vdata_stmt_object\fR can be a variable or an implied-\*CDO\fR.  In a 
\*Vdata_stmt_object\fR that is a variable; each subscript, section subscript, 
substring starting point, and substring ending point must be an expression 
where each operation is intrinsic.
.PP
The \*CDATA\fR statement implied-\*CDO\fR has the form:
.CS
  (\*Vdata_i_do_object_list\*C, \*Vdata_i_do_variable\*C = 
	\*Vscalar_int_expr\*C, \*Vscalar_int_expr\*C
	[, \*Vscalar_int_expr\*C])\fR
.CE
.PP
All operations in each subscript of a \*Vdata_i_do_object\fR subscript list must be
intrinsic.  Likewise, all operations in each loop control \*Vscalar_int_expr\fR must
be intrinsic.
.ME
$
$ Error : 0063
$msg  0063 Cannot open INCLUDE file "%s".
$nexp 0063
Error : Cannot open INCLUDE file "%s".
.PP
The compiler cannot open the specified \*CINCLUDE\fR file.
.ME
$
$ Error : 0064
$msg  0064 Recursive use of INCLUDE file "%s".
$nexp 0064
Error : Recursive use of INCLUDE file "%s".
.PP
An \*CINCLUDE\fR file must not include itself.
.ME
$
$ Error : 0065
$msg  0065 A defined operator exceeds the maximum length of 31 characters.
$nexp 0065
Error : A defined operator exceeds the maximum length of 31 characters.
.PP
A defined operator must not exceed 31 characters.
.ME
$
$ Error : 0066
$msg  0066 A defined operator is missing the "." delimiter.
$nexp 0066
Error : A defined operator is missing the "." delimiter.
.PP
A defined operator must start and end with a period delimiter.  This defined
operator is missing its end delimiter.
.ME
$
$ Error : 0067
$msg  0067 Truncating identifier which exceeds the maximum of 63 characters.
$nexp 0067
Error : Truncating identifier which exceeds the maximum of 63 characters.
.PP
Due to an internal limit, when the compiler reads an identifier, it discards
characters beyond the 63rd.
.ME
$
$ Error : 0068
$msg  0068 The label length exceeds the maximum of 5 digits.
$nexp 0068
Error : The label length exceeds the maximum of 5 digits.
.PP
The Fortran standard requires labels to be 5 or less digits.  The compiler has 
detected a label with more than 5 digits.
.ME
$
$ Error : 0069
$msg  0069 A label must have at least one nonzero digit.
$nexp 0069
Error : A label must have at least one nonzero digit.
.PP
The compiler detected a label consisting of all zeros.  A label must have at 
least one nonzero digit.
.ME
$
$ Error : 0070
$msg  0070 Each control item specifier must be specified only once.
$nexp 0070
Error : Each control item specifier must be specified only once.
.PP
An I/O statement contains the same specifier more than once.  For
example, the following would cause this message to be issued, because
\*CUNIT=\fR is specified twice.
.CS
      WRITE (UNIT=101,FMT=10,UNIT=101)
.CE
.ME
$
$ Ansi : 0071
$msg  0071 The Fortran standard requires continuation lines to contain text following the & symbol in free source form.
$nexp 0071
Ansi : The Fortran standard requires continuation lines to contain text following the & symbol in free source form.
.PP
The Fortran standard requires continuation lines to contain characters other
than the continuation character.
.ME
$
$ Internal : 0072
$msg  0072 Object "%s" is host associated to itself.  Attr index is %d.
$nexp 0072
Internal : Object "%s" is host associated to itself.  Attr index is %d.
.PP
AT_ATTR_LINK(attr_idx) = attr_idx.  This causes all kinds of internal loops
during compilation.  An attribute should never link to itself.

This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that the compiler does not try to issue this
message.
.ME
$
$ Error : 0073
$msg  0073 This control item specifier is not allowed for this I/O statement.
$nexp 0073
Error : This control item specifier is not allowed for this I/O statement.
.PP
An I/O statement contains an inappropriate or unrecognized specifier.
The specifiers allowed on a \*CREAD\fR or \*CWRITE\fR statement are: 
.CS
UNIT    FMT         END         REC
ERR     IOSTAT      NML         ADVANCE
SIZE    EOR
.CE
The specifiers allowed on an \*COPEN\fR statement are: 
.CS
UNIT     FILE       STATUS      IOSTAT
ERR      FORM       ACCESS      RECL
BLANK    POSITION   ACTION      DELIM
PAD
.CE
The specifiers allowed on a \*CCLOSE\fR statement are:
.CS
UNIT     IOSTAT     STATUS      ERR
.CE
The specifiers allowed on an \*CINQUIRE\fR statement are:
.CS
FILE     IOSTAT     ERR         EXIST
OPENED   NUMBER     NAMED       RECL
NEXTREC  NAME       ACCESS      SEQUENTIAL
FORM     DIRECT     FORMATTED   UNFORMATTED
BLANK    UNIT       POSITION    ACTION
READ     WRITE      READWRITE   DELIM
PAD
.CE
.ME
$
$ Log_Warning : 0074
$msg  0074 Assembly language output (-eS or -S filename) disables binary output.  Assembly language output selected.
$nexp 0074
Log_Warning : Assembly language output (-eS or -S filename) disables binary output.  Assembly language output selected.
.PP
The compiler cannot have an assembly language file and a binary file open at the
same time.  The binary output file is disabled, because assembly language 
output is requested on the command line (\*C-eS\fR or \*C-S\fR \*Vfilename\fR).
The binary output file is on by default; therefore, specifying an assembly
language file on the command line (without turning off the binary file)
results in this warning message.
.CS
\*C-eS \fIfile.\*Cf     \fRWarning message issued
\*C-dB -eS \fIfile.\*Cf \fRNo warning, because \*C-dB\*C 
                        \fRturns off binary output\*C
\*C-eS -dS \fIfile.\*Cf \fRWarning message for \*C-eS.\*C
                        \fRBoth assembly and binary output are disabled.
.CE
.ME
$
$ Log_Warning : 0075
$msg  0075 Rounding/truncation conflict detected.  %s option selected.
$nexp 0075
Log_Warning : Rounding/truncation conflict detected.  %s option selected.
.PP
The compiler detected a rounding/truncation conflict on the command line.
The last option specified is used.
.ME
$
$ Log_Warning : 0076
$msg  0076 Inline option conflict detected.  Automatic mode selected.
$nexp 0076
Log_Warning : Inline option conflict detected.  Automatic mode selected.
.PP
This message is issued when both \*C-O inline[0-3]\fR and \*C-I \fIfilename\fR are
specified on the command line.  Automatic inlining is selected.
.ME
$
$ Log_Error : 0077
$msg  0077 Command line has an unknown option "%c".
$nexp 0077
Log_Error : Command line has an unknown option "%c".
.PP
The compiler detected an unknown option on the command line.
.ME
$
$ Log_Error : 0078
$msg  0078 -%c option has an illegal argument "%s".
$nexp 0078
Log_Error : -%c option has an illegal argument "%s".
.PP
The argument for the specified option is illegal.
.ME
$
$ Log_Error : 0079
$msg  0079 Options are not allowed after the input file name.
$nexp 0079
Log_Error : Options are not allowed after the input file name.
.PP
Nothing must follow the source input file name on the command line.
.ME
$
$ Log_Error : 0080
$msg  0080 Name must be specified for %s file when input is from stdin.
$nexp 0080
Log_Error : Name must be specified for %s file when input is from stdin.
.PP
If the source is input from \*Cstdin\fR, a name must be specified for the binary
or the assembly language file (depending on which one is requested).
.ME
$
$ Internal : 0081
$msg  0081 Unexpected attribute entry "%s" (index = %d) has been found in the local scope during use processing.
$nexp 0081
Internal : Unexpected attribute entry "%s" (index = %d) has been found in the local scope during use processing.
.PP
Routine not_visible_semantics has found an attribute in the symbol table
The only things allowed to be in the local scope during use processing
are the procedure name, function result name, dummy argument names,
use associated objects and objects used in declaration bounds expressions.
Other items that may be present, must have a declaration error.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that the compiler does not try to issue this
message.
.ME
$
$ Log_Warning : 0082
$msg  0082 Debugging (-G) requires binary output (-eB).  Binary output is enabled.
$nexp 0082
Log_Warning : Debugging (-G) requires binary output (-eB).  Binary output is enabled.
.PP
If binary output is disabled (using the \*C-dB\fR option), the compiler can only
provide semantic and syntactic error checking.  Debugging requires binary
output to be generated.  Although binary output is on by default, it can be
turned off by specifying \*C-dB\fR or requesting an assembly language file 
(\*C-eS\fR or \*C-S\fR \*Vfilename\fR).
Following are examples to clarify this:
.nf
 \*C-dB -G0\fR   illegal - binary output is off, debug is on
 \*C-eS -G0\fR   illegal - assembly language file is on, debug is on
.fi
.ME
$
$ Error : 0083
$msg  0083 This token is missing the %c delimiter.
$nexp 0083
Error : This token is missing the %c delimiter.
.PP
The compiler detected the start of a token delimited by an apostrophe or 
a quotation mark.  The matching end delimiter is missing.
.ME
$
$ Error : 0084
$msg  0084 This Hollerith constant should have %d characters, but it only contains %d characters.
$nexp 0084
Error : This Hollerith constant should have %d characters, but it only contains %d characters.
.PP
The compiler detected an inconsistency between the count and the
number of characters actually specified in a Hollerith constant.  An example of this would be:
.CS
 4H123   \fRThe constant contains only 3 characters,\*C
         \fRbut the 4H indicates there should be 4.
.CE
.ME
$
$ Error : 0085
$msg  0085 The number of characters in the Hollerith specifier must be greater than zero.
$nexp 0085
Error : The number of characters in the Hollerith specifier must be greater than zero.
.PP
Specifying 0H is not valid.  The Hollerith constant must have a length greater than 0.
.ME
$
$ Error : 0086
$msg  0086 The END statement of this %s must be followed by %s.
$nexp 0086
Error : The END statement of this %s must be followed by %s.
.PP
The Fortran standard requires that the \*CEND\fR statement for a module 
procedure or internal procedure be \*CEND FUNCTION\fR if it is a function, and \*CEND SUBROUTINE\fR
if it is a subroutine.  The compiler detected only an \*CEND\fR statement.
.ME
$
$ Error : 0087
$msg  0087 In a real constant with a D or E following the decimal field, the D or E must be followed by an exponent.
$nexp 0087
Error : In a real constant with a D or E following the decimal field, the D or E must be followed by an exponent.
.PP
The compiler detected a missing exponent field in a real constant.
The syntax for a real constant is
.CS
   [\fIsign\*C] \fIreal_literal_constant\*C
.CE
.PP
\*Vreal_literal_constant\fR is 
\*Vsignificand\fR [\*Vexponent_letter exponent\fR] [_\*Vkind_param\fR]
.PP
     or
.PP
\*Vdigit_string exponent_letter exponent\fR [_\*Vkind_param\fR]
.PP
\*Vsignificand\fR is \*Vdigit_string\fR.[\*Vdigit_string\fR]
.PP
  or
.PP
\*Vdigit_string\fR
.PP
\*Vexponent_letter\fR is E or \*Vexponent\fR is \*Vsigned_digit_string\fR
.ME
$
$ Error : 0088
$msg  0088 The kind parameter suffix is not valid on a double precision real constant form.
$nexp 0088
Error : The kind parameter suffix is not valid on a double precision real constant form.
.PP
The compiler detected a kind parameter following a D exponent.
.PP
The syntax for a real constant is
.CS
[\*Vdigit_str\*C].\*Vdigit_str\*C[\*Vexponent_letter exponent\fR\*C]
[_\*Vkind_param\fR\*C]\fR
.CE
.PP
\*Vexponent_letter\fR is D or E; \*Vexponent\fR is \*Vsigned_digit_string\fR
with the following constraint:
if both \*Vkind_param\fR and \*Vexponent_letter\fR are present, \fIexponent_letter\fR must be E.
.ME
$
$ Error : 0089
$msg  0089 The kind parameter on a literal constant must be an integer number or named constant.
$nexp 0089
Error : The kind parameter on a literal constant must be an integer number or named constant.
.PP
If a kind parameter is specified on a literal constant, it must be an integer
number or named constant of type integer.  For example:
.CS
  r = 10.0_8    \fR! is a valid kind parameter\*C
  END

  PARAMETER (KIND=4)
  j = 10_KIND   \fR! is a valid kind parameter\*C
  END

  r = 10.0_IT   \fR! is not valid, because \*CIT\*C
                \fR! is not an integer constant value.
.CE
.ME
$
$ Ansi : 0090
$msg  0090 Boolean constants are an extension to the Fortran standard.
$nexp 0090
Ansi : Boolean constants are an extension to the Fortran standard.
.PP
Octal and hexadecimal values may be specified as Boolean constants in the
following ways:
.nf
.br
\*C1000B -->    \fRoctal   
.br
\*C'1a3'X -->   \fRhexadecimal
.fi
.PP
The Fortran standard does not support Boolean constants.
.ME
$
$ Error : 0091
$msg  0091 This constant contains %d characters.  The number of allowed characters must be > 0 and <= %d.
$nexp 0091
Error : This constant contains %d characters.  The number of allowed characters must be > 0 and <= %d.
.PP
Both nonstandard Boolean constants and Fortran BOZ constants
have limits to their size.  These limits are determined by machine type.

If the machine type has 64-bit words, the maximum lengths are as follows:

   For octal constants:        22
   For hexadecimal constants:  16
   For binary constants:       64

If the octal constant contains 22 characters, the high order (leftmost) digit
must only be a 0 or a 1.

If the machine type has 32-bit words, the maximum lengths are as follows:

   For octal constants:        11
   For hexadecimal constants:   8
   For binary constants:       32

If the octal constant contains 11 characters, the high order (leftmost) digit
must only be 0-3.
.ME
$
$ Error : 0092
$msg  0092 The value of this constant exceeds the range allowed on this machine type.
$nexp 0092
Error : The value of this constant exceeds the range allowed on this machine type.
.PP
Nonstandard Boolean constants and Fortran BOZ constants
have limits to their size.  These limits are determined by machine type.

If the machine type has 64-bit words, the maximum lengths are as follows:

   For octal constants:        22
   For hexadecimal constants:  16
   For binary constants:       64

If the octal constant contains 22 characters, the high order (leftmost) digit
must only be a 0 or a 1.

If the machine type has 32-bit words, the maximum lengths are as follows:

   For octal constants:        11
   For hexadecimal constants:   8
   For binary constants:       32

If the octal constant contains 11 characters, the high order (leftmost) digit
must only be 0-3.
.ME
$
$ Error : 0093
$msg  0093 Character '%c' is not valid in an octal constant.
$nexp 0093
Error : Character '%c' is not valid in an octal constant.
.PP
In octal constants (either BOZ or Boolean constants), the only allowed digits
are 0-7.  The compiler detected a character outside of this range.
.ME
$
$ Error : 0094
$msg  0094 A Hollerith constant must be %d characters or less when the "R" form is specified.
$nexp 0094
Error : A Hollerith constant must be %d characters or less when the "R" form is specified.
.PP
The "R" Hollerith form is specified as follows:
    '123456789'R
     9R123456789
.PP
Both of these cases are in error, because there are more than the maximum
number of characters that fit in a word specified in the Hollerith constant.
On 64 bit machines, a maximum of 8 characters is allowed. On 32 bit machines,
a maximum of 4 characters is allowed.
.ME
$
$ Error : 0095
$msg  0095 The real constant must contain digits in the whole and/or the fractional part of the constant.
$nexp 0095
Error : The real constant must contain digits in the whole and/or the fractional part of the constant.
.PP
The compiler is looking for a real constant but can only find a period.
.PP
For example:\*C
.nf
  \*C1.2       \fRLegal\*C
  \*C .2       \fRLegal\*C
  \*C1.        \fRLegal\*C
   \*C.        \fRIllegal
.fi
.ME
$
$ Ansi : 0096
$msg  0096 Hollerith constants are an extension to the Fortran standard.
$nexp 0096
Ansi : Hollerith constants are an extension to the Fortran standard.
.PP
The following specifications for a Hollerith constant are nonstandard.

    2Hab
    2Lab
    2Rab
    'ab'h
    'ab'l
    'ab'r
    "ab"H
    "ab"L
    "ab"R
.ME
$
$ Error : 0097
$msg  0097 The constant is not assignment compatible with "%s".
$nexp 0097
Error : The constant is not assignment compatible with "%s".
.PP
A variable is being initialized in either a type declaration statement or a 
\*CDATA\fR statement.  In either case, the value of the constant or structure
constructor must be compatible with its corresponding variable according to the
rules of intrinsic assignment.
.ME
$
$ Error : 0098
$msg  0098 A dot operator consisting of only letters is expected.
$nexp 0098
Error : A dot operator consisting of only letters is expected.
.PP
A dot operator is expected but not found.  A dot operator must consist of 
only letters.
.ME
$
$ Log_Error : 0099
$msg  0099 "-O %s" is specified, therefore "-O %s" must not be specified.
$nexp 0099
Log_Error : "-O %s" is specified, therefore "-O %s" must not be specified.
.PP
Two optimization options appearing on the command line request conflicting
optimizations. 
.ME
$
$ Error : 0100
$msg  0100 This statement must begin with a label, a keyword, or an identifier.
$nexp 0100
Error : This statement must begin with a label, a keyword, or an identifier.
.PP
The token that begins the statement is not valid.  The token must be either a 
label or an identifier.  An identifier can be a language keyword, a 
construct name, an identifier that is the target variable in an 
assignment statement, or the name of a statement function in a statement
function definition statement.
.ME
$
$ Error : 0101
$msg  0101 The length of the kind parameter exceeds the maximum length of 31.
$nexp 0101
Error : The length of the kind parameter exceeds the maximum length of 31.
.PP
The length of the kind parameter exceeds 31 characters.  Shorten the length
of the kind parameter.
.ME
$
$ Error : 0102
$msg  0102 Only an assignment statement can be used on a WHERE statement.
$nexp 0102
Error : Only an assignment statement can be used on a WHERE statement.
.PP
On a \*CWHERE\fR statement, the statement following the mask expression is not an
assignment statement.
.ME
$
$ Log_Summary : 0103
$msg  0103 openf95: Open64 Fortran Version %s (%s) %s
$nexp 0103
Log_Summary : openf95: Open64 Fortran Version %s (%s) %s
.PP
This identifies which version and release of the compiler is being used.
.ME
$
$ Log_Summary : 0104
$msg  0104 openf95: COMPILE TIME %f SECONDS
$nexp 0104
Log_Summary : openf95: COMPILE TIME %f SECONDS
.PP
This message reports the time, in seconds, that elapsed during the
compilation of the program.
.ME
$
$ Log_Summary : 0105
$msg  0105 openf95: MAXIMUM FIELD LENGTH %d DECIMAL WORDS
$nexp 0105
Log_Summary : openf95: MAXIMUM FIELD LENGTH %d DECIMAL WORDS
.PP
This message reports the maximum amount of memory used by the
compiler during compilation.
.ME
$
$ Log_Summary : 0106
$msg  0106 openf95: %d SOURCE LINES
$nexp 0106
Log_Summary : openf95: %d SOURCE LINES
.PP
This message reports the number of lines of source code read from the input
file and from any \*CINCLUDE\fR files specified on \*CINCLUDE\fR lines.
.ME
$
$ Log_Summary : 0107
$msg  0107 openf95: %d ERRORS, %d WARNINGS, %d OTHER MESSAGES, %d ANSI
$nexp 0107
Log_Summary : openf95: %d ERRORS, %d WARNINGS, %d OTHER MESSAGES, %d ANSI
.PP
This message gives a summary of messages issued during compilation.
.ME
$
$ Log_Summary : 0108
$msg  0108 openf95: CODE: %d WORDS, DATA: %d WORDS
$nexp 0108
Log_Summary : openf95: CODE: %d WORDS, DATA: %d WORDS
.PP
This message reports the number of words of instructions and the number of words
of data generated by the compiler.
.ME
$
$ Error : 0109
$msg  0109 Blank common must not be declared in a TASK COMMON statement or directive.
$nexp 0109
Error : Blank common must not be declared in a TASK COMMON statement or directive.
.PP
A \*CTASK COMMON\fR statement specifies the blank (unnamed) common block.
A \*CTASK COMMON\fR block must be named.  For example:
.PP
.nf
\*CTASK COMMON /\fIcbname\*C/ \fIn\*C, ...        \fR! Legal
\*CTASK COMMON // \fIn\*C, ...              \fR! Illegal
.fi
.ME
$
$ Error : 0110
$msg  0110 The SAVE attribute is specified more than once for common block "%s".
$nexp 0110
Error : The SAVE attribute is specified more than once for common block "%s".
.PP
The \*CSAVE\fR attribute must only be specified once for a common block name.
For example:
.CS
 PROGRAM MAIN
 COMMON /BLK/A
 SAVE /BLK/
 SAVE /BLK/ \fR! Illegal
.PP
This is illegal because \*CSAVE\fR is specified twice for \*CBLK\fR.
.CE
.ME
$
$ Error : 0111
$msg  0111 The PARAMETER attribute is specified, therefore an initialization expression must be specified for "%s".
$nexp 0111
Error : The PARAMETER attribute is specified, therefore an initialization expression must be specified for "%s".
.PP
The \*C= \fIinitialization_expr\fR must appear for the data object name if the
statement contains a \*CPARAMETER\fR attribute.
.ME
$
$ Ansi : 0112
$msg  0112 The arithmetic IF statement is an obsolescent feature.
$nexp 0112
Ansi : The arithmetic IF statement is an obsolescent feature.
.PP
This feature was declared obsolescent in the Fortran 2003 standard.
Use the Fortran \*CIF\fR
statement or \*CIF\fR construct in place of the arithmetic \*CIF\fR.
.ME
$
$ Error : 0113
$msg  0113 IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for data object "%s".
$nexp 0113
Error : IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for data object "%s".
.PP
When \*CIMPLICIT NONE\fR is specified, all data objects must have an explicit type.
.ME
$
$ Error : 0114
$msg  0114 A deferred-shape or assumed-shape array must not have an assumed-size specifier or an upper bound specified.
$nexp 0114
Error : A deferred-shape or assumed-shape array must not have an assumed-size specifier or an upper bound specified.
.PP
The compiler found \fIarray\fR(:10) or \*Varray\fR(:*) or a variation of this.  A
deferred-shape specifier and an assumed-shape specifier must not have an upper
bound specified.  The following bound forms are permissible:
.PP
.nf
explicit-shape array	[\fIlower-bound\fR:] \fIupper-bound\fR
.br
assumed-shape array	[\fIlower-bound\fR]:
.br
deferred-shape array	:
.br
assumed-size array	[\*Vlower-bound\fR:]*
.fi
.PP
The following example declares a 2-dimensional assumed-shape array of type integer.
.CS
    INTEGER, DIMENSION(1:,:) :: ARRAY
.CE
.ME
$
$ Error : 0115
$msg  0115 An explicit-shape array must have an upper bound specified.
$nexp 0115
Error : An explicit-shape array must have an upper bound specified.
.PP
The type of an array is determined by the first dimension.  If the 
dimension is 1:2  (where 1 and 2 are integers and 1 is optional), this
is an explicit-shape array declaration and the rest of the dimensions must
have an upper bound specified.
.ME
$
$ Error : 0116
$msg  0116 The assumed-size specifier (*) must be the upper bound for the last dimension of an explicit-shape array declaration.
$nexp 0116
Error : The assumed-size specifier (*) must be the upper bound for the last dimension of an explicit-shape array declaration.
.PP
The assumed-size specifier (*) can only be declared for the last dimension
of an explicit-shape array.  It must be the upper bound and be followed by a
right parenthesis.  The following bound forms are permissible:
.PP
.nf
explicit-shape array	[\*Vlower-bound\fR:] \*Vupper-bound\fR
.br
assumed-shape array	[\*Vlower-bound\fR]:
.br
deferred-shape array	:
.br
assumed-size array	[\*Vlower-bound\fR:]*
.fi
.ME
$
$ Error : 0117
$msg  0117 The maximum rank for an array declaration is 7.
$nexp 0117
Error : The maximum rank for an array declaration is 7.
.PP
An array can have no more than 7 dimensions.
.ME
$
$ Error : 0118
$msg  0118 Function "%s" was used in an expression.  The type is assumed to be %s.
$nexp 0118
Error : Function "%s" was used in an expression.  The type is assumed to be %s.
.PP
The compiler detected an attempt to retype a function that was previously
implicitly typed and used in an expression.  The following is an example of this
error:
.CS   
     DIMENSION ARRAY(KFUNC())
     REAL KFUNC
.CE
In this example, \*CKFUNC\fR can only be typed integer, because that is how 
it is implicitly typed in the \*CDIMENSION\fR statement.
.ME
$
$ Error : 0119
$msg  0119 This is a deferred-shape dimension specifier.  Therefore a comma or right parenthesis is expected, but a "%s" was found.
$nexp 0119
Error : This is a deferred-shape dimension specifier.  Therefore a comma or right parenthesis is expected, but a "%s" was found.
.PP
While parsing a deferred-shape array specification, the compiler expected a
comma or right parenthesis, but found something else.
A deferred-shape dimension specifier is simply a colon (:).

Other types of permissible bound forms are:
.PP
.nf
  explicit-shape array	[\*Vlower-bound\fR:] \*Vupper-bound\fR
.br
  assumed-shape array	[\*Vlower-bound\fR]:
.br
  assumed-size array	[\*Vlower-bound\fR:]*
.fi
.PP
If an array has the \*CALLOCATABLE\fR or \*CPOINTER\fR attribute, it must be specified as a
deferred-shape array.

The following example declares a 3-dimensional pointer array called \*CJOE\fR of type integer:
.CS
INTEGER, POINTER, DIMENSION(:,:,:) :: JOE
.CE     
.ME
$
$ Error : 0120
$msg  0120 The mask expression of a WHERE statement or WHERE construct must be type logical.
$nexp 0120
Error : The mask expression of a WHERE statement or WHERE construct must be type logical.
.PP
The expression used as the mask expression in a \*CWHERE\fR statement or \*CWHERE\fR 
construct is not the correct type.  It must be a logical expression.
.ME
$
$ Error : 0121
$msg  0121 If the "= initialization" expression is specified, "::" must separate the type and attributes from the identifier list.
$nexp 0121
Error : If the "= initialization" expression is specified, "::" must separate the type and attributes from the identifier list.
.PP
The \*C::\fR separator is required if \*C= \fIinitialization_expr\fR is specified
on the type declaration statement.
.ME
$
$ Error : 0122
$msg  0122 The RESULT keyword is only permitted on ENTRY statements in function subprograms.
$nexp 0122
Error : The RESULT keyword is only permitted on ENTRY statements in function subprograms.
.PP
The \*CRESULT\fR keyword is used on an \*CENTRY\fR statement that is not in a function
subprogram.  A \*CRESULT\fR keyword can only be specified on the \*CFUNCTION\fR and \*CENTRY\fR
statements in a function subprogram.
.ME
$
$ Error : 0123
$msg  0123 "%s" is a derived type type-name.  It cannot be defined as a type-name more than once.
$nexp 0123
Error : "%s" is a derived type type-name.  It cannot be defined as a type-name more than once.
.PP
A derived type type-name must only be defined once.  The compiler found
a type-name being used a subsequent time to define another derived type.
.ME
$
$ Ansi : 0124
$msg  0124 Use of %s*%d is an extension to the Fortran standard.
$nexp 0124
Ansi : Use of %s*%d is an extension to the Fortran standard.
.PP
Use of \fIdata_type\*C*\fIlength\fR is an extension to the Fortran standard for noncharacter data types.
.ME
$
$ Error : 0125
$msg  0125 "%d" is not an allowed length for %s.
$nexp 0125
Error : "%d" is not an allowed length for %s.
.PP
The length specified after the asterisk is an illegal length for the type.
For example: \*CDOUBLE PRECISION*8\fR is illegal because 16 is the only valid
star type for \*CDOUBLE PRECISION\fR.
.PP
Following are types and their legal star types:
.PP
.nf
 Integer: 1, 2, 4, 6, 8
 Real: 4, 8, 16
 Complex: 8, 16, 32
 Logical: 1, 2, 4, 8
 Double precision: 16
.fi
.ME
$
$ Error : 0126
$msg  0126 Derived type "%s" is used, but it does not have any components defined for it.
$nexp 0126
Error : Derived type "%s" is used, but it does not have any components defined for it.
.PP
The components of a derived type must be defined before they are accessed,
unless the type is specified on a \*CFUNCTION\fR statement or an \*CIMPLICIT\fR
statement.  In these cases, the derived type must be defined after the
reference.  In all cases, the derived type must be defined.  A derived type
is defined by specifying the component definitions.
.ME
$
$ Error : 0127
$msg  0127 More than one part reference of this data reference has a rank greater than zero.
$nexp 0127
Error : More than one part reference of this data reference has a rank greater than zero.
.PP
This data reference has more than one part that is array-valued.
In an array-valued data reference, there must be only one array-valued
part.  In the following example, both \*CA\fR and \*CB\fR are arrays:
.CS
	A(1:4)%B(3:4)  ! Illegal reference
.CE
.ME
$
$ Error : 0128
$msg  0128 Positional actual arguments in an actual argument list must precede any keywords.
$nexp 0128
Error : Positional actual arguments in an actual argument list must precede any keywords.
.PP
A positional argument follows an argument keyword in this actual argument list.
.ME
$
$ Error : 0129
$msg  0129 The kind type parameter must be a scalar integer constant.
$nexp 0129
Error : The kind type parameter must be a scalar integer constant.
.PP
A kind type parameter may be a \fIdigit_string\fR or the name of a scalar integer
constant.  For example:
.CS
INTEGER, PARAMETER :: large = 8
REAL, PARAMETER    :: rkind = 8.0
INTEGER            :: ikind = 8

i = 123_8      \fR! Legal\*C
i = 123_large  \fR! Legal\*C
 
i = 123_rkind  \fR! Illegal \*Crkind\fR not type integer.\*C
i = 123_ikind  \fR! Illegal \*Cikind\fR not named constant.\*C
.CE
.PP
.ME
$
$ Warning : 0130
$msg  0130 The kind type parameter value %s is not valid for type %s.
$nexp 0130
Warning : The kind type parameter value %s is not valid for type %s.
.PP
Following are valid kind types for each type:
.nf
  Character: 1
  Integer: 1, 2, 4, 6, 8
  Real: 4, 8, 16
  Complex: 4, 8, 16
  Logical: 1, 2, 4, 8
.ME
$
$ Error : 0131
$msg  0131 Explicit-shape array "%s" has nonconstant bounds so it can only be declared in a function or subroutine program unit.
$nexp 0131
Error : Explicit-shape array "%s" has nonconstant bounds so it can only be declared in a function or subroutine program unit.
.PP
Explicit-shape arrays with nonconstant bounds can only be declared in
subroutines, functions, or interface bodies which describe
subroutines and functions.  The compiler detected an array with 
nonconstant bounds, but the program unit is not a subroutine or function.
.ME
$
$ Error : 0132
$msg  0132 Function entry "%s" has result name "%s", which must be used when declaring the result to be a target.
$nexp 0132
Error : Function entry "%s" has result name "%s", which must be used when declaring the result to be a target.
.PP
If a function has a result name, the result name must be used when declaring
the result to be a target.
.ME
$
$ Ansi : 0133
$msg  0133 The Fortran standard requires a SAVE statement with no saved-entity-list to be the only SAVE in the scoping unit.
$nexp 0133
Ansi : The Fortran standard requires a SAVE statement with no saved-entity-list to be the only SAVE in the scoping unit.
.PP
If a \*CSAVE\fR statement with an omitted \fIsaved_entity_list\fR occurs in a scoping unit,
no other explicit occurrences of the \*CSAVE\fR attribute or \*CSAVE\fR statement are
permitted in the same scoping unit.
.ME
$
$ Ansi : 0134
$msg  0134 Cray pointers are an extension to the Fortran standard.
$nexp 0134
Ansi : Cray pointers are an extension to the Fortran standard.
.PP
Cray pointers are nonstandard and are declared as follows:

     \*CPOINTER\fR(\*Vpointer\*C,\*Vpointee\fR)

Fortran pointers are declared as follows:

     \*CPOINTER [::] \fR\*Vobject-name\fR

The two kinds of pointers must not be mixed.
.ME
$
$ Error : 0135
$msg  0135 A FORMAT statement is required to have a statement label.
$nexp 0135
Error : A FORMAT statement is required to have a statement label.
.PP
The syntax of a \*CFORMAT\fR statement is:
.PP
\fIformat_stmt\fR is \*CFORMAT \fIformat_specification\fR
.PP
\fIformat_specification\fR   is \*C([\fIformat_item_list\*C])\fR
.PP
The \*CFORMAT\fR statement must have a statement label.
.ME
$
$ Internal : 0136
$msg  0136 The variant active in the attribute entry passed to %s is not valid.
$nexp 0136
Internal : The variant active in the attribute entry passed to %s is not valid.
.PP
The attribute table index passed to the named routine is not valid.  It has
the wrong active variant.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0137
$msg  0137 An EQUIVALENCE list needs at least two items in the list.
$nexp 0137
Error : An EQUIVALENCE list needs at least two items in the list.
.PP
There are less than two items in an \*CEQUIVALENCE\fR list.  An \*CEQUIVALENCE\fR list
needs at least two items.
.ME
$
$ Internal : 0138
$msg  0138 Table "%s" does not line up correctly on a word boundary for this machine.
$nexp 0138
Internal : Table "%s" does not line up correctly on a word boundary for this machine.
.PP
The listed table does not line up on a word boundary for this machine.  Check
to make sure that the table is divided up into 32 bit chunks and that the
\*CNUM_...._WDS\fR macro for the table is set to the correct number of words.
.PP
This message should never be generated.  This is a bad compiler or a bad
installation.  Please notify your product support organization with this
error message number and any supporting information.  This message does
not indicate a problem with your code.
.ME
$
$ Error : 0139
$msg  0139 This I/O control list specifier requires a keyword.
$nexp 0139
Error : This I/O control list specifier requires a keyword.
.PP
All arguments used in the I/O statements \*COPEN\fR, \*CCLOSE\fR, \*CBACKSPACE\fR, \*CENDFILE\fR, 
\*CREWIND\fR, and \*CINQUIRE\fR must be identified with a \fIkeyword=\fR preceding the
argument, except for the \*CUNIT\fR, \*CFMT\fR and \*CNML\fR specifiers.  If \*CUNIT\fR is specified
without a keyword, it must be the first item in the control list.  If the
\*CFMT\fR or \*CNML\fR specifier is specified without a keyword, it must be the second 
item in the list; the first item must be the \*CUNIT\fR specified without 
its keyword.
.ME
$
$ Error : 0140
$msg  0140 The derived type being defined is sequenced so any components which are derived types must also be sequenced.
$nexp 0140
Error : The derived type being defined is sequenced so any components which are derived types must also be sequenced.
.PP
If a \*CSEQUENCE\fR statement is present in the derived type being defined, 
all derived types specified in component definitions
must be sequence types.  The compiler detected a component which is a
derived type but is not a sequence type.
.ME
$
$ Internal : 0141
$msg  0141 The beginning token of this statement has led to a bad choice of statement type.
$nexp 0141
Internal : The beginning token of this statement has led to a bad choice of statement type.
.PP
The statement parse driver has called the 
\*Cparse_bad_stmt\fR routine.  Check the beginning token and the token to
parse routines table.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0142
$msg  0142 Label variable "%s" must be scalar and of type default integer.
$nexp 0142
Error : Label variable "%s" must be scalar and of type default integer.
.PP
An \*CASSIGN\fR statement or an \*Cassigned GO TO\fR statement has a label variable
(\fIscalar_int_variable\fR) that is not scalar and/or is not of type default integer.

The syntax of the \*CASSIGN\fR statement is:
.CS
  \*CASSIGN\fR \fIlabel \*CTO\fR \fIscalar_int_variable\fR
.CE

The syntax of the \*Cassigned GO TO\fR statement is:
.CS
  \*CGO TO\fR \fIscalar_int_variable\*C [ [\fR,\*C] (\fIlabel_list\*C) ]\fR
.CE

The \fIscalar_int_variable\fR must be an unqualified name; it cannot be a
structure component, array element, or substring designator.  The variable
must be of type default integer, and it must be scalar.  For these statements,
"type default integer" also means that the integer must be a full word.  This
message is produced if an option such as \*C-i 32\fR or \*C-s default32\fR is
specified and the \fIscalar_int_variable\fR is of type default integer because
these options reduce the bit length of default integers (a Cray T3E address
is longer than 32 bits).
.ME
$
$ Error : 0143
$msg  0143 Character object "%s" has nonconstant bounds.  It can only be declared in a function or subroutine program unit.
$nexp 0143
Error : Character object "%s" has nonconstant bounds.  It can only be declared in a function or subroutine program unit.
.PP
Objects that are typed as character with nonconstant lengths can only be
declared in subroutines, functions, and interface bodies (which describe
subroutines and functions).  The compiler detected a character object
with a nonconstant length, but the declaration is not contained in a subroutine, interface body, or
function.
.ME
$
$ Error : 0144
$msg  0144 Transfer of control to the nonexecutable statement at line %d is not allowed.
$nexp 0144
Error : Transfer of control to the nonexecutable statement at line %d is not allowed.
.PP
The statement is attempting to transfer control to a nonexecutable
statement.  Control can only be transferred to an executable branch target statement.  
(See the \fICFortran Language Reference Manual\fR, publication SR-3902, 
for the definitions of executable and nonexecutable statements.)
.ME
$
$ Error : 0145
$msg  0145 Transfer of control to the %s statement at line %d is not allowed.
$nexp 0145
Error : Transfer of control to the %s statement at line %d is not allowed.
.PP
The statement is attempting to transfer control to an executable 
statement that must not be used as a branch target.  All executable statements
are branch targets except for:  \*CCASE, ELSE, ELSE IF, ELSE WHERE,
END WHERE\fR and \*CIF-THEN\fR.
Transfer of control to some executable statements is allowed only
in limited contexts:
.BL
It is permissible to branch to an \*CEND SELECT\fR statement only from within its 
\*CCASE\fR construct.
.BL
Control should only be transferred to an \*CEND IF\fR statement from within 
its \*CIF\fR construct.  Transfer of control to an \*CEND IF\fR statement from outside its \*CIF\fR 
construct is obsolescent.
.BL
Control should only be transferred to a \*CDO\fR loop termination statement from 
within its \*CDO\fR construct.  For nested \*CDO\fR loops that share a common loop 
termination statement, control should only be transferred to the loop 
termination statement from within the innermost \*CDO\fR construct.
.PP
Violations of these specific cases are diagnosed by other messages.     
.ME
$
$ Error : 0146
$msg  0146 Statement label "%s" was previously defined at line %d.
$nexp 0146
Error : Statement label "%s" was previously defined at line %d.
.PP
A statement label must only be defined once within a scoping unit.   The compiler found another 
definition for the statement label at the specified line.
.ME
$
$ Error : 0147
$msg  0147 Transfer of control into the WHERE construct starting at line %d is not allowed.
$nexp 0147
Error : Transfer of control into the WHERE construct starting at line %d is not allowed.
.PP
The statement is attempting to transfer control to a statement in the
\*CWHERE\fR block or \*CELSEWHERE\fR block of a \*CWHERE\fR construct.  In a \*CWHERE\fR construct, 
only the \*CWHERE\fR construct statement itself can be a branch target.
.ME
$
$ Error : 0148
$msg  0148 Transfer of control into the CASE block starting at line %d is not allowed.
$nexp 0148
Error : Transfer of control into the CASE block starting at line %d is not allowed.
.PP
The statement in error is attempting to transfer control to a \*CCASE\fR block from 
outside the \*CCASE\fR construct or from another \*CCASE\fR block within the same
\*CCASE\fR construct.  Transfer of control to the interior of a \*CCASE\fR block
from outside the block is not permissible. 
.ME
$
$ Ansi : 0149
$msg  0149 Transfer of control to an END IF statement from outside the IF construct is obsolescent/deleted.
$nexp 0149
Ansi : Transfer of control to an END IF statement from outside the IF construct is obsolescent/deleted.
.PP
Instead of branching to the \*CEND IF\fR statement from outside its \*CIF\fR construct, 
branch to the statement following the \*CEND IF\fR.  This may require program logic
changes if the statement following the \*CEND IF\fR is not a valid branch target
statement. 
.PP
The Fortran 95 standard has deleted this feature. The compiler supports it as
an extension.
.ME
$
$ Error : 0150
$msg  0150 Transfer of control to an END DO statement from outside its block DO construct is not allowed.
$nexp 0150
Error : Transfer of control to an END DO statement from outside its block DO construct is not allowed.
.PP
The statement in error is attempting to transfer control to an \*CEND DO\fR statement
but the branching statement is not contained in the block \*CDO\fR construct 
terminated by the labeled \*CEND DO\fR.  It is permissible to branch to an \*CEND DO\fR
statement only from within its \*CDO\fR construct.
.ME
$
$ Error : 0151
$msg  0151 This use of derived type "%s" is not valid.
$nexp 0151
Error : This use of derived type "%s" is not valid.
.PP
A derived type name is in an executable statement, but is not being used as
a derived type constructor.
.ME
$
$ Error : 0152
$msg  0152 The assignment statement of a WHERE statement or in a WHERE construct must not be a defined assignment.
$nexp 0152
Error : The assignment statement of a WHERE statement or in a WHERE construct must not be a defined assignment.
.PP
This masked array assignment resolves to a defined assignment.
.ME
$
$ Error : 0153
$msg  0153 Transfer of control to an END SELECT statement from outside its CASE construct is not allowed.
$nexp 0153
Error : Transfer of control to an END SELECT statement from outside its CASE construct is not allowed.
.PP
The statement in error is attempting to transfer control to an \*CEND SELECT\fR 
statement but the branching statement is not contained in a \*CCASE\fR block of the
\*CCASE\fR construct terminated by the labeled \*CEND SELECT\fR.  It is permissible
to branch to an \*CEND SELECT\fR statement only from within its \*CCASE\fR 
construct.
.ME
$
$ Warning : 0154
$msg  0154 Transfer of control into the DO loop starting at line %d is not safe.
$nexp 0154
Warning : Transfer of control into the DO loop starting at line %d is not safe.
.PP
The statement is transferring control into the body of a \*CDO\fR loop or to the 
termination statement of a \*CDO\fR loop but the branching statement is not 
contained in the loop.  In nested loops with a common 
termination statement, control can only be transferred to the common
termination statement from the innermost loop.
.ME
$
$ Ansi : 0155
$msg  0155 Transfer of control into the DO loop starting at line %d is nonstandard.
$nexp 0155
Ansi : Transfer of control into the DO loop starting at line %d is nonstandard.
.PP
The Fortran standard does not allow transfer of control into the body of a \*CDO\fR
loop or to the termination statement of a \*CDO\fR loop from outside the loop.
.ME
$
$ Warning : 0156
$msg  0156 Transfer of control into the %s block starting at line %d is not safe.
$nexp 0156
Warning : Transfer of control into the %s block starting at line %d is not safe.
.PP
The statement is transferring control into a \*CTHEN\fR block, \*CELSE\fR block, or \*CELSE IF\fR
block from outside the \*CIF\fR construct or from another block within the \*CIF\fR
construct.  Transfer of control to the interior of a block from outside the 
block is not allowed.
.ME
$
$ Ansi : 0157
$msg  0157 Transfer of control into the %s block starting at line %d is nonstandard.
$nexp 0157
Ansi : Transfer of control into the %s block starting at line %d is nonstandard.
.PP
The Fortran standard does not allow transfer of control into a \*CTHEN\fR block, \*CELSE IF\fR
block, or \*CELSE\fR block from outside the \*CIF\fR construct or from another block within the
\*CIF\fR construct.
.ME
$
$ Error : 0158
$msg  0158 No name can be specified on the END BLOCK DATA statement for an unnamed block data program unit.
$nexp 0158
Error : No name can be specified on the END BLOCK DATA statement for an unnamed block data program unit.
.PP
An unamed block data program unit must not have a name specified on its \*CEND\fR
statement.  The compiler detected a name on the \*CEND BLOCK DATA\fR statement
for a block data program unit that is unnamed.
.ME
$
$ Error : 0159
$msg  0159 More than one CASE DEFAULT statement was found for this CASE construct.
$nexp 0159
Error : More than one CASE DEFAULT statement was found for this CASE construct.
.PP
The compiler detected more than one \*CCASE DEFAULT\fR statement for this 
\*CCASE\fR construct.  Only one \*CCASE DEFAULT\fR statement is allowed per 
\*CCASE\fR construct.
.ME
$
$ Internal : 0160
$msg  0160 There is a problem with the block stack at block stack index %d.
$nexp 0160
Internal : There is a problem with the block stack at block stack index %d.
.PP
One of the following errors has occurred:
.BL
The compiler has found an unknown block on the block stack at the given block index.  
The block stack should never have an unknown block on it.
.BL
The block stack has an \*CIf_Blk\fR without a \*CIF_THEN_BLK\fR, \*CIF_ELSE_IF_BLK\fR, 
or \*CIF_ELSE_BLK\fR immediately following it.
.BL
The \*Cblk_stk_idx\fR = \*CNULL_IDX\fR.  The block stack should always have the program 
unit block as the first entry.
.BL
The current block is an \*CIf_Then_Blk\fR, \*CIf_Else_If_Blk\fR, or \*CIf_Else_Blk\fR 
but the preceding block entry is not an \*CIf_Blk\fR.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Ansi : 0161
$msg  0161 A character literal used as a Hollerith constant is an extension to the Fortran standard.
$nexp 0161
Ansi : A character literal used as a Hollerith constant is an extension to the Fortran standard.
.PP
A character constant is being used in a context in which a character constant
is not allowed, but a Hollerith constant is allowed.  This is an extension to
the Fortran standard.  Hollerith constants are allowed in arithmetic and
logical expressions and in assignment statements where the left side of the
equal sign is an entity of type integer or real.  This message can also occur
with the use of a literal string as a file identifier in a control
information list on an I/O statement.
.ME
$
$ Error : 0162
$msg  0162 The %s statement is not allowed in an ELSE IF block of an IF construct.
$nexp 0162
Error : The %s statement is not allowed in an ELSE IF block of an IF construct.
.PP
The following statements are allowed following the \*CELSEIF\fR of an \*CIF\fR construct:
.PP
.nf
\*CALLOCATE               ASSIGN
BACKSPACE              BUFFER
CALL                   CLOSE
CONTINUE               DATA
DEALLOCATE             DECODE
DO                     ENCODE
ENDFILE                ENDIF
ELSE                   ELSEIF
FORMAT                 GOTO
IF                     INQUIRE
NAMELIST               NULLIFY
OPEN                   PAUSE
PRINT                  READ
RETURN                 REWIND
SELECT                 STOP
WHERE                  WRITE\fR
assignment
.fi
.PP
The following statements are allowed if the listed condition is met.
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE\fR, \*CEXIT\fR, \*CEND DO\fR in a \*CDO\fR construct
.ME
$
$ Error : 0163
$msg  0163 The %s statement is not allowed in an ELSE block of an IF construct.
$nexp 0163
Error : The %s statement is not allowed in an ELSE block of an IF construct.
.PP
The following statements are allowed in the \*CELSE\fR block of an \*CIF\fR construct:
.PP
.nf
\*CALLOCATE               ASSIGN
BACKSPACE              BUFFER
CALL                   CLOSE
CONTINUE               DATA
DEALLOCATE             DECODE
DO                     ENCODE
ENDFILE                ENDIF
ELSE                   ELSEIF
FORMAT                 GOTO
IF                     INQUIRE
NAMELIST               NULLIFY
OPEN                   PAUSE
PRINT                  READ
RETURN                 REWIND
SELECT                 STOP
WHERE                  WRITE\fR
assignment
.fi
.PP
The following statements are allowed if the listed condition is met.
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE\fR, \*CEXIT\fR, \*CEND DO\fR in a \*CDO\fR construct
.ME
$
$ Error : 0164
$msg  0164 The %s statement is not allowed in an internal procedure.
$nexp 0164
Error : The %s statement is not allowed in an internal procedure.
.PP
The following statements are not allowed in an internal procedure:
.CS
    \*CCONTAINS
    \*CTASK COMMON
    \*CPUBLIC
    \*CPRIVATE
.CE
.PP
The following statements must be inside an interface block, scoping unit, or construct 
to be in an internal subprogram.
.BL
\*CSEQUENCE, END TYPE\fR in a derived type definition
.BL
\*CMODULE PROCEDURE\fR in a generic interface block  
.BL
\*CELSE, ELSEIF, END IF\fR in an \*CIF\fR construct
.BL
\*CCASE, END SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE, END WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE, EXIT, END DO\fR in a \*CDO\fR construct
.BL
\*CFUNCTION, SUBROUTINE, END INTERFACE\fR  in an interface block
.ME
$
$ Error : 0165
$msg  0165 The %s statement is not allowed in a module procedure.
$nexp 0165
Error : The %s statement is not allowed in a module procedure.
.PP
The following statements are not allowed in a module procedure:
.CS
    \*CPUBLIC
    \*CPRIVATE
    \*CTASK COMMON
.CE
.PP
The following statements must be inside 
an interface block, scoping unit, or construct to be in a module procedure:
.BL
\*CSEQUENCE, END TYPE\fR in a derived type definition
.BL
\*CMODULE PROCEDURE\fR in a generic interface block
.BL
\*CELSE, ELSEIF, END IF\fR in an \*CIF\fR construct
.BL
\*CCASE, END SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE, END WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE, EXIT, END DO\fR in a \*CDO\fR construct
.BL
\*CFUNCTION, SUBROUTINE, END INTERFACE\fR  in an interface block
.ME
$
$ Error : 0166
$msg  0166 Extraneous data follows a format specification.
$nexp 0166
Error : Extraneous data follows a format specification.
.PP
This format specification has unrecognizable characters beyond 
the end of the format specification.  The following statement would cause this 
message to be issued:
.CS
  10 FORMAT(1X,A8) X6		\fR! Illegal
.CE
.ME
$
$ Warning : 0167
$msg  0167 Field width is too small in this usage.
$nexp 0167
Warning : Field width is too small in this usage.
.PP
An improper field width is used with an E, F, D, G, I, Z, or O
edit descriptor.  In the following edit descriptors, \fIw\fR must be greater than
or equal to \fId\fR:
.CS
  \*CE\fIw.d\*C  E\fIw.d\*CE\fIe\*C
  \*CF\fIw.d\*C
  \*CG\fIw.d\*C  G\fIw.d\*CE\fIe\*C
  \*CD\fIw.d
.CE
For the following edit descriptors, \fIw\fR must be greater than or equal to \fIm\fR.
.CS
  \*CI\fIw.m\*C
  \*CB\fIw.m\*C
  \*CO\fIw.m\*C
  \*CZ\fIw.m\fR
.CE
.ME
$
$ Ansi : 0168
$msg  0168 Empty parentheses in a FORMAT are an extension to the Fortran standard and are ignored.
$nexp 0168
Ansi : Empty parentheses in a FORMAT are an extension to the Fortran standard and are ignored.
.PP
An empty format specification appears in another format specification.  The
Fortran standard requires format specifications to be in the following 
format:
.CS
  ([\fIformat_item_list\*C])
.CE
.PP
\fIformat_item_list\fR is a list of items of the following form:
.CS
  [\fIr\*C] \fIdata_edit_desc\*C
  \fIcontrol_edit_desc\*C
  [\fIr\*C] \fIchar_string_edit_desc\*C
            \fI(format_item_list)
.CE
.PP
\fIr\fR
is a nonzero, unsigned integer constant.
.ME
$
$ Ansi : 0169
$msg  0169 A comma between these format fields is required by some versions of the Fortran standard.
$nexp 0169
Ansi : A comma between these format fields is required by some versions of the Fortran standard.
.PP
This format specification has two consecutive format descriptors with no comma
separating them.  The Fortran 90 standard allows the comma to
be omitted in the following places:
.BL
Between a P descriptor and an immediately following F, E, D, or G descriptor
.BL
Before or after a slash
.BL
Before or after a colon
.PP
In addition, the Fortran 2003 standard allows the comma to be omitted between
a P descriptor and a repeat specifier.
.ME
$
$ Ansi : 0170
$msg  0170 A comma preceding a right parenthesis is an extension to the Fortran standard.
$nexp 0170
Ansi : A comma preceding a right parenthesis is an extension to the Fortran standard.
.PP
This format specification has a comma followed by another comma, or a comma
followed by a right parenthesis.  The Fortran standard does not allow a
comma to precede the right parenthesis.
.ME
$
$ Error : 0171
$msg  0171 Construct name "%s" was previously declared at line %d.
$nexp 0171
Error : Construct name "%s" was previously declared at line %d.
.PP
A construct name must only be defined once in a scoping unit.  The compiler found 
a definition for this construct name at the specified line.
.ME
$
$ Error : 0172
$msg  0172 The assign variable name must be an unqualified name.
$nexp 0172
Error : The assign variable name must be an unqualified name.
.PP
The assign variable name in an \*CASSIGN\fR statement must be a simple variable name.
The name must not contain structure component qualifications nor can it be
qualified by a subscript list or a substring range.
.ME
$
$ Error : 0173
$msg  0173 The compiler expected a NAMELIST group name following the NML specifier.
$nexp 0173
Error : The compiler expected a NAMELIST group name following the NML specifier.
.PP
The compiler expected a namelist group name to follow the \*CNML\fR specifier.
.ME
$
$ Ansi : 0174
$msg  0174 A WRITE statement with no unit I/O specifier is an extension to the Fortran standard.
$nexp 0174
Ansi : A WRITE statement with no unit I/O specifier is an extension to the Fortran standard.
.PP
A \*CWRITE\fR statement of the following form was encountered:
.CS
  WRITE \fIf\*C [,\fIiolist\*C]
.CE
.PP
This form of the \*CWRITE\fR statement, with no control information list, 
is an extension to the Fortran standard.
.ME
$
$ Error : 0175
$msg  0175 The IMPLICIT range "%c-%c" is not in alphabetical order.
$nexp 0175
Error : The IMPLICIT range "%c-%c" is not in alphabetical order.
.PP
In an \*CIMPLICIT\fR statement, if the implicit specification list is: 
\fIletter1 - letter\fR2, \fIletter2\fR must follow \fIletter1\fR alphabetically.
.ME
$
$ Error : 0176
$msg  0176 IMPLICIT NONE is specified in this scope, therefore no other IMPLICIT statements are allowed.
$nexp 0176
Error : IMPLICIT NONE is specified in this scope, therefore no other IMPLICIT statements are allowed.
.PP
If \*CIMPLICT NONE\fR is specified within the scope of a procedure, no other
\*CIMPLICIT\fR statements are allowed.
.ME
$
$ Error : 0177
$msg  0177 The following letter(s) were specified in a previous IMPLICIT statement: "%s", so must not be specified again.
$nexp 0177
Error : The following letter(s) were specified in a previous IMPLICIT statement: "%s", so must not be specified again.
.PP
A letter must only be specified once in an \*CIMPLICIT\fR statement in a scoping unit.
For example:
.CS
   IMPLICIT INTEGER (A)
   IMPLICIT REAL (A,B)          \fR! Illegal
.CE
This is illegal, because \*CA\fR was specified in the previous \*CIMPLICIT\fR statement.
.ME
$
$ Internal : 0178
$msg  0178 Array class is Unknown_Array, but there is no declaration error for the dimension spec.
$nexp 0178
Internal : Array class is Unknown_Array, but there is no declaration error for the dimension spec.
.PP
The array class should only be \*CUnknown_Array\fR if there is a declaration error
for the dimension spec.  The problem is detected at the end of \*Cparse_array_spec\fR.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Internal : 0179
$msg  0179 Chose DEFAULT case on case statement in routine %s.  Should never get here.
$nexp 0179
Internal : Chose DEFAULT case on case statement in routine %s.  Should never get here.
.PP
While processing a switch statement, a case was entered that should not be
entered.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Internal : 0180
$msg  0180 Identifier "%s" was found in table %s, but should not have been found.
$nexp 0180
Internal : Identifier "%s" was found in table %s, but should not have been found.
.PP
The name being searched for in the table should not be found in the current
scope (examples are program unit name and dummy argument identifiers).
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0181
$msg  0181 The mask expression in a WHERE statement or WHERE construct must be an array-valued expression.
$nexp 0181
Error : The mask expression in a WHERE statement or WHERE construct must be an array-valued expression.
.PP
The compiler detected a masking expression in a \*CWHERE\fR statement that is
a scalar expression.   It must be an array-valued expression.
.ME
$
$ Ansi : 0182
$msg  0182 The use of a repetition count with the slash edit descriptor is an extension to the Fortran standard.
$nexp 0182
Ansi : The use of a repetition count with the slash edit descriptor is an extension to the Fortran standard.
.PP
A repetition count was found before the slash edit descriptor.  This use 
of the repetition count is an extension to the Fortran standard.
.ME
$
$ Error : 0183
$msg  0183 The %s statement is not allowed in a CASE block of a CASE construct.
$nexp 0183
Error : The %s statement is not allowed in a CASE block of a CASE construct.
.PP
The following statements are allowed in a \*CCASE\fR construct, following a 
\*CCASE\fR statement:
.PP
.nf
\*CALLOCATE               ASSIGN
BACKSPACE              BUFFER
CALL                   CLOSE
CONTINUE               DATA
DEALLOCATE             DECODE
DO                     ENCODE
ENDFILE                ENDIF
ELSE                   ELSEIF
FORMAT                 GOTO
IF                     INQUIRE
NAMELIST               NULLIFY
OPEN                   PAUSE
PRINT                  READ
RETURN                 REWIND
SELECT                 STOP
WHERE                  WRITE\fR
assignment
.fi
.PP
The following statements are allowed if the listed condition is met.
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE\fR, \*CEXIT\fR, \*CEND DO\fR in a \*CDO\fR construct
.ME
$
$ Caution : 0184
$msg  0184 Recursive function entry name "%s" must have a result name to be called recursively when declared an array.
$nexp 0184
Caution : Recursive function entry name "%s" must have a result name to be called recursively when declared an array.
.PP
If the result of a recursive function is an array, it must have a declared
result name in order to be called recursively.  If no result name is specified,
a reference to the function name is a reference to the result variable.
This makes it impossible to call this function recursively from within its
scope.  For example:
.CS
RECURSIVE FUNCTION IARRAY(I) RESULT(I_RESULT)
REAL, DIMENSION :: I_RESULT(100)
...
! The following reference to I_RESULT
! is a reference to the 10th element of
! the result variable.

X = I_RESULT(10)

! The following reference to IARRAY
! is a recursive reference to the function

Y = IARRAY(10) 
END FUNCTION
.CE
.ME
$
$ Error : 0185
$msg  0185 Function entry "%s" has a result name.  Use result name "%s" when typing the function result.
$nexp 0185
Error : Function entry "%s" has a result name.  Use result name "%s" when typing the function result.
.PP
If a function entry has a result name, the result name must be used when
typing or specifying attributes of the function result.
.ME
$
$ Error : 0186
$msg  0186 Unexpected syntax:  Expecting "%s" to follow the END keyword, but found "%s".
$nexp 0186
Error : Unexpected syntax:  Expecting "%s" to follow the END keyword, but found "%s".
.PP
The compiler is looking for one of the following strings after the \*CEND\fR keyword:  \*CBLOCK DATA,
MODULE, PROGRAM, SUBROUTINE, \*CFUNCTION, INTERFACE, TYPE, IF, DO, SELECT,
WHERE \fRor\*C "EOS"\fR.  The compiler did not find the appropriate string.
.ME
$
$ Error : 0187
$msg  0187 The "::" separator is required on a component definition statement, if the POINTER or DIMENSION attribute is specified.
$nexp 0187
Error : The "::" separator is required on a component definition statement, if the POINTER or DIMENSION attribute is specified.
.PP
The Fortran standard requires a \*C::\fR separator to follow the specification of 
the \*CPOINTER\fR and/or \*CDIMENSION\fR attributes on a component declaration statement.
For example:
.CS
    TYPE :: NEW_TYPE
      INTEGER, POINTER :: A     ! Legal
      REAL, DIMENSION(100)  B   ! Illegal
    END TYPE
.CE
.ME
$
$ Error : 0188
$msg  0188 "%s" is already used as a component name for this derived type.
$nexp 0188
Error : "%s" is already used as a component name for this derived type.
.PP
Each component of a derived type must have a unique name.  This name is 
already a component name in this derived type.
.ME
$
$ Error : 0189
$msg  0189 Component "%s" has the POINTER attribute, therefore the DIMENSION attribute must be a deferred-shape dimension.
$nexp 0189
Error : Component "%s" has the POINTER attribute, therefore the DIMENSION attribute must be a deferred-shape dimension.
.PP
If a component has the \*CPOINTER\fR attribute and the \*CDIMENSION\fR attribute, the
dimension must be a deferred-shape dimension.
.ME
$
$ Error : 0190
$msg  0190 Component "%s" does not have a POINTER attribute so its dimension must be an explicit-shape array with constant bounds.
$nexp 0190
Error : Component "%s" does not have a POINTER attribute so its dimension must be an explicit-shape array with constant bounds.
.PP
If a component has the \*CDIMENSION\fR attribute, but does not have the \*CPOINTER\fR
attribute, the \*CDIMENSION\fR must be an explicit-shape definition with constant
bounds.
.ME
$
$ Error : 0191
$msg  0191 A character length for a component declaration must be a constant expression.
$nexp 0191
Error : A character length for a component declaration must be a constant expression.
.PP
The character length for a component definition must be a constant expression.
.ME
$
$ Error : 0192
$msg  0192 The "* char-length" must only be specified for an object of type character.
$nexp 0192
Error : The "* char-length" must only be specified for an object of type character.
.PP
A "\*C*\fIchar_length\fR" is specified for a component or data object whose
type is not character.
.ME
$
$ Internal : 0193
$msg  0193 Unexpected value %d found in %s field for entry %d.
$nexp 0193
Internal : Unexpected value %d found in %s field for entry %d.
.PP
The value found in the symbol table field is not valid.  It is issued for:
.BL
Finding something beyond Structure in AT_TYPE.
.BL
Finding a function, with a NULL_IDX in ATP_RSLT_IDX.
.BL
If the count is nonzero for a derived type, ATO_CHAR_CPNT, ATO_POINTER_CPNT
or ATO_NUMERIC_CPNT must be set.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0194
$msg  0194 The DO variable must be the name of a variable.
$nexp 0194
Error : The DO variable must be the name of a variable.
.PP
An iterative \*CDO\fR statement has the form:
.CS
DO [\fIlabel\*C] [,] \fIdo_variable\*C = \fIscalar_numeric_expr\*C,
  \fIscalar_numeric_expr\*C [, \fIscalar_numeric_expr\*C]
.CE
The \*CDO\fR variable must be a named scalar variable;  for example, it must not be
a function name nor a named constant name.
.ME
$
$ Error : 0195
$msg  0195 In a masked array assignment, the mask expression and the variable being defined must be arrays of the same shape.
$nexp 0195
Error : In a masked array assignment, the mask expression and the variable being defined must be arrays of the same shape.
.PP
All masked assignment statements must match the shape of the mask expression.
.ME
$
$ Internal : 0196
$msg  0196 Function %s was called with a token whose length exceeds %d characters.
$nexp 0196
Internal : Function %s was called with a token whose length exceeds %d characters.
.PP
The symbol table search routines are optimized and have a limit on the size of
the item being searched.  This routine was called to search for an item
which exceeded the limit of the routine.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0197
$msg  0197 Unexpected syntax: "%s" was expected but found "%s".
$nexp 0197
Error : Unexpected syntax: "%s" was expected but found "%s".
.PP
The syntax of this statement is incorrect.  During parsing, the compiler
was looking for one thing, but found another.  This is a general message
used throughout the parser.
.ME
$
$ Error : 0198
$msg  0198 The "::" separator is illegal on the EXTERNAL and the INTRINSIC statements.
$nexp 0198
Error : The "::" separator is illegal on the EXTERNAL and the INTRINSIC statements.
.PP
The syntax for the \*CINTRINSIC\fR statement is:
.CS
  INTRINSIC \fIintrinsic_procedure-name_list
.CE
.PP
The syntax for the \*CEXTERNAL\fR statement is:
.CS
  EXTERNAL \fIexternal_name_list\fR
.CE
.ME
$
$ Error : 0199
$msg  0199 The DO variable must be an unqualified name.
$nexp 0199
Error : The DO variable must be an unqualified name.
.PP
The iterative \*CDO\fR statement has the form:
.CS
  DO [\fIlabel\*C] [,] \fIdo_variable\*C = \fIscalar_numeric_expr\*C,
      \fIscalar_numeric_expr\*C [, \fIscalar_numeric_expr\*C]
.CE
.PP
The \*CDATA\fR statement, some I/O statements, and an array constructor must contain
implied-\*CDO\fRs.  The forms of the implied-\*CDO\fRs are:
.CS
 (\fIdata_i_do_object_list\*C, \fIdata_i_do_variable\*C = 
   \fIscalar_int_expr\*C, \fIscalar_int_expr\*C [, \fIscalar_int_expr\*C])

 (\fIio_implied_do_object_list\*C, \fIdo_variable\*C = 
   \fIscalar_numeric_expr\*C, \fIscalar_numeric_expr\*C 
     [, \fIscalar_numeric_expr\*C])

 (\fIac_value_list\*C, \fIac_do_variable\*C =
   \fIscalar_int_expr\*C, \fIscalar_int_expr\*C 
     [, \fIscalar_int_expr\*C])
.CE
The \*CDO\fR variable of the \*CDO\fR and I/O statements, the \fIdata_i_do_variable \fRof the 
\*CDATA\fR statement, and the \fIac_do_variable\fR of the array constructor are all 
called "\*CDO\fR variables".  In each case, the \*CDO\fR variable
must be a named variable; that is, it must be a simple, unqualified name.  For
example, it must not contain either a subscript list or the structure
component separator "%".
.ME
$
$ Ansi : 0200
$msg  0200 The use of edit descriptor "%c" is an extension to the Fortran standard.
$nexp 0200
Ansi : The use of edit descriptor "%c" is an extension to the Fortran standard.
.PP
A format specification contains an edit descriptor that is an extension to 
the Fortran standard.  The following nonstandard edit descriptors are
accepted by the compiler:
.TL 7
Zw
Hexadecimal integer
.TL
H...
Hollerith edit descriptors were deleted in the Fortran 95 standard, but the
compiler accepts them as an extension.
.TL
Rw
right-justified character
.TL
 $
suppress carriage control
.TL
[\fIn\fR]X
spacing, omitting \fIn\fR is nonstandard
.TL
\fIn\fR/
carriage control, \fIn\fR is nonstandard
.ME
$
$ Error : 0201
$msg  0201 An allocate object in an ALLOCATE statement must be either a pointer or an allocatable array.
$nexp 0201
Error : An allocate object in an ALLOCATE statement must be either a pointer or an allocatable array.
.PP
All arguments in an \*CALLOCATE\fR statement must be previously declared to be
either pointer or allocatable arrays.
.ME
$
$ Error : 0202
$msg  0202 The status variable for this ALLOCATE/DEALLOCATE statement must be a scalar integer variable.
$nexp 0202
Error : The status variable for this ALLOCATE/DEALLOCATE statement must be a scalar integer variable.
.PP
The variable used as the status variable, \*CSTAT = \fIstat_variable\fR, in an \*CALLOCATE\fR or \*CDEALLOCATE\fR
statement cannot be an array and must be of integer type.
.ME
$
$ Error : 0203
$msg  0203 The STAT= must follow all allocation items in an ALLOCATE statement.
$nexp 0203
Error : The STAT= must follow all allocation items in an ALLOCATE statement.
.PP
No allocation items in an \*CALLOCATE\fR statement must follow a status variable. 
\*CSTAT = \fIstat_variable\fR must be the last item in the list.
.ME
$
$ Error : 0204
$msg  0204 The number of subscripts is greater than the number of declared dimensions.
$nexp 0204
Error : The number of subscripts is greater than the number of declared dimensions.
.PP
An array reference of the form \fIa\*C(\fIs\*C[,\fIs\*C]...)\fR is being used.  The number of
subscripts in the reference should equal the number of dimensions
declared for the array in a \*CDIMENSION\fR, \*CCOMMON\fR, \*CPOINTER\fR or type statement.
As an extension to the Fortran standard, the compiler allows array references
to specify fewer subscripts than the declared number.  In this case,
the lower bounds of the remaining unspecified dimensions are used.
.ME
$
$ Error : 0205
$msg  0205 If an array is specified in an ALLOCATE statement, it must have an allocate shape specification list.
$nexp 0205
Error : If an array is specified in an ALLOCATE statement, it must have an allocate shape specification list.
.PP
An attempt was made to allocate an array in an \*CALLOCATE\fR statement without
supplying an allocate shape specification list.
The syntax of an allocate shape specification list is 
.CS
      [lower_bound :] upper_bound
.CE
.ME
$
$ Error : 0206
$msg  0206 The exponent in a constant initialization expression must be type integer.
$nexp 0206
Error : The exponent in a constant initialization expression must be type integer.
.PP
The Fortran standard requires that if exponentiation is used in a
constant initialization expression, it must have an integer exponent.
.ME
$
$ Internal : 0207
$msg  0207 Reset lex has attempted to reset source input to a line not in the source buffer.
$nexp 0207
Internal : Reset lex has attempted to reset source input to a line not in the source buffer.
.PP
The routine \*Creset_src_input\fR attempted to reset the source input index
to a line number that is no longer available in the source input buffer.
Source input should only be reset on the current line.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Ansi : 0208
$msg  0208 Using a Cray pointer as a DO variable is an extension to the Fortran standard.
$nexp 0208
Ansi : Using a Cray pointer as a DO variable is an extension to the Fortran standard.
.PP
A Cray pointer is being used as a \*CDO\fR
variable.  Support of Cray pointers is an extension to the Fortran standard.
.ME
$
$ Error : 0209
$msg  0209 There is no actual argument to correspond to dummy argument number %d.
$nexp 0209
Error : There is no actual argument to correspond to dummy argument number %d.
.PP
An alternate return dummy argument (\*C*\fR) does not have a matching label
actual argument.
.ME
$
$ Error : 0210
$msg  0210 "%s" has the %s attribute.  It must not be used in an initialization expression because it is not a constant.
$nexp 0210
Error : "%s" has the %s attribute.  It must not be used in an initialization expression because it is not a constant.
.PP
The compiler detected an illegal use of this object.  To be used in an
initialization expression, the object must be a constant.  The only attributes
that a constant may have are \*CPARAMETER\fR, \*CPRIVATE\fR or \*CPUBLIC\fR.
A constant may be given an explicit type and it may be declared to be an 
explicit-shape array.
.ME
$
$ Error : 0211
$msg  0211 Continuation is not allowed for this statement.
$nexp 0211
Error : Continuation is not allowed for this statement.
.PP
This continuation line in fixed source form does not follow a regular source 
line.  This error is issued when the first line of a program is a 
continuation line.
.ME
$
$ Error : 0212
$msg  0212 "%s" is typed as %s, therefore it must not have component references specified for it.
$nexp 0212
Error : "%s" is typed as %s, therefore it must not have component references specified for it.
.PP
The object in question is followed by a structure component name separator (%),
but the object has not been declared as a derived type.
.ME
$
$ Error : 0213
$msg  0213 "%s" is not a component of derived type "%s".
$nexp 0213
Error : "%s" is not a component of derived type "%s".
.PP
While parsing a component reference, the compiler 
detected a name that is not declared to be a component of the derived type.
.ME
$
$ Limit : 0214
$msg  0214 The maximum number, %d, of fatal errors has been exceeded.
$nexp 0214
Limit : The maximum number, %d, of fatal errors has been exceeded.
.PP
The compiler aborts when the maximum number of allowed
fatal errors is detected.  The maximum number is set to 100.  This can be
overridden with the \*C-dq\fR command-line option.
The \*C-dq\fR option allows for an unlimited number of fatal errors.
.ME
$
$ Error : 0215
$msg  0215 Dummy argument "%s" of statement function "%s" must have a character length that is an integer constant expression.
$nexp 0215
Error : Dummy argument "%s" of statement function "%s" must have a character length that is an integer constant expression.
.PP
If a dummy argument of a statement function is type character, its length 
must be an integer constant expression.  The compiler detected a 
statement function dummy argument whose type is character, but whose 
length is either assumed-length or variable length.
.ME
$
$ Error : 0216
$msg  0216 A character string must be less than %d characters when used as a dataset name or Boolean constant.
$nexp 0216
Error : A character string must be less than %d characters when used as a dataset name or Boolean constant.
.PP
The compiler detected a character string that will not fit in a 
machine word.  A dataset name or integer expression was expected.
A dataset name is an external file identifier in a control
information list in an I/O statement.
.ME
$
$ Error : 0217
$msg  0217 The DO loop expression must be type integer, default real, or double precision real.
$nexp 0217
Error : The DO loop expression must be type integer, default real, or double precision real.
.PP
Each expression in the \fIloop_control\fR portion of the \*CDO\fR statement must be 
numeric and of type integer, default real, or double precision real.
A \*CDO\fR loop expression of type default real or double precision real is an
obsolescent feature.  Type integer is the preferred choice.
.ME
$
$ Ansi : 0218
$msg  0218 A DO loop expression of type default real or double precision real is an obsolescent/deleted feature.
$nexp 0218
Ansi : A DO loop expression of type default real or double precision real is an obsolescent/deleted feature.
.PP
This feature was deleted in the Fortran 95 standard. The compiler supports it
as an extension.
.PP
An
expression of type integer should be used instead.
.ME
$
$ Error : 0219
$msg  0219 The DO variable must be type integer, default real, or double precision real.
$nexp 0219
Error : The DO variable must be type integer, default real, or double precision real.
.PP
The \*CDO\fR variable of the \*CDO\fR statement must be numeric and of type integer,
default real, or double precision real.  The use of a \*CDO\fR variable of
type default real or double precision real is an obsolescent feature.
Type integer is the preferred choice.
.ME
$
$ Ansi : 0220
$msg  0220 A DO variable of type default real or double precision real is an obsolescent/deleted feature.
$nexp 0220
Ansi : A DO variable of type default real or double precision real is an obsolescent/deleted feature.
.PP
This feature was deleted in the Fortran 95 standard. The compiler supports it
as an extension.
A \*CDO\fR variable of type
integer should be used instead.
.ME
$
$ Ansi : 0221
$msg  0221 Bound expressions other than integer are an extension to the Fortran standard.
$nexp 0221
Ansi : Bound expressions other than integer are an extension to the Fortran standard.
.PP
The Fortran standard requires all bound expressions to be type integer.
.ME
$
$ Error : 0222
$msg  0222 This DO statement expression must be scalar.
$nexp 0222
Error : This DO statement expression must be scalar.
.PP
In a \*CDO\fR statement of the form
.CS
   DO [\fIlabel\*C] [,] \fIdo_variable\*C = \fIscalar_numeric_expr\*C,
        \fIscalar_numeric_expr\*C [, \fIscalar_numeric_expr\*C]
.CE
.PP
each \fIscalar_numeric_expr\fR must be scalar.
.PP
In a \*CDO\fR statement of the form
.CS
   DO [\fIlabel\*C] [,] WHILE(\fIscalar_logical_expr\*C)
.CE
the \fIscalar_logical_expr\fR must be scalar.
.ME
$
$ Error : 0223
$msg  0223 The DO variable must be scalar.
$nexp 0223
Error : The DO variable must be scalar.
.PP
In a \*CDO\fR statement of the form
.CS
   DO [\fIlabel\*C] [,] \fIdo_variable\*C = \fIscalar_numeric_expr\*C,
        \fIscalar_numeric_expr\*C [, \fIscalar_numeric_expr\*C]
.CE
.PP
the \fIdo_variable\fR must be scalar.
.ME
$
$ Internal : 0224
$msg  0224 Semantic analysis of compiler generated DO loop expression failed.
$nexp 0224
Internal : Semantic analysis of compiler generated DO loop expression failed.
.PP
There are two expression trees that are generated by the compiler as part of
the representation of a \*CDO\fR statement:  the iteration count calculation and the 
calculation of the value of the \*CDO\fR variable.  Both involve only
already-established compiler temps.  Semantic analysis is only used to
propagate data types, and so on, up the expression tree therefore 
semantic analysis should not fail.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0225
$msg  0225 For [%s][%s], the error flag in table %s is TRUE (1), but the message number is 0.
$nexp 0225
Error : For [%s][%s], the error flag in table %s is TRUE (1), but the message number is 0.
.PP
This message is always followed by an internal message and compiler abort.
.PP
This message is issued by a debug routine that is called at main
initialization for debug compilations.  The routine goes through the semantic tables
and confirms that the flags and messages agree.  The tables involved are:
.TL 25
flag table
message table
.TL
\*Cobj_to_attr
attr_msg_num
.TL
obj_to_name
name_msg_num
.TL
obj_to_other
other_msg_num
.PP
Verify the entry in both tables and change one or the other so both 
agree.  If this message is issued, the combination is marked in error in the
table, but no error message is specified for the combination in the table.
.ME
$
$ Internal : 0226
$msg  0226 There are problems with the semantic tables.  See previous error messages.
$nexp 0226
Internal : There are problems with the semantic tables.  See previous error messages.
.PP
This internal error is issued if \*Cverify_semantic_tbls\fR has found any problems
in the semantic tables.  Fatal errors are issued for each problem.  This final
internal error is issued at the end to halt compilation.
.PP
This message should never be generated.  This is a bad compiler or a bad
installation.  Please notify your product support organization with this
error message number and any supporting information.  This message does
not indicate a problem with your code.
.ME
$
$ Error : 0227
$msg  0227 For [%s][%s], the error flag in table %s is FALSE (0), but the message number is %d.
$nexp 0227
Error : For [%s][%s], the error flag in table %s is FALSE (0), but the message number is %d.
.PP
This message is always followed by an internal message and compiler abort.
.PP
This message is issued by a debug routine that is called at main
initialization for debug compilations.  It goes through the semantic tables
and makes sure that the flags and messages agree.  The tables involved are:
.TL 25
flag table
message table
.TL
\*Cobj_to_attr
attr_msg_num
.TL
obj_to_name 
name_msg_num
.TL
obj_to_other
other_msg_num
.PP
Verify the entry in both tables and change one or the other so both
agree.  If this message is issued, the combination is legal, but the table has
an error message for this combination.  
.ME
$
$ Error : 0228
$msg  0228 The mode specifier for a BUFFER IN or BUFFER OUT statement must be type integer.
$nexp 0228
Error : The mode specifier for a BUFFER IN or BUFFER OUT statement must be type integer.
.PP
A \*CBUFFER IN\fR or \*CBUFFER OUT\fR statement contains an improper mode specifier.  In
the following statements, \fIm\fR, the mode specifier, must be an integer
expression:
.CS
      BUFFERIN (id,m) (bloc,eloc)
      BUFFEROUT (id,m) (bloc,eloc)
.CE
The mode specifier must not be an array expression.
.ME
$
$ Error : 0229
$msg  0229 The unit specifier or file identifier is not a valid data type or form.
$nexp 0229
Error : The unit specifier or file identifier is not a valid data type or form.
.PP
An I/O statement contains an improper unit or file identifier.  
.PP
A unit
identifier may be one of the following:
.BL
An integer expression with a value in the range 0 to 101.
.BL
An asterisk denoting the default unit.
.PP
A file identifier may be one of the following:
.BL
A character variable, array element, array, or substring specifying an internal file.
.BL
A string of up to 7 alphanumeric characters enclosed in apostrophes, specifying an external file name.
.ME
$
$ Error : 0230
$msg  0230 The mode specifier for the BUFFER IN or BUFFER OUT statement must be scalar.
$nexp 0230
Error : The mode specifier for the BUFFER IN or BUFFER OUT statement must be scalar.
.PP
The \*CBUFFER IN\fR or \*CBUFFER OUT\fR statement contains an improper mode specifier .  In
the following statements, \fIm\fR, the mode specifier, must be an integer
expression:
.CS
      BUFFERIN (\fIid\*C,\fIm\*C) (\fIbloc\*C,\fIeloc\*C)
      BUFFEROUT (\fIid\*C,\fIm\*C) (\fIbloc\*C,\fIeloc\*C)
.CE
.PP
The mode specifier must not be an array expression.
.ME
$
$ Error : 0231
$msg  0231 The character string exceeds the maximum dataset name size of %d characters.
$nexp 0231
Error : The character string exceeds the maximum dataset name size of %d characters.
.PP
The external file identifier in a control information list on an I/O
statement exceeds 7 characters.  File identifiers must consist of 1 to
7 alphanumeric characters on 64-bit architectures.  The limit is 4
characters on 32-bit architectures.
.ME
$
$ Error : 0232
$msg  0232 IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for function "%s".
$nexp 0232
Error : IMPLICIT NONE is specified in the local scope, therefore an explicit type must be specified for function "%s".
.PP
When \*CIMPLICIT NONE\fR is set, all functions must have explicit types.
.ME
$
$ Error : 0233
$msg  0233 IMPLICIT NONE is specified in the host scope, therefore an explicit type must be specified for function "%s".
$nexp 0233
Error : IMPLICIT NONE is specified in the host scope, therefore an explicit type must be specified for function "%s".
.PP
When \*CIMPLICIT NONE\fR is set in a host scope, all functions must have explicit types,
including those in any contained scopes.
.ME
$
$ Error : 0234
$msg  0234 The DO WHILE expression must be type logical.
$nexp 0234
Error : The DO WHILE expression must be type logical.
.PP
In a \*CDO\fR statement of the form:
.CS
 DO [\fIlabel\*C] [,] WHILE (\fIscalar_logical_expr\*C)
.CE
.PP
the \fIscalar_logical_expr\fR must be type logical.
.ME
$
$ Error : 0235
$msg  0235 I/O list items of a derived type which have a pointer or allocatable component are not allowed.
$nexp 0235
Error : I/O list items of a derived type which have a pointer or allocatable component are not allowed.
.PP
If a derived type ultimately contains a pointer or allocatable component,
an object of this
type must not appear as an I/O input item nor as the result of the evaluation
of an I/O output list item.
.ME
$
$ Internal : 0236
$msg  0236 NULL_IDX encountered when completing a compiler generated CONTINUE statement.
$nexp 0236
Internal : NULL_IDX encountered when completing a compiler generated CONTINUE statement.
.PP
To complete a compiler generated \*CCONTINUE\fR statement, continue_stmt_semantics
searches for a statement header (SH) following the current SH that is not
compiler generated and that has a nonzero line number.  This internal error
message is issued when the end of the chain of statement headers was
encountered without finding a statement header that satisfies the above
criteria.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Limit : 0237
$msg  0237 Maximum table size has been exceeded for the "%s".
$nexp 0237
Limit : Maximum table size has been exceeded for the "%s".
.PP
An internal error condition has occured because an internal table has grown
too large for its index field size.  Generally this is caused because the
program is too large for the compiler to handle.  Splitting your code into
multiple compilation units may allow compilation to successfully complete.
.PP
Please notify your product support organization, so that we may verify that
a compiler limit has been reached.
.ME
$
$ Error : 0238
$msg  0238 "%s" is declared as a named constant.  The type is assumed to be %s.
$nexp 0238
Error : "%s" is declared as a named constant.  The type is assumed to be %s.
.PP
The compiler detected an attempt to retype a named constant via a
type declaration statement or an \*CIMPLICIT\fR statement. This is only legal 
if the type confirms the implicit type of the object.
.CS
     PARAMETER (A = 3.0)
     INTEGER A               \fR! Illegal\*C

     PARAMETER (I = 4)
     INTEGER I               \fR! Legal\*C

     PARAMETER (I = 4)
     IMPLICIT INTEGER (I)    \fR! Legal\*C

     PARAMETER (I = 4)
     IMPLICIT CHARACTER (I)  \fR! Illegal
.CE
.ME
$
$ Error : 0239
$msg  0239 "%s" is initialized in a DATA statement.  The type is assumed to be %s.
$nexp 0239
Error : "%s" is initialized in a DATA statement.  The type is assumed to be %s.
.PP
The compiler detected an attempt to retype an object that is already    
initialized.  This is only legal if the type confirms the implicit type 
of the object.
.CS
     DATA A /3/
     INTEGER A	\fR! This is illegal\*C

     DATA I /3/
     INTEGER I	\fR! Legal
.CE
.ME
$
$ Error : 0240
$msg  0240 External function "%s" is referenced in a bounds specification expression.  It must have a nonpointer scalar result.
$nexp 0240
Error : External function "%s" is referenced in a bounds specification expression.  It must have a nonpointer scalar result.
.PP
It is an extension to the Fortran standard to allow external function calls
in array dimension bound declarators or character length declarators.
These functions must not have array-valued or pointer-valued results.
.ME
$
$ Ansi : 0241
$msg  0241 The nonblock or shared-termination DO construct is an obsolescent feature.
$nexp 0241
Ansi : The nonblock or shared-termination DO construct is an obsolescent feature.
.PP
This feature was declared obsolescent in the Fortran 2003 standard.
If a \*CDO\fR statement contains a label and the loop
termination statement is a statement other than \*CEND DO\fR or \*CCONTINUE\fR, the loop
is defined to be a nonblock \*CDO\fR construct.  The \*CCONTINUE\fR or \*CEND DO\fR statement 
should be used to terminate a \*CDO\fR loop.  If multiple \*CDO\fR loops terminate with
the same statement, a separate \*CCONTINUE\fR or \*CEND DO\fR statement should be used to
terminate each nested \*CDO\fR loop.
.ME
$
$ Error : 0242
$msg  0242 The termination statement of a nonblock DO construct must not be an unconditional GO TO statement.
$nexp 0242
Error : The termination statement of a nonblock DO construct must not be an unconditional GO TO statement.
.PP
If a \*CDO\fR statement contains a label and the loop termination statement is a
statement other than \*CEND DO\fR or \*CCONTINUE\fR, the loop is defined to be a nonblock
\*CDO\fR construct.  An unconditional \*CGO TO\fR statement must not be used as the 
termination statement of a nonblock \*CDO\fR construct.
.ME
$
$ Error : 0243
$msg  0243 The termination statement of a nonblock DO construct must not be an assigned GO TO statement.
$nexp 0243
Error : The termination statement of a nonblock DO construct must not be an assigned GO TO statement.
.PP
If a \*CDO\fR statement contains a label and the loop termination statement is a
statement other than \*CEND DO\fR or \*CCONTINUE\fR, the loop is defined to be a nonblock
\*CDO\fR construct.  An assigned \*CGO TO\fR statement must not be used as the
termination statement of a nonblock \*CDO\fR construct.
.ME
$
$ Error : 0244
$msg  0244 The termination statement of a nonblock DO construct must not be a %s statement.
$nexp 0244
Error : The termination statement of a nonblock DO construct must not be a %s statement.
.PP
If a \*CDO\fR statement contains a label and the loop termination statement is an
\fIaction_stmt\fR other than \*CEND DO\fR or \*CCONTINUE\fR, the loop is defined to be a nonblock
\*CDO\fR construct.  The following \fIaction_stmt\fRs must not be used as the termination
statement of a nonblock \*CDO\fR construct:  \*CRETURN\fR, \*CSTOP\fR, \*CEXIT\fR, \*CCYCLE\fR, \*CEND FUNCTION\fR,
\*CEND SUBROUTINE\fR, \*CEND PROGRAM\fR, or arithmetic \*CIF\fR.
.ME
$
$ Internal : 0245
$msg  0245 The number of arguments in the call list is greater than the largest list expected.
$nexp 0245
Internal : The number of arguments in the call list is greater than the largest list expected.
.PP
The number of actual arguments is greater than the allocated internal table 
size.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0246
$msg  0246 The termination statement of a nonblock DO construct must not be an outmoded IF statement.
$nexp 0246
Error : The termination statement of a nonblock DO construct must not be an outmoded IF statement.
.PP
If a \*CDO\fR statement contains a label and the loop termination statement is a
statement other than \*CEND DO\fR or \*CCONTINUE\fR, the loop is defined to be a nonblock
\*CDO\fR construct.  An outmoded \*CIF\fR statement is an indirect logical \*CIF\fR statement or
a two-branch arithmetic \*CIF\fR statement.  These obsolete features have been 
carried forward from FORTRAN 77.  These statements have been replaced by 
alternatives that enhance the portability of Fortran programs.  The \*CIF\fR 
construct or logical \*CIF\fR should be used in place of both of these outmoded
forms.  
.ME
$
$ Error : 0247
$msg  0247 A subscript or substring expression in an EQUIVALENCE statement must be an integer constant expression.
$nexp 0247
Error : A subscript or substring expression in an EQUIVALENCE statement must be an integer constant expression.
.PP
Each subscript or substring expression in an equivalence item must be a scalar integer
constant expression.
.ME
$
$ Error : 0248
$msg  0248 The DO statement label is already defined.
$nexp 0248
Error : The DO statement label is already defined.
.PP
The termination statement label specified by the \*CDO\fR statement was defined
earlier in the scoping unit.  By definition, the loop termination
statement must follow the \*CDO\fR statement.  Therefore, the label must be on a 
statement following the \*CDO\fR statement.  
.ME
$
$ Internal : 0249
$msg  0249 Compiler internal error in fold_relationals.  Unexpected types of operands.
$nexp 0249
Internal : Compiler internal error in fold_relationals.  Unexpected types of operands.
.PP
The routine fold_relationals has found operand types that are not valid.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0250
$msg  0250 Array syntax is not allowed in an EQUIVALENCE statement.
$nexp 0250
Error : Array syntax is not allowed in an EQUIVALENCE statement.
.PP
Sectioned array references are not allowed in an \*CEQUIVALENCE\fR statement.
.ME
$
$ Internal : 0251
$msg  0251 Compiler internal error in fold_relationals.  Unexpected operator.
$nexp 0251
Internal : Compiler internal error in fold_relationals.  Unexpected operator.
.PP
The routine \*Cfold_relationals\fR was called to fold an operator that it was not
intended to see.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0252
$msg  0252 The operands in this array syntax operation are not conformable.
$nexp 0252
Error : The operands in this array syntax operation are not conformable.
.PP
Both operands of an array syntax operation must be shape conformant.
The two arrays that have constant subscripts are not shape conformant.
.ME
$
$ Error : 0253
$msg  0253 The left and right hand sides of this array syntax assignment must be conformable arrays.
$nexp 0253
Error : The left and right hand sides of this array syntax assignment must be conformable arrays.
.PP
When an assignment statement is array syntax, the left side and the right
side must be shape conformant.
.ME
$
$ Caution : 0254
$msg  0254 The values of the DO statement expressions prevent the loop from executing.
$nexp 0254
Caution : The values of the DO statement expressions prevent the loop from executing.
.PP
The iterative \*CDO\fR statement has the form:
.CS
    DO [\fIlabel\*C] [,] \fIdo_variable\*C = 
	\fIstart_expr\*C, \fIend_expr\*C [, \fIinc_expr]
.CE
.PP
The iteration count for the \*CDO\fR loop is calculated from:
.CS
    MAX( INT( (\fIend_expr\*C - \fIstart_expr\*C + \fIinc_expr\*C) / 
	\fIinc_expr\fR), 0)
.CE
.PP
The iteration count is zero for both of the following cases:
.CS
    \fIstart_expr\fR > \fIend_expr\fR  and  \fIinc_expr\fR > 0,  
             or
    \fIstart_expr\fR < \fIend_expr\fR  and  \fIinc_expr\fR < 0
.CE
.PP
If the iteration count is zero, the loop will not be executed.
.ME
$
$ Error : 0255
$msg  0255 The value of the increment expression must not be zero.
$nexp 0255
Error : The value of the increment expression must not be zero.
.PP
The \*CDO\fR statement contains an increment expression that has a zero value. 
It must be greater than zero or less than zero.
.ME
$
$ Error : 0256
$msg  0256 Dummy argument "%s" has the POINTER or ALLOCATABLE attribute, so the actual argument must have the same attribute.
$nexp 0256
Error : Dummy argument "%s" has the POINTER or ALLOCATABLE attribute, so the actual argument must have the same attribute.
.PP
If a dummy argument has the \*CPOINTER\fR attribute, the actual argument must
also have the \*CPOINTER\fR attribute.
If a dummy argument has the \*CALLOCATABLE\fR attribute, the actual argument
must be an entire array having the \*CALLOCATABLE\fR attribute.
.ME
$
$ Internal : 0257
$msg  0257 An illegal item "%d" is detected in field %s of the attribute table.
$nexp 0257
Internal : An illegal item "%d" is detected in field %s of the attribute table.
.PP
An unexpected value has been found in the symbol tables.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0258
$msg  0258 This actual argument is an assumed-size array, therefore it must not be associated with an assumed-shape dummy argument.
$nexp 0258
Error : This actual argument is an assumed-size array, therefore it must not be associated with an assumed-shape dummy argument.
.PP
An assumed-shape dummy argument cannot be associated with an assumed-size
actual argument.
.ME
$
$ Error : 0259
$msg  0259 Recursive reference to "%s", a %s, is illegal in a bounds specification expression.
$nexp 0259
Error : Recursive reference to "%s", a %s, is illegal in a bounds specification expression.
.PP
Some function references are allowed in bounds specification expressions as part
of the Fortran standard, and others are allowed as an extension to the Fortran
standard. However, recursive function references are not allowed in bounds 
specification expressions.  The compiler detected something similar to the 
following example:
.CS
      FUNCTION IFUNC()
      RECURSIVE IFUNC
      DIMENSION A(IFUNC())   
.CE
.PP
The reference to \*CIFUNC\fR is illegal because a bounds expression must be
computed at the entry point.
.ME
$
$ Internal : 0260
$msg  0260 Unexpected block stack order.
$nexp 0260
Internal : Unexpected block stack order.
.PP
If the current block type is greater than Interface_Body_Blk, procedure
cif_begin_scope_rec expects it to be because one of these block types was
the first statement in the main program.  Either the Block Stack has been
corrupted or cif_begin_scope_rec needs to be corrected because it is making 
an incorrect assumption.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0261
$msg  0261 The %s statement is not allowed in a function subprogram.
$nexp 0261
Error : The %s statement is not allowed in a function subprogram.
.PP
The \*CPRIVATE\fR and \*CPUBLIC\fR statements are not allowed in a function
subprogram.
.PP
The following statements must be inside an interface block, scoping unit, or construct 
to be in a function subprogram.
.BL
\*CSEQUENCE, END TYPE\fR in a derived type definition
.BL
\*CMODULE PROCEDURE\fR in a generic interface block
.BL
\*CELSE, ELSEIF, END IF\fR in an \*CIF\fR construct
.BL
\*CCASE, END SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE, END WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE, EXIT, END DO\fR in a \*CDO\fR construct
.BL
\*CFUNCTION, SUBROUTINE, END SUBROUTINE, END INTERFACE\fR in an 
interface block
.ME
$
$ Error : 0262
$msg  0262 The %s statement is not in the range of a matching DO construct.
$nexp 0262
Error : The %s statement is not in the range of a matching DO construct.
.PP
The \*CCYCLE\fR and \*CEXIT\fR statements must be specified in the range of a \*CDO\fR construct.
If the \*CCYCLE\fR or \*CEXIT\fR statement refers to a \*CDO\fR construct name, it must be within
the range of the \*CDO\fR construct with the same \*CDO\fR construct name.
.ME
$
$ Internal : 0263
$msg  0263 The semantics pass statement driver encountered an illegal or unknown statement type.
$nexp 0263
Internal : The semantics pass statement driver encountered an illegal or unknown statement type.
.PP
The driver that applies semantic processing to each statement in the semantics
pass selects the appropriate semantics routine from an array of function 
pointers.  The index into the array is not valid.  It must be a positive value,
but is probably zero or a negative value.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Error : 0264
$msg  0264 The %s statement is not allowed following a CONTAINS statement.  Expecting FUNCTION, SUBROUTINE or END statement.
$nexp 0264
Error : The %s statement is not allowed following a CONTAINS statement.  Expecting FUNCTION, SUBROUTINE or END statement.
.PP
The only statements allowed following a \*CCONTAINS\fR statement are \*CSUBROUTINE\fR,
\*CFUNCTION\fR, and \*CEND\fR.  \*CSUBROUTINE\fR and \*CFUNCTION\fR start a new internal or 
module subprogram.
.ME
$
$ Ansi : 0265
$msg  0265 Specifying the NAMELIST statement after the first executable statement is an extension to the Fortran standard.
$nexp 0265
Ansi : Specifying the NAMELIST statement after the first executable statement is an extension to the Fortran standard.
.PP
The Fortran standard requires all \*CNAMELIST\fR statements to be in the
specification part of a program unit.  It is an extension to the
Fortran standard to allow the \*CNAMELIST\fR statement to be interspersed with
executable statements.  However, all definitions of a namelist group must 
occur before any reference to the namelist group.
.ME
$
$ Ansi : 0266
$msg  0266 The use of edit descriptor -nX is an extension to the Fortran standard.
$nexp 0266
Ansi : The use of edit descriptor -nX is an extension to the Fortran standard.
.PP
The use of a negative value with the X edit descriptor is an extension to 
the Fortran standard.
.ME
$
$ Error : 0267
$msg  0267 The %s statement is not allowed in the THEN block of an IF construct.
$nexp 0267
Error : The %s statement is not allowed in the THEN block of an IF construct.
.PP
The following statements are allowed in an \*CIF\fR construct:
.PP
.nf
\*CALLOCATE               ASSIGN
BACKSPACE              BUFFER
CALL                   CLOSE
CONTINUE               DATA
DEALLOCATE             DECODE
DO                     ENCODE
ENDFILE                ENDIF
ELSE                   ELSEIF
FORMAT                 GOTO
IF                     INQUIRE
NAMELIST               NULLIFY
OPEN                   PAUSE
PRINT                  READ
RETURN                 REWIND
SELECT                 STOP
WHERE                  WRITE\fR
assignment
.fi
.PP
The following statements are allowed if the listed condition is met.
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE\fR, \*CEXIT\fR, \*CEND DO\fR in a \*CDO\fR construct
.ME
$
$ Error : 0268
$msg  0268 The %s statement is not allowed in a DO construct.
$nexp 0268
Error : The %s statement is not allowed in a DO construct.
.PP
The following statements are allowed in a \*CDO\fR construct:
.PP
.nf
\*CALLOCATE               ASSIGN
BACKSPACE              BUFFER
CALL                   CLOSE
CONTINUE               DATA
DEALLOCATE             DECODE
DO                     ENCODE
ENDFILE                ENDIF
ELSE                   ELSEIF
FORMAT                 GOTO
IF                     INQUIRE
NAMELIST               NULLIFY
OPEN                   PAUSE
PRINT                  READ
RETURN                 REWIND
SELECT                 STOP
WHERE                  WRITE\fR
assignment
.fi
.PP
The following statements are allowed in the listed constructs.
.BL
\*CCASE\fR, \*CEND SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSE\fR, \*CELSEIF\fR, \*CEND IF\fR in an \*CIF\fR construct
.BL
\*CELSEWHERE\fR, \*CEND WHERE\fR in a \*CWHERE\fR construct
.ME
$
$ Error : 0269
$msg  0269 The %s statement is not allowed following a SELECT CASE statement.
$nexp 0269
Error : The %s statement is not allowed following a SELECT CASE statement.
.PP
The \*CCASE\fR statement is the only statement that can follow a \*CSELECT CASE\fR
statement.  The compiler found another statement (not a \*CCASE\fR statement)
following the \*CSELECT CASE\fR statement.
.ME
$
$ Error : 0270
$msg  0270 The %s statement is not allowed in a WHERE construct.
$nexp 0270
Error : The %s statement is not allowed in a WHERE construct.
.PP
Only array assignment statements are allowed in the \*CWHERE\fR block or \*CELSEWHERE\fR
block of a \*CWHERE\fR construct.  The compiler found another statement (not
an array assignment statement) in a \*CWHERE\fR or \*CELSEWHERE\fR block.
.ME
$
$ Error : 0271
$msg  0271 The %s statement is not allowed in an interface body.
$nexp 0271
Error : The %s statement is not allowed in an interface body.
.PP
The following statements are allowed in an interface body.
.TL 30
\*CALLOCATABLE
COMMON\fR 
.TL
derived type \fRdefinitions
\*CDIMENSION\fR statement
.TL
\*CEND FUNCTION
END SUBROUTINE 
.TL
EQUIVALENCE 
EXTERNAL 
.TL
IMPLICIT 
\*CIMPLICIT NONE
.TL
INTENT 
INTRINSIC 
.TL
NAMELIST 
OPTIONAL 
.TL
PARAMETER 
POINTER
.TL
SAVE 
TARGET
.TL
\fRtype declaration\*C 
USE 
.PP
\fR The following statements are allowed if the listed condition is met:
.BL
\*CSEQUENCE, END TYPE\fR in a derived type definition
.BL
\*CPUBLIC\fR, \*CPRIVATE\fR in a module program unit
.ME
$
$ Error : 0272
$msg  0272 The %s statement is not allowed in a subroutine subprogram.
$nexp 0272
Error : The %s statement is not allowed in a subroutine subprogram.
.PP
The \*CPRIVATE\fR and \*CPUBLIC\fR statements are not allowed in a subroutine
subprogram.
.PP
\fR These statements must be inside an interface block, scoping unit, or a construct
to be in a subroutine subprogram.
.BL
\*CSEQUENCE, END TYPE\fR in a derived type definition
.BL
\*CMODULE PROCEDURE\fR in a generic interface block
.BL
\*CELSE, ELSEIF, END IF\fR in an \*CIF\fR construct
.BL
\*CCASE, END SELECT\fR in a \*CCASE\fR construct
.BL
\*CELSEWHERE, END WHERE\fR in a \*CWHERE\fR construct
.BL
\*CCYCLE, EXIT, END DO\fR in a \*CDO\fR construct
.BL
\*CFUNCTION, END FUNCTION, SUBROUTINE, END SUBROUTINE, END INTERFACE\fR in an
interface block
.ME
$
$ Error : 0273
$msg  0273 The %s attribute is specified more than once for the component declaration.
$nexp 0273
Error : The %s attribute is specified more than once for the component declaration.
.PP
An attribute must only be specified once for a component.  The \*CDIMENSION\fR or
the \*CPOINTER\fR attribute has been specified more than once for the component, as in the
following example:
.CS
  TYPE :: ABC
    REAL, POINTER, POINTER :: IJK ! illegal
  END TYPE ABC
.CE
.ME
$
$ Ansi : 0274
$msg  0274 Fortran requires all intrinsic procedure actual arguments to be type integer or character for restricted expressions.
$nexp 0274
Ansi : Fortran requires all intrinsic procedure actual arguments to be type integer or character for restricted expressions.
.PP
In the Fortran standard, restricted expressions allow elemental function
references.  Each argument to the intrinsic must be type integer or character.
The compiler allows arguments to be typed other than integer or character, 
if the result of the expression is type integer or character.
.ME
$
$ Error : 0275
$msg  0275 The PUBLIC and/or PRIVATE attribute is specified more than once for derived type "%s".
$nexp 0275
Error : The PUBLIC and/or PRIVATE attribute is specified more than once for derived type "%s".
.PP
The \*CPUBLIC\fR or \*CPRIVATE\fR attribute can only be declared once for the derived type.
.ME
$
$ Error : 0276
$msg  0276 "%s" is a %s.  It must be a constant to be used in an initialization expression.
$nexp 0276
Error : "%s" is a %s.  It must be a constant to be used in an initialization expression.
.PP
The compiler detected an illegal use of the object.
It must be a constant to be in an initialization expression.
.ME
$
$ Error : 0277
$msg  0277 "%s" is not valid as a keyword for this call to "%s".
$nexp 0277
Error : "%s" is not valid as a keyword for this call to "%s".
.PP
The compiler detected an actual argument keyword that does not match a
dummy argument name in the explicit interface for the called routine.
.ME
$
$ Internal : 0278
$msg  0278 No semantics routine exists for statement type %s.
$nexp 0278
Internal : No semantics routine exists for statement type %s.
.PP
The statement level semantics processing driver encountered a statement
header that contains a statement type value for a statement for which there is
no semantic routine.
.PP
This message should never be generated.  Please notify your product support
organization with this error message number and any supporting information.
This message does not indicate a problem with your code, although you may be
able to change your code so that this error is not encountered.
.ME
$
$ Ansi : 0279
$msg  0279 The EXTERNAL attribute has been specified for "%s", the program unit being compiled.  This is nonstandard.
$nexp 0279
Ansi : The EXTERNAL attribute has been specified for "%s", the program unit being compiled.  This is nonstandard.
.PP
The compiler detected the following:   
.CS
    SUBROUTINE JOE()
    EXTERNAL JOE  ! This is nonstandard
    RETURN
    END
.CE
.ME
$
$ Ansi : 0280
$msg  0280 Fortran requires all intrinsic procedure actual arguments to be type integer for specification expressions.
$nexp 0280
Ansi : Fortran requires all intrinsic procedure actual arguments to be type integer for specification expressions.
.PP
In the Fortran standard, specification expressions allow elemental function 
references.  Each argument to the intrinsic must be type integer.  The compiler
allows arguments to be typed other than integer, as long as the result of the
expression is type integer.
.ME
$
$ Error : 0281
$msg  0281 The left hand side of an assignment statement must be a variable or a function result.
$nexp 0281
Error : The left hand side of an assignment statement must be a variable or a function result.
.PP
The left side of this assignment statement is not valid.  This could be
caused by assigning a value to a constant, external subprogram name,
namelist group name, or some other entity other than a variable or the
function result name of the current function.
.ME
$
$ Error : 0282
$msg  0282 Function "%s" is declared in an interface block, therefore it must not be typed assumed-length character.
$nexp 0282
Error : Function "%s" is declared in an interface block, therefore it must not be typed assumed-length character.
.PP
The only functions that can be typed as assumed-length character are external
functions.  If a function is declared inside an interface block, it must not
be typed as assumed-length character.  Interface blocks are only used for
explicitly declared external subprograms.
.ME
$
$ Error : 0283
$msg  0283 The name specified on the %s statement must match the %s name "%s".
$nexp 0283
Error : The name specified on the %s statement must match the %s name "%s".
.PP
\fRThe compiler detected that the name specified on the \*CEND\fR statement does
not match the name of the program unit or derived type.
.PP
The names must match for the following statements:
.nf
\*CPROGRAM/END PROGRAM \fIprogram_name\*C
MODULE/END MODULE  \fImodule_name\*C
BLOCKDATA/END BLOCK DATA \fIblock_data_name\*C
SUBROUTINE/END SUBROUTINE \fIsubroutine_name\*C
FUNCTION/END FUNCTION \fIfunction_name\*C
TYPE/END TYPE \fItype_name\fR
.fi
.ME
$
$ Error : 0284
$msg  0284 The %s construct has construct name "%s", therefore the %s statement must specify the same construct name.
$nexp 0284
Error : The %s construct has construct name "%s", therefore the %s statement must specify the same construct name.
.PP
If the \*CIF\fR, \*CDO\fR, or \*CCASE\fR construct has a construct name, the \*CEND IF\fR,
\*CEND DO\fR, or \*CEND SELECT\fR statement must specify the same construct name.  The
compiler detected a difference between the construct names or a construct
name is missing on the \*CEND\fR statement.
.ME
$
$ Error : 0285
$msg  0285 If the %s construct does not have a construct name, the %s statement must not have a construct name specified.
$nexp 0285
Error : If the %s construct does not have a construct name, the %s statement must not have a construct name specified.
.PP
If the \*CDO\fR, \*CIF\fR, or \*CCASE\fR construct does not have a construct name
specified, the corresponding \*CEND DO\fR, \*CEND IF\fR, or \*CEND SELECT\fR statement must not
have a construct name specified.
.ME
$
$ Error : 0286
$msg  0286 A derived type type-name must not be the same as the name of the intrinsic type %s.
$nexp 0286
Error : A derived type type-name must not be the same as the name of the intrinsic type %s.
.PP
The compiler detected one of the following:
.CS
     TYPE :: INTEGER
     TYPE :: REAL
     TYPE :: DOUBLE PRECISION
     TYPE :: COMPLEX
     TYPE :: CHARACTER
     TYPE :: LOGICAL
.CE
.PP
All of the above are illegal, because 
a derived type type-name must not be the
same as the name of any