Preface

This guide is part of a set of manuals that describe how to use the NVIDIA HPC Fortran, C++ and C compilers. These compilers include the NVFORTRAN, NVC++ and NVC compilers. They work in conjunction with an assembler, linker, libraries and header files on your target system, and include a CUDA toolchain, libraries and header files for GPU computing. You can use the NVIDIA HPC compilers to develop, optimize and parallelize applications for NVIDIA GPUs and x86-64, OpenPOWER and Arm Server multicore CPUs.

The NVIDIA HPC Compilers User’s Guide provides operating instructions for the NVIDIA HPC compilers command-level development environment. The NVIDIA HPC Compilers Reference Guide contains details concerning the NVIDIA compilers' interpretation of the Fortran, C++ and C language standards, implementation of language extensions, and command-level compilation. Users are expected to have previous experience with or knowledge of the Fortran, C++ and C programming languages. These guides do not teach the Fortran, C++ or C programming languages.

Audience Description

This manual is intended for scientists and engineers using the NVIDIA HPC compilers. To use these compilers, you should be aware of the role of high-level languages, such as Fortran, C++ and C as well as parallel programming models such as CUDA, OpenACC and OpenMP in the software development process, and you should have some level of understanding of programming. The NVIDIA HPC compilers are available on a variety of NVIDIA GPUs and x86-64, OpenPOWER and Arm CPU-based platforms and operating systems. You need to be familiar with the basic commands available on your system.

Compatibility and Conformance to Standards

Your system needs to be running a properly installed and configured version of the NVIDIA HPC compilers. For information on installing NVIDIA HPC compilers, refer to the Release Notes and Installation Guide included with your software.

For further information, refer to the following:

  • American National Standard Programming Language FORTRAN, ANSI X3. -1978 (1978).
  • ISO/IEC 1539-1 : 1991, Information technology – Programming Languages – Fortran, Geneva, 1991 (Fortran 90).
  • ISO/IEC 1539-1 : 1997, Information technology – Programming Languages – Fortran, Geneva, 1997 (Fortran 95).
  • ISO/IEC 1539-1 : 2004, Information technology – Programming Languages – Fortran, Geneva, 2004 (Fortran 2003).
  • ISO/IEC 1539-1 : 2010, Information technology – Programming Languages – Fortran, Geneva, 2010 (Fortran 2008).
  • ISO/IEC 1539-1 : 2018, Information technology – Programming Languages – Fortran, Geneva, 2018 (Fortran 2018).
  • Fortran 95 Handbook Complete ISO/ANSI Reference, Adams et al, The MIT Press, Cambridge, Mass, 1997.
  • The Fortran 2003 Handbook, Adams et al, Springer, 2009.
  • OpenACC Application Program Interface, Version 2.7, November 2018, http://www.openacc.org.
  • OpenMP Application Program Interface, Version 4.5, November 2015, http://www.openmp.org.
  • Programming in VAX Fortran, Version 4.0, Digital Equipment Corporation (September, 1984).
  • IBM VS Fortran, IBM Corporation, Rev. GC26-4119.
  • Military Standard, Fortran, DOD Supplement to American National Standard Programming Language Fortran, ANSI x.3-1978, MIL-STD-1753 (November 9, 1978).
  • American National Standard Programming Language C, ANSI X3.159-1989.
  • ISO/IEC 9899:1990, Information technology – Programming Languages – C, Geneva, 1990 (C90).
  • ISO/IEC 9899:1999, Information technology – Programming Languages – C, Geneva, 1999 (C99).
  • ISO/IEC 9899:2011, Information Technology – Programming Languages – C, Geneva, 2011 (C11).
  • ISO/IEC 14882:2011, Information Technology – Programming Languages – C++, Geneva, 2011 (C++11).
  • ISO/IEC 14882:2014, Information Technology – Programming Languages – C++, Geneva, 2014 (C++14).
  • ISO/IEC 14882:2017, Information Technology – Programming Languages – C++, Geneva, 2017 (C++17).

Organization

This manual contains detailed reference information about specific aspects of the compiler, such as the details of compiler options, directives, data types supported, and more. It contains these sections:

Fortran, C++ and C Data Types describes the data types that are supported by the NVIDIA HPC Fortran, C++ and C compilers.

Command-Line Options Reference provides a detailed description of most command-line options.

C++ Name Mangling describes the name mangling facility and explains the transformations of names of entities to names that include information on aspects of the entity’s type and a fully qualified name.

Runtime Environment describes details related to compiler code generation, including register conventions and calling conventions for Linux/x86-64 and Linux/OpenPOWER processor-based systems.

C++ Dialect Supported lists more details of the version of the C++ language that NVC++ supports.

x86-64 C++ and C MMX/SSE/AVX Intrinsics provides tables that list the MMX and SSE/SSE2/SSE3/SSSE3/SSE4a/ABM/AVX Inline Intrinsics supported in C++ and C programs.

Messages provides a list of Fortran compiler error messages.

Hardware and Software Constraints

This guide describes versions of the NVIDIA HPC compilers that target NVIDIA GPUs and x86-64, OpenPOWER and Arm CPUs. Details concerning environment-specific values and defaults and system-specific features or limitations are presented in the release notes delivered with the NVIDIA HPC compilers.

Conventions

This guide uses the following conventions:

italic
is used for emphasis.
Constant Width
is used for filenames, directories, arguments, options, examples, and for language statements in the text, including assembly language statements.
Bold
is used for commands.
[ item1 ]
in general, square brackets indicate optional items. In this case item1 is optional. In the context of p/t-sets, square brackets are required to specify a p/t-set.
{ item2 | item 3 }
braces indicate that a selection is required. In this case, you must select either item2 or item3.
filename ...
ellipsis indicate a repetition. Zero or more of the preceding item may occur. In this example, multiple filenames are allowed.
FORTRAN
Fortran language statements are shown in the text of this guide using a reduced fixed point size.
C++ and C
C++ and C language statements are shown in the test of this guide using a reduced fixed point size.

Terms

A number of terms related to systems, processors, compilers and tools are used throughout this guide. For example:

accelerator FMA -mcmodel=medium shared library
AVX host -mcmodel=small SIMD
CUDA hyperthreading (HT) MPI SSE
device large arrays MPICH static linking
driver linux86-64 NUMA x86-64
DWARF LLVM OpenPOWER Arm
dynamic library multicore ppc64le Aarch64

The following table lists the NVIDIA HPC compilers and their corresponding commands:

Table 1. NVIDIA HPC Compilers and Commands
Compiler or Tool Language or Function Command
NVFORTRAN ISO/ANSI Fortran 2003 nvfortran
NVC++ ISO/ANSI C++17 with GNU compatibility nvc++
NVC ISO/ANSI C11 nvc

In general, the designation NVFORTRAN is used to refer to the NVIDIA Fortran compiler, and nvfortran is used to refer to the command that invokes the compiler. A similar convention is used for each of the NVIDIA HPC compilers.

For simplicity, examples of command-line invocation of the compilers generally reference the nvfortran command, and most source code examples are written in Fortran. Use of NVC⁠+⁠+ and NVC is consistent with NVFORTRAN, though there are command-line options and features of these compilers that do not apply to NVFORTRAN, and vice versa.

There are a wide variety of x86-64 CPUs in use. Most of these CPUs are forward-compatible, but not backward-compatible, meaning that code compiled to target a given processor will not necessarily execute correctly on a previous-generation processor.

A table listing the processor options that NVIDIA HPC compilers support is available in the Release Notes. The table also includes the features utilized by the compilers that distinguish them from a compatibility standpoint.

In this manual, the convention is to use "x86-64" to specify the group of CPUs that are x86-compatible, 64-bit enabled, and run a 64-bit operating system. x86-64 processors can differ in terms of their support for various prefetch, SSE and AVX instructions. Where such distinctions are important with respect to a given compiler option or feature, it is explicitly noted in this manual.

1. Fortran, C++ and C Data Types

This section describes the scalar and aggregate data types recognized by the NVIDIA Fortran, C++ and C compilers, the format and alignment of each type in memory, and the range of values each type can have on 64-bit operating systems.

1.1. Fortran Data Types

1.1.1. Fortran Scalars

A scalar data type holds a single value, such as the integer value 42 or the real value 112.6. The next table lists Fortran scalar data types, their size, format and range. Table 3 shows the range and approximate precision for Fortran real data types. Table 4 shows the alignment for different scalar data types. The alignments apply to all scalars, whether they are independent or contained in an array, a structure or a union.

Table 2. Representation of Fortran Data Types
Fortran Data Type Format Range
INTEGER 2's complement integer -231 to 231-1
INTEGER*2 2's complement integer -32768 to 32767
INTEGER*4 2's complement integer -231 to 231-1
INTEGER*8 2's complement integer -263 to 263-1
LOGICAL 32-bit value true or false
LOGICAL*1 8-bit value true or false
LOGICAL*2 16-bit value true or false
LOGICAL*4 32-bit value true or false
LOGICAL*8 64-bit value true or false
BYTE 2's complement -128 to 127
REAL Single-precision floating point 10-37 to 1038(1)
REAL*2 Half-precision floating point (binary16) 10-4 to 10 5(1)
REAL*4 Single-precision floating point 10-37 to 10 38(1)
REAL*8 Double-precision floating point 10-307 to 10 308(1)
DOUBLE PRECISION Double-precision floating point 10-307 to 10308(1)
COMPLEX Single-precision floating point 10-37 to 1038(1)
DOUBLE COMPLEX Double-precision floating point 10-307 to 10308(1)
COMPLEX*16 Double-precision floating point 10-307 to 10308(1)
CHARACTER*n Sequence of n bytes  

(1) Approximate value

The logical constants .TRUE. and .FALSE. are all ones and all zeroes, respectively. Internally, the value of a logical variable is true if the least significant bit is one and false otherwise. When the option -⁠Munixlogical is set, a logical variable with a non-zero value is true and with a zero value is false.

Note: A variable of logical type may appear in an arithmetic context, and the logical type is then treated as an integer of the same size.
Table 3. Real Data Type Ranges
Data Type Binary Range Decimal Range Digits of Precision
REAL -2-126 to 2128 10-37 to 1038(1) 7–8
REAL*2 -2-14 to 216 10-4 to 105(1) 3–4
REAL*8 -2-1022 to 21024 10-307 to 10308(1) 15–16
Table 4. Scalar Type Alignment
This Type... ...Is aligned on this size boundary
LOGICAL*1 1-byte
LOGICAL*2 2-byte
LOGICAL*4 4-byte
LOGICAL*8 8-byte
BYTE 1-byte
INTEGER*2 2-byte
INTEGER*4 4-byte
INTEGER*8 8-byte
REAL*2 2-byte
REAL*4 4-byte
REAL*8 8-byte
COMPLEX*8 4-byte
COMPLEX*16 8-byte

1.1.2. FORTRAN real(2)

The NVFORTRAN compiler supports real(2) data type which makes it possible to declare and use data in half precision floating point. It is explicitly required to use the kind attribute with value of 2 on real data type to take advantage of this support. The following operators are supported for this data type: + , -, *, /, .lt., .le., .gt., .ge., .eq.,.ne..

There are several ways to create real(2) constants:

  ! Using kind attribute of 2 by appending _2 to the floating point value:
  real(2) :: val1 = 2.0_2
  ! Using a hexadecimal constant:
  real(2) :: val2 = z'4000'
  ! Explicitly calling real() intrinsic with the value to be converted:
  real(2) :: val3 = real(2, kind=2)
  ! Implicitly relying on compiler to convert value to real(2):
  real(2) :: val4 = 2d0
    

Half precision native support is not available on all of the architecture targets that NVFORTRAN supports. It is still possible to use this type, but be aware that implementation relies on conversion to real(4), handling operation in real(4), and then converting back to real(2). NVIDIA GPUs which support CUDA Compute Capability 6.0 and above implement operations natively and do not rely on conversion.

Half precision is represented as IEEE 754 binary16. Out of the 16-bits available to represent the floating point value, one bit is used for sign, five bits are used for exponent, and ten bits are used for significand. When encountering values that cannot be precisely represented in the format, such as when adding two real(2) numbers, IEEE 754 defines rounding rules. In the case of real(2), the default rule is round-to-nearest with ties-to-even property which is described in detail in the IEEE 754-2008 standard in section 4.3.1. This format has a small dynamic range and thus values greater than 65520 are rounded to infinity.

1.1.3. FORTRAN 77 Aggregate Data Type Extensions

The NVFORTRAN compiler supports de facto standard extensions to FORTRAN 77 that allow for aggregate data types. An aggregate data type consists of one or more scalar data type objects. You can declare the following aggregate data types:

  • An array consists of one or more elements of a single data type placed in contiguous locations from first to last.
  • A structure can contain different data types. The members are allocated in the order they appear in the definition but may not occupy contiguous locations.
  • A union is a single location that can contain any of a specified set of scalar or aggregate data types. A union can have only one value at a time. The data type of the union member to which data is assigned determines the data type of the union after that assignment.

The alignment of an array, a structure or union (an aggregate) affects how much space the object occupies and how efficiently the processor can address members. Arrays use the alignment of their members.

Array types
align according to the alignment of the array elements. For example, an array of INTEGER*2 data aligns on a 2-byte boundary. The exception to this rule is that aligment of REAL*2 arrays is on a 4-byte boundary.
Structures and Unions
align according to the alignment of the most restricted data type of the structure or union. In the next example, the union aligns on a 4-byte boundary since the alignment of c, the most restrictive element, is four.
STRUCTURE /astr/
UNION
 MAP
 INTEGER*2 a ! 2 bytes
 END MAP
 MAP
 BYTE b ! 1 byte
 END MAP
 MAP
 INTEGER*4 c ! 4 bytes
 END MAP
END UNION
END STRUCTURE

Structure alignment can result in unused space called padding. Padding between members of the structure is called internal padding. Padding between the last member and the end of the space is called tail padding.

The offset of a structure member from the beginning of the structure is a multiple of the member's alignment. For example, since an INTEGER*2 aligns on a 2-byte boundary, the offset of an INTEGER*2 member from the beginning of a structure is a multiple of two bytes.

1.1.4. Fortran 90 Aggregate Data Types (Derived Types)

The Fortran 90 standard added formal support for aggregate data types. The TYPE statement begins a derived type data specification or declares variables of a specified user-defined type. For example, the following would define a derived type ATTENDEE:

TYPE ATTENDEE
 CHARACTER(LEN=30) NAME
 CHARACTER(LEN=30) ORGANIZATION
 CHARACTER (LEN=30) EMAIL
END TYPE ATTENDEE

In order to declare a variable of type ATTENDEE and access the contents of such a variable, code such as the following would be used:

TYPE (ATTENDEE) ATTLIST(100)
. . .
ATTLIST(1)%NAME = ‘JOHN DOE’

1.2. C and C++ Data Types

1.2.1. C and C++ Scalars

Table 5 lists C and C++ scalar data types, providing their size and format. The alignment of a scalar data type is equal to its size. Table 6 shows scalar alignments that apply to individual scalars and to scalars that are elements of an array or members of a structure or union. Wide characters are supported (character constants prefixed with an L). The size of each wide character is 4 bytes.

Table 5. C/C++ Scalar Data Types
Data Type Size (bytes) Format Range
unsigned char 1 ordinal 0 to 255
signed char 1 2's complement integer -128 to 127
char 1 2's complement integer -128 to 127
char 1 ordinal 0 to 255
unsigned short 2 ordinal 0 to 65535
[signed] short 2 2's complement integer -32768 to 32767
unsigned int 4 ordinal 0 to 232 -1
[signed] int 4 2's complement integer -231 to 231-1
[signed] long [int] (win64) 4 2's complement integer -231 to 231-1
[signed] long [int] (linux86-64) 8 2's complement integer -263 to 263-1
unsigned long [int] (win64) 4 ordinal 0 to 232-1
unsigned long [int] (linux86-64) 8 ordinal 0 to 264-1
[signed] long long [int] 8 2's complement integer -263 to 263-1
unsigned long long [int] 8 ordinal 0 to 264-1
[signed] __int128 16 2's complement integer -2127 to 2127-1
unsigned __int128 16 ordinal 0 to 2128-1
float 4 IEEE single-precision floating-point 10-37 to 1038(1)
double 8 IEEE double-precision floating-point 10-307 to 10308(1)
long double 16 IEEE extended-precision floating-point 10-4931 to 104932(1)
long double 16 IBM double-double 10-307 to 10308(1)
bit field(2) (unsigned value) 1 to 32 bits ordinal 0 to 2size-1, where size is the number of bits in the bit field
bit field(2) (signed value) 1 to 32 bits 2's complement integer -2size-1 to 2size-1-1, where size is the number of bits in the bit field
pointer (32-bit operating system) 4 address 0 to 232-1
pointer 8 address 0 to 264-1
enum 4 2's complement integer -231 to 231-1

(1) Approximate value

(2) Bit fields occupy as many bits as you assign them, up to 4 bytes, and their length need not be a multiple of 8 bits (1 byte)

Table 6. Scalar Alignment
Data Type Alignment on this size boundary
char 1-byte boundary, signed or unsigned.
short 2-byte boundary, signed or unsigned.
int 4-byte boundary, signed or unsigned.
enum 4-byte boundary.
pointer 8-byte boundary.
float 4-byte boundary.
double 8-byte boundary.
long double 8-byte boundary.
long double (64-bit operating system) 16-byte boundary.
long [int] linux86-64 8-byte boundary, signed or unsigned.
long long [int] 8-byte boundary, signed or unsigned.

1.2.2. C and C++ Aggregate Data Types

An aggregate data type consists of one or more scalar data type objects. You can declare the following aggregate data types:

array
consists of one or more elements of a single data type placed in contiguous locations from first to last.
class
(C++ only) is a class that defines an object and its member functions. The object can contain fundamental data types or other aggregates including other classes. The class members are allocated in the order they appear in the definition but may not occupy contiguous locations.
struct
is a structure that can contain different data types. The members are allocated in the order they appear in the definition but may not occupy contiguous locations. When a struct is defined with member functions, its alignment rules are the same as those for a class.
union
is a single location that can contain any of a specified set of scalar or aggregate data types. A union can have only one value at a time. The data type of the union member to which data is assigned determines the data type of the union after that assignment.

1.2.3. Class and Object Data Layout

Class and structure objects with no virtual entities and with no base classes, that is just direct data field members, are laid out in the same manner as C structures. The following section describes the alignment and size of these C-like structures. C++ classes (and structures as a special case of a class) are more difficult to describe. Their alignment and size is determined by compiler generated fields in addition to user-specified fields. The following paragraphs describe how storage is laid out for more general classes. The user is warned that the alignment and size of a class (or structure) is dependent on the existence and placement of direct and virtual base classes and of virtual function information. The information that follows is for informational purposes only, reflects the current implementation, and is subject to change. Do not make assumptions about the layout of complex classes or structures.

All classes are laid out in the same general way, using the following pattern (in the sequence indicated):

  • First, storage for all of the direct base classes (which implicitly includes storage for non-virtual indirect base classes as well):
    • When the direct base class is also virtual, only enough space is set aside for a pointer to the actual storage, which appears later.
    • In the case of a non-virtual direct base class, enough storage is set aside for its own non-virtual base classes, its virtual base class pointers, its own fields, and its virtual function information, but no space is allocated for its virtual base classes.
  • Next, storage for the class’s own fields.
  • Next, storage for virtual function information (typically, a pointer to a virtual function table).
  • Finally, storage for its virtual base classes, with space enough in each case for its own non-virtual base classes, virtual base class pointers, fields, and virtual function information.

1.2.4. Aggregate Alignment

The alignment of an array, a structure or union (an aggregate) affects how much space the object occupies and how efficiently the processor can address members.

Arrays
align according to the alignment of the array elements. For example, an array of short data type aligns on a 2-byte boundary.
Structures and Unions
align according to the most restrictive alignment of the enclosing members. In the following example, the union un1 aligns on a 4-byte boundary since the alignment of c, the most restrictive element, is four:
union un1 {
 short a; /* 2 bytes */
 char b; /* 1 byte */
 int c; /* 4 bytes */
 };

Structure alignment can result in unused space, called padding. Padding between members of a structure is called internal padding. Padding between the last member and the end of the space occupied by the structure is called tail padding. Figure 1 illustrates structure alignment. Consider the following structure:

struct strc1 {
 char a; /* occupies byte 0 */
 short b; /* occupies bytes 2 and 3 */
 char c; /* occupies byte 4 */
 int d; /* occupies bytes 8 through 11 */
 };
Figure 1. Internal Padding in a Structure
png for PDF.

Figure 2 shows how tail padding is applied to a structure aligned on a doubleword (8 byte) boundary.

struct strc2{
 int m1[4]; /* occupies bytes
0 through 15 */
 double m2; /* occupies bytes 16 through 23 */
 short m3; /* occupies bytes 24 and 25 */
} st;

1.2.5. Bit-field Alignment

Bit-fields have the same size and alignment rules as other aggregates, with several additions to these rules:

  • Bit-fields are allocated from right to left.
  • A bit-field must entirely reside in a storage unit appropriate for its type. Bit-fields never cross unit boundaries.
  • Bit-fields may share a storage unit with other structure/union members, including members that are not bit-fields.
  • Unnamed bit-field's types do not affect the alignment of a structure or union.
Figure 2. Tail Padding in a Structure
png for PDF.

1.2.6. Other Type Keywords in C and C++

The void data type is neither a scalar nor an aggregate. You can use void or void* as the return type of a function to indicate the function does not return a value, or as a pointer to an unspecified data type, respectively.

The const and volatile type qualifiers do not in themselves define data types, but associate attributes with other types. Use const to specify that an identifier is a constant and is not to be changed. Use volatile to prevent optimization problems with data that can be changed from outside the program, such as memory-mapped I/O buffers.

Extended integer types __int128 and unsigned __int128 are now supported by NVC and NVC++. 128-bit integer support can be turned on with the -Mint128 flag. Note, 128-bit integer support is not supported with OpenMP, OpenACC and CUDA.

2. Command-Line Options Reference

A command-line option allows you to specify specific behavior when a program is compiled and linked. Compiler options perform a variety of functions, such as setting compiler characteristics, describing the object code to be produced, controlling the diagnostic messages emitted, and performing some preprocessor functions. Most options that are not explicitly set take the default settings. This reference section describes the syntax and operation of each compiler option. For easy reference, the options are arranged in alphabetical order.

For an overview and tips on options usage and which options are best for which tasks, refer to the ‘Using Command-line Options’ section of the HPC Compilers User Guide, which also provides summary tables of the different options.

This section uses the following notation:

[item]
Square brackets indicate that the enclosed item is optional.
{item | item}
Braces indicate that you must select one and only one of the enclosed items. A vertical bar (|) separates the choices.
...
Horizontal ellipses indicate that zero or more instances of the preceding item are valid.

2.1. HPC Compilers Option Summary

The following tables include all the HPC compiler options that are not language-specific. The options are separated by category for easier reference.

For a complete description of each option, refer to the detailed information later in this section.

2.2. Generic Compiler Options

The following descriptions are for compiler options common to the NVIDIA HPC Fortran, C++ and C compilers. For easy reference, the options are arranged in alphabetical order. For a list of options by tasks, refer to the tables in the beginning of this section.

2.2.1. -#

Displays the invocations of the compiler, assembler and linker.

Default

The compiler does not display individual phase invocations.

Usage

The following command-line requests verbose invocation information.

$ nvfortran -# prog.f

Description

The -⁠# option displays the invocations of the compiler, assembler and linker. These invocations are command-lines created by the driver from your command-line input and the default value.

2.2.2. -[no]acc

Enable [disable] OpenACC directives. The following suboptions may be used following an equals sign ("="), with multiple sub-options separated by commas:

gpu
(default) OpenACC directives are compiled for GPU execution only.
host
Compile for serial execution on the host CPU.
multicore
Compile for parallel execution on the host CPU.
legacy
Suppress warnings about deprecated NVIDIA accelerator directives.
[no]autopar
Enable [disable] loop autoparallelization within acc parallel. The default is to autoparallelize, that is, to enable loop autoparallelization.
[no]routineseq
Compile every routine for the devicee. The default behavior is to not treat every routine as a seq directive.
strict
Instructs the compiler to issue warnings for non-OpenACC accelerator directives.
sync
Ignore async clauses
verystrict
Instructs the compiler to fail with an error for any non-OpenACC accelerator directive.
[no]wait
Wait for each device kernel to finish. Kernel launching is blocked by default unless the async clause is used.

Usage

The following command-line requests that OpenACC directives be enabled and that an error be issued for any non-OpenACC accelerator directive.

$ nvfortran -acc=verystrict prog.f

2.2.3. -Bdynamic

Compiles for and links to the shared object version of the NVIDIA HPC Compilers runtime libraries.

Default

Dynamic linking is the default behavior for Linux.

Usage

% nvfortran -Bdynamic myprogram.f

When you use the NVIDIA HPC compiler flag -⁠Bdynamic to create an executable that links to the shared object form of the runtime, the executable built is smaller than one built without -⁠Bdynamic. The NVIDIA HPC Compilers runtime shared object(s), however, must be available on the system where the executable is run. The -⁠Bdynamic flag must be used when an executable is linked against a shared object built by the NVIDIA HPC compilers.

2.2.4. -byteswapio

Swaps the byte-order of data in unformatted Fortran data files on input/output.

Default

The compiler does not byte-swap data on input/output.

Usage

The following command-line requests that byte-swapping be performed on input/output.

$ nvfortran -byteswapio myprog.f

Description

Use the -⁠byteswapio option to swap the byte-order of data in unformatted Fortran data files on input/output. When this option is used, the order of bytes is swapped in both the data and record control words; the latter occurs in unformatted sequential files.

You can use this option to convert big-endian format data files produced by most legacy RISC workstations to the little-endian format used on modern Linux systems on the fly during file reads/writes.

This option assumes that the record layouts of unformatted sequential access and direct access files are the same on the systems. It further assumes that the IEEE representation is used for floating-point numbers. In particular, the format of unformatted data files produced by NVIDIA HPC Fortran compilers is identical to the format used on Sun and SGI workstations; this format allows you to read and write unformatted Fortran data files produced on those platforms from a program compiled for modern Linux platform using the -⁠byteswapio option.

2.2.5. -C

(Fortran only) Generates code to check array bounds.

Default

The compiler does not enable array bounds checking.

Usage

In this example, the compiler instruments the executable produced from myprog.f to perform array bounds checking at runtime:

$ nvfortran -C myprog.f

Description

Use this option to enable array bounds checking. If an array is an assumed size array, the bounds checking only applies to the lower bound. If an array bounds violation occurs during execution, an error message describing the error is printed and the program terminates. The text of the error message includes the name of the array, the location where the error occurred (the source file and the line number in the source), and information about the out of bounds subscript (its value, its lower and upper bounds, and its dimension).

2.2.6. -c

Halts the compilation process after the assembling phase and writes the object code to a file.

Default

The compiler produces an executable file and does not use the -⁠c option.

Usage

In this example, the compiler produces the object file myprog.o in the current directory.

$ nvfortran -c myprog.f

Description

Use the -⁠c option to halt the compilation process after the assembling phase and write the object code to a file. If the input file is filename.f, the output file is filename.o.

-c++libs

Instructs the compiler to append C⁠+⁠+ runtime libraries to the link line for programs built using NVFORTRAN.

Default

The NVFORTRAN compiler does not append the C++ runtime libraries to the link line.

Usage

In the following example the C⁠+⁠+ runtime libraries are linked with an object file compiled with NVFORTRAN

$ nvfortran main.f90 mycpp.o -c++libs

Description

Use this option to instruct the NVIDIA Fortran compiler to append C⁠+⁠+ runtime libraries to the link line.

2.2.8. -cuda

Enable CUDA; please refer to -gpu for target-specific options. The following suboptions may be used following an equals sign ("="), with multiple sub-options separated by commas:

charstring
Enable limited support for character strings in GPU kernels.
madconst
Put Module Array Descriptors in CUDA Constant Memory

Usage

The following command-line requests that CUDA interoperability be enabled and CUDA Fortran syntax be recognized and processed in all Fortran files.

$ nvfortran -cuda myprog.f

2.2.9. -cudalib

Add CUDA-optimized libraries to the link line. When no sub-option is specified the compiler will link all necessary CUDA-optimized libraries. -cudalib will use the version of the library appropriate to the CUDA version being used. The following libraries may be specified following an equals sign ("="), with multiple libraries separated by commas:

cublas
Link in the cuBLAS library.
cufft
Link in the cuFFT library.
cufftw
Link in the cuFFTW library.
curand
Link in the cuRAND library.
cusolver
Link in the cuSOLVER library.
cusparse
Link in the cuSPARSE library.
cutensor
Link in the cuTENSOR library.
nvblas
Link in the NVBLAS library.
nccl
Link in the NCCL library.
nvlamath
Link in the NVLAmath library.
nvshmem
Link in the NVSHMEM library.

Usage

The following command-line links in all necessary CUDA libraries.

$ nvfortran -acc -cudalib myprog.cpp

2.2.10. -D

Creates a preprocessor macro with a given value.

Note:

You can use the -⁠D option more than once on a compiler command line. The number of active macro definitions is limited only by available memory.

Syntax

-Dname[=value]

Where name is the symbolic name and value is either an integer value or a character string.

Default

If you define a macro name without specifying a value, the preprocessor assigns the string 1 to the macro name.

Usage

In the following example, the macro PATHLENGTH has the value 256 until a subsequent compilation. If the -⁠D option is not used, PATHLENGTH is set to 128.

$ nvfortran -DPATHLENGTH=256 myprog.F

The source text in myprog.F is this:

	#ifndef PATHLENGTH
#define PATHLENGTH 128
#endif  SUBROUTINE SUB  CHARACTER*PATHLENGTH path
        ...
END

Description

Use the -⁠D option to create a preprocessor macro with a given value. The value must be either an integer or a character string.

You can use macros with conditional compilation to select source text during preprocessing. A macro defined in the compiler invocation remains in effect for each module on the command line, unless you remove the macro with an #undef preprocessor directive or with the -⁠U option. The compiler processes all of the -⁠U options in a command line after processing the -⁠D options.

2.2.11. -d<arg>

Prints additional information from the preprocessor. [Valid only for the C compiler (nvc) ]

Default

No additional information is printed from the preprocessor.

Syntax

-d[D|I|M|N]
-dD
Print macros and values from source files.
-dI
Print include file names.
-dM
Print macros and values, including predefined and command-line macros.
-dN
Print macro names from source files.

Usage

In the following example, the compiler prints macro names from the source file.

$ nvc -dN myprog.f

Description

Use the -d<arg> option to print additional information from the preprocessor.

2.2.12. -dryrun

Displays the invocations of the compiler, assembler, and linker but does not execute them.

Default

The compiler does not display individual phase invocations.

Usage

The following command line requests verbose invocation information.

$ nvfortran -dryrun myprog.f

Description

Use the -⁠dryrun option to display the invocations of the compiler, assembler, and linker but not have them executed. These invocations are command lines created by the compiler driver from the rc files and the command-line supplied with -⁠dryrun.

2.2.13. -drystdinc

Displays the standard include directories and then exits the compiler.

Default

The compiler does not display standard include directories.

Usage

The following command line requests a display for the standard include directories.

$ nvc -drystdinc myprog.c

Description

Use the -⁠drystdinc option to display the standard include directories and then exit the compiler.

2.2.14. -E

Halts the compilation process after the preprocessing phase and displays the preprocessed output on the standard output.

Default

The compiler produces an executable file.

Usage

In the following example the compiler displays the preprocessed myprog.f on the standard output.

$ nvc -E myprog.c

Description

Use the -⁠E option to halt the compilation process after the preprocessing phase and display the preprocessed output on the standard output.

2.2.15. -F

Stops compilation after the preprocessing phase.

Default

The compiler produces an executable file.

Usage

In the following example the compiler produces the preprocessed file myprog.f in the current directory.

$ nvfortran -F myprog.F

Description

Use the -⁠F option to halt the compilation process after preprocessing and write the preprocessed output to a file. If the input file is filename.F, then the output file is filename.f.

2.2.16. -fast

Enables vectorization with SIMD instructions, cache alignment, and flushz for 64-bit targets.

Default

The compiler does not enable vectorization with SIMD instructions, cache alignment, and flushz.

Usage

In the following example the compiler produces vector SIMD code when targeting a 64-bit machine.

$ nvfortran -fast vadd.f95

Description

When you use this option, a generally optimal set of options is chosen for targets that support SIMD capability. In addition, the appropriate -⁠tp option is automatically included to enable generation of code optimized for the type of system on which compilation is performed. This option enables vectorization with SIMD instructions, cache alignment, and flushz.

Note: Auto-selection of the appropriate -⁠tp option means that programs built using the -⁠fast option on a given system are not necessarily backward-compatible with older systems.
Note: C/C++ compilers enable -⁠Mautoinline with -⁠fast.

2.2.17. --flagcheck

Causes the compiler to check that flags are correct and then exit without any compilation occuring.

Default

The compiler begins a compile without the additional step to first validate that flags are correct.

Usage

In the following example the compiler checks that flags are correct, and then exits.

$ nvfortran --flagcheck myprog.f

Description

Use this option to make the compiler check that flags are correct and then exit. If flags are all correct then the compiler returns a zero status. No compilation occurs.

2.2.18. -fortranlibs

Instructs the C++ or C compiler to append NVFORTRAN runtime libraries to the link line.

Default

The C++ and compilers do not append the NVFORTRAN runtime libraries to the link line.

Usage

In the following example a .c main program is linked with an object file compiled with nvfortran.

$ nvc main.c myfort.o -fortranlibs

Description

Use this option to instruct the C++ or C compiler to append NVFORTRAN runtime libraries to the link line.

2.2.19. -fpic

Generates position-independent code suitable for inclusion in shared object (dynamically linked library) files.

Default

The compiler does not generate position-independent code.

Usage

In the following example the resulting object file, myprog.o, can be used to generate a shared object.

$ nvfortran -fpic myprog.f

Use the -fpic option to generate position-independent code suitable for inclusion in shared object (dynamically linked library) files.

2.2.20. -fPIC

Equivalent to -⁠fpic. Provided for compatibility with other compilers.

2.2.21. -g

Instructs the compiler to include symbolic debugging information in the object module; sets the optimization level to zero unless a -⁠O option is present on the command line.

Default

The compiler does not put debugging information into the object module.

Usage

In the following example, the object file myprog.o contains symbolic debugging information.

$ nvfortran -c -g myprog.f

Description

Use the -⁠g option to instruct the compiler to include symbolic debugging information in the object module. Debuggers require symbolic debugging information in the object module to display and manipulate program variables and source code.

If you specify the -⁠g option on the command-line, the compiler sets the optimization level to -⁠O0 (zero), unless you specify the -⁠O option. For more information on the interaction between the -⁠g and -⁠O options, refer to the -⁠O entry. Symbolic debugging may give confusing results if an optimization level other than zero is selected.

Note:

Including symbolic debugging information increases the size of the object module.

2.2.22. -g77libs

Used on the link line, this option instructs the nvfortran driver to search the necessary g77 or gfortran support libraries to resolve references specific to g77- or gfortran-compiled program units.

Note: The g77 or gfortran compiler must be installed on the system on which linking occurs in order for this option to function correctly.

Default

The compiler does not search g77 or gfortran support libraries to resolve references at link time.

Usage

The following command-line requests that g77 and gfortran support libraries be searched at link time:

$ nvfortran -g77libs myprog.f g77_object.o

Description

Use the -⁠g77libs option on the link line if you are linking g77- or gfortran-compiled program units into a nvfortran-compiled main program using the nvfortran driver. When this option is present, the nvfortran driver searches the necessary g77 and gfortran support libraries to resolve references specific to g77- or gfortran-compiled program units.

2.2.23. --gcc-toolchain=<path>

Specify the gcc toolchain location for use during compilation.

Default

Compiles using the default gcc toolchain location (selected during installation).

Usage

The following examples compile using the specified gcc 9.3.0 toolchain.

$ nvc++ --gcc-toolchain=~/gcc/gcc-9.3.0/  myprog.cpp
$ nvc++ --gcc-toolchain=~/gcc/gcc-9.3.0/bin/  myprog.cpp
$ nvc++ --gcc-toolchain=~/gcc/gcc-9.3.0/bin/gcc  myprog.cpp

Description

The argument can either be gcc root directory, <root directory>/bin, or the gcc executable itself.

2.2.24. -gopt

Instructs the compiler to include symbolic debugging information in the object file, and to generate optimized code identical to that generated when -⁠g is not specified.

Default

The compiler does not put debugging information into the object module.

Usage

In the following example, the object file myprog.o contains symbolic debugging information.

$ nvfortran -c -gopt myprog.f

Description

Using -⁠g alters how optimized code is generated in ways that are intended to enable or improve debugging of optimized code. The -⁠gopt option instructs the compiler to include symbolic debugging information in the object file, and to generate optimized code identical to that generated when -⁠g is not specified.

2.2.25. -gpu

Used in combination with the -⁠acc, -⁠cuda, -⁠mp, and -⁠stdpar flags to specify options for GPU code generation. The following sub-options may be used following an equals sign ("="), with multiple sub-options separated by commas:

autocompare
Automatically compare CPU vs GPU results at execution time: implies redundant
ccXY
Generate code for a device with compute capability X.Y. Multiple compute capabilities can be specified, and one version will be generated for each. By default, the compiler will detect the compute capability for each installed GPU. Use -⁠help -⁠gpu to see the valid compute capabilities for your installation.
ccall
Generate code for all compute capabilities supported by this platform and by the selected or default CUDA Toolkit.
ccall-major
Compile for all major supported compute capabilities.
ccnative
Detects the visible GPUs on the system and generates codes for them. If no device is available, the compute capability matching NVCC's default will be used.
cudaX.Y
Use CUDA X.Y Toolkit compatibility, where installed
[no]debug
Enable [disable] debug information generation in device code
deepcopy
Enable full deep copy of aggregate data structures in OpenACC; Fortran only
fastmath
Use routines from the fast math library
[no]flushz
Enable [disable] flush-to-zero mode for floating point computations on the GPU
[no]fma
Generate [do not generate] fused multiply-add instructions; default at -⁠O3
[no]implicitsections
Change [Do not change] array element references in a data clause into an array section. In C⁠+⁠+, the implicitsections option will change update device(a[n]) to update device(a[0:n]). In Fortran, it will change enter data copyin(a(n)) to enter data copyin(a(:n)). The default behavior, noimplicitsections, can also be changed using rcfiles; for example, one could add set IMPLICITSECTIONS=0; to siterc or another rcfile.
keep
Keep the kernel files (.bin, .ptx, source)
[no]lineinfo
Enable [disable] GPU line information generation
loadcache:{L1|L2}
Choose what hardware level cache to use for global memory loads; options include the default, L1, or L2
[no]managed
Allocate [do not allocate] any dynamically allocated data in CUDA Managed memory. Use -⁠gpu=nomanaged with -⁠stdpar to prevent that flag's implicit use of -⁠gpu=managed when CUDA Managed memory capability is detected.
maxregcount:n
Specify the maximum number of registers to use on the GPU; leaving this blank indicates no limit
pinned
Use CUDA Pinned Memory
ptxinfo
Print PTX info
[no]rdc
Generate [do not generate] relocatable device code.
redundant
Redundant CPU/GPU execution
safecache
Allow variable-sized array sections in cache directives; compiler assumes they fit into CUDA shared memory
sm_XY
Generate code for a device with compute capability X.Y. Multiple compute capabilities can be specified, and one version will be generated for each. By default, the compiler will detect the compute capability for each installed GPU. Use -⁠help -⁠gpu to see the valid compute capabilities for your installation.
stacklimit:<n>nostacklimit
Sets the limit (l) of stack variables in a procedure or kernel, in KB
[no]unified
Compile [do not compile] for CUDA Unified memory capability, where system memory is accessible from the GPU. This mode utilizes system and managed memory for dynamically allocated data unless explicit behavior is set through -⁠gpu=[no]managed. Use -⁠gpu=nounified with -⁠stdpar to prevent that flag's implicit use of -⁠gpu=unified when CUDA Unified memory capability is detected. This option must appear in both the compile and link lines.
[no]unroll
Enable [disable] automatic inner loop unrolling; default at -⁠O3
zeroinit
Initialize allocated device memory with zero

Usage

In the following example, the compiler generates code for NVIDIA GPUs with compute capabilities 6.0 and 7.0.

$ nvfortran -acc -gpu=cc60,cc70 myprog.f

The compiler automatically invokes the necessary software tools to create the kernel code and embeds the kernels in the object file.

To link in the appropriate GPU libraries, you must link an OpenACC program with the -⁠acc flag, and similarly for -⁠cuda, -⁠mp, or -⁠stdpar.

DWARF Debugging Formats

Use the -⁠g option to enable generation of full DWARF information on both the host and device; in the absence of other optimization flags, -⁠g sets the optimization level to zero. If a -⁠O option raises the optimization level to one or higher, only GPU line information is generated in device code even when -⁠g is specified. To enforce full DWARF generation for device code at optimization levels above zero, use the debug sub-option to -⁠gpu. Conversely, to prevent the generation of dwarf information for device code, use the nodebug sub-option to -⁠gpu. Both debug and nodebug can be used independently of -⁠g.

2.2.26. -help

Used with no other options, -⁠help displays options recognized by the driver on the standard output. When used in combination with one or more additional options, usage information for those options is displayed to standard output.

Default

The compiler does not display usage information.

Usage

In the following example, usage information for -⁠Minline is printed to standard output.

$ nvc -⁠help -⁠Minline   
-Minline[=lib:<inlib>|<maxsize>|<func>|except:<func>|name:<func>|maxsize:<n>|
totalsize:<n>|smallsize:<n>|reshape]
                    Enable function inlining
    lib:<inlib>     Use extracted functions from inlib
    <maxsize>       Set maximum function size to inline
    <func>          Inline function func
    except:<func>   Do not inline function func
    name:<func>     Inline function func
    maxsize:<n>     Inline only functions smaller than n
    totalsize:<n>   Limit inlining to total size of n
    smallsize:<n>   Always inline functions smaller than n
    reshape         Allow inlining in Fortran even when array shapes do not
                    match
    -Minline        Inline all functions that were extracted

In the following example, usage information for -⁠help shows how groups of options can be listed or examined according to function.

$ nvc -help -help 
-help[=groups|asm|debug|language|linker|opt|other|
overall|phase|prepro|suffix|switch|target|variable]

Description

Use the -⁠help option to obtain information about available options and their syntax. You can use -⁠help in one of three ways:

  • Use -⁠help with no parameters to obtain a list of all the available options with a brief one-line description of each.
  • Add a parameter to -⁠help to restrict the output to information about a specific option. The syntax for this usage is this:
    -help <command line option>
  • Add a parameter to -⁠help to restrict the output to a specific set of options or to a building process. The syntax for this usage is this:
    -help=<subgroup>

The following table lists and describes the subgroups available with -⁠help.

Table 11. Subgroups for -⁠help Option
Use this -⁠help option To get this information...
-help=asm A list of options specific to the assembly phase.
-help=debug A list of options related to debug information generation.
-help=groups A list of available switch classifications.
-help=language A list of language-specific options.
-help=linker A list of options specific to link phase.
-help=opt A list of options specific to optimization phase.
-help=other A list of other options, such as ANSI conformance pointer aliasing for C.
-help=overall A list of options generic to any NVIDIA HPC compiler.
-help=phase A list of build process phases and to which compiler they apply.
-help=prepro A list of options specific to the preprocessing phase.
-help=suffix A list of known file suffixes and to which phases they apply.
-help=switch A list of all known options; this is equivalent to usage of -⁠help without any parameter.
-help=target A list of options specific to target processor.
-help=variable A list of all variables and their current value. They can be redefined on the command line using syntax VAR=VALUE.

For more examples of -⁠help, refer to 'Help with Command-line Options.'

2.2.27. -I

Adds a directory to the search path for files that are included using either the INCLUDE statement or the preprocessor directive #include.

Default

The compiler searches only certain directories for included files.

  • For gcc-lib includes:/usr/lib64/gcc-lib
  • For system includes:/usr/include

Syntax

-Idirectory

Where directory is the name of the directory added to the standard search path for include files.

Usage

In the following example, the compiler first searches the directory mydir and then searches the default directories for include files.

$ nvfortran -Imydir

Description

Adds a directory to the search path for files that are included using the INCLUDE statement or the preprocessor directive #include. Use the -⁠I option to add a directory to the list of where to search for the included files. The compiler searches the directory specified by the -⁠I option before the default directories.

The Fortran INCLUDE statement directs the compiler to begin reading from another file. The compiler uses two rules to locate the file:

  • If the file name specified in the INCLUDE statement includes a path name, the compiler begins reading from the file it specifies.
  • If no path name is provided in the INCLUDE statement, the compiler searches (in order):
    1. Any directories specified using the -⁠I option (in the order specified)
    2. The directory containing the source file
    3. The current directory

    For example, the compiler applies rule (1) to the following statements:

    INCLUDE '/bob/include/file1' (absolute path name)
    INCLUDE '../../file1' (relative path name)

    and rule (2) to this statement:

    INCLUDE 'file1'

2.2.28. -i2, -⁠i4, -⁠i8

(Fortran only) Treat INTEGER and LOGICAL variables as either two, four, or eight bytes.

Default

The compiler treats INTERGER and LOGICAL variables as four bytes.

Usage

In the following example, using the -⁠i8 switch causes the integer variables to be treated as 64 bits.

$ nvfortran -i8 int.f

int.f is a function similar to this:

int.f
     print *, "Integer size:", bit_size(i)
     end

Description

Use this option to treat INTEGER and LOGICAL variables as either two, four, or eight bytes. INTEGER*8 values not only occupy 8 bytes of storage, but operations use 64 bits, instead of 32 bits.

  • -i2: Treat INTEGER variables as 2 bytes.
  • -i4: Treat INTEGER variables as 4 bytes.
  • -i8: Treat INTEGER and LOGICAL variables as 8 bytes and use 64-bits for INTEGER*8 operations.

2.2.29. -K<flag>

Requests that the compiler provide special compilation semantics with regard to conformance to IEEE 754.

Default

The default is -⁠Knoieee and the compiler does not provide special compilation semantics.

Syntax

-K<flag>

Where flag is one of the following:

ieee Perform floating-point operations in strict conformance with the IEEE 754 standard. Some optimizations are disabled, and on some systems a more accurate math library is linked if -⁠Kieee is used during the link step.
noieee Default flag. Use the fastest available means to perform floating-point operations, link in faster non-IEEE libraries if available, and disable underflow traps.
PIC or pic Generate position-independent code. Equivalent to -⁠fpic. Provided for compatibility with other compilers.
trap=option

[,option]...

Controls the behavior of the processor when floating-point exceptions occur.

Possible options include:

  • fp
  • align (ignored)
  • inv
  • denorm
  • divz
  • ovf
  • unf
  • inexact
  • none
Note: Same floating-point exception functionality as -Ktrap=option[,option] can be achieved with the runtime environment variable NVCOMPILER_FPU_STATE

Usage

In the following example, the compiler performs floating-point operations in strict conformance with the IEEE 754 standard

$ nvfortran -Kieee myprog.f

Description

Use -⁠K to instruct the compiler to provide special compilation semantics.

-⁠Ktrap is only processed by the compilers when compiling main functions or programs. The options inv, denorm, divz, ovf, unf, and inexact correspond to the processor’s exception mask bits: invalid operation, denormalized operand, divide-by-zero, overflow, underflow, and precision, respectively.

Normally, the processor’s exception mask bits are on, meaning that floating-point exceptions are masked – the processor recovers from the exceptions and continues. If a floating-point exception occurs and its corresponding mask bit is off, or "unmasked", execution terminates with an arithmetic exception (C's SIGFPE signal).

-⁠Ktrap=fp is equivalent to -⁠Ktrap=inv,divz,ovf.

Note: The NVIDIA HPC compilers do not support exception-free execution for -⁠Ktrap=inexact. The purpose of this hardware support is for those who have specific uses for its execution, along with the appropriate signal handlers for handling exceptions it produces. It is not designed for normal floating point operation code support.

2.2.30. -L

Specifies a directory to search for libraries.

Note: Multiple -⁠L options are valid. However, the position of multiple -⁠L options is important relative to -⁠l options supplied.

Default

The compiler searches the standard library directory.

Syntax

-Ldirectory

Where directory is the name of the library directory.

Usage

In the following example, the library directory is /lib and the linker links in the standard libraries required by NVFORTRAN from this directory.

$ nvfortran -L/lib myprog.f

In the following example, the library directory /lib is searched for the library file libx.a and both the directories /lib and /libz are searched for liby.a.

$ nvfortran -L/lib -lx -L/libz -ly myprog.f

Description

Use the -⁠L option to specify a directory to search for libraries. Using -⁠L allows you to add directories to the search path for library files.

2.2.31. -l<library>

Instructs the linker to load the specified library. The linker searches <library>in addition to the standard libraries.

Note: The linker searches the libraries specified with -⁠l in order of appearance before searching the standard libraries.

Syntax

-llibrary

Where library is the name of the library to search.

Usage: In the following example, if the standard library directory is /lib the linker loads the library /lib/libmylib.a, in addition to the standard libraries.

$ nvfortran myprog.f -lmylib

Description

Use this option to instruct the linker to load the specified library. The compiler prepends the characters lib to the library name and adds the .a extension following the library name. The linker searches each library specified before searching the standard libraries.

2.2.32. -M

Generate make dependence lists. You can use -⁠MD,filename (nvc++ only) to generate make dependence lists and print them to the specified file.

2.2.33. -M<nvflag>

Selects options for code generation. The options are divided into the following categories:

Code generation Fortran Language Controls Optimization
Environment C/C++ Language Controls Miscellaneous
Inlining    

The following table lists and briefly describes the options alphabetically and includes a field showing the category. For more details about the options as they relate to these categories, refer to ‘-⁠M Options by Category’.

Table 12. -M Options Summary
nvflag Description Category
allocatable=95|03 Controls whether to use Fortran 95 or Fortran 2003 semantics in allocatable array assignments. Fortran Language
anno Annotate the assembly code with source code. Miscellaneous
[no]autoinline When a C/C++ function is declared with the inline keyword, inline it at -⁠O2. Inlining
[no]asmkeyword Specifies whether the compiler allows the asm keyword in C/C++ source files (nvc and nvc++ only). C/C++ Language
[no]backslash Determines how the backslash character is treated in quoted strings (nvfortran only). Fortran Language
[no]bounds Specifies whether array bounds checking is enabled or disabled. Miscellaneous
[no]builtin Do/don't compile with math subroutine builtin support, which causes selected math library routines to be inlined (nvc and nvc++ only). Optimization
byteswapio Swap byte-order (big-endian to little-endian or vice versa) during I/O of Fortran unformatted data. Miscellaneous
cache_align Where possible, align data objects of size greater than or equal to 16 bytes on cache-line boundaries. Optimization
chkptr Check for NULL pointers (nvfortran only). Miscellaneous
chkstk Check the stack for available space upon entry to and before the start of a parallel region. Useful when many private variables are declared. Miscellaneous
concur Enable auto-concurrentization of loops. Multiple processors or cores will be used to execute parallelizable loops. Optimization
cpp Run the NVIDIA cpp-like preprocessor without performing subsequent compilation steps. Miscellaneous
cray Force Cray Fortran (CF77) compatibility (nvfortran only). Optimization
cuda Enables CUDA Fortran. Fortran Language
[no]daz Do/don’t treat denormalized operands as zero (default). Code Generation
[no]dclchk Determines whether all program variables must be declared (nvfortran only). Fortran Language
[no]defaultunit Determines how the asterisk character ("*") is treated in relation to standard input and standard output, regardless of the status of I/O units 5 and 6. (nvfortran only). Fortran Language
[no]depchk Checks for potential data dependencies. Optimization
[no]dse Enables [disables] dead store elimination phase for programs making extensive use of function inlining. Optimization
[no]dlines Determines whether the compiler treats lines containing the letter "D" in column one as executable statements (nvfortran only). Fortran Language
dollar,char Specifies the character (char) to which the compiler maps the dollar sign symbol (nvfortran only). Fortran Language
[no]dwarf Specifies [not] to add DWARF debug information. Code Generation
dwarf2 When used with -⁠g, generate DWARF2 format debug information. Code Generation
dwarf3 When used with -⁠g, generate DWARF3 format debug information. Code Generation
extend Instructs the compiler to accept 132-column source code; otherwise it accepts 72-column code (nvfortran only). Fortran Language
extract invokes the function extractor. Inlining
[no]fprelaxed[=option] Perform certain floating point intrinsic functions using relaxed precision. Optimization
fixed Instructs the compiler to assume F77-style fixed format source code (nvfortran only). Fortran Language
[no]flushz Do [not] treat denormalized results as zero (default). Code Generation
[no]fpapprox Specifies not to use low-precision fp approximation operations. Optimization
free Instructs the compiler to assume F90-style free format source code (nvfortran only). Fortran Language
func32 The compiler aligns all functions to 32-byte boundaries. Code Generation
gccbug[s] Matches behavior of certain gcc bugs Miscellaneous
info Prints informational messages regarding optimization and code generation to standard output as compilation proceeds. Miscellaneous
inform Specifies the minimum level of error severity that the compiler displays. Miscellaneous
inline Invokes the function inliner. Inlining
[no]iomutex Determines whether critical sections are generated around Fortran I/O calls (nvfortran only). Fortran Language
[no]ipa Invokes interprocedural analysis and optimization. Optimization
keepasm Preserve intermediate assembly language file. Miscellaneous
[no]large_arrays Enables support for 64-bit indexing and single static data objects of size larger than 2GB. Code Generation
list Specifies whether the compiler creates a listing file. Miscellaneous
[no]loop32 Aligns [does not align] innermost loops on 32-byte boundaries. Code Generation
[no]lre Enable [disable] loop-carried redundancy elimination. Optimization
[no]m128 Recognizes [ignores] __m128, __m128d, and __m128i datatypes. (nvc only) Code Generation
fcon Instructs the compiler to treat floating-point constants as float data types rather than the default double data type (nvc and nvc++ only). C/C++ Language
neginfo Instructs the compiler to produce information on why certain optimizations are not performed. Miscellaneous
noframe Eliminates operations that set up a true stack frame pointer for functions. Optimization
[no]i4 [do not] treat INTEGER variables and constants as INTEGER(KIND=4). Optimization
nomain When the link step is called, don’t include the object file that calls the Fortran main program. (nvfortran only). Code Generation
norpath On Linux, do not add -⁠rpath paths to the link line. Miscellaneous
[no]stddef Instructs the compiler to not recognize the standard preprocessor macros. Environment
nostdinc Instructs the compiler to not search the standard location for include files. Environment
nostdlib Instructs the linker to not link in the standard libraries. Environment
[no]onetrip Determines whether each DO loop executes at least once (nvfortran only). Language
novintr Disable idiom recognition and generation of calls to optimized vector functions. Optimization
preprocess Perform cpp-like preprocessing on assembly language and Fortran input source files. Miscellaneous
[no]r8 [do not] treat REAL variables and constants as REAL(KIND=8) (nvfortran only). Optimization
[no]r8intrinsics Determines how the compiler treats the intrinsics CMPLX and REAL (nvfortran only). Optimization
[no]recursive Allocate [do not allocate] local variables on the stack; this allows recursion. SAVEd, data-initialized, or namelist members are always allocated statically, regardless of the setting of this switch (nvfortran only). Code Generation
[no]reentrant Specifies whether the compiler avoids optimizations that can prevent code from being reentrant. Code Generation
[no]ref_externals [do not] force references to names appearing in EXTERNAL statements (nvfortran only). Code Generation
safeptr Instructs the compiler to override data dependencies between pointers and arrays (nvc and nvc++ only). Optimization
safe_lastval In the case where a scalar is used after a loop, but is not defined on every iteration of the loop, the compiler does not by default parallelize the loop. However, this option tells the compiler it is safe to parallelize the loop. For a given loop, the last value computed for all scalars make it safe to parallelize the loop. Code Generation
[no]save Determines whether the compiler assumes that all local variables are subject to the SAVE statement (nvfortran only). Fortran Language
schar Specifies signed char for characters (nvc and nvc++ only – also see uchar). C/C++ Language
[no]second_underscore Do [do not] add the second underscore to the name of a Fortran global if its name already contains an underscore (nvfortran only). Code Generation
[no]signextend Do [do not] extend the sign bit, if it is set. Code Generation
[no]single Do [do not] convert float parameters to double parameter characters (nvc and nvc++ only). C/C++ Language
standard Causes the compiler to flag source code that does not conform to the ANSI standard (nvfortran only). Fortran Language
[no]stride0 Do [do not] generate alternate code for a loop that contains an induction variable whose increment may be zero (nvfortran only). Code Generation
uchar Specifies unsigned char for characters (nvc and nvc++ only – also see schar). C/C++ Language
[no]unixlogical [do not] treat any non-zero logical variable as .TRUE. . (nvfortran only). Fortran Language
[no]unroll Controls loop unrolling. Optimization
[no]upcase Determines whether the compiler preserves uppercase letters in identifiers. Fortran keywords must be in lowercase (nvfortran only). Fortran Language
varargs Forces Fortran program units to assume calls are to C functions with a varargs type interface (nvfortran only) Code Generation
[no]vect Do [do not] invoke the code vectorizer. Optimization

2.2.34. -m

Displays a link map on the standard output.

Default

The compiler does not display the link map.

Usage

When the following example is executed, nvfortran writes the link map to stdout.

$ nvfortran -m myprog.f

Description

Use this option to display a link map.

  • On Linux, the map is written to stdout.

2.2.35. -march=<target>

An alias for -tp <target>. Please see -tp <target> for details.

2.2.36. -mcmodel=medium

Generates code for the medium memory model in the Linux execution environment. Implies -⁠Mlarge_arrays.

Default: The compiler generates code for the small memory model on Arm and x86-64 targets, the medium memory model on OpenPOWER targets.

Usage

The following command line requests position independent code be generated, and the option -⁠mcmodel=medium be passed to the assembler and linker:

$ nvfortran -mcmodel=medium myprog.f

Description

The small memory model limits the combined area for a user’s object or executable to 1GB, with the Linux kernel managing usage of the second 1GB of address for system routines, shared libraries, stacks, and so on. Programs are started at a fixed address, and the program can use a single instruction to make most memory references.

The medium memory model allows for larger than 2GB data areas, or .bss sections. Program units compiled using either -⁠mcmodel=medium or -⁠fpic require additional instructions to reference memory. The effect on performance is a function of the data-use of the application. The -⁠mcmodel=medium switch must be used at both compile time and link time to create 64-bit executables. Program units compiled for the default small memory model can be linked into medium memory model executables as long as they are compiled with the option -⁠fpic, or position-independent.

The Linux environment provides static libxxx.a archive libraries, that are built both with and without -⁠fpic, and dynamic libxxx.so shared object libraries that are compiled with -⁠fpic. Using the link switch -⁠mcmodel=medium implies the -⁠fpic switch and utilizes the shared libraries by default.

2.2.37. -mcpu=<target>

An alias for -tp <target>. Please see -tp <target> for details.

2.2.38. -module <moduledir>

Allows you to specify a particular directory in which generated intermediate .mod files should be placed.

Default

The compiler places .mod files in the current working directory, and searches only in the current working directory for pre-compiled intermediate .mod files.

Usage

The following command line requests that any intermediate module file produced during compilation of myprog.f be placed in the directory mymods; specifically, the file ./mymods/myprog.mod is used.

$ nvfortran -module mymods myprog.f

Description

Use the -⁠module option to specify a particular directory in which generated intermediate .mod files should be placed. If the -⁠module <moduledir> option is present, and USE statements are present in a compiled program unit, then <moduledir> is searched for .mod intermediate files prior to a search in the default local directory.

2.2.39. -[no]mp

Enable [disable] OpenMP directives. When enabled, it instructs the compiler to interpret user-inserted OpenMP parallel programming directives and pragmas, and to generate an executable file which will utilize multiple processors in a parallel system.

Default

The compiler does not interpret user-inserted OpenMP parallel programming directives and pragmas.

Usage

The following command line requests processing of any OpenMP directives present in myprog.f:

$ nvfortran -mp myprog.f

Description

Use the -⁠mp option to instruct the compiler to interpret user-inserted OpenMP parallel programming directives and to generate an executable file which utilizes multiple processors in a parallel system.

The suboptions are one or more of the following:

[no]align
Forces loop iterations to be allocated to OpenMP processes using an algorithm that maximizes alignment of vector sub-sections in loops that are both parallelized and SIMD vectorized. This allocation can improve performance in program units that include many such loops. It can also result in load-balancing problems that significantly decrease performance in program units with relatively short loops that contain a large amount of work in each iteration.
[no]autopar
Auto-parallelization of loops within omp loop is enabled by default. To disable this optimization, use the noautopar suboption.
gpu
OpenMP directives are compiled for GPU execution as well as host fallback to the CPU. For target-specific options, refer to the documentation for -⁠gpu.
multicore
OpenMP directives are compiled for multicore CPU execution only; this sub-option is the default.
ompt
Link against the OMPT-enabled OpenMP runtime library. OMPT is an interface that helps a first-party tool monitor the execution of an OpenMP program.

For more information about how the HPC Compilers support OpenMP, refer to the "Using OpenMP" section of the HPC Compilers User Guide.

2.2.40. -mtune=<target>

-mtune= is provided for compatibility. It performs no operation.

2.2.41. -noswitcherror

Issues warnings instead of errors for unknown switches. Ignores unknown command line switches after printing a warning message.

Default

The compiler prints an error message and then halts.

Usage

In the following example, the compiler ignores unknown command line switches after printing a warning message.

$ nvfortran -noswitcherror myprog.f

Description

Use this option to instruct the compiler to ignore unknown command line switches after printing a warning message.

Tip: You can configure this behavior in the siterc file by adding: set NOSWITCHERROR=1.

2.2.42. -O<level>

Invokes code optimization at the specified level.

Default

The compiler enables classical global optimization.

Syntax

-O [level]

Where level is an integer from 0 to 4 or "fast".

Usage

In the following example, since no -⁠O option is specified, the compiler sets the optimization to level 1.

$ nvfortran myprog.f

In the following example, since no optimization level is specified and a -⁠O option is specified, the compiler enables classical global optimizations.

$ nvfortran -O myprog.f

Description

Use this option to invoke code optimization. Using the NVIDIA compiler commands with the -⁠Olevel option (the capital O is for Optimize), you can specify any of the following optimization levels:

-O0
Level zero specifies no optimization. A basic block is generated for each language statement.
-O1
Level one specifies local optimization. Scheduling of basic blocks is performed. Register allocation is performed.
-O
When no level is specified, level global optimizations are performed, including traditional scalar optimizations, induction recognition, and loop invariant motion. No SIMD vectorization is enabled.
-O2
Level two specifies all level-1 and global optimizations, and enables more advanced optimizations such as SIMD code generation, cache alignment, and partial redundancy elimination.
-O3
Level three specifies aggressive global optimization. This level performs all level-one and level-two global optimizations and enables more aggressive hoisting and scalar replacement optimizations that may or may not be profitable.
-Ofast
Enables -O3 -Mfprelaxed -Mstack_arrays. Note -Mstack_array is disabled when -Ofast is used with -stdpar.
-O4
Level four performs all level-one, level-two, and level-three optimizations and enables hoisting of guarded invariant floating point expressions.

The following table shows the interaction between the -⁠O option, -⁠g option, -⁠Mvect, and -⁠Mconcur options.

Table 13. Optimization and -⁠O, -⁠g, -⁠Mvect, and -⁠Mconcur Options
Optimize Option Debug Option -M Option Optimization Level
none none none 1
none none -Mvect 2
none none -Mconcur 2
none -g none 0
-O none or -⁠g none 2
-Olevel none or -⁠g none level
-Olevel < 2 none or -⁠g -Mvect 2
-Olevel < 2 none or -⁠g -Mconcur 2

Unoptimized code compiled using the option -⁠O0 can be significantly slower than code generated at other optimization levels. Like the -⁠Mvect option, the -⁠Munroll option sets the optimization level to level-2 if no -⁠O or -⁠g options are supplied. The -⁠gopt option is recommended for generation of debug information with optimized code. For more information on optimization, refer to the ‘Multicore CPU Optimization’ section of the HPC Compilers User Guide.

2.2.43. -o

Names the executable file. Use the -⁠o option to specify the filename of the compiler object file. The final output is the result of linking.

Default

The compiler creates executable filenames as needed. If you do not specify the -⁠o option, the default filename is the linker output file a.out.

Syntax

-o filename

Where filename is the name of the file for the compilation output. The filename should not have a .f extension.

Usage

In the following example, the executable file ismyprog instead of the default a.outmyprog.exe.

$ nvfortran myprog.f -o myprog

2.2.44. -pg

Instructs the compiler to instrument the generated executable for gprof-style gmon.out sample-based profiling trace file.

Default

The compiler does not instrument the generated executable for gprof-style profiling.

Usage:

In the following example the program is compiled for profiling using gprof.

$ nvfortran -pg myprog.c

Description

Use this option to instruct the compiler to instrument the generated executable for gprof-style sample-based profiling. You must use this option at both the compile and link steps. A gmon.out style trace is generated when the resulting program is executed, and can be analyzed using gprof.

2.2.45. -R<directory>

Instructs the linker to hard-code the pathname <directory> into the search path for generated shared object (dynamically linked library) files.

Note: There cannot be a space between R and <directory>.

Usage

In the following example, at runtime the a.out executable searches the specified directory, in this case /home/Joe/myso, for shared objects.

$ nvfortran -R/home/Joe/myso myprog.f

Description

Use this option to instruct the compiler to pass information to the linker to hard-code the pathname <directory> into the search path for shared object (dynamically linked library) files.

2.2.46. -r

Creates a relocatable object file.

Default

The compiler does not create a relocatable object file and does not use the -⁠r option.

Usage

In this example, nvfortran creates a relocatable object file.

$ nvfortran -r myprog.f

Description

Use this option to create a relocatable object file.

2.2.47. -r4 and -⁠r8

(Fortran only) Interprets DOUBLE PRECISION variables as REAL (-⁠r4), or interprets REAL variables as DOUBLE PRECISION (-⁠r8). Note that these options do not override de facto standard type declarations that explicitly declare the number of bytes in the type name (REAL*4 and REAL*8).

Usage

In this example, the double precision variables are interpreted as REAL.

$ nvfortran -r4 myprog.f

Description

Interpret DOUBLE PRECISION variables as REAL (-⁠r4) or REAL variables as DOUBLE PRECISION (-⁠r8).

2.2.48. -rc

Specifies the name of the driver startup configuration file. If the file or pathname supplied is not a full pathname, the path for the configuration file loaded is relative to the $DRIVER path (the path of the currently executing driver). If a full pathname is supplied, that file is used for the driver configuration file.

Syntax

-rc [path] filename

Where path is either a relative pathname, relative to the value of $DRIVER, or a full pathname beginning with "/". Filename is the driver configuration file.

Usage

In the following example, the file .nvfortranrctest, relative to /opt/hpc_sdk/<target>/<release>/compilers/bin, the value of $DRIVER, is the driver configuration file.

$ nvfortran -rc .nvfortranrctest myprog.f

Description

Use this option to specify the name of the compiler driver startup configuration file. If the file or pathname supplied is not a full pathname, the path for the configuration file loaded is relative to the $DRIVER path – the path of the currently executing compiler driver. If a full pathname is supplied, that file is used for the compiler driver configuration file.

2.2.49. -S

Stops compilation after the compiling phase and writes the assembly-language output to a file.

Default

The compiler does not retain a .s file.

Usage

In this example, nvfortran produces the file myprog.s in the current directory.

$ nvfortran -S myprog.f

Description

Use this option to stop compilation after the compiling phase and then write the assembly-language output to a file. If the input file is filename.f, then the output file is filename.s.

2.2.50. -s

Strips the symbol-table information from the executable file.

Default

The compiler includes all symbol-table information and does not use the -⁠s option.

Usage

In this example, nvfortran strips symbol-table information from the a.out. executable file.

$ nvfortran -s myprog.f

Description

Use this option to strip the symbol-table information from the executable.

2.2.51. -shared

Instructs the compiler to pass information to the linker to produce a shared object (dynamically linked library) file.

Default

The compiler does not pass information to the linker to produce a shared object file.

Usage

In the following example the compiler passes information to the linker to produce the shared object file:myso.so.

$ nvfortran -shared myprog.f -o myso.so

Description

Use this option to instruct the compiler to pass information to the linker to produce a shared object (dynamically linked library) file.

2.2.52. -show

Produces driver help information describing the current driver configuration.

Default

The compiler does not show driver help information.

Usage

In the following example, the driver displays configuration information to the standard output after processing the driver configuration file.

$ nvfortran -show myprog.f

Description

Use this option to produce driver help information describing the current driver configuration.

2.2.53. -silent

Do not print warning messages.

Default

The compiler prints warning messages.

Usage

In the following example, the driver does not display warning messages.

$ nvfortran -silent myprog.f

Description

Use this option to suppress warning messages.

2.2.54. -soname

The compiler recognizes the -⁠soname option and passes it through to the linker.

Default

The compiler does not recognize the -⁠soname option.

Usage

In the following example, the driver passes the soname option and its argument through to the linker.

$ nvfortran -soname library.so myprog.f

Description

Use this option to instruct the compiler to recognize the -⁠soname option and pass it through to the linker.

2.2.55. -static

Statically link all libraries, including the NVIDIA HPC Compilers runtime.

Default

Dynamic linking is the default behavior for Linux

Usage

The following command line explicitly compiles for and links to the static version of the NVIDIA HPC Compilers runtime libraries:
% nvfortran -static -c object1.f

Description

You can use this option to explicitly compile for and link to the static versions of the system libraries and NVIDIA HPC Compilers runtime libraries.

2.2.56. -static-nvidia

Linux only. Compile and statically link only to the NVIDIA HPC Compilers runtime libraries. Other libraries are dynamically linked. Implies -⁠Mnorpath.

Default

The compiler uses static libraries.

Usage

The following command line explicitly compiles for and links to the static version of the NVIDIA HPC Compilers runtime libraries:

% nvfortran -static-nvidia -c object1.f

Description

You can use this option to explicitly compile for and link to the static version of the NVIDIA HPC Compilers runtime libraries.

Note: On Linux, -⁠static-nvidia results in code that runs on most Linux systems without requiring a Portability package.

2.2.57. -stdpar

Enable ISO C⁠+⁠+17 Parallel Algorithms behavior; please refer to -⁠gpu for target-specific options. The supported sub-options may be used following an equals sign ("="), with multiple sub-options separated by commas.

Default

Without sub-options, -⁠stdpar requests generation of code for execution of C⁠+⁠+ Parallel Algorithms on the GPU.

Sub-options

gpu
Execute C⁠+⁠+ Parallel Algorithms on the GPU; the default.
multicore
Execute C⁠+⁠+ Parallel Algorithms in parallel on the CPU.

Usage

The following command-line enables parallelization of C⁠+⁠+ Parallel Algorithms for offloading to a GPU.

$ nvc++ -stdpar myprog.cpp

2.2.58. -target

Select the target device for all parallel programming paradigms used (OpenACC, OpenMP, Standard Languages). The following suboptions may be used following an equals sign ("="), with multiple sub-options separated by commas:

gpu
Globally set the target device to an NVIDIA GPU.
multicore
Globally set the target device to a multicore CPU.

Usage

The following command-line enables parallelization of C++17 Parallel Algorithms and OpenACC, and globally designates the target device as an NVIDIA GPU.

$ nvc++ -stdpar -acc -target=gpu myprog.cpp

2.2.59. -time

Print execution times for various compilation steps.

Default

The compiler does not print execution times for compilation steps.

Usage

In the following example, nvfortran prints the execution times for the various compilation steps.

$ nvfortran -time myprog.f

Description

Use this option to print execution times for various compilation steps.

2.2.60. -tp <target>

Sets the target processor.

Default

The NVIDIA HPC compilers produce code specifically targeted to the type of processor on which compilation is performed. In particular, the default is to use all supported instructions wherever possible when compiling on a given system.

The default target processor is auto-selected depending on the processor on which the compilation is performed. You can specify a target processor different than the auto-selected default, but that target must be within the same CPU family as the processor on which compilation is performed. The NVIDIA HPC Compilers support 3 different families of CPUs: x86_64, OpenPOWER, and 64-bit Arm Server CPUs.

Executables created on a given system without the -tp flag may not be usable on previous generation systems. For example, executables created on an Intel Skylake processor may use AVX-512 or other instructions that are not available on earlier Intel processors or certain AMD processors.

Usage

In the following example, nvfortran sets the target processor to an Intel Skylake Xeon processor:

$ nvfortran -tp=skylake myprog.f

Description

Use this option to set the target architecture. By default, the NVIDIA HPC compilers use all supported instructions wherever possible when compiling on a given system.

Processor-specific optimizations can be specified or limited explicitly by using the -⁠tp option. Thus, it is possible to create executables that are usable on previous-generation systems.

The following list contains the possible suboptions for -⁠tp and the processors that each suboption is intended to target.

px
generate code that is usable on any x86-64 processor-based system.
host
generate code targeted for host processor. Link native version of HPC SDK cpu math library.
native
generate code targeted for host processor. Alias for -tp host.
x86-64-v2
generate code for the x86-64 microarchitectural level including SSE.
x86-64-v3
generate code for the x86-64 microarchitectural level including AVX2.
x86-64-v4
generate code for the x86-64 microarchitectural level including some AVX512 extensions.
bulldozer
generate code for AMD Bulldozer and compatible processors.
piledriver
generate code that is usable on any AMD Piledriver processor-based system.
zen
generate code that is usable on any AMD Zen processor-based system (e.g. Naples, Ryzen).
zen2
generate code that is usable on any AMD Zen 2 processor-based system (e.g. Rome, 3rd Gen Ryzen).
zen3
generate code that is usable on any AMD Zen 3 processor-based system (e.g. Milan, Ryzen 5000).
zen4
generate code that is usable on any AMD Zen 4 processor-based system (e.g. Genoa).
sandybridge
generate code for Intel Sandy Bridge and compatible processors.
haswell
generate code that is usable on any Intel Haswell processor-based system.
skylake
generate code that is usable on an Intel Skylake Xeon processor-based system.
cascadelake
generate code that is usable on an Intel Cascadelake Xeon processor-based system.
icelake
generate code that is usable on an Intel Icelake Xeon processor-based system.
host
generate code targeted for host processor. Link native version of HPC SDK cpu math library.
native
generate code targeted for host processor. Alias for -tp host.
a64fx
generate code that is usable on any Fujitsu A64fx processor-based system (SVE x 512).
neoverse-n1
generate code that is usable on any Arm Neoverse-N1 processor-based system.
neoverse-v1
generate code that is usable on any Arm Neoverse-V1 processor-based system (SVE x 256).
neoverse-v2
generate code that is usable on any Arm Neoverse-V2 processor-based system (SVE x 128).

2.2.61. -[no]traceback

Adds debug information for runtime traceback for use with the environment variable NVCOMPILER_TERM.

Default

The compiler enables traceback for FORTRAN and disables traceback for C and C++.

Syntax

-traceback

Usage

In this example, nvfortran enables traceback for the program myprog.f.

$ nvfortran -traceback myprog.f

Description

Use this option to enable or disable runtime traceback information for use with the environment variable NVCOMPILER_TERM.

Setting set TRACEBACK=OFF; in siterc or .mynv*rc also disables default traceback.

Using ON instead of OFF enables default traceback.

2.2.62. -U

Undefines a preprocessor macro.

Syntax

-Usymbol

Where symbol is a symbolic name.

Usage

The following examples undefine the macro test.

$ nvfortran -Utest myprog.F
$ nvfortran -Dtest -Utest myprog.F

Description

Use this option to undefine a preprocessor macro. You can also use the #undef pre-processor directive to undefine macros.

2.2.63. -u

Initializes the symbol-table with <symbol>, which is undefined for the linker. An undefined symbol triggers loading of the first member of an archive library.

Default

The compiler does not use the -⁠u option.

Syntax

-usymbol

Where symbol is a symbolic name.

Usage

In this example, nvfortran initializes symbol-table with test.

$ nvfortran -utest myprog.f

Description

Use this option to initialize the symbol-table with <symbol>, which is undefined for the linker. An undefined symbol triggers loading of the first member of an archive library.

2.2.64. -V[release_number]

Displays additional information, including version messages. Further, if a release_number is appended, the compiler driver attempts to compile using the specified release instead of the default release.

Note: There can be no space between -V and release_number.

Default

The compiler does not display version information and uses the release specified by your path to compile.

Usage

The following command-line shows the output using the -⁠V option.

% nvfortran -V myprog.f

The following command-line causes nvc to compile using the 20.7 release instead of the default release.

% nvc -V20.7 myprog.c

Description

Use this option to display additional information, including version messages or, if a release_number is appended, to instruct the compiler driver to attempt to compile using the specified release instead of the default release.

The specified release must be co-installed with the default release

2.2.65. -v

Displays the invocations of the compiler, assembler, and linker.

Default

The compiler does not display individual phase invocations.

Usage

In the following example you use -⁠v to see the commands sent to compiler tools, assembler, and linker.

$ nvfortran -v myprog.f90

Description

Use the -⁠v option to display the invocations of the compiler, assembler, and linker. These invocations are command lines created by the compiler driver from the files and the -⁠W options you specify on the compiler command-line.

2.2.66. -W

Passes arguments to a specific phase.

Syntax

-W{0 | a | l },option[,option...]
Note: You cannot have a space between the -⁠W and the single-letter pass identifier, between the identifier and the comma, or between the comma and the option.
0
(the number zero) specifies the compiler.
a
specifies the assembler.
l
(lowercase letter l) specifies the linker.
option
is a string that is passed to and interpreted by the compiler, assembler or linker. Options separated by commas are passed as separate command line arguments.

Usage

In the following example the linker loads the text segment at address 0xffc00000 and the data segment at address 0xffe00000.

$ nvfortran -Wl,-k,-t,0xffc00000,-d,0xffe00000 myprog.f

Description

Use this option to pass arguments to a specific phase. You can use the -⁠W option to specify options for the assembler, compiler, or linker.

A given NVIDIA HPC compiler command invokes the compiler driver, which parses the command-line, and generates the appropriate commands for the compiler, assembler, and linker.

2.2.67. -w

Do not print warning messages.

Default

The compiler prints warning messages.

Usage

In the following example no warning messages are printed.

$ nvfortran -w myprog.f

Description

Use the -⁠w option to inhibit warning messages.

2.2.68. -Xs

Use legacy standard mode for C and C++.

Default

None.

Usage

In the following example the compiler uses legacy standard mode.

$ nvc -Xs myprog.c

Description

Use this option to use legacy standard mode for C and C++. Further, this option implies -alias=traditional.

2.2.69. -Xt

Use legacy transitional mode for C and C++.

Default

None.

Usage

In the following example the compiler uses legacy transitional mode.

$ nvc -Xt myprog.c

Description

Use this option to use legacy transitional mode for C and C++. Further, this option implies -alias=traditional.

2.2.70. -Xlinker

Pass options to the linker.

Syntax

-Xlinker option[,option...]

Default

None.

Usage

In the following example the option --trace-symbol=foo is passed to the linker, which will cause the Linux linker to list all the files that reference symbol foo.

$ nvc -Xlinker --trace-symbol=foo myprog.c

Description

Use this option pass options to the linker. This is useful when the link step needs to be customized but the compiler doesn't understand the necessary linker options. The options supported by the linker are platform dependent and are not listed here. This option has the same effect as -Wl.

2.3. C++ and C-specific Compiler Options

There are a large number of compiler options specific to the NVC++ and NVC compilers, especially NVC++. This section provides details on several of these options, but is not exhaustive. For a complete list of available options, including an exhaustive list of NVC++ options, use the -⁠help command-line option. For further detail on a given option, use -⁠help and specify the option explicitly as described in -help.

2.3.1. -A

(nvc++ only) Instructs the NVC++ compiler to accept code conforming to the ISO C++ standard, issuing errors for non-conforming code.

Default

By default, the compiler accepts code conforming to the standard C++ Annotated Reference Manual.

Usage

The following command-line requests ISO conforming C++.

	$ nvc++ -A hello.cc

Description

Use this option to instruct the NVC++ compiler to accept code conforming to the ISO C++ standard and to issues errors for non-conforming code.

2.3.2. -a

(nvc++ only) Instructs the NVC++ compiler to accept code conforming to the ISO C++ standard, issuing warnings for non-conforming code.

Default

By default, the compiler accepts code conforming to the standard C++ Annotated Reference Manual.

Usage

The following command-line requests ISO conforming C++, issuing warnings for non-conforming code.

$ nvc++ -a hello.cc

Description

Use this option to instruct the NVC++ compiler to accept code conforming to the ISO C++ standard and to issues warnings for non-conforming code.

2.3.3. -alias

Select optimizations based on type-based pointer alias rules in C and C++.

Syntax

-alias=[ansi|traditional]

Default

None.

Usage

The following command-line enables optimizations.

   $ nvc++ -alias=ansi hello.cc

Description

Use this option to select optimizations based on type-based pointer alias rules in C and C++.

ansi
Enable optimizations using ANSI C type-based pointer disambiguation
traditional
Disable type-based pointer disambiguation

2.3.4. --[no_]alternative_tokens

(nvc++ only) Enables or disables recognition of alternative tokens. These are tokens that make it possible to write C++ without the use of the comma (,) , [, ], #, &, ^, and characters. The alternative tokens include the operator keywords (e.g., and, bitand, etc.) and digraphs.

Default

The default behavior is --no_alternative_tokens, that is, to disable recognition of alternative tokens.

Usage

The following command-line enables alternative token recognition.

	$ nvc++ --alternative_tokens hello.cc

2.3.5. -B

(nvc and nvc++) Enables use of C++ style comments starting with // in C program units.

Default

The NVC C compiler does not allow C++ style comments.

Usage

In the following example the compiler accepts C++ style comments.

	$ nvc -B myprog.cc

Description

Use this option to enable use of C++ style comments starting with // in C program units.

2.3.6. --[no_]bool

(nvc++ only) Enables or disables recognition of bool.

Default

The compile recognizes bool: --bool.

Usage

In the following example, the compiler does not recognize bool.

	$ nvc++ --no_bool myprog.cc

Description

Use this option to enable or disable recognition of bool.

2.3.7. --[no_]builtin

Compile with or without math subroutine builtin support.

Default

The default is to compile with math subroutine support: --builtin.

Usage

In the following example, the compiler does not build with math subroutine support.

   $ nvc++ --no_builtin myprog.cc

Description

Use this option to enable or disable compiling with math subroutine builtin support. When you compile with math subroutine builtin support, the selected math library routines are inlined.

2.3.8. --[no_]compress_names

Compresses long function names in the file.

Default

The compiler does not compress names: --no_compress_names.

Usage

In the following example, the compiler compresses long function names.

	$ nvc++ --compress_names myprog.cc

Description

Use this option to specify to compress long function names. Highly nested template parameters can cause very long function names. These long names can cause problems for older assemblers. Users encountering these problems should compile all C++ code, including library code with --compress_names. Libraries supplied by NVIDIA work with --compress_names.

2.3.9. --create_pch filename

(nvc++ only) If other conditions are satisfied, create a precompiled header file with the specified name.

Note:

If --pch (automatic PCH mode) appears on the command line following this option, its effect is erased.

Default

The compiler does not create a precompiled header file.

Usage

In the following example, the compiler creates a precompiled header file, hdr1.

	$ nvc++ --create_pch hdr1 myprog.cc

Description

If other conditions are satisfied, use this option to create a precompiled header file with the specified name.

2.3.10. --diag_error <number>

(nvc++ only) Overrides the normal severity of the specified diagnostic messages.

Default

The compiler does not override normal diagnostics severity.

Description

Use this option to override the normal severity of the specified diagnostic messages and have them treated as errors. The message(s) may be specified using a mnemonic tag or using a diagnostic number.

2.3.11. --diag_remark <number>

(nvc++ only) Overrides the normal severity of the specified diagnostic messages.

Default

The compiler does not override normal diagnostics severity.

Description

Use this option to override the normal severity of the specified diagnostic messages and have them treated as remarks. The message(s) may be specified using a mnemonic tag or using a diagnostic number.

2.3.12. --diag_suppress <number>

(nvc++ only) Overrides the normal severity of the specified diagnostic messages.

Default

The compiler does not override normal diagnostics severity.

Usage

In the following example, the compiler suppresses the specified diagnostic messages.

	$ nvc++ --diag_suppress error_tag prog.cc

Description

Use this option to override the normal severity of the specified diagnostic messages and have them suppressed. The message(s) may be specified using a mnemonic tag or using a diagnostic number.

2.3.13. --diag_warning <number>

(nvc++ only) Overrides the normal severity of the specified diagnostic messages.

Default

The compiler does not override normal diagnostics severity.

Usage

In the following example, the compiler overrides the severity of the specified diagnostic messages and treats them as warnings.

	$ nvc++ --diag_warning an_error_tag myprog.cc

Description

Use this option to override the normal severity of the specified diagnostic messages and have them treated as warnings. The message(s) may be specified using a mnemonic tag or using a diagnostic number.

2.3.14. --display_error_number

(nvc++ only) Displays the error message number in any diagnostic messages that are generated. The option may be used to determine the error number to be used when overriding the severity of a diagnostic message.

Default

The compiler does not display error message numbers for generated diagnostic messages.

Usage

In the following example, the compiler displays the error message number for any generated diagnostic messages.

	$ nvc++ --display_error_number myprog.cc

Description

Use this option to display the error message number in any diagnostic messages that are generated. You can use this option to determine the error number to be used when overriding the severity of a diagnostic message.

2.3.15. -e<number>

(nvc++ only) Set the C++ front-end error limit to the specified <number>.

2.3.16. --no_exceptions

(nvc++ only) Disables exception handling support.

Default

Exception handling support is enabled.

Usage

In the following example, the compiler does not provide exception handling support.

	$ nvc++ --no_exceptions myprog.cc

Description

Use this option to disable exception handling support. When exception handling is turned off, any try/catch blocks or throw expressions in the code will result in a compilation error, and any exception specifications will be ignored.

2.3.17. -fvisibility=<visibility>

(nvc++ only) Sets the visibility of ELF symbols.

Default

Sets the visibility of ELF symbols.

Usage

All symbols are marked with global visibility unless overridden with this switch or with the visibility attribute.

$ nvc++ -fvisibility=default hello.cp

Description

The visibility argument can take on one of four values: default, internal, hidden, or protected.

2.3.18. --gnu_version <num>

(nvc++ only) Sets the GNU C++ compatibility version.

Default

The compiler uses the latest version installed on the system on which compilation is performed.

Usage

In the following example, the compiler sets the GNU version to 4.3.4.

    $ nvc++ --gnu_version 4.3.4 myprog.cc

Description

Use this option to set the GNU C++ compatibility version to use when you compile.

2.3.19. --[no]llalign

(nvc++ only) Enables or disables alignment of long long integers on long long boundaries.

Default

The compiler aligns long long integers on long long boundaries: --llalign.

Usage

In the following example, the compiler does not align long long integers on long long boundaries.

	$ nvc++ --nollalign myprog.cc

Description

Use this option to allow enable or disable alignment of long long integers on long long boundaries.

2.3.20. -M

Generates a list of make dependencies and prints them to stdout.

Note:

The compilation stops after the preprocessing phase.

Default

The compiler does not generate a list of make dependencies.

Usage

In the following example, the compiler generates a list of make dependencies.

	$ nvc++ -M myprog.cc

Description

Use this option to generate a list of make dependencies and print them to stdout.

2.3.21. -MD[<dfile>]

Generates a list of make dependencies and prints them to a file.

Default

The compiler does not generate a list of make dependencies.

Usage

In the following example, the compiler generates a list of make dependencies and prints them to the file myprog.d.

	$ nvc++ -MD myprog.cc

Description

Use this option to generate a list of make dependencies and print them to a file. The name of the file is determined by the name of the file under compilation, or is as specified using the optional <dfile> argument.

2.3.22. --optk_allow_dollar_in_id_chars

(nvc++ only) Accepts dollar signs ($) in identifiers.

Default

The compiler does not accept dollar signs ($) in identifiers.

Usage

In the following example, the compiler allows dollar signs ($) in identifiers.

	$ nvc++ --optk_allow_dollar_in_id_chars myprog.cc

Description

Use this option to instruct the compiler to accept dollar signs ($) in identifiers.

2.3.23. -P

Halts the compilation process after preprocessing and writes the preprocessed output to a file.

Default

The compiler produces an executable file.

Usage

In the following example, the compiler produces the preprocessed file myprog.i in the current directory.

	$ nvc++ -P myprog.cc

Description

Use this option to halt the compilation process after preprocessing and write the preprocessed output to a file. If the input file is filename.c or filename.cc., then the output file is filename.i.

2.3.24. --pch

(nvc++ only) Automatically use and/or create a precompiled header file.

Note:

If --use_pch or --create_pch (manual PCH mode) appears on the command line following this option, this option has no effect.

Default

The compiler does not automatically use or create a precompiled header file.

Usage

In the following example, the compiler automatically uses a precompiled header file.

	$ nvc++ --pch myprog.cc

Description

Use this option to automatically use and/or create a precompiled header file.

2.3.25. --pch_dir directoryname

(nvc++ only) Specifies the directory in which to search for and/or create a precompiled header file.

The compiler searches your PATH for precompiled header files / use or create a precompiled header file.

Usage

In the following example, the compiler searches in the directory myhdrdir for a precompiled header file.

	$ nvc++ --pch_dir myhdrdir myprog.cc

Description

Use this option to specify the directory in which to search for and/or create a precompiled header file. You may use this option with automatic PCH mode (-⁠-⁠pch) or manual PCH mode (-⁠-⁠create_pch or -⁠-⁠use_pch).

2.3.26. --[no_]pch_messages

(nvc++ only) Enables or disables the display of a message indicating that the current compilation used or created a precompiled header file.

The compiler displays a message when it uses or creates a precompiled header file.

In the following example, no message is displayed when the precompiled header file located in myhdrdir is used in the compilation.

	$ nvc++ --pch_dir myhdrdir --no_pch_messages myprog.cc

Description

Use this option to enable or disable the display of a message indicating that the current compilation used or created a precompiled header file.

2.3.27. --pedantic

Prints warnings from included <system header files>.

Default

The compiler does not print warnings from the included system header files.

Usage

In the following example, the compiler prints the warnings from the included system header files.

$ nvc++ --pedantic myprog.cc

2.3.28. --preinclude=<filename>

(nvc++ only) Specifies the name of a file to be included at the beginning of the compilation.

In the following example, the compiler includes the file incl_file.c at the beginning of the compilation. me

	$ nvc++ --preinclude=incl_file.c myprog.cc

Description

Use this option to specify the name of a file to be included at the beginning of the compilation. For example, you can use this option to set system-dependent macros and types.

2.3.29. --use_pch filename

(nvc++ only) Uses a precompiled header file of the specified name as part of the current compilation.

Note:

If --pch (automatic PCH mode) appears on the command line following this option, its effect is erased.

Default

The compiler does not use a precompiled header file.

In the following example, the compiler uses the precompiled header file, hdr1 as part of the current compilation.

	$ nvc++ --use_pch hdr1 myprog.cc

Use a precompiled header file of the specified name as part of the current compilation. If --pch (automatic PCH mode) appears on the command line following this option, its effect is erased.

2.3.30. --[no_]using_std

(nvc++ only) Enables or disables implicit use of the std namespace when standard header files are included.

Default

The compiler uses std namespace when standard header files are included: --using_std.

Usage

The following command-line disables implicit use of the std namespace:

	$ nvc++ --no_using_std hello.cc

Description

Use this option to enable or disable implicit use of the std namespace when standard header files are included in the compilation.

2.3.31. -Xfilename

(nvc++ only) Generates cross-reference information and places output in the specified file.

Syntax:

-Xfoo

where foo is the specified file for the cross reference information.

Default

The compiler does not generate cross-reference information.

Usage

In the following example, the compiler generates cross-reference information, placing it in the file:xreffile.

	$ nvc++ -Xxreffile myprog.cc

Description

Use this option to generate cross-reference information and place output in the specified file. This is an EDG option.

2.4. -M Options by Category

This section describes each of the options available with -⁠M by the categories:

Code Generation Fortran Language Controls Optimization Environment
C/C++ Language Controls Inlining Miscellaneous  

The following sections provide detailed descriptions of several, but not all, of the -⁠M<nvflag> options. For a complete alphabetical list of all the options, refer to Table 12. These options are grouped according to categories and are listed with exact syntax, defaults, and notes concerning similar or related options.

For the latest information and description of a given option, or to see all available options, use the -⁠help command-line option, described in -help.

2.4.1. Code Generation Controls

This section describes the -⁠M<nvflag> options that control code generation.

Default: For arguments that you do not specify, the default code generation controls are these:

daz norecursive nosecond_underscore
flushz noreentrant nostride0
noref_externals signextend

Related options:-⁠D, -⁠I, -⁠L, -⁠l, -⁠U.

The following list provides the syntax for each -⁠M<nvflag> option that controls code generation. Each option has a description and, if appropriate, any related options.

-Mdaz
Set IEEE denormalized floating point operands to zero; there is a performance benefit but misleading results can occur, such as when dividing a small normalized number by a denormalized number. To take effect, this option must be set when compiling the main program/function. [default - x86_64 and aarch64; does not apply to ppc64le]
Note: Same functionality can be achieved with the runtime environment variable NVCOMPILER_FPU_STATE
-Mnodaz
Do not treat denormalized numbers as zero. To take effect, this option must be set for the main program.
Note: Same functionality can be achieved with the runtime environment variable NVCOMPILER_FPU_STATE
-Mnodwarf
Specifies not to add DWARF debug information.
-Mdwarf2
Generate DWARF2 format debug information. To take effect, this option must be used in combination with -⁠g.
-Mdwarf3
Generate DWARF3 format debug information. To take effect, this option must be used in combination with -⁠g.
-Mflushz
Set floating point control register to flush-to-zero mode; if a floating-point underflow occurs, the result is set to zero. To take effect, this option must be set when compiling the main program/function. [default - x86_64 and aarch64; does not apply to ppc64le]
Note: Same functionality can be achieved with the runtime environment variable NVCOMPILER_FPU_STATE
-Mnoflushz
Do not set flush-to-zero mode; generate underflows. To take effect, this option must be set for the main program.
Note: Same functionality can be achieved with the runtime environment variable NVCOMPILER_FPU_STATE
-Mfunc32
Align functions on 32-byte boundaries.
-Minstrument[=functions]
Generate additional code to enable instrumentation of functions. The option -⁠Minstrument=functions is the same as -⁠Minstrument. Implies -⁠Mframe.
-Mlarge_arrays
Enable support for 64-bit indexing and single static data objects larger than 2 GB in size. This option is the default in the presence of -⁠mcmodel=medium. It can be used separately together with the default small memory model for certain 64-bit applications that manage their own memory space. For more information, refer to the ‘Programming Considerations for 64-Bit Environments’ section of the HPC Compilers User Guide
-Mnolarge_arrays
Disable support for 64-bit indexing and single static data objects larger than 2 GB in size. When this option is placed after -⁠mcmodel=medium on the command line, it disables use of 64-bit indexing for applications that have no single data object larger than 2 GB. For more information, refer to the ‘Programming Considerations for 64-Bit Environments’ section of the HPC Compilers User Guide.
-Mnomain
Instructs the compiler not to include the object file that calls the Fortran main program as part of the link step. This option is useful for linking programs in which the main program is written in C/C++ and one or more subroutines are written in Fortran (Fortran only).
-M[no]pre
enables [disables] partial redundancy elimination.
-Mrecursive
instructs the compiler to allow Fortran subprograms to be called recursively.
-Mnorecursive
Fortran subprograms may not be called recursively.
-Mref_externals
force references to names appearing in EXTERNAL statements (Fortran only).
-Mnoref_externals
do not force references to names appearing in EXTERNAL statements (Fortran only).
-Mreentrant
instructs the compiler to avoid optimizations that can prevent code from being reentrant.
-Mnoreentrant
instructs the compiler not to avoid optimizations that can prevent code from being reentrant.
-Msecond_underscore
instructs the compiler to add a second underscore to the name of a Fortran global symbol if its name already contains an underscore. This option is useful for maintaining compatibility with object code compiled using gfortran, which uses this convention by default (Fortran only).
-Mnosecond_underscore
instructs the compiler not to add a second underscore to the name of a Fortran global symbol if its name already contains an underscore (Fortran only).
-Msafe_lastval
When a scalar is used after a loop, but is not defined on every iteration of the loop, the compiler does not by default parallelize the loop. However, this option tells the compiler it’s safe to parallelize the loop. For a given loop, the last value computed for all scalars makes it safe to parallelize the loop.
-Msignextend
instructs the compiler to extend the sign bit that is set as a result of converting an object of one data type to an object of a larger signed data type.
-Mnosignextend
instructs the compiler not to extend the sign bit that is set as the result of converting an object of one data type to an object of a larger data type.
-Mstack_arrays
places automatic arrays on the stack.
-Mnostack_arrays
allocates automatic arrays on the heap. -Mnostack_arrays is the default and what traditionally has been the approach used.
-Mstride0
instructs the compiler to inhibit certain optimizations and to allow for stride 0 array references. This option may degrade performance and should only be used if zero-stride induction variables are possible.
-Mnostride0
instructs the compiler to perform certain optimizations and to disallow for stride 0 array references.
-Mvarargs
force Fortran program units to assume procedure calls are to C functions with a varargs-type interface (nvfortran only).

2.4.2. C/C++ Language Controls

This section describes the -⁠M<nvflag> options that affect C++ and C language interpretations by the NVC++ and NVC compilers. These options are only valid to the nvc++ and nvc compiler drivers.

Default: For arguments that you do not specify, the defaults are as follows:

noasmkeyword nosingle
dollar,_ schar

Usage:

In this example, the compiler allows the asm keyword in the source file.

	$ nvc -Masmkeyword myprog.c

In the following example, the compiler maps the dollar sign to the dot character.

	$ nvc -Mdollar,. myprog.c

In the following example, the compiler treats floating-point constants as float values, rather than the default double.

	$ nvc -Mfcon myprog.c

In the following example, the compiler does not convert float parameters to double parameters.

	$ nvc -Msingle myprog.c

Without -⁠Muchar or with -⁠Mschar, the variable ch is a signed character:

	char ch;
	signed char sch;

If -⁠Muchar is specified on the command line:

	$ nvc -Muchar myprog.c

char ch in the preceding declaration is equivalent to:

 unsigned char ch;

The following list provides the syntax for each -⁠M<nvflag> option that controls code generation in C++ and C. Each option has a description and, if appropriate, any related options.

-Masmkeyword
instructs the compiler to allow the asm keyword in C source files. The syntax of the asm statement is as follows:
asm("statement");
Where statement is a legal assembly-language statement. The quote marks are required.
Note: The current default is to support gcc's extended asm, where the syntax of extended asm includes asm strings. The -⁠M[no]asmkeyword switch is useful only if the target device is a Pentium 3 or older cpu type (-⁠tp piii|p6|k7|athlon|athlonxp|px).
-Mnoasmkeyword
instructs the compiler not to allow the asm keyword in C source files. If you use this option and your program includes the asm keyword, unresolved references are generated
-Mdollar,char
char specifies the character to which the compiler maps the dollar sign ($). The NVC compiler allows the dollar sign in names; ANSI C does not allow the dollar sign in names.
-M[no]eh_frame
instructs the linker to keep eh_frame call frame sections in the executable.
Note: The eh_frame option is available only on newer Linux systems that supply the system unwind libraries.
-Mfcon
instructs the compiler to treat floating-point constants as float data types, instead of double data types. This option can improve the performance of single-precision code.
-M[no]m128
instructs the compiler to recognize [ignore] __m128, __m128d, and __m128i datatypes.
-Mschar
specifies signed char characters. The compiler treats "plain" char declarations as signed char.
-Msingle
do not to convert float parameters to double parameters in non-prototyped functions. This option can result in faster code if your program uses only float parameters. However, since ANSI C specifies that routines must convert float parameters to double parameters in non-prototyped functions, this option results in non-ANSI conformant code.
-Mnosingle
instructs the compiler to convert float parameters to double parameters in non-prototyped functions.
-Muchar
instructs the compiler to treat "plain" char declarations as unsigned char.

2.4.3. Environment Controls

This section describes the -⁠M<nvflag> options that control environments.

Default: For arguments that you do not specify, the default environment option depends on your configuration.

The following list provides the syntax for each -⁠M<nvflag> option that controls environments. Each option has a description and, if appropriate, a list of any related options.

-Mnostartup
instructs the linker not to link in the standard startup routine that contains the entry point (_start) for the program.
Note: If you use the -⁠Mnostartup option and do not supply an entry point, the linker issues the following error message: Warning: cannot find entry symbol _start
-M[no]hugetlb
links in the huge page runtime library. Enables large 2-megabyte pages to be allocated. The effect is to reduce the number of TLB entries required to execute a program. This option is most effective on newer architectures; older architectures do not have enough TLB entries for this option to be beneficial. By itself, the huge suboption tries to allocate as many huge pages as required. You can also limit the pages allocated by using the environment variable NVCOMPILER_HUGE_PAGES.
-M[no]stddef
instructs the compiler not to predefine any macros to the preprocessor when compiling a C program.
-Mnostdinc
instructs the compiler to not search the standard location for include files.
-Mnostdlib
instructs the linker not to link in the standard libraries libnvf.a, libm.a, libc.a, and libnvc.a in the library directory lib within the standard directory. You can link in your own library with the -⁠l option or specify a library directory with the -⁠L option.

2.4.4. Fortran Language Controls

This section describes the -⁠M<nvflag> options that affect Fortran language interpretations by the NVIDIA Fortran compiler. These options are valid only for the nvfortran compiler driver.

Default: Before looking at all the options, let's look at the defaults. For arguments that you do not specify, the defaults are as follows:

nobackslash nodefaultunit dollar,_ noonetrip nounixlogical
nodclchk nodlines noiomutex nosave noupcase

The following list provides the syntax for each -⁠M<nvflag> option that affect Fortran language interpretations. Each option has a description and, if appropriate, a list of any related options.

-Mallocatable=95|03
controls whether Fortran 95 or Fortran 2003 semantics are used in allocatable array assignments. The default behavior is to use Fortran 95 semantics; the 03 option instructs the compiler to use Fortran 2003 semantics.
-Mbackslash
instructs the compiler to treat the backslash as a normal character, and not as an escape character in quoted strings.
-Mnobackslash
instructs the compiler to recognize a backslash as an escape character in quoted strings (in accordance with standard C usage).
-Mdclchk
instructs the compiler to require that all program variables be declared.
-Mnodclchk
instructs the compiler not to require that all program variables be declared.
-Mdefaultunit
instructs the compiler to treat "*" as a synonym for standard input for reading and standard output for writing.
-Mnodefaultunit
instructs the compiler to treat "*" as a synonym for unit 5 on input and unit 6 on output.
-Mdlines
instructs the compiler to treat lines containing "D" in column 1 as executable statements (ignoring the "D").
-Mnodlines
instructs the compiler not to treat lines containing "D" in column 1 as executable statements. The compiler does not ignore the "D".
-Mdollar,char
char specifies the character to which the compiler maps the dollar sign. The compiler allows the dollar sign in names.
-Mextend
instructs the compiler to accept 132-column source code; otherwise it accepts 72-column code.
-Mfixed
instructs the compiler to assume input source files are in FORTRAN 77-style fixed form format.
-Mfree
instructs the compiler to assume input source files are in Fortran 90/95 freeform format.
-Miomutex
instructs the compiler to generate critical section calls around Fortran I/O statements.
-Mnoiomutex
instructs the compiler not to generate critical section calls around Fortran I/O statements.
-Monetrip
instructs the compiler to force each DO loop to execute at least once. This option is useful for programs written for earlier versions of Fortran.
-Mnoonetrip
instructs the compiler not to force each DO loop to execute at least once.
-Msave
instructs the compiler to assume that all local variables are subject to the SAVE statement. This may allow older Fortran programs to run, but it can greatly reduce performance.
-Mnosave
instructs the compiler not to assume that all local variables are subject to the SAVE statement.
-Mstandard
instructs the compiler to flag non-ANSI-conforming source code.
-Munixlogical
directs the compiler to treat logical values as true if the value is non-zero and false if the value is zero (UNIX F77 convention). When -⁠Munixlogical is enabled, a logical value or test that is non-zero is .TRUE., and a value or test that is zero is .FALSE.. In addition, the value of a logical expression is guaranteed to be one (1) when the result is .TRUE..
-Mnounixlogical
directs the compiler to use the VMS convention for logical values for true and false. Even values are true and odd values are false.
-Mupcase
instructs the compiler to preserve uppercase letters in identifiers. With -⁠Mupcase, the identifiers "X" and "x" are different. Keywords must be in lower case. This selection affects the linking process. If you compile and link the same source code using -⁠Mupcase on one occasion and -⁠Mnoupcase on another, you may get two different executables – depending on whether the source contains uppercase letters. The standard libraries are compiled using the default -⁠Mnoupcase .
-Mnoupcase
instructs the compiler to convert all identifiers to lower case. This selection affects the linking process. If you compile and link the same source code using -⁠Mupcase on one occasion and -⁠Mnoupcase on another, you may get two different executables, depending on whether the source contains uppercase letters. The standard libraries are compiled using -⁠Mnoupcase.

2.4.5. Inlining Controls

This section describes the -⁠M<nvflag> options that control function inlining.

Usage: Before looking at all the options, let’s look at a few examples. In the following example, the compiler extracts functions that have 500 or fewer statements from the source file myprog.f and saves them in the file extract.il.

$ nvfortran -Mextract=500 -o extract.il myprog.f

In the following example, the compiler inlines functions with fewer than approximately 100 statements in the source file myprog.f.

$ nvfortran -Minline=maxsize:100 myprog.f

Related options: -⁠o, -⁠Mextract

The following list provides the syntax for each -⁠M<nvflag> option that controls function inlining. Each option has a description and, if appropriate, a list of any related options.

-M[no]autoinline[=option[,option,...]]
instructs the compiler to inline [not to inline] a C++ and C functions at -⁠O2, where the option can be any of these:
maxsize:n
instructs the compiler not to inline functions of size > n. The default size is 100.
nostatic
do not inline static functions without the inline keyword
totalsize:n
instructs the compiler to stop inlining when the size equals n. The default size is 800.
-Mextract[=option[,option,...]]
Extracts functions from the file indicated on the command line and creates or appends to the specified extract directory where option can be any of the following:
name:func
instructs the extractor to extract function func from the file.
size:number
instructs the extractor to extract functions with number or fewer statements from the file.
lib:filename.ext
instructs the extractor to use directory filename.ext as the extract directory, which is required to save and re-use inline libraries.

If you specify both name and size, the compiler extracts functions that match func, or that have number or fewer statements. For examples of extracting functions, refer to the ‘Using Function Inlining’ section of the HPC Compilers User Guide.

-Minline[=option[,option,...]]
instructs the compiler to pass options to the function inliner, where the option can be any of the following:
except:func
Inlines all eligible functions except func, a function in the source text. You can use a comma-separated list to specify multiple functions.
[name:]func
Inlines all functions in the source text whose name matches func. You can use a comma-separated list to specify multiple functions.

The function name should be a non-numeric string that does not contain a period. You can also use a name: prefix followed by the function name. If name: is specified, what follows is always the name of a function.

[maxsize:]number
A numeric option is assumed to be a size. Functions of size number or less are inlined. If both number and function are specified, then functions matching the given name(s) or meeting the size requirements are inlined.

The size number need not exactly equal the number of statements in a selected function; the size parameter is merely a rough gauge.

[no]reshape
instructs the inliner to allow [disallow] inlining in Fortran even when array shapes do not match. The default is -⁠Minline=noreshape, except with -⁠Mconcur or -⁠mp, where the default is -⁠Minline=reshape.
smallsize:number
Always inline functions of size smaller than number regardless of other size limits.
totalsize:number
Stop inlining in a function when the function's total inlined size reaches the number specified.
[lib:]filename.ext
instructs the inliner to inline the functions within the library file filename.ext. The compiler assumes that a filename.ext option containing a period is a library file.
Tip: Create the library file using the -⁠Mextract option. You can also use a lib: prefix followed by the library name.
  • If lib: is specified, no period is necessary in the library name. Functions from the specified library are inlined.
  • If no library is specified, functions are extracted from a temporary library created during an extract prepass.

If you specify both func and number, the compiler inlines functions that match the function name or have number or fewer statements.

Inlining can be disabled with -⁠Mnoinline.

For examples of inlining functions, refer to the ‘Using Function Inlining’ section of the HPC Compilers User Guide.

2.4.6. Optimization Controls

This section describes the -⁠M<nvflag> options that control optimization.

Default: Before looking at all the options, let's look at the defaults. For arguments that you do not specify, the default optimization control options are as follows:

depchk noipa nounroll nor8
i4 nolre novect nor8intrinsics
nofprelaxed noprefetch    

Usage: In this example, the compiler invokes the vectorizer with use of packed SIMD instructions enabled.

$ nvfortran -Mvect=simd -Mcache_align myprog.f
Note: If you do not supply any sub-options to -⁠Mvect, the compiler uses defaults that are dependent upon the target system. Not all sub-options are valid on all target systems.

Related options:-⁠g, -⁠O

The following list provides the syntax for each -⁠M<nvflag> option that controls optimization. Each option has a description and, if appropriate, a list of any related options.

-Mcache_align
Align unconstrained objects of length greater than or equal to 16 bytes on cache-line boundaries. An unconstrained object is a data object that is not a member of an aggregate structure or common block. This option does not affect the alignment of allocatable or automatic arrays. To effect cache-line alignment of stack-based local variables, the main program or function must be compiled with -⁠Mcache_align.
-Mconcur[=option [,option,...]]
Instructs the compiler to enable auto-parallelization of loops for multicore CPUs. If -⁠Mconcur is specified, multiple CPU cores will be used to execute loops that the compiler determines to be parallelizable. option is one of the following:
allcores
Instructs the compiler to use all available cores. Use this option at link time.
[no]altcode:n
Instructs the parallelizer to generate alternate serial code for parallelized loops.
  • If altcode is specified without arguments, the parallelizer determines an appropriate cutoff length and generates serial code to be executed whenever the loop count is less than or equal to that length.
  • If altcode:n is specified, the serial altcode is executed whenever the loop count is less than or equal to n.
  • If noaltcode is specified, the parallelized version of the loop is always executed regardless of the loop count.
cncall
Indicates that calls in parallel loops are safe to parallelize. Also, no minimum loop count threshold must be satisfied before parallelization will occur, and last values of scalars are assumed to be safe.
[no]innermost
Instructs the parallelizer to enable parallelization of innermost loops. The default is to not parallelize innermost loops, since it is usually not profitable on dual-core processors.
levels:n
Parallelize loops nested at most n levels deep.
noassoc
Instructs the parallelizer to disable parallelization of loops with reductions.
When linking, the -⁠Mconcur switch must be specified or unresolved references result.
Note: This option applies only on shared-memory multi-processor (SMP) or multicore CPU-based systems.
-Mcray[=option[,option,...]]
(Fortran only) Force Cray Fortran compatibility with respect to the listed options. Possible values of option include:
pointer
for purposes of optimization, it is assumed that pointer-based variables do not overlay the storage of any other variable.
-Mdepchk
instructs the compiler to assume unresolved data dependencies actually conflict.
-Mnodepchk
Instructs the compiler to assume potential data dependencies do not conflict. However, if data dependencies exist, this option can produce incorrect code.
-Mdse
Enables a dead store elimination phase that is useful for programs that rely on extensive use of inline function calls for performance. This is disabled by default.
-Mnodse
Disables the dead store elimination phase. This is the default.
-M[no]fpapprox[=option]
Perform certain floating point operations using low-precision approximation. -⁠Mnofpapprox specifies not to use low-precision fp approximation operations. By default -⁠Mfpapprox is not used. If -⁠Mfpapprox is used without suboptions, it defaults to use approximate div, sqrt, and rsqrt. The available suboptions are these:
div
Approximate floating point division
sqrt
Approximate floating point square root
rsqrt
Approximate floating point reciprocal square root
-M[no]fpmisalign
Instructs the compiler to allow (not allow) vector arithmetic instructions with memory operands that are not aligned on 16-byte boundaries. The default is -⁠Mnofpmisalign on all processors.
-M[no]fprelaxed[=option]
Instructs the compiler to use [not use] relaxed precision in the calculation of some intrinsic functions. Can result in improved performance at the expense of numerical accuracy. The possible values for option are:
div
Perform divide using relaxed precision.
intrinsic
Enables use of relaxed precision intrinsics.
noorder
Do not allow expression reordering or factoring.
order
Allow expression reordering, including factoring.
recip
Perform reciprocal using relaxed precision.
rsqrt
Perform reciprocal square root (1/sqrt) using relaxed precision.
sqrt
Perform square root with relaxed precision.
With no options, -⁠Mfprelaxed generates relaxed precision code for those operations that generate a significant performance improvement, depending on the target processor. The default is -⁠Mnofprelaxed which instructs the compiler to not use relaxed precision in the calculation of intrinsic functions.
-Mi4
(Fortran only) instructs the compiler to treat INTEGER variables as INTEGER*4.
-Mlre[=array | assoc | noassoc]
Enables loop-carried redundancy elimination, an optimization that can reduce the number of arithmetic operations and memory references in loops. The available suboptions are:
assoc
allow expression re-association. Specifying this suboption can increase opportunities for loop-carried redundancy elimination but may alter numerical results.
noassoc
disallow expression re-association.
-Mnolre
Disable loop-carried redundancy elimination.
-Mnoframe
Eliminate operations that set up a true stack frame pointer for every function. With this option enabled, you cannot perform a traceback on the generated code and you cannot access local variables.
-Mnoi4
(Fortran only) instructs the compiler to treat INTEGER variables as INTEGER*2.
-Mpre
Enables partial redundancy elimination.
-Mprefetch[=option [,option...]]
enables generation of prefetch instructions on processors where they are supported. Possible values for option include:
d:m
set the fetch-ahead distance for prefetch instructions to m cache lines.
n:p
set the maximum number of prefetch instructions to generate for a given loop to p.
nta
use the prefetch instruction.
plain
use the prefetch instruction (default).
t0
use the prefetcht0 instruction.
w
use the AMD-specific prefetchw instruction.
-Mnoprefetch
Disables generation of prefetch instructions.
-M[no]propcond
Enables or disables constant propagation from assertions derived from equality conditionals. The default is enabled.
-Mr8
(Fortran only) The compiler promotes REAL variables and constants to DOUBLE PRECISION variables and constants, respectively. DOUBLE PRECISION elements are 8 bytes in length.
-Mnor8
(Fortran only) The compiler does not promote REAL variables and constants to DOUBLE PRECISION. REAL variables will be single precision (4 bytes in length).
-Mr8intrinsics
(Fortran only) The compiler treats the intrinsics CMPLX and REAL as DCMPLX and DBLE, respectively.
-Mnor8intrinsics
(Fortran only) The compiler does not promote the intrinsics CMPLX and REAL to DCMPLX and DBLE, respectively.
-Msafeptr[=option[,option,...]]
(C++ and C only) instructs the C++ or C compiler to override data dependencies between pointers of a given storage class. Possible values of option include:
all
assume all pointers and arrays are independent and safe for aggressive optimizations, and in particular that no pointers or arrays overlap or conflict with each other.
arg
instructs the compiler to treat arrays and pointers with the same copyin and copyout semantics as Fortran dummy arguments.
global
instructs the compiler that global or external pointers and arrays do not overlap or conflict with each other and are independent.
local/auto
instructs the compiler that local pointers and arrays do not overlap or conflict with each other and are independent.
static
instructs the compiler that static pointers and arrays do not overlap or conflict with each other and are independent.
-M[no]target_temps
instructs the compiler to enable [disable] using temporaries when passing an array for a callee assumed-shape variable with the target attribute.
-Munroll[=option [,option...]]
invokes the loop unroller to execute multiple instances of the loop during each iteration. This also sets the optimization level to 2 if the level is set to less than 2, or if no -⁠O or -⁠g options are supplied. The option is one of the following:
c:m
instructs the compiler to completely unroll loops with a constant loop count less than or equal to m, a supplied constant. If this value is not supplied, the m count is set to 1.
m:<n>
instructs the compiler to unroll multi-block loops n times. This option is useful for loops that have conditional statements. If n is not supplied, then the default value is 1. The default setting is not to enable -⁠Munroll=m.
n:<n>
instructs the compiler to unroll single-block loops n times, a loop that is not completely unrolled, or has a non-constant loop count. If n is not supplied, the unroller computes the number of times a candidate loop is unrolled.
-Mnounroll
instructs the compiler not to unroll loops.
-M[no]vect[=option [,option,...]]
enable [disable] the code vectorizer, where option is one of the following:
[no]altcode
Enable [disable] generation of alternate code (altcode) for vectorized loops when appropriate. For each vectorized loop the compiler decides whether to generate altcode and what type or types to generate, which may be any or all of: altcode without iteration peeling, altcode with non-temporal stores and other data cache optimizations, and altcode based on array alignments calculated dynamically at runtime. The compiler also determines suitable loop count and array alignment conditionals for executing the altcode. This option is enabled by default.
[no]assoc
Enable [disable] certain associativity conversions that can change the results of a computation due to roundoff error. A typical optimization is to change an arithmetic operation to an arithmetic operation that is mathematically correct, but can be computationally different, due to round-off error.
cachesize:n
Instructs the vectorizer, when performing cache tiling optimizations, to assume a cache size of n. The default is set per processor type, either using the -⁠tp switch or auto-detected from the host computer.
[no]fuse
Enable [disable] automatic loop fusion by the vectorizer.
[no]gather
Enable [disable] vectorization of loops containing indirect array references, such as this one:
sum = 0.d0
do k=d(j),d(j+1)-1
     sum = sum + a(k)*b(c(k))
enddo
The default is gather.
[no]idiom
Enable [disable] idiom recognition by the vectorizer.
levels:n
Maximum nest level of loops to optimize
nocond
Disable vectorization of loops with conditionals.
[no]partial
Enable [disable] partial loop vectorization through innermost loop distribution.
prefetch
Instructs the vectorizer to search for vectorizable loops and, wherever possible, make use of prefetch instructions.
[no]short
Enable [disable] short vector operations. -Mvect=short enables generation of packed SIMD instructions for short vector operations that arise from scalar code outside of loops or within the body of a loop iteration.
[no]simd[:{128|256|512}]
Enable [disable] vectorization using SIMD instructions and data, either 128 bits, 256 bits or 512 bits wide, on processors where there is a choice.
[no]simdresidual
Enable [disable] vectorization using SIMD instructions of the residual iterations of a vectorized loop.
[no]sizelimit:n
Instructs the vectorizer to generate vector code for all loops where possible regardless of the number of statements in the loop. This overrides a heuristic in the vectorizer that ordinarily prevents vectorization of loops with a number of statements that exceeds a certain threshold. The default is nosizelimit.
[no]uniform
Instructs the vectorizer to perform the same optimizations in the vectorized and residual loops.
Note: This option may affect the performance of the residual loop.
-Mnovintr
instructs the compiler not to perform idiom recognition or introduce calls to hand-optimized vector functions.

2.4.7. Miscellaneous Controls

This section describes the -⁠M<nvflag> options that do not easily fit into one of the other categories of -⁠M<nvflag> options.

Default: Before looking at all the options, let’s look at the defaults. For arguments that you do not specify, the default miscellaneous options are as follows:

inform nobounds nolist warn

Related options: -⁠m, -⁠S, -⁠V, -⁠v

Usage: In the following example, the compiler includes Fortran source code with the assembly code.

 $ nvfortran -Manno -S myprog.f

In the following example, the assembler does not delete the assembly file myprog.s after the assembly pass.

 $ nvfortran -Mkeepasm myprog.f

In the following example, the compiler displays information about inlined functions with fewer than approximately 20 source lines in the source file myprog.f.

 $ nvfortran -Minfo=inline -Minline=20 myprog.f

In the following example, the compiler creates the listing file myprog.lst.

 $ nvfortran -Mlist myprog.f

In the following example, array bounds checking is enabled.

 $ nvfortran -Mbounds myprog.f

The following list provides the syntax for each miscellaneous -⁠M<nvflag> option. Each option has a description and, if appropriate, a list of any related options.

-Manno
annotate the generated assembly code with source code. Implies -⁠Mkeepasm.
-M[no]bounds
Enables [disables] array bounds checking.
  • If an array is an assumed size array, the bounds checking only applies to the lower bound.
  • If an array bounds violation occurs during execution, an error message describing the error is printed and the program terminates. The text of the error message includes the name of the array, the location where the error occurred (the source file and the line number in the source), and information about the out of bounds subscript (its value, its lower and upper bounds, and its dimension).
The following is a sample error message:
NVFTN-F-Subscript out of range for array a (a.f: 2)
subscript=3, lower bound=1, upper bound=2, dimension=2
-Mbyteswapio
swap byte-order from big-endian to little-endian or vice versa upon input/output of Fortran unformatted data files.
-Mchkptr
instructs the compiler to check for pointers that are dereferenced while initialized to NULL (Fortran only).
-Mchkstk
instructs the compiler to check the stack for available space in the prologue of a function and before the start of a parallel region. Prints a warning message and aborts the program gracefully if stack space is insufficient. This option is useful when many local and private variables are declared in an OpenMP program.
-Mcpp[=option [,option,...]]
run the cpp-like preprocessor without execution of any subsequent compilation steps. This option is useful for generating dependence information to be included in makefiles.
Note: Only one of the m, md, mm or mmd options can be present; if multiple of these options are listed, the last one listed is accepted and the others are ignored.
The option is one or more of the following:
m
print makefile dependencies to stdout.
md
print makefile dependencies to filename.d, where filename is the root name of the input file being processed, ignoring system include files.
mm
print makefile dependencies to stdout, ignoring system include files.
mmd
print makefile dependencies to filename.d, where filename is the root name of the input file being processed, ignoring system include files.
[no]comment
do [do not] retain comments in output.
[suffix:]<suff>
use <suff> as the suffix of the output file containing makefile dependencies.
-Mgccbug[s]
instructs the compiler to match the behavior of certain gcc bugs.
-Miface[=option]
adjusts the calling conventions for Fortran, where option is one of the following:
cref
uses CREF calling conventions, no trailing underscores.
mixed_str_len_arg
places the lengths of character arguments immediately after their corresponding argument. Has affect only with the CREF calling convention.
nomixed_str_len_arg
places the lengths of character arguments at the end of the argument list. Has affect only with the CREF calling convention.
-Minfo[=option [,option,...]]
instructs the compiler to produce information on standard error, where option is one of the following:
all
instructs the compiler to produce all available -⁠Minfo information. Implies a number of suboptions:
-Minfo=accel,inline,ipa,loop,lre,mp,opt,par,vect,stdpar 
accel
instructs the compiler to enable accelerator information.
ftn
instructs the compiler to enable Fortran-specific information.
inline
instructs the compiler to display information about extracted or inlined functions. This option is not useful without either the -⁠Mextract or -⁠Minline option.
intensity
instructs the compiler to provide informational messages about the intensity of the loop. Specify <n> to get messages on nested loops.
  • For floating point loops, intensity is defined as the number of floating point operations divided by the number of floating point loads and stores.
  • For integer loops, the loop intensity is defined as the total number of integer arithmetic operations, which may include updates of loop counts and addresses, divided by the total number of integer loads and stores.
  • By default, the messages just apply to innermost loops.
loop
instructs the compiler to display information about loops, such as information on vectorization.
lre
instructs the compiler to enable LRE, loop-carried redundancy elimination, information.
mp
instructs the compiler to display information about parallelization.
opt
instructs the compiler to display information about optimization.
par
instructs the compiler to enable parallelizer information.
stdpar
instructs the compiler to emit information about parallelization of C⁠+⁠+ parallel algorithms and Fortran DO CONCURRENT loops.
time
instructs the compiler to display compilation statistics.
unroll
instructs the compiler to display information about loop unrolling.
vect
instructs the compiler to enable vectorizer information.
-Minform=level
instructs the compiler to display error messages at the specified and higher levels, where level is one of the following:
fatal
instructs the compiler to display fatal error messages.
[no]file
instructs the compiler to print or not print source file names as they are compiled. The default is to print the names: -⁠Minform=file.
inform
instructs the compiler to display all error messages (inform, warn, severe and fatal).
severe
instructs the compiler to display severe and fatal error messages.
warn
instructs the compiler to display warning, severe and fatal error messages.
-Minstrumentation=option
Specifies the level of instrumentation calls generated, implies -Mframe. option is one of the following:
level
specifies the level of instrumentation calls generated.
function (default)
generates instrumentation calls for entry and exit to functions. Just after function entry and just before function exit, the following profiling functions are called with the address of the current function and its call site. (x86-64 only).
void __cyg_profile_func_enter (void *this_fn, void *call_site);
void __cyg_profile_func_exit (void *this_fn, void *call_site);
In these calls, the first argument is the address of the start of the current function.
-Mkeepasm
instructs the compiler to keep the assembly file as compilation continues. Normally, the assembler deletes this file when it is finished. The assembly file has the same filename as the source file, but with a .s extension.
-Mlist
instructs the compiler to create a listing file. The listing file is filename.lst, where the name of the source file is filename.f.
-Mnames={lowercase|uppercase}
specifies the case for the names of Fortran externals.
  • lowercase - Use lowercase for Fortran externals.
  • uppercase - Use uppercase for Fortran externals.
-Mneginfo[=option[,option,...]]
instructs the compiler to produce information on standard error, where option is one of the following:
all
instructs the compiler to produce all available information on why various optimizations are not performed.
accel
instructs the compiler to enable accelerator information.
concur
instructs the compiler to produce all available information on why loops are not automatically parallelized. In particular, if a loop is not parallelized due to potential data dependence, the variable(s) that cause the potential dependence are listed in the messages that you see when using the option -⁠Mneginfo.
ftn
instructs the compiler to enable Fortran-specific information.
inline
instructs the compiler to display information about extracted or inlined functions. This option is not useful without either the -⁠Mextract or -⁠Minline option.
loop
instructs the compiler to display information about loops, such as information on vectorization.
lre
instructs the compiler to enable LRE, loop-carried redundancy elimination, information.
mp
instructs the compiler to display information about parallelization.
opt
instructs the compiler to display information about optimization.
par
instructs the compiler to enable parallelizer information.
vect
instructs the compiler to enable vectorizer information.
-Mnolist
the compiler does not create a listing file. This is the default.
-Mnorpath
Do not add -⁠rpath to the link line.
-Mpreprocess
instruct the compiler to perform cpp-like preprocessing on assembly and Fortran input source files.
-Mwritable_strings
stores string constants in the writable data segment.
Note: Options -⁠Xs and -⁠Xst include -⁠Mwritable_strings.

3. C++ Name Mangling

Name mangling transforms the names of entities so that the names include information on aspects of the entity’s type and fully qualified name. This ability is necessary since the intermediate language into which a program is translated contains fewer and simpler name spaces than there are in the C++ language; specifically:

  • Overloaded function names are not allowed in the intermediate language.
  • Classes have their own scopes in C++, but not in the generated intermediate language. For example, an entity x from inside a class must not conflict with an entity x from the file scope.
  • External names in the object code form a completely flat name space. The names of entities with external linkage must be projected onto that name space so that they do not conflict with one another. A function f from a class A, for example, must not have the same external name as a function f from class B.
  • Some names are not names in the conventional sense of the word, they're not strings of alphanumeric characters, for example: operator=.

There are two main problems here:

  1. Generating external names that will not clash.
  2. Generating alphanumeric names for entities with strange names in C++.

Name mangling solves these problems by generating external names that will not clash, and alphanumeric names for entities with strange names in C++. It also solves the problem of generating hidden names for some behind-the-scenes language support in such a way that they match up across separate compilations.

You see mangled names if you view files that are translated by NVC++ or NVC, and you do not use tools that demangle the C++ names. Intermediate files that use mangled names include the assembly and object files created by the NVC++ command. To view demangled names, use the tool nvdecode, which takes input from stdin. nvdecode demangles NVC++ names.

prompt> nvdecode
_ZN1A1gEf
A::g(float)

The name mangling algorithm for the NVC++ compiler is IA-64 ABI compliant and is described at http://mentorembedded.github.io/cxx-abi. Refer to this document for a complete description of the name mangling algorithm.

Pre-defined Compiler Macros

The HPC compilers will pre-define certain compiler macros. The macros are defined as follows:

#define __NVCOMPILER_MAJOR__ 24
#define __NVCOMPILER_MINOR__ 1
#define __NVCOMPILER_PATCHLEVEL__ 0
#define __NVCOMPILER_CLANG_SSE_INTRINSICS_VERSION__ 60000
#define __NVCOMPILER 1
#define __NVCOMPILER_LLVM__ 1
    

5. Runtime Environment

This section describes details related to compiler code generation, including register conventions and calling conventions for x86-64 and OpenPOWER processor-based systems. It addresses these conventions for processors running Linux operating systems.

Note: In this section we sometimes refer to word, halfword, and double word. The equivalent byte information is word (4 byte), halfword (2 byte), and double word (8 byte).

5.1. Linux Programming Model

This section defines compiler and assembly language conventions for the use of certain aspects of an x86-64 or OpenPOWER processor running a Linux operating system. These standards must be followed to guarantee that compilers, application programs, and operating systems written by different people and organizations will work together. The conventions supported by the NVC ISO/ANSI C compiler implement the application binary interface (ABI) as defined in the System V Application Binary Interface: AMD64 Architecture Processor Supplement and the OpenPOWER for Linux Supplement, Power Architecture 64-Bit ELF V2 ABI Specification listed in the Preface.

5.1.1. x86-64 Function Calling Sequence

This section describes the standard function calling sequence, including the stack frame, register usage, and parameter passing.

x86-64 Register Usage Conventions

The following table defines the standard for register allocation. The x86-64 Architecture provides a variety of registers. All the general purpose registers, x87 registers, XMM registers, SSE registers and AVX registers are visible to all procedures in a running program.

Table 14. x86-64 Register Allocation
Type Name Purpose
General %rax 1st return register. When callee has a variable number of arguments, %al specifies the number of vector registers passed.
  %rbx callee-saved; optional base pointer
  %rcx pass 4th argument to functions
  %rdx pass 3rd argument to functions; 2nd return register
  %rsp stack pointer
  %rbp callee-saved; optional stack frame pointer
  %rsi pass 2nd argument to functions
  %rdi pass 1st argument to functions
  %r8 pass 5th argument to functions
  %r9 pass 6th argument to functions
  %r10 temporary register; pass a function's static chain pointer
  %r11 temporary register
  %r12-r15 callee-saved registers
XMM %xmm0-%xmm1 pass and return floating point arguments
  %xmm2-%xmm7 pass floating point arguments
  %xmm8-%xmm15 temporary registers
x87 %st(0) temporary register; return long double arguments
  %st(1) temporary register; return long double arguments
  %st(2) - %st(7) temporary registers

x86-64 Stack Frame Organization

In addition to the registers, each function has a frame on the run-time stack. This stack grows downward from high addresses. Table 15 shows the stack frame organization.

Table 15. Standard Stack Frame
Position Contents Frame
8n+16 (%rbp) argument eightbyte n previous
  . . .  
16 (%rbp) argument eightbyte 0  
8 (%rbp) return address current
0 (%rbp) caller's %rbp current
-8 (%rbp) unspecified  
  . . .  
0 (%rsp) variable size  
-128 (%rsp) red zone  

x86-64 Stack Usage Conventions

Key points concerning the stack frame:

  • The end of the input argument area is aligned on a 16-byte boundary.
  • The 128-byte area beyond the location of %rsp is called the red zone and can be used for temporary local data storage. This area is not modified by signal or interrupt handlers.
  • A call instruction pushes the address of the next instruction (the return address) onto the stack. The return instruction pops the address off the stack and effectively continues execution at the next instruction after the call instruction. A function must preserve non-volatile registers, a register whose contents must be preserved across subroutine calls. Additionally, the called function must remove the return address from the stack, leaving the stack pointer (%rsp) with the value it had before the call instruction was executed.

All registers on an x86-64 system are visible to both a calling and a called function. Registers %rbx, %rsp, %rbp, %r12, %r13, %r14, and %r15 are non-volatile across function calls. Therefore, a function must preserve these registers' values for its caller. Remaining registers are volatile (scratch) registers, that is a register whose contents need not be preserved across subroutine calls. If a calling function wants to preserve such a register value across a function call, it must save its value explicitly.

Registers are used extensively in the standard calling sequence. The first six integer and pointer arguments are passed in these registers (listed in order): %rdi, %rsi, %rdx, %rcx, %r8, %r9. The first eight floating point arguments are passed in the first eight XMM registers: %xmm0, %xmm1, ..., %xmm7. The registers %rax and %rdx are used to return integer and pointer values. The registers %xmm0 and %xmm1 are used to return floating point values.

Additional registers with assigned roles in the standard calling sequence:

%rsp
The stack pointer holds the limit of the current stack frame, which is the address of the stack's bottom-most, valid word. The stack must be 16-byte aligned.
%rbp
The frame pointer holds a base address for the current stack frame. Consequently, a function has registers pointing to both ends of its frame. Incoming arguments reside in the previous frame, referenced as positive offsets from %rbp, while local variables reside in the current frame, referenced as negative offsets from %rbp. A function must preserve this register value for its caller.
RFLAGS
The flags register contains the system flags, such as the direction flag and the carry flag. The direction flag must be set to the "forward" (i.e., zero) direction before entry and upon exit from a function. Other user flags have no specified role in the standard calling sequence and are not preserved.
Floating Point Control Word
The control word contains the floating-point flags, such as the rounding mode and exception masking. This register is initialized at process initialization time and its value must be preserved.

Signals can interrupt processes. Functions called during signal handling have no unusual restriction on their use of registers. Moreover, if a signal handling function returns, the process resumes its original execution path with registers restored to their original values. Thus, programs and compilers may freely use all registers without danger of signal handlers changing their values.

x86-64 Functions Returning Scalars or No Value

  • A function that returns an integral or pointer value places its result in the next available register of the sequence %rax, %rdx.
  • A function that returns a floating point value that fits in the XMM registers returns this value in the next available XMM register of the sequence %xmm0, %xmm1.
  • An X87 floating-point return value appears on the top of the floating point stack in %st(0) as an 80-bit X87 number. If this X87 return value is a complex number, the real part of the value is returned in %st(0) and the imaginary part in %st(1).
  • A function that returns a value in memory also returns the address of this memory in %rax.
  • Functions that return no value (also called procedures or void functions) put no particular value in any register.

x86-64 Functions Returning Structures or Unions

A function can use either registers or memory to return a structure or union. The size and type of the structure or union determine how it is returned. If a structure or union is larger than 16 bytes, it is returned in memory allocated by the caller.

To determine whether a 16-byte or smaller structure or union can be returned in one or more return registers, examine the first eight bytes of the structure or union. The type or types of the structure or union’s fields making up these eight bytes determine how these eight bytes will be returned. If the eight bytes contain at least one integral type, the eight bytes will be returned in %rax even if non-integral types are also present in the eight bytes. If the eight bytes only contain floating point types, these eight bytes will be returned in %xmm0.

If the structure or union is larger than eight bytes but smaller than 17 bytes, examine the type or types of the fields making up the second eight bytes of the structure or union. If these eight bytes contain at least one integral type, these eight bytes will be returned in %rdx even if non-integral types are also present in the eight bytes. If the eight bytes only contain floating point types, these eight bytes will be returned in %xmm1.

If a structure or union is returned in memory, the caller provides the space for the return value and passes its address to the function as a "hidden" first argument in %rdi. This address will also be returned in %rax.

x86-64 Integral and Pointer Arguments

Integral and pointer arguments are passed to a function using the next available register of the sequence %rdi, %rsi, %rdx, %rcx, %r8, %r9. After this list of registers has been exhausted, all remaining integral and pointer arguments are passed to the function via the stack.

x86-64 Floating-Point Arguments

Float and double arguments are passed to a function using the next available XMM register taken in the order from %xmm0 to %xmm7. After this list of registers has been exhausted, all remaining float and double arguments are passed to the function via the stack.

x86-64 Structure and Union Arguments

Structure and union arguments can be passed to a function in either registers or on the stack. The size and type of the structure or union determine how it is passed. If a structure or union is larger than 16 bytes, it is passed to the function in memory.

To determine whether a 16-byte or smaller structure or union can be passed to a function in one or two registers, examine the first eight bytes of the structure or union. The type or types of the structure or union’s fields making up these eight bytes determine how these eight bytes will be passed. If the eight bytes contain at least one integral type, the eight bytes will be passed in the first available general purpose register of the sequence %rdi, %rsi, %rdx, %rcx, %r8, %r9 even if non-integral types are also present in the eight bytes. If the eight bytes only contain floating point types, these eight bytes will be passed in the first available XMM register of the sequence from %xmm0 to %xmm7.

If the structure or union is larger than eight bytes but smaller than 17 bytes, examine the type or types of the fields making up the second eight bytes of the structure or union. If the eight bytes contain at least one integral type, the eight bytes will be passed in the next available general purpose register of the sequence %rdi, %rsi, %rdx, %rcx, %r8, %r9 even if non-integral types are also present in the eight bytes. If these eight bytes only contain floating point types, these eight bytes will be passed in the next available XMM register of the sequence from %xmm0 to %xmm7.

If the first or second eight bytes of the structure or union cannot be passed in a register for some reason, the entire structure or union must be passed in memory.

x86-64 Passing Arguments on the Stack

If there are arguments left after every argument register has been allocated, the remaining arguments are passed to the function on the stack. The unassigned arguments are pushed on the stack in reverse order, with the last argument pushed first.

x86-64 Parameter Passing

Table 16 shows the register allocation and stack frame offsets for the function declaration and call shown in the following example. Both table and example are adapted from System V Application Binary Interface: AMD64 Architecture Processor Supplement.

typedef struct {
    int a, b;
    double d; 
    } 
    structparam; 
    structparam s;
    int e, f, g, h, i, j, k; 
    float flt;  
    double m, n; 
    extern void func(int e, int f, structparam s, int g, int h,  
    float flt, double m, double n, int i, int j, int k);
    void func2() 
    {  
    func(e, f, s, g, h, flt, m, n, i, j, k); 
    }
Table 16. Register Allocation for Example A-2
General Purpose Registers Floating Point Registers Stack Frame Offset
%rdi: e %xmm0: s.d 0: j
%rsi: f %xmm1: flt 8: k
%rdx: s.a,s.b %xmm2: m  
%rcx: g %xmm3: n  
%r8: h    
%r9: i    

x86-64 Implementing a Stack

In general, compilers and programmers must maintain a software stack. The stack pointer, register %rsp, is set by the operating system for the application when the program is started. The stack must grow downwards from high addresses.

A separate frame pointer enables calls to routines that change the stack pointer to allocate space on the stack at run-time (e.g. alloca). Some languages can also return values from a routine allocated on stack space below the original top-of-stack pointer. Such a routine prevents the calling function from using %rsp-relative addressing for values on the stack. If the compiler does not call routines that leave %rsp in an altered state when they return, a frame pointer is not needed and may not be used if the compiler option -⁠Mnoframe is specified.

The stack must be kept aligned on 16-byte boundaries.

x86-64 Variable Length Parameter Lists

Parameter passing in registers can handle a variable number of parameters. The C language uses a special method to access variable-count parameters. The stdarg.h and varargs.h files define several functions to access these parameters. A C routine with variable parameters must use the va_start macro to set up a data structure before the parameters can be used. The va_arg macro must be used to access the successive parameters.

For calls that use varargs or stdargs, the register %rax acts as a hidden argument whose value is the number of XMM registers used in the call.

x86-64 C Parameter Conversion

In C, for a called prototyped function, the parameter type in the called function must match the argument type in the calling function. If the called function is not prototyped, the calling convention uses the types of the arguments but promotes char or short to int, and unsigned char or unsigned short to unsigned int and promotes float to double, unless you use the -⁠Msingle option. For more information on the -⁠Msingle option, refer to -M Options by Category .

x86-64 Calling Assembly Language Programs

The following example shows a C program calling an assembly-language routine sum_3.

C Program Calling an Assembly-language Routine

/* File: testmain.c */
#include <stdio.h>
int
main() {
 long l_para1 = 2;
 float f_para2 = 1.0;
 double d_para3 = 0.5;
 float f_return;
 extern float sum_3(long para1, float para2, double para3);
 f_return = sum_3(l_para1, f_para2, d_para3);
 printf("Parameter one, type long = %ld\n", l_para1);
 printf("Parameter two, type float = %f\n", f_para2);
 printf("Parameter three, type double = %f\n", d_para3);
 printf("The sum after conversion = %f\n", f_return);
 return 0;
}
# File: sum_3.s
# Computes ( para1 + para2 ) + para3
	.text
	.align	16
	.globl	sum_3
sum_3:
	pushq	%rbp
	movq	%rsp, %rbp
	cvtsi2ssq %rdi, %xmm2
	addss	%xmm0, %xmm2
	cvtss2sd %xmm2,%xmm2
	addsd %xmm1, %xmm2
	cvtsd2ss %xmm2, %xmm2
	movaps	%xmm2, %xmm0
	popq	%rbp
	ret
	.type	sum_3, @function
	.size	sum_3,.-sum_3

5.1.2. OpenPOWER Function Calling Sequence

OpenPOWER Register Usage Conventions

The following table defines the standard for register allocation. The OpenPOWER Architecture provides a variety of registers. All the general purpose registers, vector scalar registers, and vector registers are visible to all procedures in a running program.

In the 64-bit OpenPOWER Architecture, there are always 32 general-purpose registers, each 64 bits wide. Throughout this document the symbol rN is used, where N is a register number, to refer to general-purpose register N.

Table 17. OpenPOWER Register Allocation
Type Name Preservation Rules Purpose
General r0 Volatile Optional use in function linkage. Used in function prologues.
r1 Nonvolatile Stack frame pointer.
r2 Nonvolatile(1) TOC pointer.
r3–r10 Volatile Parameter and return values.
r11 Volatile Optional use in function linkage. Used as an environment pointer in languages that require environment pointers.
r12 Volatile Optional use in function linkage. Function entry address at the global entry point.
r13 Reserved Thread pointer.
r14–r31(2) Nonvolatile Local variables.
Floating-point f0 Volatile Local variables.
f1–f13 Volatile Used for parameter passing and return values of binary float types.
f14–f31 Nonvolatile Local variables.
Vector v0–v1 Volatile Local variables.
v2–v13 Volatile Used for parameter passing and return values.
v14–v19 Volatile Local variables.
v20–v31 Nonvolatile Local variables.

(1) Register r2 is nonvolatile with respect to calls between functions in the same compilation unit. It is saved and restored by code inserted by the linker resolving a call to an external function.

(2) If a function needs a frame pointer, assigning r31 to the role of the frame pointer is recommended.

In OpenPOWER-compliant processors, floating-point and vector functions are implemented using a unified vector-scalar model. As shown in Figure 3 and Figure 4, there are 64 vector-scalar registers; each is 128 bits wide.

The vector-scalar registers can be addressed with vector-scalar instructions, for vector and scalar processing of all 64 registers, or with the “classic” Power floating-point instructions to refer to a 32-register subset of 64 bits per register. They can also be addressed with VMX instructions to refer to a 32-register subset of 128-bit wide registers.

Figure 3. Floating-point Registers as Part of Vector Scalar Registers
Floating-point registers take up a quarter               portion of the vector scalar space.

Figure 4. Vector Registers as Part of Vector Scalar Registers
Showing how vector registers take up a half                     portion of the vector scalar space.

The classic floating-point repertoire consists of 32 floating-point registers, each 64 bits wide, and an associated special-purpose register to provide floating-point status and control. Throughout this document, the symbol fN is used, where N is a register number, to refer to floating-point register N.

For the purpose of function calls, the right half of VSX registers, corresponding to the classic floating-point registers (that is, vsr0–vsr31), is volatile.

Single-precision and double-precision shall be passed in the floating-point registers. Single-precision decimal floating-point shall occupy the lower half of a floating-point register. When a floating-point register is skipped during input parameter allocation, words in the corresponding GPR or memory doubleword in the parameter list are not skipped.

The OpenPOWER vector-category instruction repertoire provides the ability to reference 32 vector registers, each 128 bits wide, of the vector-scalar register file, and a special-purpose register VSCR. Throughout this document, the symbol vN is used, where N is a register number, to refer to vector register N.

Parameters in the long double format with a pair of two double-precision floating-point values shall be passed in two successive floating-point registers.

If only one value can be passed in a floating-point register, the second parameter will be passed in a GPR or in memory in accordance with the parameter passing rules for structure aggregates.

OpenPOWER Stack Frame Organization

Figure 5. Stack Frame Organization
Relative layout of an allocated stack frame                     following a nonleaf function call.

OpenPOWER Stack Usage Conventions

  • The stack shall be quadword aligned.
  • The minimum stack frame size shall be 32 bytes. A minimum stack frame consists of the first 4 doublewords (back-chain doubleword, CR save word and reserved word, LR save doubleword, and TOC pointer doubleword), with padding to meet the 16-byte alignment requirement.
  • There is no maximum stack frame size defined.
  • Padding shall be added to the Local Variable Space of the stack frame to maintain the defined stack frame alignment.
  • The stack pointer, r1, shall always point to the lowest address doubleword of the most recently allocated stack frame.
  • The stack shall start at high addresses and grow downward toward lower addresses.
  • The lowest address doubleword (the back-chain word in Figure 5) shall point to the previously allocated stack frame when a back chain is present. As an exception, the first stack frame shall have a value of 0 (NULL).
  • If required, the stack pointer shall be decremented in the called function's prologue and restored in the called function's epilogue.
  • The stack pointer shall be updated atomically so that, at all times, it points to a valid back-chain doubleword if a back chain is maintained.
  • Before a function calls any other functions, it shall save the value of the LR register into the LR save doubleword of the caller's stack frame.

Back Chain Doubleword

When a back chain is not present, alternate information compatible with the ABI unwind framework to unwind a stack must be provided by the compiler, for all languages, regardless of language features. A compiler that does not provide such system-compatible unwind information must generate a back chain. All compilers shall generate back chain information by default, and default libraries shall contain a back chain.

CR Save Word

If a function changes the value in any nonvolatile field of the condition register, it shall first save at least the value of those nonvolatile fields of the condition register, to restore before function exit. The caller frame CR Save Word may be used as the save location. This location in the current frame may be used as temporary storage, which is volatile over function calls.

Reserved Word

This word is reserved for system functions. Modifications of the value contained in this word are prohibited unless explicitly allowed by future ABI amendments.

LR Save Doubleword

If a function changes the value of the link register, it must first save the old value to restore before function exit. The caller frame LR Save Doubleword may be used as the save location. This location in the current frame may be used as temporary storage, which is volatile over a function call.

TOC Pointer Doubleword

If a function changes the value of the TOC pointer register, it shall first save it in the TOC pointer doubleword.

OpenPOWER Optional Save Areas

This ABI provides a stack frame with a number of optional save areas. These areas are always present, but may be of size 0. This section indicates the relative position of these save areas in relation to each other and the primary elements of the stack frame.

Because the back-chain word of a stack frame must maintain quadword alignment, a reserved word is introduced above the CR save word to provide a quadword-aligned minimal stack frame and align the doublewords within the fixed stack frame portion at doubleword boundaries.

An optional alignment padding to a quadword-boundary element might be necessary above the Vector Register Save Area to provide 16-byte alignment, as shown in Figure 5.

Floating-Point Register Save Area

If a function changes the value in any nonvolatile floating-point register fN, it shall first save the value in fN in the Floating-Point Register Save Area and restore the register upon function exit.

The Floating-Point Register Save Area is always doubleword aligned. The size of the Floating-Point Register Save Area depends upon the number of floating-point registers that must be saved. If no floating-point registers are to be saved, the Floating-Point Register Save Area has a zero size.

General-Purpose Register Save Area

If a function changes the value in any nonvolatile general-purpose register rN, it shall first save the value in rN in the General-Purpose Register Save Area and restore the register upon function exit.

The General-Purpose Register Save Area is always doubleword aligned. The size of the General-Purpose Register Save Area depends upon the number of general registers that must be saved. If no general-purpose registers are to be saved, the General-Purpose Register Save Area has a zero size.

Vector Register Save Area

If a function changes the value in any nonvolatile vector register vN, it shall first save the value in vN in the Vector Register Save Area and restore the register upon function exit.

The Vector Register Save Area is always quadword aligned. If necessary to ensure suitable alignment of the vector save area, a padding doubleword may be introduced between the vector register and General- Purpose Register Save Areas, and/or the Local Variable Space may be expanded to the next quadword boundary. The size of the Vector Register Save Area depends upon the number of vector registers that must be saved. It ranges from 0 bytes to a maximum of 192 bytes (12 × 16). If no vector registers are to be saved, the Vector Register Save Area has a zero size.

Local Variable Space

The Local Variable Space is used for allocation of local variables. The Local Variable Space is located immediately above the Parameter Save Area, at a higher address. There is no restriction on the size of this area.

Note: Sometimes a register spill area is needed. It is typically positioned above the Local Variable Space.

The Local Variable Space also contains any parameters that need to be assigned a memory address when the function's parameter list does not require a save area to be allocated by the caller.

Parameter Save Area

The Parameter Save Area shall be allocated by the caller for function calls unless a prototype is provided for the callee indicating that all parameters can be passed in registers. (This requires a Parameter Save Area to be created for functions where the number and type of parameters exceeds the registers available for parameter passing in registers, for those functions where the prototype contains an ellipsis to indicate a variadic function, and functions are declared without prototype.)

When the caller allocates the Parameter Save Area, it will always be automatically quadword aligned because it must always start at SP + 32. It shall be at least 8 doublewords in length. If a function needs to pass more than 8 doublewords of arguments, the Parameter Save Area shall be large enough to spill all register-based parameters and to contain the arguments that the caller stores in it.

The calling function cannot expect that the contents of this save area are valid when returning from the callee.

The Parameter Save Area, which is located at a fixed offset of 32 bytes from the stack pointer, is reserved in each stack frame for use as an argument list when an in-memory argument list is required. For example, a Parameter Save Area must be allocated by the caller when calling functions with the following characteristics:

  • Prototyped functions where the parameters cannot be contained in the parameter registers
  • Prototyped functions with variadic arguments
  • Functions without a suitable declaration available to the caller to determine the called function's characteristics (for example, functions in C without a prototype in scope).

Under these circumstances, a minimum of 8 doublewords are always reserved. The size of this area must be sufficient to hold the longest argument list being passed by the function that owns the stack frame. Although not all arguments for a particular call are located in storage, when an in-memory parameter list is required, consider the parameters to be forming a list in this area. Each argument occupies one or more doublewords.

More arguments might be passed than can be stored in the parameter registers. In that case, the remaining arguments are stored in the Parameter Save Area. The values passed on the stack are identical to the values placed in registers. Therefore, the stack contains register images for the values that are not placed into registers.

This ABI uses a simple va_list type for variable lists to point to the memory location of the next parameter. Therefore, regardless of type, variable arguments must always be in the same location so that they can be found at runtime. The first 8 doublewords are located in general registers r3–r10. Any additional doublewords are located in the stack Parameter Save Area. Alignment requirements such as those for vector types may require the va_list pointer to first be aligned before accessing a value.

Follow these rules for parameter passing:

  • Map each argument to enough doublewords in the Parameter Save Area to hold its value.
  • Map single-precision floating-point values to the least-significant word in a single doubleword.
  • Map double-precision floating-point values to a single doubleword.
  • Map simple integer types (char, short, int, long, enum) to a single doubleword. Sign or zero extend values shorter than a doubleword to a doubleword based on whether the source data type is signed or unsigned.
  • When 128-bit integer types are passed by value, map each to two consecutive GPRs, two consecutive doublewords, or a GPR and a doubleword. The required alignment of int128 data types is 16 bytes. Therefore, by-value parameters must be copied to a new location in the local variable area of the callee's stack frame before the address of the type can be provided (for example, using the address-of operator, or when the variable is to be passed by reference), when the incoming parameter is not aligned at a 16-byte boundary.
  • Map long double to two consecutive doublewords. The required alignment of long double data types is 16 bytes. Therefore, by-value parameters must be copied to a new location in the local variable area of the callee's stack frame before the address of the type can be provided (for example, using the address-of operator, or when the variable is to be passed by reference), when the incoming parameter is not aligned at a 16-byte boundary.
  • Map complex floating-point and complex integer types as if the argument was specified as separate real and imaginary parts.
  • Map pointers to a single doubleword.
  • Map vectors to a single quadword, quadword aligned. This might result in skipped doublewords in the Parameter Save Area.
  • Map fixed-size aggregates and unions passed by value to as many doublewords of the Parameter Save Area as the value uses in memory. Align aggregates and unions as follows:
    • Aggregates that contain qualified floating-point or vector arguments are normally aligned at the alignment of their base type. For more information about qualified arguments, see OpenPOWER Parameter Passing in Registers.
    • Other aggregates are normally aligned in accordance with the aggregate's defined alignment.
    • The alignment will never be larger than the stack frame alignment (16 bytes).

    This might result in doublewords being skipped for alignment. When a doubleword in the Parameter Save Area (or its GPR copy) contains at least a portion of a structure, that doubleword must contain all other portions mapping to the same doubleword. (That is, a doubleword can either be completely valid, or completely invalid, but not partially valid and invalid, except in the last doubleword where invalid padding may be present.)

  • Pad an aggregate or union smaller than one doubleword in size so that it is in the least-significant bits of the doubleword. Pad all others, if necessary, at their tail. Variable size aggregates or unions are passed by reference.
  • Map other scalar values to the number of doublewords required by their size.
  • Future data types that have an architecturally defined quadword-required alignment will be aligned at a quadword boundary.
  • If the callee has a known prototype, arguments are converted to the type of the corresponding parameter when loaded to their parameter registers or when being mapped into the Parameter Save Area. For example, if a long is used as an argument to a float double parameter, the value is converted to double precision and mapped to a doubleword in the Parameter Save Area.

OpenPOWER Protected Zone

The 288 bytes below the stack pointer are available as volatile program storage that is not preserved across function calls. Interrupt handlers and any other functions that might run without an explicit call must take care to preserve a protected zone, also referred to as the red zone, of 512 bytes that consists of:

  • The 288-byte volatile program storage region that is used to hold saved registers and local variables
  • An additional 224 bytes below the volatile program storage region that is set aside as a volatile system storage region for system functions

If a function does not call other functions and does not need more stack space than is available in the volatile program storage region (that is, 288 bytes), it does not need to have a stack frame. The 224-byte volatile system storage region is not available to compilers for allocation to saved registers and local variables.

OpenPOWER Parameter Passing in Registers

For the OpenPOWER Architecture, it is more efficient to pass arguments to functions in registers rather than through memory. For more information about passing parameters through memory, see Parameter Save Area under OpenPOWER Optional Save Areas. For the OpenPOWER ABI, the following parameters can be passed in registers:

  • Up to eight arguments can be passed in general-purpose registers r3–r10.
  • Up to thirteen qualified floating-point arguments can be passed in floating-point registers f1–f13 or up to twelve in vector registers v2–v13.
  • Up to thirteen single-precision or double-precision decimal floating-point arguments can be passed in floating-point registers f1–f13.
  • Up to six quad-precision decimal floating-point arguments can be passed in even-odd floating-point register pairs f2–f13.
  • Up to 12 qualified vector arguments can be passed in v2–v13.

A qualified floating-point argument corresponds to:

  • A scalar floating-point data type
  • Each member of a complex floating-point type
  • A member of a homogeneous aggregate of multiple like data types passed in up to eight floating-point registers

    A homogeneous aggregate can consist of a variety of nested constructs including structures, unions, and array members, which shall be traversed to determine the types and number of members of the base floating-point type. (A complex floating-point data type is treated as if two separate scalar values of the base type were passed.)

    Homogeneous floating-point aggregates can have up to four long double members or eight members of floating-point types. (Unions are treated as their largest member. For homogeneous unions, different union alternatives may have different sizes, provided that all union members are homogeneous with respect to each other.) They are passed in floating-point registers if parameters of that type would be passed in floating-point registers. They are passed in vector registers if parameters of that type would be passed in vector registers. They are passed as if each member was specified as a separate parameter.

A qualified vector argument corresponds to:

  • A vector data type
  • A member of a homogeneous aggregate of multiple like data types passed in up to eight vector registers
  • Any future type requiring 16-byte alignment (see OpenPOWER Optional Save Areas) or processed in vector registers

    A homogeneous aggregate can consist of a variety of nested constructs including structures, unions, and array members, which shall be traversed to determine the types and number of members of the base vector type. Homogeneous vector aggregates with up to eight members are passed in up to eight vector registers as if each member was specified as a separate parameter. (Unions are treated as their largest member. For homogeneous unions, different union alternatives may have different sizes, provided that all union members are homogeneous with respect to each other.)

Note: Floating-point and vector aggregates that contain padding words and integer fields with a width of 0 should not be treated as homogeneous aggregates.

A homogeneous aggregate is either a homogeneous floating-point aggregate or a homogeneous vector aggregate. This ABI does not specify homogeneous aggregates for integer types.

Long double numbers in are passed using two successive floating-point registers. A floating-point register might be skipped to allocate an even/odd register pair when necessary. When a floating-point register is skipped, no corresponding memory word is skipped in the natural home location; that is, the corresponding GPR or memory doubleword in the parameter list.

All other aggregates are passed in consecutive GPRs, in GPRs and in memory, or in memory.

When a parameter is passed in a floating-point or vector register, a number of GPRs are skipped, in allocation order, commensurate to the size of the corresponding in-memory representation of the passed argument's type.

Each parameter is allocated to at least one doubleword.

Full doubleword rule:

When a doubleword in the Parameter Save Area (or its GPR copy) contains at least a portion of a structure, that doubleword must contain all other portions mapping to the same doubleword. (That is, a doubleword can either be completely valid, or completely invalid, but not partially valid and invalid, except in the last doubleword where invalid padding may be present.)

Long Double

Long double parameters are passed as if they were a struct consisting of separate double parameters.

Long double parameters shall be considered as a distinct type for the determination of homogeneous aggregates.

If fewer arguments are needed, the unused registers defined previously will contain undefined values on entry to the called function.

If there are more arguments than registers or no function prototype is provided, a function must provide space for all arguments in its stack frame. When this happens, only the minimum storage needed to contain all arguments (including allocating space for parameters passed in registers) needs to be allocated in the stack frame.

General-purpose registers r3–r10 correspond to the allocation of parameters to the first 8 doublewords of the Parameter Save Area. Specifically, this requires a suitable number of general-purpose registers to be skipped to correspond to parameters passed in floating-point and vector registers.

If a parameter corresponds to an unnamed parameter that corresponds to the ellipsis, a caller shall promote float values to double. If a parameter corresponds to an unnamed parameter that corresponds to the ellipsis, the parameter shall be passed in a GPR or in the Parameter Save Area.

If no function prototype is available, the caller shall promote float values to double and pass floating-point parameters in both available floating-point registers and in the Parameter Save Area. If no function prototype is available, the caller shall pass vector parameters in both available vector registers and in the Parameter Save Area. (If the callee expects a float parameter, the result will be incorrect.)

It is the callee's responsibility to allocate storage for the stored data in the local variable area. When the callee's parameter list indicates that the caller must allocate the Parameter Save Area (because at least one parameter must be passed in memory or an ellipsis is present in the prototype), the callee may use the preallocated Parameter Save Area to save incoming parameters.

OpenPOWER Parameter Passing Register Selection Algorithm

The following algorithm describes where arguments are passed for the C language. In this algorithm, arguments are assumed to be ordered from left (first argument) to right. The actual order of evaluation for arguments is unspecified.

  • gr contains the number of the next available general-purpose register.
  • fr contains the number of the next available floating-point register.
  • vr contains the number of the next available vector register.
Note: The following types refer to the type of the argument as declared by the function prototype. The argument values are converted (if necessary) to the types of the prototype arguments before passing them to the called function.

If a prototype is not present, or it is a variable argument prototype and the argument is after the ellipsis, the type refers to the type of the data objects being passed to the called function.

  • INITIALIZE: If the function return type requires a storage buffer, set gr = 4; else set gr = 3.
    Set fr = 1
    Set vr = 2
  • SCAN: If there are no more arguments, terminate. Otherwise, allocate as follows based on the class of the function argument:
    switch(class(argument))
    
    integer:
    pointer:
    
        if gr > 10
            goto mem_argument
        pass (GPR, gr, argument);
        gr++
    
        break;
    
    aggregate:
        if (homogeneous(argument,float) and regs_needed(members(argument)) <= 8)
            n_fregs = n_fregs_for_type(member_type(argument,0))
            agg_size = members(argument * n_fregs
            reg_size = min(agg_size, 15-fr)
            pass(FPR,fr,first_n_DW(argument,reg_size)
            fr += reg_size;
            gr += size_in_DW (first_n_DW(argument,reg_size))
    
            if remaining_members
                argument = after_n_DW(argument,reg_size))
                goto gpr_struct
        break;
    
        if (homogeneous(argument,vector) and members(argument) <= 8)
            use_vrs:
                agg_size = members(argument)
                reg_size = min(agg_size, 14-vr)
                if (gr&1 = 0) // align vector in memory
                    gr++
                pass(VR,vr,first_n_elements(argument,reg_size);
                vr += reg_size
                gr += size_in_DW (first_n_elements(argument,reg_size)
    
                if remaining_members
                    argument = after_n_elements(argument,reg_size))
                    goto gpr_struct
    
        break;
    
        if gr > 10
            goto mem_argument
    
        size = size_in_DW(argument)
    
    gpr_struct:
        reg_size = min(size, 11-gr)
        pass (GPR, gr, first_n_DW (argument, reg_size));
        gr += size_in_DW (first_n_DW (argument, reg_size))
    
        if remaining_members
            argument = after_n_DW(argument,reg_size))
            goto mem_argument
    
        break;
    
    float:
    
    // float is passed in one FPR.
    // double is passed in one FPR.
    
        if (register_type_used (type (argument)) == vr)
            goto use_vr;
        if fr > 14
            goto mem_argument
    
        n_fregs = n_fregs_for_type(argument) // Assumes n_fregs_for_type == 2
                                             // for long double == 1 for float
                                             // or double
        pass(FPR,fr,argument)
        fr += n_fregs
        gr += size_in_DW(argument)
    
        break;
    
    vector:
        Use vr:
            if vr > 13
                goto mem_argument
    
            if (gr&1 = 0) // align vector in memory
                gr++
    
            pass(VR,vr,argument)
            vr ++
            gr += 2
    
            break;
    
    next argument;
    
    mem_argument:
        need_save_area = TRUE
        pass (stack, gr, argument)
        gr += size_in_DW(argument)
    
    next argument;

All complex data types are handled as if two scalar values of the base type were passed as separate parameters.

If the callee takes the address of any of its parameters, values passed in registers are stored to memory. It is the callee's responsibility to allocate storage for the stored data in the local variable area. When the callee's parameter list indicates that the caller must allocate the Parameter Save Area (because at least one parameter must be passed in memory, or an ellipsis is present in the prototype), the callee may use the preallocated Parameter Save Area to save incoming parameters. (If an ellipsis is present, using the preallocated Parameter Save Area ensures that all arguments are contiguous.) If the compilation unit for the caller contains a function prototype, but the callee has a mismatching definition, this may result in the wrong values being stored.

Note: If the declaration of a function that is used by the caller does not match the definition for the called function, corruption of the caller's stack space can occur.

OpenPOWER Variable Argument Lists

C programs that are intended to be portable across different compilers and architectures must use the header file <stdarg.h> to deal with variable argument lists. This header file contains a set of macro definitions that define how to step through an argument list. The implementation of this header file may vary across different architectures, but the interface is the same.

C programs that do not use this header file for the variable argument list and assume that all the arguments are passed on the stack in increasing order on the stack are not portable, especially on architectures that pass some of the arguments in registers. The OpenPOWER Architecture is one of the architectures that passes some of the arguments in registers.

The parameter list may be zero length and is only allocated when parameters are spilled, when a function has unnamed parameters, or when no prototype is provided. When the Parameter Save Area is allocated, the Parameter Save Area must be large enough to accommodate all parameters, including parameters passed in registers.

OpenPOWER Return Values

Functions that return a value shall place the result in the same registers as if the return value was the first named input argument to a function unless the return value is a nonhomogeneous aggregate larger than 2 doublewords or a homogeneous aggregate with more than eight registers. For a definition of homogeneous aggregates, see OpenPOWER Parameter Passing in Registers. (Homogeneous aggregates are arrays, structs, or unions of a homogeneous floating-point or vector type and of a known fixed size.) Therefore, long double functions are returned in f1:f2.

Homogeneous floating-point or vector aggregate return values that consist of up to eight registers with up to eight elements will be returned in floating-point or vector registers that correspond to the parameter registers that would be used if the return value type were the first input parameter to a function.

Aggregates that are not returned by value are returned in a storage buffer provided by the caller. The address is provided as a hidden first input argument in general-purpose register r3.

Functions that return values of the following types shall place the result in register r3 as signed or unsigned integers, as appropriate, and sign extended or zero extended to 64 bits where necessary:

  • char
  • enum
  • short
  • int
  • long
  • pointer to any type
  • _Bool

5.1.3. Linux Fortran Supplement

Sections A2.4.1 through A2.4.4 of the ABI for Linux defines the Fortran supplement. The register usage conventions set forth in that document remain the same for Fortran.

Fortran Fundamental Types

Table 18. Linux Fortran Fundamental Types
Fortran Type Size (bytes) Alignment (bytes)
INTEGER 4 4
INTEGER*1 1 1
INTEGER*2 2 2
INTEGER*4 4 4
INTEGER*8 8 8
LOGICAL 4 4
LOGICAL*1 1 1
LOGICAL*2 2 2
LOGICAL*4 4 4
LOGICAL*8 8 8
BYTE 1 1
CHARACTER*n n 1
REAL 4 4
REAL*4 4 4
REAL*8 8 8
DOUBLE PRECISION 8 8
COMPLEX 8 4
COMPLEX*8 8 4
COMPLEX*16 16 8
DOUBLE COMPLEX 16 8

A logical constant is one of:

  • .TRUE.
  • .FALSE.

The logical constants .TRUE. and .FALSE. are defined to be the four-byte values -1 and 0 respectively. A logical expression is defined to be .TRUE. if its least significant bit is 1 and .FALSE. otherwise.

Note that the value of a character is not automatically NULL-terminated.

Naming Conventions

By default, all globally visible Fortran symbol names (subroutines, functions, common blocks) are converted to lower-case. In addition, an underscore is appended to Fortran global names to distinguish the Fortran name space from the C/C⁠+⁠+ name space.

Argument Passing and Return Conventions

Arguments are passed by reference (i.e., the address of the argument is passed, rather than the argument itself). In contrast, C/C⁠+⁠+ arguments are passed by value.

When passing an argument declared as Fortran type CHARACTER, an argument representing the length of the CHARACTER argument is also passed to the function. This length argument is a four-byte integer passed by value, and is passed at the end of the parameter list following the other formal arguments. A length argument is passed for each CHARACTER argument; the length arguments are passed in the same order as their respective CHARACTER arguments.

A Fortran function, returning a value of type CHARACTER, adds two arguments to the beginning of its argument list. The first additional argument is the address of the area created by the caller for the return value; the second additional argument is the length of the return value. If a Fortran function is declared to return a character value of constant length, for example CHARACTER*4 FUNCTION CHF(), the second extra parameter representing the length of the return value must still be supplied.

On Linux86-64 systems a Fortran complex function returns its value in memory. The caller provides space for the return value and passes the address of this storage as if it were the first argument to the function. On OpenPOWER systems a Fortran complex function returns its value in the same manner as complex functions.

Alternate return specifiers of a Fortran function are not passed as arguments by the caller. The alternate return function passes the appropriate return value back to the caller in %rax on Linux86-64 and in r1 on OpenPOWER.

The handling of the following Fortran 90 features is implementation-defined: internal procedures, pointer arguments, assumed-shape arguments, functions returning arrays, and functions returning derived types.

Inter-language Calling

Inter-language calling between Fortran and C/C⁠+⁠+ is possible if function/subroutine parameters and return values match types.

  • If a C/C⁠+⁠+ function returns a value, call it from Fortran as a function, otherwise, call it as a subroutine.
  • If a Fortran function has type CHARACTER (or COMPLEX on Linux86-64), call it from C/C⁠+⁠+ as a void function.
  • If a Fortran subroutine has alternate returns, call it from C/C⁠+⁠+ as a function returning int; the value of such a subroutine is the value of the integer expression specified in the alternate RETURN statement.
  • If a Fortran subroutine does not contain alternate returns, call it from C/C⁠+⁠+ as a void function.

Fortran 2003 also provides a mechanism to support interoperability with C. This mechanism inclues the ISO_C_BINDING intrinsic module, binding labels, and the BIND attribute.

Table 19 provides the C/C⁠+⁠+ data type corresponding to each Fortran data type.

Table 19. Fortran and C/C⁠+⁠+ Data Type Compatibility
Fortran Type C/C++ Type Size (bytes)
CHARACTER*n x char x[n] n
REAL x float x 4
REAL*4 x float x 4
REAL*8 x double x 8
DOUBLE PRECISION x double x 8
INTEGER x int x 4
INTEGER*1 x signed char x 1
INTEGER*2 x short x 2
INTEGER*4 x int x 4
INTEGER*8 x long x, or long long x 8
LOGICAL x int x 4
LOGICAL*1 x char x 1
LOGICAL*2 x short x 2
LOGICAL*4 x int x 4
LOGICAL*8 x long x, or long long x 8
Table 20. Fortran and C/C++ Representation of the COMPLEX Type
Fortran Type (lower case) C/C++ Type Size (bytes)
complex x struct {float r,i;} x; 8
  float complex x;  
complex*8 x struct {float r,i;} x; 8
  float complex x; 8
double complex x struct {double dr,di;} x; 16
  double complex x; 16
complex *16 x struct {double dr,di;} x; 16
  double complex x; 16
Note: For C/C++, the complex type implies C99 or later.

Arrays

C/C++ arrays and Fortran arrays use different default initial array index values. By default, C/C⁠+⁠+ arrays start at 0 and Fortran arrays start at 1. A Fortran array can be declared to start at zero.

Another difference between Fortran and C/C⁠+⁠+ arrays is the storage method used. Fortran uses column-major order and C/C⁠+⁠+ use row-major order. For one-dimensional arrays, this poses no problems. For two-dimensional arrays, where there are an equal number of rows and columns, row and column indexes can simply be reversed. Inter-language function mixing is not recommended for arrays other than single dimensional arrays and square two-dimensional arrays.

Structures, Unions, Maps, and Derived Types

Fields within Fortran structures and derived types, and multiple map declarations within a Fortran union, conform to the same alignment requirements used by C structures.

Common Blocks

A named Fortran common block can be represented in C/C⁠+⁠+ by a structure whose members correspond to the members of the common block. The name of the structure in C/C⁠+⁠+ must have the added underscore.

For example, the Fortran common block:

INTEGER I, J
COMPLEX C
DOUBLE COMPLEX CD
DOUBLE PRECISION D
COMMON /COM/ i, j, c, cd, d

is represented in C with the following equivalent:

extern struct {
    int i;
    int j;
    struct {float real, imag;} c;
    struct {double real, imag;} cd;
    double d;
} com_;

and in C++ with the following equivalent:

extern "C" struct {
    int i;
    int j;
    struct {float real, imag;} c;
    struct {double real, imag;} cd;
    double d;
} com_;
Note: The compiler-provided name of the BLANK COMMON block is implementation specific.

Calling Fortran COMPLEX and CHARACTER functions from C/C⁠+⁠+ is not as straightforward as calling other types of Fortran functions. Additional arguments must be passed to the Fortran function by the C/C⁠+⁠+ caller. A Fortran COMPLEX function returns its value in memory; the first argument passed to the function must contain the address of the storage for this value. A Fortran CHARACTER function adds two arguments to the beginning of its argument list. The following example of calling a Fortran CHARACTER function from C/C⁠+⁠+ illustrates these caller-provided extra parameters:

CHARACTER*(*) FUNCTION CHF(C1, I)
CHARACTER*(*) C1
INTEGER I
END
extern void chf_();
char tmp[10];
char c1[9];
int i;
chf_(tmp, 10, c1, &i, 9);

The extra parameters tmp and 10 are supplied for the return value, while 9 is supplied as the length of c1.

6. C++ Dialect Supported

The NVC++ compiler accepts the C++ language of the ISO/IEC standards up to and including the 14882:2017 standard, plus substantially all GNU C++ extensions.

Command-line options provide full support of many C++ variants, including strict standard conformance. NVC++ provides the --c++XY command-line options to enable the user to specify the version of C++ accepted, where XY is one of {17 | 14 | 11 | 03}. The C++ version accepted by default is determined by and matches that of the version of the GCC toolchain used for compilation.

6.1. C++17 Language Features Accepted

The NVC⁠+⁠+ compiler includes support for the C⁠+⁠+⁠17 language standard. Enable this support by compiling with -⁠-⁠c⁠+⁠+17 or -⁠std=c⁠+⁠+17.

Supported C⁠+⁠+⁠17 core language features are available on Linux systems using a GCC 7 or later toolchain.

The following C⁠+⁠+⁠17 language features are supported:

  • Structured bindings
  • Selection statements with initializers
  • Compile-time conditional statements, a.k.a. constexpr if
  • Fold expressions
  • Inline variables
  • Constexpr lambdas
  • Lambda capture of *this by value
  • Class template deduction
  • Auto non-type template parameters
  • Guaranteed copy elision

The NVC++ compiler installation does not include a C⁠+⁠+ standard library, so support for C⁠+⁠+⁠17 additions to the standard library depends on the C⁠+⁠+ library provided on your system. On Linux, GCC 7 is the first GCC release with significant C⁠+⁠+⁠17 support.

The following C⁠+⁠+ library changes are supported when building against GCC 7 or later:

  • std::string_view
  • std::optional
  • std::variant
  • std::any
  • Variable templates for metafunctions

The following C⁠+⁠+ library changes are supported when building against GCC 9 or later:

  • Parallel algorithms
  • Filesystem support
  • Polymorphic allocators and memory resources

7. x86-64 C++ and C MMX/SSE/AVX Intrinsics

An intrinsic is a function available in a given language whose implementation is handled specifically by the compiler. Typically, an intrinsic substitutes a sequence of automatically-generated instructions for the original function call. Since the compiler has an intimate knowledge of the intrinsic function, it can better integrate it and optimize it for the situation.

NVIDIA provides support for MMX and SSE/SSE2/SSE3/SSSE3/SSE4a/ABM/AVX intrinsics in C++ and C programs.

Intrinsics make the use of processor-specific enhancements easier because they provide a C++ and C language interface to assembly instructions. In doing so, the compiler manages things that the user would normally have to be concerned with, such as register names, register allocations, and memory locations of data.

This section contains these tables associated with inline intrinsics:

  • A table of MMX inline intrinsics (mmintrin.h)
  • A table of SSE inline intrinsics (xmmintrin.h)
  • A table of SSE2 inline intrinsics (emmintrin.h)
  • A table of SSE3 inline intrinsics (pmmintrin.h)
  • A table of SSSE3 inline intrinsics (tmmintrin.h)
  • A table of SSE4a inline intrinsics (ammintrin.h)
  • A table of ABM inline intrinsics (intrin.h)
  • A table of AVX inline intrinsics (immintrin.h)

7.1. Using Intrinsic functions

The definitions of the intrinsics are provided in the corresponding header files.

7.1.1. Required Header File

To call these intrinsic functions from a C/C++ source, you must include the corresponding header file – one of the following:

  • For MMX, use mmintrin.h
  • For SSE, use xmmintrin.h
  • For SSE2, use emmintrin.h
  • For SSE3, use pmmintrin.h
  • For SSSE3 use tmmintrin.h
  • For SSE4a use ammintrin.h
  • For ABM use intrin.h
  • For AVX use intrin.h

7.1.2. Intrinsic Data Types

The following table describes the data types that are defined for intrinsics:

Data Types Defined in Description
__m64 mmintrin.h For use with MMX intrinsics, this 64-bit data type stores one 64-bit or two 32-bit integer values.
__m128 xmmintrin.h For use with SSE intrinsics, this 128-bit data type, aligned on 16-byte boundaries, stores four single-precision floating point values.
__m128d emmintrin.h For use with SSE2/SSE3 intrinsics, this 128-bit data type, aligned on 16-byte boundaries, stores two double-precision floating point values.
__ m128i emmintrin.h For use with SSE2/SSE3 intrinsics, this 128-bit data type, aligned on 16-byte boundaries, stores two 64-bit integer values.
__m256 immintrin.h For use with AVX intrinsics, this 256-bit data type, aligned on 31-byte boundaries, stores eight single-precision floating point values.
__m256d immintrin.h For use with AVX intrinsics, this 256-bit data type, aligned on 32-byte boundaries, stores four double-precision floating point values.
__m256i immintrin.h For use with AVX intrinsics, this 256-bit data type, aligned on 16-byte boundaries, stores four 64-bit integer values.

7.1.3. Intrinsic Example

The MMX/SSE intrinsics include functions for initializing variables of the types defined in the preceding table. The following sample program, example.c, illustrates the use of the SSE intrinsics _mm_add_ps and _mm_set_ps.

#include<xmmintrin.h>
 int main(){
 __m128 A, B, result;
 A = _mm_set_ps(23.3, 43.7, 234.234, 98.746); /* initialize A */
 B = _mm_set_ps(15.4, 34.3, 4.1, 8.6); /* initialize B */
 result = _mm_add_ps(A, B);
 return 0;
 }
To compile this program, use the following command:
$ nvc example.c -o myprog

7.2. x86-64 MMX Intrinsics

NVC++ and NVC support a set of MMX Intrinsics which allow the use of the MMX instructions directly from C++ and C code, without writing the assembly instructions. The following table lists the MMX intrinsics supported.

Note: Intrinsics with a * are only available on 64-bit systems.
Table 21. MMX Intrinsics (mmintrin.h)
_mm_empty _m_paddd _m_psllw _m_pand
_m_empty _mm_add_si64 _mm_slli_pi16 _mm_andnot_si64
_mm_cvtsi32_si64 _mm_adds_pi8 _m_psllwi _m_pandn
_m_from_int _m_paddsb _mm_sll_pi32 _mm_or_si64
_mm_cvtsi64x_si64* _mm_adds_pi16 _m_pslld _m_por
_mm_set_pi64x* _m_paddsw _mm_slli_pi32 _mm_xor_si64
_mm_cvtsi64_si32 _mm_adds_pu8 _m_pslldi _m_pxor
_m_to_int _m_paddusb _mm_sll_si64 _mm_cmpeq_pi8
_mm_cvtsi64_si64x* _mm_adds_pu16 _m_psllq _m_pcmpeqb
_mm_packs_pi16* _m_paddusw _mm_slli_si64 _mm_cmpgt_pi8
_m_packsswb _mm_sub_pi8 _m_psllqi _m_pcmpgtb
_mm_packs_pi32 _m_psubb _mm_sra_pi16 _mm_cmpeq_pi16
_m_packssdw _mm_sub_pi16 _m_psraw _m_pcmpeqw
_mm_packs_pu16 _m_psubw _mm_srai_pi16 _mm_cmpgt_pi16
_m_packuswb _mm_sub_pi32 _m_psrawi _m_pcmpgtw
_mm_unpackhi_pi8 _m_psubd _mm_sra_pi32 _mm_cmpeq_pi32
_m_punpckhbw _mm_sub_si64 _m_psrad _m_pcmpeqd
_mm_unpackhi_pi16 _mm_subs_pi8 _mm_srai_pi32 _mm_cmpgt_pi32
_m_punpckhwd _m_psubsb _m_psradi _m_pcmpgtd
_mm_unpackhi_pi32 _mm_subs_pi16 _mm_srl_pi16 _mm_setzero_si64
_m_punpckhdq _m_psubsw _m_psrlw _mm_set_pi32
_mm_unpacklo_pi8 _mm_subs_pu8 _mm_srli_pi16 _mm_set_pi16
_m_punpcklbw _m_psubusb _m_psrlwi _mm_set_pi8
_mm_unpacklo_pi16 _mm_subs_pu16 _mm_srl_pi32 _mm_setr_pi32
_m_punpcklwd _m_psubusw _m_psrld _mm_setr_pi16
_mm_unpacklo_pi32 _mm_madd_pi16 _mm_srli_pi32 _mm_setr_pi8
_m_punpckldq _m_pmaddwd _m_psrldi _mm_set1_pi32
_mm_add_pi8 _mm_mulhi_pi16 _mm_srl_si64 _mm_set1_pi16
_m_paddb _m_pmulhw _m_psrlq _mm_set1_pi8
_mm_add_pi16 _mm_mullo_pi16 _mm_srli_si64  
_m_paddw _m_pmullw _m_psrlqi  
_mm_add_pi32 _mm_sll_pi16 _mm_and_si64  

7.3. x86-64 SSE Intrinsics

NVC++ and NVC support a set of SSE Intrinsics which allows the use of the SSE instructions directly from C++ and C code, without writing the assembly instructions. The following tables list the SSE intrinsics supported.

Note: Intrinsics with a * are only available on 64-bit systems.
Table 22. SSE Intrinsics (xmmintrin.h)
_mm_add_ss _mm_comige_ss _mm_load_ss
_mm_sub_ss _mm_comineq_ss _mm_load1_ps
_mm_mul_ss _mm_ucomieq_ss _mm_load_ps1
_mm_div_ss _mm_ucomilt_ss _mm_load_ps
_mm_sqrt_ss _mm_ucomile_ss _mm_loadu_ps
_mm_rcp_ss _mm_ucomigt_ss _mm_loadr_ps
_mm_rsqrt_ss _mm_ucomige_ss _mm_set_ss
_mm_min_ss _mm_ucomineq_ss _mm_set1_ps
_mm_max_ss _mm_cvtss_si32 _mm_set_ps1
_mm_add_ps _mm_cvt_ss2si _mm_set_ps
_mm_sub_ps _mm_cvtss_si64x* _mm_setr_ps
_mm_mul_ps _mm_cvtps_pi32 _mm_store_ss
_mm_div_ps _mm_cvt_ps2pi _mm_store_ps
_mm_sqrt_ps _mm_cvttss_si32 _mm_store1_ps
_mm_rcp_ps _mm_cvtt_ss2si _mm_store_ps1
_mm_rsqrt_ps _mm_cvttss_si64x* _mm_storeu_ps
_mm_min_ps _mm_cvttps_pi32 _mm_storer_ps
_mm_max_ps _mm_cvtt_ps2pi _mm_move_ss
_mm_and_ps _mm_cvtsi32_ss _mm_extract_pi16
_mm_andnot_ps _mm_cvt_si2ss _m_pextrw
_mm_or_ps _mm_cvtsi64x_ss* _mm_insert_pi16
_mm_xor_ps _mm_cvtpi32_ps _m_pinsrw
_mm_cmpeq_ss _mm_cvt_pi2ps _mm_max_pi16
_mm_cmplt_ss _mm_movelh_ps _m_pmaxsw
_mm_cmple_ss _mm_setzero_ps _mm_max_pu8
_mm_cmpgt_ss _mm_cvtpi16_ps _m_pmaxub
_mm_cmpge_ss _mm_cvtpu16_ps _mm_min_pi16
_mm_cmpneq_ss _mm_cvtpi8_ps _m_pminsw
_mm_cmpnlt_ss _mm_cvtpu8_ps _mm_min_pu8
_mm_cmpnle_ss _mm_cvtpi32x2_ps _m_pminub
_mm_cmpngt_ss _mm_movehl_ps _mm_movemask_pi8
_mm_cmpnge_ss _mm_cvtps_pi16 _m_pmovmskb
_mm_cmpord_ss _mm_cvtps_pi8 _mm_mulhi_pu16
_mm_cmpunord_ss _mm_shuffle_ps _m_pmulhuw
_mm_cmpeq_ps _mm_unpackhi_ps _mm_shuffle_pi16
_mm_cmplt_ps _mm_unpacklo_ps _m_pshufw
_mm_cmple_ps _mm_loadh_pi _mm_maskmove_si64
_mm_cmpgt_ps _mm_storeh_pi _m_maskmovq
_mm_cmpge_ps _mm_loadl_pi _mm_avg_pu8
_mm_cmpneq_ps _mm_storel_pi _m_pavgb
_mm_cmpnlt_ps _mm_movemask_ps _mm_avg_pu16
_mm_cmpnle_ps _mm_getcsr _m_pavgw
_mm_cmpngt_ps _MM_GET_EXCEPTION_STATE _mm_sad_pu8
_mm_cmpnge_ps _MM_GET_EXCEPTION_MASK _m_psadbw
_mm_cmpord_ps _MM_GET_ROUNDING_MODE _mm_prefetch
_mm_cmpunord_ps _MM_GET_FLUSH_ZERO_MODE _mm_stream_pi
_mm_comieq_ss _mm_setcsr _mm_stream_ps
_mm_comilt_ss _MM_SET_EXCEPTION_STATE _mm_sfence
_mm_comile_ss _MM_SET_EXCEPTION_MASK _mm_pause
_mm_comigt_ss _MM_SET_ROUNDING_MODE _MM_TRANSPOSE4_PS
  _MM_SET_FLUSH_ZERO_MODE  

Table 23 lists the SSE2 intrinsics that are supported and available in emmintrin.h.

Table 23. SSE2 Intrinsics (emmintrin.h)
_mm_load_sd _mm_cmpge_sd _mm_cvtps_pd _mm_srl_epi32
_mm_load1_pd _mm_cmpneq_sd _mm_cvtsd_si32 _mm_srl_epi64
_mm_load_pd1 _mm_cmpnlt_sd _mm_cvtsd_si64x* _mm_slli_epi16
_mm_load_pd _mm_cmpnle_sd _mm_cvttsd_si32 _mm_slli_epi32
_mm_loadu_pd _mm_cmpngt_sd _mm_cvttsd_si64x* _mm_slli_epi64
_mm_loadr_pd _mm_cmpnge_sd _mm_cvtsd_ss _mm_srai_epi16
_mm_set_sd _mm_cmpord_sd _mm_cvtsi32_sd _mm_srai_epi32
_mm_set1_pd _mm_cmpunord_sd _mm_cvtsi64x_sd* _mm_srli_epi16
_mm_set_pd1 _mm_comieq_sd _mm_cvtss_sd _mm_srli_epi32
_mm_set_pd _mm_comilt_sd _mm_unpackhi_pd _mm_srli_epi64
_mm_setr_pd _mm_comile_sd _mm_unpacklo_pd _mm_and_si128
_mm_setzero_pd _mm_comigt_sd _mm_loadh_pd _mm_andnot_si128
_mm_store_sd _mm_comige_sd _mm_storeh_pd _mm_or_si128
_mm_store_pd _mm_comineq_sd _mm_loadl_pd _mm_xor_si128
_mm_store1_pd _mm_ucomieq_sd _mm_storel_pd _mm_cmpeq_epi8
_mm_store_pd1 _mm_ucomilt_sd _mm_movemask_pd _mm_cmpeq_epi16
_mm_storeu_pd _mm_ucomile_sd _mm_packs_epi16 _mm_cmpeq_epi32
_mm_storer_pd _mm_ucomigt_sd _mm_packs_epi32 _mm_cmplt_epi8
_mm_move_sd _mm_ucomige_sd _mm_packus_epi16 _mm_cmplt_epi16
_mm_add_pd _mm_ucomineq_sd _mm_unpackhi_epi8 _mm_cmplt_epi32
_mm_add_sd _mm_load_si128 _mm_unpackhi_epi16 _mm_cmpgt_epi8
_mm_sub_pd _mm_loadu_si128 _mm_unpackhi_epi32 _mm_cmpgt_epi16
_mm_sub_sd _mm_loadl_epi64 _mm_unpackhi_epi64 _mm_srl_epi16
_mm_mul_pd _mm_store_si128 _mm_unpacklo_epi8 _mm_cmpgt_epi32
_mm_mul_sd _mm_storeu_si128 _mm_unpacklo_epi16 _mm_max_epi16
_mm_div_pd _mm_storel_epi64 _mm_unpacklo_epi32 _mm_max_epu8
_mm_div_sd _mm_movepi64_pi64 _mm_unpacklo_epi64 _mm_min_epi16
_mm_sqrt_pd _mm_move_epi64 _mm_add_epi8 _mm_min_epu8
_mm_sqrt_sd _mm_setzero_si128 _mm_add_epi16 _mm_movemask_epi8
_mm_min_pd _mm_set_epi64 _mm_add_epi32 _mm_mulhi_epu16
_mm_min_sd _mm_set_epi32 _mm_add_epi64 _mm_maskmoveu_si128
_mm_max_pd _mm_set_epi64x* _mm_adds_epi8 _mm_avg_epu8
_mm_max_sd _mm_set_epi16 _mm_adds_epi16 _mm_avg_epu16
_mm_and_pd _mm_set_epi8 _mm_adds_epu8 _mm_sad_epu8
_mm_andnot_pd _mm_set1_epi64 _mm_adds_epu16 _mm_stream_si32
_mm_or_pd _mm_set1_epi32 _mm_sub_epi8 _mm_stream_si128
_mm_xor_pd _mm_set1_epi64x* _mm_sub_epi16 _mm_stream_pd
_mm_cmpeq_pd _mm_set1_epi16 _mm_sub_epi32 _mm_movpi64_epi64