, as described below
c99
conforms to ISO/IEC 9899:1999 standard for C programs
c11
conforms to ISO/IEC 9899:2011 standard for C programs
c17
conforms to ISO/IEC 9899:2017 standard for C programs
c18
conforms to ISO/IEC 9899:2018 standard for C programs
c++11
enables C++11 support for C++ programs
c++14
enables C++14 support for C++ programs
c++17
enables C++17 support for C++ programs
c++20
enables C++20 support for C++ programs
c89
conforms to ISO/IEC 9899:1990 standard for C programs
gnu89
conforms to ISO C90 plus GNU extensions
gnu99
conforms to ISO C99 plus GNU extensions
gnu++98
conforms to 1998 ISO C++ standard plus GNU extensions
gnu++11
conforms to 2011 ISO C++ standard plus GNU extensions
gnu++14
conforms to 2014 ISO C++ standard plus GNU extensions
gnu++17
conforms to 2017 ISO C++ standard plus GNU extensions
gnu++20 c
onforms to 2020 ISO C++ standard plus GNU extensions
-strict-ansi
Implement a strict ANSI conformance dialect
Preprocessing
Unless explicitly disabled by the file extension as described above files are preprocessed. If you pass the -E option the file will be preprocessed only and will not be compiled. The output is sent to the standard output
OpenMP support
-fopenmp
Enable handling of OpenMP directives
-qopenmp-stubs
Compile OpenMP programs in sequential mode
-parallel
Auto parallelize
OpenACC support
Offlading will not work on all platforms and may require additional options.
- Important compiler specific options
-Wall
This enables all the warnings about constructions that some
users consider questionable, and that are easy to avoid (or
modify to prevent the warning)
-Wextra
This enables some extra warning flags that are not enabled by
-Wall.
-help [category] print full or category help message
Valid categories include
advanced - Advanced Optimizations
codegen - Code Generation
compatibility - Compatibility
component - Component Control
data - Data
deprecated - Deprecated Options
diagnostics - Compiler Diagnostics
float - Floating Point
help - Help
inline - Inlining
ipo - Interprocedural Optimization (IPO)
language - Language
link - Linking/Linker
misc - Miscellaneous
opt - Optimization
output - Output
pgo - Profile Guided Optimization (PGO)
preproc - Preprocessor
reports - Optimization Reports
openmp - OpenMP and Parallel Processing
Moving to Intel's new compiler icx
The Intel compilers icc and icpc are being retired and being replaced with icx and icpx.
Other than the name change many people will not notice significant differences.
The document https://www.intel.com/content/www/us/en/developer/articles/guide/porting-guide-for-icc-users-to-dpcpp-or-icx.html
has details. Here are some important blurbs from that page.
ICX and ICC Classic use different compiler drivers. The Intel® C++ Compiler Classic
compiler drivers are icc, icpc, and icl. The Intel® oneAPI DPC++/C++ Compiler drivers
are icx and icpx. Use icx to compile and link C programs, and icpx for C++ programs.
Unlike the icc driver, icx does not use the file extension to determine whether to
compile as C or C+. Users must invoke icpx to compile C+ files. . In addition to
providing a core C++ Compiler, ICX/ICPX is also used to compile SYCL/DPC++ codes for the
Intel® oneAPI Data Parallel C++ Compiler when we pass an additional flag “-fsycl”.
The major changes in compiler defaults are listed below:
- The Intel® oneAPI DPC++/C++ Compiler drivers are icx and icpx.
- Intel® C++ Compiler Classic uses icc, icpc or icl drivers but this compiler will be deprecated in the upcoming release.
- DPC++/SYCL users can use the icx/icpx driver along with the -fsycl flag which invokes ICX with SYCL extensions.
- Unlike Clang, the ICX Default floating point model was chosen to match ICC behavior and by default it is -fp-model=fast .
- MACRO naming is changing. Please be sure to check release notes for future macros to be included in ICX.
- No diagnostics numbers are listed for remarks, warnings, or notes. Every diagnostic is emitted with the corresponding compiler option to disable it.
- Compiler intrinsics cannot be automatically recognized without processor targeting options, unlike the behavior in Intel® C++ Compiler Classic. If you use intrinsics, read more on the documentation about intrinsic behavior changes.
ifort
This discussion is for version 2021.6.0. Ifort will be replaced with a clang backend based alternative in the near future, ifx. Ifx will have most of the same options as ifort with some clang additions. In the Cray environment if PrgEnv-intel is loaded the "cc" maps to icc.
Normal invocation
# Compile and link a program with the executable sent to the indicated
file
ifort mycode.f90 -o myexec
# Compile a file but don't link
ifort -c mycode.c
Default optimization
The default optimization level is -O2.
-O1 optimize for maximum speed, but disable some optimizations which
increase code size for a small speed benefit
-O2 optimize for maximum speed (DEFAULT)
-O3 optimize for maximum speed and enable more aggressive
optimizations that may not improve performance on some programs
-O same as -O2
-Os enable speed optimizations, but disable some optimizations which
increase code size for small speed benefit
-O0 disable optimizations
-Ofast enable -O3 -no-prec-div -fp-model fast=2 optimizations
-fno-alias assume no aliasing in program
-fno-fnalias assume no aliasing within functions, but assume
aliasing across calls
-fast enable -xHOST -ipo -no-prec-div -O3 -static -fp-model=fast=2
optimizations
-opt_report Generate and optimization report
You can learn more about optimizations are at various levels of optimization as shown below.
-g[n]
0 Disables generation of symbolic debug information.
1 Produces minimal debug information for performing stack traces.
2 Produces complete debug information. This is the same as specifying -g with no n.
3 Produces extra information that may be useful for some tools.
none Disables all check options.
arg_temp_created Determines whether checking occurs for actual
arguments copied into temporary storage before routine calls.
assume Determines whether checking occurs to test that the
scalar-Boolean-expression in the ASSUME directive is true, or
that the addresses in the ASSUME_ALIGNED directive are aligned
on the specified byte boundaries.
bounds Determines whether checking occurs for array subscript and
character s ubstring expressions.
contiguous Determines whether the compiler checks pointer
contiguity at pointer-assignment time.
format Determines whether checking occurs for the data type of an
item being formatted for output.
output_conversion Determines whether checking occurs for the fit
of data items within a designated format descriptor field.
pointers Determines whether checking occurs for certain
disassociated or uninitialized pointers or unallocated
allocatable objects.
shape Determines whether array conformance checking is performed.
stack Determines whether checking occurs on the stack frame.
teams Determines whether the run-time system diagnoses
non-standard coarray team usage.
udio_iostat Determines whether conformance checking occurs when
user-defined derived type input/output routines are executed.
uninit Determines whether checking occurs for uninitialized
variables.
all Enables all check options.
-Os Generate extra code to write profile information suitable for
the analysis program gprof
Some file extensions
Filenames with the suffix .f90 are interpreted as free-form Fortran
95/90 source files.
Filenames with the suffix .f, .for, or .ftn are interpreted as
fixed-form Fortran source files.
Filenames with the suffix .fpp, .F, .FOR, .FTN, or .FPP are
interpreted as fixed-form Fortran source files, which must be
preprocessed by the fpp preprocessor before being compiled.
Filenames with the suffix .F90 are interpreted as free-form Fortran
source files, which must be pre-processed by the fpp preprocessor
before being compiled.
You can specify explicitly the language for file indepenent of the extension using the -x option. For example icc -x c file.cc will complie the program as C instead of C++.
Language standard settings (Dialect)
-stand
none Tells the compiler to issue no messages for nonstandard
language elements. This is the same as specifying nostand.
f90 Tells the compiler to issue messages for language elements
that are not standard in Fortran 90.
f95 Tells the compiler to issue messages for language elements
that are not standard in Fortran 95.
f03 Tells the compiler to issue messages for language elements
that are not standard in Fortran 2003.
f08 Tells the compiler to issue messages for language elements
that are not standard in Fortran 2008.
f18 Tells the compiler to issue messages for language elements
that are not standard in Fortran 2018. This option is set if you
specify warn stderrors.
Generate Listing
Preprocessing
Unless explicitly enabled by the file extension as described above files are not preprocessed. If you pass the -E option the file will be preprocessed only and will not be compiled. The output is sent to the standard output. The option **-fpp ** will force running the preprocessor.
OpenMP support
-fopenmp
Enable handling of OpenMP directives
-qopenmp-stubs
Compile OpenMP programs in sequential mode
-parallel
Auto parallelize
OpenACC support
Coarray Fortran
-coarray[=keyword] Enables the coarray feature where keyword
Specifies the memory system where the coarrays will be
implemented. Possible values are:
shared Indicates a shared memory system. This is the default.
distributed Indicates a distributed memory system.
single Indicates a configuration where the image does not
contain self-replication code. This results in an executable with
a single running image. This configuration can be useful for
debugging purposes, even though there are no inter-image
interactions.
- Important compiler specific options
-save Causes variables to be placed in static memory.
Default: This option saves all variables in static allocation
except local variables within a recursive routine and variables
declared as AUTOMATIC.
-auto-scalar Scalar variables of intrinsic types INTEGER, REAL,
COMPLEX, and LOGICAL are allocated to the run-time stack unless
the routine is recursive of OpenMP For Fortran 95 and later
variables are not saved by default and allocatable arrays are
deallocated. This appears to be true ifort even if the standard
is set to f90. However, it is poor practice to rely on this
behavior.
-Wall. This enables all the warnings about constructions that some
users consider questionable, and that are easy to avoid (or
modify to prevent the warning)
-warn declarations Generate warnings for variables that are not
explicitly typed.
-Wextra This enables some extra warning flags that are not
enabled by -Wall.
-save Causes variables to be placed in static memory.
Default: This option saves all variables in static allocation
except local variables within a recursive routine and variables
declared as AUTOMATIC.
-auto-scalar Scalar variables of intrinsic types INTEGER, REAL,
COMPLEX, and LOGICAL are allocated to the run-time stack unless
the routine is recursive of OpenMP For Fortran 95 and later
variables are not saved by default and allocatable arrays are
deallocated. This appears to be true ifort even if the standard
is set to f90. However, it is poor practice to rely on this
behavior.
-Wall. This enables all the warnings about constructions that some
users consider questionable, and that are easy to avoid (or
modify to prevent the warning)
-warn declarations Generate warnings for variables that are not
explicitly typed.
-Wextra This enables some extra warning flags that are not
enabled by -Wall.
-help [category] print full or category help message
Valid categories include
advanced - Advanced Optimizations
codegen - Code Generation
compatibility - Compatibility
component - Component Control
data - Data
deprecated - Deprecated Options
diagnostics - Compiler Diagnostics
float - Floating Point
help - Help
inline - Inlining
ipo - Interprocedural Optimization (IPO)
language - Language
link - Linking/Linker
misc - Miscellaneous
opt - Optimization
output - Output
pgo - Profile Guided Optimization (PGO)
preproc - Preprocessor
reports - Optimization Reports
openmp - OpenMP and Parallel Processing
Moving to Intel's new compiler ifx
Intel® Fortran Compiler Classic (ifort) is now deprecated and will be discontinued in late 2024.
Intel recommends that customers transition now to using the LLVM-based Intel® Fortran Compiler (ifx).
Other than the name change some people will not notice significant differences. The new compiler
supports offloading to Intel GPU. Kestrel and Swift do not have Intel GPUs so this is not at NREL.
One notable deletion from the new compiler is dropping of auto-parilization. With ifort the
-parallel compiler option auto-parallelization is enabled. That is not true for ifx; there
is no auto-parallelization feature with ifx.
For complete details please see: https://www.intel.com/content/www/us/en/developer/articles/guide/porting-guide-for-ifort-to-ifx.html
Cray CC
In the Cray environment cc is a generic call for several different compilers. The compile actually called is determined by the modules loaded. Here we discuss Cray C : Version 14.0.4. cc will detect if the program being compiled calls MPI routines. If so, it will call the program as MPI. Cray C : Version 14.0.4 is clang based with Cray enhancements
Normal invocation
# Compile and link a program with the executable sent to the indicated
file
cc mycode.c -o myexec
# Compile a file but don't link
cc -c mycode.c
Default optimization
The default optimization level is -O0.
-O0, -O1, -O2, -O3, -Ofast, -Os, -Oz, -Og, -O, -O4 Specify which
optimization level to use:
-O0 Means "no optimization": this
level compiles the fastest and generates the most debuggable
code.
-O1 Somewhere between -O0 and -O2.
-O2 Moderate level of optimization which enables most
optimizations.
-O3 Like -O2, except that it enables optimizations that take
longer to perform or that may generate larger code (in an attempt
to make the program run faster).
-Ofast Enables all the optimizations from -O3 along with other
aggressive optimizations that may violate strict compliance with
language standards.
-Os Like -O2 with extra optimizations to reduce code size.
-Oz Like -Os (and thus -O2), but reduces code size further.
-Og Like -O1. In future versions, this option might disable
different optimizations in order to improve debuggability.
-O Equivalent to -O1.
-O4 and higher Currently equivalent to -O3
Option Run Time (sec)
-O0 10.30
-O1 3.19
-O2 2.99
-O3 2.04
-Ofast 1.88
-Ofast -flto 1.49
-Os 3.19
-Oz 3.31
-Og 3.19
-O 3.20
-fstandalone-debug
Turn off the stripping of some debug information that might be useful to some debuggers
-feliminate-unused-debug-types
By default, Clang does not emit type information for types that are defined but not
used in a program. To retain the debug info for these unused types, the negation
-fno-eliminate-unused-debug-types can be used.
-fexceptions
Enable generation of unwind information. This allows exceptions to be thrown through
Clang compiled stack frames. This is on by default in x86-64.
-ftrapv
Generate code to catch integer overflow errors. Signed integer overflow is undefined
in C. With this flag, extra code is generated to detect this and abort when it happens.
Some file extensions
file.c
C source code that must be preprocessed.
file.i
C source code that should not be preprocessed.
file.ii
C++ source code that should not be preprocessed.
file.cc
file.cp
file.cxx
file.cpp
file.CPP
file.c++
file.C
C++ source code that must be preprocessed.
file.upc
UPC
Language standard settings (Dialect)
Standards are determined by the file extension as given above. Some addttional checks can be performed.
Specify the language standard to compile for.
Supported values for the C language are:
-
ISO C 1999 with GNU extensions
-
iso9899:1990
-
ISO C 2011
-
ISO C 2011 with GNU extensions
-
ISO C 2017
-
ISO C 2017 with GNU extensions
The default C language standard is gnu17
Supported values for the C++ language are:
-
ISO C++ 1998 with amendments
-
ISO C++ 1998 with amendments and GNU extensions
-
ISO C++ 2011 with amendments
-
ISO C++ 2011 with amendments and GNU extensions
-
ISO C++ 2014 with amendments
-
ISO C++ 2014 with amendments and GNU extensions
-
ISO C++ 2017 with amendments
-
ISO C++ 2017 with amendments and GNU extensions
-
Working draft for ISO C++ 2020
-
Working draft for ISO C++ 2020 with GNU extensions
-
The default OpenCL language standard is cl1.0.
Supported values for the CUDA language are:
Generating listing
-fsave-loopmark Generate a loopmark listing file (.lst) that shows which optimizations
were applied to which parts of the source code.
-floopmark-style=<style> Specifies the style of the loopmark listing file.
Valid values for <style> are:
''grouped'' Places all messages at the end of the listing.
''interspersed'' Places each message after the relevant source code line.
Preprocessing
Automatic preprocessing is determined by the file name extension as discussed above. You can manually turn it on/off via the options
-E with output going to standard out
The compiler predefines the macro cray in addition to all of the usual Clang predefined macros.
OpenMP support
-fopenmp Enables OpenMP and links in OpenMP libraries
OpenACC support
Important compiler specific options
Unified Parallel C (UPC) Options
-hupc, -hdefault -hupc Configures the compiler driver to expect
UPC source code. Source files with a .upc extension are
automatically treated as UPC code, but this option permits a file
with any other extension (typically .c) to be understood as UPC
code. -hdefault cancels this behavior; if both -hupc and
-hdefault appear in a command line, whichever appears last takes
precedence and applies to all source files in the command line.
-fupc-auto-amo, -fno-upc-auto-amo Automatically use network
atomics for remote updates to reduce latency. For example, x +=
1 can be performed as a remote atomic add. If an update is
recognized as local to the current thread, then no atomic is
used. These atomics are intended as a performance optimization
only and shall not be relied upon to prevent race conditions.
Enabled at -O1 and above.
-fupc-buffered-async, -fno-upc-buffered-async Set aside memory in
the UPC runtime library for aggregating random remote accesses
designated with "#pragma pgas buffered_async". Disabled by
default.
-fupc-pattern, -fno-upc-pattern Identify simple communication
loops and aggregate the remote accesses into a single function
call which replaces the loop. Enabled at -O1 and above.
-fupc-threads=<N> Set the number of threads for a static THREADS
translation. This option causes __UPC_STATIC_THREADS__ to be
defined instead of __UPC_DYNAMIC_THREADS__ and replaces all uses
of the UPC keyword THREADS with the value N.
Cray ftn
In the Cray environment ftn is a generic call for several different compilers. The compile actually called is determined by the modules loaded. Here we discuss Cray Fortran : Version 14.0.4. Ftn will detect if the program being compiled calls MPI routines. If so, it will call the program as MPI.
Normal invocation
# Compile and link a program with the executable sent to the indicated
file
ftn mycode.f90 -o myexec
# Compile a file but don't link
ftn -c mycode.f90
Default optimization
The default optimization level is -O 2.
-O
0 Specifies no automatic cache management; all memory
references are allocated to cache. Both automatic cache blocking
and manual cache blocking (by use of the BLOCKABLE directive) are
shut off. Characteristics include low compile time. This option
is compatible with all optimization levels.
1 Specifies conservative automatic cache management.
Characteristics include moderate compile time. Symbols are
placed in the cache when the possibility of cache reuse exists
and the predicted cache footprint of the symbol in isolation is
small enough to experience reuse.
2 Specifies moderately aggressive automatic cache management.
Characteristics include moderate compile time. Symbols are
placed in the cache when the possibility of cache reuse exists
and the pre†dicted state of the cache model is such that the
symbol will be reused. (Default)
3 Specifies aggressive automatic cache management.
Characteristics include potentially high compile time. Symbols
are placed in the cache when the possibility of cache reuse
exists and the allocation of the symbol to the cache is
predicted to increase the number of cache hits.
fast Same as 3.
-G (level)
0 Full information is available for debugging, but at the cost
of a slower and larger executable. Breakpoints can be set at
each line. Most optimizations are disabled.
1 Most information is available with partial optimization. Some
optimizations make tracebacks and limited breakpoints available
in the debugger. Some scalar optimizations and all loop nest
re†structuring is disabled, but the source code will be
visible and most symbols will be available.
2 Partial information. Most optimizations, tracebacks and very
limited breakpoints are available in the debugger. The source
code will be visible and some symbols will be available.
-R runchk Specifies any of a group of runtime checks for your
program. To specify more than one type of checking, specify
consecutive runchk arguments, as follows: -R bs.
b Enables checking of array bounds. Bounds checking is not
performed on arrays dimensioned as (1). Enables -Ooverindex.
c Enables conformance checking of array operands in array
expressions.
d Enables a run time check for the !dir$ collapse directive and
checks the validity of the loop_info count information.
p Generates run time code to check the association or allocation
status of referenced POINTER variables, ALLOCATABLE arrays, or
assumed-shape arrays.
s Enables checking of character substring bounds.
Some file extensions
The default is fixed for source files that have .f, .F, .for, or .FOR
The default is free for source files that have .f90, .F90, .f95, .F95, .f03, .F03, .f08, .F08, .f18, .F18, .ftn, or .FTN
The upper-case file extensions, .F, .FOR, .F90, .F95, .F03, .F08, .F18, or .FTN, will enable source preprocessing by default.
Language standard settings (Dialect)
Standards are determined by the file extension as given above. Some addttional checks can be performed.
-e enable
b If enabled, issue a warning message rather than an error
message when the compiler detects a call to a procedure
with one or more dummy arguments having the TARGET,
VOLATILE or ASYNCHRONOUS attribute and there is not an
explicit interface definition.
c Interface checking: use Cray system modules to check
library calls in a compilation. If you have a procedure
with the same name as one in the library, you will get
errors, as the compiler does not skip user- specified
procedures when performing checks.
C Enable/disable some types of standard call site
checking. The current Fortran standard requires that the
number and types of arguments must agree between the caller
and callee. These constraints are enforced in cases where
the compiler can detect them, however, specifying -dC
disables some of this error-checking, which may be
necessary in order to get some older Fortran codes to
compile.
-f source_form free or fixed
Language standard settings (Save)
-e v Allocate variables to static storage. These variables
are treated as if they had appeared in a SAVE statement. Variables
that are explicitly or implicitly defined as automatic variables are
not allocated to static storage. The following types of variables are
not allocated to static storage: automatic variables (explicitly or
implicitly stated), variables declared with the AUTOMATIC attribute,
variables allocated in an ALLOCATE statement, and local
variables in explicit recursive procedures. Variables with the
ALLOCATABLE attribute remain allocated upon procedure exit, unless
explicitly deallocated, but they are not allocated in static memory.
Variables in explicit recursive procedures consist of those in
functions, in subroutines, and in internal procedures within
functions and subroutines that have been defined with the RECURSIVE
attribute. The STACK compiler directive overrides this option.
Generating listing
-h list=a
Preprocessing
Automatic preprocessing is determined by the file name extension as discussed above. You can manually turn it on/off via the options
-E Preprocess and compile
-eZ Preprocess and compile
-eP Preprocess don' compile
gfortran -cpp -E file.F90 > file.f90
ftn file.f80
OpenMP support
-homp Enables OpenMP and links in OpenMP libraries when possible
using CCE-Classic.
-hnoomp Disables OpenMP and links in non-OpenMP libraries when
using CCE-classic.
THE FOLLOWING APPLIE IF THE BACKEND COMPILER IS NOT CRAY FORTRAN.
-fopenmp Enables OpenMP and links in OpenMP libraries when possible
using CCE, AOCC, and GNU.
-openmp Enables OpenMP and links in OpenMP libraries when
possible.
-noopenmp Disables OpenMP.
-mp Enables OpenMP and links in OpenMP libraries when
possible using PGI.
-Mnoopenmp Disables OpenMP and links in non-OpenMP libraries when
using PGI.
-qopenmp Enables OpenMP and links in OpenMP libraries when
possible when using Intel.
-qno-openmp Disables OpenMP and links in non-OpenMP libraries
when possible when using Intel.
OpenACC support
Coarray
The -h pgas_runtime option directs the compiler driver to link with the runtime libraries required when linking programs that use UPC or coarrays. In general, a resource manager job launcher such as aprun or
srun must be used to launch the resulting executable.
Important compiler specific options
-e I Treat all variables as if an IMPLICIT NONE statement had been specified.