24小时热门版块排行榜

南方科技大学公共卫生及应急管理学院2025级博士研究生招生报考通知

返回列表

【奖励】本帖被评价12次，作者dxf_dxf_dxf增加金币 9.4 个

本帖产生 1 个 1ST强帖，点击这里进行查看

dxf_dxf_dxf

银虫 (小有名气)

1ST强帖: 4
应助: 15 (小学生)
金币: 269.1
帖子: 104
在线: 29.1小时
虫号: 1449737

[资源] [VASP编译]五 Makefile中fortran Flags的解释

Gfortran编译器
略

Intel编译器

vasp.5.lib中的Makefile
makefile.linux_ifc_P4
−O0 −FI –FR
−O0 [cite man ifort]
−O[n]
Speciﬁes the code optimization for applications.
Architectures: IA−32, Intel® 64 architectures
Arguments:
n
Is the optimization level. Possible values are 1, 2, or 3. On Linux* OS and Mac OS* X systems, you can also specify 0.
Default:
O2
Optimizes for code speed. This default may change depending on which other compiler options are speciﬁed. For details, see below.
Description:
This option speciﬁes the code optimization for applications.
Option Description
O(Linux* OS and Mac OS* X)
This is the same as specifying O2.
O0 (Linux OS and Mac OS X)
Disables all optimizations.
This option may set other options. This is determined by the compiler,depending on which operating system and architecture you are using. The options that are set may change from release to release.
This option causes certain warn options to be ignored. This is the default if you specify option −debug (with no keyword).
O1
Enables optimizations for speed and disables some optimizations that
increase code size and affect speed. Tolimit code size, this option:
•Enables global optimization; this includes data−ﬂowanalysis, code motion, strength reduction and test replacement, split−lifetime analysis, and instruction scheduling.
This option may set other options. This is determined by the compiler,depending on which operating system and architecture you are using. The options that are set may change from release to release.
The O1 option may improve performance for applications with very large code size, many branches, and execution time not dominated by code within loops.
O2
Enables optimizations for speed. This is the generally recommended optimization level. Vectorization is enabled at O2 and higher levels.
On systems using IA−32 architecture: Some basic loop optimizations such as Distribution, Predicate Opt, Interchange, multi−versioning, and scalar replacements are performed.
This option also enables:
•Inlining of intrinsics
•Intra−ﬁle interprocedural optimization, which includes:
•inlining
•constant propagation
•forward substitution
•routine attribute propagation
•variable address−taken analysis
•dead static function elimination
•removalofunreferenced variables
•The following capabilities for performance gain:
•constant propagation
•copypropagation
•dead−code elimination
•global register allocation
•global instruction scheduling and control speculation
•loop unrolling
•optimized code selection
•partial redundancyelimination
•strength reduction/induction variable simpliﬁcation
•variable renaming
•exception handling optimizations
•tail recursions
•peephole optimizations
•structure assignment lowering and optimizations
•dead store elimination
This option may set other options, especially options that optimize for code speed. This is determined by the compiler,depending on which operating system and architecture you are using. The options that are set may change from release to release.
On Windows systems, this option is the same as the Ox option.
On Linux and Mac OS X systems, if −g is speciﬁed, O2 is turned offand O0 is the default unless O2 (or O1 or O3) is explicitly speciﬁed in the command line together with −g. Manyroutines in the shared libraries are more highly optimized for Intel® microprocessors than for non−Intel microprocessors.
O3 Performs O2 optimizations and enables more aggressive loop transforma tions such as Fusion, Block−Unroll−and−Jam, and collapsing IF statements.
This option may set other options. This is determined by the compiler,depending on which operating system and architecture you are using. The options that are set may change from release to release.
When O3 is used with options −ax or −x (Linux OS) or with options /Qax or /Qx (Windows OS), the compiler performs more aggressive data dependencyanalysis than for O2, which may result in longer compilation times.
The O3 optimizations may not cause higher performance unless loop and memory access transformations takeplace. The optimizations may slowdowncode in some cases compared to O2 optimizations.
The O3 option is recommended for applications that have loops that heavily use ﬂoating−point calculations and process large data sets. Manyroutines in the shared libraries are more highly optimized for Intel® microprocessors than for non−Intel microprocessors.
The last O option speciﬁed on the command line takes precedence overany others.
−FI [cite man ifort]
−noﬁxed
Speciﬁes source ﬁles are in ﬁxed format.
Architectures: IA−32, Intel® 64 architectures
Arguments:
None
Default:
OFF The source ﬁle format is determined from the ﬁle extension.
Description:
This option speciﬁes source ﬁles are in ﬁxed format. If this option is not speciﬁed, format is determined as follows:
•Files with an extension of .f90, .F90, or .i90 are free−format source ﬁles.
•Files with an extension of .f, .for,.FOR, .ftn, .FTN, .fpp, .FPP,or.iare ﬁxed−format ﬁles.
Note that on Linux and Mac OS X systems, ﬁle names and ﬁle extensions are case sensitive.
Alternate Options:
Linux and Mac OS X: −FI
−FR [cite man ifort]
−free
−nofree
Speciﬁes source ﬁles are in free format.
Architectures: IA−32, Intel® 64 architectures
Arguments:
None
Default:
OFF The source ﬁle format is determined from the ﬁle extension.
Description:
This option speciﬁes source ﬁles are in free format. If this option is not speciﬁed, format is deter-
mined as follows:
•Files with an extension of .f90, .F90, or .i90 are free−format source ﬁles.
•Files with an extension of .f, .for,.FOR, .ftn, or .i are ﬁxed−format ﬁles.
Alternate Options:
Linux and Mac OS X: −FR

makefile.linux_ifc_P4
−FR –lowercase –assume byterecl
–lowercase [cite man ifort]
−names keyword
Speciﬁes howsource code identiﬁers and external names are interpreted.
Architectures:
IA−32, Intel® 64 architectures
Arguments:
keyword
Speciﬁes howtointerpret the identiﬁers and external names in source code.
Possible values are:
lowercase
Causes the compiler to ignore case differences in identiﬁers and to convert external names to lowercase.
uppercase
Causes the compiler to ignore case differences in identiﬁers and to convert external names to uppercase.
as_is
Causes the compiler to distinguish case differences inidentiﬁers and to preservethe case of external names.
Default:
lowercase
This is the default on Linux and Mac OS X systems. The compiler ignorescase differences in identiﬁers and converts external names to lowercase.
uppercase
This is the default on Windows systems. The compiler ignores case differences in identiﬁers and converts external names to uppercase.
Description:
This option speciﬁes howsource code identiﬁers and external names are interpreted. It can be useful in mixed−language programming.
This naming convention applies whether names are being deﬁned or referenced.
Youcan use the ALIAS directive tospecify an alternate external name to be used when referring to external subprograms.
CAUTION:
On Windows systems, if you specify option /iface:cref, it overrides the default for external names and causes them to be lowercase. It is as if you speciﬁed "!dec$ attributes c, reference" for the external name.
If you specify option /iface:cref and want external names to be uppercase, you must explicitly specify option /names:uppercase.
Alternate Options:
names lowercase
Linux and Mac OS X: −lowercase (this is a deprecated option)
names uppercase
Linux and Mac OS X: −uppercase (this is a deprecated option)
−assume byterecl [cite man ifort]
−assume keyword
Tells the compiler to makecertain assumptions.
Architectures:
IA−32, Intel® 64 architectures
Arguments:
keyword Speciﬁes the assumptions to be made.
Possible values are:
none
Disables all assume options.
[no]bscc
Determines whether the backslash character is treated asaC−style control character syntax in character literals.
[no]buffered_io
Determines whether data is immediately written to disk or accumulated in a buffer.
[no]byterecl
Determines whether units for the OPEN statement RECL speciﬁer (record length) value in unformatted ﬁles are in bytes or longwords (four−byte units).
[no]cc_omp
Determines whether conditional compilation as deﬁned by the OpenMP Fortran API is enabled or disabled.
[no]dummy_aliases
Determines whether the compiler assumes that dummy arguments to procedures share memory locations with other dummy arguments or with COMMON variables that are assigned.
[no]fpe_summary
Determines whether a ﬂoating−point exceptions summary is displayed when a STOP statement is encountered.
[no]ieee_fpe_ﬂags
Determines whether the ﬂoating−point exception and status ﬂags are savedonroutine entry and restored on routine exit.
[no]minus0
Determines whether the compiler uses Fortran 2003 or Fortran 90/77 standard semantics in the SIGN intrinsic when treating −0.0 and +0.0 as 0.0, and howitwrites the value on formatted output.
[no]old_boz
Determines whether the binary,octal, and hexadecimal constant arguments in intrinsic functions INT,REAL, DBLE, and CMPLX are treated as signed integer constants.
[no]old_ldout_format
Determines the output of integer and real values in list−directed and namelist−directed output.
[no]old_logical_ldio
Determines whether NAMELIST and list−directed input accept logical values for numeric IO−list items.
[no]old_maxminloc
Determines the results of intrinsics MAXLOC and MINLOC when givenanempty array as an argument.
      nobyterecl
Units for OPEN statement RECL values with unfor matted files are in four-byte (longword) units.
      Description:
This option specifies assumptions to be made by the compiler.
      Option
Description
assume byterecl
Specifies that the units for the OPEN statement RECL specifier (record length) value are in bytes for unformatted data files, not long words (four-byte units). For formatted files, the RECL value is always in bytes.
If a file is open for unformatted data and assume byterecl is specified, INQUIRE returns RECL in bytes; otherwise, it returns RECL in long words. An INQUIRE returns RECL in bytes if the unit is not open.
−assume byterecl [cite vasp.5.2/makefile.linux_ifc_P4]
#-----------------------------------------------------------------------
# general fortran flags  (there must a trailing blank on this line)
# byterecl is strictly required for ifc, since otherwise
# the WAVECAR file becomes huge
#-----------------------------------------------------------------------

PGI编译器
vasp.5.lib中的Makefile
makefile.linux_pg & makefile.linux_pgi_opt
-Mx,119,0x200000 -O1 –Mfree
-Mx,119,0x200000 [cite vasp.5.2/makefile.linux_pg]
#-----------------------------------------------------------------------
# general fortran flags  (there must a trailing blank on this line)
# the -Mx,119,0x200000 is required if you use older pgf90 versions
# on a more recent LINUX installation
# the option will not do any harm on other 3.X pgf90 distributions
#-----------------------------------------------------------------------
译：
一般的fortran编译选项（本行结尾必须有一个空格）。
如果你在比较新的LINUX发行版中使用较旧的pgf90版本，你需要“-Mx,119,0x200000”选项。
这个选项不会对其它 3.X pgf90发行版产生任何害处。
-O1[cite man pgf90]
−O[level]
Set the optimization level. If −O is not speciﬁed, then the default level is 1 if −g is not speciﬁed,and 0 if −g is speciﬁed. If a number is not supplied with −O then the optimization levelisset to 2.The optimization levels and their meanings are as follows:
设定优化等级。如果 -O 没有被专门设定，那么如果没有设定 -g ，则默认等级为1，反之为0。如果你设定了一个不被 –O 支持的数字，则优化等级被设为2.优化等级与他们的含义如下：
0 A basic block is generated for each C statement. No scheduling is done between statements. No global optimizations are performed.
0 每个C语句都产生一个基本语句块。不对语句块的顺序重新排列。不进行全局优化。
1 Scheduling within extended basic blocks is performed. Some register allocation is performed. No global optimizations are performed.
1 执行扩展的基本语句块内的重排。某些寄存器被分配以使用。不进行全局优化。
2 All level 1 optimizations are performed. In addition, traditional scalar optimizations such as induction recognition and loop invariant motion are performed by the global optimizer.
执行所有的1级优化。此外，传统的“标量优化”，例如“诱发识别”和“循环不变”等通过全局优化实现。
3 All level 1 and 2 optimizations are performed. In addition, this level enables more
aggressive code hoisting and scalar replacement optimizations that may or may not be proﬁtable.
执行所有的1级和2级优化。此外，这个等级允许更多的、不知是否合适的积极的代码优化和标量取代优化被执行。
4 All level 1,2and 3 optimizations are performed. In addition, hoisting of guarded
invariant ﬂoating point expressions is enabled.
执行所有的1，2和3级优化。此外，这个等级允许对对浮点数表达式的优化。

makefile.linux_pgi_opt
-tp k8-64 –i8
-tp k8-64 [cite man pgf90]
−tp target
Specify the type of the target processor; possibilities are
−tp amd64
AMD Opteron or Athlon-64 in 64-bit mode
−tp amd64e
AMD Opteron revision E or later,in64-bit mode; includes SSE3 instructions
−tp athlon
AMD Athlon processor
−tp athlonxp
AMD AthlonXP and AthlonMP processor
−tp barcelona −tp barcelona−32
AMD Barcelona processor in 32-bit mode
−tp barcelona−64
AMD Barcelona processor in 64-bit mode
−tp core2 −tp core2−32
Intel core2 processor in 32-bit mode
−tp core2−64
Intel core2 processor in 64-bit mode
−tp istanbul −tp istanbul−32
AMD Istanbul architecture Opteron processor,32-bit mode
−tp istanbul−64
AMD Istanbul architecture Opteron processor,64-bit mode
−tp k7 AMD Athlon processor
−tp k8−32
AMD Opteron or Athlon-64 in 32-bit mode
−tp k8−64
AMD Opteron or Athlon-64 in 64-bit mode
−tp k8−64e
AMD Opteron revision E or later,in64-bit mode; includes SSE3 instructions
−tp nehalem −tp nehalem−32
Intel Nehalem architecture Core processor,32-bit mode
−tp nehalem−64
Intel Nehalem architecture Core processor,64-bit mode
−tp p5 Pentium processor
−tp p6 Pentium Pro / Pentium II
−tp p7 −tp p7−32
Pentium 4 processor
−tp p7−64
IA32/EM64T processor in 64-bit mode
−tp penryn −tp penryn−32
Intel Penryn architecture Pentium processor,32-bit mode
−tp penryn−64
Intel Penryn architecture Pentium processor,64-bit mode
−tp piii Pentium III processor
−tp piv Pentium 4 processor
−tp px −tp px−32
Blended code generation that will work on anyx86-compatible processor
−tp px−64
Blended code generation that will work on any64-bit x86 processor
−tp shanghai −tp shanghai−32
AMD Shanghai architecture Opteron processor,32-bit mode
−tp shanghai−64
AMD Shanghai architecture Opteron processor,64-bit mode
−tp x64
Single binary where each procedure is optimized for both the AMD Opteron and the Intel IA32/EM64T in 64-bit mode; the selection of which optimized copy to execute is made at run time depending on the machine executing the code.
The default in the absence of the −tp ﬂag is to compile for the type of CPU on which the compiler is running.

−i8 [cite man pgf90]
−i2 Treat INTEGER variables as twobytes.
−i4 Treat INTEGER variables as four bytes.
−i8 Treat default INTEGER and LOGICAL variables as eight bytes. Foroperations involving
integers, use 64-bits for computations.
vasp.5.2中的Makefile
makefile.linux_pg
−Mfree –Mx,119,0x200000 –O2 –tp p6
−Mfree [cite man pgf90]
−Mfree −Mfreeform −Mnofree −Mnofreeform
Process Fortran source using free form speciﬁcations. The −Mnofree and −Mﬁxed options specify ﬁxed form formatting. By default ﬁles with a .f90, .F90, .f95 or .F95 extension use freeform formatting.
KEY WORDS：
FFLAGS fortran编译选项 ifort Intel PGI pgf90[ Last edited by dxf_dxf_dxf on 2012-4-10 at 15:55 ]

回复此楼

» 本帖附件资源列表

欢迎监督和反馈：小木虫仅提供交流平台，不对该内容负责。
本内容由用户自主发布，如果其内容涉及到知识产权问题，其责任在于用户本人，如对版权有异议，请联系邮箱：xiaomuchong@tal.com
附件 1 : ifort.pdf

2012-04-10 15:42:46, 407.79 K

附件 2 : pgf90.pdf

2012-04-10 15:42:47, 60.08 K

1楼2012-04-10 15:43:18

已阅回复此楼关注TA 给TA发消息送TA红花 TA的回帖

回帖置顶 ( 共有1个 )

dxf_dxf_dxf

银虫 (小有名气)

1ST强帖: 4
应助: 15 (小学生)
金币: 269.1
帖子: 104
在线: 29.1小时
虫号: 1449737

简略版
Gfortran编译器
略
Intel编译器
vasp.5.lib中的Makefile
makefile.linux_ifc_P4
−O0 −FI –FR
−O0 [cite man ifort]
−O[n] n=0,1,2,3
Speciﬁes the code optimization for applications.
−FI [cite man ifort]
−noﬁxed
Speciﬁes source ﬁles are in ﬁxed format.
Linux and Mac OS X: −FI
−FR [cite man ifort]
−free
−nofree
Speciﬁes source ﬁles are in free format.
Alternate Options:
Linux and Mac OS X: −FR
vasp.5.2中的Makefile
makefile.linux_ifc_P4
−FR –lowercase –assume byterecl
–lowercase [cite man ifort]
−names keyword
Speciﬁes howsource code identiﬁers and external names are interpreted.
keyword
Speciﬁes how to interpret the identiﬁers and external names in source code.
Possible values are:
lowercase
Causes the compiler to ignore case differences in identiﬁers and to convert external names to lowercase.
Alternate Options:
names lowercase
Linux and Mac OS X: −lowercase (this is a deprecated option)
names uppercase
Linux and Mac OS X: −uppercase (this is a deprecated option)
−assume byterecl [cite man ifort]
−assume keyword
Tells the compiler to make certain assumptions.
keyword Speciﬁes the assumptions to be made.
[no]byterecl
Determines whether units for the OPEN statement RECL speciﬁer (record length) value in unformatted ﬁles are in bytes or longwords (four−byte units).
−assume byterecl [cite vasp.5.2/makefile.linux_ifc_P4]
#-----------------------------------------------------------------------
# general fortran flags (there must a trailing blank on this line)
# byterecl is strictly required for ifc, since otherwise
# the WAVECAR file becomes huge
#-----------------------------------------------------------------------
−Mfree [cite man pgf90]
PGI编译器
vasp.5.lib中的Makefile
makefile.linux_pg & makefile.linux_pgi_opt
-Mx,119,0x200000 -O1 –Mfree
-Mx,119,0x200000 [cite vasp.5.2/makefile.linux_pg]
#-----------------------------------------------------------------------
# general fortran flags (there must a trailing blank on this line)
# the -Mx,119,0x200000 is required if you use older pgf90 versions
# on a more recent LINUX installation
# the option will not do any harm on other 3.X pgf90 distributions
#-----------------------------------------------------------------------
译：
一般的fortran编译选项（本行结尾必须有一个空格）。
如果你在比较新的LINUX发行版中使用较旧的pgf90版本，你需要“-Mx,119,0x200000”选项。
这个选项不会对其它 3.X pgf90发行版产生任何害处。
-O1[cite man pgf90]
−O[level] 0~4
makefile.linux_pgi_opt
-tp k8-64 –i8
-tp k8-64 [cite man pgf90]
−tp core2−64
Intel core2 processor in 64-bit mode
−i8 [cite man pgf90]
−i8 Treat default INTEGER and LOGICAL variables as eight bytes. Foroperations involving
integers, use 64-bits for computations.
vasp.5.2中的Makefile
makefile.linux_pg
−Mfree –Mx,119,0x200000 –O2 –tp p6
−Mfree [cite man pgf90]
−Mfree −Mfreeform −Mnofree −Mnofreeform
Process Fortran source using free form speciﬁcations. The −Mnofree and −Mﬁxed options specify ﬁxed form formatting. By default ﬁles with a .f90, .F90, .f95 or .F95 extension use freeform formatting.

[ Last edited by dxf_dxf_dxf on 2012-4-10 at 16:06 ]