Home

PGI Compiler User`s Guide

1. 11 4 Using Environment Modules on Linux On Linux if you use the Environment Modules package that is the nodule load command PGI includes a script to set up the appropriate module files Assuming your installation base directory is opt pgi and your MODULEPATH environment variable is usr local Modules modulefiles execute this command opt pgi linux86 15 10 etc modulefiles pgi module install all install usr local Modules modulefiles This command creates module files for all installed versions of the PGI compilers You must have write permission to the nodulefiles directory to enable the module commands module load pgi32 15 10 module load pgi64 15 10 module load pgi 15 10 oe oe oe PGI Compiler User s Guide 146 Using Environment Variables where pgi 15 10 uses the 32 bit compilers on a 32 bit system and uses 64 bit compilers on a 64 bit system To see what versions are available use this command 9 module avail pgi The module load command sets or modifies the environment variables as indicated in the following table This Environment Variable Is set or modified by the module load command Fc er oi an O LI PGI does not provide support for the Environment Modules package For more information about the package go to http modules sourceforge net 11 5 Stack Traceback and JIT Debugging When
2. PGI Compiler User s Guide 137 Environment Variable OMP_NUM_THREADS OMP_SCHEDULE OMP_STACKSIZE OMP_WAIT_POLICY PATH PGI PGI CONTINUE PGI OBJSUFFIX PGI STACK USAGE PGI TERM PGI TERM DEBUG PGROUPD LICENSE FILE PWD STATIC RANDOM SEED TMP TMPDIR Using Environment Variables Description Specifies the number of threads to use during execution of parallel regions Default is Specifies the type of iteration scheduling and optionally the chunk size to use for omp for and omp parallel for loops that include the runtime schedule clause The default is STATIC with chunk size 1 Overrides the default stack size for a newly created thread Sets the behavior of idle threads defining whether they spin or sleep when idle The values are ACTIVE and PASSIVE The defaultis ACTIVE Determines which locations are searched for commands the user may type Specifies at compile time the root directory where the PGI compilers and tools are installed If set when a program compiled with Mch k fps t k is executed the stack is automatically cleaned up and execution then continues Specifies the location of the PGI license This variable is set in the registry on Windows machines and is specific to PGI products On Windows PGROUPD LICENSE FILE does not need to be set Allows you to display the current directory
3. BUG variable may be set to override the default behavior when PGI TERM The value of PGI TERM D For example I BUG should be set to the command line used to invoke the program gdb quiet pid d The first occurrence of 3d in the PGI TERM DEBUG string is replaced by the process id The program named in the PGI TERM DEBUG string must be found on the current PATH or specified with a full path name 11 3 22 PGROUPD LICENSE FILE You can use the PGROUPD LICENSE FILE to specifies the location of the PGI license This variable is set in the registry on Windows machines and is specific to PGI products J The system environment variable PGROUPD_LICENSE_ FILE is not required by PGI products on Windows but you can use it to override the default location that is searched for the license dat file To use the system environment variable PEGROUPD LICENSE FILE set it to the full path of the license keys file To do this follow these steps 1 Open the System Properties dialog from Control Panel System 2 Select the Advanced tab 3 Click the Environment Variables button PGI Compiler User s Guide 145 Using Environment Variables gt If PGROUPD LICENSE FILE is not already an environment variable create a new system variable for it Set its value to the full pat
4. Forces the seed returned by RANDOM SEED to be constant Sets the directory to use for temporary files created during execution of the PGI compilers and tools interchangeable with TMPDI R Sets the directory to use for temporary files created during execution of the PGI compilers and tools 11 3 PGI Environment Variables You use the environment variables listed in Table 27 to alter the default behavior of the PGI compilers and the executables which they generate This section provides more detailed descriptions about the variables in this table that are not OpenMP environment variables 11 3 1 FLEXLM_BATCH By default on Windows the license server creates interactive pop up messages to issue warning and errors You can use the environment variable FLEXLM BATCH to prevent interactive pop up windows To do this set the environment variable FLEXLM BATCH to 1 The following csh example prevents interactive pop up messages for licensing warnings and errors set FLEXLM BATCH 1 PGI Compiler User s Guide 138 Using Environment Variables 11 3 2 FORTRANOPT FORTRANOPT allows the user to adjust the behavior of the PGI Fortran compilers gt If FORTRANOPT exists and contains the value vaxio the record length in the open statement is in units of 4 byte words and the edit descriptor only has an effect for lines beginning with a space or a plus sign gt If FORTRANOPT
5. seen 171 14 7 Medium Memory Model and Large Array in C nne 172 14 8 Medium Memory Model and Large Array in Fortran 173 14 9 Large Array and Small Memory Model in Fortran sssseeeeeneenneenenennns 174 Chapter 15 C C Inline Assembly and Intrinsics eerte ntn 175 15 1 Inline Assembly iccirco rk iiri nian EB EE Rv a 175 15 2 Extended Inline Assembly essen ieeeteneetentnn iE i be serbe ds 175 15 21 Output Operands irta ertet aderire andiae aaa oiio eas dr n reboot dead 176 15 2 2 Input Operands serre eno eere e ra Dade Re UR ee 178 15 2 3 6lODDGT DIS Ue ert etin mattis E T SEE A EN TT EE E ET SUE RAT 180 15 24 Additional Constralnits a roten eer o pe HUE iate D ER n Hen RC nee 181 15 2 5 Simple Gonstralhis ccrte bere o eoa bate te i ON OD Re dee rur ceu eg 181 15 2 6 Machine Constraints as 5 toe ot t re tente qun acte t n Be De odo RO d ti a 182 15 2 7 Multiple Alternative Constraints iet tete rrt tori bireete rte eire n Peor trei nt 184 15 2 8 Constraint MOdIfIers o tnr i I PEG DU a ai a E EU RE SPERM AREE S 185 15 3 Operanid Aliases titt om t i n bed tt d D ta i c Het dh be HE RR RR 186 1544 Assembly String MOdINGES sti screw cte ecd perte cie actin 187 15 5 Extended Asm Macros nennen tenete nnne tnter tenente tenerent tentent teens 189 15 6 ININSICS EU 189 Chapter 16 Contact Information s sisien FoU ROSEO RS RS
6. PGI Compiler User s Guide 131 Creating and Using Libraries 1 Use Mmakeimp1ib with the PGI compilers to build an import library for the first DLL without building the DLL itself pgfortran Bdynamic c object2 f95 pgfortran Mmakeimplib o obj2 lib object2 obj oe gs o Tip The de lt def file gt option can also be used with Mnakeimplib Usea def file when you need to export additional symbols from the DLL A de f file is not needed in this example because all symbols are exported using DLLEXPORT 2 Use the import library obj 2 lib created in Step 1 to link the second DLL oe pgfortran Bdynamic c object3 f95 pgfortran Mmakedll o obj3 dll object3 obj defaultlib obj2 oe 3 Use the import library obj 3 lib created in Step 2 to link the first DLL pgfortran Mmakedll o obj2 dll object2 obj defaultlib 0obj3 4 Compile the main program and link against the import libraries for the two DLLs jOCRE IEEIeciN Beyonce prog 19S proca clitemibcliloge isy 2 cx elsuseubte Lio gS 5 Execute prog2 to ensure that the DLLs were create properly prog2 iue e Ceillilaimig a routine im Clos Chil TODE Silo TUME Sel Calling e zowtine alin 91692 ULL func 2b 10 7 5 Import a Fortran module from a DLL In this example we import a Fortran module from a DLL We use the source file defmod 90 to create a DLL containing a Fortran module We then use the source file use mod
7. 2 2 cicatrice tierce cce caccia edo cud 18 2 2 Help with Command line Options enne tentent tenerent tentent tenerent 18 2 3 Getting Started with Performance us co crier rer rave ttem poet arava do c n gr eve d te tre 20 2 3 1 Using fast and fastsse Options sse R 20 2 3 2 Other Performance Related Options tentent tenentes 21 24 Targeting Multiple Systems Using the tp Option 21 PGI Compiler User s Guide ii 25 Freguenty used ODptIOns ierit ret fire rre diner ci edere FE ER ok Ee PER 22 Chapter 3 Optimizing and Parallelizing iterat tenter dete eiecti ctetu tn 24 3 1 Overview o Optimi eT E 24 3 1 1 Local Optimization rt i eR eai ed On n RERO RR SE RE eau 25 3 1 2 Global Optimization nitide tete e t tte e tad rr leen veta e p te e ER RA 25 3 1 3 Loop Optimization Unrolling Vectorization and Parallelization sss 25 3 1 4 Interprocedural Analysis IPA and Optimization sseeeeeneeen teen 25 3 1 5 Function Inlimihgi sas inet ien tbe tee reri eter tte eter tnnt reete Ore Ee n pere e ER 26 3 1 6 Profile Feedback Optimization PFO seen nennt 26 3 2 Getting Started with Optimization seen tentent tentent te tenen tenentes 26 M 27 E m 27 9 2 9 MNE Oc e ert tote trt or to
8. Returns the type of accelerator device used to run the next accelerator region if one is selected Returns the number of the device being used to execute an accelerator region Returns the total available free memory on the attached accelerator device Returns the total memory on the attached accelerator device Returns the number of accelerator devices of the given type attached to the host Returns the number of accelerator kernels launched since the start of the program Returns the number of accelerator regions entered since the start of the program Returns the number of microseconds spent in accelerator compute regions and in moving data for accelerator data regions 7 12 Environment Variables PGI supports environment variables that modify the behavior of accelerator regions This section defines the user setable environment variables used to control behavior of accelerator enabled programs at execution These environment variables must comply with these rules gt The names of the environment variables must be upper case gt The values assigned environment variables are case insensitive and may have leading and trailing white space gt The behavior is implementation defined if the values of the environment variables change after the program has started even if the program itself modifies the values Table 20 lists and briefly describes the Accelerator environment variables that PGI supports PGI Compiler User s Gu
9. This works but again has the side effect of compiling each source file and recompiling each object file at link time 3 9 6 Building a Program with IPA Several Steps Just by adding the Mipa command line switch it is possible to use individual pgcc commands to compile each source file followed by a command to link the resulting object files into an executable oe eek Mpa ASer eo ira Il S ies ijs c i3 10 92 59 ipee iWeefesi cc i303 pocek Mipa ast eo mggoui filed o dule2 g9 Iule3 39 oe oe oe PGI Compiler User s Guide 45 Optimizing and Parallelizing The pgcc driver invokes the compiler and assembler as required to process each source file and invokes the IPA linker for the final link command If you modify one of the source files the executable can be rebuilt by compiling just that file and then relinking pgcc Mipa fast c filel c pgcc Mipa fast o a out filel o file2 o file3 o oe When the IPA linker is invoked it will determine that the IPA optimized object for filel o filel ipa5 a out oo o is stale since it is older than the object i1e1 0 and hence it needs to be rebuilt and reinvokes the compiler to generate it In addition depending on the nature of the changes to the source file ile1 c the interprocedural optimizations previously performed for 11e2 and file3 may now be inaccurate For instance IPA may have propagated a constant argument value in a call from a function in filel ctoa func
10. epu tale PGI Compiler User s Guide 114 Using Directives and Pragmas When this source is compiled using the Mvect command line option func2 is vectorized but funcl is not vectorized In the following example the global novector pragma turns off vectorization for the entire file include math h funcl pragma global novector closer alioi LOO of LOO LOO 7 sloer CLOG LOO CLO LOO Tg Gime 35535 Fow SLO p aL lt ILO 0 2 alate for q 0 7 9510075 42 eus gg alata a lolita Sli ale eli ial e ela ta a loli cli ale fume 20 Float ALZO00 ZOO 21919 21909 19 Elon C200 200 LEZIONI T 24909 9 aumlE 353p Ror Ste p 3E 20 0 Fates for j50 j3j 200 j e E ep s fep 3 e tale eis celal fal biai Ubri DS Special Scope Rules Special rules apply for a pragma with loop routine and global scope When the pragma is placed within a routine it applies to the routine from its point in the routine to the end of the routine The same rule applies for one of these pragmas with global scope However there are several pragmas for which only routine and global scope applies and which affect code immediately following the pragma gt bounds and fcon The bounds and fcon pragmas behave in a similar manner to pragmas with loop scope That is they apply to the code following the pragma gt opt and safe When the opt or safe pragmas are placed within a routin
11. seen 146 11 9 Stack Traceback and JIT Debugging irri retenti rn ii tese 147 Chapter 12 Distributing Files Deployment eene tentent tenter tnter intent intent 148 12 1 Deploying Applications on LinUx sssssssssseeeenenneentn tnter trennen tenentes 148 12 1 1 Runtime Library Consideratioris ir ntt rte ttd rt ett ve rev te dd 148 12 1 2 64 bit Linux Consideratioris ertet intct reta nce edicit 149 12 1 3 Linux Redistributable Files sssseeeenenneen nennen tenente tenens 149 12 144 Restrictions on Linux Portability ye 21 rct ctim nr tes cim ned tenen an tege 149 12 1 5 Licensing for Redistributable Files seen 149 12 2 Deploying Applications on Windows sese tnter tenen 149 PGI Compiler User s Guide vii 12 2 4 PO Redistrbutables intr rt eb boire e err eme ee re E uS pev 150 12 2 2 Microsoft Redistributables seen tentent tenente tnter tnter 150 12 3 Code Generation and Processor Architecture sse enn 150 12 31 Generating Generic X86 Code me tede ten reet ied ee emi ginis 151 12 3 2 Generating Code for a Specific Processor sense ne ntente tette tette tentent tnn 151 12 4 Generating One Executable for Multiple Types of Processors 151 12 4 1 PGI Unified Binary Command line Switches sees 151 12 4 2 PGI Unified Binary Di
12. 4 1 42 Output FileS cde rr reet Preterea cer Papeete t o oc E p crt eH RE Ta Eee eR ER 6 1 5 Fortran C and C Data TVp8s iar ree eet irem iare RO B Mire iren EROS 7 1 6 Parallel Programming Using the PGI Compilers esee 7 1 6 1 Run SMP Parallel Programs tra eden dene ti eee i ee 8 1 7 Platform specific considerations seen tette tete tete tenente nens 8 1 1 Using the POL Compilers on LINUX iss itinere tree fete rq e nie 9 1 7 2 Using the PGI Compilers on Windows ree recette ain vie yen rp sd EAR Pega 9 1 7 3 PGI on the Windows Desktop tnnt tenente tentent tenens 11 1 4 4 Using the PGI Compilers on OS X ertet tete ttr tte rire rien ra Dea pe eir eee n 12 1 8 Site Specific Customization of the Compilers sssssssssssseeeeneent tentent 14 diol UsesSitero F Ies cese ro m re nete Hrs B SERERE HORIS nud RS a DOR ILE RERIU EE RE ERU RE Ple ania Lig 14 EE PANUEDDPESE did 14 1 9 Common Development Tasks eene tentent te tete tenens 15 Chapter 2 Use Command bine Options aana aeara adaa raana 17 2 1 Command Line Option Overview ssssssseseeeeeenenetn teet nnne trennen tenerent tete tnt teen 17 2 1 1 Command line Options SVMaxt av dene te tete tton tential oda meet deter eo de E ERE 17 2412 Command line SubOpliOrIs 22 rrr tn det tei be n D gg v E redet 18 2 1 3 Command line Contlictinig Optioris
13. Mconcur innermost Enable parallelization of innermost loops Mconcur levels cn Parallelize loops nested at most n levels deep Mconcur n Use thread processors affinity when running on a NUMA architecture Specifying Mconcur nonuma disables this option The environment variable NCPUS is checked at runtime for a parallel program If NCPUS is set to 1 a parallel program runs serially but will use the parallel routines generated during compilation If NCPUS is set to a value greater than 1 the specified number of processors are used to execute the program Setting NCPUS to a value exceeding the number of physical processors can produce inefficient execution Executing a program on multiple processors in an environment where some of the processors are being time shared with another executing job can also result in inefficient execution As with the vectorizer the Mconcur option can speed up code if it contains well behaved countable loops and or computationally intensive nested loops that operate on arrays However it is possible that some codes show a decrease in performance on multi processor systems when compiled with Mconcur due to parallelization overheads memory bandwidth limitations in the target system false sharing of cache lines or other architectural or code generation factors For this reason it is recommended that you check carefully whether particular program units or loops show improved performance when compiled usin
14. Several Minline suboptions let you determine the selection criteria for functions to be inlined These suboptions include PGI Compiler User s Guide 51 Using Function Inlining 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 size n Inlines functions with a statement count less than or equal to n the specified size The size n may not exactly equal the number of statements in a selected function the size parameter is merely a rough gauge levels n Inlines n level of function calling levels The default number is one 1 Using a level greater than one indicates that function calls within inlined functions may be replaced with inlined code This approach allows the function inliner to automatically perform a sequence of inline and extract processes lib file ext Instructs the inliner to inline the functions within the library file 1le ext If no inline library is specified functions are extracted from a temporary library created during an extract prepass cy Tip Create the library file using the Mextract option If you specify both a function name and a size n the compiler inlines functions that match the function name or have n or fewer statements If a name is used withou
15. TABLE OF CONTENTS PIG TACO t svat suavatuasesaut snes savssace anys eassasuenssayeans tourbacs avaaicasetreamseuenn naman an ie xii Audience Descriptioria oti niaaa cer epo ER CL ERR ERE H EE PETER EHE re Ye Re iin avant Xii Compatibility and Conformance to Standards ssssssseeeeeeneeete nennen Xii Ero nag Xiii Hardware and Software Constraints sese senes tens xiv CONVENTIONS ics nectare Dub re edt HER AR ETSI ROS EROR ORE LAE REL QUA ERR ote aa HERE SERO tse ted enero eres xiv TEMS a o eE tete dent ebd hte mter EE A NE AS E a A A ir Ga RR antennas XV Related Publications sisrate ana rien matin Ee He oh eer nara ainnemnasienaanel xvi Chapter 1 Getting Started enenatis cites ete ta cratere Ero ek facra te iet te E era sese E etes ess Sapa R enne 1 J Me TU 4 stresneciiinainrrinaiinidinsirn ni E A inal aaa cae Dadnani enna 1 1 2 Creating an Example aei aber ete i it cial ane Et dc ed a 2 1 3 Invoking the Command level PG Compilers seen tenente enis 2 13 1 Command line Syrntax ucc ertr tt tereti cipere dea voee edi 2 1 3 2 Command line Options sese tentent tenente tenente tete tenerent teen nenne 3 1 3 3 Fortran Directives and C C t Pragimas i rr rt tant v t eve i v c v 3 1 4 Filename Conventions rne ni tne eene ie Re rre rela e ne e be 3 JU MIU ce
16. allocate a m n b m n c m n clo 2 100 am i Co 3 100 im 1 a i j 10000 0D0 dble i dble j b i j 20000 0D0 dble i dble j enddo enddo call mat add a b c m n print we M _m N z n prime Sy Yeu Ww qmm end subroutine mat add a b c m n integer m n i j double precision a m n b m n c m n do i 1 m glo 3 d m e t 3 F a i j ar b i j enddo enddo return end 9 NDS O met allo met SUISSE Mace cav NES PGI Compiler User s Guide 174 Chapter 15 C C INLINE ASSEMBLY AND INTRINSICS 15 1 Inline Assembly Inline Assembly lets you specify machine instructions inside a C function The format for an inline assembly instruction is this esn _ _ csm i Vetring z The asm statement begins with the asm or asm keyword The __asm__ keyword is typically used in header files that may be included in ISO C programs string is one or more machine specific instructions separated with a semi colon or newline vi character These instructions are inserted directly into the compiler s assembly language output for the enclosing function Some simple asm statements are evs UC SERE ames stt These asm statements disable and enable system interrupts respectively In the following example the eax register is set to zero asm pushl eax n t movl 0 eax n t popl eax Notice that eax is pushed on the stack so that it is it not clobbered When the
17. level loop If the outer loop were a timing loop this would be a practical use for a loop scoped directive 9 5 Scope of C C Pragmas and Command Line Options During compilation a pragma either turns an option on or turns an option off Pragmas apply to the section of code corresponding to the specified scope either the entire file the following loop or the following or current routine This section presents several examples showing the effect of pragmas and the use of the pragma scope indicators cy In all cases pragmas override a corresponding command line option For pragmas that have only routine and global scope there are two rules for determining the scope of the pragma We cover these special scope rules at the end of this section Consider the following program main Eloee elio LOO Jol lOO LOO e roo LOO 2 int time maxtime n i j maxtime 10 m O for time 0 time lt maxtime time fow J 07 Jams j FOr 0r cn bs ekiti ep 5 s bliri When this is compiled using the Mvect command line option both interior loops are interchanged with the outer loop Pragmas alter this behavior either globally or on a routine or PGI Compiler User s Guide 113 Using Directives and Pragmas loop by loop basis To ensure that vectorization is not applied use the novector pragma with global scope main pragma global novector Eloet xev MOON MO ON Peto MEO 3 E3E9 1 ei roo TLO IONE TMS ESNS
18. Further it does not cover automatic detection and offloading of regions of code to an accelerator by a compiler or other tool While future versions of the PGI compilers may allow for automatic offloading this feature is not currently supported 7 2 Terminology Clear and consistent terminology is important in describing any programming model This section provides definitions of the terms required for you to effectively use this section and the associated programming model Accelerator a special purpose co processor attached to a CPU and to which the CPU can offload data and executable kernels to perform compute intensive calculations Compute intensity for a given loop region or program unit the ratio of the number of arithmetic operations performed on computed data divided by the number of memory transfers required to move that data between two levels of a memory hierarchy Compute region a structure block defined by an OpenACC compute construct A compute construct is a structured block containing loops which are compiled for the accelerator A compute region may require device memory to be allocated and data to be copied from host to device upon region entry and data to be copied from device to host memory and device memory deallocated upon exit The dynamic range of a compute construct including any code in procedures called from within the construct is the compute region In this release compute regions may not contain other comp
19. Mac OS Debugging Requirements Both the g and Mkeepobj switches play important roles when compiling a program on Apple Mac OS for debugging gt To debug a program with symbol information on the OS X files must be compiled with the g switch to keep the program s object files the files with a o extension Further these object files must remain in the same directory in which they were created gt Ifa program is built with separate compile and link steps by compiling with the c switch which generates the o object files then using the g switch guarantees the required object files are available for debugging Use the following command sequence to compile and then link your code To compile the programs use these commands Ogee NEIGE pacc ca EGO pocok ca opare To link use this command POCE G melno sO oars Linking on OS X On the OS X the PGI Workstation 15 10 compilers do not support static linking of user binaries For compatibility with future Apple updates the compilers support dynamic linking of user binaries For more information on dynamic linking refer to Creating and Using Dynamic Libraries on Mac OS X Running Parallel Programs on OS X You may encounter difficulties running auto parallel or OpenMP programs on OS X systems when the per thread stack size is set to the default MB If you have unexplained failures please try setting the environment variable OMP_STACKSIZE to a larger value such as 16
20. You can use a number of options to specify the optimization levels including O Mvect Mipa and Mconcur In addition you can use several of the M pgf1ag switches to control specific types of optimization and parallelization This chapter describes these optimization options fast Minline O Munroll Mconcur Mipa fast Mpfi Mvect Minfo Mneginfo Mpfo Msafeptr Mipa fast inline This chapter also describes how to choose optimization options to use with the PGI compilers This overview is helpful if you are just getting started with one of the PGI compilers or wish to experiment with individual optimizations Complete specifications of each of these options is available in the Command Line Options Reference section of the PGI Compiler Reference Manual 3 1 Overview of Optimization In general optimization involves using transformations and replacements that generate more efficient code This is done by the compiler and involves replacements that are independent of the PGI Compiler User s Guide 24 Optimizing and Parallelizing particular target processor s architecture as well as replacements that take advantage of the x86 or x64 architecture instruction set and registers For discussion purposes we categorize optimization Local Optimization Global Optimization Loop Optimization Interprocedural Analysis IPA and Optimization Optimization Through Function Inlining Profile Feedback Optimiz
21. clo ak i sg elo aj d 3m al apt Jos z enddo k i 2 aie Gl ems m 2 k in Gh 1 2 enddo If the outer loop is parallelized conflicting values are stored into k by the various processors The variable k cannot be made local to each processor because its value must remain coherent among the processors It is possible the loop could be parallelized if all assignments to k are placed in PGI Compiler User s Guide 39 Optimizing and Parallelizing critical sections However it is not clear where critical sections should be introduced because in general the value for k could depend on another scalar or on k itself and code to obtain the value of other scalars must reside in the same critical section In the previous example the assignment to k within a conditional at label 2 prevents k from being recognized as an induction variable If the conditional statement at label 2 is removed k would be an induction variable whose value varies linearly with j and the loop could be parallelized Scalar Last Values During parallelization scalars within loops often need to be privatized that is each execution thread has its own independent copy of the scalar Problems can arise if a privatized scalar is accessed outside the loop For example consider the following loops in C C and Fortran 7 C C version oye ai dLg SNA sbspesr di ait xe a 54 ig excors ET Fortran version do I 1 N stem dee amp O0
22. code size savings of 10 to 90 compared to generating full copies of code for each target Programs can use the PGI Unified Binary even if all of the object files and libraries are not compiled as unified binaries Like any other object file you can use PGI Unified Binary object PGI Compiler User s Guide 42 Optimizing and Parallelizing files to create programs or libraries No special start up code is needed support is linked in from the PGI libraries The Mipa option disables generation of PGI Unified Binary Instead the default target auto detect rules for the host are used to select the target processor 3 9 Interprocedural Analysis and Optimization using Mipa The PGI Fortran C and C compilers use interprocedural analysis IPA that results in minimal changes to makefiles and the standard edit build run application development cycle Other than adding Mipa to the command line no other changes are required For reference and background the process of building a program without IPA is described later in this section followed by the minor modifications required to use IPA with the PGI compilers While the PGCC compiler is used here to show how IPA works similar capabilities apply to each of the PGI Fortran C and C compilers The examples use Linux file naming conventions On Windows o files would be obj files and a out files would be exe files 3 9 1 Building a Program Without IPA Single Step U
23. cy C on Windows does not support Bdynamic Bstatic Compile for and link to the static version of the PGI runtime libraries This flag corresponds to the MT flag used by Microsoft s c1 compilers On Windows you must use Bstatic for both compiling and linking Mmakedll Generate a dynamic link library or DLL Implies Bdynamic Mmakeimplib Generate an import library without generating a DLL Use this flag when you want to generate an import library for a DLL but are not yet ready to build the DLL itself This situation might arise for example when building DLLs with mutual imports as shown in Build DLLs Containing Mutual Imports Fortran o lt file gt Passed to the linker Name the DLL or import library lt file gt PGI Compiler User s Guide 126 Creating and Using Libraries def lt file gt When used with Mmaked11 this flag is passed to the linker and a def file named file is generated for the DLL The def file contains the symbols exported by the DLL Generating a def file is not required when building a DLL but can be a useful debugging tool if the DLL does not contain the symbols that you expect it to contain When used with Mmakeimplib this flag is passed to 1ib which requires a def file to create an import library The def file can be empty if the list of symbols to export are passed to 1ib on the command line or explicitly marked as DLLEXPORT in the source code implib
24. followed by the operand number For example 0 references operand 0 or the output item amp a w in function example16 in the previous example Extended asm also supports operand aliasing which PGI Compiler User s Guide 186 C C Inline Assembly and Intrinsics allows use of a symbolic name instead of a number for specifying operands as illustrated in this example void examplel7 int w 1 z 0 asm movl 1 outputl1 n addl input S outputl n movil input output2 output1 amp a w output2 r z input r w In example16 0 and output1 both represent the output operand 15 4 Assembly String Modifiers Special character sequences in the assembly string affect the way the assembly is generated by the compiler For example the is an escape sequence for specifying an operand 9696 produces a percent for hard coded registers and n specifies a new line Table 39 summarizes these modifiers known as Assembly String Modifiers Table 39 Assembly String Modifier Characters Modifier Description Same as Vin printf format strings fm Adds a in the assembly string Adds a in the assembly string YA Adds a in front of an operand in the assembly string For example A0 adds a in front of operand 0 in the assembly output Produces the byte op code suffix for this operand For example b0 produces b on x86 and x86 64 Produces the word op code
25. glo u i im Clo xp d ia a i j 10000 0D0 dble i dble j b i j3 20000 0D0 dble i dble j enddo enddo omp parallel omp do Ce a ip im clo 3 1 m aga e ali I a bliri enddo enddo omp do do i 1 m gio g 1 m d 30000 0D0 dble i dble j dble j act el mes cGu53 acm print Ker qu ee print 57 Ue ars pe ve Gas print re Ug Eu el Stop endif enddo enddo Somp end parallel print We M PAM we N u iN print wy Xe n VU E end PGI Compiler User s Guide 173 Programming Considerations for 64 Bit Environments When compiled with the PGFORTRAN compiler using mcmodel medium pgfortran Mfree mp o mat mat f i8 mcmodel medium setenv OMP NUM THREADS 2 mat 16000 N 16000 M N 480032000 0000000 Q Z oe oe oe 14 9 Large Array and Small Memory Model in Fortran The following example uses large dynamically allocated arrays The code is divided into a main and subroutine so you could put the subroutine into a shared library Dynamic allocation of large arrays saves space in the size of executable and saves time initializing data Further the routines can be compiled with 32 bit compilers by just decreasing the parameter size Large Array and Small Memory Model in Fortran 2 car mat alle T90 program mat_allo integer i j integer size m n parameter size 16000 parameter m size n size coulole jrecisiem ellllocataloilessa 8 8 plo 258 pels n 8
26. has an Accelerator tab that displays profiling information provided by the accelerator This information is available in the file pgprof out and is collected by using pgcollect on an executable binary compiled for an accelerator target For more information on pgcollect refer to the pgcollect Reference section of the PGPROF Profiler Guide 7 15 2 NVIDIA CUDA Profile You can use the NVIDIA CUDA Profiler with PGI generated code for the NVIDIA GPUs You may download the CUDA Profiler from the same website as the CUDA software www nvidia com cuda Documentation and support is provided by NVIDIA 7 15 3 TAU Tuning and Analysis Utility You can use the TAU Tuning and Analysis Utility version 2 18 1 with PGI generated accelerator code TAU instruments code at the function or loop level and version 2 18 1 is enhanced with support to track performance in accelerator regions TAU software and documentation is available at this website http tau uoregon edu PGI Compiler User s Guide 102 Using an Accelerator 7 16 Supported 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 Intrinsics make the use of
27. kind and or rank of the actual arguments when checking all the specifics in a generic call for ambiguities 9 8 1 IGNORE TKR Directive Syntax The syntax for the IGNORE TKR directive is this IDIR IGNORE TKR letter dummy arg letter is one or any combination of the following T type K kind R rank PGI Compiler User s Guide 117 Using Directives and Pragmas For example KR indicates to ignore both kind and rank rules and TKR indicates to ignore the type kind and rank arguments lt dummy_arg gt if specified indicates the dummy argument for which TKR rules should be ignored If not specified TKR rules are ignored for all dummy arguments in the procedure that contains the directive 9 8 2 IGNORE_TKR Directive Format Requirements The following rules apply to this directive gt IGNORE_TKR must not specify dummy arguments that are allocatable Fortran 90 pointers or assumed shape arrays gt IGNORE_TKR may appear in the body of an interface block or in the body of a module procedure and may specify dummy argument names only gt IGNORE_TKR may appear before or after the declarations of the dummy arguments it specifies gt f dummy argument names are specified IGNORE_TKR applies only to those particular dummy arguments gt f no dummy argument names are specified IGNORE_TKR applies to all dummy arguments except those that are allocatable objects Fortran 90 pointers or assumed shap
28. lt file gt Passed to the colinker Generate an import library named file for the DLL A DLL s import library is the interface used when linking an executable that depends on routines in a DLL To use the PGI compilers to create an executable that links to the DLL form of the runtime use the compiler flag Bdynamic The executable built will be smaller than one built without Bdynamic the PGI runtime DLLs 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 DLL built by the PGI compilers The following examples outline how to use Bdynamic Mmakedll and Mmakeimplib to build and use DLLs with the PGI compilers cy C on Windows does not support Bdynamic 10 7 1 Build a DLL Fortran This example builds a DLL from a single source file object1 f which exports data and a subroutine using DLLEXPORT The source file prog1 f uses DLLIMPORT to import the data and subroutine from the DLL objectl f subroutine subl i DECS ATTRIBUTES DLLEXPORT subl integer i common acommon adata integer adata DECS ATTRIBUTES DLLEXPORT acommon ponme Savioli acleica acca joraime 7 Wrewgodl a Wy at adata i end progl f program progl common acommon adata integer adata external subl DECS ATTRIBUTES DLLIMPORT subi acommon adata 11 call sub1 12 De
29. or as complex as a subroutine depending on the accelerator hardware 7 6 1 Host Functions Even in accelerator targeted regions the host must orchestrate the execution it allocates memory on the accelerator device initiates data transfer sends the kernel code to the accelerator v v v y passes kernel arguments queues the kernel waits for completion transfers results back to the host v v v yvy deallocates memory In most cases the host can queue a sequence of kernels to be executed on the device one after the other 7 6 2 Levels of Parallelism Most current GPUs support two levels of parallelism gt an outer doall fully parallel loop level gt an inner synchronous SIMD or vector loop level Each level can be multidimensional with 2 or 3 dimensions but the domain must be strictly rectangular The synchronous level may not be fully implemented with SIMD or vector operations so explicit synchronization is supported and required across this level No synchronization is supported between parallel threads across the doall level The execution model on the device side exposes these two levels of parallelism and the programmer is required to understand the difference between for example a fully parallel loop and a loop that is vectorizable but requires synchronization across iterations All fully parallel loops can be scheduled for either doall or synchronous parallel execution but by definition SIMD vector loo
30. 2 at the next nested parallel level OMP_SCHEDULE STATIC with Specifies the type of iteration scheduling and optionally the chunk size chunk size of 1 to use for omp for and omp parallel for loops that include the runtime schedule clause The supported schedule types which can be specified in upper or lower case are static dynamic guided and auto OMP PROC BIND FALSE Specifies whether executing threads should be bound to a core during execution Allowable values are true and false OMP STACKSIZE BEEN Overrides the default stack size for a newly created thread OMP THREAD LIMIT 64 Specifies the absolute maximum number of threads that can be used in a program OMP_WAIT_POLICY ACTIVE Sets the behavior of idle threads defining whether they spin or sleep when idle The values are ACTIVE and PASSIVE PGI Compiler User s Guide 73 Chapter 6 USING MPI Message Passing Interface MPI is an industry standard application programming interface designed for rapid data exchange between processors in a cluster application MPI is computer software used in computer clusters that allows the processes of a parallel application to communicate with one another PGI provides MPI support with PGI compilers and tools PGI compilers provide explicit support to build MPI applications on Windows using Microsoft s implementation of MPI MS MPI on OS X using MPICH and on Linux using MPICH MVAPICH2 and Open MPI Of course you may always b
31. 90 to build a program that imports and uses the Fortran module from de mod 90 defmod f90 module testm EVIOS GL uS dUsE eNyese 9 gum LNE nd type a typ type a type a b DECS ATTRIBUTES DLLEXPORT a b contains subroutine print a DECS ATTRIBUTES DLLEXPORT prime a werte p avan Lme end subroutine subroutine print_b DECS ATTRIBUTES DLLEXPORT print b write b an int end subroutine end module usemod f90 use testm coen int an int cali print caltiprint Io i S E end PGI Compiler User s Guide 132 Creating and Using Libraries 1 Create the DLL pgf90 Mmakedll o defmod dll defmod f90 Creating library defmod lib and object defmod exp 2 Create the exe and link against the import library for the imported DLL pgf90 Bdynamic o usemod usemod f90 defaultlib defmod lib 6 3 Run the exe to ensure that the module was imported from the DLL properly usemod d 2 10 8 Using LIB3F The PGI Fortran compilers include complete support for the de facto standard LIB3F library routines on both Linux and Windows operating systems See the PGI Fortran Language Reference manual for a complete list of available routines in the PGI implementation of LIB3F 10 9 LAPACK BLAS and FFTs Pre compiled versions of the public domain LAPACK and BLAS libraries are included with the PGI compilers The LAPACK library is
32. B result A i SSE PSE 2353 Adat 234 234 88 745 6 B mn see jse la 4 34 3 Gal 8 6 2 result mm add ps A B i amp wEQH NE Op PGI Compiler User s Guide 190 Chapter 16 CONTACT INFORMATION You can contact PGI at 20400 NW Amberwood Drive Suite 100 Beaverton OR 97006 Or electronically using any of the following means Fax 1 503 682 2637 Sales sales pgroup com Support trs pgroup com WWW http www pgroup com The PGI User Forum is monitored by members of the PGI engineering and support teams as well as other PGI customers The forum newsgroups may contain answers to commonly asked questions Log in to the PGI website to access the forum http www pgroup com userforum index php Many questions and problems can be resolved by following instructions and the information available at our frequently asked questions FAQ site http www pgroup com support fag htm All technical support is by e mail or submissions using an online form at http www pgroup com support Phone support is not currently available PGI documentation is available at http www pgroup com resources docs htm or in your local copy of the documentation in the release directory doc index htm PGI Compiler User s Guide 191 Notice ALL NVIDIA DESIGN SPECIFICATIONS REFERENCE BOARDS FILES DRAWINGS DIAGNOSTICS LISTS AND OTHER DOCUMENTS TOGETHER AND SEPARATELY MATERIALS ARE BEING PROVIDED AS IS NVIDIA MAK
33. Bika w i 4 i enddo do j 1 200000 calil Jiexemp 5xp wv pd s OS IN enddo preia sp xb px VIL eLa p 6498 7 xx 9999 end subroutine loop a b c d s n integer i n real 4 ain Din eun gms Clo at IL in erat bi e s clih enddo end Assume the preceding program is compiled as follows where Mvect nosimd disables SSE vectorization 9 pgfortran fast Mvect nosimd Minfo vadd f Mfree o vadd vector op 4 Loop unrolled 16 times Generates 1 prefetches in scalar loop 9 Loop not vectorized parallelized contains call loop 18 Loop unrolled 4 times The following output shows a sample result if the generated executable is run and timed on a 4 Core Intel Sandybridge 2 5 GHz system bin time vadd 1 000000 771 000 color 000 65 298 00 9969 00 1 35user 0 00system 0 01 35elapsed 99 CPU Qavgtext 0avgdata 3936maxresident k Oinputst Ooutputs 0major 290minor pagefaults Oswaps PGI Compiler User s Guide 35 Optimizing and Parallelizing Now recompile with SSE vectorization enabled and you see results similar to these 9 pgfortran fast Minfo vadd f Mfree o vadd VECEOr COD 4 Loop not vectorized may not be beneficial Unrolled inner loop 8 times Residual loop unrolled 7 times completely unrolled Generated 1 prefetches in scalar loop 9 Loop not vectorized parallelized contains call 17 Generated 4 alternate versions of the loop Generated vector sse code for
34. Considerations for 64 Bit Environments In the presence of Mlarge arrays the compiler may silently promote 32 bit integers to 64 bits which can have unexpected side effects On Linux86 64 the Mlarge arrays option also enables single static data objects larger than 2 GB This option is the default in the presence of mcmodel medium cy On Win64 static data may not be larger than 2GB 14 5 Compiler Options for 64 bit Programming The usual switches that apply to 64 bit programmers seeking to increase the data range of their applications are in the following table Table 32 64 bit Compiler Options mcmodel medium Mlargeaddressaware Mlarge arrays Purpose Enlarge object size Allow for declared data the size of larger than 2GB Win64 only Generates code that allows for addresses greater than 2GB using RIP relative addressing Perform all array location to address calculations using 64 bit integer arithmetic Position independent code Necessary for shared libraries All INTEGER functions data and constants not explicitly declared INTEGER 4 are assumed to be INTEGER 8 Considerations Linux86 64 only Slower execution Cannot be used with fPIC Objects cannot be put into shared libraries Use Mlargeaddressawarezno for a direct addressing mechanism that restricts the total addressable memory This is not applicable if the object file is placed in a DLL Further if an object file is compile
35. Directive and Pragma ClallSes s x aeter rere trm ete ni itt eie quee e ee E ln a LERRA 64 5 6 Runtime Library ROUEITIGS c ote rre rtt i Exe e ibo ir p br x pa e 67 5 9 Environment Variables dieere three t ie ete th e dl ete b endet Het i b Et ns T2 Chapter 6 ung 74 6 1 MPIOVeIVIBW c hee ani e res une arie ase an De wd de UR darren dae EXER ie uu d 74 6 2 Compiling and Linking MPI Applications sssseeeeenennnnnt tentent tenerent 75 6 3 Debugging MPI Applications isis 3 eii tco cett tee tere oa cert re olen rene i ER e le 75 6 4 Profiling MPI Applications ertt trees tree ctt aah thee een det ta ek e tup arb t maaan epu 76 6 5 Using MPICEL oni Einux and OS X uoti tet ie tt eh wien tla Rn Re b TT 6 6 Using MPICH1 MPICH2 and MVAPICH1 on Linux trennen ene 77 6 7 Using MVAPICH2 on LINUX asessorina tenente tenete tentent tette nens 78 6 8 Using MS MPl Oni WirldOWs citi ciet rette anomie rent E d e tie t st edet e He eren 78 6 9 Using Open MPI On LINUX assassin adseatanaedun ra eii nan dean ium edo ei edita io dia fae 78 6 10 Using SGI MPI on UNUK EE 79 6 11 Using MPI Compller Wrappers ctt ic trei ert ee cerea mnie 79 612 HMS nna e ee ain dent 79 6 13 Jesting and Benchmarking i sc circi d a a ro enda i ero et ta 80 Chapter 7 Using an Accelerator nicis ciectcceceaceste retten che ceste Hiactsseceaceate ciactasechecestechach
36. For example asm call MyFunc memory This asm statement contains a memory flag because it contains a call The callee may otherwise clobber registers in use by the caller without the memory flag PGI Compiler User s Guide 180 C C Inline Assembly and Intrinsics The following function uses extended asm and the cc flag to compute a power of 2 that is less than or equal to the input parameter n pragma noinline int asmDivideConquer int n int ax 0 amne lope Lp asm LogLoop n emos lms jnle Donen Uime Sopa Wackel SE c alint jmp LogLoopn Done n dec 0n r ax W EqeN bx 2 Ui ge n 8 Mee A return ax The cc flag is used because the asm statement contains some control flow that may alter the control code register The pragma noinline statement prevents the compiler from inlining the asmDivideConquer function If the compiler inlines asmDivideConquer then it may illegally duplicate the labels LogLoop and Done in the generated assembly 15 2 4 Additional Constraints Operand constraints can be divided into four main categories gt Simple Constraints gt Machine Constraints gt Multiple Alternative Constraints gt Constraint Modifiers 15 2 5 Simple Constraints The simplest kind of constraint is a string of letters or characters known as Simple Constraints such as the r and m constraints introduced in Output Operands Table 35 describes these
37. MPI applications with some MPI distributions more convenient by adding the MPI include and library directories to the compiler s include and library search paths The compiler determines the location of these directories using various mechanisms described in the MPI distribution specific sections later in this section Table 15 lists the sub options supported by Mmpi Table 15 MPI Distribution Options This MPI implementation Requires compiling and linking with this option MPICH1 Deprecated Mmpi mpichl MPICH2 Deprecated Mmpi mpich2 MPICH v3 Mmpi mpich For distributions of MPI that are not supported by the Mmpi compiler option use the MPI distribution supplied compiler wrappers mpicc mpic mpif77 or mpif90 to compile and link 6 3 Debugging MPI Applications The PGI debugger PGDBG provides support for symbolic debugging of MPI applications The number and location of processes that can be debugged is limited by your license PGI Workstation licenses limit processes to a single system whereas PGI CDK licenses support general development on clusters For all distributions of MPI except MPICHI support for which has been deprecated and SGI MPI you can initiate an MPI debugging session from either the command line or from within PGDBG For MPICHI and SGI MPI debugging must be initiated at the command line For specific information on how to initiate a debugging session for a particular version of MPI refer to the PG
38. O The PGI runtime libraries make use of several additional system library routines On 64 bit Linux systems the system library routines that PGI supports include these aio error aio write pthread mutex init sleep aio read calloc pthread mutex lock aio return getrlimit pthread mutex unlock aio suspend pthread attr init setrlimit On 32 bit Linux systems the system library routines that PGI supports include these aio error aio suspend getrlimit sleep aio read aio write pthread attr init aio return calloc setrlimit 10 3 Creating and Using Shared Object Files on Linux All of the PGI Fortran C and C compilers support creation of shared object files Unlike statically linked object and library files shared object files link and resolve references with an executable at runtime via a dynamic linker supplied with your operating system The PGI compilers must generate position independent code to support creation of shared objects by the linker However this is not the default You must create object files with position independent code and shared object files that will include them 10 3 1 Procedure to create a use a shared object file The following steps describe how to create and use a shared object file 1 Create an object file with position independent code To do this compile your code with the appropriate PGI compiler using the pic option or one of the equivalent options such as PIC Kpic and KPIC which are
39. POR ERR A EUER XA a iaaa REPOS 191 PGI Compiler User s Guide ix Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15 Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Table 22 Table 23 Table 24 LIST OF TABLES PGI Compilers and Commands uiii intere tn eee eei Pih eL de Pere es E Rege fe nn XV Option Descriptions ccccccccecsecesseceseesesecscsesceseseseescseeceaesesecsecseseaeseeacseeeeseacseensaeeceaeassusseeeeseatsneasaeeeeseatsneesaeeeeaees 6 Examples of Usine siterc and User rc Files sse 14 Typical fast and fastsse ODptiOris re err terr e E e re E ERE ieee 20 Additional fast and fastsse Options ioo tte imn t t ee ec cm n a rat 20 Commonly Used Command Line Options seen terns 22 Example of Effect of Code Unrolling ssssseeeneennetn nennen ener 32 MVECL SUDOPUONS imina e R ERU orem aaa aaa Tae uu san cdd etc 33 Meonc r SUbODlIOTIS cocotte eere credet c PURO en Aen anata EE RUE ERR ERR 37 Optimization and O g and M opt Options sssssssssssseeeeennne tenente ttes 48 Directive and Pragma Summary Table sssssssseseeeeenenene tentent tentent tentent tenete nnns 63 Directive and Pragma Summary Table sssssessseeeeee ener 64 Runtime Library Routines Summa
40. Properties then select C C General then Code Analysis uncheck the top level categories to deselect everything The PGI plugin follows the same rules for creating building and running a project as any other compiler supported by Eclipse For more information refer to Eclipse documentation and tutorials at http www eclipse org documentation PGI Compiler User s Guide 108 Chapter 9 USING DIRECTIVES AND PRAGMAS It is often useful to be able to alter the effects of certain command line options or default behavior of the compiler Fortran directives and C C pragmas provide pragmatic information that control the actions of the compiler in a particular portion of a program without affecting the program as a whole That is while a command line option affects the entire source file that is being compiled directives and pragmas apply or disable the effects of a command line option to selected subprograms or to selected loops in the source file for example to optimize a specific area of code Use directives and pragmas to tune selected routines or loops 9 1 PGI Proprietary Fortran Directives PGI Fortran compilers support proprietary directives that may have any of the following forms pgi g directive pgi r directive pgi l1 directive pgi directive Li If the input is in fixed format the comment character must begin in column 1 and either or C is allowed in place of The scope indicator controls the scope of the
41. addition you specify Minfo and your code contains loops that can be vectorized the compiler reports relevant information on the optimizations applied When a PGI compiler command is invoked with the Mvect option the vectorizer scans code searching for loops that are candidates for high level transformations such as loop distribution loop interchange cache tiling and idiom recognition replacement of a recognizable code sequence such as a reduction loop with optimized code sequences or function calls When the vectorizer finds vectorization opportunities it internally rearranges or replaces sections of loops the vectorizer changes the code generated your source code s loops are not altered In addition to performing these loop transformations the vectorizer produces extensive data dependence information for use by other phases of compilation and detects opportunities to use vector or packed Streaming SIMD Extensions SSE instructions on processors where these are supported The Mvect option can speed up code which contains well behaved countable loops which operate on large REAL REAL 4 REAL 8 INTEGER INTEGER 4 COMPLEX 4 or COMPLEX 8 arrays in Fortran and their C C counterparts However it is possible that some codes will show a decrease in performance when compiled with the Mvect option due to the generation of conditionally executed code segments inability to determine data alignment and other code generation factors F
42. and parallel for Enables you to specify which loops the compiler should parallelize PARALLEL END PARALLEL and Supports a fork join execution model in which a single thread executes all statements parallel until a parallel region is encountered PARALLEL SECTIONS and parallel Defines a non iterative work sharing construct without the need to define an enclosing sections parallel region PARALLEL WORKSHARE END Provides a short form method for including a WORKSHARE directive inside a PARALLEL WORKSHARE PARALLEL construct SECTIONS END SECTIONS and Defines a non iterative work sharing construct within a parallel region sections SINGLE END SINGLE and single Designates code that executes on a single thread and that is skipped by the other threads TASK andtask o and task Defines an explicittask 0 an Defines an explicittask 0 task EE and taskyield Specifies a scheduling point for a task where the currently executing task may be yielded and a different deferred task may be executed PGI Compiler User s Guide 63 Fortran Directive and C Pragma Description Using OpenMP TASKWAIT and taskwait Specifies a wait on the completion of child tasks generated since the beginning of the current task THREADPRIVATE and threadprivate When a common block or variable that is initialized appears in this directive or pragma each thread s copy is initialized once prior to its first use WORKSHARE END WO
43. are combined using the reduction operator and Parallel the result combined with the value of the original variable and stored in the original variable self list Update Specifies that the items in the list are to be copied from the accelerator device memory to the local memory The se 1 f clause is a synonym for the host clause seq Loop Tells the compiler to execute this loop sequentially on the accelerator There is no maximum number of iterations for a seq schedule shortloop Loop If the loop has the vector clause this tells the compiler that the trip count for the loop is less than or equal to the number of vector lanes created for this kernel as specified by the vector length on the parallel construct or vector clause in a kernels construct PGI extension use device list Host Data Tells the compiler to use the device address of any item in the list in E code within the construct vector length int expr Loop Tells the compiler to execute this loop in vector or SIMD mode on the accelerator vector length int expr Parallel Defines the number of vector lanes that will be active after a worker transitions from vector single mode to vector partitioned mode wait int expr list The compute data or update operation may not be launched or executed until all operations enqueued up to this point by this thread on the associated asynchronous device activity queues have completed Parallel Kernels Enter Data Exit Dat
44. asm addl 1 0 n aU x Vo uw YB Compiling with the tp p7 switch chooses a 32 bit target 9 pgcc examplel2 c PTIC c tp p7 PGC S 0354 Can t find a register in class BREG for extended ASM operand 1 examplel2 c 3 PGC x86 Linux x86 Rel Dev compilation completed with severe errors 15 2 7 Multiple Alternative Constraints Sometimes a single instruction can take a variety of operand types For example the x86 permits register to memory and memory to register operations To allow this flexibility in inline assembly use multiple alternative constraints An alternative is a series of constraints for each operand To specify multiple alternatives separate each alternative with a comma Table 37 Multiple Alternative Constraints Constraint Description Y Separates each alternative for a particular operand NEENL MEMMMMMMMNNNN PGI Compiler User s Guide 184 C C Inline Assembly and Intrinsics The following example uses multiple alternatives for an add operation void examplel3 int x 1 int y 1 asm addl 1 0 n Worl CON ox Meld Gel uu Jg The preceding example13 has two alternatives for each operand ab cd for the output operand and db cam for the input operand Each operand must have the same number of alternatives however each alternative can have any number of constraints for example the output operand in example 13 has two constraints for its seco
45. code to execute the iterations in parallel without synchronization PGI Compiler User s Guide 96 Using an Accelerator Use this clause In these directives To do this num_gangs int expr Parallel Defines the number of parallel gangs that will execute the region num_workers int expr Defines the number of workers within each gang that will be active after a gang transitions from worker single mode to worker partitioned mode Parallel present list Tells the implementation that the items in the list are already present Parallel in device memory Kernels Data Declare private list Specifies that a copy of each item in the list will be created for each thread that executes one or more iterations of the associated loop or loops private list Parallel Declares that a copy of each item on the list will be created for each parallel gang reduction operator list For each variable in the list a private copy is created for each thread that executes iterations of the associated loop or loops and initialized for the operator At the end of the loop the values for each thread are combined using the reduction operator and the result combined with the value of the original variable and stored in the original variable reduction operator list For each variable in the list a private copy is created for each parallel gang and initialized for the operator At the end of the region the values for each gang
46. directive This indicator occurs after the Some directives ignore the scope indicator The valid scopes shown in the previous forms of the directive are these global indicates the directive applies to the end of the source file routine indicates the directive applies to the next subprogram loop indicates the directive applies to the next loop but not to any loop contained within the loop body Loop scoped directives are only applied to DO loops blank indicates that the default scope for the directive is applied PGI Compiler User s Guide 109 Using Directives and Pragmas The body of the directive may immediately follow the scope indicator Alternatively any number of blanks may precede the name of the directive Any names in the body of the directive including the directive name may not contain embedded blanks Blanks may surround any special characters such as a comma or an equal sign The directive name including the directive prefix may contain upper or lower case letters and the case is not significant Case is significant for any variable names that appear in the body of the directive if the command line option Mupcase is selected For compatibility with other vendors directives the prefix cpgi may be substituted with cdirS or cvdS 9 2 PGI Proprietary C and C Pragmas Pragmas may be supplied in a C C source file to provide information to the compiler Many pragmas have a corresponding command lin
47. encountered during execution Task construct A task directive plus a structured block Task region The dynamic sequence of instructions produced by the execution of a task by a thread 5 3 Fortran Parallelization Directives Parallelization directives are comments in a program that are interpreted by the PGI Fortran compilers when the option mp is specified on the command line The form of a parallelization directive is sentinel directive name clauses With the exception of the SGI compatible DOACROSS directive the sentinel must comply with these rules gt Be one of these 5OMP CSOMP or OMP gt Must start in column 1 one gt Must appear as a single word without embedded white space gt The sentinel marking a DOACROSS directive is C The directive name can be any of the directives listed in Directive and Pragma Summary Table The valid clauses depend on the directive Directive and Pragma Clauses provides a list of clauses the directives and pragmas to which they apply and their functionality In addition to the sentinel rules the directive must also comply with these rules gt Standard Fortran syntax restrictions such as line length case insensitivity and so on apply to the directive line gt Initial directive lines must have a space or zero in column six gt Continuation directive lines must have a character other than a space or a zero in column six Continuation lines for CSDOACROSS directi
48. exists and contains the value format_relaxed an I O item corresponding to a numerical edit descriptor such as F E I and so on is not required to be a type implied by the descriptor gt JInanon Windows environment if FORTRANOPT exists and contains the value crif a sequential formatted or list directed record is allowed to be terminated with the character sequence r n carriage return newline This approach is useful when reading records from a file produced on a Window s system The following example causes the PGI Fortran compilers to use VAX I O conventions setenv FORTRANOPT vaxio 11 3 3 GMON OUT PREFIX GMON OUT PREFIX specifies the name of the output file for programs that are compiled and linked with the pg option The default name is gmon out If GHON OUT PREFIX is set the name of the output file has GMON OUT PREFIX as a prefix Further the suffix is the pid of the running process The prefix and suffix are separated by a dot For example if the output file is mygmon then the full filename may look something similar to this mnygmon 0012348567 The following example causes the PGI Fortran compilers to use pgout as the output file for programs compiled and linked with the pg option setenv GMON OUT PREFIX pgout 11 3 4 LD_LIBRARY_PATH The LD LIBRARY PATH variable is a colon separated set of directories specifying where libraries should first be searched prior to searchin
49. expression pairs that specify what C values are needed by the asm statement The input constraints specify how the data is delivered to the asm statement For example r x says that the input operand is a register that has a copy of the value stored in C variable x Another example is m x which says that the input item is the memory location associated with variable x Other constraint types are discussed in Additional Constraints An example follows void example5 int x 1 int y 2 int z2 3 ewewn Wevelolil 32 Sua Wevelolil 32 GOW Ware ex aaa iw 2 Wiel qe yp The previous example adds variable z item 2 to variable x and variable y The resulting values for x and y are 4 and 5 respectively PGI Compiler User s Guide 178 C C Inline Assembly and Intrinsics Another type of input constraint worth mentioning here is the matching constraint A matching constraint is used to specify an operand that fills both an input as well as an output role An example follows int x 1 void example6 PE Wevelelb Sil sil ac a OH px g The previous example is equivalent to the example2 function shown in Output Operands The constraint expression pair 0 x tells the compiler to initialize output item 0 with variable x at the beginning of the asm statement The resulting value for x is 2 Also note that 267 in the asm string means the same thing as 0 in this case That is because there is
50. following commands SE JgoOSNECNSOCGONI SUIS S Perr CAO marn O CACO Evine Cc Executing the resulting a out file should produce the following output main a 8 b 2 ptr c Oxbffffb94 func a 8b 2 ptr c bffffb94 func res 4 main res 4 You cannot use the extern C form of declaration for an object s member functions 13 8 5 Example Fortran Calling C The Fortran main program shown in Fortran Main Program f2cp_main f calling a C function calls the C function shown in C function f2cp_func C Notice gt Each argument is defined as a pointer in the C function since Fortran passes by reference gt The C function name uses all lower case and a trailing _ Fortran Main Program f2cp_main f calling a C function Jbexentrereudb db Texoxeubd character letterl integer 4 numintl numint2 real numfloatl double precision numdoubl integer 2 numshorl external f2cpfunc call epe ooI M lerccsri mumimwel numint2 numfloatl numdoub1 numshor1 wedge Cu ibe AZ r5 15 2 WG Eos 15 tT OOO ete mil numinicle a MEENA Mael oew numdoubl numshorl end PGI Compiler User s Guide 162 Inter language Calling C function f2cp_func C define TRUE Oxff define FALSE 0 extern C externi od dete MN mcus TJoxli lcs inte numinel numines flea Mm esito diis double numdoubl SIMONA neat lr ime lem lecteri booll TRUE letterl v numintl 1
51. following sections provide a quick overview of the menu selections To access the main PGI menu from the Start menu select Start All Programs PGI Workstation PGI Compiler User s Guide 9 Getting Started Command Shell Submenus From the PGI Workstation menu you have access to PGI command shells for each version of PGI installed on your system For example if you have both PGI 14 1 and PGI 13 9 installed then you have a submenu for each of these versions The PGI submenus for each version include the following gt PGI Bash 64 Select this option to launch a Cygwin bash shell in which the environment is pre initialized to use the 64 bit PGI compilers and tools The default environment variables are already set and available Available only on x64 systems with Cygwin installed gt PGI Bash Select this option to launch a Cygwin bash shell in which the environment is pre initialized to use the 32 bit PGI compilers and tools The default environment variables are already set and available Available only on systems with Cygwin installed gt PGI Cmd 64 Select this option to launch a Microsoft command shell in which the environment is pre initialized to use the 64 bit PGI compilers and tools The default environment variables are already set and available Available only on x64 systems gt PGI Cmd Select this option to launch a Microsoft command shell in which the environment is pre initialized to use the 32 bi
52. h asm posix types h and others Also PGI s version of stdarg h supports changes in newer versions of Linux If you are using the PGI C or C compilers please make sure that the supplied versions of these include files are found before the system versions This hierarchy happens by default unless you explicitly add a I option that references one of the system include directories Running Parallel Programs on Linux You may encounter difficulties running auto parallel or OpenMP programs on Linux systems when the per thread stack size is set to the default 2MB If you have unexplained failures please try setting the environment variable OMP STACKSIZE to a larger value such as 8MB For information on setting environment variables refer to Setting Environment Variables If your program is still failing you may be encountering the hard 8 MB limit on main process stack sizes in Linux You can work around the problem by issuing the following command In csh limit stacksize unlimited In bash sh zsh or ksh use ulimit s unlimited 1 7 2 Using the PGI Compilers on Windows LI The PGI C compiler for Windows is deprecated and will no longer be available as of the PGI 16 1 release PGI on the Windows Start Menu PGI provides a Start menu entry that provides access to different versions of PGI command shells as well as easy access to the PGI Debugger the PGI Profiler documentation and licensing The
53. how to use a product and often then find that they need more information and facts about specific areas of the product Knowing how as well as why you might use certain options or perform certain tasks is key to using the PGI compilers and tools effectively and efficiently However once you have this knowledge and understanding you very likely might find yourself wanting to know much more about specific areas or specific topics This book contains the essential information on how to use the compiler and is divided into these sections Getting Started provides an introduction to the PGI compilers and describes their use and overall features Use Command Line Options provides an overview of the command line options as well as task related lists of options Optimizing and Parallelizing describes standard optimization techniques that with little effort allow users to significantly improve the performance of programs Using Function Inlining describes how to use function inlining and shows how to create an inline library Using OpenMP provides a description of the OpenMP Fortran parallelization directives and of the OpenMP C and C parallelization pragmas and shows examples of their use Using MPI describes how to use MPI with PGI Workstation and PGI server PGI Compiler User s Guide Xiii Preface Using an Accelerator describes how to use the PGI Accelerator compilers Eclipse describes how to use the PGI C C compilers from withi
54. in the default output file a out pgfortran Minline name proc lib lib il myprog f The following command line is equivalent to the preceding line with the exception that in the following example does not use the keywords name and 1ib You typically use keywords to avoid name conflicts when you use an inline library name that does not contain a period Otherwise without the keywords a period informs the compiler that the file on the command line is an inline library pgfortran Minline proc lib il myprog f 4 3 Creating an Inline Library You can create or update an inline library using the Mextract command line option If you do not specify selection criteria with the Mextract option the compiler attempts to extract all subprograms Several Mextract options let you determine the selection criteria for creating or updating an inline library These selection criteria include func Extracts the function func you can use a comma separated list to specify multiple functions name func Extracts the functions whose name matches func a function in the source text size Limits the size of the extracted functions to functions with a statement count less than or equal to n the specified size The size n may not exactly equal the number of statements in a selected function the size parameter is merely a rough gauge lib Fext 1ib Stores the extracted information in the library directory ext 1ib If no inline library is s
55. interaction between the O level g and M lt opt gt options In the table level can be 0 1 2 3 or 4 and opt can be vect concur unroll or ipa The default optimization level is dependent upon these command line options Table 10 Optimization and O g and M opt Options Optimize Option Debug Option M opt Option Optimization Level e CNN CNN pe m S em ESL ES ee mm T Code that is not optimized yet compiled using the option 00 can be significantly slower than code generated at other optimization levels The M opt option where opt is vect concur unroll or ipa sets the optimization level to 2 if no O options are supplied The fast option sets the optimization level to a target dependent optimization level if no O options are supplied 3 12 Local Optimization Using Directives and Pragmas Command line options let you specify optimizations for an entire source file Directives supplied within a Fortran source file and pragmas supplied within a C or C source file provide information to the compiler and alter the effects of certain command line options or the default behavior of the compiler Many directives have a corresponding command line option While a command line option affects the entire source file that is being compiled directives and pragmas let you do the following gt A Apply or disable the effects of a particular command line option to selected subprograms or to selected lo
56. line options and provides a brief overview of a few of the more common options For a complete list of command line options their descriptions and use refer to the Command Line Options Reference section of the PGI Compiler s Reference Guide 2 1 Command Line Option Overview Before looking at all the command line options first become familiar with the syntax for these options There are a large number of options available to you yet most users only use a few of them So start simple and progress into using the more advanced options By default the PGI compilers generate code that is optimized for the type of processor on which compilation is performed the compilation host Before adding options to your command line review Help with Command line Options and Frequently used Options 2 1 1 Command line Options Syntax On a command line options need to be preceded by a hyphen If the compiler does not recognize an option you get an unknown switch error The error can be downgraded to a warning by adding the noswitcherror option This document uses the following notation when describing options 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 I separates the choices PGI Compiler User s Guide 17 Use Command Line Options Horizontal ellipses indicate that zero or more instances of the precedi
57. link the first DLL the import library for the second DLL must be available Usually an import library is created when a DLL is linked In this case however the second DLL cannot be linked without the import library for the first DLL When such circular imports exist an import library for one of the DLLs must be created in a separate step without creating the DLL The PGI drivers call the Microsoft 1ib tool to create import libraries in this situation Once the DLLs are built we can use them to build the main program object2 f95 subroutine func 2a external func 3b DECS ATTRIBUTES DLLEXPORT func 2a DEC ATTRIBUTES DLLIMPORT func 3b peime Mitauec 2a Calling a routine im goJo CULLY call func 3b end subroutine subroutine func 2b IDECS ATTRIBUTES DLLEXPORT func 2b joxeiusn p Mizu Zo end subroutine object3 f95 subroutine func 3a external func 2b DEC ATTRIBUTES DLLEXPORT func 3a DECS ATTRIBUTES DLLIMPORT func 2b joexgsu 9 Ufruwae Sei Calling a routine im 91972 cll iL Call fune Zo end subroutine subroutine func 3b DECS ATTRIBUTES DLLEXPORT func 3b pirme MENS Suo end subroutine prog2 f95 program prog2 external func 2a external func 3a DECS ATTRIBUTES DLLIMPORT func 2a DECS ATTRIBUTES DLLIMPORT func 3a call func 2a Celi sewuae Se end program
58. local optimization vectorization and scheduling of instructions Performance for loops on systems with multiple processors may also improve using the parallelization features of the PGI compilers 3 1 4 Interprocedural Analysis IPA and Optimization Interprocedural analysis IPA allows use of information across function call boundaries to perform optimizations that would otherwise be unavailable For example if the actual argument to a function is in fact a constant in the caller it may be possible to propagate that constant into the callee and perform optimizations that are not valid if the dummy argument is treated as a variable A wide range of optimizations are enabled or improved by using IPA including but not limited to data alignment optimizations argument removal constant propagation pointer disambiguation pure function detection F90 F95 array shape propagation data placement PGI Compiler User s Guide 25 Optimizing and Parallelizing vestigial function removal automatic function inlining inlining of functions from pre compiled libraries and interprocedural optimization of functions from pre compiled libraries 3 1 5 Function Inlining This optimization allows a call to a function to be replaced by a copy of the body of that function This optimization will sometimes speed up execution by eliminating the function call and return overhead Function inlining may also create opportunities for other types of optimization
59. m c m return 0 9 pgcc mcmodel medium o bigadd bigadd c When SIZE is greater than 2G 4 and the arrays are of type float with 4 bytes per element the size of each array is greater than 2GB With pgcc using the mcmodelzmedium switch a static data object can now be 2GB in size If you execute with these settings in your environment you may see the following bigadd Segmentation fault PGI Compiler User s Guide 172 Programming Considerations for 64 Bit Environments Execution fails because the stack size is not large enough You can most likely correct this error by using the limit stacksize command to reset the stack size in your environment limit stacksize 3000M The command limit stacksize unlimited probably does not provide as large a stack as we are using in the this example bigadd a 0 1 b 0 2 c 0 3 n 599990000 a 599990000 5 9999e 08 b 599990000 1 19998e 09 c 599990000 1 79997e 09 14 8 Medium Memory Model and Large Array in Fortran The following example works with the PGFORTRAN PGF95 and PGF77 compilers included in Release 2015 Both compilers use 64 bit addresses and index arithmetic when the mcmodel medium option is used Consider the following example Medium Memory Model and Large Array in Fortran 9 picari nace program mat integer i j k size 1 m n parameter size 16000 gt 2GB parameter m size n size eel El Gin ia rotar ia ret Cap pel
60. modifier include oma lib kinds h integer kind omp sched kind kind integer modifier double omp set schedule omp get schedule PGI Compiler User s Guide 70 Using OpenMP Runtime Library Routines with Examples Retrieve the value of the run sched var subroutine omp get schedule kind modifier include omp lib kinds h integer kind omp sched kind kind integer modifier C C double omp get schedule omp get wtime Returns the elapsed wall clock time in seconds as a DOUBLE PRECISION value for directives and as a floating point double value for pragmas Times returned are per thread times and are not necessarily globally consistent across all threads Fortran double precision function omp get wtime C C double omp get wtime void omp_get_wtick Returns the resolution of omp_get_wtime in seconds as a DOUBLE PRECISION value for Fortran directives and as a floating point double value for C C pragmas Fortran double precision function omp get wtick C C double omp get wtick omp_init_lock Initializes a lock associated with the variable lock for use in subsequent calls to lock routines The initial state of the lock is unlocked If the variable is already associated with a lock it is illegal to make a call to this routine Fortran subroutine omp init lock lock include omp lib kinds h integer kind omp lock kind lock void omp init lock omp lock t lock void
61. movl 0 eax movl eax x rip lineno 0 popq rbp rer In the generated assembly shown notice that the compiler generated two statements for the asm statement at line number 5 The compiler generated movl 0 eax from the asm string Also notice that eax appears in place of 0 because the compiler assigned the eax register to variable x Since item 0 is an output operand the result must be stored in its expression x In addition to write only output operands there are read write output operands designated with a instead of a For example r x tells the compiler to initialize the output operand with variable x at the beginning of the asm statement To illustrate this point the following example increments variable x by 1 int x 1 void example2 eum Wacklil Sil 49 te o Y To perform the increment the output operand must be initialized with variable x The read write constraint modifier instructs the compiler to initialize the output operand with its expression The compiler generates the following assembly code for the example2 function example2 uestis pushq rbp DCENOR movq rsp rbp 5 odbIeaeab dlc ENIE lineno 5 movl x rip eax addl 1 eax movl eax x rip 14 lineno 0 popq rbp EXE From the example2 code two extraneous moves are generated in the assembly one movl for initializing the output register and a second movl to write it to
62. objects in a running program TS 14 6 Practical Limitations of Large Array Programming The 64 bit addressing capability of the Linux86 64 and Win64 environments can cause unexpected issues when data sizes are enlarged significantly The following table describes the most common occurrences of practical limitations of large array programming Table 34 64 Bit Limitations array initialization Initializing a large array with a data statement may result in very large assembly and object files where a line of assembler source is required for each element in the initialized array Compilation and linking can be very time consuming as well To avoid this issue consider initializing large arrays in a loop at runtime rather than in a data statement stack space Stack space can be a problem for data that is stack based In Win64 stack space can be increased by using this link time switch where N is the desired stack size W1 stack N In linux86 64 stack size is increased in the environment Setting stacksize to unlimited often is not large enough limiet stecksize new size l im es ulimit s new size in bash page swapping If your executable is much larger than the physical size of memory page swapping can cause it to run dramatically slower it may even fail This is not a compiler problem Try smaller data sets to determine whether or not a problem is due to page thrashing configured space Be sure your linux86 64 system is configured
63. omp init nest lock omp nest lock t lock omp destroy lock Disassociates a lock associated with the variable Fortran subroutine omp destroy lock lock include omp lib kinds h integer kind omp lock kind lock C C void omp destroy lock omp lock t lock void omp destroy nest lock omp nest lock t lock omp set lock Causes the calling thread to wait until the specified lock is available PGI Compiler User s Guide 71 Using OpenMP Runtime Library Routines with Examples The thread gains ownership of the lock when it is available If the variable is not already associated with a lock it is illegal to make a call to this routine Fortran subroutine omp set lock lock include omp lib kinds h integer kind omp lock kind lock C C wosdEompbESct e oc Opa o c Eee c void omp set nest lock omp nest lock t lock omp unset lock Causes the calling thread to release ownership of the lock associated with integer var If the variable is not already associated with a lock it is illegal to make a call to this routine Fortran subroutine omp unset lock lock include omp lib kinds h integer kind omp lock kind lock C C include lt omp h gt void omp unset lock omp lock t lock void omp unset nest lock omp nest lock t lock omp test lock Causes the calling thread to try to gain ownership of the lock associated with the variable The function retums TRUE for directives and non zero for pragmas
64. or C function returns a value call it from Fortran as a function gt When aC or C function does not return a value call it as a subroutine For a C C program calling a Fortran function the call should return a similar type Table 29 Fortran and C C Data Type Compatibility lists compatible types If the call is to a Fortran subroutine a Fortran CHARACTER function or a Fortran COMPLEX function call it from C C as a function that returns void The exception to this convention is when a Fortran subroutine has alternate returns call such a subroutine from C C as a function returning int whose value is the value of the integer expression specified in the alternate RETURN statement PGI Compiler User s Guide 154 Inter language Calling 13 4 Upper and Lower Case Conventions Underscores By default on Linux Win64 and OSX systems all Fortran symbol names are converted to lower case C and C are case sensitive so upper case function names stay upper case When you use inter language calling you can either name your C C functions with lower case names or invoke the Fortran compiler command with the option Mupcase in which case it will not convert symbol names to lower case When programs are compiled using one of the PGI Fortran compilers on Linux Win64 and OSX systems an underscore is appended to Fortran global names names of functions subroutines and common blocks This mechanism distinguishes Fortran name s
65. processor based systems The PGI Compiler User s Guide provides operating instructions for the PGI command level development environment The PGI Compiler Reference Manual contains details concerning the PGI compilers interpretation of the Fortran language implementation of Fortran language extensions and command level compilation Users are expected to have previous experience with or knowledge of the Fortran programming language Neither guide teaches the Fortran programming language Audience Description This manual is intended for scientists and engineers using the PGI 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 assembly language in the software development process and you should have some level of understanding of programming The PGI compilers are available on a variety of x86 or x64 hardware 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 this PGI product For information on installing PGI compilers and tools refer to the Release Notes and Installation Guide included with your software For further information refer to the following gt American National Standard Programming Language FORTRAN ANSI X3 1978 1978 gt ISO IEC 1539 1 1991 Information te
66. referenced in the executable along with the pathname of the directory from which they will be extracted 9 ldd myprog If the pathname is not hard coded using the R option and if LD LIBRARY PATH is not initialized the pathname is listed as not found For more information on 1dd its options and usage see the online man page for 1dd 10 4 Creating and Using Dynamic Libraries on Mac OS X PGI compilers for Mac OS X do not support static linking on user executables Apple only ships dynamic versions of its system libraries not static versions You can create static libraries however you cannot create 100 static executables The 32 bit version of PGI Workstation for Mac OS X supports generation of dynamic libraries To create the dynamic library you use the Hdynamiclib switch to invoke the libtool utility program provided by Mac OS X For more information refer to the 1ibtool man page The following example creates and uses a dynamic library 1 Create the object files world f90 subroutine world print Hello World end hello f90 program hello call world end 2 Build the dynamic library 9 pgfortran dynamiclib world f90 o world dylib 3 Build the program that uses the dynamic library pgfortran hello f90 world dylib o hello 6 4 Run the program hello Hello World 10 5 PGI Runtime Libraries on Windows Both statically and dynamically linked library DLL versi
67. register name for an operand For example if operand 0 is in register a then w0 will produce ax z Produces an op code suffix based on the size of an operand For example b for byte w for half word I for word and q for quad word Yt C 96D F 0 X f Yh l Yon s y are not supported These modifiers begin with either a backslash or a percent The modifiers that begin with a backslash V e g n have the same effect as they do ina printf format string The modifiers that are preceded with a are used to modify a particular operand These modifiers begin with either a backslash or a percent For example b0 means produce the byte or 8 bit version of operand 0 If operand 0 is a register it will produce a byte register such as al bl cl and so on Consider this example void examplel9 ame Gl Ex JLB int p amp a asm add zO Sql a0 amp p p sel a giu p Ve On an x86 target the compiler produces the following instruction for the asm string shown in the preceding example addl ecx eax The z0 modifier produced an T lower case L suffix because the size of pointer p is 32 bits on x86 The q1 modifier produced the word register name for variable a The a0 instructs the compiler to add parentheses around operand 0 hence eax On an x86 64 target the compiler produces the following instruction for the asm string shown in
68. since the overhead in setting up and starting a parallel loop will likely outweigh the potential benefits In addition the default is not to parallelize innermost loops since these often by definition are vectorizable using SSE instructions and it is seldom profitable to both vectorize and parallelize the same loop especially on multi core processors Compiler switches and directives are available to let you override most of these restrictions on auto parallelization 3 7 1 Auto Parallelization Sub options The parallelizer performs various operations that can be controlled by arguments to the Mconcur command line option The following sections describe these arguments that affect the operation of the parallelizer In addition these parallelizer operations can be controlled from within code using directives and pragmas For details on the use of directives and pragmas refer to Using Directives and Pragmas By default Mconcur without any sub options is equivalent to Mconcur dist block This enables parallelization of loops with blocked iteration allocation across the available threads of execution These defaults may vary depending on the target system The following table lists and briefly describes some of the Mconcur suboptions Table 9 Mconcur Suboptions Use this option To instruct the parallelizer to do this Use all available cores Specify this option at link time Mconcur no altcode Generate do not generate
69. some common development tasks that you might wish to perform When you compile code you can specify a number of options on the command line that define specific characteristics related to how the program is compiled and linked typically enhancing or overriding the default behavior of the compiler For a list of the most common command line options and information on all the command line options refer to Use Command Line Options Code optimization and parallelization allows the compiler to organize your code for efficient execution While possibly increasing compilation time and making the code more difficult to debug these techniques typically produce code that runs significantly faster than code that does not use them For more information on optimization and parallelization refer to Optimizing and Parallelizing Function inlining a special type of optimization replaces a call to a function or a subroutine with the body of the function or subroutine This process can speed up execution by eliminating parameter passing and the function or subroutine call and return overhead In addition function inlining allows the compiler to optimize the function with the rest of the code However function inlining may also result in much larger code size with no increase in execution speed For more information on function inlining refer to Using Function Inlining Directives and pragmas allow users to place hints in the source code to help the compile
70. sss teens 132 10 8 WSINGUEIBSE oco nee ciet rte emo e te t tr etu Ote EUR reU UR UM EIE Perte etr e ED 133 10 9 EAPACK BIAS and FF TS scant sr th tan dace recette pa cn dtc en t vine 133 10 10 Linking with ScaEAPACK s iicet re eth ee rr n i tr ba e pda 133 10 11 The C Standard Template Library nennen 134 PGI Compiler User s Guide vi Chapter 14 Using Environment Variables cdsiscccsctcccccseuceuttcehcesccuterattcbectetieuceuceaehctes mettere tix ie utet sere acess de 135 11 1 Setting Environment Variables ssseseeeeeenennenene tenente nnne terrens 135 11 1 1 Setting Environment Variables on Linux sss 135 11 1 2 Setting Environment Variables on Windows sssseeeeeeennetnnen nnns 136 11 1 3 Setting Environment Variables on Mac OSX sse nentes 136 11 2 PGl Related Environment Variables nter cnt reet iEn prd i e Mee e Le 137 11 3 PGI Environment Variables sse tenente tete tentent nter tenen 138 11 3 1 EEEXEM BATOR Side eder trot re tet torrente de ret deren teritur Bey ett pr tet nena 138 TES 22 EORERANOPT 2t itta Eom intet dine de etre inco tL 139 Uu uclo mot i C 139 QUIBAS 139 11 95 EMEBICENSE FILE eret eme a ee ee a ata ede ea atn 139 11 9 6 MANPATLEL odere Ro pne fe eet mete mee eene Fei pu Bee de nu eb
71. suffix for this operand For example LO produces l on x86 and x86 64 0 0 0 0 Produces a quad word op code suffix for this operand if it is supported by the target Otherwise it produces a word op code suffix For example Q0 produces q on x86_64 and l on x86 JoP If producing Position Independent Code PIC the compiler adds the PIC suffix for this operand For example P0 produces PLT on x86 and x86 64 90S Produces s suffix for this operand For example 9650 produces s on x86 and x86 64 Produces t suffix for this operand For example 9650 produces t on x86 and x86 64 Produces the half word op code suffix for this operand For example 96 WO produces w on x86 and x86 64 Adds open and close parentheses around the operand b Produces the byte register name for an operand For example if operand 0 is in register a then b0 will produce al Cuts the character from an immediate operand k Produces the word register name for an operand For example if operand 0 is in register a then k0 will produce eax PGI Compiler User s Guide 187 C C Inline Assembly and Intrinsics Modifier Description q Produces the quad word register name for an operand if the target supports quad word Otherwise it produces a word register name For example if operand 0 is in register a then q0 produces rax on X86 64 or eax on x86 Yow Produces the half word
72. supported for PGI Compiler User s Guide 121 Creating and Using Libraries compatibility with other systems For example use the following command to create an object file with position independent code using pgfortran 9 pgfortran c fpic tobeshared f 2 Produce a shared object file To do this use the appropriate PGI compiler to invoke the linker supplied with your system It is customary to name such files using a so filename extension On Linux you do this by passing the shared option to the linker o pgfortran shared o tobeshared so tobeshared o Compilation and generation of the shared object can be performed in one step using both the fpi c option and the appropriate option for generation of a shared object file 3 Use a shared object file To do this use the appropriate PGI compiler to compile and link the program which will reference functions or subroutines in the shared object file and list the shared object on the link line as shown here pgfortran o myprog myprog f tobeshared so 4 Make the executable available You now have an executable myprog which does not include any code from functions or subroutines in ctobeshared so but which can be executed and dynamically linked to that code By default when the program is linked to produce myprog no assumptions are made on the location of tobeshared so Therefore for myprog to execute correctly you must initialize the environment variab
73. system refer to Fortran C and C Data Types section of the PGI Compiler s Reference Guide For more information on x86 specific data representation refer to the System V Application Binary Interface Processor Supplement by AT amp T UNIX System Laboratories Inc Prentice Hall Inc For more information on x64 processor based systems and the application binary interface ABI for those systems see http www x86 64 org documentation abi pdf 1 6 Parallel Programming Using the PGI Compilers The PGI compilers support many styles of parallel programming gt Automatic shared memory parallel programs compiled using the Mconcur option to pgf77 pgf95 pgfortran pgcc or pgc Parallel programs of this variety can be run on shared memory parallel SMP systems such as dual core or multi processor workstations gt OpenMP shared memory parallel programs compiled using the mp option to pgf77 pgf95 pgfortran pgcc or pgc Parallel programs of this variety can be run on SMP systems Carefully coded user directed parallel programs using OpenMP directives can often achieve significant speed ups on dual core workstations or large numbers of processors on SMP server systems Using OpenMP contains complete descriptions of user directed parallel programming PGI Compiler User s Guide T Getting Started gt Distributed computing using an MPI message passing library for communication between distributed processes gt Accelerate
74. the filename argument on the command line However the placement of some options is significant such as the 1 option in which the order of the filenames determines the search order cy If two or more options contradict each other the last one in the command line takes precedence 1 3 3 Fortran Directives and C C Pragmas You can insert Fortran directives and C C pragmas in program source code to alter the effects of certain command line options and to control various aspects of the compilation process for a specific routine or a specific program loop For more information on Fortran directives and C C pragmas refer to Using OpenMP and Using Directives and Pragmas 1 4 Filename Conventions The PGI compilers use the filenames that you specify on the command line to find and to create input and output files This section describes the input and output filename conventions for the phases of the compilation process PGI Compiler User s Guide 3 Getting Started 1 4 1 Input Files You can specify assembly language files preprocessed source files Fortran C C source files object files and libraries as inputs on the command line The compiler driver determines the type of each input file by examining the filename extensions For systems with a case insensitive file system use the Mpreprocess option described in Command Line Options Reference section of the PGI Compiler s Reference Manual under the commands for Fort
75. the loop Generated 3 prefetch instructions for the loop Notice the informational message for the loop at line 17 gt The first two lines of the message indicate that the loop was vectorized SSE instructions were generated and four alternate versions of the loop were also generated The loop count and alignments of the arrays determine which of these versions is executed gt The last line of the informational message indicates that prefetch instructions have been generated for three loads to minimize latency of data transfers from main memory Executing again you should see results similar to the following bin time vadd 1 500009 77 7315 OOM Cilmi 00 521 9X9 99199 d 0 60user 0 00system 0 00 61elapsed 99 CPU Oavgtextt0avgdata 3920maxresident k Oinputst Ooutputs 0major 4289minor pagefaults Oswaps The SIMD result is 2 25 times faster than the equivalent non SIMD version of the program Speed up realized by a given loop or program can vary widely based on a number of factors gt When the vectors of data are resident in the data cache performance improvement using vector SSE or SSE2 instructions is most effective gt If data is aligned properly performance will be better in general than when using vector SSE operations on unaligned data gt If the compiler can guarantee that data is aligned properly even more efficient sequences of SSE instructions can be generated gt The efficiency of l
76. the output to a specific set of options or to a building process The syntax for this usage is this help subgroup By using the command pgfortran help help as previously shown we can see output that shows the available subgroups You can use the following command to restrict the output on the help command to information about only the options related to only one group such as debug information generation pgfortran help debug The output you see is similar to this Debugging switches M no bounds Generate code to check array bounds Mchkfpstk Check consistency of floating point stack at subprogram calls SiZ Isabie Omis Mchkstk Check for sufficient stack space upon subprogram entry Mcoff Generate COFF format object Mdwarfl Generate DWARF1 debug information with g Mdwarf2 Generate DWARF2 debug information with g Mdwarf3 Generate DWARF3 debug information with g Melf Generate ELF format object g Generate information for debugger gopt Generate information for debugger without disabling optimizations For a complete description of subgroups refer to the help description in the Command Line Options Reference section of the PGI Compiler Reference Manual PGI Compiler User s Guide 19 Use Command Line Options 2 3 Getting Started with Performance One of the top priorities of most users is performance and optimization This section provides a quick overview of a few of the command line optio
77. the preceding example addq rcx rax The z0 modifier produced a q suffix because the size of pointer p is 64 bit on x86 64 Because x86 64 supports quad word registers the q1 modifier produced the quad word register name rax for variable a PGI Compiler User s Guide 188 C C Inline Assembly and Intrinsics 15 5 Extended Asm Macros As with traditional inline assembly described in Inline Assembly extended asm can be used in a macro For example you can use the following macro to access the runtime stack pointer define GET SP x asm mov sp 0 m x sp void example20 VOLE w sitrack porn een GET SP stack pointer The GET_SP macro assigns the value of the stack pointer to whatever is inserted in its argument for example stack_pointer Another C extension known as statement expressions is used to write the GET_SP macro another way define GET SP2 vore my Stack peri y asi noy sss 20 3 Yau my stack pire E MU Sr MD my stack ptr void example21 woud stack pouncenr CENIO The statement expression allows a body of code to evaluate to a single value This value is specified as the last instruction in the statement expression In this case the value is the result of the asm statement my_stack_ptr By writing an asm macro with a statement expression the asm result may be assigned directly to another variable for example void stack_
78. to control the behavior of the PGDBG debugger or PGPROF profiler For a description of environment variables that affect these tools refer to the PGDBG Debugger Manual and PGPROF Profiler Manual respectively 11 1 Setting Environment Variables Before we look at the environment variables that you might use with the PGI compilers and tools let s take a look at how to set environment variables To illustrate how to set these variables in various environments let s look at how a user might initialize the shell environment prior to using the PGI compilers and tools 11 1 1 Setting Environment Variables on Linux Let s assume that you want access to the PGI products when you log in Let s further assume that you installed the PGI compilers in opt pgi and that the license file isin opt pgi license dat For access at startup you can add the following lines to your startup file In csh use these commands setenv PGI opt pgi setenv MANPATH SMANPATH SPGI linux86 15 10 man setenv LM LICENSE FILE PGI license dat set path PGI linux86 15 10 bin path oe oe oe oe PGI Compiler User s Guide 135 Using Environment Variables In bash sh zsh or ksh use these commands PGI opt pgi export PGI MANPATH MANPATH PGI linux86 15 10 man export MANPATH LM LICENSE FILE PGI license dat export LM LICENSE FILE PATH S PGI linux86 15 10 bin PATH export PATH 4 4 UF Ur 11 1 2 Setting En
79. ui nl tn dni 53 43 4 Working with Inline LIDTaftes doin eorr toti tira conc et ren EE 54 4 3 2 Dependencles uote Pre Le mer re an eri E E rA Re rre re eR E 54 4 3 3 Updating Inline Libraries Makefiles sssssseeeeeeeeteeentetntnt nne 54 44 Error Detection during InlininGt siti se oie ur retener trit rte eee tenet tcr aeree ederet eerta 55 ASS SCHO m H 55 4 6 Restrictions on Inlining ssseseeeeee ieai 55 Chapter 5 Using OpenMP eee 57 5 1 OpenMP OVGIVIGW cure tatio rr ERO EHE PELO TER D rei ede a Rez eue HERI RES iaaa 57 5 1 1 OpenMP Shared Memory Parallel Programming Model eene 58 5 1 2 Terminology teret trt tet tcr b tere ten Bi rete cr he rh Hr de vu ed re er d d Hc 58 5 13 lt OpenMP Example set recesso iet ria UO tesi mendi Bir df veda t d fran aina 59 5 2 Task OVON EW iiss tote terme oc t tete ot e ten b rv ntt vt tte e teet treu tbt rp tmb 60 5 3 Fortran Parallelization Ditectives i anc acciri dana ect terti e etr reir e 61 54 C C Parallelization Pragmas sse nennen tentent tenente tenens 62 5 5 Directive and Pragma ROeCOGFIODI a ccr ce titt re nnde cep tdeo ible d e n t 62 5 6 Directive and Pragma Summary Table sss teens 62 5 6 1 Directive and Pragma Summary Table sssssssssseeeene enne tenente nens 63 S
80. v tad 140 Te MPS ne 140 11 93 89 MP BIND ives ss carats eio acini Gm nominamus bumana m reta famed ca Eos 140 11 3 9 MP BLIST cerno e eei gode to ote sb tone od o ie asad t esed be to e ivre dod 141 110 MP SPIN E M 141 WARS Le IMPS WARN E E 141 132123 S JU RET 142 11 9 19 NCPUS MAX otro tre Patre entire dea dtes M Dore inue Do B on SOM e rua dis 142 11 3 14 NO STOP MESSAGE netten tentent tentent tentent te tenente tente tete en te tenen tentent ten 142 11319 PA m 142 11 3 16 SPOL ois tm xem en metet b nuance ess Ud a 142 11 347 POL CONTINUE dti tci rtr rtt re retenti tene tienne etn 143 11318 POIL OBJISUFFIX e Aan a ee a aie 143 11 3 19 PGI STACK USAGES chante de ioi nei mop ano mot de e ODE de KO OX ELO EO OR OLEO RO DR RD RS 143 14 93 20 P GLTTEBM oreet trace et in o toma etu om dolet da Ue e eee Te 143 11 321 PGL TERM DEBUG jis heh t Gestor tid nre e peer t p cie d GER RA LEE d 145 11 93 22 PGROUPD LICENSE FILE 1 1 122 1 rite tiet tette iacet runter tiet ea roit ates 145 JURA SUP 146 11 9 24 STATIC RANDOMC SEED ince acte a pde e cet 146 od e EE 146 1423 26 TMPDIR 146 11 4 Using Environment Modules on Linux
81. variable x To eliminate these PGI Compiler User s Guide 177 C C Inline Assembly and Intrinsics moves use a memory constraint type instead of a register constraint type as shown in the following example int x 1 void example2 asmi mwelell Si QM g Was ox je The compiler generates a memory reference in place of a memory constraint This eliminates the two extraneous moves Because the assembly uses a memory reference to variable x it does not have to move x into a register prior to the asm statement nor does it need to store the result after the asm statement Additional constraint types are found in Additional Constraints example2 so DEE 3 pushq rbp DEMOS movq rsp rbp ao DEn NIE lineno 5 addl 1 x rip lineno 0 popq rbp Tet The examples thus far have used only one output operand Because extended asm accepts a list of output operands asm statements can have more than one result as shown in the following example void example4 dime sedg aime Yep asmi Urol Sil siyn Yace Sil SOU Ware se Wa Gy je This example increments variable y by then adds it to variable x Multiple output operands are separated with a comma The first output operand is item 0 0 and the second is item 1 1 in the asm string The resulting values for x and y are 4 and 3 respectively 15 2 2 Input Operands The input operands are an optional list of input constraint and
82. w Here t is privatizable but the use of t outside the loop may yield incorrect results since the compiler may not be able to detect on which iteration of the parallelized loop t is last assigned The compiler detects the previous cases When a scalar is used after the loop but is not defined on every iteration of the loop parallelization does not occur When the programmer knows that the scalar is assigned on the last iteration of the loop the programmer may use a directive or pragma to let the compiler know the loop is safe to parallelize The directive or pragma safe_lastval informs the compiler that for a given loop all scalars PGI Compiler User s Guide 41 Optimizing and Parallelizing are assigned in the last iteration of the loop thus it is safe to parallelize the loop We could add the following line to any of our previous examples pgi l safe lastval Fortran Version pragma loop safe lastval ERC Version The resulting code looks similar to this E VE pragma loop safe lastval roe Gul ALENA esa d ime Gases 0 etg y E Fortran Version pgi l safe lastv In addition a command line option Msafe lastval provides this information for all loops within the routines being compiled which essentially provides global scope 3 8 Processor Specific Optimization amp the Unified Binary Every brand of processor has differences some subtle in hardware features such as instruction set
83. with swap space sufficiently large to support the data sets used in your application s If your memory swap space is not sufficiently large your application will likely encounter a segmentation fault at runtime PGI Compiler User s Guide 171 Programming Considerations for 64 Bit Environments support for large Arrays that are not dynamically allocated are limited by how the compiler can express the distance address offsets in object between them when generating code A field in the object file stores this distance value which is file format limited to 32 bits on Win32 Win64 linux86 and linux86 64 with mcmodel small It is 64 bits on linux86 64 with mcmodel medium Without the 64 bit offset support in the object file format large arrays cannot be declared statically or locally stack based 14 7 Medium Memory Model and Large Array in C Consider the following example where the aggregate size of the arrays exceeds 2GB Medium Memory Model and Large Array in C 9 cat bigadd c include lt stdio h gt define SIZE 600000000 gt 2GB 4 Siccicwe wiko eke eS TZEI EDS IZE SLIME main Leino Lome a im mp float ci SEAEI 7 Goes on stack n SIZE m 0 itoue GL e Of a lt me a 4e 10000 4 amp lx a ar is Ip 2 0 amp a ae 1g ez ala a biriz i g b 0 Olsan all DIG SIO g C lld a lld g b 11d gc 11d 2 gNn m m a m m b m
84. 1 numint2 44 numfloatl 39 6 numdoubl 39 2 numshortl 981 Assuming the Fortran program is in a file fmain f and the C function is in a file cpfunc C create an executable using the following command lines S Perk e F200 we oC 9 pgfozcran EZCO FUNGO 2GP Wied Sogierrililos Executing the a out file should produce the following output wy di a 3956 39 2 Yel 13 8 6 Example C Calling Fortran Fortran Subroutine cp2f_func f shows a Fortran subroutine called by the C main program shown in C main program cp2f_main C Notice that each call uses the amp operator to pass by reference Also notice that the call to the Fortran subroutine uses all lower case and a trailing C main program cp2f main C include lt iostream gt extern CY extern void c2 zune Char Char Spint pime p tlost couble short J P main char booll letterl aba iahbiadsliicil ELEA A float numtloatili double numdoub1 short numshorl cp2f func amp booll amp letterl amp numintl amp numint2 amp numfloatl amp numdoubl amp numshor1 cour lt lt isoollil Wg IXOOLIL eos lt lt W nsum Seoul lt lt VEALSE Ve come lt lt eimclils coub lt lt Letterl lt lt lebter sendl coub lt lt o manmel lt lt n mintEL lt lt encl cout sc muamipDLZ U nomintz ssendls cout lt numtloatl lt nmumfloat sendl cout lt lt numdoubl lt
85. 12 9 5 Scope of C C Pragmas and Command Line Options cccccccccececsescceeseseeeseeceeescseseeeeeeseesnseeeeeeeestsneaseeeeeets 113 9 6 Prefetch Directives and Pragmas iscine ioter candi Gandia ure n dei n e red 115 9 6 1 Prefetch Directive Syntax in Fortran nene tentent nnne nte tentent nen 116 9 6 2 Prefetch Directive Format Requirements ssssssssssssseeeen tnter 116 9 6 3 Sample Usage of Prefetch Directive sse 116 9 64 Prefetch Pragma Syntax in C C ouest eterni eet d eet ie i ee ii dee ie guns 116 9 6 5 Sample Usage of Prefetch Pragma eeeeeene tenente tnnt tnter tenete tnt tet tete tnt tnb eb teens 117 sre dq uaa atee 117 9 8 IGNORE TKR Directive acidic tana ie Co a die OH red e re cuoc 117 9 8 1 IGNORE TKR Directive Sy Max etr rt t er i e en De rd t 117 9 8 2 IGNORE TKR Directive Format Requirement cccecccsscecessesesesesescseseeeeseseeseeeeseatsnseseeeeeneenenseeeeteeeeneees 118 9 8 3 Sample Usage of IGNORE TKR Directive sees 118 9 9 DECS Directives coca re pote eem te n retenta teret R 118 9 9 1 IDEC Directive Syntax cca cce etes n e RW n TR E NK et 119 9 92 Format Requirements oot peer Un Ce re ae VERRE Pe CU wad a a eaaa NA 119 9 9 3 Summary TADS EU 119 Chapter TO Creating and Using Eibraries 2 rer retro menie tenti emeret cette 120 10 1 Using builtin Math Fun
86. 15 Related Accelerator Programming Tools ssssseseeeeeee entente tentent tenens 102 1 54 PGPROF pgcollBGLs secte tenere niter E erede ann 102 7 15 2 NVIDIA CUDA Profile sse nennt tenen 102 1 15 3 TAU Tuning and Analysis Utility 1 tractet ceret teh creer 102 416 Supported Jhtfinsics rm mehr o ta rre iom ete e p a p Rea 103 7 16 1 Supported Fortran Intrinsics Summary Table sse 103 7 16 2 Supported C Intrinsics Summary Table eene tenen 104 7 17 References related to Accelerators sssssssssssssseseeee nennen tenete tnnt enn 105 Chapter 8 ECliDSe sio aiia PERSE RO REEF EUH MeY Rr YR REEREFO ERFURT e EVEN OY FER e ei ure RETE 107 8 1 Install Eclipse CDT ciere pertinet rmt nta peer eir here er tte REP rre DHL RR nae 107 8 2 WSO Eclipse CDT feces crine rtt trt et ee tre edere n Pride aac elc reete aata Een eoe ele e Pr de REESE EIER 108 Chapter 9 Using Directives and Pragimassciiccs ccicsccscsccsceccstcscoteascecesesttccacenscste eatuscosearcece ceatuscoreancateceatunteteascececestuscoscaves 109 9 1 PGI Proprietary Fortran Directives 109 9 2 PGI Proprietary C and C Pragmas sse nter tnter entente tentent tenen 110 PGI Compiler User s Guide V 9 3 PGI Proprietary Optimization Directive and Pragma Summary seen 110 9 4 Scope of Fortran Directives and Command Line Options sse 1
87. 16 Using Directives and Pragmas where lt varn gt is any valid variable member or array element reference 9 6 5 Sample Usage of Prefetch Pragma Prefetch Pragma in C This example uses the prefetch pragma to prefetch data from the source vector x for eight iterations beyond the current iteration ror G 07 i lt 2 1 pragma mem prefetch x i 8 svat svat ar asedii 9 7 CSPRAGMA C When programs are compiled using one of the PGI Fortran compilers on Linux Win64 and OSX systems an underscore is appended to Fortran global names including names of functions subroutines and common blocks This mechanism distinguishes Fortran name space from C C name space You can use CSPRAGMA C in the Fortran program to call a C C function from Fortran The statement would look similar to this CSPRAGMA C name name This statement directs the compiler to recognize the routine name as a C function thus preventing the Fortran compiler from appending an underscore to the routine name On Win32 systems the CSPRAGMA C as well as the attributes C and STDCALL may effect other changes on argument passing as well as on the names of the routine For more information on this topic refer to Win32 Calling Conventions 9 8 IGNORE_TKR Directive This directive indicates to the compiler to ignore the type kind and or rank TKR of the specified dummy arguments in an interface of a procedure The compiler also ignores the type
88. 162 13 8 6 Example Ctr Calling Fortran uie e tet t tae ev A cue ra tere uad 163 13 9 Win32 Calling Conventions 2 1 xerit ione eene eren et ener centes cereos i reet eo c ee RR 164 13 9 1 Win32 Fortran Calling Conventions seeeeeeee tenentes 164 13 9 2 Symbol Name Construction and Calling Example senes 165 13 9 3 Using the Default Calling Convention sssrinin nentes 166 13 9 4 Using the STDCALL Calling Convention sese 166 13 9 5 Using the C Calling Convention cincti rtc ete chic 167 13 9 6 Using the UNIX Calling Convention tenentes 167 13 9 7 Using the CREF Calling Convention 22 22 2 1 nitri re tti ret ane crt c reet 167 Chapter 14 Programming Considerations for 64 Bit Environments eene 168 14 1 Data Types in the 64 Bit Environment eene rnnt 168 14 151 C C Data Types oerte parere eere da pae dne de da evt ev re dt 169 14 12 Fortran Data Types tenete terere pnt iere cepe ciere o RED 169 14 2 Large Static Data in Linux tentent tentent tentent tenete tenter nes 169 14 3 Large Dynamically Allocated Data terrestre itti ce en et nto oe gos 169 144 64 Bit Array Indexing b teme i tte be ato bant ob bane io nb a taxed inne ie 170 14 5 Compiler Options for 64 bit Programming seen 170 PGI Compiler User s Guide viii 14 6 Practical Limitations of Large Array Programming
89. 2GB The medium memory model specified by mcmodel medium allows combined code and static data areas text and bss sections larger than 2GB The mcmodel medium option must be used on both the compile command and the link command in order to take effect The Win64 and 64 bit Mac OS X operating systems do not have any support for large static data declarations There are two drawbacks to using mcmodel medium First there is increased addressing overhead to support the large data range This can affect performance though the compilers seek to minimize the added overhead through careful instruction generation Second mcmodel medium cannot be used for objects in shared libraries because there is no OS support for 64 bit dynamic linkage 14 3 Large Dynamically Allocated Data Dynamically allocated data objects in programs compiled by the 64 bit PGI compilers can be larger than 2GB No special compiler options are required to enable this functionality The size of the allocation is only limited by the system However to correctly access dynamically allocated arrays with more than 2G elements you should use the Mlarge arrays option described in the following section PGI Compiler User s Guide 169 14 4 64 Bit Array Indexing The 64 bit PGI compilers provide an option Mlarge arrays that enables 64 bit indexing of arrays This means that as necessary 64 bit INTEGER constants and variables are used to index arrays Programming
90. DBG Debugger Guide PGDBG can display the contents of message queues for instances of MPI that have been configured to support that feature The version of MPICHI provided with PGI Workstation on Linux and OS X is configured with this support At this time MS MPI does not support displaying message queue contents For more information on MPI and displaying message queues refer to the documentation for your specific distribution of MPI PGI Compiler User s Guide 75 Using MPI 6 4 Profiling MPI Applications The PGI performance profiler PGPROF provides support for profiling MPI applications The number of processes that can be profiled is limited by your license PGI Workstation licenses limit processes to a single system whereas PGI CDK licenses support general development on clusters PGPROF instrumentation is inserted into the program by the compiler and after the program is executed the PGPROF profiler can display MPI message count statistics as they relate to the source code of the application and the time spent in those portions of the application To create and view a performance profile of your MPI application you must first build an instrumented version of the application using the Mpro f option to specify one of the MPI distributions The Mp ro f option requires that you use another profiling sub option in conjunction with the MPI distribution sub options listed in Table 16 Table 16 MPI Profiling Options This MPI distr
91. ERM gt In the Windows Command Prompt C gt set PGI TERM option option Table 28 lists the supported values for opt ion Following the table is a complete description of each option that indicates specifically how you might apply the option By default all of these options are disabled Table 28 Supported PGI TERM Values no debug Enables disables just in time debugging debugging invoked on error no trace Enables disables stack traceback on error no abort Enables disables calling the system termination routine abort mjsgnd 0 Enables disables establishment of signal handlers for common signals that cause program termination no debug This enables disables just in time debugging The default is nodebug When PGI TERM is set to debug the following command is invoked on error unless you use PGI TERM DEBUG to override this default Hj pgdbg text attach pid pid is the process ID of the process being debugged The PGI TERM DEBUG environment variable may be set to override the default setting For more information refer to PGI TERM DEBUG no trace This enables disables stack traceback on error no signal This enables disables establishing signal handlers for the most common signals that cause program termination The default is nosignal Setting trace and debug automatically enables signal Specifically setting nosignal allows you to override this be
92. ES NO WARRANTIES EXPRESSED IMPLIED STATUTORY OR OTHERWISE WITH RESPECT TO THE MATERIALS AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE Information furnished is believed to be accurate and reliable However NVIDIA Corporation assumes no responsibility for the consequences of use of such information or for any infringement of patents or other rights of third parties that may result from its use No license is granted by implication of otherwise under any patent rights of NVIDIA Corporation Specifications mentioned in this publication are subject to change without notice This publication supersedes and replaces all other information previously supplied NVIDIA Corporation products are not authorized as critical components in life support devices or systems without express written approval of NVIDIA Corporation Trademarks PGI Workstation PGI Server PGI Accelerator PGF95 PGF90 PGFORTRAN and PGI Unified Binary are trademarks and PGI PGHPF PGF77 PGCC PGC PGI Visual Fortran PVF PGI CDK Cluster Development Kit PGPROF PGDBG and The Portland Group are registered trademarks of NVIDIA Corporation in the U S and other countries Other company and product names may be trademarks of the respective companies with which they are associated Copyright 2013 2015 NVIDIA Corporation All rights reserved PGI
93. For strings declared in Fortran as type CHARACTER an argument representing the length of the string is also passed to a calling function On the following systems the compiler places the length argument s at the end of the parameter list following the other formal arguments gt On Linux and Mac OS X systems gt On Win32 systems when using the UNIX calling convention on Windows that is using the option Miface unix gt On Win64 systems except when using the option Miface cref The length argument is passed by value not by reference 13 6 1 Passing by Value V6 VAL When passing parameters from a Fortran subprogram to a C C function it is possible to pass by value using the VAL function If you enclose a Fortran parameter with VAL the parameter is passed by value For example the following call passes the integer i and the logical bvar by value integer 1 i logical i ibam call cvalue VAL i VAL bvar 13 6 2 Character Return Values Functions and Subroutines describes the general rules for return values for C C and Fortran inter language calling There is a special return value to consider When a Fortran function returns a character two arguments need to be added at the beginning of the C C calling function s argument list gt The address of the return character or characters PGI Compiler User s Guide 157 Inter language Calling gt The length of the return character The following example i
94. Function inlining is not always beneficial When used improperly it may increase code size and generate less efficient code 3 1 6 Profile Feedback Optimization PFO Profile feedback optimization PFO makes use of information from a trace file produced by specially instrumented executables which capture and save information on branch frequency function and subroutine call frequency semi invariant values loop index ranges and other input data dependent information that can only be collected dynamically during execution of a program By definition use of profile feedback optimization is a two phase process compilation and execution of a specially instrumented executable followed by a subsequent compilation which reads a trace file generated during the first phase and uses the information in that trace file to guide compiler optimizations 3 2 Getting Started with Optimization Your first concern should be getting your program to execute and produce correct results To get your program running start by compiling and linking without optimization Add 00 to your compile line to select no optimization or add g to allow you to debug your program easily and isolate any coding errors exposed during porting to x86 or x64 platforms For more information on debugging refer to the PGDBG Debugger Guide If you want to get started quickly with optimization a good set of options to use with any of the PGI compilers is fast Mipa fast inline F
95. GI OBJSUFFIX You can set the PGI_OBJSUFFIX environment variable to generate object files that have a specific suffix For example if you set PGI_OBJSUFFIX to o the object files have a suffix of o rather than obj 11 3 19 PGI STACK USAGE Windows only The PGI STACK USAGE variable allows you to explicitly set stack properties for your program When the user compiles a program with the Mchkstk option and sets the PGI STACK USAGE environment variable to any value the program displays the stack space allocated and used after the program exits You might see something similar to the following message thread 0 stack max 8180KB used 48KB This message indicates that the program used 48KB of a 8180KB allocated stack For more information on the Mchkstk option refer to Mchkstk in the PGI Compiler s Reference Manual 11 3 20 PGI TERM The PGI TERM environment variable controls the stack traceback and just in time debugging functionality The runtime libraries use the value of PGI_TERM to determine what action to take when a program abnormally terminates The value of PGI TERM is a comma separated list of options The commands for setting the environment variable follow PGI Compiler User s Guide 143 Using Environment Variables gt Incsh setenv PGI TERM option option gt In bash sh zsh or ksh PGI TERM option option export PGI T
96. GI Unified Binary object files to create programs or libraries No special start up code is needed support is linked in from the PGI libraries The Mp i option disables generation of PGI Unified Binary object files Instead the default target auto detect rules for the host are used to select the target processor 12 4 1 PGI Unified Binary Command line Switches The PGI Unified Binary command line switch is an extension of the target processor switch tp which may be applied to individual files during compilation PGI Compiler User s Guide 151 Distributing Files Deployment The target processor switch tp accepts a comma separated list of 64 bit targets and generates code optimized for each listed target The following example generates optimized code for three targets tp k8 64 p7 64 core2 64 A special target switch tp x64 is the same as tp k8 64 p7 64 12 4 2 PGI Unified Binary Directives and Pragmas PGI Unified binary directives and pragmas may be applied to functions subroutines or whole files The directives and pragmas cause the compiler to generate PGI Unified Binary code optimized for one or more targets No special command line options are needed for these pragmas and directives to take effect The syntax of the Fortran directive is ESSI NECPONESSSSHENE where the scope is g global r routine or blank The default is r routine For example the following syntax indicates that the whole file repre
97. Jo il IWF maxtime 10 n 100 for time 0 time lt maxtime timet t ror G07 spem e itoue Oe wos alae Se alligi biiliiis In this version the compiler does not perform vectorization for the entire source file Another use of the pragma scoping mechanism turns an option on or off locally either for a specific procedure or for a specific loop The following example shows the use of a loop scoped pragma main sloeg allo LOO 19 350971 LOO elLOO L LOO IME TMS eeN iM y A maxtime 10 n 100 pragma loop novector for time 0 time lt maxtime time for j 0 j lt n j Eo S09 sb a eb ex Jefa 518 Loop level scoping does not apply to nested loops That is the pragma only applies to the following loop In this example the pragma turns off vector transformations for the top level loop If the outer loop were a timing loop this would be a practical use for a loop scoped pragma The following example shows routine pragma scope include math h 1E uae 1L O fpragma routine novector float a 100 100 b 100 100 ElOere Cto troo elro TrO0 9 aN 3 8 ron t 07 aE 1007F for j3 0 3 lt 100 j aad E33 ala ts se Tea elallals ea 31 Cla ts s Jed eg tals func2 sloer 29011 T200 1912199 12007 loa E200 1200 amp 1200 1200 7 IME tk 375 rown t 9 ak lt O09 i for j 0 3 lt 200 j alti ig ala ig e orir eria eli ig Cla li s sft ig
98. LLEL single SECTIONS PARALLEL WORKSHARE SECTIONS SINGLE READ ATOMIC Specifies that the atomic action is reading a value DO for REDUCTION Specifies that one or more variables that are private to each thread are the subject of a reduction operation at the end of the parallel region PARALLEL parallel operator PARALLEL DO parallel for END PARALLEL DO parallel sections PARALLEL SECTIONS sections END PARALLEL list SECTIONS PARALLEL WORKSHARE SECTIONS intrinsic SCHEDULE DO END DO for Applies to the FOR directive allowing PARALLEL DO parallel for the user to specify the chunking method type for parallelization Work is assigned to WE END PARALLEL DO threads in different manners depending chunk on scheduling type or chunk size used SHARED PARALLEL parallel Ma s i or more bes PARALLEL DO parallel for Pip P NIAE DANN threads within a team access the same END PARALLEL DO parallel sections storage area for shared variables PARALLEL SECTIONS END PARALLEL SECTIONS PGI Compiler User s Guide 66 Using OpenMP Applies to this This clause Applies to this directive pragma Has this functionality PARALLEL WORKSHARE UNTIED TASK task Specifies ee un vues in team TASKWAIT tackwdit can resume the task region after a suspension UPDATE ATOMIC atomic Specifies that the atomic action is updating a value WRITE ATOMIC atomic Specifies that th
99. MB For information on how to set environment variables refer to Setting Environment Variables PGI Compiler User s Guide 13 Getting Started 1 8 Site Specific Customization of the Compilers If you are using the PGI compilers and want all your users to have access to specific libraries or other files there are special files that allow you to customize the compilers for your site 1 8 1 Use siterc Files The PGI compiler drivers utilize a file named siterc to enable site specific customization of the behavior of the PGI compilers The siterc file is located in the bin subdirectory of the PGI installation directory Using siterc you can control how the compiler drivers invoke the various components in the compilation tool chain 1 8 2 Using User rc Files In addition to the siterc file user rc files can reside in a given user s home directory as specified by the user s HOME environment variable You can use these files to control the respective PGI compilers All of these files are optional On Linux and OS X these files are named mypgf77rc mypgf90rc mypgccro and mypgc t trc On Windows these files are named mypgf 77rc mypgf90rc mypgf95rc mypgfortranrc mypgccrc andmypgcttre The following examples show how you can use these rc files to tailor a given installation for a particular purpose Table 3 Examples of Usine siterc and User rc Files To do this Add the line shown to the indicated file s Make availa
100. PA is not compatible with parallel make environments 3 10 Profile Feedback Optimization using Mpfi Mpfo The PGI compilers support many common profile feedback optimizations including semi invariant value optimizations and block placement These are performed under control of the Mpfi Mpfo command line options When invoked with the Mp i option the PGI compilers instrument the generated executable for collection of profile and data feedback information This information can be used in subsequent compilations that include the Mpfo optimization option Mpfi must be used at both compile time and link time Programs compiled with Mp fi include extra code to collect runtime PGI Compiler User s Guide 4T Optimizing and Parallelizing statistics and write them out to a trace file When the resulting program is executed a profile feedback trace file pgfi out is generated in the current working directory Programs compiled and linked with Mpf i execute more slowly due to the instrumentation and data collection overhead You should use executables compiled with Mp f i only for execution of training runs When invoked with the Mp o option the PGI compilers use data from a pgf i out profile feedback tracefile to enable or enhance certain performance optimizations Use of this option requires the presence of a pg i out trace file in the current working directory 3 11 Default Optimization Levels The following table shows the
101. PARALLEL parallel PARALLEL DO parallel for PGI Compiler User s Guide Specifies that each thread should have its own instance of a variable and that each variable in the list should be initialized with the value of the original variable because it exists before the parallel construct Specifies whether a loop should be executed in parallel or in serial Specifies that the enclosing context s version of the variable is set equal to the private version of whichever thread executes the final iteration of a for loop construct or last section of pragma sections Specifies that this task will run with the same data environment including OpenMP internal control variables as when it is encountered Eliminates the barrier implicit at the end of a parallel region Sets the number of threads in a thread team 65 Using OpenMP Applies to this This clause Applies to this directive pragma Has this functionality END PARALLEL DO parallel sections PARALLEL SECTIONS END PARALLEL SECTIONS PARALLEL WORKSHARE ORDERED DO END DO parallel for ee e s within Hp PARALLEL DO parallel DO or FOR region needs to be execute serially in the same order END PARALLEL DO indicated by the enclosing loop PRIVATE DO for Specifies that each thread should have PARALLEL parallel its own instance of a variable PARALLEL DO parallel for END PARALLEL DO parallel sections PARALLEL SECTIONS sections END PARA
102. PGI gt On Mac OS X the default value of this variable is opt pgi PGI Compiler User s Guide 142 Using Environment Variables In most cases if the PGI environment variable is not set the PGI compilers and tools dynamically determine the location of this root directory based on the instance of the compiler or tool that was invoked However there are still some dependencies on the PGI environment variable and you can use it as a convenience when initializing your environment for use of the PGI compilers and tools For example assuming you use csh and want the 64 bit linux86 64 versions of the PGI compilers and tools to be the default you would use this syntax oe setenv PGI opt pgi setenv MANPATH SMANPATH S PGI linux86 15 10 man setenv LM LICENSE FILE PGI license dat set path PGI linux86 64 15 10 bin path 11 3 17 PGI CONTINUE You set the PGI CONTINUE variable to specify the actions to take before continuing with execution For example if the PGI CONTINUE environment variable is set and then a program that is compiled with Mchkfpstk is executed the stack is automatically cleaned up and execution then continues If PGI CONTINUE is set to verbose the stack is automatically cleaned up a warning message is printed and then execution continues oe oe oe cy There is a performance penalty associated with the stack cleanup 11 3 18 P
103. RKSHARE Provides a mechanism to effect parallel execution of non iterative but implicitly data parallel constructs 5 7 Directive and Pragma Clauses Some directives and pragmas accept clauses that further allow a user to control the scope attributes of variables for the duration of the directive or pragma Not all clauses are allowed on all directives so the clauses that are valid are included with the description of the directive and pragma The following table provides a brief summary of the clauses associated with OPENMP directives and pragmas that PGI supports For complete information on these clauses refer to the OpenMP documentation available on the World Wide Web Table 12 Directive and Pragma Summary Table This clause CAPTURE COLLAPSE n COPYIN list COPYPRIVATE list PGI Compiler User s Guide Applies to this directive Applies to this pragma ATOMIC atomic DO END DO PARALLEL DO PARALLEL WORKSHARE parallel for PARALLEL PARALLEL DO PARALLEL SECTIONS PARALLEL WORKSHARE parallel parallel for SINGLE single Has this functionality Specifies that the atomic action is reading and updating or writing and updating a value capturing the intermediate state Specifies how many loops are associated with the loop construct Allows threads to access the master thread s value for a threadprivate variable You assign the same value to threadprivate variables f
104. Sets the number of threads to use for the next parallel region PGI Compiler User s Guide 67 Using OpenMP Runtime Library Routines with Examples This subroutine or function can only be called from a serial region of code If it is called from within a parallel region or from within a subroutine or function that is called from within a parallel region the results are undefined Further this subroutine or function has precedence over the OMP NUM THREADS environment variable subroutine omp set num threads scalar integer exp void omp set num threads int num threads omp get thread num Returns the thread number within the team The thread number lies between 0 and omp get num threads 1 When called from a serial region this function returns 0 A nested parallel region is the same as a single parallel region integer function omp get thread num int omp get thread num void omp get ancestor thread num Returns for a given nested level of the current thread the thread number of the ancestor Fortran integer function omp get ancestor thread num level integer level me Gwo ger eneestor hte memme devel omp_get_active_level Returns the number of enclosing active parallel regions enclosing the task that contains the call PGI currently supports only one level of active parallel regions so the return value currently is 1 integer function omp get active level omp get level Returns the number of parallel reg
105. Syntax eacee cecleare elewe I elewa as Update Directive async int expr wait int expr device type devtype list if condition self list host list device list C Syntax fpragma acc update clause clause Fortran Syntax acc update clause clause Routine Directive gang worker vector seq PGI Compiler User s Guide Used during the lifetime of accelerator data to update all or part of a host memory array with values from the corresponding array in device memory or to update all or part of a device memory array with values from the corresponding array in host memory new line Used to tell the compiler to compile a given procedure for an accelerator as well as the host In a file or routine with a procedure call the rout ine directive tells the implementation the attributes of the procedure when called on the accelerator 94 Using an Accelerator This directive Accepts these clauses Has this functionality C Syntax pragma acc routine clause clause new lin or pragma acc routine name clause clause Fortran Syntax Sacc routine clause clause or Sacc routine name clause clause Wait Directive Specifies to wait until all operations on a specific device if condition async queue or all async queues are complete device list C Syntax pragma acc wait int expr list clause clause new line Fort
106. a Update PGI Compiler User s Guide 97 Using an Accelerator Use this clause In these directives To do this worker num int expr Specifies that the iterations of the associated loop or loops are to be executed in parallel by distributing the iterations among the multiple workers within a single gang Loop 7 11 OpenAcc Runtime Libraries This section provides an overview of the user callable functions and library routines that are available for use by programmers to query the accelerator features and to control behavior of accelerator enabled programs at runtime In Fortran none of the OpenACC runtime library routines may be called from a PURE or ELEMENTAL procedure 7 11 1 Runtime Library Definitions There are separate runtime library files for C and for Fortran C Runtime Library Files In C prototypes for the runtime library routines are available in a header file named accel h All the library routines are extern functions with C linkage This file defines gt The prototypes of all routines in this section gt Any data types used in those prototypes including an enumeration type to describe types of accelerators Fortran Runtime Library Files In Fortran interface declarations are provided in a Fortran include file named accel lib h and in a Fortran module named accel lib These files define gt Interfaces for all routines in this section gt nteger parameters to define
107. a programming error results in a runtime error message or an application exception a program will usually exit perhaps with an error message The PGI runtime library includes a mechanism to override this default action and instead print a stack traceback start a debugger or on Linux create a core file for post mortem debugging The stack traceback and just in time debugging functionality is controlled by an environment variable PGI_ TERM described in PGI TERM The runtime libraries use the value of PGI TERM to determine what action to take when a program abnormally terminates When the PGI runtime library detects an error or catches a signal it calls the routine pgi stop here prior to generating a stack traceback or starting the debugger The pgi stop here routine is a convenient spot to set a breakpoint when debugging a program PGI Compiler User s Guide 147 Chapter 12 DISTRIBUTING FILES DEPLOYMENT Once you have successfully built debugged and tuned your application you may want to distribute it to users who need to run it on a variety of systems This section addresses how to effectively distribute applications built using PGI compilers and tools The application must be installed in such a way that it executes accurately on a system other than the one on which it was built and which may be configured differently 12 1 Deploying Applications on Linux To successfully deploy your application on Linux som
108. able append the path to the license file to the variable s current value using a semi colon to separate entries 11 3 6 MANPATH The MANPATH variable sets the directories that are searched for manual pages associated with the commands that the user types When using PGI products it is important that you set your PATH to include the location of the PGI products and then set the MANPATH variable to include the man pages associated with the products MANPATH is not required for OS X The following csh example targets x64 linux86 64 version of the compilers and tool s and allows the user access to the manual pages associated with them set path opt pgi linux86 64 15 10 bin Spath setenv MANPATH SMANPATH opt pgi linux86 64 15 10 man 11 3 7 MPSTKZ MPSTKZ increases the size of the stacks used by threads executing in parallel regions You typically use this variable with programs that utilize large amounts of thread local storage in the form of private variables or local variables in functions or subroutines called within parallel regions The value should be an integer n concatenated with M or m to specify stack sizes of n megabytes For example the following setting specifies a stack size of 8 megabytes setenv MPSTKZ 8M 11 3 8 MP BIND You can set MP_BIND to yes or y to bind processes or threads executing in a parallel region to physical processor Set it to no or n to disable such binding The defa
109. 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 lf al tcode nis specified the serial altcode is executed whenever the loop count is less than or equal to n Specifying noaltcode disables this option and no alternate serial code is generated Mconcur no assoc Enable disable parallelization of loops with associative reductions Bind threads to cores Specify this option at link time Mconcur cncall Specifies that it is safe to parallelize loops that contain subroutine or function calls By default such loops are excluded from consideration for auto parallelization Also no minimum loop count threshold must be satisfied before PGI Compiler User s Guide 37 Optimizing and Parallelizing Use this option To instruct the parallelizer to do this parallelization occurs and last values of scalars are assumed to be safe Mconcur dist block cyclic Specifies whether to assign loop iterations to the available threads in blocks or in a cyclic round robin fashion Block distribution is the default If cyclic is specified iterations are allocated to processors cyclically That is processor 0 performs iterations 0 3 6 etc processor 1 performs iterations 1 4 7 etc and processor 2 performs iterations 2 5 8 etc
110. amples Showing Effect of Unrolling The following side by side examples show the effect of code unrolling on a segment that computes a dot product This example is only meant to represent how the compiler can transform the loop it is not meant to imply that the programmer needs to manually change code In fact manually unrolling your code can sometimes inhibit the compiler s analysis and optimization PGI Compiler User s Guide 31 Optimizing and Parallelizing Table 7 Example of Effect of Code Unrolling Dot Product Code Unrolled Dot Product Code REAL 4 A 100 B 100 Z LOO 7 IAW z INTEGE DO I 1 B i A itl B i 1 Using the Minfo option the compiler informs you when a loop is being unrolled For example a message similar to the following indicating the line number and the number of times the code is unrolled displays when a loop is unrolled doek 5 Loop unrolled 5 times Using the c lt m gt and n lt m gt sub options to Munroll or using Mnounroll you can control whether and how loops are unrolled on a file by file basis Using directives or pragmas you can precisely control whether and how a given loop is unrolled For more information on Munroll refer to Use Command Line Options 3 6 Vectorization using Mvect The Mvect option is included as part of fast on all x86 and x64 targets If your program contains computationally intensive loops the Mvect option may be helpful If in
111. ance data and cannot be used concurently Windows Users To ensure that all the performance information is collected we recommend that acc shutdown is called before your application is finished and main exits PGI Compiler User s Guide 101 Using an Accelerator Accelerator Kernel Timing Data bb04 90 sl 15 region entered 1 times time us total 1490738 init 1489138 region 1600 kernels 155 data 1445 w o init total 1600 max 1600 min 1600 avg 1600 18 kernel launched 1 times time us total 155 max 155 min 155 avg 155 In this example a number of things are occurring gt For each accelerator region the file name bb04 90 and subroutine or function name s1 is printed with the line number of the accelerator region which in the example is 15 gt The library counts how many times the region is entered 1 in the example and the microseconds spent in the region in this example 1490738 which is split into initialization time in this example 1489138 and execution time in this example 1 600 gt The execution time is then divided into kernel execution time and data transfer time between the host and GPU gt Foreach kernel the line number is given 18 in the example along with a count of kernel launches and the total maximum minimum and average time spent in the kernel all of which are 155 in this example 7 15 Related Accelerator Programming Tools 7 15 1 PGPROF pgcollect The PGI profiler PGPROF
112. anecencestectectaructects 81 MOT I M M 81 EAN Components 2 rite tr tent n n Ete endo trn rn dd m a Ee A ui re TAI 81 GAZ Avallability in tio nie on tee on tene ette E AE net on ttn ee 82 7 1 3 User directed Accelerator Programming sssssseeeeeeee tenete nte tentent nter 82 PGI Compiler User s Guide iv 7 1 4 Features Not Covered or Implemented cceccececsesesessesesecsescesesesucseeeeseatsnsuceeeeeacacsnseceeeceasansesaceeeesatenseeases 82 2 Terminology sr tet o eite te EE dente rre dete Rt Pede nra eI qe ee ER EE rrt FRU ERI dEa 82 7 3 System Requirements ice ete ais die A E ee Pet b edie e e eee dl aera 84 7 4 Supported Processors and GPUS ii is aaaea ariaa aiaa EEE aaia raai SEa aaa 84 Fo Installation and EICenslrig 25 0 dai dicta E rete d i reti EE oid ee d oe ende A 85 1 5 1 Required Iles rre m ttem et rec te ree deoa n ERR E CH maine eines 85 1 5 2 Command Line Flag euis ist eins eieestietdentr entend endete aiei Ea rhe ndn ede 85 T6 Execution Modell rette ier rue trarre derer Rt tata HR de Peter erbe a 85 EG Host FUNCIONS oerte ER URER RR RH REDE EROR ERRARE HORDE REA EHE E KE REDE tien 86 1 0 2 Levels of Parallelism i m tmr deti i tta e ere rp hr bee ME Me eH ERREUR EN ERETS ann 86 ele Memory Model detet erdt ideo cna dec nn PE n ecd t te e e Tg 86 7 7 1 Separate Host and Accelerator Memory Consider
113. ard 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 PGI supports is available in the Release Notes The table also includes the features utilized by the PGI compilers that distinguish them from a compatibility standpoint In this manual the convention is to use x86 to specify the group of processors that are 32 bit but not 64 bit The convention is to use x64 to specify the group of processors that are both 32 bit and 64 bit x86 processor based systems can run only 32 bit operating systems x64 processor based systems can run either 32 bit or 64 bit operating systems and can execute all 32 bit x86 binaries in either case x64 processors have additional registers and 64 bit addressing capabilities that are utilized by the PGI compilers and tools when running on a 64 bit operating system The prefetch SSE1 SSE2 SSE3 and AVX processor features further distinguish the various processors Where such distinctions are important with respect to a given compiler option or feature it is explicitly noted in this manual The default for performing scalar floating point arithmetic is to use SSE instructions on targets that support SSE1 and SSE2 Related Publications The following documents contain additional information related to the x86 and x64 architectures and the compile
114. ation PFO 3 1 1 Local Optimization Local optimization is performed on a block by block basis within a program s basic blocks A basic block is a sequence of statements in which the flow of control enters at the beginning and leaves at the end without the possibility of branching except at the end The PGI compilers perform many types of local optimization including algebraic identity removal constant folding common sub expression elimination redundant load and store elimination scheduling strength reduction and peephole optimizations 3 1 2 Global Optimization This optimization is performed on a program unit over all its basic blocks The optimizer performs control flow and data flow analysis for an entire program unit All loops including those formed by IFs and GOTOs are detected and optimized Global optimization includes constant propagation copy propagation dead store elimination global register allocation invariant code motion and induction variable elimination 3 1 3 Loop Optimization Unrolling Vectorization and Parallelization The performance of certain classes of loops may be improved through vectorization or unrolling options Vectorization transforms loops to improve memory access performance and make use of packed SSE instructions which perform the same operation on multiple data items concurrently Unrolling replicates the body of loops to reduce loop branching overhead and provide better opportunities for
115. ations seen 87 ald Accelerator Memory oerte ot Re tot tet t e Hot Up te bove e tod tenu BO v tn 87 L3 Cache Management deri b onini tn ia R ree rede rbi i e re vi reed 87 17 8 Running an Accelerator Progfam ie o tre ro eO OE GRO RU ROLE RA ROO UU FURORE 87 1 9 Accelerator DIrectiV6S u tero tm omo nie o at e a eta de Rer o t e er db c d 88 1 931 Enable Accelerator Directives erem mte P tret ep rere ve re HO ai Ni Te HERES 88 9 2 3 0 aloes d reote ctae etude ertet tenerte adiecit ure reseen fer test Dt ef oT UR 88 1932 CDIRECIVES M 89 7 9 4 Free Form Fortran Directives teen tette tenente tenete nennen 89 7 9 5 Fixed Form Fortran Directives sese tenente tentent tenerent 90 1 9 6 OpenACC Directive SUMMA seriea ette er ete ete te ee eee eee eate eoe 91 7 10 Accelerator Directive Clauses nennen tette tentent tenete rnnt tenerent tenens 95 LAV OpenAcc Runtime LIDEarIes dette tritt ti erbe e tenet i re tede gn 98 31 1 Runtime Library DefIRiftiona s ete rtr dpt je rem ote re a re dip th ren bd a 98 7 41 2 Runtime Library ROUNE S isc imei imn ime Gre e eG de ene Gane n Pe ri e 98 1 42 Environment Varables eter te ite e neam Ure bre tes Rt LP PL Eh P leaned 99 7 43 Applicable Command Line Optlons te aiana Rer e Eg eels 100 1 14 Profiling Accelerator Kemels sese nennen nenne tnnt tn tente tn ttt tt ta tet to tetto te detto t dont nn 101 7
116. ations of Microsoft and several other vendors the C calling convention supported by the PGI Fortran compilers for Windows is not strictly compatible with the C calling convention used by most C C compilers In particular symbol names produced by PGI Fortran compilers using the C convention are all lower case The standard C convention is to preserve mixed case symbol names You can cause any of the PGI Fortran compilers to preserve mixed case symbol names using the Mupcase option but be aware that this could have other ramifications on your program 13 9 2 Symbol Name Construction and Calling Example This section presents an example of the rules outlined in Calling Conventions Supported by the PGI Fortran Compilers In the pseudocode shown in the following examples addr refers to the address of a data item while va1 refers to the value of that data item Subroutine and function names are converted into symbol names according to the rules outlined in Table 31 Consider the following subroutine call where a is a double precision scalar b is a real vector of size n and n is an integer call work YER e Zi lo ia gt Default The symbol name for the subroutine is constructed by pre pending an underscore converting to all upper case and appending an sign followed by an integer indicating the total number of bytes occupied by the argument list Byte counts for character arguments appear immediately following the corresponding argume
117. ay be clobbered by the asm statement By placing registers in this list the programmer does not have to explicitly save and restore them as required in traditional inline assembly described in Inline Assembly The compiler takes care of any required saving and restoring of the registers in this list Each machine register in the clobber list is a string separated by a comma The leading is optional in the register name For example eax is equivalent to eax When specifying the register inside the asm string you must include two leading characters in front of the name for example eax Otherwise the compiler will behave as if a bad input output operand was specified and generate an error message An example follows void examples IONE xA int y 2 asm movl 1 eax n movl 1 edx n addl edx eax n addl eax 0 y x UO Cy ii y eax des pe Ud This code uses two hard coded registers eax and edx It performs the equivalent of 3 y and assigns it to x producing a result of 6 In addition to machine registers the clobber list may contain the following special flags cc The asm statement may alter the control code register memory The asm statement may modify memory in an unpredictable fashion When the memory flag is present the compiler does not keep memory values cached in registers across the asm statement and does not optimize stores or loads to that memory
118. bal optimizations are performed including traditional scalar optimizations induction recognition and loop invariant motion No SIMD vectorization is enabled 02 Level two specifies global optimization This level performs all level one local optimization as well as level two global optimization described in O In addition more advanced optimizations such as SIMD code generation cache alignment and partial redundancy elimination are enabled PGI Compiler User s Guide 29 Optimizing and Parallelizing 03 Level three specifies aggressive global optimization This level performs all level one and level two optimizations and enables more aggressive hoisting and scalar replacement optimizations that may or may not be profitable 04 Level four performs all level one level two and level three optimizations and enables hoisting of guarded invariant floating point expressions Types of Optimizations The PGI compilers perform many different types of local optimizations including but not limited to Algebraic identity removal Constant folding Common subexpression elimination Local register optimization Peephole optimizations Redundant load and store elimination Strength reductions Level two optimization O2 or O specifies global optimization The fast option generally specifies global optimization however the fast switch varies from release to release depending on a reasonable selection of switches for any on
119. bblas a PGI Compiler User s Guide 133 Creating and Using Libraries 10 11 The C Standard Template Library On Linux and OS X the GNU compatible pgc compiler uses the GNU g header files and Standard Template Library STL directly The versions used are dependent on the version of the GNU compilers installed on your system or specified when makelocalrc was run during installation of the PGI compilers On Linux OS X and Windows the deprecated pgcpp compiler includes a bundled copy of the STLPort Standard C Library See the online Standard C Library tutorial and reference manual at www stlport com for further details and licensing PGI Compiler User s Guide 134 Chapter 11 USING ENVIRONMENT VARIABLES Environment variables allow you to set and pass information that can alter the default behavior of the PGI compilers and the executables which they generate This section includes explanations of the environment variables specific to PGI compilers Other environment variables are referenced and documented in other sections of this User s Guide the accompanying Reference Manual the PGDBG Debugger Guide and the PGPROF Profiler Manual gt You use OpenMP environment variables to control the behavior of OpenMP programs For consistency related to the OpenMP environment the details of the OpenMP related environment variables are included in OpenMP section Environment Variables gt You can use environment variables
120. be removed in a future release pgfortran Error The environment variable SMPIDIR must be set to use MPICH2 Once MPIDIR is set you can use Mmpi and Mprof with the mpichl mpich2 and mvapichl sub options as you have in previous releases PGI Compiler User s Guide 77 Using MPI 6 7 Using MVAPICH2 on Linux PGI CDK users can download a PGI built version of MVAPICH2 for Linux that includes everything required to compile execute profile and debug MPI programs using MVAPICH2 To build an application using MVAPICH2 use the MVAPICH2 compiler wrappers mpicc mpic mpif77 and mpif90 These wrappers autonmatically set up the compiler commands with the correct include file search paths library directories and link libraries To build an application using MVAPICH2 for debugging add g to the compiler wrapper command line arguments PGI provides two profiling configuration files with MV APICH2 to enable profiling When using the compilers wrappers mpicc or mpic add the profile profcc option When using the wrappers mpif77 or mpif90 add the profile proffer option To all the wrappers add one of the Mprof func Mprof lines or Mprof time options 6 8 Using MS MPI on Windows PGI products on Windows include a version of Microsoft s MPI You can compile run debug and profile locally on your system using this instance of MS MPI To compile the application use the Mmpi msmpi option This option automatically sets up th
121. ble to all linux86 64 set SITELIB opt newlibs 64 compilations the libraries found in opt newlibs 64 to opt pgi linux86 64 15 10 bin siterc Make available to all linux86 compilations set SITELIB opt newlibs 32 the libraries found in opt newlibs 32 to opt pgi linux86 15 10 bin siterc Add to all linux86 64 compilations a append SITELIB opt local fast new library path opt 1ocal fast to opt pgi linux86 64 15 10 bin siterc Make available to all linux86 compilations set SITELIB opt acml include the include path I opt acml include to opt pgi linux86 15 10 bin sitercand opt pgi linux86 64 15 10 bin siterc PGI Compiler User s Guide 14 Getting Started To do this Add the line shown to the indicated file s With linux86 64 compilations change set MPILIBDIR opt mympi 64 Mmp ii to link in opt mympi 64 libmpix a set MPILIBNAME mpix to opt pgi linux86 64 15 10 bin siterc With linux86 64 compilations always add set SI TEDEF IS64BIT AMD D S64BIT DAMD to opt pgi linux86 64 15 10 bin siterc Build an F90 or F95 executable for set set RPATH REDIST linux86 64 or linux86 that resolves PGI shared objects in the relative directory to mypgfortranrc REDIST Note This only affects the behavior of PGFORTRAN for the given user 1 9 Common Development Tasks Now that you have a brief introduction to the compiler let s look at
122. c float _declspec dllexport fdata The PGI Fortran compilers support the DEC ATTRIBUTES extensions DLLIMPORT and DLLEXPORT cDECS ATTRIBUTES DLLEXPORT object object cDECS ATTRIBUTES DLLIMPORT object object Here c is one of C c or object is the name of the subprogram or common block that is exported or imported Further common block names are enclosed within slashes as shown here cDECS ATTRIBUTES DLLIMPORT intfunc DECS ATTRIBUTES DLLEXPORT fdata For more information on these extensions refer to DECS Directives The examples in this section further illustrate the use of these extensions To create a DLL from the command line use the Mmaked11 option The following switches apply to making and using DLLs with the PGI compilers Bdynamic Compile for and link to the DLL version of the PGI runtime libraries This flag is required when linking with any DLL built by the PGI compilers This flag corresponds to the MD flag used by Microsoft s c1 compilers When you use the PGI compiler flag Bdynamic to create an executable that links to the DLL form of the runtime the executable built is smaller than one built without Bdynamic The PGI runtime DLLs however must be available on the system where the executable is run You must use the Bdynamic flag when linking an executable against a DLL built by the PGI compilers
123. c Static local pointers are safe If your C C program has pointer aliasing and you also want automating inlining then compiling with Mipa fast or Mipa fast inline includes pointer aliasing optimizations IPA may be able to optimize some of the alias references in your program and leave intact those that cannot be safely optimizied 34 2 O Using the PGI compiler commands with the Olevel option the capital O is for Optimize you can specify any integer level from 0 to 4 00 Level zero specifies no optimization A basic block is generated for each language statement At this level the compiler generates a basic block for each statement Performance will almost always be slowest using this optimization level This level is useful for the initial execution of a program It is also useful for debugging since there is a direct correlation between the program text and the code generated To enable debugging include g on your compile line 01 Level one specifies local optimization Scheduling of basic blocks is performed Register allocation is performed Local optimization is a good choice when the code is very irregular such as code that contains many short statements containing IF statements and does not contain loops DO or DO WHILE statements Although this case rarely occurs for certain types of code this optimization level may perform better than level two O2 0 When no level is specified level two glo
124. called 11blapack a or on Windows liblapack lib The BLAS library is called 1ibblas a oron Windows 1ibblas lib These libraries are installed to PGI target lib where target is replaced with the appropriate target name linux86 linux86 64 osx86 osx86 64 win32 or win64 To use these libraries simply link them in using the 1 option when linking your main program 9 pgfortran myprog f llapack lblas Highly optimized assembly coded versions of BLAS and certain FFT routines may be available for your platform In some cases these are shipped with the PGI compilers See the current release notes for the PGI compilers you are using to determine if these optimized libraries exist where they can be downloaded if necessary and how to incorporate them into your installation as the default 10 10 Linking with SCaLAPACK The ScaLAPACK libraries are automatically installed with each MPI library version which accompanies a PGI installation You can link with the ScaLAPACK libraries by specifying Mscalapack on any of the PGI compiler command lines For example 9 mpif90 myprog f Mscalapack A pre built version of the BLAS library is automatically added when the Mscalapack switch is specified If you wish to use a different BLAS library and still use the Mscalapack switch then you can list the set of libraries explicitly on your link line Alternately you can copy your BLAS library into SPGI linux86 15 10 1ib li
125. cc host data clause structured block clause use_device list Fortran Syntax acc host data clause clause structured block acc end host data Loop Directive collapse n gang gang arg list worker num int expr vector length int expr seq independent private list reduction operator list shortloop C Syntax pragma acc loop clause clause MEO OT Fortran Syntax acc loop clause clause do loop Combined Parallel Loop Directive Any clause that is allowed on a parallel directive or a loop directive is allowed on a combined parallel loop directive C Syntax pragma acc parallel loop clause To OOD Fortran Syntax acc parallel loop clause do loop Combined Kernels Loop Directive Any clause that is allowed on a kernels directive or a loop directive is allowed on a combined kernels loop directive PGI Compiler User s Guide clause eram cS CU GWseg m Using an Accelerator Has this functionality new line Makes the address of the device data available on the host CLAWS a5 new line Describes what type of parallelism to use to execute the loop and declare loop private variables and arrays and reduction operations Applies to a loop which must appear on the following line new line Is a shortcut for specifying a loop directive nested immediately inside an accelerator parallel directive The mean
126. ccurs in three phases gt Collection Create a summary of each function or procedure collecting the useful information for interprocedural optimizations This is done during the compile step if the Mipa switch is present on the command line summary information is collected and stored in the object file gt Propagation Process all the object files to propagate the interprocedural summary information across function and file boundaries This is done during the link step when all the object files are combined if the Mipa switch is present on the link command line gt Recompile Optimization Recompile each of the object files with the propagated interprocedural information producing a specialized object file This process is also performed during the link step when the Mipa switch is present on the link command line PGI Compiler User s Guide 44 Optimizing and Parallelizing When linking with Mipa the PGI compilers automatically regenerate IPA optimized versions of each object file essentially recompiling each file If there are IPA optimized objects from a previous build the compilers will minimize the recompile time by reusing those objects if they are still valid They will still be valid if the IPA optimized object is newer than the original object file and the propagated IPA information for that file has not changed since it was optimized After each object file has been recompiled the regular linker is invoked to build th
127. cedural Analysis Mipa fast is the recommended option to get best performances for global optimization You can also add the suboption inline to enable automatic global inlining across files You might consider using Mipa fast inline This option for interprocedural analysis and global optimization can improve performance You may also obtain further performance improvements by experimenting with the M lt pgflag gt options described in the section M Options by Category section of the PGI Compiler s Reference Guide These options include but are not limited to Mvect Munroll Minline Mconcur Mpfi and Mpfo However performance improvements using these options are typically application and system dependent It is important to time your application carefully when using these options to ensure no performance degradations occur For more information on optimization refer to Optimizing and Parallelizing For specific information about these options refer to the Optimization Controls section of the PGI Compiler s Reference Guide 2 4 Targeting Multiple Systems Using the tp Option The tp option allows you to set the target architecture By default the PGI compiler uses all supported instructions wherever possible when compiling on a given system As a result executables created on a given system may not be usable on previous generation systems For example executables created on a Pentium 4 may fail to execute
128. chnology Programming Languages Fortran Geneva 1991 Fortran 90 PGI Compiler User s Guide Xii Preface gt ISO IEC 1539 1 1997 Information technology Programming Languages Fortran Geneva 1997 Fortran 95 gt ISO IEC 1539 1 2004 Information technology Programming Languages Fortran Geneva 2004 Fortran 2003 gt ISO IEC 1539 1 2010 Information technology Programming Languages Fortran Geneva 2010 Fortran 2008 gt Fortran 95 Handbook Complete ISO ANSI Reference Adams et al The MIT Press Cambridge Mass 1997 gt The Fortran 2003 Handbook Adams et al Springer 2009 gt OpenMP Application Program Interface Version 3 1 July 2011 http www openmp org gt Programming in VAX Fortran Version 4 0 Digital Equipment Corporation September 1984 gt IBM VS Fortran IBM Corporation Rev GC26 4119 gt Military Standard Fortran DOD Supplement to American National Standard Programming Language Fortran ANSI x 3 1978 MIL STD 1753 November 9 1978 gt American National Standard Programming Language C ANSI X3 159 1989 gt ISO IEC 9899 1999 Information technology Programming Languages C Geneva 1999 C99 gt ISO IEC 9899 2011 Information Technology Programming Languages C Geneva 2011 C11 gt ISO IEC 14882 2011 Information Technology Programming Languages C Geneva 2011 C11 Organization Users typically begin by wanting to know
129. code keep Keep the kernel files kepler Generate code for Kepler Architecture equivalent to NVIDIA compute capability 3 x maxregcount n Specify the maximum number of registers to use on the GPU Leaving this blank indicates no limit no fma Control generation of fused multiply add instructions noL1 Prevent the use of L1 hardware data cache to cache global variables no rdc Control generation of relocatable device code default on host Generate host code for accelerator regions radeon Select AMD Radeon GPU as the accelerator target This option has a number of suboptions keep Keep the kernel files spectre Generate code for the Radeon Spectre architecture tahiti Generate code for the Radeon Tahiti architecture buffercount n Set the number of OpenCL buffers to use default is 3 The compiler automatically invokes the necessary software tools to create the kernel code and embeds the kernels in the object file To access accelerator libraries you must link an accelerator program with the t a flag 7 14 Profiling Accelerator Kernels This release supports the environment variable PGI ACC TIME for 64 bits Setting this environment variable to a nonzero value enables collection and printing of simple timing information about the accelerator regions and generated kernels Turn off all CUDA Profilers NVIDIA s Visual Profiler NVPROF CUDA PROFILE etc when enabling PGI ACC TIME they use the same library to gather perform
130. compound statement a sequence of statements beginning with and ending with that has a single entry and a single exit No statement or compound statement is a C C structured block if there is a jump into or out of that statement 9 9 Directive and Pragma Recognition The compiler option mp enables recognition of the parallelization directives and pragmas The use of this option also implies Mreentrant Local variables are placed on the stack and optimizations such as Mnoframe that may result in non reentrant code are disabled Miomutex For directives critical sections are generated around Fortran I O statements For pragmas calls to I O library functions are system dependent and are not necessarily guaranteed to be thread safe I O library calls within parallel regions should be protected by critical regions to ensure they function correctly on all systems 5 6 Directive and Pragma Summary Table The following table provides a brief summary of the directives and pragmas that PGI supports In the table the values in uppercase letters are Fortran directives while the names in lowercase letters are C C pragmas PGI Compiler User s Guide 62 Using OpenMP 5 6 1 Directive and Pragma Summary Table Table 11 Directive and Pragma Summary Table Fortran Directive and C Pragma Description ATOMIC TYPE END ATOMIC Semantically equivalent to enclosing a single statement in the CRITCIAL END and atomic CRITICAL dir
131. considered for parallelization even if they contain calls to user defined subroutines or functions or if their loop counts do not exceed usual thresholds Do don t enable auto concurrentization of loops depchk nodepchk Do don t ignore potential data dependencies eqvchk noeqvchk Do don t check EQUIVALENCE for data dependencies fcon nofcon Do don t assume unsuffixed real constants are single precision invarif noinvarif Do don t remove invariant if constructs from loops Ignore potential data dependencies PGI Compiler User s Guide 111 Using Directives and Pragmas Directive or Fortran C C pragma Functionality Default Scope Scope safe lastval Parallelize when loop contains a scalar used not enabled L L outside of loop safeptr nosafeptr Do don t ignore potential data dependencies to sadi al aci pointers single nosingle nosingle single nosingle Doldon t convert float parameters to double t convert float Doldon t convert float parameters to double to double EN la specified targets 9 4 Scope of Fortran Directives and Command Line Options During compilation the effect of a directive may be to either turn an option on or turn an option off Directives apply to the section of code following the directive corresponding to the specified scope which may include the following loop the following routine or the rest of the program This section presents several examples that
132. constraints Table 35 Simple Constraints Constraint Description Whitespace characters are ignored An immediate floating point operand EN wvemenpsergremyrmememep wien p o emememmewemt 0 uc Same as i LEE Same as m PGI Compiler User s Guide 181 C C Inline Assembly and Intrinsics Constraint Description An operand that is a valid memory address The expression associated with the constraint is expected to evaluate to an address for example p amp x A general purpose register operand 0 1 2 9 Matching Constraint See Output Operands for a description The following example uses the general or g constraint which allows the compiler to pick an appropriate constraint type for the operand the compiler chooses from a general purpose register memory or immediate operand This code lets the compiler choose the constraint type for y void example9 RIM 3 YAn asm movl 1 0 n r 3 s g yw De This technique can result in more efficient code For example when compiling example9 the compiler replaces the load and store of y with a constant 2 The compiler can then generate an immediate 2 for the y operand in the example The assembly generated by pgcc for our example is as follows example9 xa DED pushq rbp ao DEELS movq rsp rbp 5 WIGNER Ig 5 o HNL S lineno 3 movl 2 eax lineno 6 popq rbp fet In this example notice the use of 2 for the y op
133. constructed by pre pending an underscore converting to all lower case and appending an underscore Byte counts for character strings appear in sequence following the last argument in the argument list The following is an example of the pseudocode for the work subroutine call using UNIX conventions call work addr ERR addr a addr b addr n 3 13 9 3 Using the Default Calling Convention The Default calling convention is used if no directives are inserted to modify calling conventions and if neither the Miface unix or Munix compilation flag is used Refer to Symbol Name Construction and Calling Example for a complete description of the Default calling convention 13 9 4 Using the STDCALL Calling Convention Using the STDCALL calling convention requires the insertion of a compiler directive into the declarations section of any Fortran program unit which calls the STDCALL program unit You cannot mix UNIX style argument passing and STDCALL calling conventions within the same file In the following example syntax for the directive work is the name of the subroutine to be called using STDCALL conventions DEC ATTRIBUTES STDCALL work PGI Compiler User s Guide 166 Inter language Calling You can list more than one subroutine separating them by commas Refer to Symbol Name Construction and Calling Example for a complete description of the implementation of STDCALL gt The directive prefix DEC requ
134. ctions in C C ssssseseeeeneeeenne nnne tenentes 120 10 2 Using System Library Routines sssseeeeene ener 121 10 3 Creating and Using Shared Object Files on Linux senes 121 10 3 1 Procedure to create a use a shared object file ssssessseseeenns 121 10 3 2 Idd COmmalnd td ertet retenti ti d ene tb od re c c ete den RI b P ee 122 10 4 Creating and Using Dynamic Libraries on Mac OS X 123 10 5 PGI Runtime Libraries on WindoOws sse tenente tente tenens 123 10 6 Creating and Using Static Libraries on Windows ssseeeeenennennnnnnnn 124 CoTr commande E A 124 SYM EEE EEE AE EE TE TEE TE PATNE E AE TETEE E E A TATAE re ttc 124 e siaina N ead ANEDE Naaa e a Faaa 124 10 6 2ranlib COMMANG a apetece di cade E ie anaes dati 125 SY e MM A 125 OpllONS t 125 10 7 Creating and Using Dynamic Link Libraries on Windows seennnnnn 125 10 1 1 Build a DLL Fortra iii ctore ct rette ve ete vec teen o et eo at 127 10 7 2 Build a DUL uisi ate e etel terc fee E RE Red RH CHEER HEBR vd 128 10 7 3 Build DLLs Containing Circular Mutual Imports C seen 129 10 7 4 Build DLLs Containing Mutual Imports Fortran seeeeeennennnnns 131 10 7 5 Import a Fortran module from a DLL
135. d reference information on commands and graphical interfaces gt Release Notes Select this option to display the latest PGI Server and Workstation Release Notes This document describes changes between previous releases and the current release gt User s Guide Select this option to display the PGI User s Guide This document provides operating instructions for the PGI command level development environment as well as details concerning the PGI compilers interpretation of the Fortran language implementation of Fortran language extensions and command level compilation Licensing Submenu From the Licensing menu you have access to the PGI License Agreement and an automated license generating tool gt Generate License Select this option to display the PGI License Setup dialog that walks you through the steps required to download and install a license for PGI Workstation or PGI Server To complete this process you need an internet connection gt License Agreement Select this option to display the license agreement that is associated with use of PGI software 1 7 3 PGI on the Windows Desktop By default a PGI Workstation installation creates a shortcut on the Windows desktop This shortcut launches a Cygwin bash shell if Cygwin is installed otherwise it launches a Microsoft command shell The environment for this shell is pre configured to use PGI compilers and tools On 64 bit systems the 64 bit compilers are targeted while
136. d with this option it must also be used when linking Slightly slower execution Is implicit with mcmodel medium Can be used with option mcmodel small Win64 does not support Mlarge arrays for static objects larger than 2GB Dynamic linking restricted to a 32 bit offset External symbol references should refer to other shared lib routines rather than the program calling them Users should take care to explicitly declare INTEGER functions as INTEGER 4 The following table summarizes the limits of these programming models under the specified conditions The compiler options you use vary by processor PGI Compiler User s Guide 170 Programming Considerations for 64 Bit Environments Table 33 Effects of Options on Memory and Array Sizes Addr Math Max Size Gbytes Condition A l AS DS 32 bit linux86 programs 64 bit addr limited by option mcmodel small fpic incompatible with mcmodel medium Enable full support for 64 bit data addressing I Address Type size in bits of data used for address calculations 32 bit or 64 bit Index Arithmetic bit size of data used to index into arrays and other aggregate data structures If 32 bit total range of any single data object is limited to 2GB Maximum Array Size the maximum size in gigabytes of any single data object max size in gigabytes combined of all data objects in bss Maximum Total Size max size in gigabytes in aggregate of all executable code and data
137. d computing using either a low level model such as CUDA Fortran or a high level model such as the PGI Accelerator model or OpenACC to target a many core GPU or other attached accelerator The first two types of parallel programs are collectively referred to as SMP parallel programs On a single silicon die today s CPUs incorporate two or more complete processor cores functional units registers level 1 cache level 2 cache and so on These CPUs are known as multi core processors For purposes of threads or OpenMP parallelism these cores function as two or more distinct processors However the processing cores are on a single chip occupying a single socket on a system motherboard For purposes of PGI software licensing a multi core processor is treated as a single CPU 1 6 1 Run SMP Parallel Programs When you execute an SMP parallel program by default it uses only one processor To run on more than one processor set the NCPUS environment variable to the desired number of processors For information on how to set environment variables refer to Setting Environment Variables If you set NCPUS to a number larger than the number of physical processors your program may execute very slowly 1 7 Platform specific considerations The following list are the platforms supported by the PGI Workstation and PGI Server compilers and tools gt 32 bit Linux supported on 32 bit Linux operating systems running on either a 32 bit x86 co
138. d for output operands only The output operand is initialized with its expression before the first instruction in the asm statement If specified the must appear as the first character of the constraint string may be modified before the asm statement finishes using all of the input operands The compiler will not place this operand in a register that may be used as an input operand or part of any memory address Characters following a up to the first comma if present are to be ignored in the constraint The character that follows the is to be ignored in the constraint A constraint or an alternative constraint as defined in Multiple Alternative Constraints containing an amp indicates that the output operand is an early clobber operand This type operand is an output operand that The and modifiers apply to the operand regardless of the number of alternatives in the constraint string For example the in the output operand of example13 appears once and PGI Compiler User s Guide 185 C C Inline Assembly and Intrinsics applies to both alternatives in the constraint string The amp and modifiers apply only to the alternative in which they appear Normally the compiler assumes that input operands are used before assigning results to the output operands This assumption lets the compiler reuse registers as needed inside the asm statement However if the asm statement does not follow this c
139. d subroutines in Fortran C and C gt Naming and case conversion conventions gt Compatible data types gt Argument passing and special return values gt Arrays and indexes gt Win32 calling conventions The sections Inter language Calling Considerations through Example C Calling Fortran describe how to perform inter language calling using the Linux Mac OSX or Win64 convention Default Fortran calling conventions for Win32 differ although Win32 programs compiled using the Miface unix Fortran command line option use the Linux Win64 convention rather than the default Win32 conventions All information in those sections pertaining to compatibility of arguments applies to Win32 as well For details on the symbol name and argument passing conventions used on Win32 platforms refer to Win32 Calling Conventions PGI Compiler User s Guide 153 Inter language Calling 13 2 Inter language Calling Considerations In general when argument data types and function return values agree you can call a C or C function from Fortran as well as call a Fortran function from C or C When data types for arguments do not agree you may need to develop custom mechanisms to handle them For example the Fortran COMPLEX type has a matching type in C99 but does not have a matching type in C89 however it is still possible to provide inter language calls but there are no general calling conventions for such cases gt Ifa C function conta
140. d with libnuma the differences between systems is masked by the different versions of 1ibnuma When a program is deployed to the target system the proper set of libraries real or stub should be deployed with the program This facility requires that the program be dynamically linked with 1ibnuma 3 14 1 libnuma Not all systems have 1ibnuma Typically only numa systems have this library PGI supplies a stub version of 1ibnuma which satisfies the calls from the PGI runtime to 1ibnuma libnuma is a shared library that is linked dynamically at runtime The reason to have a numa library on all systems is to allow multi threaded programs such as programs compiled with Mconcur or mp to be compiled linked and executed without regard to whether the host or target systems has a numa library When the numa library is not available a multi threaded program still runs because the calls to the numa library are satisfied by the PGI stub library During installation the installation procedure checks for the existence of a real 1ibnuma among the system libraries If the real library is not found the PGI stub version is substituted PGI Compiler User s Guide 50 Chapter 4 USING FUNCTION INLININC Function inlining replaces a call to a function or a subroutine with the body of the function or subroutine This can speed up execution by eliminating parameter passing and function subroutine call and return overhead It also allows the compil
141. e clause cSacc directive name clause clause Sace directive nane clause cllause Rules In addition to the general directive rules the following rules apply to Accelerator Fixed Form Fortran directives gt The sentinel acc c acc or acc must occupy columns 1 5 gt Fixed form line length white space continuation and column rules apply to the directive line gt Initial directive lines must have a space or zero in column 6 and continuation directive lines must have a character other than a space or zero in column 6 Comments may appear on the same line as a directive starting with an exclamation point on or after column 7 and continuing to the end of the line gt Directives are case insensitive gt Directives cannot be embedded within continued statements gt Statements must not be embedded within continued directives PGI Compiler User s Guide 90 Using an Accelerator 7 9 6 OpenACC Directive Summary PGI currently supports these types of accelerator directives Parallel Directive Kernels Directive Loop Directive Combined Directive Data Directive Enter Data and Exit Data Directives Host_Data Directive Cache Directive Declare Directive Update Directive Routine Directive Wait Directive Table 17 lists and briefly describes each of the accelerator directives that PGI currently supports For a complete description of each directive refer to PGI Accelerator Directives in th
142. e appropriate include and library paths to use the MS MPI headers and libraries To compile for debugging add g To build an application that generates MPI profile data use the Mprof msmpi option This option performs MPICH style profiling for Microsoft MPI Using this option implies the option Mmpi msmpi The profile data generated by running an application built with the option Mprof msmpi contains information about the number of sends and receives as well as the number of bytes sent and received correlated with the source location associated with the sends and receives You must use Mprof msmpi in conjunction with either Mprof func or Mprof lines 6 9 Using Open MPI on Linux PGI CDK users can download a PGI built version of Open MPI for Linux that includes everything required to compile execute profile and debug MPI programs using Open MPI To build an application using Open MPI use the Open MPI compiler wrappers mpicc mpic mpif77 and mpif90 These wrappers automatically set up the compiler commands with the correct include file search paths library directories and link libraries To build an application using Open MPI for debugging add g to the compiler wrapper command line arguments PGI Compiler User s Guide 78 Using MPI To build an application that generates MPI profile data suitable for use with PGPROF use the Open MPI compiler wrappers with the Mprof func Mprof lines or Mprof time option PGI ha
143. e e riiez pgcc o a out filel o file2 o file3 o oe oe oe The pgcc driver invokes the compiler and assembler as required to process each source file and invokes the linker for the final link command If you modify one of the source files the executable can be rebuilt by compiling just that file and then relinking oe gee e idle paee 9 e oe dex hel o rileg LlG 3 9 3 Building a Program Without IPA Using Make The program compilation and linking process can be simplified greatly using the make utility on systems where it is supported Suppose you create a makefile containing the following lines oe a out filel o file2 o file3 o oc NESSUN oM owes edes OMEN 2 OMEN I GNO is ded gs deste ue pec Q Orr e iulgd lg PAWS 0 9 iile e o eS SOL e reale oe ELIAS deLILe SG OPT c file3 c It is then possible to type a single make command make The make utility determines which object files are out of date with respect to their corresponding source files and invokes the compiler to recompile only those source files and to relink the executable If you subsequently edit one or more source files the executable can be rebuilt with the minimum number of recompilations using the same single make command 3 9 4 Building a Program with IPA Interprocedural analysis and optimization IPA by the PGI compilers alters the standard and make utility command level interfaces as little as possible IPA o
144. e preprocessed filename s indicates an assembly language file filename o Linux and OS X indicates an object file filename obj Windows systems only indicates an object file filename a Linux and OS X indicates a library of object files filename lib Windows systems only indicates a statically linked library of object files or an import library filename so Linux only indicates a library of shared object files filename dll Windows systems only indicates a dynamically linked library filename dylib OS X systems only indicates a dynamically linked library The driver passes files with s extensions to the assembler and files with o obj so dll a and lib extensions to the linker Input files with unrecognized extensions or no extension are also passed to the linker Files with a F Capital F or FOR suffix are first preprocessed by the Fortran compilers and the output is passed to the compilation phase The Fortran preprocessor functions like cpp for C programs but is built in to the Fortran compilers rather than implemented through an invocation of cpp This design ensures consistency in the preprocessing step regardless of the type or revision of operating system under which you are compiling Any input files not needed for a particular phase of processing are not processed For example if on the command line you specify an assembly language file filename s and the S option to
145. e they apply to the entire routine as if they had been placed at the beginning of the routine 9 6 Prefetch Directives and Pragmas Today s processors are so fast that it is difficult to bring data into them quickly enough to keep them busy Prefetch instructions can increase the speed of an application substantially by bringing data into cache so that it is available when the processor needs it When vectorization is enabled using the Mvect or Mprefetch compiler options or an aggregate option such as fast that incorporates Mvect the PGI compilers selectively emit instructions to explicitly prefetch data into the data cache prior to first use You can control how these prefetch instructions are emitted by using prefetch directives and pragmas For a list of processors that support prefetch instructions refer to the PGI Release Notes PGI Compiler User s Guide 115 Using Directives and Pragmas 9 6 1 Prefetch Directive Syntax in Fortran The syntax of a prefetch directive is as follows cSmem prefetch lt varl gt lt var2 gt where lt varn gt is any valid variable member or array element reference 9 6 2 Prefetch Directive Format Requirements The sentinel for prefetch directives is c mem which is distinct from the cpgiS sentinel used for optimization directives Any prefetch directives that use the cpgi sentinel are ignored by the PGI compilers gt The c must be in column 1 gt Either or is al
146. e Common Compiler Feedback Format CCFF For more information on PGPROF refer to the PGPROF Profiler Guide 3 3 Common Compiler Feedback Format CCFF Using the Common Compiler Feedback Format CCFF PGI compilers save information about how your program was optimized or why a particular optimization was not made in the executable file To append this information to the object file use the compiler option Minfo ccff If you choose to use PGPROF to aid with your optimization PGPROF can extract this information and associate it with source code and other performance data allowing you to view all of this information simultaneously in one of the available profiler panels 3 4 Local and Global Optimization This section describes local and global optimization 3 4 1 Msafeptr The Msafeptr option can significantly improve performance of C C programs in which there is known to be no pointer aliasing For obvious reasons this command line option must be used carefully There are a number of suboptions for Msafeptr gt Msafeptr all All pointers are safe Equivalent to the default setting Msafeptr gt Msafeptr arg Function formal argument pointers are safe Equivalent to Msafeptr dummy gt Msafeptr global Global pointers are safe PGI Compiler User s Guide 28 Optimizing and Parallelizing gt Msafeptr local Local pointers are safe Equivalent to Msafeptr auto gt Msafeptr stati
147. e PGI Compiler s Reference Guide Table 17 PGI Accelerator Directive Summary Table This directive Accepts these clauses Has this functionality Parallel Directive Defines the region of the program that should be compiled for async int expr J parallel execution on the accelerator device wait int expr list num gangs int expr num workers int expr vector length int expr if condition reduction operator list copy list copyin list copyout list create list present list deviceptr list private list firstprivate list C Syntax pragma acc parallel clause clause new line SGU ete ClO EO Gtk Fortran Syntax acc parallel clause clause structured block acc end parallel PGI Compiler User s Guide 91 Using an Accelerator This directive Accepts these clauses Has this functionality Kernels Directive Defines the region of the program that should be compiled into a sequence of kernels for execution on the accelerator device async int expr wait int expr list if condition copy list copyin list copyout list create list present list deviceptr list C Syntax pragma acc kernels clause clause structured block Fortran Syntax acc kernels clause clause Sreiieicwaerel Joilorel lt acc end kernels Data Directive Defines data typically arrays that should be allocat
148. e application with the IPA optimized object files The IPA optimized object files are saved in the same directory as the original object files for use in subsequent program builds 3 9 5 Building a Program with IPA Single Step By adding the Mi pa command line switch several source files can be compiled and linked with interprocedural optimizations with one command 9 pgcc Mipa fast o a out filel c file2 c file3 c Just like compiling without Mi pa the driver executes several steps to produce the assembly and object files to create the executable oe jose gse ss SS o zilelis zilele as O ied rileks pocek Mipa raS ti fS go ixle25 e i 1e2 6 as o file2 0 file2 s pocek Mipa raste S9 o iiless unless as g iulesSs3go Enles s pgcc Mipa fast o a out filel o file2 o file3 o oe oe oe oe oe oe In the last step an IPA linker is invoked to read all the IPA summary information and perform the interprocedural propagation The IPA linker reinvokes the compiler on each of the object files to recompile them with interprocedural information This creates three new objects with mangled names EME WSS aout 000 Slez ipao a ONE GOO Eiles USD Els 0UT CG0 The system linker is then invoked to link these IPA optimized objects into the final executable Later if one of the three source files is edited the executable can be rebuilt with the same command line 9 pgcc Mipa fast o a out filel c file2 c file3 c
149. e arrays 9 8 3 Sample Usage of IGNORE TKR Directive Consider this subroutine fragment subroutine example A B C D DIRS IGNORE TKR A R B TK C K D Table 25 indicates which rules are ignored for which dummy arguments in the preceding sample subroutine fragment Table 25 IGNORE TKR Example Dummy Argument Ignored Rules Notice that no letters were specified for A so all type kind and rank rules are ignored 9 9 IDEC Directives PGI Fortran compilers for Microsoft Windows support several de facto standard Fortran directives that help with inter language calling and importing and exporting routines to and from DLLs PGI Compiler User s Guide 118 Using Directives and Pragmas 9 9 1 IDEC Directive Syntax These directives all take the form DECS directive 9 9 2 Format Requirements You must follow the following format requirements for the directive to be recognized in your program gt The directive must begin in column 1 when the file is fixed format or compiled with Mfixed gt The directive prefix DECS requires a space between the prefix and the directive keyword such as ATTRIBUTES gt The must begin the prefix when compiling Fortran 90 95 free form format gt The characters C or can be used in place of in either form of the prefix when compiling F77 style fixed form format gt The directives are completely case insensitive 9 9 3 Summary Table The foll
150. e atomic action is writing a value 5 8 Runtime Library Routines User callable functions are available to the programmer to query and alter the parallel execution environment Any C C program unit that invokes these functions should include the statement include lt omp h gt The omp h include file contains definitions for each of the C C library routines and the required type definitions For example to use the omp get num threads function use this syntax include lt omp h gt int omp get num threads void Unlimited OpenMP thread counts are available in all PGI configurations The number of threads is unlicensed in the OpenMP runtime libraries up to the hard limit of 256 threads The following table summarizes the runtime library calls cy The Fortran call is shown first followed by the equivalent C C call Table 13 Runtime Library Routines Summary Runtime Library Routines with Examples omp_get_num_threads Returns the number of threads in the team executing the parallel region from which it is called When called from a serial region this function returns 1 A nested parallel region is the same as a single parallel region By default the value returned by this function is equal to the value of the environment variable OMP NUM THREADS orto the value set by the last previous call to omp set num threads integer function omp get num threads l int omp get num threads void omp set num threads
151. e default for fast with 03 while maintaining all other elements of fast simply compile as follows pgfortran fast 03 prog f 3 5 Loop Unrolling using Munroll This optimization unrolls loops which reduces branch overhead and can improve execution speed by creating better opportunities for instruction scheduling A loop with a constant count may be completely unrolled or partially unrolled A loop with a non constant count may also be unrolled A candidate loop must be an innermost loop containing one to four blocks of code The following example shows the use of the Munroll option pgfortran Munroll prog f The Munroll option is included as part of fast on all x86 and x64 targets The loop unroller expands the contents of a loop and reduces the number of times a loop is executed Branching overhead is reduced when a loop is unrolled two or more times since each iteration of the unrolled loop corresponds to two or more iterations of the original loop the number of branch instructions executed is proportionately reduced When a loop is unrolled completely the loop s branch overhead is eliminated altogether Loop unrolling may be beneficial for the instruction scheduler When a loop is completely unrolled or unrolled two or more times opportunities for improved scheduling may be presented The code generator can take advantage of more possibilities for instruction grouping or filling instruction delays found within the loop Ex
152. e of the issues to consider include gt Runtime Libraries gt 64 bit Linux Systems gt Redistribution of Files gt Licensing 12 1 1 Runtime Library Considerations On Linux systems the system runtime libraries can be linked to an application either statically or dynamically For example for the C runtime library 1ibc you can use either the static version libc a or the shared object version 1ibc so If the application is intended to run on Linux systems other than the one on which it was built it is generally safer to use the shared object version of the library This approach ensures that the application uses a version of the library that is compatible with the system on which the application is running Further it works best when the application is linked on a system that has an equivalent or earlier version of the system software than the system on which the application will be run LJ Building on a newer system and running the application on an older system may not produce the desired output To use the shared object version of a library the application must also link to shared object versions of the PGI runtime libraries To execute an application built in such a way on a system PGI Compiler User s Guide 148 Distributing Files Deployment on which PGI compilers are not installed those shared objects must be available To build using the shared object versions of the runtime libraries use the Bdynamic option as shown h
153. e option Pragmas may also toggle an option selectively enabling and disabling the option The general syntax of a pragma is pragma scope pragma body The optional scope field is an indicator for the scope of the pragma some pragmas ignore the scope indicator The valid scopes are global indicates the pragma applies to the entire source file routine indicates the pragma applies to the next function loop indicates the pragma applies to the next loop but not to any loop contained within the loop body Loop scoped pragmas are only applied to for and while loops If a scope indicator is not present the default scope if any is applied Whitespace must appear after the pragma keyword and between the scope indicator and the body of the pragma Whitespace may also surround any special characters such as a comma or an equal sign Case is significant for the names of the pragmas and any variable names that appear in the body of the pragma 9 3 PGI Proprietary Optimization Directive and Pragma Summary The following table summarizes the supported Fortran directives and C C pragmas The following terms are useful in understanding the table gt Functionality is a brief summary of the way to use the directive or pragma For a complete description refer to the Directives and Pragmas Reference section of the PGI Compiler s Reference Guide gt Many of the directives and pragmas can be preceded by NO The default entry
154. e particular release The O or O2 level performs all level one local optimizations as well as global optimizations Control flow analysis is applied and global registers are allocated for all functions and subroutines Loop regions are given special consideration This optimization level is a good choice when the program contains loops the loops are short and the structure of the code is regular The PGI compilers perform many different types of global optimizations including but not limited to Branch to branch elimination Constant propagation Copy propagation Dead store elimination Global register allocation Induction variable elimination Invariant code motion You can explicitly select the optimization level on the command line For example the following command line specifies level two optimization which results in global optimization pgfortran 02 prog f PGI Compiler User s Guide 30 Optimizing and Parallelizing The default optimization level changes depending on which options you select on the command line For example when you select the g debugging option the default optimization level is set to level zero O0 However if you need to debug optimized code you can use the gopt option to generate debug information without perturbing optimization For a description of the default levels refer to Default Optimization Levels The fast option includes 02 on all x86 and x64 targets If you want to override th
155. eaeaesasseeeeseassneasaeeseesatseeneaeeetees 144 Fortran and C C Data Type Compatibility seen 155 Fortran and C C Representation of the COMPLEX Type sssssssessssseee eene 156 Calling Conventions Supported by the PGI Fortran Compilers sss 165 64sbit Compiler Options Eoo re bre toe ee o i ae m b ob en teo ode te ander 170 Effects of Options on Memory and Array Sizes 171 64 Bit Limitations 55 c Rt epus 171 oimple Constralnts oit iE RR HUE R a E ayia ie ae aa HEURE RR ee 181 x06 x86 64 Machine Constraints 2 1e Neid tet ate See tete a coven oe Pe e rege d 183 Multiple Alternative Constralnts e ri ee te rre ie tete e edere b etr ie pidhudn 184 Constraint Modifier Characters sse tentent renes 185 Assembly String Modifier Characters seen 187 Intrinsic Header File Organization 1 2 tritt et lettre rt t ee te 190 PGI Compiler User s Guide xi PREFACE This guide is part of a set of manuals that describe how to use The Portland Group PGI Fortran C and C compilers and program development tools These compilers and tools include the PGF77 PGFORTRAN PGC and PGCC compilers the PGPROF profiler and the PGDBG debugger They work in conjunction with an x86 or x64 assembler and linker You can use the PGI compilers and tools to compile debug optimize and profile serial and parallel applications for x86
156. eater CPUs 12 3 2 Generating Code for a Specific Processor You can use the tp option to request that the compiler generate code optimized for a specific processor The PGI Release Notes contains a list of supported processors or you can look at the tp entry in the compiler output generated by using the help option described in help in the PGI Compiler Reference Manual 12 4 Generating One Executable for Multiple Types of Processors PGI unified binaries provide a low overhead method for a single program to run well on a number of hardware platforms All 64 bit PGI compilers can produce PGI Unified Binary programs that contain code streams fully optimized and supported for both AMD64 and Intel 64 processors using the tp target option The compilers generate and combine multiple binary code streams into one executable where each stream is optimized for a specific platform At runtime this one executable senses the environment and dynamically selects the appropriate code stream Executable size is automatically controlled via unified binary culling Only those functions and subroutines where the target affects the generated code have unique binary images resulting in a code size savings of 10 90 compared to generating full copies of code for each target Programs can use PGI Unified Binary technology even if all of the object files and libraries are not compiled as unified binaries Like any other object file you can use P
157. ective or critical pragma TYPE may be empty or one of the following UPDATE READ WRITE or CAPTURE The END ATOMIC directive is only allowed when ending ATOMIC CAPTURE regions o Only certain statements are allowed BARRIER and barrier Synchronizes all threads at a specific point in a program so that all threads complete work to that point before any thread continues CRITICAL END CRITICAL and Defines a subsection of code within a parallel region a critical section which is executed critical one thread at a time DO END DO and for Provides a mechanism for distribution of loop iterations across the available threads in a parallel region C DOACROSS Specifies that the compiler should parallelize the loop to which it applies even though that loop is not contained within a parallel region FLUSH and flush When this appears all processor visible data items or when a list is present FLUSH list only those specified in the list are written to memory thus ensuring that all the threads in a team have a consistent view of certain objects in memory MASTER END MASTER and Designates code that executes on the master thread and that is skipped by the other master threads ORDERED and ordered Defines a code block that is executed by only one thread at a time and in the order of the loop iterations this makes the ordered code block sequential while allowing parallel execution of statements outside the code block PARALLEL DO
158. ed This example shows the compilation informational messages and runtime results using the SSE instructions on a 4 Core Intel Sandybridge 2 5 GHz system along with issues that affect SSE performance Loops vectorized using SSE instructions operate much more efficiently when processing vectors that are aligned to a cache line boundary You can cause unconstrained data objects of size 16 bytes or greater to be cache aligned by compiling with the Mcache align switch An PGI Compiler User s Guide 34 Optimizing and Parallelizing unconstrained data object is a data object that is not a common block member and not a member of an aggregate data structure For stack based local variables to be properly aligned the main program or function must be compiled with Mcache align The Mcache align switch has no effect on the alignment of Fortran allocatable or automatic arrays If you have arrays that are constrained such as vectors that are members of Fortran common blocks you must specifically pad your data structures to ensure proper cache alignment You can use Mcache align for only the beginning address of each common block to be cache aligned The following examples show the results of compiling the sample code in Vector operation using SIMD instructions both with and without the option Mvect simd Vector operation using SIMD instructions program vector op parameter N 9999 real 4 x N y N z N W N doi 1 n yu i Z G
159. ed copy list in the device memory for the duration of the data region Py whether data should be copied from the host to the device copyin list memory upon region entry and copied from the device to copyout list host memory upon region exit create list deviceptr list if condition C Syntax pragma acc data clause clause new line structured block Fortran Syntax eacee cerca lelewa i causes structured block acc end data Enter Data Directive Defines data typically arrays that should be allocated in the device memory for the duration of the program or until an exit data regdirective that deallocates the data and whether data should be copied from the host to the device if condition async int expr wait int expr list copyin list memoryatthe enter data directive create list C Syntax pragma acc enter data clause clause new line FortranSyntax acc enter data clause clause Exit Data Directive Defines data typically arrays that should be allocated in the device memory and whether data should be copied from the device to the host memory if condition async int expr wait int expr list copyout list delete list PGI Compiler User s Guide 92 This directive Accepts these clauses C Syntax pragma acc exit data clause Fortran Syntax WaceWexieidatalllerausemy Host_Data Directive C Syntax pragma a
160. ed in Runtime Library Routines PGI provides libraries that export C interfaces by using Fortran modules On Windows PGI also provides additions to the supported library functionality for runtime functions included in DFLIB This section has examples that include the following options related to creating and using libraries Bdynamic def file implib file Mmakeimplib Bstatic dynamiclib l O fpic Mmakedll shared 10 1 Using builtin Math Functions in C C The name of the math header file is math h Include the math header file in all of your source files that use a math library routine as in the following example which calculates the inverse cosine of 3 5 include lt math h gt include lt stdio h gt define PI 3 1415926535 void main double x y ss ec den OR y acos 0 5 joresiniese Y Bie EEN Loos p PGI Compiler User s Guide 120 Creating and Using Libraries Including math h causes PGI C and C to use builtin functions which are much more efficient than library calls In particular if you include math h the following intrinsics calls are processed using builtins abs acosf asinf atan atan2 atan2f atanf COS cosf exp expf fabs fabsf fmax fmaxf fmin fminf log log10 log10f logf pow powf sin sinf sqrt sqrtf tan tanf 10 2 Using System Library Routines Release 15 10 of the PGI runtime libraries makes use of Linux system libraries to implement for example OpenMP and Fortran I
161. el region that is serialized PGI Compiler User s Guide 69 Using OpenMP Runtime Library Routines with Examples for example in the presence of an IF clause evaluating FALSE for directives and zero for pragmas the function returns FALSE for directives and zero for pragmas logical function omp in parallel int omp in parallel void omp set dynamic Allows automatic dynamic adjustment of the number of threads used for execution of parallel regions This function is recognized but currently has no effect subroutine omp set dynamic scalar logical exp void omp set dynamic int dynamic threads omp get dynamic Allows the user to query whether automatic dynamic adjustment of the number of threads used for execution of parallel regions is enabled This function is recognized but currently always returns FALSE for directives and zero for pragmas logical function omp get dynamic void omp get dynamic void omp set nested Allows enabling disabling of nested parallel regions Fortran subroutine omp set nested nested logical nested C C void omp set nested int nested omp get nested Allows the user to query whether dynamic adjustment of the number of threads available for execution of parallel regions is enabled Fortran logical function cmo get nested Y C C int omp get nested void omp set schedule Set the value of the run sched var subroutine omp set schedule kind
162. ence Manual 3 2 4 dryrun The dryrun option can be useful as a diagnostic tool if you need to see the steps used by the compiler driver to preprocess compile assemble and link in the presence of a given set of command line inputs When you specify the dr yrun option these steps are printed to standard error stderr but are not actually performed For example you can use this option to inspect the PGI Compiler User s Guide 27 Optimizing and Parallelizing default and user specified libraries that are searched during the link phase and the order in which they are searched by the linker 3 2 5 V The v option is similar to dryrun except each compilation step is performed 3 2 6 PGPROF PGPROF is a profiling tool that provides a way to visualize the performance of the components of your program Using tables and graphs PGPROF associates execution time and resource utilization data with the source code and instructions of your program This association allows you to see where a program s execution time is spent Through resource utilization data and compiler analysis information PGPROF helps you to understand why certain parts of your program have high execution times This information may help you with selecting which optimization options to use with your program PGPROF also allows you to correlate the messages produced by Minfo and Mneginfo described above to your program s source code This feature is known as th
163. er PGCC auto parallelizing accelerator enabled C11 and K amp R C compiler PGC auto parallelizing accelerator enabled C 11 compiler NVIDIA CUDA Toolkit components A simple command line tool to detect whether the system has an appropriate GPU or accelerator card v v v yvy v No accelerator enabled debugger is included with this release PGI Compiler User s Guide 81 Using an Accelerator 7 1 2 Availability The PGI 15 10 Fortran amp C Accelerator compilers are available only on x86 processor based workstations and servers with an attached NVIDIA CUDA enabled GPU or Tesla card These compilers target all platforms that PGI supports All examples included in this section are developed and presented on such a platform For a list of supported GPUs refer to the Accelerator Installation and Supported Platforms list in the latest PGI Release Notes 7 1 3 User directed Accelerator Programming In user directed accelerator programming the user specifies the regions of a host program to be targeted for offloading to an accelerator device The bulk of a user s program as well as regions containing constructs that are not supported on the targeted accelerator are executed on the host This section concentrates on specification of loops and regions of code to be offloaded to an accelerator 7 1 4 Features Not Covered or Implemented This section does not describe features or limitations of the host programming environment as a whole
164. er to optimize the function with the rest of the code Note that using function inlining indiscriminately can result in much larger code size and no increase in execution speed The PGI compilers provide two categories of inlining gt Automatic inlining During the compilation process a hidden pass precedes the compilation pass This hidden pass extracts functions that are candidates for inlining The inlining of functions occurs as the source files are compiled gt Inline libraries You create inline libraries for example using the pgfortran compiler driver and the o and Mextract options There is no hidden extract pass but you must ensure that any files that depend on the inline library use the latest version of the inline library There are important restrictions on inlining Inlining only applies to certain types of functions Refer to Restrictions on Inlining for more details on function inlining limitations This section describes how to use the following options related to function inlining Mextract Minline Mrecursive 4 1 Invoking Function Inlining To invoke the function inliner use the Minline option If you do not specify an inline library the compiler performs a special prepass on all source files named on the compiler command line before it compiles any of them This pass extracts functions that meet the requirements for inlining and puts them in a temporary inline library for use by the compilation pass
165. er value for example using set omp num threads will cause the number of processes or threads to be set at the value of NCPUS MAX rather than the value specified in the function call 11 3 14 NO STOP MESSAGE Ifthe NO STOP MESSAGE variable exists the execution of a plain STOP statement does not produce the message FORTRAN STOP The default behavior of the PGI Fortran compilers is to issue this message 11 3 15 PATH The PATH variable sets the directories that are searched for commands that the user types When using PGI products it is important that you set your PATH to include the location of the PGI products You can also use this variable to specify that you want to use only the linux86 version of the compilers and tools or to target linux86 as the default The following csh example targets x64 linux86 64 version of the compilers and tools set path opt pgi linux86 64 15 10 bin path 11 3 16 PGI The PGI environment variable specifies the root directory where the PGI compilers and tools are installed This variable is recognized at compile time If it is not set the default value depends on your system as well as which compilers are installed gt On Linux the default value of this variable is opt pgi gt On Windows the default value is C Program Files PGI where C represents the system drive If both 32 and 64 bit compilers are installed the 32 bit compilers are in C Program Files x86
166. erand Of course if y is always 2 then the immediate value may be used instead of the variable with the 1 constraint as shown here void examplel0 Suge Xp asm movl 1 0 n lp x wa 2 Na Compiling example10 with pgcc produces assembly similar to that produced for example9 15 2 6 Machine Constraints Another category of constraints is Machine Constraints The x86 and x86 64 architectures have several classes of registers To choose a particular class of register you can use the x86 x86 64 machine constraints described in Table 36 PGI Compiler User s Guide 182 C C Inline Assembly and Intrinsics Table 36 x86 x86 64 Machine Constraints Constraint Description a register e g al ax eax Yorax Specifies a or d registers This is used primarily for holding 64 bit integer values on 32 bit targets The d register holds the most significant bits and the a register holds the least significant bits b register e g 9obl bx ebx 9orbx c register e g Yocl cx ecx rcx Not supported d register e g dl dx edx Yordx di register e g 9odil di edi ordi Constant in range of Oxffffffff to Ox7fffffff Not supported Floating point constant in range of 0 0 to 1 0 Constant in range of 0 to 31 e g for 32 bit shifts Constant in range of 0 to 63 e g for 64 bit shifts Constant in range of 0 to 65535 Constant in range of 0 to 3 constant e g shift
167. erates SSE instructions Generates scalar SSE code with xmm registers implies Mf lushza Mcache align Aligns long objects on cache line boundaries Note On 32 bit systems if one file is compiled with the CMcache align option then all files should be compiled with it This is not true on 64 bit systems Mflushz Sets SSE to flush to zero mode PGI Compiler User s Guide 20 Use Command Line Options Use this option To do this M no vect Controls automatic vector pipelining For best performance on processors that support SSE instructions use the PGFORTRAN compiler even for FORTRAN 77 code and the fas t option To see the specific behavior of fast for your target use the following command pgfortran help fast 2 3 2 Other Performance Related Options While fast and fastsse are options designed to be the quickest route to best performance they are limited to routine boundaries Depending on the nature and writing style of the source code the compiler often can perform further optimization by knowing the global context of usage of a given routine For instance determining the possible value range of actual parameters of a routine could enable a loop to be vectorized similarly determining static occurrence of calls helps to decide which routine is beneficial to inline These types of global optimizations are under control of Interprocedural Analysis IPA in PGI compilers Option Mi pa enables Interpro
168. erator Directives PGI Accelerator compilers enable accelerator directives with the acc and ta command line option For more information on this option as it relates to the Accelerator refer to Applicable Command Line Options The syntax used to define directives allows compilers to ignore accelerator directives if support is disabled or not provided _ACCEL macro The ACCEL macro name is defined to have a value yyyymm where yyyy is the year and mm is the month designation of the version of the Accelerator directives supported by the implementation For example the version for May 2009 is 200905 The PGI compilers define this macro when accelerator directives are enabled OPENACC macro The OPENACC macro name is defined to have a value yyy ymm where yyyy is the year and mm is the month designation of the version of the OpenACC directives supported by the implementation For example the version for June 2013 is 201306 All OpenACC compilers define this macro when OpenACC directives are enabled 7 9 2 Format The specific format of the directive depends on the language and the format or form of the source Directives include a name and clauses and the format of the directive depends on the type gt C directives described in C Directives gt Free form Fortran directives described in Free Form Fortran Directives PGI Compiler User s Guide 88 Using an Accelerator gt Fixed form Fortran directives descr
169. ere pgf90 Bdynamic myprog f90 12 1 2 64 bit Linux Considerations On 64 bit Linux systems 64 bit applications that use the mcmode1l medium option sometimes cannot be successfully linked statically Therefore users with executables built with the momodel medium option may need to use shared libraries linking dynamically Also runtime libraries built using the pic option use 32 bit offsets so they sometimes need to reside near other runtime 1ibs in a shared area of Linux program memory If your application is linked dynamically using shared objects then the shared object versions of the PGI runtime are required 12 1 3 Linux Redistributable Files The method for installing the shared object versions of the runtime libraries required for applications built with PGI compilers and tools is manual distribution When the PGI compilers are installed there are directories that have a name that begins with REDIST for each platform linux86 and linux86 64 these directories contain the redistributed shared object libraries These may be redistributed by licensed PGI customers under the terms of the End User License Agreement 12 1 4 Restrictions on Linux Portability You cannot expect to be able to run an executable on any given Linux machine Portability depends on the system you build on as well as how much your program uses system routines that may have changed from Linux release to Linux release For example
170. ers appended to the end of the common block gt An argument mismatch exists in other words the number and type size of actual and formal parameters must be equal gt A name clash exists such as a call to subroutine xyz in the extracted subprogram and a variable named xyz in the caller The following types of C and C functions cannot be inlined gt Functions containing switch statements gt Functions which reference a static variable whose definition is nested within the function gt Functions which accept a variable number of arguments Certain C C functions can only be inlined into the file that contains their definition gt Static functions gt Functions which call a static function gt Functions which reference a static variable PGI Compiler User s Guide 56 Chapter 5 USING OPENMP The PGF77 and PGFORTRAN Fortran compilers support the OpenMP Fortran Application Program Interface The PGCC and PGC compilers support the OpenMP C C Application Program Interface OpenMP is a specification for a set of compiler directives an applications programming interface APD and a set of environment variables that can be used to specify shared memory parallelism in FORTRAN and C C programs OpenMP may be used to obtain most of the parallel performance you can expect from your code or it may serve as a stepping stone to parallelizing an entire application with MPI This section provides information on OpenMP as
171. es section of the PGI Compiler s Reference Guide PGI Compiler User s Guide 168 Programming Considerations for 64 Bit Environments 14 1 1 C C Data Types On 32 bit Windows int is 4 bytes long is 4 bytes and pointers are 4 bytes On 64 bit windows the size of an int is 4 bytes a long is 4 bytes and a pointer is 8 bytes On the 32 bit Linux and Mac OS X operating systems the size of an int is 4 bytes a long is 4 bytes and a pointer is 4 bytes On the 64 bit Linux and Mac OS X operating systems the size of an int is 4 bytes a long is 8 bytes and a pointer is 8 bytes 14 1 2 Fortran Data Types In Fortran the default size of the INTEGER type is 4 bytes The i8 compiler option may be used to make the default size of all INTEGER data in the program 8 bytes When using the Mlarge arrays option described in 64 Bit Array Indexing any 4 byte INTEGER variables that are used to index arrays are silently promoted by the compiler to 8 bytes This promotion can lead to unexpected consequences so 8 byte INTEGER variables are recommended for array indexing when using the option Mlarge arrays 14 2 Large Static Data in Linux Linux86 64 operating systems support two different memory models The default model used by PGI compilers is the small memory model which can be specified using mcmodel small This is the 32 bit model which limits the size of code plus statically allocated data including system and user libraries to
172. f useful mathematical routines optimized for AMD processors PGI Compiler User s Guide 10 Getting Started gt CUDA Fortran Reference Select this option to display the CUDA Fortran Programming Guide and Reference This document describes CUDA Fortran a small set of extensions to Fortran that support and build upon the CUDA computing architecture gt Fortran Language Reference Select this option to display the PGI Fortran Reference This document describes The Portland Group s implementation of the FORTRAN 77 and Fortran 90 95 languages and presents the Fortran language statements intrinsics and extension directives gt Installation Guide Select this option to display the PGI Server and Workstation Installation Guide This document provides an overview of the steps required to successfully install and license PGI Server and PGI Workstation gt PGDBG Debugger Guide Select this option to display the PGDBG Debugger Guide This guide describes how to use the PGDBG debugger to debug serial and parallel applications built with PGI compilers It contains information about how to use PGDBG as well as detailed reference information on commands and graphical interfaces gt PGPROF Profiler Guide Select this option to display the PGPROF Profiler Guide This guide describes how to use the PGPROF profiler to tune serial and parallel applications built with PGI compilers It contains information about how to use the profiler as well as detaile
173. fied with the acc mechanism Syntax The syntax of directives in free form source files is Sacc directive name clause clause PGI Compiler User s Guide 89 Using an Accelerator Rules In addition to the general directive rules the following rules apply to OpenACC Free Form Fortran directives gt The comment prefix may appear in any column but may only be preceded by white space spaces and tabs gt The sentinel acc must appear as a single word with no intervening white space gt Line length white space and continuation rules apply to the directive line gt Initial directive lines must have a space after the sentinel gt Continued directive lines must have an ampersand amp as the last nonblank character on the line prior to any comment placed in the directive gt Comments may appear on the same line as the directive starting with an exclamation point and extending to the end of the line gt fthe first nonblank character after the sentinel is an exclamation point the line is ignored gt Directives are case insensitive gt Directives cannot be embedded within continued statements gt Statements must not be embedded within continued directives 7 9 5 Fixed Form Fortran Directives Fixed Form Accelerator directives are specified using one of three formats Syntax The syntax of directives in fixed form source files is one these three formats Sacc directive nane claus
174. file gt Maintains the inline library utils il gt Updates the library whenever you change utils f or one of the include files it uses gt Compiles parser f and alloc f whenever you update the library PGI Compiler User s Guide 54 Using Function Inlining Sample Makefile SRC mydir FC pgfortran FFLAGS 02 main o SRC main f SRC global h S FC S FFLAGS c SRC main f uestis SIC ucila re 9C global hms 97C utile hn S FC S S EEDAGS c SRC utils uEsLIS 3L 8 SEC wx S SINC oileloail m S SINC mess Im SI ORGOESSICEIBISAG S Mexc raice liom uisisits eiie COE wie I SIS parsenesor SRI I SC ENSE Iob de Us E FC FFLAGS Minline utils il c SRC parser f alloc SIS BS alloes S SRC alobal a wcilas ii S FC S FFLAGS Minline utils il c SRC alloc f myprog main o utils o parser o alloc o S FC o myprog main o utils o parser o alloc o 4 4 Error Detection during Inlining You can specify the Minfo inline option to request inlining information from the compiler when you invoke the inliner For example pgfortran Minline mylib il Minfo inline myext f 4 5 Examples Assume the program dh ry consists of a single source file dh ry f The following command line builds an executable file for dh ry in which proc is inlined wherever it is called pgfortran dhry f Minline proc7 The following command lines build an executable file for dh ry in which proc7 plus any function
175. fined for CCFLAGS but disables vectorization CCFLAGS c Mvect simd CCNOVECTFLAGS CCFLAGS Mnovect 2 2 Help with Command line Options If you are just getting started with the PGI compilers and tools it is helpful to know which options are available when to use them and which options most users find effective Using help The help option is useful because it provides information about all options supported by a given compiler PGI Compiler User s Guide 18 Use Command Line Options You can use help in one of three ways gt Use help with no parameters to obtain a list of all the available options with a brief one line description of each gt Add a parameter to help to restrict the output to information about a specific option The syntax for this usage is help command line option Suppose you use the following command to restrict the output to information about the fast option pgfortran help fast The output you see is similar to fast Common optimizations includes 02 Munroll c 1 Mnoframe Mlre In the following example we add the help parameter to restrict the output to information about the help command The usage information for help shows how groups of options can be listed or examined according to function pgfortran help help help groups asm debug language linker opt other overall phase prepro suffix switch target variable gt Add a parameter to help to restrict
176. for C C is identical between most compilers on Win32 Linux Mac OS X and Win64 However Fortran calling conventions vary widely between legacy Win32 Fortran compilers and Linux or Win64 Fortran compilers 13 9 1 Win32 Fortran Calling Conventions Four styles of calling conventions are supported using the PGI Fortran compilers for Win32 Default C STDCALL and UNIX gt Default Used in the absence of compilation flags or directives to alter the default gt Cor STDCALL Used if an appropriate compiler directive is placed in a program unit containing the call The C and STDCALL conventions are typically used to call routines coded in C or assembly language that depend on these conventions PGI Compiler User s Guide 164 Inter language Calling gt UNIX Used in any Fortran program unit compiled using the Miface unix or the Munix compilation flag The following table outlines each of these calling conventions Table 31 Calling Conventions Supported by the PGI Fortran Compilers Convention Default STDCALL C UNIX Leading underscore Yes Yes Yes Yes Trailing underscore Argument byte count added Arguments passed by reference Yes Character argument length passed After each char No End of argument argument list First character of character string is No Yes Yes passed by value Except arrays which are always passed by reference even in the STDCALL and C conventions While it is compatible with the Fortran implement
177. functions in this file generate no C exceptions allowing more optimal code generation lo Names the output file Specifies code optimization level where lt level gt is 0 1 2 3 or 4 tp target target Specify the target processor s for the 64 bit compilers more than one target is allowed and enables generation of PGI Unified Binary executables PGI Compiler User s Guide 22 Use Command Line Options Use this option To do this W1 option Compiler driver passes the specified options to the linker PGI Compiler User s Guide 23 Chapter 3 OPTIMIZING AND PARALLELIZING Source code that is readable maintainable and produces correct results is not always organized for efficient execution Normally the first step in the program development process involves producing code that executes and produces the correct results This first step usually involves compiling without much worry about optimization After code is compiled and debugged code optimization and parallelization become an issue Invoking one of the PGI compiler commands with certain options instructs the compiler to generate optimized code Optimization is not always performed since it increases compilation time and may make debugging difficult However optimization produces more efficient code that usually runs significantly faster than code that is not optimized The compilers optimize code according to the specified optimization level
178. g the standard set of directories This variable is useful when debugging a new library or using a nonstandard library for special purposes The following csh example adds the current directory to your LD LIBRARY PATH variable setenv LD LIBRARY PATH SLD LIBRARY PATH 11 3 5 LM LICENSE FILE The LM LICENSE FILE variable specifies the full path of the license file that is required for running the PGI software For example once the license file is in place you can execute the following csh commands to make the products you have purchased accessible and to initialize your environment for use of FLEXIm These commands assume that you use the default installation directory opt pgi setenv PGI opt pgi setenv LM LICENSE FILE SLM LICENSE FILE opt pgi license dat oe oe PGI Compiler User s Guide 139 Using Environment Variables To set the environment variable LM LICENSE FILE to the full path of the license key file do this 1 Open the System Properties dialog Start Control Panel System 2 Select the Advanced tab 3 Click the Environment Variables button gt fLM LICENSE FILE is not already an environment variable create a new system variable for it Set its value to the full path including the name of the license key file license dat gt fLM LICENSE FILE already exists as an environment vari
179. g this option If the compiler is not able to successfully auto parallelize your application you should refer to Using OpenMP It is possible that insertion of explicit parallelization directives or pragmas and use of the compiler option mp might enable the application to run in parallel 3 7 2 Loops That Fail to Parallelize In spite of the sophisticated analysis and transformations performed by the compiler programmers may notice loops that are seemingly parallel but are not parallelized In this subsection we look at some examples of common situations where parallelization does not occur Innermost Loops As noted earlier in this section the PGI compilers will not parallelize innermost loops by default because it is usually not profitable However you can override this default using the Mconcur innermost command line option PGI Compiler User s Guide 38 Optimizing and Parallelizing Timing Loops Often loops occur in programs that are similar to timing loops The outer loop in the following example is one such loop domt dX 2 Clo 5k e b 1 aj enddo enddo The outer loop in the preceding example is not parallelized because the compiler detects a cross iteration dependence in the assignment to a i Suppose the outer loop were parallelized Then both processors would simultaneously attempt to make assignments into a 1 n Now in general the values computed by each processor for a 1 n will differ so that
180. h including the name of the file for the license keys file gt If PGROUPD LICENSE FILE already exists as an environment variable append the path to the license file to the variable s current value using a semi colon to separate entries 11 3 23 PWD The PWD variable allows you to display the current directory 11 3 24 STATIC_RANDOM_SEED You can use STATIC RANDOM SEED to force the seed returned by the Fortran 90 95 RANDOM_SEED intrinsic to be constant The first call to RANDOM_SEED without arguments resets the random seed to a default value then advances the seed by a variable amount based on time Subsequent calls to RANDOM SEED without arguments reset the random seed to the same initial value as the first call Unless the time is exactly the same each time a program is run a different random number sequence is generated Setting the environment variable STATIC RANDOM SEED to YES forces the seed returned by RANDOM SEED to be constant thereby generating the same sequence of random numbers at each execution of the program 11 3 25 TMP You can use TMP to specify the directory to use for placement of any temporary files created during execution of the PGI compilers and tools This variable is interchangeable with TMPDIR 11 3 26 TMPDIR You can use TMPDIR to specify the directory to use for placement of any temporary files created during execution of the PGI compilers and tools
181. havior no abort This enables disables calling the system termination routine abort The default is noabort When noabort is in effect the process terminates by calling exit 127 On Linux when abort is in effect the abort routine creates a core file and exits with code 127 PGI Compiler User s Guide 144 Using Environment Variables On Windows when abort is in effect the abort routine exits with the status of the exception received For example if the program receives an access violation abort exits with status 0xC0000005 A few runtime errors just print an error message and call exit 127 regardless of the status of PGI TERM These are mainly errors such as specifying an invalid environment variable value where a traceback would not be useful If it appears that abort does not generate core files on a Linux system be sure to unlimit the coredumpsize You can do this in these ways gt Using csh limit coredumpsize unlimited setenv PGI TERM abort gt Using bash sh zsh or ksh ulimit c unlimited export PGI_TERM abort 9 o 2 o To debug a core file with pgdbg start pgdbg with the core option For example to view a core file named core for a program named a out pgdbg core core a out For more information on why to use this variable refer to Stack Traceback and JIT Debugging 11 3 21 PGI TERM DEBUG The PGI TERM D is set to debug I
182. he current thread cannot be changed In the PGI implementation all OpenMP or auto parallelization threads are created just prior to the first parallel region therefore only calsto omp set stack size thatoccur prior to the first region have an effect Fortran subroutine omp set stack size integer KIND OMP STACK SIZI C C vorc cmo set stack SH ENSE omp_get_team_size Returns for a given nested level of the current thread the size of the thread team to which the ancestor belongs Fortran integer function omp get team size level integer level C C integer omp get team size int level Returns whether or not the call is within a final task Returns TRUE for directives and non zero for pragmas if called from within a parallel region and FALSE for directives and zero for pragmas if called outside of a parallel region When called from within a parallel region that is serialized for example in the presence of an IF clause evaluating FALSE for directives and zero for pragmas the function returns FALSE for directives and zero for pragmas integer function omp in final int omp in final void omp in parallel Returns whether or not the call is within a parallel region Returns TRUE for directives and non zero for pragmas if called from within a parallel region and FALSE for directives and zero for pragmas if called outside of a parallel region When called from within a parall
183. i mpich bin PATH pgfortran Mmpi mpich mpihello f o mpihello Ht 4 Ut Ut 1 mpiexec mpihello Hello world I m node 0 mpiexec np 4 mpihello Hello world I m node 0 Hello world I m node 2 Hegstomwosld rim mode Hello world I m node 3 If you want to build your MPI application using the instance of MPICH installed with the PGI compilers just use the Mmpi mpich option Add g for debugging or use Mprof mpich instead to instrument for MPICH profiling To use a different instance of MPICH instead of the PGI provided one use the MPIDIR environment variable Before compiling set MPIDIR to the root of the MPICH installation directory that you want to use that is the directory that contains bin include lib and so on You can still use the Mmpi mpich and Mprof mpich options but now the compilers use the MPIDIR specified version of MPICH instead of the PGI default 6 6 Using MPICH1 MPICH2 and MVAPICH1 on Linux The Mmpi and Mprof sub options mpichl mpich2 and mvapich1 have been deprecated in the PGI 2015 release and will be disabled in a future release To use these deprecated options you must first set the environment variable MPIDIR to the root of the MPI installation directory you intend to use that is the directory that contains bin include lib and so on If you do not set MPIDIR you will see an error like the following pgfortran Mmpi mpich2 pi f90 o fpi Mmpi mpich2 has been deprecated and will
184. ibed in Fixed Form Fortran Directives This document uses free form for all PGI Accelerator compiler Fortran directive examples Rules The following rules apply to all OpenACC compiler directives gt Only one directive name can be specified per directive gt The order in which clauses appear is not significant gt Clauses may be repeated unless otherwise specified gt Forclauses that have a list argument a list is a comma separated list of variable names array names or in some cases subarrays with subscript ranges 7 9 3 C Directives In C OpenACC compiler directives are specified using pragma Syntax The syntax of an OpenACC compiler directive is pragma acc directive name clause clause new line Rules In addition to the general directive rules the following rules apply to OpenACC C directives gt Each directive starts with pragma acc gt The remainder of the directive follows the C conventions for pragmas gt White space may be used before and after the white space may be required to separate words in a directive gt Preprocessing tokens following the pragma acc are subject to macro replacement gt C directives are case sensitive gt An Accelerator directive applies to the immediately following structured block or loop 7 9 4 Free Form Fortran Directives OpenACC Fortran directives can be either Free Form or Fixed Form directives Free Form Accelerator directives are speci
185. ibution Requires compiling and linking with these options MVAPICH2 Use MVAPICH2 compiler wrappers profile profcc proffer Mprof func lines time MS MPI Mprof msmpi func lines Open MPI Use with Open MPI compiler wrappers Mprof func lines time SGI MPI Mprof sgimpi func lines time For example you can use the following command to compile for profiling with MPICH v3 pgfortran fast Mprof mpich func my mpi app f90 The default versions of the compiler wrappers i e mp icc and mpi 90 provided by MPI distributions do not correctly support the Mp ro f option and may need to be modified to enable profiling Once you have built an instrumented version of your MPI application running it produces the profile data For specific details on using PGPROF to view the profile data refer to the PGPROF Profiler Guide PGI Compiler User s Guide 76 Using MPI 6 5 Using MPICH on Linux and OS X PGI Workstation for Linux and OS X and PGI CDK for Linux include MPICH header files libraries and tools required to compile execute profile and debug MPI programs PGI Workstation can be installed on a single system and that system can be treated as if it is a small cluster by using this version of MPI Example MPI Hello World Example The following MPI example program uses MPICH cd my example dir cp r PGI linux86 64 14 0 EXAMPLES MPI mpihello cd mpihello export PATH S PGI linux86 64 2015 mp
186. ide 99 Using an Accelerator Table 20 Accelerator Environment Variables This environment variable Does this ACC_DEVICE_TYPE Controls which accelerator device to use when executing accelerator regions if the program has been compiled to use more than one different type of device The value of this environment variable is implementation defined and currently may be the string NVIDIA RADEON or HOST ACC_DEVICE_NUM Controls the default device number to use when executing accelerator regions The value of this environment variable must be a nonnegative integer between zero and the number of devices attached to the host PGI ACC NOTIFY When set to an integer value the value is used as a bit mask to print information about kernel launches value 1 data transfers value 2 wait operations or synchronizations with the device value 4 region entry exit value 8 and data allocate free value 16 PGI ACC TIME Enables a lightweight profiler to measure data movement and accelerator kernel execution time and print a summary at the end of program execution PGI ACC BUFFERSIZE For NVIDIA CUDA devices this defines the size of the pinned buffer used to transfer data between host and device PGI ACC GANGLIMIT For NVIDIA CUDA devices this defines the maximum number of gangs CUDA thread blocks that will be launched by a kernel PGI ACC DEV MEMORY For AMD Radeon devices this defines the maximum size OpenCL buffer to allocate The maxim
187. if the thread gains ownership of the lock otherwise it returns FALSE for directives and zero for pragmas If the variable is not already associated with a lock it is illegal to make a call to this routine Fortran logical function omp test lock lock include omp lib kinds h integer kind omp lock kind lock C C Hin omo cest lock Omo loek 106k int Omo test nest lock onmo mosic lock ic lock 5 9 Environment Variables You can use OpenMP environment variables to control the behavior of OpenMP programs These environment variables allow you to set and pass information that can alter the behavior of directives and pragmas The following summary table is a quick reference for the OpenMP environment variables that PGI uses Table 14 OpenMP related Environment Variable Summary Table Environment Variable Default Description OMP_DYNAMIC FALSE Currently has no effect Typically enables TRUE or disables FALSE the dynamic adjustment of the number of threads OMP_MAX_ACTIVE_LEVELS Specifies the maximum number of nested parallel regions OMP_NESTED FALSE Enables TRUE or disables FALSE nested parallelism PGI Compiler User s Guide 72 Using OpenMP Environment Variable Default Description OMP_NUM_THREADS 1 Specifies the number of threads to use during execution of parallel regions at the corresponding nested level For example OMP_NUM_THREADS 4 2 uses 4 threads at the first nested parallel level and
188. ilding the DLL itself pacc Bdynamic c object3 c pacc Mmakeimplib o obj3 lib object3 0bj oe ce o Tip The de f lt def file gt option can also be used with Mnakeimplib Usea def file when you need to export additional symbols from the DLL A def file is not needed in this example because all symbols are exported using dec1spec dllexport Step 2 Use the import library obj 3 lib created in Step 1 to link the second DLL pacc Bdynamic c object4 c pacc Mmakedll o obj4 dll object4 0bj defaultlib obj3 oe oe Step 3 Use the import library obj 4 1ib created in Step 2 to link the first DLL pgcc Mmakedll o obj3 dll object3 obj defaultlib obj4 Step 4 Compile the main program and link against the import libraries for the two DLLs pgcc Bdynamic prog3 c o prog3 defaultlib obj3 defaultlib obj4 6 PGI Compiler User s Guide 130 Creating and Using Libraries 5 Step 5 Execute prog3 exe to ensure that the DLLs were create properly prog3 BUNCE Sa Calling van cout ime mobs 4 cis func 4b func 4a calling a routine in obj3 dll FUME JS 10 7 4 Build DLLs Containing Mutual Imports Fortran In this example we build two DLLs when each DLL is dependent on the other and use them to build the main program In the following source files obj ect2 95 makes calls to routines defined in object3 f95 and vice versa This situation of mutual imports requires two steps to build each DLL To
189. indicates the default for the directive or pragma N A appears if a default does not apply PGI Compiler User s Guide 110 Using Directives and Pragmas gt The scope entry indicates the allowed scope indicators for each directive or pragma with L for loop R for routine and G for global The default scope is surrounded by parentheses and N A appears if the directive or pragma is not available in the given language The in the scope indicates this For routine scoped directive The scope includes the code following the directive or pragma until the end of the routine For globally scoped directive The scope includes the code following the directive or pragma until the end of the file rather than for the entire file The name of a directive or pragma may also be prefixed with M For example you can use the directive Mbounds which is equivalent to the directive bounds and you can use Mopt which is equivalent to opt For pragmas you can use the directive Mnoassoc which is equivalent to the pragma noassoc and Mvintr which is equivalent to vintr Table 24 Proprietary Optimization Related Fortran Directive and C C Pragma Summary Directive or Fortran C C pragma Functionality Default Scope altcode noaltcode Do don t generate alternate code for vectorized and parallelized loops Do don t perform associative transformations bounds nobounds Do don t perform array bounds checking cncall nocncall Loops are
190. ing is identical to explicitly specifying a parallel construct containing a loop directive new line Is a shortcut for specifying a loop directive nested immediately inside an accelerator kernels directive The meaning is identical to explicitly specifying a kernels construct containing a loop directive 93 This directive Accepts these clauses C Syntax pragma acc kernels loop clause clause for loop Fortran Syntax acc kernels loop clause clause do loop Cache Directive C Syntax pragma acc cache list Fortran Syntax acc cache list Declare Directive copy list copyin list copyout list create list deviceptr list device resident list C Syntax Using an Accelerator Has this functionality new line Specifies array elements or subarrays that should be fetched into the highest level of the cache for the body of a loop Must appear at the top of inside of the loop Specifies that an array or arrays are to be allocated in the device memory for the duration of the implicit data region of a function subroutine or program Specifies whether the data values are to be transferred from the host to the device memory upon entry to the implicit data region and from the device to the host memory upon exit from the implicit data region Creates a visible device copy of the variable or array pragma acc declare clause clause Fortran
191. ins objects with constructors and destructors calling such a function from either C or Fortran is not possible unless the initialization in the main program is performed from a C program in which constructors and destructors are properly initialized gt In general you can call a C or Fortran function from C without problems as long as you use the extern C keyword to declare the function in the C program This declaration prevents name mangling for the C function name If you want to call a C function from C or Fortran you also have to use the extern C keyword to declare the C function This keeps the C compiler from mangling the name of the function gt You can use the __cplusplus macro to allow a program or header file to work for both C and C For example the following defines in the header file stdio h allow this file to work for both C and C ifndef STDIO H define STDIO H abolesr colusplus swquesem WEH i endif cplusplus Functions and data types defined ifdef cplusplus endif _ cplusplus endif gt C member functions cannot be declared extern since their names will always be mangled Therefore C member functions cannot be called from C or Fortran 13 3 Functions and Subroutines Fortran C and C define functions and subroutines differently For a Fortran program calling a C or C function observe the following return value convention gt When aC
192. integer kinds for arguments to those routines gt Integer parameters to describe types of accelerators gt The integer parameter accel version witha value yyy ymm where yyyy and mm are the year and month designations of the version of the Accelerator programming model supported This value matches the value of the preprocessor variable ACCEL 7 11 2 Runtime Library Routines Table 19 lists and briefly describes the runtime library routines supported by PGI in addition to the standard OpenACC runtine API routines PGI Compiler User s Guide 98 Using an Accelerator Table 19 Accelerator Runtime Library Routines This Runtime Library Routine acc_allocs acc_bytesalloc acc_bytesin acc_bytesout acc_copyins acc_copyouts acc_disable_time acc_enable_time acc_exec_time acc_frees acc_get_device acc_get_device_num acc_get_free_memory acc_get_memory acc_get_num_devices acc_kernels acc_regions acc_total_time Does this Returns the number of arrays copied in to the accelerator by data or compute regions Returns the number of arrays copied out from the accelerator by data or compute regions Tells the runtime to stop profiling accelerator regions and kernels Tells the runtime to start profiling accelerator regions and kernels if it is not already doing so Returns the number of microseconds spent on the accelerator executing kernels Returns the number of arrays freed or deallocated in data or compute regions
193. ions enclosing the task that contains the call integer function omp get level omp get max threads Returns the maximum value that can be returned by calls to omp get num threads lf omp set num threads is used to change the number of processors subsequent calls to omp get max threads return the new value Further this function returns the maximum value whether executing from a parallel or serial region of code integer function omp get max threads void omp get max threads void omp get num procs Returns the number of processors that are available to the program integer function omp get num procs int omp get num procs void PGI Compiler User s Guide 68 Using OpenMP Runtime Library Routines with Examples omp_get_stack_size Returns the value of the OpenMP internal control variable that specifies the size that is used to create a stack for a newly created thread This value may not be the size of the stack of the current thread omp get stack size interface function omp get stack size use omp lib kinds integer kind OMP STACK SIZE KIND omp get stack size end function omp get stack size engineers C C size t omp get stack size void omp set stack size Changes the value of the OpenMP internal control variable that specifies the size to be used to create a stack for a newly created thread The integer argument specifies the stack size in kilobytes The size of the stack of t
194. ires a space between the prefix and the directive keyword ATTRIBUTES gt The must begin the prefix when compiling using Fortran 90 freeform format gt The characters C or can be used in place of in either form of the prefix when compiling with fixed form format gt The directives are completely case insensitive 13 9 5 Using the C Calling Convention Using the C calling convention requires the insertion of a compiler directive into the declarations section of any Fortran program unit which calls the C program unit You cannot mix UNIX style argument passing and C calling conventions within the same file Syntax for the directive is as follows DECS ATTRIBUTES C work Where work is the name of the subroutine to be called using C conventions More than one subroutine may be listed separated by commas Refer to Symbol Name Construction and Calling Example for a complete description of the implementation of the C calling convention 13 9 6 Using the UNIX Calling Convention Using the UNIX calling convention is straightforward Any program unit compiled using Miface unix or the Munix compilation flag uses the UNIX convention 13 9 7 Using the CREF Calling Convention Using the CREF calling convention is straightforward Any program unit compiled using Miface cref compilation flag uses the CREF convention PGI Compiler User s Guide 167 Chapter 14 PROGRAMMING CONSIDERATIONS FOR 64 BIT ENVIRONMENTS PGI
195. it is supported by PGI compilers Currently all PGI compilers support the version 3 1 OpenMP specification Use the mp compiler switch to enable processing of the OpenMP pragmas listed in this section As of the PGI 2011 Release the OpenMP runtime library is linked by default Note that GNU pthreads are not completely interoperable with OpenMP threads When using pgc on Linux or OS X the GNU STL is thread safe to the extent listed in the GNU documentation as required by the C 11 standard If an STL thread safe issue is suspected the suspect code can be run sequentially inside of an OpenMP region using pragma omp critical sections When using the deprecated pgcpp compiler on Linux OS X or Windows the C Standard Template Library is thread safe to the extent allowed in the STLport code simultaneous accesses to distinct containers are safe simultaneous reads to shared containers are also safe However simultaneous writes to shared containers must be protected by pragma omp critical sections This section describes how to use the following option related to using OpenMP mp 5 1 OpenMP Overview Let s look at the OpenMP shared memory parallel programming model and some common OpenMP terminology PGI Compiler User s Guide 57 Using OpenMP 5 1 1 OpenMP Shared Memory Parallel Programming Model The OpenMP shared memory programming model is a collection of compiler directives or pragmas library routines and environment variab
196. l optimizations Question Can I build multiple applications in the same directory with Mipa Answer Yes Suppose you have three source files mainl c main2 c and sub c where sub c is shared between the two applications Suppose you build the first application with Mipa using this command gt pgce Mrpa fast o appl menal E sub c The IPA linker creates two IPA optimized object files and uses them to build the first application mainl ipa4 appl oo sub ipa4 appl oo Now suppose you build the second application using this command 9 pgcc Mipa fast o app2 main2 c sub c The IPA linker creates two more IPA optimized object files main2 ipa4 app2 oo sub ipa4 app2 oo There are now three object files for sub c the original sub o and two IPA optimized objects one for each application in which it appears Question How is the mangled name for the IPA optimized object files generated Answer The mangled name has ipa appended followed by the decimal number of the length of the executable file name followed by an underscore and the executable file name itself The suffix is changed to oo so that linking o does not pull in the IPA optimized objects If the IPA linker determines that the file would not benefit from any interprocedural optimizations it does not have to recompile the file at link time and uses the original object Question Can I use parallel make environments e g pmake with IPA Answer No I
197. l region that is executed by a team consisting of more than one thread The definition of an active parallel region changed between OpenMP 2 5 and OpenMP 3 1 In OpenMP 2 5 the definition was a parallel region whose IF clause evaluates to true To examine the significance of this change look at the following example program test logical omp in parallel omp parallel print omp in parallel omp end parallel stop end Suppose we run this program with OMP NUM THREADS set to one In OpenMP 2 5 this program yields T while in OpenMP 3 1 the program yields F In OpenMP 3 1 execution is not occurring by more than one thread Therefore change in this definition may mean previous programs require modification Task A specific instance of executable code and its data environment generated when a thread encounters a task construct or a parallel construct 5 1 3 OpenMP Example Look at the following simple OpenMP example involving loops PGI Compiler User s Guide 59 Using OpenMP OpenMP Loop Example PROGRAM MAIN INTEGER I N OMP GET THREAD NUM REAL 8 V 1000 GSUM LSU GSUM 0 0D0 N 1000 DO I 1 N V I DBLE I ENDDO SOMP PARALLEL PRIVATE I LSUM SHARED V GSUM N LSUM 0 0D0 S OMP DO BO x e i N SUM LSUM V I ENDDO OMP END DO SOMP CRITICAL print Thread OMP GET THREAD NUM Voca
198. l sum LSUM GSUM GSUM LSUM 1S 0MP END CRITICAL SOMP END PARALLEL PRINT Global Sum GSUM STOP END If you execute this example with the environment variable OMP NUM THREADS set to 4 then the output looks similar to this Thread local swns 31375 00000000000 Thread iL local sums 93875 00000000000 Thread 20k ca sums 156375 0000000000 Thread 3 local suns 218875 0000000000 Global Sum 500500 0000000000 FORTRAN STOP 5 2 Task Overview Every part of an OpenMP program is part of a task A task whose execution can be performed immediately or delayed has these characteristics gt Code to execute gt A data environment that is it owns its data gt An assigned thread that executes the code and uses the data There are two activities associated with tasks packaging and execution gt Packaging Each encountering thread packages a new instance of a task code and data gt Execution Some thread in the team executes the task at some later time In the following sections we use this terminology Task The package of code and instructions for allocating data created when a thread encounters a task construct A task can be implicit or explicit gt An explicit task is a task generated when a task construct is encountered during execution PGI Compiler User s Guide 60 Using OpenMP gt An implicit task is a task generated by the implicit parallel region or generated when a parallel construct is
199. le LD LIBRARY PATH to include the directory containing tobeshared so If LD LIBRARY PATH is already initialized it is important not to overwrite its contents If you have placed tobeshared so in directory home myusername bin you can initialize LD LIBRARY PATH to include that directory and preserve its existing contents as shown in the following setenv LD LIBRARY PATH SLD LIBRARY PATH home myusername bin If you know that tobeshared so always resides in a specific directory you can create the executable myprog in a form that assumes this directory by using the R link time option For example you can link as follows 9 pgfortran o myprog myprof f tobeshared so R home myusername bin As with the T option there is no space between R and the directory name If the R option is used itis not necessary to initialize LD LIBRARY PATH In the previous example the dynamic linker always looks in home myusername bin to resolve references to tobeshared so By default if the LD LIBRARY PATH environment variable is not set the linker only searches usr 1ib and 1ib for shared objects 10 3 2 Idd Command The 1dd command is a useful tool when working with shared object files and executables that reference them When applied to an executable as shown in the following example 1dd lists all PGI Compiler User s Guide 122 Creating and Using Libraries shared object files
200. les that can be used to specify shared memory parallelism in Fortran and in C C programs Fortran directives and C C pragmas Allow users to mark sections of code that can be executed in parallel when the code is compiled using the mp switch When this switch is not present the compiler ignores these directives and pragmas OpenMP Fortran directives begin with SOMP CSOMP or SOMP beginning in column 1 OpenMP pragmas for C C begin with pragma omp This format allows the user to have a single source for use with or without the mp switch as these lines are then merely viewed as comments when mp is not present These directives and pragmas allow the user to create task loop and parallel section work sharing constructs and synchronization constructs They also allow the user to define how data is shared or copied between parallel threads of execution Fortran directives and C C pragmas include a parallel region construct for writing coarse grain SPMD programs work sharing constructs which specify that DO loop iterations or C C for loop iterations should be split among the available threads of execution and synchronization constructs The data environment is controlled either by using clauses on the directives or pragmas or with additional directives or pragmas Runtime library routines Are available to query the parallel runtime environment for example to determine how many threads are participating in execution
201. llustrates the extra parameters tmp and 10 supplied by the caller Character Return Parameters Fortran function returns a character CHARACTER FUNCTION CHF C1 I CHARACTER Cl INTEGER I END C declaration of Fortran function exceteur om cc s MOF ole deme L1L T g clavaie il 9g ague aia Car tmo 10 ci exu 995 If the 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 must still be supplied but is not used The value of the character function is not automatically NULL terminated 13 6 3 Complex Return Values When a Fortran function returns a complex value an argument needs to be added at the beginning of the C C calling function s argument list this argument is the address of the complex return value COMPLEX Return Values illustrates the extra parameter cp1x supplied by the caller COMPLEX Return Values COMPLEX FUNCTION CE C E ENTE CER END PyEeuem vord iE Lg typedef struct float real imag cplx Colles sige abe eut Gel fi ep 13 7 Array Indices C C arrays and Fortran arrays use different default initial array index values By default arrays in C C start at 0 and arrqays in Fortran start at 1 If you adjust your array comparisons so that a Fortran second element is compared to a C C first ele
202. lowed in place of c gt The scope indicators g r and used with the cpgi sentinel are not supported gt The directive name including the directive prefix may contain upper or lower case letters and is case insensitive case is not significant gt Ifthe command line option Mupcase is used any variable names that appear in the body of the directive are case sensitive 9 6 3 Sample Usage of Prefetch Directive Prefetch Directive Use This example uses prefetch directives to prefetch data in a matrix multiplication inner loop where a row of one source matrix has been gathered into a contiguous vector eale ala Tex tat Nae mp 199 5 cese exe Gab do J l cSmem prefetch arow 1 b 1 j cSmem prefetch arow 5 b 5 j cSmem prefetch arow 9 b 9 j gio Je e Le imp 4 c mem prefetch arow k 12 b k 12 j eli I e a arowlik DI ei ey 4r arowa o ei ey ar arowa Tess 3p ei e ar arowa o enddo enddo This pattern of prefetch directives the compiler emits prefetch instructions whereby elements of arow and b are fetched into the data cache starting four iterations prior to first use By varying the prefetch distance in this way it is sometimes possible to reduce the effects of main memory latency and improve performance 9 6 4 Prefetch Pragma Syntax in C C The syntax of a prefetch pragma is as follows pragma mem prefetch lt varl gt lt var2 gt PGI Compiler User s Guide 1
203. lt numdoubl endl cout lt numshorl lt lt numshorl endi PGI Compiler User s Guide 163 Inter language Calling Fortran Subroutine cp2f_func f SUOLOUEIMS COZ wm lyoclil Jiexcuescibo mumutinel numint2 numfloatl numdoubl numshor1 logrecealsil booli character letterl integer numintl numint2 double precision numdoub1 real numfloatl integer 2 numshorl booll true letterl v numintl 11 numint2 44 numdoubl 902 numfloatl 39 6 numshorl 299 return end To compile this Fortran subroutine and C program use the following command lines 9 parorcrenm CPE CUNE i S paet COA TWNG O GOZE mein esp cis ONES Executing this C main should produce the following output booll TRUE letterl v numintl 11 numint2 44 numfloatl 39 6 numdoubl 902 numshorl 299 You must explicitly link in the PGFORTRAN runtime support libraries when linking pgfortran compiled program units into C or C main programs When linking pgf77 compiled program units into C or C main programs you need only linkin Lpgftnrtl 13 9 Win32 Calling Conventions A calling convention is a set of conventions that describe the manner in which a particular routine is executed A routine s calling conventions specify where parameters and function results are passed For a stack based routine the calling conventions determine the structure of the routine s stack frame The calling convention
204. m to t v P o o te fee es tee et t dee ve ote tv te 27 PPM mS 27 UL E A E E E 28 9 2 6 PGPROF E saaana eaaa TE Ea ADEE a Eaa SE a anata aaa aD Eea SIA 28 3 3 Common Compiler Feedback Format CCFF ssssseseeeee netten tnter trennen tenens 28 3 4 Local and Global Optimization sse tenente tenente tenerent nennen 28 SIM CM P 28 342 O A A niue nme Mh e n P ea e Fe n Es E 29 3 5 Loop Unrolling using Munroll seen nente nter tenente tentent tentent tenen 31 3 6 Vectorization using Mvo reses rien te teet dian et ete a died een cee ees 32 33 0 1 Vectorization SUD ODUIONS fs asics tr rie tte e ec r reda a Rer rer etd 33 3 6 2 Vectorization Example Using SIMD Instructions ssseeeneennnennnn 34 3 7 Auto Parallelization using Mconcur essent nter tentent tentent tentent tenen tenen 36 3 7 1 Auto Parallelization Sub options sese tenentes 37 3 1 2 Loops That Fail to Paralleli26 55 2 ro ort ea tete ete P o a trc ene qon 38 3 8 Processor Specific Optimization amp the Unified Binary senes 42 3 9 Interprocedural Analysis and Optimization using Mipa senes 43 3 9 1 Building a Program Without IPA Single Step 43 3 9 2 Building a Program Without IPA Several Steps 44 3 9 3 Building a Program Withou
205. ment and adjust similarly for other elements you should not have problems working with this difference If this is not satisfactory you can declare your Fortran arrays 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 uses 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 PGI Compiler User s Guide 158 Inter language Calling row and column indexes can simply be reversed For arrays other than single dimensional arrays and square two dimensional arrays inter language function mixing is not recommended 13 8 Examples This section contains examples that illustrate inter language calling 13 8 1 Example Fortran Calling C There are other solutions to calling C from Fortran than the one presented in this section For example you can use the iso c binding intrinsic module which PGI does support For more information on this module and for examples of how to use it search the web using the keyword iso_c_binding C function f2c_func_ shows a C function that is called by the Fortran main program shown in Fortran Main Program f2c_main f Notice that each argument is defined as a pointer since Fortran passes by reference Also notice that the C function name uses all lower case and a trailing _ Fortran Main Program f2c_main f logicei booli cha
206. mory Specifies that the memory of the named variables should be allocated in the accelerator device memory and not in the host memory deviceptr list Declares that the items in the list are device pointers so the data need not be allocated or moved between the host and device Parallel Kernels Data Declare firstprivate list Parallel Declares that a copy of each item on the list will be created for each parallel gang and that the copy will be initialized with the value of that item on the host when the parallel construct is encountered gang gang arg list Loop Specifies that the iterations of the associate loop or loops are to be Routine executed in parallel by distributing the iterations among the gangs created by the parallel construct host list Update Copies the visible device copies of the variables arrays or subarrays in the list argument to the associated host memory locations The copy occurs after completion of the compute or data region if condition When present tells the compiler to generate two copies of the region one for the accelerator one for the host and to generate code to decide which copy to execute Parallel Kernels if condition Conditionally allocate memory on and move data to and or from the Data Region device Enter Data Exit Data Update independent Tells the compiler that the iterations of this loop are data independent of each other thus allowing the compiler to generate
207. mpatible or an x64 compatible processor gt 64 bit 32 bit Linux includes all features and capabilities of the 32 bit Linux version and is also supported on 64 bit Linux operating systems running on an x64 compatible processor gt 32 bit Windows supported on 32 bit Windows operating systems running on either a 32 bit x86 compatible or an x64 compatible processor gt 64 bit 32 bit Windows includes all features and capabilities of the 32 bit Windows version also supported on 64 bit Windows operating systems running an x64 compatible processor gt 32 bit OS X supported on 32 bit OS X operating systems running on either a 32 bit or 64 bit Intel based Mac system gt 64 bit OS X supported on 64 bit OS X operating systems running on a 64 bit Intel based Mac system The following sections describe the specific considerations required to use the PGI compilers on the various platforms Linux Windows and OS X PGI Compiler User s Guide Getting Started 1 7 1 Using the PGI Compilers on Linux Linux Header Files The Linux system header files contain many GNU gcc extensions PGI supports many of these extensions thus allowing the PGI C and C compilers to compile most programs that the GNU compilers can compile A few header files not interoperable with the PGI compilers have been rewritten These files are included in PGI linux86 include and in PGI linux86 include gcc suchas sigset h asm byteorder h stddef
208. n Eclipse a free open source integrated software development environment Using Directives and Pragmas provides a description of each Fortran optimization directive and C C optimization pragma and shows examples of their use Creating and Using Libraries discusses PGI support libraries shared object files and environment variables that affect the behavior of the PGI compilers Using Environment Variables describes the environment variables that affect the behavior of the PGI compilers Distributing Files Deployment describes the deployment of your files once you have built debugged and compiled them successfully Inter language Calling provides examples showing how to place C language calls in a Fortran program and Fortran language calls in a C program Programming Considerations for 64 Bit Environments discusses issues of which programmers should be aware when targeting 64 bit processors C C Inline Assembly and Intrinsics describes how to use inline assembly code in C and C programs as well as how to use intrinsic functions that map directly to x86 and x64 machine instructions Hardware and Software Constraints This guide describes versions of the PGI compilers that produce assembly code for x86 and x64 processor based systems Details concerning environment specific values and defaults and system specific features or limitations are presented in the release notes delivered with the PGI compilers Conventions Thi
209. nd alternative and the input operand has three for its second alternative The compiler first tries to satisfy the left most alternative of the first operand for example the output operand in example13 When satisfying the operand the compiler starts with the left most constraint If the compiler cannot satisfy an alternative with this constraint for example if the desired register is not available it tries to use any subsequent constraints If the compiler runs out of constraints it moves on to the next alternative If the compiler runs out of alternatives it issues an error similar to the one mentioned in example 2 If an alternative is found the compiler uses the same alternative for subsequent operands For example if the compiler chooses the c register for the output operand in example13 then it will use either the a or m constraint for the input operand 15 2 8 Constraint Modifiers Characters that affect the compiler s interpretation of a constraint are known as Constraint Modifiers Two constraint modifiers the and the were introduced in Output Operands The following table summarizes each constraint modifier Table 38 Constraint Modifier Characters Constraint Modifier Description This operand is write only It is valid for output operands only If specified the must appear as the first character of the constraint string This operand is both read and written by the instruction It is vali
210. ng item are valid Some options do not allow a space between the option and its argument or within an argument When applicable the syntax section of the option description in the Command Line Options Reference section of the PGI Compiler s Reference Guide 2 1 2 Command line Suboptions Some options accept several suboptions You can specify these suboptions either by using the full option statement multiple times or by using a comma separated list for the suboptions The following two command lines are equivalent pgfortran Mvect simd Mvect noaltcode pgfortran Mvect simd noaltcode 2 1 3 Command line Conflicting Options Some options have an opposite or negated counterpart For example both Mvect and Mnovect are available Mvect enables vectorization and Mnovect disables it If you used both of these commands on a command line they would conflict Rule When you use conflicting options on a command line the last encountered option takes precedence over any previous one The conflicting options rule is important for a number of reasons gt Some options such as fast include other options Therefore it is possible for you to be unaware that you have conflicting options gt You can use this rule to create makefiles that apply specific flags to a set of files as shown in the following example Example Makefiles with Options In this makefile fragment CCFLAGS uses vectorization CCNOVECTFLAGS uses the flags de
211. ni o mme o wetiesmedfesmm ma mensem O me webietensresedde 7 16 2 Supported C Intrinsics Summary Table This section contains two alphabetical summaries one for double functions and a second for float functions These lists contain only those C intrinsics that the accelerator supports Table 22 Supported C Intrinsic Double Functions This intrinsic Returns this value ee emestemdende PGI Compiler User s Guide 104 Using an Accelerator This intrinsic Returns this value fer seemesesed e m mesresmetewmm 000 Table 23 Supported C Intrinsic Float Functions This intrinsic Returns this value 7 17 References related to Accelerators gt ISO IEC 1539 1 1997 Information Technology Programming Languages Fortran Geneva 1997 Fortran 95 gt American National Standard Programming Language C ANSI X3 159 1989 ANSI C gt ISO IEC 9899 1999 Information Technology Programming Languages C Geneva 1999 C99 PGI Compiler User s Guide 105 Using an Accelerator gt ISO IEC 9899 2011 Information Technology Programming Languages C Geneva 2011 C11 gt ISO IEC 14882 2011 Information Technology Programming Languages C Geneva 2011 C11 gt PGDBG Dubugger Manual The Portland Group Release 15 1 January 2015 Available online at http www pgroup com doc pgdbg pdf gt PGPROF Profiler Manual The Portland Group Release 15 1 January 2015 Available online a
212. nies M maim acad ceara end PGI Compiler User s Guide 127 Creating and Using Libraries 1 Create the DLL obj1 d11 and its import library obj1 1ib using the following series of commands pgfortran Bdynamic c objectl f pgfortran Mmakedll objectl obj o objl dll 2 Compile the main program pgfortran Bdynamic o progl progl f defaultlib objl The Bdynamic and Mmakedl11 switches cause the compiler to link against the PGI runtime DLLs instead of the PGI runtime static libraries The Bdynamic switch is required when linking against any PGI compiled DLL such as obj1 d11 The defaultlib switch specifies that ob31 1ib the DLL s import library should be used to resolve imports 3 Ensure that obj 1 d11 is in your path then run the executable prog1 to determine if the DLL was successfully created and linked progl subl adata 11 subl 3 12 main adata 12 Should you wish to change ob31 d11 without changing the subroutine or function interfaces no rebuilding of prog1 is necessary Just recreate obj 1 d11 and the new 0b31 d11 is loaded at runtime 10 7 2 Build a DLL C In this example we build a DLL out of a single source file obj ect2 c which exports data and a subroutine using declspec dllexport The main source file prog2 c uses declspec dllimport to import the data and subroutine from the DLL object2 c int declspec dllexport data void _declspec dllexport une2 ani L p
213. not corresponding device memory will be allocated and the data ata PGI Compiler User s Guide 95 Using an Accelerator Use this clause In these directives To do this Declare will be copied to the device at entry to the region and the memory Enter Data deallocated at exit from the region For compatibility with OpenACC 2 0 this may also be spelled present_or_copyin or pcopyin copyout list At execution time the implementation will test whether each of the items in the list are already present in the device memory If not corresponding device memory will be allocated at entry to the region and data will be copied back to the host and the memory deallocated Parallel Kernels Data Declare at exit from the region For compatibility with OpenACC 2 0 this may Exit Data also be spelled present_or_copyout or pcopyout create list Parallel At execution time the implementation will test whether each of the items in the list are already present in the device memory If not Kernels j corresponding device memory will be allocated at entry to the region Data and the memory deallocated at exit from the region For compatibility Declare with OpenACC 2 0 this may also be spelled present or create or Enter Data pcreate corresponding device memory device list Update Copies the variables arrays or subarrays in the list argument from host memory to the visible device copy of the variables arrays or subarrays in device me
214. ns that are useful in improving performance 2 3 1 Using fast and fastsse Options PGI compilers implement a wide range of options that allow users a fine degree of control on each optimization phase When it comes to optimization of code the quickest way to start is to use the options fast or astsse These options create a generally optimal set of flags for x86 targets They incorporate optimization options to enable use of vector streaming SIMD SSE instructions for 64 bit targets They enable vectorization with SSE instructions cache alignment and SSE arithmetic to flush to zero mode The contents of the fast or fastsse options are host dependent Further you should use these options on both compile and link command lines The following table shows the typical fast and fastsse options Table4 Typical fast and astsse Options Use this option To do this Specifies a code optimization level of 2 Munroll c 1 Unrolls loops executing multiple instances of the original loop during each iteration Mnoframe Indicates to not generate code to set up a stack frame Note With this option a stack trace does not work Indicates loop carried redundancy elimination Indicates partial redundancy elimination fast for 64 bit targets and fastsse for both 32 and 64 bit targets also typically include the options shown in this table Table 5 Additional fast and fastsse Options Use this option To do this Gen
215. nt in the argument list PGI Compiler User s Guide 165 Inter language Calling The following example is pseudocode for the preceding subroutine call using Default conventions call WORK 20 addr ERR 3 addr a addr b addr n gt STDCALL The symbol name for the subroutine is constructed by pre pending an underscore converting to all lower case and appending an sign followed by an integer indicating the total number of bytes occupied by the argument list Character strings are truncated to the first character in the string which is passed by value as the first byte in a 4 byte word The following is an example of the pseudocode for the work subroutine call using STDCALL conventions call work 20 val E val a addr b val n Notice in this case that there are still 20 bytes in the argument list However rather than five 4 byte quantities as in the Default convention there are three 4 byte quantities and one 8 byte quantity the double precision value of a gt C The symbol name for the subroutine is constructed by pre pending an underscore and converting to all lower case Character strings are truncated to the first character in the string which is passed by value as the first byte in a 4 byte word The following is an example of the pseudocode for the work subroutine call using C conventions call work val E val a addr b val n gt UNIX The symbol name for the subroutine is
216. nterprocedural optimizations There may be additional information when inlining is enabled The total size of the object file may be 5 10 times its original size The extra sections are not added to the final executable Question What if I compile with Mipa and link without Mipa Answer The PGI compilers generate a legal object file even when the source file is compiled with Mipa If you compile with Mipa and link without Mi pa the linker is invoked on the original object files A legal executable is generated While this executable does not have the benefit of interprocedural optimizations any other optimizations do apply PGI Compiler User s Guide 46 Optimizing and Parallelizing Question What if I compile without Mipa and link with Mipa Answer At link time the IPA linker must have summary information about all the functions or routines used in the program This information is created only when a file is compiled with Mipa If you compile a file without Mi pa and then try to get interprocedural optimizations by linking with Mipa the IPA linker will issue a message that some routines have no IPA summary information and will proceed to run the system linker using the original object files If some files were compiled with Mipa and others were not it will determine the safest approximation of the IPA summary information for those files not compiled with Mi pa and use that to recompile the other files using interprocedura
217. oatl double numdoubl short numshorl Essi naso tol c sts NN 7 c2f sub amp booll letterl amp numintl numint2 numfloatl amp numdoubl amp numshorl 1 prime e Se Sel Sl So LE SO Scl a DOLL Panu o iS lorte p Mnu nE NEA y numfloatl numdoubl numshorl Fortran Subroutine c2f_sub f SUbBOuIEINEeE C2 rume boost es mowimel MUMINEA numfloatl numdoubl numshorl logical i booli character letter1 integer numintl numint2 double precision numdoubl real numfloatl integer 2 numshorl booll true letterl v numintl 11 numint2 44 numdoubl 902 numfloatl 39 6 numshorl 299 return end To compile this Fortran subroutine and C program use the following commands S pee e CLE meim E pgfortran Mnomain c2f main o c2 sub f Executing the resulting a out file should produce the following output TRUE v 11 44 39 6 902 299 PGI Compiler User s Guide 160 Inter language Calling 13 8 3 Example C Calling C C Main Program cp2c_main C Calling a C Function shows a C main program that calls the C function shown in Simple C Function c2cp_func c C Main Program cp2c_main C Calling a C Function extern WCW woe joie Funente mp Aine im aee 7939 9 include lt iostream gt main abge aD el a 8 b 2 c 0 cout lt lt main a a p lt lt b lt lt ptr c lt lt hex lt lt amp c lt lt endl CpZicmEUMC ay bree co
218. obj 3 d11 and obj 4 d11 each of which imports a routine that is exported by the other To link the first DLL the import library for the second DLL must be available Usually an import library is created when a DLL is linked In this case however the second DLL cannot be linked without the import library for the first DLL When such circular imports exist an import library for one of the DLLs must be created in a separate PGI Compiler User s Guide 129 Creating and Using Libraries step without creating the DLL The PGI drivers call the Microsoft 1ib tool to create import libraries in this situation Once the DLLs are built we can use them to build the main program objecto e void declspec dllimport func 4b void void declspec dllexport func 3a void priat rwie Sa Calling ei mewediaS in obj cll Wat p func 4b void declspec dllexport func 3b void primer Urune Sig im 7 IES ooeec e E void declspec dllimport func 3b void void declspec dllexport func 4a void printf func 4a calling a routine in obj3 dll in func 3b void declspec dllexport func 4b void primer Eur eatenus A OO e void _declspec dllimport func 3a void void _declspec dllimport func 4a void age main 4 func 3a func 4a return 0 Step 1 Use Mmakeimplib with the PGI compilers to build an import library for the first DLL without bu
219. of a parallel region Environment variables Are available to control the execution behavior of parallel programs For more information on OpenMP see www openmp org Macro substitution C and C omp pragmas are subject to macro replacement after pragma omp 9 1 2 Terminology For OpenMP 3 1 there are a number of terms for which it is useful to have common definitions Thread An execution entity with a stack and associated static memory called threadprivate memory gt An OpenMP thread is a thread that is managed by the OpenMP runtime system gt A thread safe routine is a routine that performs the intended function even when executed concurrently that is by more than one thread PGI Compiler User s Guide 58 Using OpenMP Region All code encountered during a specific instance of the execution of a given construct or of an OpenMP library routine A region includes any code in called routines as well as any implicit code introduced by the OpenMP implementation Regions are nested if one region is dynamically enclosed by another region that is a region is encountered during the execution of another region PGI supports non lexically nested parallel regions Parallel region In OpenMP 3 1 there is a distinction between a parallel region and an active parallel region A parallel region can be either inactive or active gt An inactive parallel region is executed by a single thread An active parallel region is a paralle
220. on 32 bit systems the 32 bit compilers are targeted PGI Compiler User s Guide 11 Getting Started BASH Shell Environment Cygwin A UNIX like shell environment Cygwin is bundled with PGI compilers and tools for Windows to provide a familiar development environment for Linux or UNIX users After installation of PGI Workstation or PGI Server you have a PGI Workstation icon on your Windows desktop Double left click on this icon to launch an instance of the Cygwin bash command shell window Working within BASH is very much like working within the sh or ksh shells on a Linux system yet BASH has a command history feature similar to csh and several other unique features Shell programming is fully supported The BASH shell window is pre initialized for usage of the PGI compilers and tools so there is no need to set environment variables or modify your command path when the command window comes up In addition to the PGI compiler commands within BASH you have access to over 100 common commands and utilities including but not limited to the following vi gzip gunzip ftp tar untar grep egrep fgrep awk sed cksum cp cat diff du date kill ls find mv printenv env more less touch wc rm rmdir make If you are familiar with program development in a Linux environment editing compiling and executing programs within bash will be very comfortable If you have not previously used such an environment you might want to familia
221. on AMD GPUS the program looks for and dynamically loads the AMD OpenCL libraries If the libraries are not available or if they are in a different directory than they were when the program was compiled you may need to append the appropriate library directory to your LD LIBRARY PATH environment variable on Linux or to the PATH environment variable on Windows gt On Linux if you have no server running on your NVIDIA GPU when your program reaches its first accelerator region there may be a 0 5 to 1 5 second pause to warm up the GPU from a power off audience You can avoid this delay by running the pgcudainit program in the background which keeps the GPU powered on PGI Compiler User s Guide 87 Using an Accelerator gt If you compile a program for a particular accelerator type then run the program on a system without that accelerator or on a system where the target libraries are not in a directory where the runtime library can find them the program may fail at runtime with an error message gt Ifyou set the environment variable PGI ACC NOTIFY to a nonzero integer value the runtime library prints a line to standard error every time it launches a kernel on the accelerator 7 9 Accelerator Directives This section provides an overview of the Fortran and C directives used to delineate accelerator regions and to augment information available to the compiler for scheduling of loops and classification of data 7 9 1 Enable Accel
222. on a Pentium III or Pentium II 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 Using a tp flag option of k8 or p7 produces an executable that runs on most x86 hardware in use today PGI Compiler User s Guide 21 Use Command Line Options For more information about the tp option refer to the tp target target description in the Command Line Options Reference section of the PGI Compiler s Reference Guide 2 0 Frequently used Options In addition to overall performance there are a number of other options that many users find useful when getting started The following table provides a brief summary of these options For more information on these options refer to the complete description of each option available in the Command Line Options Reference section of the PGI Compiler s Reference Guide Also there are a number of suboptions available with each of the M options listed For more information on those options refer to the specific section on M Options by Category Table 6 Commonly Used Command Line Options Use this option To do this fast These options create a generally optimal set of flags for targets that support SIMD capability They incorporate optimization options to enable use of vector streaming SIMD instructions 64 bit targets fastsse and enable vectorization
223. one area of significant change between some versions of Linux is in 1ibpthread so PGI compilers use this shared object for both the option Mconcur auto parallel and the option mp OpenMP programs Typically portability is supported for forward execution meaning running a program on the same or a later version of Linux but not for backward compatibility that is running on a prior release For example a user who compiles and links a program under Suse 9 1 should not expect the program to run without incident on a Red Hat 9 0 system which is an earlier version of Linux It may run but it is less likely Developers might consider building applications on earlier Linux versions for wider usage 12 1 5 Licensing for Redistributable Files The files in the REDIST directories may be redistributed under the terms of the End User License Agreement for the product in which they were included 12 2 Deploying Applications on Windows Windows programs may be linked statically or dynamically PGI Compiler User s Guide 149 Distributing Files Deployment gt A statically linked program is completely self contained created by linking to static versions of the PGI and Microsoft runtime libraries gt A dynamically linked program depends on separate dynamically linked libraries DLLs that must be installed on a system for the application to run on that system Although it may be simpler to install a statically linked executable there a
224. only one operand with both an input and an output role Matching constraints are very similar to the read write output operands mentioned in Output Operands However there is one key difference between read write output operands and matching constraints The matching constraint can have an input expression that differs from its output expression The following example uses different values for the input and output roles E Sep int y 2 void example7 eem cc ME Sil Bi y x ou wo Ye The compiler generates the following assembly for example7 example7 oo DEDU pushq rbp o IDeigat 0 8 movq rsp rbp 5 y DEEL dl a ao RANILE lineno 8 movl y rip eax addl 1 eax movl eax x rip lineno 0 popq rbp reg Variable x gets initialized with the value stored in y which is 2 After adding the resulting value for variable x is 3 Because matching constraints perform an input role for an output operand it does not make sense for the output operand to have the read write modifier In fact the compiler disallows matching constraints with read write output operands The output operand must have a write only modifier PGI Compiler User s Guide 179 C C Inline Assembly and Intrinsics 15 2 3 Clobber List The clobber list is an optional list of strings that hold machine registers used in the asm string Essentially these strings tell the compiler which registers m
225. ons are available with the PGI runtime libraries on Windows The static libraries are used by default PGI Compiler User s Guide 123 Creating and Using Libraries gt You can use the dynamically linked version of the runtime by specifying Bdynamic at both compile and link time cy C on Windows does not support Bdynamic gt You can explicitly specify static linking the default by using Bstatic at compile and link time For details on why you might choose one type of linking over another type refer to Creating and Using Dynamic Link Libraries on Windows 10 6 Creating and Using Static Libraries on Windows The Microsoft Library Manager LIB EXE is the tool that is typically used to create and manage a static library of object files on Windows LIB is provided with the PGI compilers as part of the Microsoft Open Tools Refer to www msdn2 com for a complete LIB reference search for LIB EXE For a list of available options invoke LIB with the switch For compatibility with legacy makefiles PGI provides a wrapper for LIB and LINK called ar This version of ar is compatible with Windows and object file formats PGI also provides ranlib as a placeholder for legacy makefile support 10 6 1 ar command The ar command is a legacy archive wrapper that interprets legacy ar command line options and translates these to LINK LIB options Y
226. onvention the compiler may indiscriminately clobber a result register with an input operand To prevent this behavior apply the early clobber amp modifier An example follows void examplel5 int w 1 aisle AA asm movl 1 0 n addl 2 0 n ew g2 Qd Meare N w eee Mu z 8 isst w ie The previous code example presents an interesting ambiguity because w appears both as an output and as an input operand So the value of z can be either 1 or 2 depending on whether the compiler uses the same register for operand 0 and operand 2 The use of constraint r for operand 2 allows the compiler to pick any general purpose register so it may or may not pick na register a for operand 2 This ambiguity can be eliminated by changing the constraint for operand 2 from r to a so the value of z will be 2 or by adding an early clobber amp modifier so that z will be 1 The following example shows the same function with an early clobber amp modifier void examplel6 int w 1 Sum VAR asm movl 1 0 n WeyelelL 952 Spyro movl 2 1 dime uU w JE ET n 8 Viggo w A Adding the early clobber amp forces the compiler not to use the a register for anything other than operand 0 Operand 2 will therefore get its own register with its own copy of w The result for z in example16 is 1 15 3 Operand Aliases Extended asm specifies operands in assembly strings with a percent
227. oops that operate on single precision data can be higher SSE2 vector instructions can operate on four single precision elements concurrently but only two double precision elements Compiling with CMvec ts imd can result in numerical differences from the executables generated with less optimization Certain vectorizable operations for example dot products are sensitive to order of operations and the associative transformations necessary to enable vectorization or parallelization 3 7 Auto Parallelization using Mconcur With the Mconcur option the compiler scans code searching for loops that are candidates for auto parallelization Mconcur must be used at both compile time and link time When the parallelizer finds opportunities for auto parallelization it parallelizes loops and you are informed PGI Compiler User s Guide 36 Optimizing and Parallelizing of the line or loop being parallelized if the Minfo option is present on the compile line For a complete specification of Mconcur refer to the Optimization Controls section of the PGI Compiler Reference Manual A loop is considered parallelizable if it doesn t contain any cross iteration data dependencies Cross iteration dependencies from reductions and expandable scalars are excluded from consideration enabling more loops to be parallelizable In general loops with calls are not parallelized due to unknown side effects Also loops with low trip counts are not parallelized
228. ops in the source file for example an optimization gt Globally override command line options PGI Compiler User s Guide 48 Optimizing and Parallelizing gt Tune selected routines or loops based on your knowledge or on information obtained through profiling Using Directives and Pragmas provides details on how to add directives and pragmas to your source files 3 13 Execution Timing and Instruction Counting As this chapter describes once you have a program that compiles executes and gives correct results you may optimize your code for execution efficiency Selecting the correct optimization level requires some thought and may require that you compare several optimization levels before arriving at the best solution To compare optimization levels you need to measure the execution time for your program There are several approaches you can take for timing execution gt You can use shell commands that provide execution time statistics gt You can include function calls in your code that provide timing information gt You can profile sections of code Timing functions available with the PGI compilers include these gt 3Ftiming routines gt The SECNDS pre declared function in PGF77 PGF95 or PGFORTRAN gt The SYSTEM CLOCK or CPU CLOCK intrinsics in PGF95 or PGFORTRAN In general when timing a program you should try to eliminate or reduce the amount of system level activities such as I O program loading and
229. or each thread in the team executing the parallel region Then for each variable specified the value of the variable in the master thread of the team is copied to the threadprivate copies at the beginning of the parallel region Specifies that one or more variables should be shared among all threads This clause provides a mechanism to use a private variable to broadcast a 64 Applies to this This clause Applies to this directive pragma Using OpenMP Has this functionality value from one member of a team to the other members DEFAULT PARALLEL parallel PARALLEL DO parallel for PARALLEL SECTIONS PARALLEL WORKSHARE Specifies the behavior of unscoped variables in a parallel region such as the data sharing attributes of variables FINAL TASK task Specifies that all subtasks of this task will be run immediately FIRSTPRIVATE list M du PARALLEL DO parallel for PARALLEL SECTIONS sections PARALLEL single WORKSHARE SECTIONS SINGLE PARALLEL END parallel PARALLEL parallel for PARALLEL DO parallel sections END PARALLEL DO PARALLEL SECTIONS END PARALLEL SECTIONS PARALLEL WORKSHARE LASTPRIVATE list DO PARALLEL DO parallel for END PARALLEL DO parallel sections PARALLEL SECTIONS sections END PARALLEL SECTIONS SECTIONS parallel MERGEABLE TASK NOWAIT DO END DO for SECTIONS sections SINGLE single WORKSHARE END WORKSHARE NUM THREADS
230. or example pgfortran fast Mipa fast inline prog f For all of the PGI Fortran C and C compilers the fast Mipa fast inline options generally produce code that is well optimized without the possibility of significant slowdowns due to pathological cases gt The ast option is an aggregate option that includes a number of individual PGI compiler options which PGI compiler options are included depends on the target for which compilation is performed gt The Mipa fast inline option invokes interprocedural analysis IPA including several IPA suboptions The inline suboption enables automatic inlining with IPA If you do not wish to use automatic inlining you can compile with Mipa fast and use several IPA suboptions without inlining By experimenting with individual compiler options on a file by file basis further significant performance gains can sometimes be realized However depending on the coding style PGI Compiler User s Guide 26 Optimizing and Parallelizing individual optimizations can sometimes cause slowdowns and must be used carefully to ensure performance improvements There are other useful command line options related to optimization and parallelization such as help Minfo Mneginfo dryrun and v 3 2 1 help As described in Help with Command Line Options you can see a specification of any command line option by invoking any of the PGI compilers with help in combination with the o
231. or this reason it is recommended that you check carefully whether particular program units or loops show improved performance when compiled with this option enabled PGI Compiler User s Guide 32 Optimizing and Parallelizing 3 6 1 Vectorization Sub options The vectorizer performs high level loop transformations on countable loops A loop is countable if the number of iterations is set only before loop execution and cannot be modified during loop execution Some of the vectorizer transformations can be controlled by arguments to the Mvect command line option The following sections describe the arguments that affect the operation of the vectorizer In addition some of these vectorizer operations can be controlled from within code using directives and pragmas For details on the use of directives and pragmas refer to Using Directives and Pragmas The vectorizer performs the following operations gt Loop interchange gt Loop splitting gt Loop fusion gt Memory hierarchy cache tiling optimizations gt Generation of SSE instructions on processors where these are supported gt Generation of prefetch instructions on processors where these are supported gt Loop iteration peeling to maximize vector alignment Alternate code generation By default Mvect without any sub options is equivalent to Mvect assoc cachesize c where c is the actual cache size of the machine This enables the options for nested loop transf
232. ore efficient code if they know the specific x86 processor architecture on which the program will run When preparing to deploy your application you should determine whether you want the application to run on the widest possible set of x86 processors or if you want to restrict the application to run on a specific processor or set of processors The restricted approach allows you to optimize performance for that set of processors Different processors have differences some subtle in hardware features such as instruction sets and cache size The compilers make architecture specific decisions such as instruction selection instruction scheduling and vectorization all of which can have a profound effect on the performance of applications Processor specific code generation is controlled by the tp option described in the section tp lt target gt target of the PGI Compiler Reference Manual When an application is compiled without any tp options the compiler generates code for the type of processor on which the compiler is run PGI Compiler User s Guide 150 Distributing Files Deployment 12 3 1 Generating Generic x86 Code To generate generic x86 code use one of the following forms of the t p option on your command line tp px generate code for any x86 cpu type tp p6 generate code for Pentium 2 or greater While both of these examples are good choices for portable execution most users have Pentium 2 or gr
233. ormation and various other vectorizer options These defaults may vary depending on the target system The following table lists and briefly describes some of the Mvect suboptions Table 8 Mvect Suboptions Use this option To instruct the vectorizer to do this Mvect altcode Generate appropriate code for vectorized loops Mvect no assoc Perform disable associativity conversions that can change the results of a computation due to a round off error For example a typical optimization is to change one arithmetic operation to another arithmetic operation that is mathematically correct but can be computationally different and generate faster code This option is provided to enable or disable this transformation since a round off error for such associativity conversions may produce unacceptable results Mvect cachesize n Tiles nested loop operations assuming a data cache size of n bytes By default the vectorizer attempts to tile nested loop operations such as matrix multiply using multi dimensional strip mining techniques to maximize re use of items in the data cache Mvect fuse Enable loop fusion Mvect gather Enable vectorization of indirect array references PGI Compiler User s Guide 33 Optimizing and Parallelizing Use this option To instruct the vectorizer to do this M vect prefetch Automatically generate prefetch instructions when vectorizable loops are encountered even in case
234. ortran options path filename PGI Compiler User s Guide Getting Started Where options is One or more command line options all of which are described in detail in Use Command Line Options path is the pathname to the directory containing the file named by filename If you do not specify the path for a filename the compiler uses the current directory You must specify the path separately for each filename not in the current directory filename is the name of a source file preprocessed source file assembly language file object file or library to be processed by the compilation system You can specify more than one path filename 1 3 2 Command line Options The command line options control various aspects of the compilation process For a complete alphabetical listing and a description of all the command line options refer to Use Command Line Options The following list provides important information about proper use of command line options gt Caseis significant for options and their arguments gt The compiler drivers recognize characters preceded by a hyphen as command line options For example the M1ist option specifies that the compiler creates a listing file The convention for the text of this manual is to show command line options using a dash instead of a hyphen for example you see M1 i st gt The order of options and the filename is flexible That is you can place options before and after
235. ou can use it to create libraries of object files Syntax The syntax for the ar command is this ar options archive object file Where gt The first argument must be a command line switch and the leading dash on the first option is optional gt The single character options such as d and v may be combined into one option such as dv Thus ar dv ar dv andar d v all mean the same thing gt The first non switch argument must be the library name gt Exactly one of d r t or x must appear on the command line Options The options available for the ar command are these PGI Compiler User s Guide 124 Creating and Using Libraries lt This switch is for compatibility it is ignored d Deletes the named object files from the library r Replaces in or adds the named object files to the library t Writes a table of contents of the library to standard out v Writes a verbose file by file description of the making of the new library to standard out x Extracts the named files by copying them into the current directory 10 6 2 ranlib command The ranlib command is a wrapper that allows use of legacy scripts and makefiles that use the ranlib command The command actually does nothing it merely exists for compatibility Syntax The syntax for the ranlib command is this ranlib options archive Options The options available for the ar command are these help Sho
236. owing table summarizes the supported DEC directives For a complete description of each directive refer to the DECS Directives section of the Directives and Pragmas Reference section in the PGI Compiler s Reference Guide Table 26 DEC Directives Summary Table Directive Functionality ALIAS Specifies an alternative name with which to resolve a routine ATTRIBUTES Lets you specify properties for data objects and procedures DECORATE Specifies that the name specified in the ALIAS directive should have the prefix and postfix decorations performed on it that are associated with the calling conventions that are in effect This directive has no effect if ALIAS is not specified DISTRIBUTE Tells the compiler at what point within a loop to split into two loops PGI Compiler User s Guide 119 Chapter 10 CREATING AND USING LIBRARIES A library is a collection of functions or subprograms that are grouped for reference and ease of linking This section discusses issues related to PGI supplied compiler libraries Specifically it addresses the use of C C builtin functions in place of the corresponding libc routines creation of dynamically linked libraries known as shared objects or shared libraries and math libraries This section does not duplicate material related to using libraries for inlining described in Creating an Inline Library or information related to runtime library routines available to OpenMP programmers describ
237. pabilities by adding a line similar to the following one to the sitenvrc file This example sets the compute capability to enable code generation for all of the supported compute capabilities Notice that the compute capabilities are separated by a space set COMPUTECAP 10 11 12 13 20 30 35 Place the si tenvrc file in the following directory where PGI is the PGI installation directory which is typically opt pgi or usr pgi SPGI linux86 64 14 1 bin 7 5 2 Command Line Flag After acquiring the PGI Accelerator compilers license key you can use the acc or ta option with the pgfortran pgcc or pgc commands For more information on the ta flag and the suboptions that relate to the target accelerators refer to Applicable Command Line Options The compiler automatically invokes the necessary CUDA software tools to create the kernel code and embeds the kernels in the Linux object file cy To access the accelerator libraries you must link an accelerator program with the t a flag as well 7 6 Execution Model The execution model targeted by the PGI Accelerator compilers is host directed execution with an attached accelerator device such as a GPU The bulk of a user application executes on the host Compute intensive regions are offloaded to the accelerator device under control of the host PGI Compiler User s Guide 85 Using an Accelerator The accelerator device executes kernels which may be as simple as a tightly nested loop
238. pace from C C name space Use these naming conventions gt Ifyou call a C C function from Fortran you should rename the C C function by appending an underscore or use CSPRAGMA C in the Fortran program For more information on CSPRAGMA C refer to CSPRAGMA C gt f you call a Fortran function from C C you should append an underscore to the Fortran function name in the calling program 13 5 Compatible Data Types Table 29 shows compatible data types between Fortran and C C Table 30 Fortran and C C Representation of the COMPLEX Type shows how the Fortran COMPLEX type may be represented in C C Tip If you can make your function subroutine parameters as well as your return values match types you should be able to use inter language calling Table 29 Fortran and C C Data Type Compatibility Fortran Type lower case C C Type Size bytes real x real 4 x 4 real 8 x double x PGI Compiler User s Guide 155 Inter language Calling Fortran Type lower case C C Type Size bytes Table 30 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 float complex x complex x struct float r i x float complex x double complex x complex 16 x struct double dr di x double complex x 8 8 8 8 double complex x struct double dr di x 16 16 16 16 cy For C C the complex type implies C99 o
239. pecified functions are extracted to a temporary library created during an extract prepass for use during the compilation stage When you use the Mextract option only the extract phase is performed the compile and link phases are not performed The output of an extract pass is a library of functions available for inlining This output is placed in the inline library file specified on the command line with the o filename specification If the library file exists new information is appended to it If the file does not exist it is created You can use a command similar to the following pgfortran Mextract lib lib il myfunc f You can use the Minline option with the Mextract option In this case the extracted library of functions can have other functions inlined into the library Using both options enables you to obtain more than one level of inlining In this situation if you do not specify a library with the Minline option the inline process consists of two extract passes The first pass is a hidden PGI Compiler User s Guide 53 Using Function Inlining pass implied by the Minline option during which the compiler extracts functions and places them into a temporary library The second pass uses the results of the first pass but puts its results into the library that you specify with the o option 4 3 1 Working with Inline Libraries An inline library is implemented as a directory with each inline function in the library stored a
240. phases of compilation preprocessing compiling assembling and linking Once a file is compiled and an executable file is produced you can execute debug or profile the program on your system Executables produced by the PGI compilers are unconstrained meaning they can be executed on any compatible x86 or x64 processor based system regardless of whether the PGI compilers are installed on that system In general using a PGI compiler involves three steps 1 Produce program source code in a file containing a f extension or another appropriate extension as described in Input Files This program may be one that you have written or one that you are modifying 2 Compile the program using the appropriate compiler command 3 Execute debug or profile the executable file on your system You might also want to deploy your application though this is not a required step The PGI compilers allow many variations on these general program development steps These variations include the following gt Stop the compilation after preprocessing compiling or assembling to save and examine intermediate results gt Provide options to the driver that control compiler optimization or that specify various features or limitations gt Include as input intermediate files such as preprocessor output compiler output or assembler output PGI Compiler User s Guide 1 Getting Started 1 2 Creating an Example Let s look at a simple example of u
241. pointer GET SP2 or included in a larger expression such as void stack_pointer GET SP2 sizeof long Which style of macro to use depends on the application If the asm statement needs to be a part of an expression then a macro with a statement expression is a good approach Otherwise a traditional macro like GET SP x will probably suffice 15 6 Intrinsics Inline intrinsic functions map to actual x86 or x64 machine instructions Intrinsics are inserted inline to avoid the overhead of a function call The compiler has special knowledge of intrinsics so with use of intrinsics better code may be generated as compared to extended inline assembly code The PGI Workstation version 7 0 or higher compiler intrinsics library implements MMX SSE SS2 SSE3 SSSE3 SSE4a ABM and AVX instructions The intrinsic functions are available to C and C programs on Linux and Windows Unlike most functions which are in libraries intrinsics are implemented internally by the compiler A program can call the intrinsic functions from C C source code after including the corresponding header file The intrinsics are divided into header files as follows PGI Compiler User s Guide 189 C C Inline Assembly and Intrinsics Table 40 Intrinsic Header File Organization Instructions Header File Instructions Header File The following is a simple example program that calls XMM intrinsics include lt xmmintrin h gt int main m128 A
242. pose co processor attached to a CPU and to which the CPU can offload data and executable kernels to perform compute intensive calculations This section describes a collection of compiler directives used to specify regions of code in Fortran and C programs that can be offloaded from a host CPU to an attached accelerator 7 1 Overview The programming model and directives described in this section allow programmers to create high level host accelerator programs without the need to explicitly initialize the accelerator manage data or program transfers between the host and accelerator or initiate accelerator startup and shutdown Rather all of these details are implicit in the programming model and are managed by the PGI Fortran C and C accelerator compilers The method described provides a model for accelerator programming that is portable across operating systems and various host CPUs and accelerators The directives allow a programmer to migrate applications incrementally to accelerator targets using standards compliant Fortran or C This programming model allows the programmer to augment information available to the compilers including specification of data local to an accelerator region guidance on mapping of loops onto an accelerator and similar performance related details 7 1 1 Components The PGI Accelerator compiler technology includes the following components PGFORTRAN auto parallelizing accelerator enabled Fortran 2003 compil
243. processor specific enhancements easier because they provide a 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 an overview of the Fortran and C intrinsics that the accelerator supports 7 16 1 Supported Fortran Intrinsics Summary Table Table 21 is an alphabetical summary of the supported Fortran intrinsics that the accelerator supports These functions are specific to Fortran 90 95 unless otherwise specified For complete descriptions of these intrinsics refer to Fortran Intrinsics of the PG Fortran Reference Manual In most cases PGI provides support for all the data types for which the intrinsic is valid When support is available for only certain data types the middle column of the table specifies which ones using the following codes for integer S for single precision real C for single precision complex D for double precision real Z for double precision complex Table 21 Supported Fortran Intrinsics This intrinsic Returns this value O pee eama PGI Compiler User s Guide 103 Using an Accelerator This intrinsic Returns this value mp eessembsuesNOmmewee o LNNNNNNLIIILICTOIITDT NK woo NNNM toc emae O CC S NN hw o mnemestenmieiee o woof Ymca NN hr o memmemeremnene O or eesseseenemseesum Em o memmestmered
244. provides 64 bit compilers for the 64 bit Linux Windows and Mac OS X operating systems running on the x64 architecture You can use these compilers to create programs that use 64 bit memory addresses However there are limitations to how this capability can be applied With the exception of Linux86 64 the object file formats on all of the operating systems limit the total cumulative size of code plus static data to 2GB This limit includes the code and statically declared data in the program and in system and user object libraries Linux86 64 implements a mechanism that overcomes this limitations as described in Large Static Data in Linux This section describes the specifics of how to use the PGI compilers to make use of 64 bit memory addressing The 64 bit Windows Linux and OS X environments maintain 32 bit compatibility which means that 32 bit applications can be developed and executed on the corresponding 64 bit operating system The 64 bit PGI compilers are 64 bit applications which cannot run on anything but 64 bit CPUs running 64 bit Operating Systems This section describes how to use the following options related to 64 bit programming fPIC mcemodel medium Mlarge arrays i8 Mlargeaddressaware tp 14 1 Data Types in the 64 Bit Environment The size of some data types can be different in a 64 bit environment This section describes the major differences For detailed information refer to the Fortran C and C Data Typ
245. ps that require synchronization can only be scheduled for synchronous parallel execution 7 7 Memory Model The most significant difference between a host only program and a host accelerator program is that the memory on the accelerator can be completely separate from host memory which is the case on most current GPUs For example gt The host cannot read or write accelerator memory by reference because it is not mapped into the virtual memory space of the host gt All data movement between host memory and accelerator memory must be performed by the host through runtime library calls that explicitly move data between the separate memories PGI Compiler User s Guide 86 Using an Accelerator gt tis not valid to assume the accelerator can read or write host memory though this may be supported by accelerators in the future 7 7 1 Separate Host and Accelerator Memory Considerations The programmer must be aware of the potentially separate memories for many reasons including but not limited to gt Memory bandwidth between host memory and accelerator memory determines the compute intensity required to effectively accelerate a given region of code gt Limited size of accelerator memory may prohibit offloading of regions of code that operate on very large amounts of data 7 7 2 Accelerator Memory On the accelerator side current GPUs implement a weak memory model In particular they do not support memory coherence between
246. ption in question without specifying any input files For example you might want information on O pgfortran help O The resulting output is similar to this O Set opt level All O1 optimizations plus traditional scheduling and global scalar optimizations performed Or you can see the full functionality of help itself which can return information on either an individual option or groups of options pgfortran help help The resulting output is similar to this help groups asm debug language linker opt other overall phase prepro suffix switch target variable Show compiler switches 3 2 2 Minfo You can use the Min fo option to display compile time optimization listings When this option is used the PGI compilers issue informational messages to standard error stderr as compilation proceeds From these messages you can determine which loops are optimized using unrolling SSE instructions vectorization parallelization interprocedural optimizations and various miscellaneous optimizations You can also see where and whether functions are inlined For more information on Minfo refer to Optimization Controls section of the PGI Compiler Reference Manual 3 2 3 Mneginfo You can use the Mneginfo option to display informational messages to standard error stderr that explain why certain optimizations are inhibited For more information on Mneginfo refer to Optimization Controls section of the PGI Compiler Refer
247. r generate better assembly code You typically use directives and pragmas to control the actions of the compiler in a particular portion of a program without affecting the program as a whole You place them in your source code where you want them to take effect A directive or pragma typically stays in effect from the point where included until the end of the compilation unit or until another directive or pragma changes its status For more information on directives and pragmas refer to Using OpenMP and Using Directives and Pragmas PGI Compiler User s Guide 15 Getting Started gt A library is a collection of functions or subprograms used to develop software Libraries contain helper code and data which provide services to independent programs allowing code and data to be shared and changed in a modular fashion The functions and programs in a library are grouped for ease of use and linking When creating your programs it is often useful to incorporate standard libraries or proprietary ones For more information on this topic refer to Creating and Using Libraries gt Environment variables define a set of dynamic values that can affect the way running processes behave on a computer It is often useful to use these variables to set and pass information that alters the default behavior of the PGI compilers and the executables which they generate For more information on these variables refer to Using Environment Variables gt Deploymen
248. r later 13 5 1 Fortran Named 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 here is a Fortran common block INTEGER I COMPLEX C DOUBLE COMPLEX CD DOUBLE PRECISION D COMMONS AO MET CE C This Fortran Common Block is represented in C with the following equivalent extern struct WME a9 struct float real imag c struct double real imag cd double d com PGI Compiler User s Guide 156 Inter language Calling This same Fortran Common Block is represented in C with the following equivalent extern C struct dou ab 9 struct float real imag c struct double real imag cd doublieral com cy Tip For global or external data sharing ex tern C is not required 13 6 Argument Passing and Return Values In Fortran arguments are passed by reference that is the address of the argument is passed rather than the argument itself In C C arguments are passed by value except for strings and arrays which are passed by reference Due to the flexibility provided in C C you can work around these differences Solving the parameter passing differences generally involves intelligent use of the amp and operators in argument passing when C C calls Fortran and in argument declarations when Fortran calls C C
249. r of the program DMA Direct Memory Access a method to move data between physically separate memories this is typically performed by a DMA engine separate from the host CPU that can access the host physical memory as well as an IO device or GPU physical memory GPU a Graphics Processing Unit one type of accelerator device GPGPU General Purpose computation on Graphics Processing Units Host the main CPU that in this context has an attached accelerator device The host CPU controls the program regions and data loaded into and executed on the device Loop trip count the number of times a particular loop executes OpenACC a parallel programming standard describing a set of compiler directives which can be applied to standard C C and Fortran to specify regions of code for offloading from a host CPU to an attached accelerator OpenCL Open Compute Language a standard C like programming environment similar to CUDA that enables portable low level general purpose programming on GPUs and other accelerators Private data with respect to an iterative loop data which is used only during a particular loop iteration With respect to a more general region of code data which is used within the region but is not initialized prior to the region and is re initialized prior to any use after the region Region the dynamic range of a construct including any procedures invoked from within the construct Structured block in C an executable statement possibl
250. racter letter1 integer 4 numintl numint2 real numfloatl double precision numdoubl integer 2 numshorl external EcmEune call f2c func booll letterl numintl numint2 numfloatl numdoubl numshor1 weite be m2 r5 US Healy Wei 15 V booll letterl numintl numint2 numfloatl numdoubl numshorl end C function f2c_func_ define TRUE Oxff define FALSE 0 void f2c func booll letterl numintl numint2 numfloatl1 numdoubl numshorl len letterl ene oyoxolliL Eeee i Hide AE a usque iblloxeue mibrellhoeic il p double numdoubl short numshorl ime lem ileicieexeil booll TRUE letterl v numintl 11 numint2 44 numfloatl 39 6 numdoubl 39 2 numshorl 981 PGI Compiler User s Guide 159 Inter language Calling Compile and execute the program 2c_main f with the call to 2c func using the following command lines S PCO C T20 TUNGE 9 pgfortran f2c func o f2c main f Executing the a out file should produce the following output IECIT MCCC MESS 13 8 2 Example C Calling Fortran The example C Main Program c2f main c shows a C main program that calls the Fortran subroutine shown in Fortran Subroutine c2f sub f gt Each call uses the amp operator to pass by reference gt The call to the Fortran subroutine uses all lower case and a trailing C Main Program c2f main c void main char booll letterl int numintl numint2 float numfl
251. ran preprocessing The drivers use the following conventions filename f indicates a Fortran source file filename F indicates a Fortran source file that can contain macros and preprocessor directives to be preprocessed filename FOR indicates a Fortran source file that can contain macros and preprocessor directives to be preprocessed filename F95 indicates a Fortran 90 95 source file that can contain macros and preprocessor directives to be preprocessed filename f90 indicates a Fortran 90 95 source file that is in freeform format filename f95 indicates a Fortran 90 95 source file that is in freeform format filename cuf indicates a Fortran 90 95 source file in free format with CUDA Fortran extensions filename CUF indicates a Fortran 90 95 source file in free format with CUDA Fortran extensions and that can contain macros and preprocessor directives to be preprocessed filename c indicates a C source file that can contain macros and preprocessor directives to be preprocessed filename C indicates a C source file that can contain macros and preprocessor directives to be preprocessed filename i indicates a preprocessed C or C source file PGI Compiler User s Guide 4 Getting Started filename cc indicates a C source file that can contain macros and preprocessor directives to be preprocessed filename cpp indicates a C source file that can contain macros and preprocessor directives to b
252. ran Syntax acc wait int expr list clause clause 7 10 Accelerator Directive Clauses Table 18 provides an alphabetical listing and brief description of each clause that is applicable for the various Accelerator directives The table also indicates for which directives the clause is applicable Table 18 Directive Clauses Summary Use this clause In these directives To do this async int expr The parallel or kernels region or data operations may be processed asynchronously while the local thread continues with the code following the construct or directive The async clause on the data construct is a PGI extension Parallel Kernels Data Enter Data Exit Data Update collapse n Specifies how many tightly nested loops are associated with the loop construct copy list Parallel At execution time the implementation will test whether each of the items in the list are already present in the device memory If not corresponding device memory will be allocated and the data will be copied to the device at entry to the region and data will be Declare copied back to the host and the memory deallocated at exit from the region For compatibility with OpenACC 2 0 this may also be spelled present_or_copy or pcopy Kernels Data Region copyin list At execution time the implementation will test whether each of Parallel the items in the list are already present in the device memory If Kernels n Dat
253. re advantages to using the DLL versions of the runtime including gt Executable binary file size is smaller gt Multiple processes can use DLLs at once saving system resources gt New versions of the runtime can be installed and used by the application without rebuilding the application Dynamically linked Windows programs built with PGI compilers depend on dynamic runtime library files DLLs These DLLs must be distributed with such programs to enable them to execute on systems where the PGI compilers are not installed These redistributable libraries include both PGI runtime libraries and Microsoft runtime libraries 12 2 1 PGI Redistributables PGI redistributable directories contain all of the PGI Linux runtime library shared object files or Windows dynamically linked libraries that can be re distributed by PGI 15 10 licensees under the terms of the End User License Agreement EULA 12 2 2 Microsoft Redistributables The PGI products on Windows include Microsoft Open Tools The Microsoft Open Tools directory contains a subdirectory named redist PGI licensees may redistribute the files contained in this directory in accordance with the terms of the End User License Agreement Microsoft supplies installation packages vcredist_x86 exe and vcredist x64 exe containing these runtime files These files are available in the redist directory 12 3 Code Generation and Processor Architecture The PGI compilers can generate much m
254. rectives and Pragmas sssseeeeennennenn nennen 152 Chapter 13 Inter language Callitig nerit AE 153 13 1 Overview of Calling Conventions seen tentent tentent tentent 153 13 2 Inter language Calling Considerations seen 154 13 3 Eunetions and SUbEOLTITIeS s ede RERO E oe RE ca cede aes 154 13 4 Upper and Lower Case Conventions Underscores sssssssss sese tnn 155 13 5 Compatible Data Types cotta ette etait epe Rem Pe tert eere E Haake Dee Er Pee P VEU RH vv 155 13 5 1 Fortran Named Common BlockS m eot into afe eb aie a cam dum a m nii 156 13 6 Argument Passing and Returni ValU8s reote reti tct nb et e o rt t 157 13 631 Passing by Val e VAL sssrini eret citet reta ete i rid ete nitet trt ted 157 13 6 2 Character Return Values ssssssssseeseeeneeete ener tentent netten 157 13 6 3 Complex Return Values cetera cr te ce tte t rmt tret hide qe rte t petat 158 T9 Atray Indies uec i ren he aen O 158 Ibn EXAMS cC r E 159 138 1 Example Fortran Calling C 5 310 ductae tre t ettet eerte aie ea 159 13 8 2 Example C Calling Fortran ssssseeeeenennenene nennen tne 160 13 8 3 Example C Calling 51 attinent e reri ert t t ec i e 161 138 4 CUEeEe nerl m A a EE 161 13 6 5 Example Fortran Calling Gt 5 2 nete etie ten eai een nde eo nt m and
255. rintf func2 data d n data printf func2 i d n i data i prog2 c int declspec dllimport data void declspec dllimport func2 int aae main data Hip Func 2 C122 8 printf main data d n data ceturn Or PGI Compiler User s Guide 128 Creating and Using Libraries 1 Step 1 Create the DLL 0532 d11 and its import library obj2 1ib using the following series of commands DECE Eeyan Neb NC pgcc Mmakedll object2 0bj o obj2 dll 2 Step 2 Compile the main program 9 pgcc Bdynamic o prog2 prog2 c defaultlib obj2 The Bdynamic switch causes the compiler to link against the PGI runtime DLLs instead of the PGI runtime static libraries The Bdynamic switch is required when linking against any PGI compiled DLL such as obj 2 d11 The defaultlib switch specifies that obj2 1ib the DLL s import library should be used to resolve the imported data and subroutine in prog2 c 3 Step 3 Ensure that obj 2 d11 is in your path then run the executable prog2 to determine if the DLL was successfully created and linked prog2 func2 data 11 func2 i 12 main data 12 Should you wish to change ob32 d11 without changing the subroutine or function interfaces no rebuilding of prog2 is necessary Just recreate obj 2 d11 and the new 0b32 d11 is loaded at runtime 10 7 3 Build DLLs Containing Circular Mutual Imports C In this example we build two DLLs
256. rize yourself with vi or other editors and with makefiles The Web has an extensive online tutorial available for the vi editor as well as a number of thorough introductions to the construction and use of makefiles ar or ranlib For library compatibility PGI provides versions of ar and ranlib that are compatible with native Windows object file formats For more information on these commands refer to Creating and Using Static Libraries on Windows 1 7 4 Using the PGI Compilers on OS X PGI Workstation 15 10 for OS X supports most of the features of the 32 and 64 bit versions for Linux environments Typically the PGI compilers and tools on OS X function identically to their Linux counterparts PGI Compiler User s Guide 12 Getting Started OS X Header Files The OS X header files contain numerous non standard extensions PGI supports many of these extensions thus allowing the PGI C and C compilers to compile most programs that the GNU compilers can compile A few header files not interoperable with the PGI compilers have been rewritten These files are included in PGI osx86 15 10 include or PGI osx86 64 15 10 include These files are stdarg h stddef h and others If you are using the PGI C or C compilers please make sure that the supplied versions of these include files are found before the system versions This will happen by default unless you explicitly add a I option that references one of the system include directories
257. rs and tools available from The Portland Group gt PGI Fortran Reference manual describes the FORTRAN 77 Fortran 90 95 Fortran 2003 statements data types input output format specifiers and additional reference material related to use of the PGI Fortran compilers gt System V Application Binary Interface Processor Supplement by AT amp T UNIX System Laboratories Inc Prentice Hall Inc gt System V Application Binary Interface X86 64 Architecture Processor Supplement http www Xx86 64 0rg abi pdf PGI Compiler User s Guide xvi Preface gt Fortran 95 Handbook Complete ISO ANSI Reference Adams et al The MIT Press Cambridge Mass 1997 gt Programming in VAX Fortran Version 4 0 Digital Equipment Corporation September 1984 gt IBM VS Fortran IBM Corporation Rev GC26 4119 gt The C Programming Language by Kernighan and Ritchie Prentice Hall gt C A Reference Manual by Samuel P Harbison and Guy L Steele Jr Prentice Hall 1987 gt The Annotated C Reference Manual by Margaret Ellis and Bjarne Stroustrup AT amp T Bell Laboratories Inc Addison Wesley Publishing Co 1990 PGI Compiler User s Guide xvii Preface PGI Compiler User s Guide xviii Chapter 1 GETTING STARTED This section describes how to use the PGI compilers 1 1 Overview The command used to invoke a compiler such as the pgfortran command is called a compiler driver The compiler driver controls the following
258. rt help information is printed out V Version information is printed out 10 7 Creating and Using Dynamic Link Libraries on Windows There are several differences between static and dynamic link libraries on Windows Libraries of either type are used when resolving external references for linking an executable but the process differs for each type of library When linking with a static library the code needed from the library is incorporated into the executable When linking with a DLL external references are resolved using the DLL s import library not the DLL itself The code in the DLL associated with the external references does not become a part of the executable The DLL is loaded when the executable that needs it is run For the DLL to be loaded in this manner the DLL must be in your path Static libraries and DLLs also handle global data differently Global data in static libraries is automatically accessible to other objects linked into an executable Global data in a DLL can only be accessed from outside the DLL if the DLL exports the data and the image that uses the data imports it PGI Compiler User s Guide 125 Creating and Using Libraries To access global data the C compilers support the Microsoft storage class extensions declspec dllimport and declspec dllexport These extensions may appear as storage class modifiers and enable functions and data to be imported and exported extern int _declspec dllimport intfun
259. ry sse tentent ntes 67 OpenMP related Environment Variable Summary Table ssssseeeeeeeneenenn 72 MPI Distribution Options sssssseseeenne entente tenerent tentent tenerent te tenens 75 MPI Profiling OPUNE oreinen cane dent caste dats deen manatee demain ead 76 PGI Accelerator Directive Summary Table ssssssesseeneeenentete entente tenentes 91 Directive Clauses SUMMA M 95 Accelerator Runtime Library Routines essei nennen tnter tentent tette tet aineen 99 Accelerator Environment Variables sse tenente tenes 100 Supported Fortran Intrinsics corteo terea tr recette e d 103 Supported C Intrinsic Double Functions sse tnnt tenens 104 Supported C Intrinsic Float FUNCUONS iine teret tem i e ee tete ee 105 Proprietary Optimization Related Fortran Directive and C C Pragma Summary s s s 111 PGI Compiler User s Guide X Table 25 Table 26 Table 27 Table 28 Table 29 Table 30 Table 31 Table 32 Table 33 Table 34 Table 35 Table 36 Table 37 Table 38 Table 39 Table 40 IGNORE TKR Exarmple tti eed ert er e rie de tll 118 DECS Directives Summary Table ht e ng E ETE RA EE Re 119 PGI Related Environment Variable Summary sees 137 Supported PGI TERM Values c cccecscseceseesesesseeeseesesececseeeeeseseeacaeeceaeansusseeees
260. ry of terms which you can access at http www pgroup com support definitions htm The following table lists the PGI compilers and tools and their corresponding commands Table 1 PGI Compilers and Commands Compiler or Tool Language or Function Command PGF77 ANSI FORTRAN 77 pgf77 PGCC ISO ANSI C11 and K amp R C pgcc PGC ISO ANSI C 11 with GNU compatibility pgc on Linux and OS X ISO ANSI C 03 DEPRECATED pgcpp on Linux OS X and Windows DEPRECATED PGDBG Source code debugger pgdbg PGI Compiler User s Guide XV Preface Compiler or Tool Language or Function Command PGPROF Performance profiler pgprof In general the designation PGI Fortran is used to refer to The Portland Group s Fortran 2003 compiler and pgfortran is used to refer to the command that invokes the compiler A similar convention is used for each of the PGI compilers and tools For simplicity examples of command line invocation of the compilers generally reference the pgfortran command and most source code examples are written in Fortran Usage of the PGF77 compiler whose features are a subset of PGFORTRAN is similar Usage of PGC and PGCC is consistent with PGFORTRAN and PGF77 though there are command line options and features of these compilers that do not apply to PGFORTRAN and PGF77 and vice versa There are a wide variety of x86 compatible processors in use All are supported by the PGI compilers and tools Most of these processors are forw
261. s a file using an encoded form of the inlinable function A special file named TOC in the inline library directory serves as a table of contents for the inline library This is a printable ASCII file which you can examine to locate information about the library contents such as names and sizes of functions the source file from which they were extracted the version number of the extractor which created the entry and so on Libraries and their elements can be manipulated using ordinary system commands gt Inline libraries can be copied or renamed gt Elements of libraries can be deleted or copied from one library to another gt The 1s or dir command can be used to determine the last change date of a library entry 4 3 2 Dependencies When a library is created or updated using one of the PGI compilers the last change date of the library directory is updated This allows a library to be listed as a dependence in a makefile and ensures that the necessary compilations are performed when a library is changed 4 3 3 Updating Inline Libraries Makefiles If you use inline libraries you must be certain that they remain up to date with the source files into which they are inlined One way to assure inline libraries are updated is to include them in a makefile The makefile fragment in the following example assumes the file utils f contains a number of small functions used in the files parser f and alloc f This portion of the make
262. s and cache size The compilers make architecture specific decisions about things such as instruction selection instruction scheduling and vectorization By default the PGI compilers produce code specifically targeted to the type of processor on which the compilation is performed That is the default is to use all supported instructions wherever possible when compiling on a given system As a result executables created on a given system may not be usable on previous generation systems For example executables created on a Pentium 4 may fail to execute on a Pentium III or Pentium II All PGI compilers have the capability of generating unified binaries which provide a low overhead means for generating a single executable that is compatible with and has good performance on more than one hardware platform You can use the tp option to control compilation behavior by specifying the processor or processors with which the generated code is compatible The compilers generate and combine into one executable multiple binary code streams each optimized for a specific platform At runtime the executable senses the environment and dynamically selects the appropriate code stream For specific information on the tp option refer to the PGI Compiler Reference Manual Executable size is automatically controlled via unified binary culling Only those functions and subroutines where the target affects the generated code have unique binary images resulting in a
263. s for lea instruction Constant in range of 0 to 255 e g for out instruction Same as r simple constraint Same as r simple constraint Same as r simple constraint si register e g sil si Yedi 9orsi Constant in range of Oto 127 The following example uses the x result in a or XMM register constraint to subtract c from b and store the double examplell double a double b 400 99 double c 300 98 asm Uwe 2 Sp 2 Nay a won icy p Marn cy return a PGI Compiler User s Guide 183 C C Inline Assembly and Intrinsics The generated assembly for this example is this examplell Pee stato E pushq rbp co IDXG3Eat 0 movq rsp rbp Derm oo ENIS lineno 4 movsd C00128 rip xmm1 movsd C00130 rip xmm2 movapd Sxmml xmm0O subpd xmm2 xmm0 lineno 10 lineno 11 popq rbp rot If a specified register is not available the pgcc and pgc compilers issue an error message For example pgcc and pgc reserves the ebx register for Position Independent Code PIC on 32 bit system targets If a program has an asm statement with a b register for one of the operands the compiler will not be able to obtain that register when compiling for 32 bit with the fPIC switch which generates PIC To illustrate this point the following example is compiled for a 32 bit target using PIC void examplel2 int x 1 int y 1
264. s 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 PGI Compiler User s Guide xiv Preface 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 C C C language statements are shown in the test of this guide using a reduced fixed point size The PGI compilers and tools are supported on both 32 bit and 64 bit variants of the Linux OS X and Windows operating systems on a variety of x86 compatible processors There are a wide variety of releases and distributions of each of these types of operating systems Terms A number of terms related to systems processors compilers and tools are used throughout this guide For example For a complete definition of these terms and other terms in this guide with which you may be unfamiliar PGI provides a glossa
265. s number The first operand is number 0 the second is number 1 the third is number 2 and so on In the preceding example 0 references the output operand and 1 references the input operand The asm string also contains eax which references machine register eax Hard coded registers like eax should be specified in the clobber list to prevent conflicts with other instructions in the compiler s assembly language output output operands input operands and clobber list items are described in more detail in the following sections 15 2 1 Output Operands The output operands are an optional list of output constraint and expression pairs that specify the result s of the asm statement An output constraint is a string that specifies how a result is PGI Compiler User s Guide 176 C C Inline Assembly and Intrinsics delivered to the expression For example r x says the output operand is a write only register that stores its value in the C variable x at the end of the asm statement An example follows dH OSA void example ewe mrowil SO Cu ca ere dep Xp The previous example assigns 0 to the C variable x For the function in this example the compiler produces the following assembly If you want to produce an assembly listing compile the example with the pgcc S compiler option example ao DEROO pushq rbp 4 IDIGIE SLO S movq rsp rbp ou Dear aL dL o ENIES lineno 8
266. s of approximately 10 or fewer statements are inlined one level only The specified functions are inlined only if they are previously placed in the inline library temp i1 during the extract phase pgfortran dhry f Mextract lib temp il pgfortran dhry f Minline 10 proc7 temp il Using the same source file dhry f the following example builds an executable for dh ry in which all functions of roughly ten or fewer statements are inlined Two levels of inlining are performed This means that if function A calls function B and B calls C and both B and C are inlinable then the version of B which is inlined into A will have had C inlined into it pgfortran dhry f Minline size 10 levels 2 4 6 Restrictions on Inlining The following Fortran subprograms cannot be extracted gt Main or BLOCK DATA programs gt Subprograms containing alternate return assigned GO TO DATA SAVE or EQUIVALENCE statements PGI Compiler User s Guide 55 Using Function Inlining gt Subprograms containing FORMAT statements gt Subprograms containing multiple entries A Fortran subprogram is not inlined if any of the following applies gt tis referenced in a statement function gt Acommon block mismatch exists in other words the caller must contain all common blocks specified in the callee and elements of the common blocks must agree in name order and type except that the caller s common block can have additional memb
267. s pre configured these wrappers for use with Mprof 6 10 Using SGI MPI on Linux PGI compilers and tools support SGI s version of MPI If you want to build your MPI application using an instance of SGI MPI you must first set either MPIDIR or MPI_ROOT to the root of the SGI MPI installation directory that you want to use that is the directory that contains bin include lib and so on Then you can compile with the Mmpi sgimpi option Add g for debugging or use Mprof sgimpi instead to instrument for profiling To run programs built with SGI MPI you must include the SGI MPI lib directory in the LD LIBRARY PATH environment variable 6 11 Using MPI Compiler Wrappers When you use MPI compiler wrappers to build with the pic or mcmodel medium options then you must specify sh1ib to link with the correct libraries Here are a few examples For a static link to the MPI libraries use this command 9 mpicc hello f For a dynamic link to the MPI libraries use this command 9 mpicc hello f shlib To compile with pic which by default invokes dynamic linking use this command 9 S oee mre S e To compile with mcmodel medium use this command 9 mpicc mcmodel medium shlib hello f 6 12 Limitations The Open Source Cluster utilities in particular the MPICH and ScaLAPACK libraries are provided with support necessary to build and define their proper use However use of these libraries on linu
268. s where SSE instructions are not generated Mvect short Enable short vector operations Mvect simd Automatically generate packed SSE Streaming SIMD Extensions and prefetch instructions when vectorizable loops are encountered SIMD instructions first introduced on Pentium Ill and AthlonXP processors operate on single precision floating point data Mvect sizelimit n Limit the size of vectorized loops Mvect tile Enable loop tiling Mvect uniform Perform consistent optimizations in both vectorized and residual loops Be aware that this may affect the performance of the residual loop Inserting no in front of the option disables the option For example to disable the generation of SSE or SIMD instructions compile with Mvect nosimd 3 6 2 Vectorization Example Using SIMD Instructions One of the most important vectorization options is Mvect simd When you use this option the compiler automatically generates SSE instructions where possible when targeting processors on which these instructions are supported This process can improve performance by several factors compared with the equivalent scalar code All of the PGI Fortran C and C compilers support this capability The PGI Release Notes show which x86 and x64 processors PGI supports In the program in Vector operation using SIMD instructions the vectorizer recognizes the vector operation in subroutine loop when either the compiler switch Mvect simd or fast is us
269. se cdt8 and so on 5 Click OK in the Add Repository dialog PGI Compiler User s Guide 107 Eclipse The install form now shows The Portland Group C C Compiler Plugin as an option to install 6 Check the box next to The Portland Group option and select Next to get to the Install Details view 7 Click Next again 8 Review and accept the End User License agreement 9 Click Finish You are prompted to restart Select Restart to complete installation of the plugin 8 2 Use Eclipse CDT To use the Eclipse plugin for the PGI C and C compilers the directory containing PGI compilers and tools should be included in your PATH prior to starting the Eclipse IDE For details on how to include this directory in your PATH environment variable refer to Using Environment Variables and specifically to PATH This plugin does not currently support the Code Analysis feature of Eclipse CDT This feature is not disabled by default for PGI projects and is the cause of spurious syntax errors during pre compilation This feature can be disabled manually at either the Project or Workspace level any actual compilation or link errors are reported at build time gt To disable Code Analysis at the Workspace level select the menu item Window Preferences then select C C then Code Analysis uncheck all category items which should deselect all sub items gt To disable Code Analysis at the Project level select the menu item Project
270. sem SIE endif enddo v t call Ei The value of t may not be computed on the last iteration of the loop Normally if a scalar is assigned within a loop and used following the loop the PGI compilers save the last value of the scalar However if the loop is parallelized and the scalar is not assigned on every iteration it may be difficult without resorting to costly critical sections to determine on what iteration t is last assigned Analysis allows the compiler to determine that a scalar is assigned on each iteration and hence that the loop is safe to parallelize if the scalar is used later as illustrated in the following C C and Fortran examples PGI Compiler User s Guide 40 C CHY version oue wail Paap alse 4 ie x25 4 t 2 0 else t 3 0 y Dac seg ae Fortran version do I 1 N aam CL 18 On O ite era t 2 0 else t 3 0 y i t endif enddo v t Optimizing and Parallelizing Notice that t is assigned on every iteration of the loop However there are cases where a scalar may be privatizable but if it is used after the loop it is unsafe to parallelize Examine the following loops in which each use of t within the loop is reached by a definition from the same Iteration C C Version igne GL al p ste o ber dt ase ise E363 SO O 4 t ex i y lboc een Mesbr f v Fortran Version do I 1 N if x I gt 0 0 then t x I y i t endif enddo v t ras ar
271. sented by g should be optimized for both k8 64 and p7 64 pgi g pgi tp k8 64 p7 64 The syntax of the C C pragma is pragma global routine tp target where the scope is global routine or blank The default is routine For example the following syntax indicates that the next function should be optimized for k8_64 p7_64 and core2_64 pragma routine tp k8 64 p7 64 core2 64 PGI Compiler User s Guide 152 Chapter 13 INTER LANGUAGE CALLING This section describes inter language calling conventions for C C and Fortran programs using the PGI compilers Fortran 2003 provides a mechanism to support the interoperability with C This includes the ISO_C_Binding intrinsic module binding labels and the BIND attribute In the absence of this mechanism the following sections describe how to call a Fortran function or subroutine from a C or C program and how to call a C or C function from a Fortran program For information on calling assembly language programs refer to the Runtime Environment section of the PGI Compilers Reference Guide This section provides examples that use the following options related to inter language calling For more information on these options refer to the Command Line Options Reference section of the PGI Compiler Reference Guide x Mnomain Miface Mupcase 13 1 Overview of Calling Conventions This section includes information on the following topics gt Functions an
272. show the effect of directives as well as their scope Consider the following Fortran code integer maxtime time parameter n 1000 maxtime 10 double precision a n n b n n c n n do time 1 maxtime do i l n Clo J i m eri a0 b73 enddo enddo enddo When compiled with Mvect both interior loops are interchanged with the outer loop pgfortran Mvect dirvectl f Directives alter this behavior either globally or on a routine or loop by loop basis To assure that vectorization is not applied use the novector directive with global scope cpgi g novector integer maxtime time parameter n 1000 maxtime 10 double precision a n n b n n c n n do time 1 maxtime glor ak db im oe 3 d m G r T a i j is b i j enddo enddo enddo end PGI Compiler User s Guide 112 Using Directives and Pragmas In this version the compiler disables vectorization for the entire source file Another use of the directive scoping mechanism turns an option on or off locally either for a specific procedure or for a specific loop integer maxtime time parameter n 1000 maxtime 10 double precision a n n b n n c n n cpgi l novector do time 1 maxtime clo ab dL in do j 1 n eli e alii s 3 53 enddo enddo enddo Loop level scoping does not apply to nested loops That is the directive only applies to the following loop In this example the directive turns off vector transformations for the top
273. simultaneous assignment into a 1 n will produce values different from sequential execution of the loops In this example values computed for a 1 n don t depend on j so that simultaneous assignment by both processors does not yield incorrect results However it is beyond the scope of the compilers dependence analysis to determine that values computed in one iteration of a loop don t differ from values computed in another iteration So the worst case is assumed and different iterations of the outer loop are assumed to compute different values for a 1 n Is this assumption too pessimistic If j doesn t occur anywhere within a loop the loop exists only to cause some delay most probably to improve timing resolution It is not usually valid to parallelize timing loops to do so would distort the timing information for the inner loops Scalars Quite often scalars will inhibit parallelization of non innermost loops There are two separate cases that present problems In the first case scalars appear to be expandable but appear in non innermost loops as in the following example There are a number of technical problems to be resolved in order to recognize expandable scalars in non innermost loops Until this generalization occurs scalars like x in the preceding code segment inhibit parallelization of loops in which they are assigned In the following example scalar k is not expandable and it is not an accumulator for a reduction k 1
274. sing the PGI compiler to create compile and execute a program that prints hello 1 Create your program For this example suppose you enter the following simple Fortran program in the file hello f IX ME a Vello end Compile the program When you created your program you called it hello f In this example we compile it from a shell command prompt using the default pgfortran driver option Use the following syntax EGIS pgfortran hello f By default the executable output is placed in the file a out or on Windows platforms in a filename based on the name of the first source or object file on the command line However you can specify an output file name by using the o option To place the executable output in the file hello use this command PGIS pgfortran o hello hello f Execute the program To execute the resulting hello program simply type the filename at the command prompt and press the Return or Enter key on your keyboard PGI hello hello 3 Invoking the Command level PGI Compilers To translate and link a Fortran C or C program the pgf77 pgf95 pgfortran pgcc and pgc commands do the following 1 2 3 4 i Preprocess the source text file Check the syntax of the source text Generate an assembly language file Pass control to the subsequent assembly and linking steps 3 1 Command line Syntax The compiler command line syntax using pgfortran as an example is pgf
275. sing the pgcc command level compiler driver multiple source files can be compiled and linked into a single executable with one command The following example compiles and links three source files 9 PCS a Owe ilele tilez e tiles e In actuality the pgcc driver executes several steps to produce the assembly code and object files corresponding to each source file and subsequently to link the object files together into a single executable file This command is roughly equivalent to the following commands performed individually o 5 jie SNC ON iio S MEERSIKCTNN 5 as o filel o filel s paee NONEM COMICI COM as o file2 o file2 s paee e ileg s IEINI COP sas oO FeSO fries is a oee O a OU mielo riea Ge NP ode oe oe oe o If any of the three source files is edited the executable can be rebuilt with the same command line 9 Pee a OU tiLlel c ilez ESRB This always works as intended but has the side effect of recompiling all of the source files even if only one has changed For applications with a large number of source files this can be time consuming and inefficient PGI Compiler User s Guide 43 Optimizing and Parallelizing 3 9 2 Building a Program Without IPA Several Steps It is also possible to use individual pgcc commands to compile each source file into a corresponding object file and one to link the resulting object files into an executable oe pee Enikail o jee ied s ge
276. sion filename f indicates a preprocessed file if you compiled a Fortran file using the F option filename i indicates a preprocessed file if you compiled using the P option PGI Compiler User s Guide 6 Getting Started filename 1lst indicates a listing file from the Ml ist option filename o or filename obj indicates a object file from the c option filename s indicates an assembly language file from the S option Unless you specify otherwise the destination directory for any output file is the current working directory If the file exists in the destination directory the compiler overwrites it The following example demonstrates the use of output filename extensions SESSEL C DIE OC DIE ONES This produces the output files proto o and protol o or on Windows proto obj and protol obj allof which are binary object files Prior to compilation the file protol F is preprocessed because it has a F filename extension 1 5 Fortran C and C Data Types The PGI Fortran C and C compilers recognize scalar and aggregate data types A scalar data type holds a single value such as the integer value 42 or the real value 112 6 An aggregate data type consists of one or more scalar data type objects such as an array of integer values For information about the format and alignment of each data type in memory and the range of values each type can have on x86 or x64 processor based systems running a 32 bit operating
277. software On Windows LM LICENSE FILE does not need to be set MANPATH Sets the directories that are searched for manual pages associated with the command that the user types On OS X MANPATH does not need to be set MPSTKZ Increases the size of the stacks used by threads executing in parallel regions The value should be an integer n concatenated with M or m to specify stack sizes of n megabytes MP_BIND Specifies whether to bind processes or threads executing in a parallel region to a physical processor MP_BLIST WhenMP BINDis yes this variable specifically defines the thread CPU relationship overriding the default values MP_SPIN Specifies the number of times to check a semaphore before calling sched_yield on Linux or Mac OS X or _sleep on Windows MP_WARN Allows you to eliminate certain default warning messages NCPUS Sets the number of processes or threads used in parallel regions NCPUS_MAX Limits the maximum number of processors or threads that can be used in a parallel region NO_STOP_MESSAGE If used the execution of a plain STOP statement does not produce the message FORTRAN STOP OMP_DYNAMIC Currently has no effect Enables TRUE or disables FALSE the dynamic adjustment of the number of threads The default is FALSE OMP_MAX_ACTIVE_LEVELS Specifies the maximum number of nested parallel regions OMP_NESTED Currently has no effect Enables TRUE or disables FALSE nested parallelism The default is FALSE
278. statement is done with eax it is restored with the popl instruction Typically a program uses macros that enclose asm statements The following two examples use the interrupt constructs created previously in this section Wee HSESISSbI Im ST SUERTE define enabletne _ asm SERE 15 2 Extended Inline Assembly Inline Assembly explains how to use inline assembly to specify machine specific instructions inside a C function This approach works well for simple machine operations such as disabling and enabling system interrupts However inline assembly has three distinct limitations PGI Compiler User s Guide 175 C C Inline Assembly and Intrinsics 1 The programmer must choose the registers required by the inline assembly 2 To prevent register clobbering the inline assembly must include push and pop code for registers that get modified by the inline assembly 3 There is no easy way to access stack variables in an inline assembly statement Extended Inline Assembly was created to address these limitations The format for extended inline assembly also known as extended asm is as follows cS NEVOISCHSECEENEV clas Sy string output operands input operands clobberlist gt Extended asm statements begin with the asm or __asm__ keyword Typically the __asm__ keyword is used in header files that may be included by ISO C programs gt An optional volatile or __volatile__ Keyword may appear after the asm key
279. stop before the assembly phase the compiler takes no action on the assembly language file Processing stops after compilation and the assembler does not run In this scenario the compilation must have been completed in a previous pass which created the s file For a complete description of the S option refer to Output Files In addition to specifying primary input files on the command line code within other files can be compiled as part of include files using the INCLUDE statement in a Fortran source file or PGI Compiler User s Guide 5 Getting Started the preprocessor include directive in Fortran source files that use a F extension or C and C source files When linking a program with a library the linker extracts only those library components that the program needs The compiler drivers link in several libraries by default For more information about libraries refer to Create and Use Libraries 1 4 2 Output Files By default an executable output file produced by one of the PGI compilers is placed in the file a out or on Windows in a filename based on the name of the first source or object file on the command line As the Hello example shows you can use the o option to specify the output file name If you use one of the options F Fortran only P C C only S or c the compiler produces a file containing the output of the last completed phase for each input file as specified by the option supplied The outp
280. t though possibly an infrequent task can present some unique issues related to concerns of porting the code to other systems Deployment in this context involves distribution of a specific file or set of files that are already compiled and configured The distribution must occur in such a way that the application executes accurately on another system which may not be configured exactly the same as the system on which the code was created For more information on what you might need to know to successfully deploy your code refer to Distributing Files Deployment gt An intrinsic is a function available in a given language whose implementation is handled specially by the compiler Intrinsics make using processor specific enhancements easier because they provide a C C language interface to assembly instructions In doing so the compiler manages details that the user would normally have to be concerned with such as register names register allocations and memory locations of data For C C programs PGI provides support for MMX SSE SSE2 SSE3 SSSE3 SSE4A ABM and AVX intrinsics For more information on these intrinsics refer to the C C MMX SSE Inline Intrinsics section of the PGI Compiler s Reference Manual PGI Compiler User s Guide 16 Chapter 2 USE COMMAND LINE OPTIONS A command line option allows you to control specific behavior when a program is compiled and linked This section describes the syntax for properly using command
281. t IPA Using Make sseeeneenenenete nene 44 3 944 Building a Program with PA cocta ertt tereti ath cau edere titt Geri 44 3 9 5 Building a Program with IPA Single Step seen tenente tenens 45 3 9 6 Building a Program with IPA Several Steps tenentes 45 3 9 7 Building a Program with IPA Using Make sees 46 3 9 8 Questions about IA a cert pre nae te o ertet Gal Dep e e teet ate Rete ant d Rd 46 3 10 Profile Feedback Optimization using Mpfi Mpfo ccccececceccceesesessesesceeeseseeseeeeeaesnsaceeeceaeansusaeeeseeatensaseeeeeates 4T 3 41 Default Optimization Eevels icai rtt tre rrr creer te retire eat epe Co ere PEL Ere Pe edd 48 3 12 Local Optimization Using Directives and Pragmas sese 48 3 13 Execution Timing and Instruction Counting esee tenente tenen 49 3 14 Portability of Multi Threaded Programs on Linux seen 49 Os VA Ass lIDRUTTI US secet retro titre Seiten ida ted ettet Mes Lem ET pate eet LTD EE 50 PGI Compiler User s Guide iii Chapter 4 Using Function Inlihing era cete ER ter ar obrem bore pere bec Re Aber EDT aac 51 4 1 Invoking Function Inlining seseeeeneee eene tenerent tenente tentent tentent tenen 51 4 2 Using an Inline LBN s c terere tercera eere ae ann nei A E EE 52 4 3 Creating arn Inline EIbrary o emi b nhe ehe i re pne nad
282. t PGI compilers and tools The default environment variables are already set and available The command window launched by PGI Workstation can be customized using the Properties selection on the menu accessible by right clicking the window s title bar Debugger amp Profiler Submenu From the Debugger amp Profiler menu you have access to the PGI debugging and profiling tools PGDBG is a symbolic debugger for Fortran C C and assembly language programs It provides debugger features such as execution control using breakpoints single stepping and examination and modification of application variables memory locations and registers gt PGDBG Debugger Select this option to launch the PGI debugger PGDBG for use with both 32 bit and 64 bit applications gt PGPROF Performance Profiler Select this option to launch the PGPROF Performance Profiler PGPROF provides a way to visualize and diagnose the performance of the components of your program and provides features for helping you to understand why certain parts of your program have high execution times Documentation Submenu From the Documentation menu you have access to all PGI documentation that is useful for PGI users The documentation that is available includes the following gt AMD Core Math Library Select this option to display documentation that describes elements of the AMD Core Math Library a software development library released by AMD that includes a set o
283. t a keyword then a name with a period is assumed to be an inline library and a name without a period is assumed to be a function name If a number is used without a keyword the number is assumed to be a size In the following example the compiler inlines functions with fewer than approximately 100 statements in the source file myprog f and writes the executable code in the default output file a out pgfortran Minline size 100 myprog f Refer to M Options by Category in the PGI Compiler s Reference Guide For more information on the Minline options refer to M Options by Category section of the PGI Compiler s Reference Guide 4 2 Using an Inline Library If you specify one or more inline libraries on the command line with the Minline option the compiler does not perform an initial extract pass The compiler selects functions to inline from the specified inline library If you also specify a size or function name all functions in the inline library meeting the selection criteria are selected for inline expansion at points in the source text where they are called PGI Compiler User s Guide 52 Using Function Inlining If you do not specify a function name or a size limitation for the Minline option the compiler inlines every function in the inline library that matches a function in the source text In the following example the compiler inlines the function proc from the inline library lib il and writes the executable code
284. t http www pgroup com doc pgprof pdf gt PGI Fortran Reference The Portland Group Release 15 1 January Available online at http www pgroup com doc pgifortref pdf PGI Compiler User s Guide 106 Chapter 8 ECLIPSE This document explains how to install and use the PGI plugin for Eclipse CDT C C development tool PGI Eclipse integration is only available on Linux 8 1 Install Eclipse CDT To install the Eclipse plugin for the PGI C and C compilers 1 Download the Eclipse plugin from the PGI website 2 Before you install check your CDT version 1 Go to Help gt About Eclipse 2 Click the Eclipse CDT button You might need to hover the mouse pointer on the button to see the hint 3 Select Eclipse C C Development Tools The first number in the feature version specifies which plugin version is selected 3 Go to Help gt Install New software 4 Click the Add button to add a new software repository 5 In the Add Repository dialog box Click Local Select your PGI installation directory such as opt pgi Browse inside 2015 eclipse and select the directory matching your CDT version Click OK mo p The Add Repository dialog should show the path to the local directory containing the plugin for your CDT version For example if PGI compilers are installed in opt pgi then the CDT 7 plugin is located in opt pgi os version 2015 eclipse cdt7 the CDT 8 plugin is in opt pgi os version 2015 eclip
285. task switching The following example shows a fragment that indicates how to use SYSTEM CLOCK effectively within a Fortran program unit Using SYSTEM CLOCK code fragment Areces SS injec ln dedioxebs Clocki real time camis ys cenie rocki count aec hu call system clock count clock0 lt do work gt call system clock count clock1l ig elei eloco time real t real hz Or you can use the F90 cpu_ time subroutine wee CECI EE EET call cpu time tl do work call cpu time t2 time t2 tl 3 14 Portability of Multi Threaded Programs on Linux PGI created the library 1ibnuma to handle the variations between various implementations of Linux PGI Compiler User s Guide 49 Optimizing and Parallelizing Some older versions of Linux are lacking certain features that support multi processor and multi core systems in particular the system call sched setaffinity and the numa library 1ibnuma The PGI runtime library uses these features to implement some Mconcur and mp operations These variations led to the creation of the PGI library 1ibnuma which is used on all 32 bit and 64 bit Linux systems but is not needed on Windows or OS X When a program is linked with the system 1ibnuma library the program depends on that library to run On systems without a 1ibnuma library the PGI version of 1ibnuma provides the required stubs so that the program links and executes properly If the program is linke
286. te certain default warning messages By default a warning is printed to standard error if you execute an OpenMP or auto parallelized program with NCPUS or OMP NUM THREADS set to a value larger than the number of physical processors in the system For example if you produce a parallelized executable a out and execute as follows on a system with only one processor you get a warning message setenv OMP NUM THREADS 2 a out Warning OMP NUM THREADS or NCPUS 2 greater than available cpus 1 FORTRAN STOP Setting MP WARN to NO eliminates these warning messages PGI Compiler User s Guide 141 Using Environment Variables 11 3 12 NCPUS You can use the NCPUS environment variable to set the number of processes or threads used in parallel regions The default is to use only one process or thread which is known as serial mode OMP NUM THREADS has the same functionality as NC PUS For historical reasons PGI supports the environment variable NCPUS If both OMP NUM THREADS and NCPUS are set the value of OMP NUM THREADS takes precedence Setting NCPUS to a value larger than the number of physical processors or cores in your system can cause parallel programs to run very slowly 11 3 13 NCPUS MAX You can use the NCPUS MAX environment variable to limit the maximum number of processes or threads used in a parallel program Attempts to dynamically set the number of processes or threads to a high
287. ted in the release to specify the target processor Use the acc flag to enable OpenACC directives and the ta lt target gt flag to target NVIDIA and AMD GPU You can then use the generated code on any supported system with CUDA installed that has a CUDA enabled GeForce Quadro or Tesla card or any supported system with a supported AMD Radeon GPU For more information on these flags as they relate to accelerator technology refer to Applicable Command Line Options For a complete list of supported CUDA GPUs refer to the NVIDIA website at www nvidia com object cuda learn products html PGI Compiler User s Guide 84 Using an Accelerator You can detect whether the system has CUDA properly installed and has an attached GPU by running the pgaccelinfo command which is delivered as part of the PGI Accelerator compilers software package 7 9 Installation and Licensing The PGI Accelerator compilers have a different license key than the x64 only version of the PGI Workstation PGI Server or PGI CDK products 7 9 1 Required Files If you are installing on Windows the required files are built for you The default NVIDIA Compute Capability for generated code in this release is cc2 or fermi enabling code generation for NVIDIA Fermi and Kepler GPUs You can use the ta flag to specify other compute capabilities including compute capability 1 0 through 1 3 You can also change the default to one or more of the supported compute ca
288. threads unless those threads are parallel only at the synchronous level and the memory operations are separated by an explicit barrier Otherwise if one thread updates a memory location and another reads the same location or two threads store a value to the same location the hardware does not guarantee the results While the results of running such a program might be inconsistent it is not accurate to say that the results are incorrect By definition such programs are defined as being in error While a compiler can detect some potential errors of this nature it is nonetheless possible to write an accelerator region that produces inconsistent numerical results 7 7 3 Cache Management Some current GPUs have a software managed cache some have hardware managed caches and most have hardware caches that can be used only in certain situations and are limited to read only data In low level programming models such as CUDA or OpenCL it is up to the programmer to manage these caches However in the PGI Accelerator programming model the compiler manages these caches using hints from the programmer in the form of directives 7 8 Running an Accelerator Program Running a program that has accelerator directives and was compiled and linked with the ta flag is the same as running the program compiled without the ta flag gt When running programs on NVIDIA GPUs the program looks for and dynamically loads the CUDA libraries When running programs
289. tion in file2 c if the value of the argument has changed any optimizations based on that constant value are invalid The IPA linker determines which if any of the previously created IPA optimized objects need to be regenerated and as appropriate reinvokes the compiler to regenerate them Only those objects that are stale or which have new or different IPA information are regenerated This approach saves compile time 3 9 7 Building a Program with IPA Using Make As shown earlier programs can be built with IPA using the make utility Just add the command line switch Mipa as shown here OPT Mipa fast cou aceite oM Ete EESO poe S O 0 Grobe edie o iI OREEHNII OS ELI OF ile oc pace Seu e Titled ELIe2 03 Erle Ce pace Sem e Tile2 e rleS359s tiles opgee S ORW se rile c Using the single make command invokes the compiler to generate any of the object files that are out of date then invokes pgcc to link the objects into the executable At link time pgcc calls the IPA linker to regenerate any stale or invalid IPA optimized objects make 3 9 8 Questions about IPA Question Why is the object file so large Answer An object file created with Mipa contains several additional sections One is the summary information used to drive the interprocedural analysis In addition the object file contains the compiler internal representation of the source file so the file can be recompiled at link time with i
290. uild using an arbitrary version of MPI to do this use the I L and 1 option PGI Workstation for Linux and OS X includes MPICH while on Windows PGI Workstation includes MS MPI The PGI CDK on Linux also includes MPICH and CDK users can download PGI built versions of MVAPICH2 and Open MPI from pgroup com This section describes how to use the MPI capabilities of PGI compilers and how to configure PGI compilers so these capabilities can be used with custom MPI installations The debugger and profiler are enabled to support MPI applications running locally with a limited number of processes The PGPROF Profile Guide and the PGDBG Debugger Guide describe the MPI enabled tools in detail gt PGPROF graphical MPI OpenMP multi thread performance profiler gt PGDBG graphical MPI OpenMP multi thread symbolic debugger 6 1 MPI Overview This section contains general information applicable to various MPI distributions For distribution specific information refer to the sections later in this section MPI is a set of function calls and libraries that are used to send messages between multiple processes These processes can be located on the same system or on a collection of distributed servers Unlike OpenMP the distributed nature of MPI allows it to work in almost any parallel environment PGI Compiler User s Guide 74 Using MPI 6 2 Compiling and Linking MPI Applications The PGI compilers provide an option Mmpi to make building
291. ult is to not bind processes to processors This variable is an execution time environment variable interpreted by the PGI runtime support libraries It does not affect the behavior of the PGI compilers in any way cy The MP_ BIND environment variable is not supported on all platforms setenv MP BIND y PGI Compiler User s Guide 140 Using Environment Variables 11 3 9 MP BLIST MP BLIST allows you to specifically define the thread CPU relationship This variable is only in effect when MP BINDiS yes While the MP BIND variable binds processors or threads to a physical processor MP BLIST allows you to specifically define which thread is associated with which processor The list defines the processor thread relationship order beginning with thread 0 This list overrides the default binding For example the following setting for MP BLIST maps CPUs 3 2 1 and 0 to threads 0 1 2 and 3 respectively pNSCHOml EE UST SETS 11 3 10 MP SPIN When a thread executing in a parallel region enters a barrier it spins on a semaphore You can use MP SPIN to specify the number of times it checks the semaphore before calling sched yield on Linux or MAC OS X or sleep on Windows These calls cause the thread to be re scheduled allowing other processes to run The default value is 1000000 setenv MP SPIN 200 11 3 11 MP WARN MP WARN allows you to elimina
292. um size may also be limited by the target device 7 13 Applicable Command Line Options The following command line options are applicable specifically when working with accelerators ta Use this option to enable recognition of the SACC directives in Fortran and pragma acc directives in C Use this option to specify the target host processor architecture acc Use this option to enable OpenACC directives You can use the acc suboptions to specify loop autoparallelization how the compiler reports compute regions failures to accelerate and whether to issue a warning or an error for non OpenACC accelerator directives Minfo or Minfo accel Use this option to see messages about the success or failure of the compiler in translating the accelerator region into GPU kernels The ta flag has the following accelerator related suboptions tesla Select NVIDIA accelerator target This option has a number of suboptions CC20 cc30 cc35 Generate code for compute capability 2 0 3 0 or 3 5 respectively multiple selections are valid cuda6 5 or 6 5 Specify the CUDA 6 5 version of the toolkit This is the default cuda7 0 or 7 0 Specify the CUDA 7 0 version of the toolkit PGI Compiler User s Guide 100 Using an Accelerator fastmath Use routines from the fast math library fermi Generate code for Fermi Architecture equivalent to NVIDIA compute capability 2 x no flushz Control flush to zero mode for floating point computations in the GPU
293. ut lt lt main res lt lt c lt lt endl Simple C Function c2cp_func c void cp2c func numl num2 res int numl num2 res printf func a d b d ptr c x n numl num2 res res numl num2 printf func res d n res To compile this C function and C main program use the following commands S gee C CIC cue S pacrt Co2e meim C CPo2e TUNGO Executing the resulting a out file should produce the following output main a 8 b 2 ptr c Oxbffffb94 tunes e 8 lp 2 joel JOY Ee func res 4 main res 4 13 8 4 Example C Calling C The example in C Main Program c2cp_main c Calling a C Function shows a C main program that calls the C function shown in Simple C Function c2cp func C with Extern C C Main Program c2cp_main c Calling a C Function extera WLC ce iue am el amie Id ihe e pg include lt stdio h gt main imera DAE a 8 b 2 printf main a d b d ptr c ena D Se C2 Com Fune laog EE s printf main res dWMn c PGI Compiler User s Guide 161 Inter language Calling Simple C Function c2cp_func C with Extern C include lt iostream gt externi WOU yore CAC TUNC slide men wise imi dione Dres eroxbuE ENG eL se lt lt crnmil lt ey c Wess iere el cuu resa endih res numl num2 cout lt lt func res lt lt res lt lt endl To compile this C function and C main program use the
294. ut file is a preprocessed source file an assembly language file or an unlinked object file respectively Similarly the E option does not produce a file but displays the preprocessed source file on the standard output Using any of these options the o option is valid only if you specify a single input file If no errors occur during processing you can use the files created by these options as input to a future invocation of any of the PGI compiler drivers The following table lists the stop after options and the output files that the compilers create when you use these options It also indicates the accepted input files Table 2 Option Descriptions Option Stop After Input Output E preprocessing Source files This option is not valid for preprocessed file pgcc or pgc preprocessing Source files This option is not valid for preprocessed file i pgf77 pgf95 or pgfortran Source files or preprocessed files assembly language file S C assembly Source files or preprocessed files or unlinked object file o or obj assembly language files linking Source files or preprocessed files executable file a out or exe assembly language files object files or libraries If you specify multiple input files or do not specify an object filename the compiler uses the input filenames to derive corresponding default output filenames of the following form where filename is the input filename without its exten
295. ute regions or data regions Construct a structured block identified by the programmer or implicitly defined by the language Certain actions may occur when program execution reaches the start and end of a construct such as PGI Compiler User s Guide 82 Using an Accelerator device memory allocation or data movement between the host and device memory Loops in a compute construct are targeted for execution on the accelerator The dynamic range of a construct including any code in procedures called from within the construct is called a region CUDA stands for Compute Unified Device Architecture the CUDA environment from NVIDIA is a C like programming environment used to explicitly control and program an NVIDIA GPU Data region a region defined by an OpenACC data construct or an implicit data region for a function or subroutine containing OpenACC directives Data regions typically require device memory to be allocated and data to be copied from host to device memory upon entry and data to be copied from device to host memory and device memory deallocated upon exit Data regions may contain other data regions and compute regions Device a general reference to any type of accelerator Device memory memory attached to an accelerator which is physically separate from the host memory Directive in C a pragma or in Fortran a specially formatted comment statement that is interpreted by a compiler to augment information about or specify the behavio
296. ves are specified using the C amp sentinel gt Directives which are presented in pairs must be used in pairs Clauses associated with directives have these characteristics gt The order in which clauses appear in the parallelization directives is not significant gt Commas separate clauses within the directives but commas are not allowed between the directive name and the first clause gt Clauses on directives may be repeated as needed subject to the restrictions listed in the description of each clause PGI Compiler User s Guide 61 Using OpenMP 5 4 C C Parallelization Pragmas Parallelization pragmas are pragma statements in a C or C program that are interpreted by the PGI C and C compilers when the option mp is specified on the command line The form of a parallelization pragma is pragma omp pragma_name clauses The format for pragmas include these standards gt The pragmas follow the conventions of the C and C standards Whitespace can appear before and after the Preprocessing tokens following the pragma omp are subject to macro replacement The order in which clauses appear in the parallelization pragmas is not significant v v v yvy Spaces separate clauses within the pragmas Clauses on pragmas may be repeated as needed subject to the restrictions listed in the description of each clause v For the purposes of the OpenMP pragmas a C C structured block is defined to be a statement or
297. vironment Variables on Windows In Windows when you access PGI Workstation 15 10 for example using Start ALL Programs PGI Workstation Command Shells 15 10 you have options that PGI provides for setting your environment variables either the DOS command environment or the Cygwin Bash environment When you open either of these shells available to you the default environment variables are already set and available to you You may want to use other environment variables such as the OpenMP ones This section explains how to do that Suppose that your home directory is C tmp The following examples show how you might set the temporary directory to your home directory and then verify that it is set Command prompt Once you have launched a command shell for the version of PGI that you are using 32 bit or 64 bit enter the following DOS set TMPDIR C tmp DOS echo TMPDIR C Ntmp DOS Cygwin Bash prompt From PGI Workstation 15 10 select PGI Workstation 32 bit or 64 bit and at the Cygwin Bash prompt enter the following PGIS export TMPDIR C tmp PGIS echo STMPDIR CENEO PGI 11 1 3 Setting Environment Variables on Mac OSX Let s assume that you want access to the PGI products when you log in Let s further assume that you installed the PGI compilers in opt pgi and that the license file isin opt pgi license dat For access at startup you can add the following lines to your startup file For
298. with SEE instructions cache aligned and flushz lg Instructs the compiler to include symbolic debugging information in the object module 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 Provides information about available options moemodel medium Enables medium model core generation for 64 bit targets which is useful when the data space of the program exceeds 4GB Mconcur Instructs the compiler to enable auto concurrentization of loops If specified the compiler uses multiple processors to execute loops that it determines to be parallelizable thus loop iterations are split to execute optimally in a multithreaded execution context Instructs the compiler to produce information on standard error Minie Enables function inlining Mipa fast inline Enables interprocedural analysis and optimization Also enables automatic procedure inlining j Mpfi or Mpfo Enable profile feedback driven optimizations Keeps the generated assembly files Munroll Invokes the loop unroller to unroll loops executing 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 M no vect Enables Disables the code vectorizer no exceptions Removes exception handling from user code For C declares that the
299. word This keyword instructs the compiler not to delete move significantly or combine with any other asm statement Like __asm_ the __volatile__ keyword is typically used with header files that may be included by ISO C programs gt string is one or more machine specific instructions separated with a semi colon or newline vi character The string can also contain operands specified in the output operands input operands and clobber list The instructions are inserted directly into the compiler s assembly language output for the enclosing function gt The output operands input operands and clobber list items each describe the effect of the instruction for the compiler For example asm movl 1 eax n movl eax 0 r x r y eax where r x is an output operand r y is an input operand eax is the clobber list consisting of one register eax The notation for the output and input operands is a constraint string surrounded by quotes followed by an expression and surrounded by parentheses The constraint string describes how the input and output operands are used in the asm string For example r tells the compiler that the operand is a register The tells the compiler that the operand is write only which means that a value is stored in an output operand s expression at the end of the asm statement Each operand is referenced in the asm string by a percent 96 and it
300. x64 osx86 64 in a csh set path opt pgi osx86 64 15 10 bin Spath For x64 osx86 64 in a bash sh zsh or ksh PATH opt pgi osx86 64 15 10 bin PATH export PATH PGI Compiler User s Guide 136 Using Environment Variables 11 2 PGI Related Environment Variables For easy reference the following table provides a quick listing of some OpenMP and all PGI compiler related environment variables This section provides more detailed descriptions of the environment variables specific to PGI compilers and the executables they generate For information specific to OpenMP environment variables refer to Table 14 and to the complete descriptions in OpenMP Environment Variables in the PGI Compiler s Reference Manual Table 27 PGl Related Environment Variable Summary Environment Variable Description FLEXLM_BATCH Windows only When set to 1 prevents interactive pop ups from appearing by sending all licensing errors and warnings to standard out rather than to a pop up window FORTRANOPT Allows the user to specify that the PGI Fortran compilers user VAX I O conventions GMON_OUT_PREFIX Specifies the name of the output file for programs that are compiled and linked with the pg option LD_LIBRARY_PATH Specifies a colon separated set of directories where libraries should first be searched prior to searching the standard set of directories LM_LICENSE_FILE Specifies the full path of the license file that is required for running the PGI
301. x86 64 systems is subject to the following limitations gt MPI libraries are limited to Messages of length lt 2GB and integer arguments are INTEGER 4 in FORTRAN and int in C gt nteger arguments for ScaLAPACK libraries are INTEGER 4 in FORTRAN and int in C gt Arrays passed must be lt 2GB in size PGI Compiler User s Guide 79 Using MPI 6 13 Testing and Benchmarking The Examples directory contains various benchmarks and tests Copy this directory into a local working directory by issuing the following command cp r PGI linux86 14 1 EXAMPLES MPI NAS Parallel Benchmarks The NPB2 3 subdirectory contains version 2 3 of the NAS Parallel Benchmarks in MPI Issue the following commands to run the BT benchmark on 4 nodes of your cluster 3 cd MPI NPB2 3 BT 5 make BT NPROCS 4 CLASS W 6 el o c u 5 mpirun np 4 bt W 4 jo ge oe o There are several other NAS parallel benchmarks available in this directory Similar commands are used to build and run each of them If you want to run a larger problem try building the Class A version of BT by substituting A for W in the previous commands ScaLAPACK The ScaLAPACK test times execution of the 3D PBLAS parallel BLAS on your cluster To run this test execute the following commands oe cd scalapack make 9 5 mpirun np 4 pdbla3tim PGI Compiler User s Guide 80 Chapter 7 USING AN ACCELERATOR An accelerator is a special pur
302. y compound with a single entry at the top and a single exit at the bottom In Fortran a block of executable statements with a single entry at the top and a single exit at the bottom PGI Compiler User s Guide 83 Using an Accelerator Vector operation a single operation or sequence of operations applied uniformly to each element of an array Visible device copy a copy of a variable array or subarray allocated in device memory that is visible to the program unit being compiled 7 3 System Requirements For NVIDIA GPUs To use the PGI Accelerator compiler features on NVIDIA GPUs you must install the NVIDIA drivers You may download these components from the NVIDIA website at www nvidia com cuda These are not PGI products They are licensed and supported by NVIDIA You must be using an operating system that is supported by both the current PGI release and by the CUDA software and drivers For AMD Radeon GPUs To use the PGI Accelerator compiler features on AMD Radeon GPUs you must install the AMD Catalyst drivers You may download these components from the AMD website at www amd com These are not PGI products They are licensed and supported by AMD You must be using an operating system that is supported by both the current PGI release and by the Catalyst drivers 7 4 Supported Processors and GPUs This PGI Accelerator compiler release supports all AMD64 and Intel 64 host processors Use the tp target flag as documen

PGI Compiler User`s Guide

Contents

Download Pdf Manuals

Related Search

Related Contents