The January 2013 Update for XL Fortran Advanced Edition for Blue Gene/P, V11.1 (VRMF: 22.214.171.124) is now available at the following links and on Fix Central . January 2013 Update for XL Fortran Advanced Edition for Blue Gene/P, V11.1 Can't find the updates for your compilers? Check out our Fortran compilers latest updates page here .
2012 Update for XL Fortran for Linux, V13.1 (VRMF: 126.96.36.199) is now
available at the following links and on Fix Central. October 2012 Update for XL Fortran for Linux, V13.1 October 2012 Runtime for XL Fortran for Linux, V13.1 Can't find the updates for your compilers? Check out our Fortran compilers latest updates page here .
FREE newsletter - get IBM Software news, special #PowerSystems features - subscribe: http://bit.ly/rs1rZn #POWER7 We just published a special Power Systems Edition of the newsletter, which you can see here - http://www-148.ibm.com/tela/webmail/NlDynamicPage/9540/32113/cosmetic?web_page_id=546427&pre=0 - and we're also including Power Systems-related content in each monthly issue.
IBM marks its centennial today. As part of our 100 days of celebration, IBM highlighted 100 of its best achievements in the Icons of Progress site. FORTRAN is featured as the March 3rd icon. Since its release in 1957, the Fortran language has continued to evolve, with the latest versions adding object-orientation, parameterized derived types, and distributed programming. No one knows what innovations the next 100 years will bring, but who knows, Fortran's innovation, longevity, and high performance might be highlighted in the next... [More]
The XL Fortran for Linux, 15.1 and XL Fortran for AIX, 15.1 introduced the -qpath option , which is functionally similar to the -t and -B combination but more convenient. You can use -qpath to specify substitute path names for XL Fortran components, such as the assembler, C preprocessor, and linker. The -qpath option is more feasible than the -t and -B combination, as you can use the -qpath option to specify more than one component, each with a different path. This can bring convenience as you needn’t copy all specified components to the... [More]
In IBM XL Fortran for Linux, V15.1 and IBM XL Fortran for AIX, V15.1, a new Fortran 2008 intrinsic function has been added to let you inquire the location of the first or last matching of the wanted value in an array along the specified dimension. Function FINDLOC returns the subscript of the element of wanted value using positive integers. The complete format is FINDLOC(ARRAY, VALUE, DIM, MASK, KIND, BACK) where DIM, MASK, KIND and BACK are optional. The result indicates the subscript of the location of the masked element of ARRAY whose value... [More]
The XL Fortran for BlueGene/Q Fix Pack 7 (January 2014 Update) for 14.1 is now available at the following links and on Fix Central: XL Fortran for BlueGene/Q Fix Pack 7 (February 2014 Update) for 1 4 .1 Fix list for XL Fortran for BlueGene/Q Can't find the updates for your compiler? Check out our Fortran compiler latest update page here
The Fortran 2008 standard relaxes some rules for internal procedures. Procedure pointers can now point to internal procedures. In addition, you can use internal procedures and pointers to internal procedures as actual arguments. Because an internal procedure has access to the variables of its host procedure, procedure pointers to an internal procedure can only be used before the host procedure completes its execution. Here is an example of a procedure pointer to an internal procedure: module m implicit none... [More]
The history of the Fortran language can be traced straight to IBM. Fortran was originally proposed by John Backus in 1953, and the initial Fortran compiler, for IBM mainframe computers, was delivered in 1957. Fast-forward to today, and Fortran is widely used on the field of High-Performance Computing. It has become an ANSI standard, and has been implemented on all major platforms on this space. What makes Fortran successful on this space is its ability to provide developers with very good performance for numerical computation withour requiring... [More]
The XL Fortran for AIX Fix Pack 19 (August 2014 PTF) for 13.1 is now available at the following links and on Fix Central : XL F ort r an f or A IX F ix P ack 19 (August 20 1 4 P T F) f or V 13 . 1 F i x l i s t f o r X L F o r t r a n f o r A I X XL F ort r an R unt i me f or A IX F ix P ack 19 (August 20 1 4 U p dat e ) f o r 1 3 . 1 F i x l i s t f o... [More]
The Fortran 2008 standard has the following exception to argument presence: A dummy argument or an entity that is host associated with a dummy argument is not present if the dummy argument does not have the ALLOCATABLE or POINTER attribute, and corresponds to an actual argument that has the ALLOCATABLE attribute and is not allocated, or has the POINTER attribute and is disassociated. In XL Fortran, V14.1, the -qxlf2008=checkpresence option is added so users can detect actual argument presence correctly. The default of the option depends on your... [More]
The April 2012 Update for XL Fortran Advanced Edition for Blue Gene/P, V11.1 is now available at the following link and on Fix Central: April 2012 Update for XL Fortran Advanced Edition for Blue Gene/P, V11.1 Can't find the updates for your compilers? Check out our C/C++ compilers latest updates page here .
Authors: Bardia Mahjour, Christine Li The STOP statement has been enhanced in XL Fortran v14.1 to comply with the F2008 standard. The stop code can now be specified using any scalar integer constant expression or a scalar character constant expression. For instance, you can now successfully compile and run the following short program: INTEGER, PARAMETER :: rc = 2 STOP rc END The same program would have been invalid in previous standards, since the error code is... [More]
If your code contains a large number of loops, you can use option -qassert=minitercnt= n1 and -qassert=maxitercnt= n2 provided in XL Fortran V14.1 to specify the expected minimum and maximum iteration counts of the loops in the program. The minitercnt and maxitercnt values are not required to be accurate since the values only affect performance, never correctness. The values will be provided to the optimizer. n1 and n2 must be positive integers. If you specify the two suboptions at the same time, the values must follow the rule... [More]
The POPCNT intrinsic function is new for XL Fortran V14.1. It counts the number of set bits in a data object. The data object can be of type integer, byte, logical, or real with REAL(16) excluded. The result value is the number of bits set to 1 in the sequence of bits of the given data object. For example, the binary representation for decimal number 4 is 0100, so the value of POPCNT(4) is 1. Similarly, the binary representation for decimal number 3 is 0011, so the value of POPCNT(3) is 2. For more information, see POPCNT . Note: This blog is... [More]