A From:	SMTP%"RELAY-INFO-VAX@CRVAX.SRI.COM"  7-MAY-1993 18:19:10.50  To:	EVERHART CC:	 Subj:	Re: fortran V6.0   X-Newsgroups: comp.os.vms . From: lionel@quark.enet.dec.com (Steve Lionel) Subject: Re: fortran V6.0 5 Message-ID: <1993May7.020641.6694@dbased.nuo.dec.com> 4 Sender: news@dbased.nuo.dec.com (USENET News System)+ Organization: Digital Equipment Corporation " Date: Fri, 7 May 1993 02:18:07 GMT
 Lines: 274 To: Info-VAX@KL.SRI.COM  X-Gateway-Source-Info: USENET     } In article <1993May6.202632.18720@butch.lmsc.lockheed.com>, burkhardt@force.ssd.lmsc.lockheed.com (Spike Burkhardt) writes... K >Hello fortran masters!  I am thinking about upgrading fortran from V5.5 to I >6.0.  How stable is the new version?  Are there any installation gotchas  >or problems with it?  >   E How about the straight dope from the source?  (Or the straight source  from the dope - your choice...)   A The most significant "gotcha" you may run into is if you have any D Fortran programs around that were linked on VMS V3 or earlier.  ThisC is due to an incompatibility between the VMSRTL.EXE provided by VMS : and the new FORRTL and MTHRTL provided on the Fortran kit.  G The other important thing to be aware of is that the MTHRTL provided on G the Fortran kit has had its GSMATCH ident bumped as the transfer vector H had to be extended to include the large number of new entry points.  TheF side effect of this is that images linked against the new MTHRTL won'tF run on systems that don't have Fortran V6.0 installed.  There are waysD around this, detailed in the Release Notes, but it is important thatA all your system users be aware of it.  The installation preserves 2 the VMS-supplied RTL images so you won't be stuck.  G As for installation itself, some users have failed to read the warnings E in the Installation Guide and displayed by the installation procedure C itself regarding installing the kit on a VMScluster.  If you do, be F *SURE* to run the SYS$UPDATE:FORTRAN$POST_INSTALL.COM procedure on allD other cluster nodes after installation.  The IG has complete details on this.  J Now as for stability...  we ran a six-month field test at 10 high-activityC sites, and have our own 9000-program test system.  But no amount of E testing prepares one for what users really do...   There is currently E the first update kit, called FORTECO01060, available from the support A centers.  I highly recommend that you install this instead of the B SSB version.  The ECO01 kit is edit level V6.0-14 and it fixes theD various RTL-related problems.  This was made available at the end ofB March.  I'm about to release a second update, ECO02, which will beH edit level V6.0-31.  This may sound like a lot of fixes in a short time,G but almost all of these are minor annoyances (and a few are relaxations G of restrictions and improved error reporting).  I'll attach the list of % the fixes to the end of this message.   H V6.0 has a lot of new features and contains some significant performanceF optimizations.  One feature you may like is detection of uninitializedB variables; I was astonished when I turned on this feature, ran ourC test system which contains many industry standard applications, and C found dozens of uninitialized variables.  How many hidden bugs will I this feature find in your code?  Another new feature is recursion, though I you'll want to get ECO02 to make the most of it; this should be available D by the end of next week.  Also there's the ability to read and writeB unformatted data in non-VAX formats such as IEEE floating (big and( little-endian), Cray and IBM System-370.  J Pull the release notes from the kit (they're available in text, PostScriptE and Bookreader form) and read them to get a good feel for what's new. C The March ConOLD CD-ROM also has the new documentation - look under @ "DEC Fortran for OpenVMS VAX Systems", our new name, rather than VAX FORTRAN.  K I'm always interested in comments and questions on the VAX Fortran product; J I'll answer them in the newsgroup or by mail, as you prefer.   And lastly,N I'll be at DECUS in Atlanta giving a presentation on "DEC Fortran Directions",D including full details on the new V6.0 for OpenVMS VAX.  Stop by and
 say hello.  E     Table_A-1_DEC_Fortran_Maintenance_Changes________________________   E     Version_____Description__________________________________________   B     V6.0-1      DEC Fortran V6.0 Release; contains all corrections>                 to VAX FORTRAN through version V5.9-175 and to8                 VAX FORTRAN-HPO through version V1.3-252  ;     V6.0-2      The compiler now detects as an extension to A                 FORTRAN-77 the use of a non-integer expression as :                 a logical unit number in an I/O statement.  D     V6.0-3      A number of problems related to the "non-native data:                 in I/O" feature were corrected, including:  E                 o  The FOR$CONVERTnnn logical names were not properly                     recognized.  =                 o  Big-endian IEEE and all big-endian integer 5                    conversion did not work correctly.   E                 o  Conversion in I/O statements with implied DO-loops 1                    did not always work correctly.   @                 o  Conversion of COMPLEX variables did not work.  A                 o  Applications which executed an ENDFILE or FIND C                    statement on a unit which was not open could get 2                    an INVARGFOR error at run-time.  <     V6.0-4      The compiler no longer goes into a loop if a@                 statement function has two or more formals whichA                 are declared as RECORDs using the same STRUCTURE.   D     V6.0-5      The compiler now issues the new diagnostic EQVSAVCOM?                 if EQUIVALENCE is used to attempt to put a SAVE %                 variable into COMMON.   C     V6.0-6      The DEC Fortran kit now provides an updated copy of A                 the VMSRTL.EXE shareable image, which is required A                 by images linked against VAX/VMS versions earlier A                 than V4.0. Without the updated VMSRTL.EXE, images D                 which use it may fail unpredictably due to registersC                 not being properly saved across calls to Fortran or B                 Math Run-Time Library routines. The new VMSRTL.EXED                 specifies that all registers are to be saved so that@                 changes to register usage in routines now in theC                 BASRTL, COBRTL, FORRTL, LIBRTL and MTHRTL shareable C                 images will not break images linked against VMSRTL. A                 Images linked on VAX/VMS V4.0 or later do not use ,                 VMSRTL and are not affected.  E                 The updated VMSRTL.EXE is not provided on OpenVMS VAX @                 Version V6.0 or later as it is not needed there.  A     V6.0-7      A problem in which the compiler would fail with a 9                 bugcheck in the PACK phase was corrected.   2     V6.0-8      The Digital-provided INCLUDE filesE                 SYS$LIBRARY:FORDEF.FOR, SYS$LIBRARY:FORIOSDEF.FOR and D                 the $FORMSG module of SYS$LIBRARY:FORSYSDEF.TLB haveE                 been updated to include the new FOR$_FLOCONFAI status ?                 value and associated IOSTAT number definitions.   C     V6.0-9      The compiler no longer issues the CRX error message B                 BADRGTJUST for a CDD non-text field that specifiesC                 right-justification. The justification attribute is '                 ignored by DEC Fortran.   @     V6.0-10     The correct line number is now displayed for the=                 EXTCHASOU and TOOMANCON compiler diagnostics.   ?     V6.0-11     The compiler no longer gives spurious USEUNIVAR B                 warnings for references to COMPLEX variables which@                 were data-initialized using character constants,C                 and now only gives one warning per COMPLEX variable =                 instead of separate warnings for the real and                   imaginary parts.  E     V6.0-12     The compiler now properly displays the attempted file A                 specification when it gives an INCOPEFAI error on 3                 failing to open an INCLUDE library.r  E     V6.0-13     The compiler now allows record variables to appear inc6                 AUTOMATIC, SAVE and STATIC statements.  =     V6.0-14     The compiler was changed to emit the quadwords@                 stack alignment routine entry code sequence (seeD                 Section 1.5.9) only when the /VECTOR compile command'                 qualifier is specified.<  @     V6.0-15     The compiler no longer generates bad code when aB                 LOGICAL*2 or LOGICAL*4 array element is used in an(                 arithmetic-IF statement.  @     V6.0-16     A nested STRUCTURE with a data-initialized field@                 in in a program unit compiled with /RECURSIVE noC                 longer causes the linker to issue an UDEFPSC error.dA                 In addition, the compiler now issues the warnings >                 AUTDATINI and AUTSAVALL if a variable declared>                 AUTOMATIC was data-initialized or appears in aC                 program unit which specifies that all variables arer                 to be SAVEd.  D     V6.0-17     The compiler no longer generates incorrect code whenA                 the only store to a formal argument occurs by itsvD                 reference in an IOSTAT specification. In some cases,C                 the modified value was not written back through the                  argument list.  A     V6.0-18     The compiler no longer causes the linker to issuetA                 a UDEFPSC error for program units which referenceeB                 a local variable in a READ or ACCEPT statement but?                 don't pass that variable as an actual argument.f  =     V6.0-19     The compiler now properly passes a copy of anoC                 actual argument when that argument is an expressiongB                 consisting of a reference to an intrinsic functionC                 that returns its argument unchanged, such as INT ofe$                 an INTEGER argument.  C     V6.0-20     The Run-Time Library no longer reports an INVARGFORoD                 error when the PDP-11 compatibility routines ERRSET,>                 ERRSNS or ERRTST are used for error number 95,                 FLOCONFAI.  C     V6.0-21     The compiler no longer fails with a bugcheck in theoC                 CSE phase when /VECTOR is specified and the programs<                 contains certain uses of COMPLEX intrinsics.  E     V6.0-22     The compiler now allows recursive function referenceseE                 as well as recursive references to ENTRY points whichi@                 are defined after the function reference or CALLD                 statement. The new error message ROUREFREC, "Routine?                 referenced recursively; /RECURSIVE required" isiC                 given if the compiler detects a recursive reference B                 and /RECURSIVE was not specified. The compiler nowD                 produces correct cross-reference and SCA information2                 for recursive function references.  B     V6.0-23     The compiler no longer generates incorrect code inB                 a situation whose symptom was that register R0 wasC                 believed to still hold a value that had been loadedr=                 prior to a subroutine call which modified R0.   <     V6.0-24     The compiler no longer issues the EXCCHASOU,=                 "Extra characters in source" warning, enablede?                 by /WARNINGS=TRUNCATED_SOURCE, for source linese<                 containing whole-line comments or only blankD                 characters, even if those lines exceed the statement@                 field width. The compiler issues the new warning?                 message FUNVALUND, "Function value undefined atrA                 end of routine" if a function return value is not.C                 defined at one or more exit points of the function.nA                 The standards extension diagnostic EXT_UNDFUNC iseB                 no longer issued, being replaced by FUNVALUND. The?                 compiler now more accurately displays statement C                 line numbers associated with error messages and theo-                 /SHOW=LOOP_STRUCTURE display._  C     V6.0-25     The compiler no longer generates incorrect code for_@                 certain unusual cases involving a routine with aA                 character formal argument whose address was takenn@                 with %LOC and another argument was used as a KEYC                 value in an indexed READ statement. The symptom was D                 that one argument's address was loaded into registerE                 R0 or R1 for later use, but the copy of the characteru@                 argument's descriptor overwrote those registers.  ?     V6.0-26     The compiler's limit for the amount of internale=                 representation allowed per statement has been <                 doubled, making it much less likely than the@                 STATOOCOM, "Statement too complex" error will be                 generated.  C     V6.0-27     The compiler no longer fails with a bugcheck during D                 LEXSYN when a type declaration statement has an end-A                 of line comment, is followed by a blank line, andn,                 /ANALYSIS_DATA is specified.  D     V6.0-28     The compiler now generates correct code for the caseD                 where a subprogram has one or more ENTRY statements,E                 a non-CHARACTER dummy argument appears in two or morehA                 of the dummy argument lists, and that argument isDE                 passed as an actual argument to another routine using                  %DESCR.C  C     V6.0-29     The compiler no longer fails with a bugcheck during @                 PACK in certain cases where the double-precision?                 MOD intrinsic is used in conjunction with arrayM                 references.c  /     V6.0-30     The compiler now recognizes thea?                 /OPTIMIZE=[NO]COMPLEX_EXPANSION compile commandtE                 qualifier which, if specified as NOCOMPLEX_EXPANSION, =                 disables the complex expansion optimization. r  <     V6.0-31     The compiler no longer fails with a bugcheckA                 during LOCAL when the %LOC operator is used on an E                 external procedure which has been typed as COMPLEX orrE                 COMPLEX*16.                                          O  9 Steve Lionel                    lionel@quark.enet.dec.coml SDT Languages Groupn Digital Equipment Corporation  110 Spit Brook Roadn Nashua, NH 03062  