AMBER Archive (2008)

Subject: Re: AMBER: Compiling AMBER 10 on SGI MIPS and Sun SPARC

From: Atro Tossavainen (atro.tossavainen+amber_at_helsinki.fi)
Date: Sun Dec 14 2008 - 14:11:53 CST

> > "dpinit.F90", line 240: error: missing '.
>
> Try removing the initial "&" from line 240. That should not be there, but
> apparently most compilers don't complain.

Oddly enough the leading ampersand on a continuation line is not a
problem on lines 139-140, 146-147, 153-154, 161-162 etc. of the same
file.

Removing the leading ampersand from line 240 changes nothing, the
compiler still barfs. Commenting out the entire line

  & ' Will not use the radii specified in block.F'

causes the error message to remain otherwise the same but indicate the
next line instead.

Printing out something that doesn't end in & !! on line 239 makes the
problem go away.

My FORTRAN-fu is regrettably far too weak to analyse this one any further.

I have another Sun issue now that the first one was circumvented:

f90 -c -O1 -free -M../../sander -I../src/include -DHAS_LAPACK -DQMMM -o ../obj/tsqna.o ../src/ts/tsqna.F90

        & '('' Lambda too large& ! Scaled Down'', &
                                   ^
"../src/ts/tsqna.F90", Line = 453, Column = 36: ERROR: Unknown edit descriptor "M" has been detected.
                                                ^
"../src/ts/tsqna.F90", Line = 453, Column = 49: ERROR: Unknown edit descriptor "v" has been detected.
                                                                 ^
"../src/ts/tsqna.F90", Line = 453, Column = 66: ERROR: Missing right parenthesis or unexpected end of FORMAT specifier.

f90comp: 857 SOURCE LINES
f90comp: 3 ERRORS, 0 WARNINGS, 0 OTHER MESSAGES, 0 ANSI
*** Error code 1

Regarding the SGI bit:

> I can't see any problem with the code here, and this syntax is used in many
> places in the "ncsu" modules. Did you successfully compile any of the ncsu-
> files?

ncsu-umbrella.f was compiled successfully. ncsu-abmd-hooks is the next
one and nothing else had been tried so far.

> You could try the following experiment: UMBRELLA_MAX_NEXTENTS is defined this
> way:

The experiment you suggested allows ncsu-abmd-hooks.f to compile. None
of the other ncsu-* files present any problems.

> It could be that the SGI f90 code is not correctly handling these sorts of
> indirections.

It could be. The likelihood of any further patches to the MIPSpro
compilers seems a little slim, unfortunately. 

-- 
Atro Tossavainen (Mr.)               / The Institute of Biotechnology at
Systems Analyst, Techno-Amish &     / the University of Helsinki, Finland,
+358-9-19158939  UNIX Dinosaur     / employs me, but my opinions are my own.
< URL : http : / / www . helsinki . fi / %7E atossava / > NO FILE ATTACHMENTS
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu