/[MITgcm]/MITgcm/verification/cpl_aim+ocn/code_atm/CPP_EEOPTIONS.h
ViewVC logotype

Contents of /MITgcm/verification/cpl_aim+ocn/code_atm/CPP_EEOPTIONS.h

Parent Directory Parent Directory | Revision Log Revision Log | View Revision Graph Revision Graph


Revision 1.4 - (show annotations) (download)
Fri Jan 23 19:19:31 2004 UTC (20 years, 4 months ago) by adcroft
Branch: MAIN
CVS Tags: checkpoint52l_pre, checkpoint52j_post, checkpoint52l_post, checkpoint52k_post, checkpoint52j_pre
Changes since 1.3: +7 -4 lines
File MIME type: text/plain
Updating so USE_W2 is undef'd in eesupp/inc and teh remaining CPP_EEOPTIONS.h
in the cubed sphere expts are similar in structure to that in eesupp/inc .
 - this is primarily to get JMC out of my office!

1 C $Header: /u/gcmpack/MITgcm/verification/cpl_aim+ocn/code_atm/CPP_EEOPTIONS.h,v 1.3 2004/01/22 14:45:01 adcroft Exp $
2 C $Name: $
3
4 C
5 C /==========================================================\
6 C | CPP_EEOPTIONS.h |
7 C |==========================================================|
8 C | C preprocessor "execution environment" supporting |
9 C | flags. Use this file to set flags controlling the |
10 C | execution environment in which a model runs - as opposed |
11 C | to the dynamical problem the model solves. |
12 C | Note: Many options are implemented with both compile time|
13 C | and run-time switches. This allows options to be |
14 C | removed altogether, made optional at run-time or |
15 C | to be permanently enabled. This convention helps |
16 C | with the data-dependence analysis performed by the |
17 C | adjoint model compiler. This data dependency |
18 C | analysis can be upset by runtime switches that it |
19 C | is unable to recoginise as being fixed for the |
20 C | duration of an integration. |
21 C | A reasonable way to use these flags is to |
22 C | set all options as selectable at runtime but then |
23 C | once an experimental configuration has been |
24 C | identified, rebuild the code with the appropriate |
25 C | options set at compile time. |
26 C \==========================================================/
27
28 #ifndef _CPP_EEOPTIONS_H_
29 #define _CPP_EEOPTIONS_H_
30
31 C In general the following convention applies:
32 C ALLOW - indicates an feature will be included but it may
33 C CAN have a run-time flag to allow it to be switched
34 C on and off.
35 C If ALLOW or CAN directives are "undef'd" this generally
36 C means that the feature will not be available i.e. it
37 C will not be included in the compiled code and so no
38 C run-time option to use the feature will be available.
39 C
40 C ALWAYS - indicates the choice will be fixed at compile time
41 C so no run-time option will be present
42
43 C Flag used to indicate whether Fortran formatted write
44 C and read are threadsafe. On SGI the routines can be thread
45 C safe, on Sun it is not possible - if you are unsure then
46 C undef this option.
47 #undef FMTFTN_IO_THREADSAFE
48
49 C-- Control MPI based parallel processing
50 CXXX We no longer select the use of MPI via this file (CPP_EEOPTIONS.h)
51 CXXX To use MPI, use an appropriate genmake2 options file or use
52 CXXX genmake2 -mpi .
53 CXXX #undef ALLOW_USE_MPI
54 CXXX #undef ALWAYS_USE_MPI
55
56 C-- Control use of communication with other component:
57 C allow to import and export from/to Coupler interface.
58 #define COMPONENT_MODULE
59
60 C-- Control use of communication that might overlap computation.
61 C Under MPI selects/deselects "non-blocking" sends and receives.
62 #define ALLOW_ASYNC_COMMUNICATION
63 #undef ALLOW_ASYNC_COMMUNICATION
64 #undef ALWAYS_USE_ASYNC_COMMUNICATION
65 C-- Control use of communication that is atomic to computation.
66 C Under MPI selects/deselects "blocking" sends and receives.
67 #define ALLOW_SYNC_COMMUNICATION
68 #undef ALWAYS_USE_SYNC_COMMUNICATION
69
70 C-- Control use of JAM routines for Artic network
71 C These invoke optimized versions of "exchange" and "sum" that
72 C utilize the programmable aspect of Artic cards.
73 #undef LETS_MAKE_JAM
74 #undef JAM_WITH_TWO_PROCS_PER_NODE
75
76 C-- Control storage of floating point operands
77 C On many systems it improves performance only to use
78 C 8-byte precision for time stepped variables.
79 C Constant in time terms ( geometric factors etc.. )
80 C can use 4-byte precision, reducing memory utilisation and
81 C boosting performance because of a smaller working
82 C set size. However, on vector CRAY systems this degrades
83 C performance.
84 #define REAL4_IS_SLOW
85
86 C-- Control use of "double" precision constants.
87 C Use D0 where it means REAL*8 but not where it means REAL*16
88 #define D0 d0
89
90 C-- Control XY periodicity in processor to grid mappings
91 C Note: Model code does not need to know whether a domain is
92 C periodic because it has overlap regions for every box.
93 C Model assume that these values have been
94 C filled in some way.
95 #undef ALWAYS_PREVENT_X_PERIODICITY
96 #undef ALWAYS_PREVENT_Y_PERIODICITY
97 #define CAN_PREVENT_X_PERIODICITY
98 #define CAN_PREVENT_Y_PERIODICITY
99
100 #endif /* _CPP_EEOPTIONS_H_ */
101
102 #include "CPP_EEMACROS.h"

  ViewVC Help
Powered by ViewVC 1.1.22