File: MPI_Reduce.3

package info (click to toggle)
mpich 1.2.5.3-5
  • links: PTS
  • area: main
  • in suites: sarge
  • size: 53,880 kB
  • ctags: 44,904
  • sloc: ansic: 260,029; cpp: 91,556; sh: 42,421; java: 33,448; makefile: 8,959; fortran: 4,601; tcl: 3,548; f90: 3,517; perl: 2,251; asm: 999; csh: 856
file content (161 lines) | stat: -rw-r--r-- 3,421 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
.TH MPI_Reduce 3 "2/19/2002" " " "MPI"
.SH NAME
MPI_Reduce \-  Reduces values on all processes to a single value 
.SH SYNOPSIS
.nf
#include "mpi.h"
int MPI_Reduce ( void *sendbuf, void *recvbuf, int count, 
                MPI_Datatype datatype, MPI_Op op, int root, MPI_Comm comm )
.fi
.SH INPUT PARAMETERS
.PD 0
.TP
.B sendbuf 
- address of send buffer (choice) 
.PD 1
.PD 0
.TP
.B count 
- number of elements in send buffer (integer) 
.PD 1
.PD 0
.TP
.B datatype 
- data type of elements of send buffer (handle) 
.PD 1
.PD 0
.TP
.B op 
- reduce operation (handle) 
.PD 1
.PD 0
.TP
.B root 
- rank of root process (integer) 
.PD 1
.PD 0
.TP
.B comm 
- communicator (handle) 
.PD 1

.SH OUTPUT PARAMETER
.PD 0
.TP
.B recvbuf 
- address of receive buffer (choice, 
significant only at 
.I root
) 
.PD 1

.SH ALGORITHM
This implementation currently uses a simple tree algorithm.

.SH NOTES FOR FORTRAN
All MPI routines in Fortran (except for 
.I MPI_WTIME
and 
.I MPI_WTICK
) have
an additional argument 
.I ierr
at the end of the argument list.  
.I ierr
is an integer and has the same meaning as the return value of the routine
in C.  In Fortran, MPI routines are subroutines, and are invoked with the
.I call
statement.

All MPI objects (e.g., 
.I MPI_Datatype
, 
.I MPI_Comm
) are of type 
.I INTEGER
in Fortran.

.SH NOTES ON COLLECTIVE OPERATIONS

The reduction functions (
.I MPI_Op
) do not return an error value.  As a result,
if the functions detect an error, all they can do is either call 
.I MPI_Abort
or silently skip the problem.  Thus, if you change the error handler from
.I MPI_ERRORS_ARE_FATAL
to something else, for example, 
.I MPI_ERRORS_RETURN
,
then no error may be indicated.

The reason for this is the performance problems in ensuring that
all collective routines return the same error value.

.SH ERRORS

All MPI routines (except 
.I MPI_Wtime
and 
.I MPI_Wtick
) return an error value;
C routines as the value of the function and Fortran routines in the last
argument.  Before the value is returned, the current MPI error handler is
called.  By default, this error handler aborts the MPI job.  The error handler
may be changed with 
.I MPI_Errhandler_set
; the predefined error handler
.I MPI_ERRORS_RETURN
may be used to cause error values to be returned.
Note that MPI does 
.B not
guarentee that an MPI program can continue past
an error.

.PD 0
.TP
.B MPI_SUCCESS 
- No error; MPI routine completed successfully.
.PD 1
.PD 0
.TP
.B MPI_ERR_COMM 
- Invalid communicator.  A common error is to use a null
communicator in a call (not even allowed in 
.I MPI_Comm_rank
).
.PD 1
.PD 0
.TP
.B MPI_ERR_COUNT 
- Invalid count argument.  Count arguments must be 
non-negative; a count of zero is often valid.
.PD 1
.PD 0
.TP
.B MPI_ERR_TYPE 
- Invalid datatype argument.  May be an uncommitted 
MPI_Datatype (see 
.I MPI_Type_commit
).
.PD 1
.PD 0
.TP
.B MPI_ERR_BUFFER 
- Invalid buffer pointer.  Usually a null buffer where
one is not valid.
.PD 1
.PD 0
.TP
.B MPI_ERR_BUFFER 
- This error class is associcated with an error code that
indicates that two buffer arguments are 
.B aliased
; that is, the 
describe overlapping storage (often the exact same storage).  This
is prohibited in MPI (because it is prohibited by the Fortran 
standard, and rather than have a separate case for C and Fortran, the
MPI Forum adopted the more restrictive requirements of Fortran).
.PD 1
.SH LOCATION
reduce.c