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
|
DOUBLE PRECISION FUNCTION DMFLOP( OPS, TIME, INFO )
*
* -- LAPACK timing routine (version 3.0) --
* Univ. of Tennessee, Univ. of California Berkeley, NAG Ltd.,
* Courant Institute, Argonne National Lab, and Rice University
* February 29, 1992
*
* .. Scalar Arguments ..
INTEGER INFO
DOUBLE PRECISION OPS, TIME
* ..
*
* Purpose
* =======
*
* DMFLOP computes the megaflop rate given the number of operations
* and time in seconds. This is basically just a divide operation,
* but care is taken not to divide by zero.
*
* Arguments
* =========
*
* OPS - DOUBLE PRECISION
* On entry, OPS is the number of floating point operations
* performed by the timed routine.
*
* TIME - DOUBLE PRECISION
* On entry, TIME is the total time in seconds used by the
* timed routine.
*
* INFO - INTEGER
* On entry, INFO specifies the return code from the timed
* routine. If INFO is not 0, then DMFLOP returns a negative
* value, indicating an error.
*
* =====================================================================
*
* .. Parameters ..
DOUBLE PRECISION ZERO
PARAMETER ( ZERO = 0.0D+0 )
* ..
* .. Intrinsic Functions ..
INTRINSIC ABS, DBLE
* ..
* .. Executable Statements ..
*
IF( TIME.LE.ZERO ) THEN
DMFLOP = ZERO
ELSE
DMFLOP = OPS / ( 1.0D6*TIME )
END IF
IF( INFO.NE.0 )
$ DMFLOP = -ABS( DBLE( INFO ) )
RETURN
*
* End of DMFLOP
*
END
|
