1 ! { dg-do run { xfail *-*-mingw* spu-*-* } }
6 ! For mingw targets this test is disabled as the MS implementation
7 ! of BESSEL_YN(n,x) has different results. It returns NAN rather than
8 ! -INF for "x=0.0" and all "n".
10 ! XFAILed for SPU targets since we don't have an accurate library
11 ! implementation of the single-precision Bessel functions.
13 ! Run-time tests for transformations BESSEL_YN
16 real,parameter :: values(*) = [0.0, 0.5, 1.0, 0.9, 1.8,2.0,3.0,4.0,4.25,8.0,34.53, 475.78]
17 real,parameter :: myeps(size(values
)) = epsilon(0.0) &
18 * [2, 3, 4, 5, 8, 2, 12, 6, 7, 6, 31, 168 ]
19 ! The following is sufficient for me - the values above are a bit
21 ! * [0, 0, 0, 3, 3, 0, 9, 0, 2, 1, 22, 130 ]
22 integer,parameter :: nit(size(values
)) = &
23 [100, 100, 100, 25, 15, 100, 10, 31, 7, 100, 7, 25 ]
24 integer, parameter :: Nmax
= 100
25 real :: rec(0:Nmax
), lib(0:Nmax
)
28 do i
= 1, ubound(values
,dim
=1)
29 call compare(values(i
), myeps(i
), nit(i
), 6*epsilon(0.0))
34 subroutine compare(X
, myeps
, nit
, myeps2
)
39 rec
= BESSEL_YN(0, Nmax
, X
)
40 lib
= [ (BESSEL_YN(i
, X
), i
=0,Nmax
) ]
42 !print *, 'YN for X = ', X, ' -- Epsilon = ',epsilon(x)
44 ! print '(i2,2e17.9,e12.2,f14.10,2l3)', i, rec(i), lib(i), &
45 ! rec(i)-lib(i), ((rec(i)-lib(i))/rec(i))/epsilon(x), &
46 ! i > nit .or. rec(i) == lib(i) &
47 ! .or. abs((rec(i)-lib(i))/rec(i)) < myeps2, &
48 ! rec(i) == lib(i) .or. abs((rec(i)-lib(i))/rec(i)) < myeps
49 if (.not
. (i
> nit
.or
. rec(i
) == lib(i
) &
50 .or
. abs((rec(i
)-lib(i
))/rec(i
)) < myeps2
)) &
52 if (.not
. (rec(i
) == lib(i
) .or
. abs((rec(i
)-lib(i
))/rec(i
)) < myeps
)) &