Hello r-devel,
When compiling Fortran code containing BLAS functions and calling it using
dyn.load, I am getting incorrect results. A small example with which I can
reproduce the problem is below.
I am running on OSX Mavericks (upgraded R, Xcode, etc per instructions on
this list), but I do not think it is related since the check on my
"blupsurv" package on r-forge seems to be showing similar behavior, and the
problem predates Mavericks. The package used to build and check without
errors when I originally put it on r-forge. Would appreciate suggestions.
Thanks,
Emmanuel
Example: this program creates two double vectors, takes the dot product
with ddot, and prints the result:
test.f:
subroutine testddot(n,x,y)
integer i,n
double precision x,y,d1(n),d2(n)
do 100 i=1,n
d1(i)=x
d2(i)=y
100 continue
print *,ddot(n,d1,1,d2,1)
end
Compiling
R CMD SHLIB test.f
Running:
> dyn.load("test.so")
> test<-.Fortran("testddot",as.integer(3),as.double(1),as.double(1.2))
-1.0737418E+08
The result should be 3.6, not -1.07e8. Most other inputs give incorrect
results too.
Here is my session info. I upgraded to the latest devel to see if it would
fix the problem.
> sessionInfo()
R Under development (unstable) (2014-01-10 r64727)
Platform: x86_64-apple-darwin10.8.0 (64-bit)
locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
