Re: [PATCH, gfortran testsuite]: Do not load denormals in gfortran.fortran_torture/intrinsic_nearest.f90

[Tim Prince <n8tm at aol dot com>]

N.M. Maclaren wrote:
> On Aug 23 2009, IainS wrote:
> > On 23 Aug 2009, at 01:26, Tim Prince wrote:
> >
> > > I suppose the least controversial part of this would be to supply  
> > > the IEEE_ARITHMETIC subroutines IEEE_SET_UNDERFLOW_MODE and  
> > > IEEE_GET_UNDERFLOW_MODE in libgfortran, as it looks straightforward  
> > > in current gcc/gfortran with iso_c_interop for the i386/x64  
> > > platforms.  I don't have any other targets available to test, and  it 
> > > clearly requires target-dependent code.
> >
> > I'm willing to stick my neck out to try and replicate what you do on  
> > i386/x64 on PowerPC (at least G4 and G5 which I have access too).
> > ( but I might need help :-) )
>
> If you tell me precisely what you are trying to implement/support, I am
> willing to try to make time to write a few test cases against the Fortran
> standard and/or sanity.  Please note: the FORTRAN standard - I know that
> aspect of IEEE 754 and C99 as well as anyone, but the latter is close to
> untestable.
>
> I don't currently have access to much except x86, though I can run quick
> tests on Alphas and probably a SiCortex.
>
> Regards,
> Nick Maclaren.
I didn't see any C99 support in gcc for IEEE_ARITHMETIC, so I'm assuming 
the choice is between methods currently in gcc/config/fpu* and use of 
SSE intrinsics or inline asm.
I'll be interested to see your analysis of whether Fortran standard is 
at all specific about what should happen with UNDERFLOW_MODE.
ifort has implemented these but they aren't documented.  I don't see how 
this stuff could be proprietary, but I can't account for viewpoints of 
others.
xlf document says there is no non-IEEE underflow mode.  HP Fortran doc 
says the function is present, but that's all.  Nag says they are present 
but do nothing.
At the moment I have a seemingly trivial C programming mental block.