Re: [PATCH] Fix strtod for small DBL_DIG

[Christian Bruel <christian dot bruel at st dot com>]

On 05/16/2011 03:35 PM, Corinna Vinschen wrote:
> On May 16 15:27, Peter Rosin wrote:
> > Den 2011-05-16 14:33 skrev Corinna Vinschen:
> > > Thanks for the testcase.  However, I can still not reproduce the
> > > problem.  I tried again with float and strtof on Cygwin, but
> > > regardless of running it with or without your patch, it results
> > > in printing
> > >
> > >    12.3456789 12.3456789
> > >
> > > The strtod_r code is plain C so I'd expect that the conversion
> > > is target independent.  So why would it fail for arm but not for
> > > i386?
> >
> > Perhaps the reason for that is that in the arm case the double inside
> > strtod_r is only 32 bits? Try replacing the local doubles with floats?
> >
> > There's also a whole bunch of code conditional on _DOUBLE_IS_32BITS
> > to consider.
>
> You have a point there.  Ok, I can't test that easily and since it
> works fine for sizeof(double)==8 systems, and since Christian has
> apparently tested his patch, I checked it in.

Many thanks,

Just note for the record that the original report was for SH4, not ARM. 
(although I did test on ARM as well).

Best Regards

Christian

> Thanks,
> Corinna