[Insight-developers] Nonportable Code

Sat Aug 15 13:31:37 EDT 2009

Le 15 août 09 à 18:31, Luis Ibanez a écrit :

> Hi Gaetan,
>
> The logic of the computation in :
>
> http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/Common/itkImageBase.txx?root=Insight&r1=1.56&r2=1.60
>
> we flawed anyways.
>
> If the purpose is to verify whether the request for a
> number in pixels along every dimension results in
> a combined allocation that is larger that the largest
> number of pixels that can be allocated

It's rather *manipulated* than allocated. That's why this check can be  
done in ImageBase. It is valid for all the image types (sparse, label  
map, component trees, ...).

> , then such
> evaluation can't simply be performed at the output
> of the for loop.
>
> Instead it should have been tested at every increment
> of the "num" value.
>
> Otherwise, by the time the num value gets larger that
> the container type, it will wrap around and it will be
> presented as a negative number.
>
> The check should have been of the form:
>
> Instead of:
>         num *= bufferSize[i];
>
> do:
>
>       if ( NumericTraits<SizeValueType>::max() / buffersize[i] > num )

You're right. That's a lot better. I think there is a small mistake  
though.

   if ( NumericTraits<SizeValueType>::max() / buffersize[i] < num )

seems better :-)

>          {
>          itkExceptionMacro();
>          }
>      num *= bufferSize[i];
>
>
> all operations in an integer type are equivalent
> to the same operation module Integer::Max.
>
> Checking after the fact will not reveal an overflow.
> (unless we had a type that is largest that size_t,
> which is not the case).
>
>

Using a larger type was the (non portable) idea.

Thanks for the tip. I'll try it... after the release :-)

Gaëtan

>      Luis
>
>
> -------------------------------------------------------------
> 2009/8/15 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>
>
> Le 15 août 09 à 05:30, Bill Lorensen a écrit :
>
>
> vxl_int_64 must be void on some platforms.
>
> http://www.cdash.org/CDash/viewBuildError.php?buildid=402232
>
> I repeat my concerns about such fundamental changes so close to a  
> release.
>
> I have commented the changes and filed bug reports to keep a track.
>
>  http://www.itk.org/Bug/view.php?id=9427
>  http://www.itk.org/Bug/view.php?id=9426
>
> We are back to the previous situation.
>
>  http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/Common/itkImageBase.txx?root=Insight&r1=1.56&r2=1.60
>
> Gaëtan
>
>
>
> Bill
>
> On Fri, Aug 14, 2009 at 1:25 PM, Tom Vercauteren<tom.vercauteren at m4x.org 
> > wrote:
> Hi Sean,
>
> Yes c++0x will have cstdint. It is already in boost.
>
> In the meantime we might want to rely on pstdint.h (a portable  
> stdint):
> http://www.azillionmonkeys.com/qed/pstdint.h
>
> See wikipedia for more information
> http://en.wikipedia.org/wiki/Stdint.h
>
> But it looks like a large change to do right before the release.
>
> Tom
>
> On Fri, Aug 14, 2009 at 19:21, Sean McBride<sean at rogue-research.com>  
> wrote:
> On 8/14/09 6:56 PM, Gaëtan Lehmann said:
>
> I've used that option, but I'm quite sure the types defined there are
> wrong.
> For example,
>
>  #ifdef _WIN32
>    typedef long      ITK_INT64;
>  #endif
>
> and
>
>  #ifdef _WIN32
>    typedef unsigned long ITK_UINT64;
>  #endif
>
> don't like right to me.
>
> Agreed.
>
> Tom's way - vxl_int_64 - seems to be a better option.
>
> Should we fix the content of itkIntType.h by using vxl_int_64 types?
> And do we even need itkIntType.h?
>
> This whole issue of fixed size types was solved and standardised with
> C99 (ISO/IEC 9899:1999), see:
> <http://en.wikipedia.org/wiki/Stdint.h>
>
> I know ITK is C++, not C.  Anyone know if the next C++ standard C++-0x
> will incorporate those types from C99?
>
> If so, seems to me we should use int32_t instead of creating an
> INT_INT32.  (I know some compilers do not have stdint.h, but we can
> create compatible typedefs in those cases.)
>
> --
> ____________________________________________________________
> Sean McBride, B. Eng                 sean at rogue-research.com
> Rogue Research                        www.rogue-research.com
> Mac Software Developer              Montréal, Québec, Canada
>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>
> -- 
> Gaëtan Lehmann
> Biologie du Développement et de la Reproduction
> INRA de Jouy-en-Josas (France)
> tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
> http://voxel.jouy.inra.fr  http://www.itk.org
> http://www.mandriva.org  http://www.bepo.fr
>
>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the ITK FAQ at: http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>
>

-- 
Gaëtan Lehmann
Biologie du Développement et de la Reproduction
INRA de Jouy-en-Josas (France)
tel: +33 1 34 65 29 66    fax: 01 34 65 29 09
http://voxel.jouy.inra.fr  http://www.itk.org
http://www.mandriva.org  http://www.bepo.fr

-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 203 bytes
Desc: Ceci est une signature ?lectronique PGP
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20090815/ba3c8e32/attachment.pgp>