[Insight-developers] Failing tests from DashFest

[Gaëtan Lehmann]

Le 19 nov. 09 à 17:43, Bradley Lowekamp a écrit :

> Hello Gaetan,
>

Hi Brad,

> This looks like this is a bug with the ccache.
>
> It appears that using the SSE2 intrinsics, with optimization and  
> with ccache when certain order of rounding methods occur incorrect  
> results are computed. This may only occur with the tests and not  
> actual ITK code. All appears to work in Debug mode or with the /usr/ 
> bin/gcc.
>
> SSE2 can be disabled with the following cmake variable:
>  VNL_CONFIG_ENABLE_SSE2_ROUNDING
>
> Or just don't use ccache.

Thanks to have looked at that. It must be a bug with the ccache/gcc  
versions installed on that host, because I'm using ccache on all my  
host without seeing that problem anywhere else.

>
> BTW, I submitting that patch for the fenv issue and that appears to  
> be working now on your Suns.

Yes, I've seen the green build this morning :-)

Thanks a lot!

Gaëtan

>
> Brad
>
> On Nov 9, 2009, at 1:48 PM, Gaëtan Lehmann wrote:
>
>>
>> Le 9 nov. 09 à 15:58, Bradley Lowekamp a écrit :
>>
>>> Hello Gaetan,
>>>
>>
>> Hi Brad,
>>
>>> 	On your system voxel.jouy.inra.fr I see that you pointed out that  
>>> itkMathRoundTest failed, also vnltest_math failed on several of  
>>> the rounding related methods. Can you please give us a little more  
>>> information about this system, compiler, processor etc.
>>
>> sure
>>
>> [glehmann at voxel ~]$ cat /proc/cpuinfo | grep model
>> model		: 23
>> model name	: Intel(R) Xeon(R) CPU           E5405  @ 2.00GHz
>> model		: 23
>> model name	: Intel(R) Xeon(R) CPU           E5405  @ 2.00GHz
>> model		: 23
>> model name	: Intel(R) Xeon(R) CPU           E5405  @ 2.00GHz
>> model		: 23
>> model name	: Intel(R) Xeon(R) CPU           E5405  @ 2.00GHz
>> [glehmann at voxel ~]$ cat /etc/redhat-release
>> CentOS release 4.7 (Final)
>> Nahant
>> [glehmann at voxel ~]$ gcc --version
>> gcc (GCC) 3.4.6 20060404 (Red Hat 3.4.6-10)
>> Copyright (C) 2006 Free Software Foundation, Inc.
>> This is free software; see the source for copying conditions.   
>> There is NO
>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR  
>> PURPOSE.
>>
>>> I would also like to know if you are using the SSE versions of the  
>>> rounding methods, ie is  VNL_CONFIG_ENABLE_SSE2_ROUNDING turned on?
>>>
>>
>> It must be the default - I made this build only for the fest, with  
>> a fresh cvs checkout
>>
>> [glehmann at voxel ~]$ grep VNL_CONFIG_ENABLE_SSE2_ROUNDING /tmp/ 
>> itkfest/Insight/build/CMakeCache.txt
>> VNL_CONFIG_ENABLE_SSE2_ROUNDING:BOOL=ON
>> //ADVANCED property for variable: VNL_CONFIG_ENABLE_SSE2_ROUNDING
>> VNL_CONFIG_ENABLE_SSE2_ROUNDING-ADVANCED:INTERNAL=1
>>
>> Gaëtan
>>
>> -- 
>> 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
>>
>
> ========================================================
> Bradley Lowekamp
> Lockheed Martin Contractor for
> Office of High Performance Computing and Communications
> National Library of Medicine
> blowekamp at mail.nih.gov
>
>

-- 
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/20091119/5f000ec7/attachment.pgp>