Sorry but I am new to PETSc. This is out of my ability to find why this happens in PETSc.<div>I am happy to do some tests to look for the tests if you can instruct me how to do that.</div><div><br></div><div>There is an option in config/configure.py, which I do not use and may be relevant to check.</div>
<div><div> --with-64-bit-pointers=<bool></div><div> : Use 64 bit compilers and libraries</div><div><br></div><div>Thanks,</div><div>Mengda</div><br><div class="gmail_quote">On Mon, Jan 17, 2011 at 2:01 PM, Barry Smith <span dir="ltr"><<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><br>
Mengda,<br>
<br>
Thanks for the report. At the website it says "You perform arithmetic on 64-bit pointers and then pass the results to an inline function that expects an "int" data type." Do you know specifically where this is happening in PETSc? We don't intend to in PETSc " perform arithmetic on 64-bit pointers and then pass the results to an inline function that expects an "int" data type.<br>
<br>
Thanks<br>
<font color="#888888"><br>
Barry<br>
</font><div><div></div><div class="h5"><br>
<br>
<br>
On Jan 17, 2011, at 3:43 PM, Mengda Wu wrote:<br>
<br>
> This indeed is caused by a bug in Visual c++ 2005 64bit compiler when using<br>
> optimization.<br>
> The result is correct after installing the hotfix:<br>
> <a href="http://support.microsoft.com/kb/976617/" target="_blank">http://support.microsoft.com/kb/976617/</a><br>
> Thanks a lot!<br>
><br>
> Mengda<br>
><br>
> On Mon, Jan 17, 2011 at 12:27 PM, Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>> wrote:<br>
><br>
>><br>
>> Compiler bug. Immediately before the call to MatMult() in the code add the<br>
>> two lines<br>
>><br>
>> ierr = VecView(u,0);<br>
>> ierr = MatView(A,0);<br>
>><br>
>> how large are the two objects? Given the code it is inconceivable that<br>
>> suddenly the vector length becomes 57.<br>
>><br>
>> Barry<br>
>><br>
>><br>
>> On Jan 17, 2011, at 2:43 AM, Mengda Wu wrote:<br>
>><br>
>>> Hi all,<br>
>>><br>
>>> I just compiled the debugged and optimized versions of petsc-3.1-p7.<br>
>>> Both are successful. I am running on Windows Vista 64bit machine.<br>
>>> The C/C++ compiler is cl.exe from visual studio 2005 (64 bit) and no<br>
>> FORTRAN<br>
>>> compiler is used. BLAS/LAPACK<br>
>>> support comes from Intel MKL-10.1.3.028 (under em64t\lib). No MPI is<br>
>> used.<br>
>>><br>
>>> The debugged petsc was configured with:<br>
>>> $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
>>> --with-mpi=<br>
>>> 0<br>
>>><br>
>> --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_thread.lib,mkl_core.lib<br>
>>> ,libiomp5mt.lib] -CFLAGS='-MDd -W3 -Z7' -CXXFLAGS='-MDd -W3 -Z7'<br>
>>><br>
>>> The optimized petsc was configured with:<br>
>>> $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
>>> --with-mpi=<br>
>>> 0 --with-debugging=0<br>
>>> --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_threa<br>
>>> d.lib,mkl_core.lib,libiomp5mt.lib] -CFLAGS='-MD -wd4996 -O2'<br>
>> -CXXFLAGS='-MD<br>
>>> -w<br>
>>> d4996 -O2'<br>
>>><br>
>>> When I run ksp/examples/tutorials/Ex2.c. The result with debugged petsc<br>
>>> is<br>
>>> =================================================================<br>
>>> Norm of error 0.000156044 iterations 6<br>
>>> =================================================================<br>
>>><br>
>>> However, there are errors with the optimized petsc with the output as<br>
>>> follows:<br>
>>> =================================================================<br>
>>> [0]PETSC ERROR: --------------------- Error Message<br>
>>> ----------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: Nonconforming object sizes!<br>
>>> [0]PETSC ERROR: Mat mat,Vec y: global dim 56 57!<br>
>>> [0]PETSC ERROR:<br>
>>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: Petsc Release Version 3.1.0, Patch 7, Mon Dec 20 14:26:37<br>
>>> CST 20<br>
>>> 10<br>
>>> [0]PETSC ERROR: See docs/changes/index.html for recent updates.<br>
>>> [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.<br>
>>> [0]PETSC ERROR: See docs/index.html for manual pages.<br>
>>> [0]PETSC ERROR:<br>
>>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR:<br>
>>> D:\Develop\Test\PETSc\petsc-3.1-p7\src\ksp\ksp\examples\tutorial<br>
>>> s\ex2.exe on a cygwin-c- named CVBRL-38 by mengda Mon Jan 17 00:35:15<br>
>> 2011<br>
>>> [0]PETSC ERROR: Libraries linked from<br>
>>> /cygdrive/d/Develop/Test/PETSc/petsc-3.1-p<br>
>>> 7/cygwin-c-opt/lib<br>
>>> [0]PETSC ERROR: Configure run at Sun Jan 16 23:34:25 2011<br>
>>> [0]PETSC ERROR: Configure options --with-cc=cl --with-fc=0 --with-cxx=cl<br>
>>> --with-<br>
>>> mpi=0 --with-debugging=0<br>
>>> --with-blas-lapack-lib="[mkl_intel_lp64.lib,mkl_intel_t<br>
>>> hread.lib,mkl_core.lib,libiomp5mt.lib]" -CFLAGS="-MD -wd4996 -O2"<br>
>>> -CXXFLAGS="-MD<br>
>>> -wd4996 -O2" --useThreads=0<br>
>>> [0]PETSC ERROR:<br>
>>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: MatMult() line 1888 in<br>
>>> src/mat/interface/D:\Develop\Test\PETSc\P<br>
>>> ETSC-~1.1-P\src\mat\INTERF~1\matrix.c<br>
>>> [0]PETSC ERROR: main() line 146 in<br>
>>> src/ksp/ksp/examples/tutorials/D:\Develop\Tes<br>
>>> t\PETSc\PETSC-~1.1-P\src\ksp\ksp\examples\TUTORI~1\ex2.c<br>
>>><br>
>>> This application has requested the Runtime to terminate it in an unusual<br>
>>> way.<br>
>>> Please contact the application's support team for more information.<br>
>>> =================================================================<br>
>>><br>
>>> I am wondering what problems may lead to the errors. Please let me know<br>
>>> if you need more<br>
>>> information.<br>
>>><br>
>>> Thanks,<br>
>>> Mengda<br>
>>><br>
>>> Hi all,<br>
>>><br>
>>> I just compiled the debugged and optimized versions of petsc-3.1-p7.<br>
>> Both are successful. I am running on Windows Vista 64bit machine.<br>
>>> The C/C++ compiler is cl.exe from visual studio 2005 (64 bit) and no<br>
>> FORTRAN compiler is used. BLAS/LAPACK<br>
>>> support comes from Intel MKL-10.1.3.028 (under em64t\lib). No MPI is<br>
>> used.<br>
>>><br>
>>> The debugged petsc was configured with:<br>
>>> $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
>> --with-mpi=<br>
>>> 0<br>
>> --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_thread.lib,mkl_core.lib<br>
>>> ,libiomp5mt.lib] -CFLAGS='-MDd -W3 -Z7' -CXXFLAGS='-MDd -W3 -Z7'<br>
>>><br>
>>> The optimized petsc was configured with:<br>
>>> $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
>> --with-mpi=<br>
>>> 0 --with-debugging=0<br>
>> --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_threa<br>
>>> d.lib,mkl_core.lib,libiomp5mt.lib] -CFLAGS='-MD -wd4996 -O2'<br>
>> -CXXFLAGS='-MD -w<br>
>>> d4996 -O2'<br>
>>><br>
>>> When I run ksp/examples/tutorials/Ex2.c. The result with debugged<br>
>> petsc is<br>
>>> =================================================================<br>
>>> Norm of error 0.000156044 iterations 6<br>
>>> =================================================================<br>
>>><br>
>>> However, there are errors with the optimized petsc with the output as<br>
>> follows:<br>
>>> =================================================================<br>
>>> [0]PETSC ERROR: --------------------- Error Message<br>
>> ----------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: Nonconforming object sizes!<br>
>>> [0]PETSC ERROR: Mat mat,Vec y: global dim 56 57!<br>
>>> [0]PETSC ERROR:<br>
>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: Petsc Release Version 3.1.0, Patch 7, Mon Dec 20 14:26:37<br>
>> CST 20<br>
>>> 10<br>
>>> [0]PETSC ERROR: See docs/changes/index.html for recent updates.<br>
>>> [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.<br>
>>> [0]PETSC ERROR: See docs/index.html for manual pages.<br>
>>> [0]PETSC ERROR:<br>
>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR:<br>
>> D:\Develop\Test\PETSc\petsc-3.1-p7\src\ksp\ksp\examples\tutorial<br>
>>> s\ex2.exe on a cygwin-c- named CVBRL-38 by mengda Mon Jan 17 00:35:15<br>
>> 2011<br>
>>> [0]PETSC ERROR: Libraries linked from<br>
>> /cygdrive/d/Develop/Test/PETSc/petsc-3.1-p<br>
>>> 7/cygwin-c-opt/lib<br>
>>> [0]PETSC ERROR: Configure run at Sun Jan 16 23:34:25 2011<br>
>>> [0]PETSC ERROR: Configure options --with-cc=cl --with-fc=0 --with-cxx=cl<br>
>> --with-<br>
>>> mpi=0 --with-debugging=0<br>
>> --with-blas-lapack-lib="[mkl_intel_lp64.lib,mkl_intel_t<br>
>>> hread.lib,mkl_core.lib,libiomp5mt.lib]" -CFLAGS="-MD -wd4996 -O2"<br>
>> -CXXFLAGS="-MD<br>
>>> -wd4996 -O2" --useThreads=0<br>
>>> [0]PETSC ERROR:<br>
>> ----------------------------------------------------------------<br>
>>> --------<br>
>>> [0]PETSC ERROR: MatMult() line 1888 in<br>
>> src/mat/interface/D:\Develop\Test\PETSc\P<br>
>>> ETSC-~1.1-P\src\mat\INTERF~1\matrix.c<br>
>>> [0]PETSC ERROR: main() line 146 in<br>
>> src/ksp/ksp/examples/tutorials/D:\Develop\Tes<br>
>>> t\PETSc\PETSC-~1.1-P\src\ksp\ksp\examples\TUTORI~1\ex2.c<br>
>>><br>
>>> This application has requested the Runtime to terminate it in an unusual<br>
>> way.<br>
>>> Please contact the application's support team for more information.<br>
>>> =================================================================<br>
>>><br>
>>> I am wondering what problems may lead to the errors. Please let me<br>
>> know if you need more<br>
>>> information.<br>
>>><br>
>>> Thanks,<br>
>>> Mengda<br>
>><br>
>><br>
><br>
> This indeed is caused by a bug in Visual c++ 2005 64bit compiler when using optimization.<br>
> The result is correct after installing the hotfix: <a href="http://support.microsoft.com/kb/976617/" target="_blank">http://support.microsoft.com/kb/976617/</a><br>
> Thanks a lot!<br>
><br>
> Mengda<br>
><br>
> On Mon, Jan 17, 2011 at 12:27 PM, Barry Smith <<a href="mailto:bsmith@mcs.anl.gov">bsmith@mcs.anl.gov</a>> wrote:<br>
><br>
> Compiler bug. Immediately before the call to MatMult() in the code add the two lines<br>
><br>
> ierr = VecView(u,0);<br>
> ierr = MatView(A,0);<br>
><br>
> how large are the two objects? Given the code it is inconceivable that suddenly the vector length becomes 57.<br>
><br>
> Barry<br>
><br>
><br>
> On Jan 17, 2011, at 2:43 AM, Mengda Wu wrote:<br>
><br>
> > Hi all,<br>
> ><br>
> > I just compiled the debugged and optimized versions of petsc-3.1-p7.<br>
> > Both are successful. I am running on Windows Vista 64bit machine.<br>
> > The C/C++ compiler is cl.exe from visual studio 2005 (64 bit) and no FORTRAN<br>
> > compiler is used. BLAS/LAPACK<br>
> > support comes from Intel MKL-10.1.3.028 (under em64t\lib). No MPI is used.<br>
> ><br>
> > The debugged petsc was configured with:<br>
> > $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
> > --with-mpi=<br>
> > 0<br>
> > --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_thread.lib,mkl_core.lib<br>
> > ,libiomp5mt.lib] -CFLAGS='-MDd -W3 -Z7' -CXXFLAGS='-MDd -W3 -Z7'<br>
> ><br>
> > The optimized petsc was configured with:<br>
> > $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl'<br>
> > --with-mpi=<br>
> > 0 --with-debugging=0<br>
> > --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_threa<br>
> > d.lib,mkl_core.lib,libiomp5mt.lib] -CFLAGS='-MD -wd4996 -O2' -CXXFLAGS='-MD<br>
> > -w<br>
> > d4996 -O2'<br>
> ><br>
> > When I run ksp/examples/tutorials/Ex2.c. The result with debugged petsc<br>
> > is<br>
> > =================================================================<br>
> > Norm of error 0.000156044 iterations 6<br>
> > =================================================================<br>
> ><br>
> > However, there are errors with the optimized petsc with the output as<br>
> > follows:<br>
> > =================================================================<br>
> > [0]PETSC ERROR: --------------------- Error Message<br>
> > ----------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: Nonconforming object sizes!<br>
> > [0]PETSC ERROR: Mat mat,Vec y: global dim 56 57!<br>
> > [0]PETSC ERROR:<br>
> > ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: Petsc Release Version 3.1.0, Patch 7, Mon Dec 20 14:26:37<br>
> > CST 20<br>
> > 10<br>
> > [0]PETSC ERROR: See docs/changes/index.html for recent updates.<br>
> > [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.<br>
> > [0]PETSC ERROR: See docs/index.html for manual pages.<br>
> > [0]PETSC ERROR:<br>
> > ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR:<br>
> > D:\Develop\Test\PETSc\petsc-3.1-p7\src\ksp\ksp\examples\tutorial<br>
> > s\ex2.exe on a cygwin-c- named CVBRL-38 by mengda Mon Jan 17 00:35:15 2011<br>
> > [0]PETSC ERROR: Libraries linked from<br>
> > /cygdrive/d/Develop/Test/PETSc/petsc-3.1-p<br>
> > 7/cygwin-c-opt/lib<br>
> > [0]PETSC ERROR: Configure run at Sun Jan 16 23:34:25 2011<br>
> > [0]PETSC ERROR: Configure options --with-cc=cl --with-fc=0 --with-cxx=cl<br>
> > --with-<br>
> > mpi=0 --with-debugging=0<br>
> > --with-blas-lapack-lib="[mkl_intel_lp64.lib,mkl_intel_t<br>
> > hread.lib,mkl_core.lib,libiomp5mt.lib]" -CFLAGS="-MD -wd4996 -O2"<br>
> > -CXXFLAGS="-MD<br>
> > -wd4996 -O2" --useThreads=0<br>
> > [0]PETSC ERROR:<br>
> > ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: MatMult() line 1888 in<br>
> > src/mat/interface/D:\Develop\Test\PETSc\P<br>
> > ETSC-~1.1-P\src\mat\INTERF~1\matrix.c<br>
> > [0]PETSC ERROR: main() line 146 in<br>
> > src/ksp/ksp/examples/tutorials/D:\Develop\Tes<br>
> > t\PETSc\PETSC-~1.1-P\src\ksp\ksp\examples\TUTORI~1\ex2.c<br>
> ><br>
> > This application has requested the Runtime to terminate it in an unusual<br>
> > way.<br>
> > Please contact the application's support team for more information.<br>
> > =================================================================<br>
> ><br>
> > I am wondering what problems may lead to the errors. Please let me know<br>
> > if you need more<br>
> > information.<br>
> ><br>
> > Thanks,<br>
> > Mengda<br>
> ><br>
> > Hi all,<br>
> ><br>
> > I just compiled the debugged and optimized versions of petsc-3.1-p7. Both are successful. I am running on Windows Vista 64bit machine.<br>
> > The C/C++ compiler is cl.exe from visual studio 2005 (64 bit) and no FORTRAN compiler is used. BLAS/LAPACK<br>
> > support comes from Intel MKL-10.1.3.028 (under em64t\lib). No MPI is used.<br>
> ><br>
> > The debugged petsc was configured with:<br>
> > $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl' --with-mpi=<br>
> > 0 --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_thread.lib,mkl_core.lib<br>
> > ,libiomp5mt.lib] -CFLAGS='-MDd -W3 -Z7' -CXXFLAGS='-MDd -W3 -Z7'<br>
> ><br>
> > The optimized petsc was configured with:<br>
> > $ ./config/configure.py --with-cc='cl' --with-fc=0 --with-cxx='cl' --with-mpi=<br>
> > 0 --with-debugging=0 --with-blas-lapack-lib=[mkl_intel_lp64.lib,mkl_intel_threa<br>
> > d.lib,mkl_core.lib,libiomp5mt.lib] -CFLAGS='-MD -wd4996 -O2' -CXXFLAGS='-MD -w<br>
> > d4996 -O2'<br>
> ><br>
> > When I run ksp/examples/tutorials/Ex2.c. The result with debugged petsc is<br>
> > =================================================================<br>
> > Norm of error 0.000156044 iterations 6<br>
> > =================================================================<br>
> ><br>
> > However, there are errors with the optimized petsc with the output as follows:<br>
> > =================================================================<br>
> > [0]PETSC ERROR: --------------------- Error Message ----------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: Nonconforming object sizes!<br>
> > [0]PETSC ERROR: Mat mat,Vec y: global dim 56 57!<br>
> > [0]PETSC ERROR: ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: Petsc Release Version 3.1.0, Patch 7, Mon Dec 20 14:26:37 CST 20<br>
> > 10<br>
> > [0]PETSC ERROR: See docs/changes/index.html for recent updates.<br>
> > [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.<br>
> > [0]PETSC ERROR: See docs/index.html for manual pages.<br>
> > [0]PETSC ERROR: ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: D:\Develop\Test\PETSc\petsc-3.1-p7\src\ksp\ksp\examples\tutorial<br>
> > s\ex2.exe on a cygwin-c- named CVBRL-38 by mengda Mon Jan 17 00:35:15 2011<br>
> > [0]PETSC ERROR: Libraries linked from /cygdrive/d/Develop/Test/PETSc/petsc-3.1-p<br>
> > 7/cygwin-c-opt/lib<br>
> > [0]PETSC ERROR: Configure run at Sun Jan 16 23:34:25 2011<br>
> > [0]PETSC ERROR: Configure options --with-cc=cl --with-fc=0 --with-cxx=cl --with-<br>
> > mpi=0 --with-debugging=0 --with-blas-lapack-lib="[mkl_intel_lp64.lib,mkl_intel_t<br>
> > hread.lib,mkl_core.lib,libiomp5mt.lib]" -CFLAGS="-MD -wd4996 -O2" -CXXFLAGS="-MD<br>
> > -wd4996 -O2" --useThreads=0<br>
> > [0]PETSC ERROR: ----------------------------------------------------------------<br>
> > --------<br>
> > [0]PETSC ERROR: MatMult() line 1888 in src/mat/interface/D:\Develop\Test\PETSc\P<br>
> > ETSC-~1.1-P\src\mat\INTERF~1\matrix.c<br>
> > [0]PETSC ERROR: main() line 146 in src/ksp/ksp/examples/tutorials/D:\Develop\Tes<br>
> > t\PETSc\PETSC-~1.1-P\src\ksp\ksp\examples\TUTORI~1\ex2.c<br>
> ><br>
> > This application has requested the Runtime to terminate it in an unusual way.<br>
> > Please contact the application's support team for more information.<br>
> > =================================================================<br>
> ><br>
> > I am wondering what problems may lead to the errors. Please let me know if you need more<br>
> > information.<br>
> ><br>
> > Thanks,<br>
> > Mengda<br>
><br>
><br>
<br>
</div></div></blockquote></div><br></div>