On Tue, Mar 27, 2012 at 10:44 PM, David Fuentes <span dir="ltr"><<a href="mailto:fuentesdt@gmail.com">fuentesdt@gmail.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi Satish,<div><br></div><div>what version of cusp/thrust are you using with cuda 4.1 ? </div></blockquote><div><br></div><div>Thrust is packaged with 4.1 and that is the one you should use (just --with-thrust).</div><div>
We use the latest pull from CUSP, which is 2.00</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">
<div><br><div class="gmail_quote">On Tue, Mar 27, 2012 at 10:23 PM, Satish Balay <span dir="ltr"><<a href="mailto:balay@mcs.anl.gov" target="_blank">balay@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">Chetan, thanks for the note. I guess I should have first searched<br>
petsc archives :).<br>
<br>
Looks like we default to -g3 for gcc, but -g for g++/gfortran [due to<br>
bugs in some version of these compilers]. However -g3 is usefull for<br>
debugging macros - so its a good default.<br>
<br>
For now - I'm switching the default from -g3 to -g for cuda<br>
builds [one can always override this with COPTFLAGS options]<br>
<span><font color="#888888"><br>
Satish<br>
</font></span><div><div><br>
On Tue, 27 Mar 2012, Chetan Jhurani wrote:<br>
<br>
> Satish,<br>
><br>
> This is because of -g3 option to nvcc. Attaching my<br>
> earlier mail to petsc-maint.<br>
><br>
> Chetan<br>
><br>
> > -----Original Message-----<br>
> > From: <a href="mailto:petsc-dev-bounces@mcs.anl.gov" target="_blank">petsc-dev-bounces@mcs.anl.gov</a> [mailto:<a href="mailto:petsc-dev-bounces@mcs.anl.gov" target="_blank">petsc-dev-bounces@mcs.anl.gov</a>] On Behalf Of Satish Balay<br>
> > Sent: Tuesday, March 27, 2012 7:31 PM<br>
> > To: <a href="mailto:petsc-dev@mcs.anl.gov" target="_blank">petsc-dev@mcs.anl.gov</a><br>
> > Subject: [petsc-dev] cuda 4.1 errors with sm_20<br>
> ><br>
> > I get errors with building petsc using cuda-4.1 with 'sm_20' on<br>
> > linux. Reverting to cuda-4.0 [with cusp-v0.3.1, thrust-1.6.0]<br>
> > overcomes this issue.<br>
> ><br>
> > >From<br>
> > <a href="http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/docs/CUDA_Toolkit_Release_Notes_And_" target="_blank">http://developer.download.nvidia.com/compute/cuda/4_1/rel/toolkit/docs/CUDA_Toolkit_Release_Notes_And_</a><br>
> > Errata.txt<br>
> > * For target architecture sm_2x, a new compiler component 'cicc' is used instead of 'nvopencc'.<br>
> ><br>
> > So the new compiler used for sm20 - 'cicc' is perhaps buggy.<br>
> ><br>
> > Matt,Sean or Barry, Do you see the same behavior on Mac?<br>
> ><br>
> > Satish<br>
> > ---------------------------<br>
> > balay@bb30:~/petsc-dev/src/mat/impls/aij/seq/seqcusp>nvcc -g -arch=sm_20 -c --compiler-options="-Wall<br>
> > -Wwrite-strings -Wno-strict-aliasing -Wno-unknown-pragmas -g3 -fno-inline -O0 -I/home/balay/petsc-<br>
> > dev/include -I/home/balay/petsc-dev/arch-cuda-double/include -I/usr/local/cuda/include -<br>
> > I/home/balay/petsc-dev/arch-cuda-double/include/txpetscgpu/include -I/usr/local/cusp-v0.3.1/ -<br>
> > D__INSDIR__=src/mat/impls/aij/seq/seqcusp/" <a href="http://aijcusp.cu" target="_blank">aijcusp.cu</a><br>
> > <built-in>(2): error: "__STDC_HOSTED__" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(8): error: "__WCHAR_TYPE__" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(115): error: "__x86_64" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(116): error: "__x86_64__" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(126): error: "__linux__" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(129): error: "__unix__" is predefined; attempted redefinition ignored<br>
> ><br>
> > <built-in>(130): error: "unix" is predefined; attempted redefinition ignored<br>
> ><br>
> > 7 errors detected in the compilation of "/tmp/tmpxft_00003334_00000000-7_aijcusp.cpp3.i".<br>
><br>
<br>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener<br>