OK, I have a vanilla configure+make running on both SPARC/Solaris-10 and AMD64/Solaris-11.<div>I am using the 12.3 Oracle compilers in both cases to match the original report.<br><div>I&#39;ll post the results when they complete.</div>
<div><br></div><div>In the meantime, I took a quick look at the code and have a pretty reasonable guess as to the cause.</div><div>Looking at ompi/mca/coll/ml/coll_ml.h I see:</div><div><br></div><div>   827  int mca_coll_ml_memsync_intra(mca_coll_ml_module_t *module, int bank_index);</div>
<div>[...]</div><div><div>   996  static inline __opal_attribute_always_inline__</div><div>   997          int mca_coll_ml_buffer_recycling(mca_coll_ml_collective_operation_progress_t *ml_request)</div><div>   998  {</div>
</div><div>[...]</div><div>  1023                  rc = mca_coll_ml_memsync_intra(ml_module, ml_memblock-&gt;memsync_counter);</div><div>[...]</div><div>  1041  }</div><div><br></div><div>Based on past experience w/ the Sun/Oracle compilers on another project (See <a href="http://bugzilla.hcs.ufl.edu/cgi-bin/bugzilla3/show_bug.cgi?id=193">http://bugzilla.hcs.ufl.edu/cgi-bin/bugzilla3/show_bug.cgi?id=193</a> ), I suspect that this static-inline-always function is being emitted by the compiler in every object which includes this header even if they don&#39;t call it..  The call on line 1023 then results in the undefined reference to mca_coll_ml_memsync_intra.  Basically it is not safe for an inline function in a header to call an extern function that isn&#39;t available to every object that includes the header REGARDLESS of whether the object invokes the inline function or not.</div>
<div><br></div><div>-Paul</div><div><br></div><div><br><br><div class="gmail_quote">On Fri, Aug 24, 2012 at 4:40 PM, Ralph Castain <span dir="ltr">&lt;<a href="mailto:rhc@open-mpi.org" target="_blank">rhc@open-mpi.org</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="word-wrap:break-word">Oracle uses an abysmally complicated configure line, but nearly all of it is irrelevant to the problem here. For this, I would suggest just doing a vanilla ./configure - if the component gets pulled into libmpi, then we know there is a problem.<div>
<br></div><div>Thanks!</div><div><br></div><div>Just FYI: here is there actual configure line, just in case you spot something problematic:</div><div><br></div><div><pre>CC=cc CXX=CC F77=f77 FC=f90  --with-openib  --enable-openib-connectx-xrc  --without-udapl 
--disable-openib-ibcm  --enable-btl-openib-failover   --without-dtrace  --enable-heterogeneous
--enable-cxx-exceptions --enable-shared --enable-orterun-prefix-by-default --with-sge
--enable-mpi-f90 --with-mpi-f90-size=small  --disable-peruse --disable-state 
--disable-mpi-thread-multiple   --disable-debug  --disable-mem-debug  --disable-mem-profile 
CFLAGS=&quot;-xtarget=ultra3 -m32 -xarch=sparcvis2 -xprefetch -xprefetch_level=2 -xvector=lib -Qoption
cg -xregs=no%appl -xdepend=yes -xbuiltin=%all -xO5&quot;  CXXFLAGS=&quot;-xtarget=ultra3 -m32
-xarch=sparcvis2 -xprefetch -xprefetch_level=2 -xvector=lib -Qoption cg -xregs=no%appl -xdepend=yes
-xbuiltin=%all -xO5 -Bstatic -lCrun -lCstd -Bdynamic&quot;  FFLAGS=&quot;-xtarget=ultra3 -m32 -xarch=sparcvis2
-xprefetch -xprefetch_level=2 -xvector=lib -Qoption cg -xregs=no%appl -stackvar -xO5&quot; 
FCFLAGS=&quot;-xtarget=ultra3 -m32 -xarch=sparcvis2 -xprefetch -xprefetch_level=2 -xvector=lib -Qoption
cg -xregs=no%appl -stackvar -xO5&quot;  
--prefix=/workspace/euloh/hpc/mtt-scratch/burl-ct-t2k-3/ompi-tarball-testing/installs/JA08/install 
--mandir=${prefix}/man  --bindir=${prefix}/bin  --libdir=${prefix}/lib 
--includedir=${prefix}/include   --with-tm=/ws/ompi-tools/orte/torque/current/shared-install32 
--enable-contrib-no-build=vt --with-package-string=&quot;Oracle Message Passing Toolkit &quot;
--with-ident-string=&quot;@(#)RELEASE VERSION 1.9openmpi-1.5.4-r1.9a1r27092&quot;</pre><div><br></div><div>and the error he gets is:</div><div><br></div><div><pre>make[2]: Entering directory
`/workspace/euloh/hpc/mtt-scratch/burl-ct-t2k-3/ompi-tarball-testing/mpi-install/s3rI/src/openmpi-1.9a1r27092/ompi/tools/ompi_info&#39;
  CCLD     ompi_info
Undefined			first referenced
 symbol  			    in file
mca_coll_ml_memsync_intra           ../../../ompi/.libs/libmpi.so
ld: fatal: symbol referencing errors. No output written to .libs/ompi_info
make[2]: *** [ompi_info] Error 2
make[2]: Leaving directory
`/workspace/euloh/hpc/mtt-scratch/burl-ct-t2k-3/ompi-tarball-testing/mpi-install/s3rI/src/openmpi-1.9a1r27092/ompi/tools/ompi_info&#39;
make[1]: *** [install-recursive] Error 1
make[1]: Leaving directory
`/workspace/euloh/hpc/mtt-scratch/burl-ct-t2k-3/ompi-tarball-testing/mpi-install/s3rI/src/openmpi-1.9a1r27092/ompi&#39;
make: *** [install-recursive] Error 1
</pre><div><br></div></div><div>On Aug 24, 2012, at 4:30 PM, Paul Hargrove &lt;<a href="mailto:phhargrove@lbl.gov" target="_blank">phhargrove@lbl.gov</a>&gt; wrote:</div><div><br><blockquote type="cite">I have access to a few different Solaris machines and can offer to build the trunk if somebody tells me what configure flags are desired.<div>
<br></div><div>-Paul<br><br><div class="gmail_quote">On Fri, Aug 24, 2012 at 8:54 AM, Ralph Castain <span dir="ltr">&lt;<a href="mailto:rhc@open-mpi.org" target="_blank">rhc@open-mpi.org</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Eugene - can you confirm that this is only happening on the one Solaris system? In other words, is this a general issue or something specific to that one machine?<br>


<br>
I&#39;m wondering because if it is just the one machine, then it might be something strange about how it is setup - perhaps the version of Solaris, or it is configuring --enable-static, or...<br>
<br>
Just trying to assess how general a problem this might be, and thus if this should be a blocker or not.<br>
<br>
On Aug 24, 2012, at 8:00 AM, Eugene Loh &lt;<a href="mailto:eugene.loh@oracle.com" target="_blank">eugene.loh@oracle.com</a>&gt; wrote:<br>
<br>
&gt; On 08/24/12 09:54, Shamis, Pavel wrote:<br>
&gt;&gt; Maybe there is a chance to get direct access to this system ?<br>
&gt; No.<br>
&gt;<br>
&gt; But I&#39;m attaching compressed log files from configure/make.<br>
&gt;<br>
&gt; &lt;tarball-of-log-files.tar.bz2&gt;_______________________________________________<br>
&gt; devel mailing list<br>
&gt; <a href="mailto:devel@open-mpi.org" target="_blank">devel@open-mpi.org</a><br>
&gt; <a href="http://www.open-mpi.org/mailman/listinfo.cgi/devel" target="_blank">http://www.open-mpi.org/mailman/listinfo.cgi/devel</a><br>
<br>
<br>
_______________________________________________<br>
devel mailing list<br>
<a href="mailto:devel@open-mpi.org" target="_blank">devel@open-mpi.org</a><br>
<a href="http://www.open-mpi.org/mailman/listinfo.cgi/devel" target="_blank">http://www.open-mpi.org/mailman/listinfo.cgi/devel</a><br>
</blockquote></div><br><br clear="all"><span class="HOEnZb"><font color="#888888"><div><br></div>-- <br><font face="courier new, monospace"><div>Paul H. Hargrove                          <a href="mailto:PHHargrove@lbl.gov" target="_blank">PHHargrove@lbl.gov</a></div>

<div>Future Technologies Group</div><div>Computer and Data Sciences Department     Tel: <a href="tel:%2B1-510-495-2352" value="+15104952352" target="_blank">+1-510-495-2352</a></div><div>Lawrence Berkeley National Laboratory     Fax: <a href="tel:%2B1-510-486-6900" value="+15104866900" target="_blank">+1-510-486-6900</a></div>
</font><br>
</font></span></div><span class="HOEnZb"><font color="#888888">
_______________________________________________<br>devel mailing list<br><a href="mailto:devel@open-mpi.org" target="_blank">devel@open-mpi.org</a><br><a href="http://www.open-mpi.org/mailman/listinfo.cgi/devel" target="_blank">http://www.open-mpi.org/mailman/listinfo.cgi/devel</a></font></span></blockquote>
</div><br></div></div><br>_______________________________________________<br>
devel mailing list<br>
<a href="mailto:devel@open-mpi.org">devel@open-mpi.org</a><br>
<a href="http://www.open-mpi.org/mailman/listinfo.cgi/devel" target="_blank">http://www.open-mpi.org/mailman/listinfo.cgi/devel</a><br></blockquote></div><br><br clear="all"><div><br></div>-- <br><font face="courier new, monospace"><div>
Paul H. Hargrove                          <a href="mailto:PHHargrove@lbl.gov" target="_blank">PHHargrove@lbl.gov</a></div><div>Future Technologies Group</div><div>Computer and Data Sciences Department     Tel: +1-510-495-2352</div>
<div>Lawrence Berkeley National Laboratory     Fax: +1-510-486-6900</div></font><br>
</div></div>