I guess the root of the problem is when the autotools start up - we&#39;re having autoconf 2.68,<br>which is not liked by gcc 4.7.1<br><br><a href="http://configure.ac:33">configure.ac:33</a>: error: Please use exactly Autoconf 2.64 instead of 2.68.<br>
config/override.m4:12: _GCC_AUTOCONF_VERSION_CHECK is expanded from...<br><a href="http://configure.ac:33">configure.ac:33</a>: the top level<br>autom4te: m4 failed with exit status: 1<br><br><br>I&#39;m gonna downgrade to autoconf 2.64 and see what happens.<br>
<br>Alex <br><br><div class="gmail_quote">On Wed, Jul 4, 2012 at 6:21 PM, Martin Jansa <span dir="ltr">&lt;<a href="mailto:martin.jansa@gmail.com" target="_blank">martin.jansa@gmail.com</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">On Wed, Jul 04, 2012 at 06:06:55PM +0300, Damian, Alexandru wrote:<br>
&gt; I guess the problem is no lmld, (skips over), Mihai says that<br>
&gt;<br>
&gt; - library search path &quot;/usr/lib/cmplrs/cc2.11&quot; is unsafe for<br>
&gt; cross-compilation<br>
<br>
</div>It&#39;s one of issues I sometimes see during gcc upgrades.. and it was<br>
there with e.g. gcc-4.6.3 too<br>
<a href="http://build.shr-project.org/tests/jama/gcc-issue/gcc.issue" target="_blank">http://build.shr-project.org/tests/jama/gcc-issue/gcc.issue</a> (from Jan<br>
14)<br>
<div class="HOEnZb"><div class="h5"><br>
&gt; is the cause.<br>
&gt;<br>
&gt;<br>
&gt; I&#39;m unable to trace back the cc2.11 file name (path).<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt; Let&#39;s take it bit by bit, we have two issues here.<br>
&gt; 1. - why the configure hangs from time to time (what is that c2.11 ??)<br>
&gt;<br>
&gt; 2. - why sometimes it triggers massive rebuild even if nothing is changed.<br>
&gt;<br>
&gt; Regarding point 2, does anybody else sees it ?<br>
&gt;<br>
&gt; Alex<br>
&gt;<br>
&gt; On Wed, Jul 4, 2012 at 6:02 PM, Richard Purdie &lt;<br>
&gt; <a href="mailto:richard.purdie@linuxfoundation.org">richard.purdie@linuxfoundation.org</a>&gt; wrote:<br>
&gt;<br>
&gt; &gt; On Wed, 2012-07-04 at 17:50 +0300, Damian, Alexandru wrote:<br>
&gt; &gt; &gt; The question wasn&#39;t why it fails, but why sometimes it goes through,<br>
&gt; &gt; &gt; why sometimes fail, and why is it retriggered<br>
&gt; &gt; &gt; without any changes to the config or files (run 5 times the<br>
&gt; &gt; &gt; core-image-sato-sdk build in a row, and at least once it will fail).<br>
&gt; &gt; &gt;<br>
&gt; &gt; &gt; I don&#39;t trust systems that are not deterministic :).<br>
&gt; &gt;<br>
&gt; &gt; The check runs at the end of the do_configure task.<br>
&gt; &gt;<br>
&gt; &gt; The config.log is generated by do_compile. If do_configure is re-run,<br>
&gt; &gt; you&#39;ll hit the failure.<br>
&gt; &gt;<br>
&gt; &gt; &gt; This being said:<br>
&gt; &gt; &gt;<br>
&gt; &gt; &gt; grep -C5 -e &#39;CROSS COMPILE Badness:&#39; -e &#39;is unsafe for<br>
&gt; &gt; &gt; cross-compilation&#39;<br>
&gt; &gt; /home/ddalex/ssd/yocto/build/tmp/work/i586-poky-linux/gcc-4.7.1+svnr188658-r2/gcc-4_7-branch/build.i586-poky-linux.i586-poky-linux/gcc/config.log<br>
&gt; &gt; &gt; | return ldopen ();<br>
&gt; &gt; &gt; |   ;<br>
&gt; &gt; &gt; |   return 0;<br>
&gt; &gt; &gt; | }<br>
&gt; &gt; &gt; configure:8836: i586-poky-linux-gcc  -m32   -march=i586<br>
&gt; &gt; &gt; --sysroot=/home/ddalex/ssd/yocto/build/tmp/sysroots/qemux86 -o<br>
&gt; &gt; &gt; conftest -pipe -g -feliminate-unused-debug-types   -Wl,-O1<br>
&gt; &gt; &gt; -Wl,--hash-style=gnu -Wl,--as-needed conftest.c<br>
&gt; &gt; &gt; -L/usr/lib/cmplrs/cc2.11 -lmld &gt;&amp;5<br>
&gt; &gt; &gt;<br>
&gt; &gt; /home/ddalex/ssd/yocto/build/tmp/sysroots/x86_64-linux/usr/libexec/i586-poky-linux/gcc/i586-poky-linux/4.7.2/ld:<br>
&gt; &gt; warning: library search path &quot;/usr/lib/cmplrs/cc2.11&quot; is unsafe for<br>
&gt; &gt; cross-compilation<br>
&gt; &gt; &gt;<br>
&gt; &gt; /home/ddalex/ssd/yocto/build/tmp/sysroots/x86_64-linux/usr/libexec/i586-poky-linux/gcc/i586-poky-linux/4.7.2/ld:<br>
&gt; &gt; cannot find -lmld<br>
&gt; &gt; &gt; collect2: error: ld returned 1 exit status<br>
&gt; &gt; &gt; configure:8836: $? = 1<br>
&gt; &gt; &gt; configure: failed program was:<br>
&gt; &gt; &gt; | /* confdefs.h */<br>
&gt; &gt;<br>
&gt; &gt; So we need it to stop looking for libmld whatever that is. Hopefully<br>
&gt; &gt; there is a --disable-xxxx option we can pass in?<br>
&gt; &gt;<br>
&gt; &gt; Cheers,<br>
&gt; &gt;<br>
&gt; &gt; Richard<br>
&gt; &gt;<br>
&gt; &gt;<br>
&gt; &gt; _______________________________________________<br>
&gt; &gt; Openembedded-core mailing list<br>
&gt; &gt; <a href="mailto:Openembedded-core@lists.openembedded.org">Openembedded-core@lists.openembedded.org</a><br>
&gt; &gt; <a href="http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core" target="_blank">http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core</a><br>
&gt; &gt;<br>
<br>
&gt; _______________________________________________<br>
&gt; Openembedded-core mailing list<br>
&gt; <a href="mailto:Openembedded-core@lists.openembedded.org">Openembedded-core@lists.openembedded.org</a><br>
&gt; <a href="http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core" target="_blank">http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core</a><br>
<br>
<br>
</div></div><span class="HOEnZb"><font color="#888888">--<br>
Martin &#39;JaMa&#39; Jansa     jabber: <a href="mailto:Martin.Jansa@gmail.com">Martin.Jansa@gmail.com</a><br>
</font></span><br>_______________________________________________<br>
Openembedded-core mailing list<br>
<a href="mailto:Openembedded-core@lists.openembedded.org">Openembedded-core@lists.openembedded.org</a><br>
<a href="http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core" target="_blank">http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core</a><br>
<br></blockquote></div><br>