<p dir="ltr">Compare the output of configure run on working and non working versions. Usually this is very helpful. See if its enabling or disabling something new. </p>
<div class="gmail_quote">On Mar 10, 2016 10:55 PM, "Jens Rehsack" <<a href="mailto:rehsack@gmail.com">rehsack@gmail.com</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
> Am 10.03.2016 um 06:13 schrieb Gary Thomas <<a href="mailto:gary@mlbassoc.com">gary@mlbassoc.com</a>>:<br>
><br>
> I'm working on a package (amanda - the Advanced Maryland Archiving<br>
> system) that is written heavily in perl with swig interfaces to C.<br>
> This code ran great until the update to perl 5.22; it now dies a<br>
> horrible death on almost every activity. These failures seem to<br>
> always be in the swig generated wrappers, but that may just be<br>
> where most of the work gets done.<br>
><br>
> I've narrowed this down to exactly the change to perl 5.22 from<br>
> 5.20. Using bisect as well as experimentation (e.g. trying all<br>
> the compiler combinations that have occurred since a last good<br>
> version) and I can go from working to failing by only the change<br>
> in perl.<br>
><br>
> The interesting (scary) thing is that I've built amanda for my<br>
> target natively on my board running debian, including perl 5.22.<br>
> This means I can't say definitively that perl 5.22 is the culprit<br>
> as on debian it runs fine. So, it's got something to do with the<br>
> OE environment/porting/packaging of perl and not just the revision.<br>
><br>
> I've also tested this on multiple architectures (ARM, PowerPC) with<br>
> the same results - with perl 5.20 amanda works, with perl 5.22 it fails.<br>
><br>
> I've compared the actual 5.22.1 sources used by OE-core and debian<br>
> and they are subtly different, although I can't pinpoint any change<br>
> that might be responsible.<br>
><br>
> For the moment, I can just fall back to perl 5.20 for my target<br>
> that needs to run amanda, but this isn't a real solution (e.g.<br>
> in this state I can't propose my recipe to any layer as it's<br>
> totally broken with the current OE-core). I'd like to see this<br>
> fixed but the amanda code (swig wrappers) are horribly complex<br>
> which makes debugging quite the challenge, not to mention they<br>
> may be about the only way to uncover the bug, whether it's in<br>
> amanda or perl.<br>
><br>
> Any suggestions on how to move forward?<br>
<br>
Since I have no clue what's wrong and how it fails (backtrace<br>
would point in some directions), several ideas might work:<br>
<br>
How clean is your build location (we realize that often between<br>
updates some files remain in our target images until we wipe<br>
tmp/ - cleansstate for image doesn't help ...)?<br>
<br>
Did you prove the library path's of your *.so's? Perl does<br>
almost everything within libperl.so - build against wrong version<br>
causes in weird crashes (scan DBI mailing list for admin's<br>
build issues of DBI on AIX/HP-UX ...).<br>
<br>
Maybe share your recipe can help to reproduce the problem<br>
elsewhere and debug locally.<br>
<br>
Cheers<br>
--<br>
Jens Rehsack - <a href="mailto:rehsack@gmail.com">rehsack@gmail.com</a><br>
<br>
<br>--<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.openembedded.org/mailman/listinfo/openembedded-core" rel="noreferrer" target="_blank">http://lists.openembedded.org/mailman/listinfo/openembedded-core</a><br>
<br></blockquote></div>