On Tuesday, 13 September 2016 12:04:35 UTC+1, Jeffrey Walton wrote:
>
> [snip]
> > This did not resolve the problem. 
>
> That's a _really_ bad sign. We are quickly approaching the "compiler 
> bug" wall since you just tried to use a straight C/C++ implementation. 
>

Looks that way. :-(
 

> If its not a compiler bug, then it could be a hardware issue. But my 
> money is on a compiler bug. 


So is mine.
 

> The compiler is aligning to 2-bytes and we 
> are not doing anything special that needs the alignment increased. 
>
> One last thing to try... Use the Crypto++ 5.6.2 makefile to build the 
> latest sources. This uses the older CXXFLAGS from 5.6.2; and not the 
> newer ones from 5.6.3. 
>
>
I tried just building 562 and that gives the same problem, same place.
Then I tried copying the 562 makefile to 564 but that gave a build error - 
No rule to make target 'bench.o'.
The compiler flags were: -DNDEBUG -O -g0 -native -template=no%extdef -m64 
-c so I tried the 564 makefile with those options.
This gave me the pipe warning back but I ignored it. This gave the same 
error.

This is very unlikely to be a hardware problem because I see the same core 
dump on our solaris 10 machine as well an on the newer solaris 11 machine 
where I am working now.


 

> I added instructions for using the 5.6.2 makefile to build latest 
> sources at 
> https://cryptopp.com/wiki/Solaris_(Command_Line)#It_used_to_work.21.21.21 
> . 
>
> Jeff 
>

-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to