[cfe-dev] Is boost python tested?

[Sebastian Redl]

On 07.10.2010 11:57, Riccardo Rossi wrote:
> well...
>        i was asking if the thing was tested with clang. We have been
> using Boost-Python during years and we know it works.
>
> Furthermore our code was tested using many different compilers and on
> different OSs. The segfault only appears with clang, and this is a bit
> surprising to us, so i was asking if anyone is using routinely clang
> together with boost python. It would not be that strange that some
> errors are still in there since it is a relatively short time boost is
> supported...
>    
The Boost regression tests show no failures for the Clang builder that 
don't also occur with other compilers. On the other hand, Boost.Python 
was one of the last libraries to work, and it's rather complex, so there 
might well be issues left for more complicated usages.

It would be ideal if you managed to reduce a minimal test case for the 
failures you're seeing.

Sebastian