[lsb-discuss] what is the current libchk status in LSB 4.0 and 4.1?

[Jeff Licquia]

On 05/20/2013 08:54 PM, Benjamin De Kosnik wrote:
> 
>> The warnings you're seeing come from C++ virtual function changes; for
>> example, where a subclass implements a virtual function where it
>> didn't before.  These are perfectly legal, but libchk doesn't
>> currently handle them well.  That was the question I brought up on
>> the conference call; we should handle these cases better.
> 
> It's a general accounting error with abstract virtuals in instantiated
> templates in libchk. FWICT, This has been an issue for years on ppc64.
> 
> There's an additional issue in that one of these libstdc++ exports is
> not necessary (but harmless): the ctype<char> vtable explicitly elides
> the abstract base, but it is explicitly instantiated and exported just
> the same. Because it's never used, it doesn't really matter what the
> values are.
> 
> Please waive this Jeff.

That sounds like the right way forward here.  Objections?

Just so I understand, it's your belief that ctype<char>'s abstract base
should explicitly not be a part of the LSB?  If we waive this, that
implies that there's a bug to fix, and I want to get the bug right.

Thanks for putting some time on this.
-- 
Jeff Licquia
The Linux Foundation
+1 (317) 915-7441
licquia at linuxfoundation.org

Linux Foundation Events Schedule:  events.linuxfoundation.org
Linux Foundation Training Schedule: training.linuxfoundation.org