You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@stdcxx.apache.org by Martin Sebor <se...@roguewave.com> on 2006/12/04 19:04:03 UTC

XLC++ 8.0 build results

...for the weekend are reported as DATA (incomplete
data) even though the library failed to build.

Also, the log shows that the build script continued
to try to build examples and other pieces after the
library failed to build. Didn't we use to short-circuit
the rest of the build when this happened? If not, we
probably should since there is little point in trying
to test anything when there's no library to link with.

Andrew, could you please look into it when you have
a chance?

Thanks
Martin

Re: XLC++ 8.0 build results

Posted by Martin Sebor <se...@roguewave.com>.
Andrew Black wrote:

> Greetings Martin
> 
> This behavior change was caused by the changes made to the glue script 
> that translates paths into variable names.  Resolving this was fairly 
> simple, and has been checked into version control.  When the nightly 
> archive is rebuilt tonight, the change will be included.

I see. The old "pipeline exit status" problem.

Thanks
Martin

> 
> --Andrew Black
> 
> Martin Sebor wrote:
> 
>> ...for the weekend are reported as DATA (incomplete
>> data) even though the library failed to build.
>>
>> Also, the log shows that the build script continued
>> to try to build examples and other pieces after the
>> library failed to build. Didn't we use to short-circuit
>> the rest of the build when this happened? If not, we
>> probably should since there is little point in trying
>> to test anything when there's no library to link with.
>>
>> Andrew, could you please look into it when you have
>> a chance?
>>
>> Thanks
>> Martin

Re: XLC++ 8.0 build results

Posted by Andrew Black <ab...@roguewave.com>.
Greetings Martin

This behavior change was caused by the changes made to the glue script 
that translates paths into variable names.  Resolving this was fairly 
simple, and has been checked into version control.  When the nightly 
archive is rebuilt tonight, the change will be included.

--Andrew Black

Martin Sebor wrote:
> ...for the weekend are reported as DATA (incomplete
> data) even though the library failed to build.
> 
> Also, the log shows that the build script continued
> to try to build examples and other pieces after the
> library failed to build. Didn't we use to short-circuit
> the rest of the build when this happened? If not, we
> probably should since there is little point in trying
> to test anything when there's no library to link with.
> 
> Andrew, could you please look into it when you have
> a chance?
> 
> Thanks
> Martin