This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: LSB patch to gcc 3.3 tree

You Wrote Michael Matz
> Hi again,
>
> On Tue, 29 Jun 2004, Michael Matz wrote:
>
>> > >> +  _ZNSt23__codecvt_abstract_base*;
>> > >> +
>> > >>  } GLIBCPP_3.2.3;
>> >
>> > This should be in a new section, GLIBCPP_3.2.5
>> >
>> > Please just use the 3.4. style instead of these blanket exports.
>>
>> Okay.
>
> Hmm, now looking at 3.4 and head.  Both contain:
> ----------------
>     # std::__codecvt_abstract_base*
>     _ZNStSt23__codecvt_abstract_base*;
> ----------------
>
> I.e. first it's also a glob (in difference to for instance the typeinfo
> struct for this class), and second it's wrong.  Namely the '_ZNStSt23...'
> contains 'St' twice where the symbols are named '_ZNSt23...' .
>
> So, question is, is >=3.4 indeed broken, and can I nevertheless use those
> globs for std::__codecvt_abstract_base members?

Yes.  See Benjmain's message

    http://gcc.gnu.org/ml/libstdc++/2004-06/msg00360.html

Thanks,
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]