This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ 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: target-specific contents for libstdc++: how to? [with actual patch!]



On 25/08/2005, at 3:00 PM, Paolo Carlini wrote:


Hi,


It looks like other similar compatibility stuff goes in
src/compatibility.cc.  Is that right?  More generally, is this patch
right?  (The ugly bit is at the end.)  If not right, how should this
be done?


Benjamin and Jakub introduced compatibility.cc to fix an annoying
problem and in my opinion are the most qualified people to answer your
question.

My personal point of view is that that file has been introduced for a
*very specific* reason (maybe not fully transparent from the name), that
is, exploit the alias extension in order to export some new symbols,
already exported (by mistake) as @GLIBCXX_3.4, also, correctly, as
@GLIBCXX_3.4.5. I think we all want to keep this generic file as small
as possible, in preparation for the eventual complete elimination.
Therefore, I would definitely encourage you to localize the
darwin-specific additions in the darwin-specific files.

I did think of that (several times), but automake requires that you tell it about all the source files you plan to use in advance, so there's no way to just say "on Darwin, please compile this file too", or at least I haven't thought of one. If it was possible, that would be what I would like to do; have a config/.../darwin/eprintf.cc file.


I could, though, very easily make a 'src/darwin.cc' file, or similar.

Attachment: smime.p7s
Description: S/MIME cryptographic signature


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