This is the mail archive of the
cygwin-developers@sources.redhat.com
mailing list for the Cygwin project.
Re: current winsup/cygwin/*.h files policy
- To: Christopher Faylor <cygwin-developers at cygwin dot com>
- Subject: Re: current winsup/cygwin/*.h files policy
- From: Egor Duda <deo at logos-m dot ru>
- Date: Wed, 6 Dec 2000 22:51:40 +0300
- Organization: DEO
- References: <15228862171.20001206222335@logos-m.ru><20001206142931.H8257@redhat.com>
- Reply-To: Egor Duda <cygwin-developers at cygwin dot com>
Hi!
Wednesday, 06 December, 2000 Christopher Faylor cgf@redhat.com wrote:
>> if i understand correctly, *.h files in winsup/cygwin/ were
>>redesigned some time ago from single monolithic winsup.h to a
>>bunch of small *.h files which don't use
>>#ifndef _FOO_H_
>>#define _FOO_H_
>>...
>>#endif
>>construct, and should be included in *.cc files in correct
>>order to satisfy all dependencies. there're, however 4 *.h files
>>left, which include other "*.h" files -- heap.h, pinfo.h,
>>shared_info.h and winsup.h. was it intentional?
CF> Yes. I'm not trying to be overly dogmatic about this. If it makes sense
CF> to include perthread.h in thread.h, then go ahead and do that. Just make
CF> sure that you remove any duplication from any source files.
ok.
>>btw, does usage of c++'s multiple inheritance allowed in cygwin?
CF> I'm not sure what you mean. Aren't some fhandler classes doing this already?
no. 'grep "class.*:.*,.*" *' shows only lines in ChangeLogs. i was
meaning that cygwin used to avoid some g++ features (-fno-exceptions
-fno-rtti), so i wanted to make sure that multiple inheritance is ok
in current gcc.
Egor. mailto:deo@logos-m.ru ICQ 5165414 FidoNet 2:5020/496.19