[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Problem with configure falsely detecting presence ofsystemroutines(w
Kelly F. Hickel
RE: Problem with configure falsely detecting presence ofsystemroutines(while porting to NSK)
Fri, 3 Oct 2003 13:56:40 -0500
> -----Original Message-----
> From: Derek Robert Price [mailto:address@hidden
> Sent: Friday, October 03, 2003 1:47 PM
> To: Kelly F. Hickel
> Cc: address@hidden
> Subject: Re: Problem with configure falsely detecting presence
> ofsystemroutines(while porting to NSK)
> Then I think the immediate workaround for your problem then is to plan
> on configuring all packages with `./configure CC=/usr/local/bin/cc
> CPPFLAGS=-D_XOPEN_SOURCE_EXTENDED=1' (alternatively, configure should
> find your `cc' if it is the first `cc' in your PATH). The long term
> for the `cc' bit is to work out a way for configure to determine that
> Floss is available and needed on your system and to use it with the
> Autoconf folks (<address@hidden> to ask questions,
> <address@hidden> for bug reports and patch submissions) and get
> the changes incorporated there.
[Kelly Hickel] I fully understand and support the fact that this doesn't
have to be in the autoconf, I have no problem with telling people they
have to get Floss and then do the CC above. The problem is that this
still generates a broken cvs, in that it can't handle files over 52k.
The patch that I'm proposing to submit only deals with the 52k issue,
not fixing autoconf. I have previously submitted patches for detecting
the NSK/OSS environment, I do not believe they've been accepted, but
they haven't been bounced either. They've vanished into the bit bucket.
> It is possible that the Autoconf folks will tell you that it is not
> their business that your OS comes with a broken linker, especially
> the solution is so simple as putting a working linker (the Floss `cc'
> script) in your PATH. If there were options that a script could
> determine that needed to be added to LDFLAGS on your system to get the
> correct results then the story might be different. Of course it is
> possible that they might decide that all of the Floss functionality is
> useful on other systems or that there would be a particular interest
> Tandem there, or even that the problem has already been solved and you
> only need to update your Autoconf version.
[Kelly Hickel] I suspect that getting a fully functioning AutoConf on
NSK/OSS is beyond my time availability right now, even if I could
interest them in it.
> What is _XOPEN_SOURCE_EXTENDED supposed to do and how could it be
> detected that it is needed? Is this something that is needed for
> compiling on your system in general or only for particular parts of
[Kelly Hickel] As far as the _XOPEN_SOURCE_EXTENDED stuff goes, my quick
googling shows that it's present on several systems (including hpux,
z-os, tru-64). On the NSK it was needed to resolve some socket calls.
It may be needed for other things, but it was definitely needed for