This is the mail archive of the gdb-testers@sourceware.cygnus.com mailing list for the GDB project. See the GDB home page for more information.


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

Re: 4.17.85 Feedback


>>>>> "Stan" == Stan Shebs <shebs@cygnus.com> writes:
Stan> The pre-release snapshot has been up for a week, and statistics
Stan> show many downloads, but there's been very little feedback so
Stan> far.  
Stan>
Stan> So if you've run into problems that need fixing, please speak up
Stan> now!

Is the tui interface supposed to be usable on all platforms?  

I ask because when I enabled it, a got a conflict between the macro IN
#defined in m2-exp.tab.c the variable of the same name in NetBSD 1.3's
curses.h.  This occurs because defs.h includes the tui headers, which
in turn include curses.h.

A quick hack for this particular problem would be to change m2-exp.y.
However, I would not be surprised if there are other identifiers in
some host/target specific part of GDB that conflict with some other
host's curses.h.  curses.h is namespace unfriendly --- the curses
library was never subjected to the same type of compatibility pressure
that the C libraries, etc. had before ANSI/ISO/POSIX/etc.  Even the
current X/Open XPG4.2 curses standard is all to quiet about namespace
issues.

Perhaps the #include of tui headers needs to be removed from defs.h
and included in the handful of source files that interface with the
tui library.

	--jtc

-- 
J.T. Conklin
RedBack Networks