From mboxrd@z Thu Jan 1 00:00:00 1970 From: Randolph Chung To: Eli Zaretskii Cc: gdb-patches@sources.redhat.com Subject: Re: [patch/rfa] configure check for curses library Date: Sun, 16 May 2004 05:17:00 -0000 Message-id: <20040516051748.GG566@tausq.org> References: <20040515180123.GP566@tausq.org> X-SW-Source: 2004-05/msg00456.html > Does the TUI code in GDB actually use wattr_on or any symbols that > reference wattr_on? If not, we could defeat curses implementations > that don't have wattr_on, but are otherwise good enough for what GDB > needs. tausq@hiauly3:~/src/build/gdb$ nm tui-wingeneral.o |grep wattr U wattr_off U wattr_on this comes from box_win(). there's a call to wattron that is getting translated into wattr_on(). actually, the problem seems to be: on this particular HPUX system where i'm doing the gdb build, there are both ncurses headers and native HPUX curses headers. The include search path for GCC is such that the build picks up the ncurses curses.h header which does the wattron() -> wattr_on() translation. :-( so i guess my fix is not really correct; we should make configure use the correct header. But how? randolph -- Randolph Chung Debian GNU/Linux Developer, hppa/ia64 ports http://www.tausq.org/