From mboxrd@z Thu Jan 1 00:00:00 1970 From: Romain Naour Date: Sat, 15 Feb 2014 15:06:29 +0100 Subject: [Buildroot] [PATCH 1/1] ncftp: fix cross-compilation issues In-Reply-To: <20140214095330.5b632ac0@skate> References: <1392342348-28066-1-git-send-email-romain.naour@openwide.fr> <20140214095330.5b632ac0@skate> Message-ID: <52FF7465.7090004@openwide.fr> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hi Thomas, All, Le 14/02/2014 09:53, Thomas Petazzoni a ?crit : > Dear Romain Naour, > > On Fri, 14 Feb 2014 02:45:48 +0100, Romain Naour wrote: >> ncftp is unable to find ncurses library installed by ncurses package. >> So ncftpbookmarks is not build and install fails. >> >> If ncurses is installed on the host machine then ncftp find it >> and build ncftpbookmarks. >> >> To avoid that, we need to remove the cross-compilation test in >> configure script and set ac_cv_prog_cc_cross=yes in ncftp.mk > Can you expand a bit on the relation between ac_cv_prog_cc_cross=yes, > and this story about ncurses? I don't quite see the connection between > the two. Sure, it was (very) late and I was not very clear in my explanations, sorry. So, the problem is that the configure script assumes that it is doing native compilation due to a false result of the cross-compilation test. In this case an additional test is performed and fails if ncurses is not installed on the host machine. (test #line 5893 "configure") This test is skipped for cross-compilation. This has resulted that the ncurses library support is disabled beneath the feet of ncftp's package. Thus ncftpbootmarks is not build and install fails. > That being said, I agree that the cross-compilation test is stupid: it > builds a program, and tries to run it. If it runs, then the configure > script concludes that we're doing native compilation, if it doesn't > run, we're doing cross-compilation. Except that of course when you're > building x86 or x86-64 on x86-64 and both the target and host use > glibc, the program may very well run. > > Now, maybe we could simply patch the configure script to use something > like what all other configure scripts are doing: > > cross_compiling=no > > if test "x$host_alias" != x; then > if test "x$build_alias" = x; then > cross_compiling=maybe > $as_echo "$as_me: WARNING: If you wanted to set the --build type, don't use --host. > If a cross compiler is detected then cross compile mode will be used." >&2 > elif test "x$build_alias" != "x$host_alias"; then > cross_compiling=yes > fi > fi Ok, thanks ! I will rework my patch to do thatbut since we're cross-compiling, should I take into account the "maybe" case ? Best regards, Romain Naour -------------- next part -------------- An HTML attachment was scrubbed... URL: