Date: Tue, 13 Aug 2013 11:31:52 -0600 From: Ian Lepore <ian@FreeBSD.org> To: Glen Barber <gjb@FreeBSD.org> Cc: svn-src-head@FreeBSD.org, svn-src-all@FreeBSD.org, src-committers@FreeBSD.org Subject: Re: svn commit: r254217 - head/sys/conf Message-ID: <1376415112.3320.229.camel@revolution.hippie.lan> In-Reply-To: <20130813171302.GQ2432@glenbarber.us> References: <201308111357.r7BDvF7H065442@svn.freebsd.org> <1376407389.3320.209.camel@revolution.hippie.lan> <20130813153222.GL2432@glenbarber.us> <1376413337.3320.224.camel@revolution.hippie.lan> <20130813171302.GQ2432@glenbarber.us>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 2013-08-13 at 13:13 -0400, Glen Barber wrote: > On Tue, Aug 13, 2013 at 11:02:17AM -0600, Ian Lepore wrote: > > It fails for me during buildworld (I haven't gotten to buildkernel yet). > > This is a cross-build of armv6 from i386, if that matters. > > > > It looks like the build is failing in sys/boot/fdt. > > ===> sys/boot/fdt (installincludes) > 1 error > *** Error code 2 > 1 error > *** Error code 2 > > > When I use "make -s" the build fails immediately in "building an up to > > date make" because it somehow has $DESTDIR/usr/lib/libc.a as a > > dependency for bmake/unit-tests, and clearly that isn't going to exist > > before installworld has created it (when DESTDIR is other than / which > > is always is for me). For some reason I can't even begin to figure out, > > I don't get that error with the -s left off. > > > > Odd. '-s' should suppress build output (similar to adding "@" to the > beginning of the Makefile command lines). > > > Hrm, I just noticed I get the same bmake build error at the beginning > > with and without the -s, the difference is that without -s it ignores > > the error and keeps building. Seems like a different bug. > > > > Yes, that is strange. > > > So, leaving the make -s problems aside, a regular "make buildworld" > > results in http://people.freebsd.org/~ian/make.buildworld.log (the > > failure is around line 6178). > > > > I don't know what's different between make with and without -s but > > perhaps it somehow toggles this problem. > > > > BTW, I'm seeing this problem at r254281, and did not see it at r253965. > > > > At this point, I am just ready to revert all of the changes that allow > use of svnliteversion from base to determine the svn version. In my > buildkernel, newvers.sh does correctly expand '$(realpath ${0})' to > $SYSBASE/sys/conf/newvers.sh. I do not know why it is not the same for > the cross build. Although, I could just call svn{,lite}version on > $SYSBASE/Makefile. > > root@nucleus:/usr/src # make -j8 -s buildkernel > -------------------------------------------------------------- > >>> Kernel build for NUCLEUS started on Tue Aug 13 13:09:53 EDT 2013 > -------------------------------------------------------------- > ===> NUCLEUS > -------------------------------------------------------------- > >>> stage 1: configuring the kernel > -------------------------------------------------------------- > Kernel build directory is /usr/obj/usr/src/sys/NUCLEUS > Don't forget to do ``make cleandepend && make depend'' > -------------------------------------------------------------- > >>> stage 2.1: cleaning up the object tree > -------------------------------------------------------------- > -------------------------------------------------------------- > >>> stage 2.2: rebuilding the object tree > -------------------------------------------------------------- > -------------------------------------------------------------- > >>> stage 2.3: build tools > -------------------------------------------------------------- > -------------------------------------------------------------- > >>> stage 3.1: making dependencies > -------------------------------------------------------------- > -------------------------------------------------------------- > >>> stage 3.2: building everything > -------------------------------------------------------------- > + [ ! -z '' ] > + [ -x /usr/bin/svnversion ] > + [ ! -z '' ] > + [ -x /usr/local/bin/svnversion ] > + [ -z ] > + realpath /usr/src/sys/conf/newvers.sh > + /usr/local/bin/svnversion /usr/src/sys/conf/newvers.sh > + [ 0 -eq 0 ] > + svnversion=/usr/local/bin/svnversion > + break > + [ -z /usr/local/bin/svnversion ] > + [ -x /usr/bin/p4 ] > + [ -x /usr/local/bin/p4 ] > + [ -d /usr/src/sys/conf/../../.git ] > + [ -n /usr/local/bin/svnversion ] > + cd /usr/src/sys/conf/.. > + /usr/local/bin/svnversion > + svn=253815:254284M > + svn=' r253815:254284M' > + [ -n '' ] > + [ -n '' ] > + cat > + echo 159 > linking kernel.debug > text data bss dec hex filename > 9082608 1113780 1035592 11231980 ab62ec kernel.debug > -------------------------------------------------------------- > >>> Kernel build for NUCLEUS completed on Tue Aug 13 13:11:49 EDT 2013 > -------------------------------------------------------------- > > Glen > I'm not even sure what $0 *should* expand to in a script that was sourced in. The manpage doesn't say anything meaningful to me about it. Maybe it's a kind of "indeterminate results" thing which is pretty much what we're seeing. I was thinking that $(realpath ${PARAMFILE}) might be a good way to tap dance around the problem, but PARAMFILE can be set by default from $SYSDIR and it's not clear to me that that'll always be right either. -- Ian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1376415112.3320.229.camel>