Confirmed update procedure for a self-hosted armv6hf system

[Ian Lepore]

On Mon, 2016-05-23 at 20:18 -0700, bob prohaska wrote:
> On Mon, May 23, 2016 at 08:49:57PM -0600, Ian Lepore wrote:
> > 
> > Ut oh.  It just popped into my head that the standard freebsd
> > update
> > procedure is to install the kernel, then reboot, then install the
> > world, right?  (I've never once done it that way in 20 years of
> > using
> > freebsd, I always install both at once then reboot.)
> > 
> My custom has been to build world and kernel, install world and
> kernel,
> then reboot promptly. Not sure where I got the  habit from, the idea 
> was to minimize mismatch between kernel and userland. 
> 
> > I think that may be the problem here, if you did it the right way. 
> >  Once you're running on the new kernel, TARGET_ARCH=armv6 isn't a
> > crossbuild anymore because it thinks armv6 is the current arch.  I
> > wonder if it would work for you to try:
> > 
> >   setenv UNAME_P armv6hf
> >   make -j6 installworld TARGET_ARCH=armv6
> > 
> If I did this right, no luck...
> 
> root at www:/usr/src # uname -p
> armv6hf
> root at www:/usr/src # setenv UNAME_P armv6hf
> root at www:/usr/src # make -j6 installworld TARGET_ARCH=armv6 >
> installworld.log &
> [1] 69457
> root at www:/usr/src # 1 error
> 
> [1]    Exit 2                        make -j6 installworld
> TARGET_ARCH=armv6 > installworld.log
> root at www:/usr/src # 
> root at www:/usr/src # tail -n100 installworld.log
> --- installworld ---
> --- __installcheck_UGID ---
> --- installworld ---
> mkdir -p /tmp/install.XvgDGNcF
> progs=$(for prog in [ awk cap_mkdb cat chflags chmod chown cmp cp 
>  date echo egrep find grep id install   ln make mkdir mtree mv
> pwd_mkdb  rm sed services_mkdb sh strip sysctl test true uname wc zic
> tzsetup   makewhatis; do  if progpath=`which $prog`; then  echo
> $progpath;  else  echo "Required tool $prog not found in PATH." >&2; 
>  exit 1;  fi;  done);  libs=$(ldd -f "%o %p\n" -f "%o %p\n" $progs
> 2>/dev/null | sort -u |  while read line; do  $line;  if [ "$2 $3" !=
> "not found" ]; then  echo $2;  else  echo "Required library $1 not
> found." >&2;  exit 1;  fi;  done);  cp $libs $progs
> /tmp/install.XvgDGNcF
> cp -R ${PATH_LOCALE:-"/usr/share/locale"}
> /tmp/install.XvgDGNcF/locale
> cd /usr/src; MAKEOBJDIRPREFIX=/usr/obj  MACHINE_ARCH=armv6 
>  MACHINE=arm  CPUTYPE=
> GROFF_BIN_PATH=/usr/obj/usr/src/tmp/legacy/usr/bin 
>  GROFF_FONT_PATH=/usr/obj/usr/src/tmp/legacy/usr/share/groff_font 
>  GROFF_TMAC_PATH=/usr/obj/usr/src/tmp/legacy/usr/share/tmac CC="cc "
> CXX="c++  "  CPP="cpp "  AS="as" AR="ar" LD="ld" NM=nm 
>  OBJDUMP=objdump OBJCOPY="objcopy"  RANLIB=ranlib STRINGS= 
>  SIZE="size"
> PATH=/usr/obj/usr/src/tmp/legacy/usr/sbin:/usr/obj/usr/src/tmp/legacy
> /usr/bin:/usr/obj/usr/src/tmp/legacy/bin:/usr/obj/usr/src/tmp/usr/sbi
> n:/usr/obj/usr/src/tmp/usr/bin:/tmp/install.XvgDGNcF 
>  LD_LIBRARY_PATH=/tmp/install.XvgDGNcF 
>  PATH_LOCALE=/tmp/install.XvgDGNcF/locale make -f Makefile.inc1   
>  __MAKE_SHELL=/tmp/install.XvgDGNcF/sh reinstall; 
>  MAKEOBJDIRPREFIX=/usr/obj  MACHINE_ARCH=armv6  MACHINE=arm  CPUTYPE=
> GROFF_BIN_PATH=/usr/obj/usr/src/tmp/legacy/usr/bin 
>  GROFF_FONT_PATH=/usr/obj/usr/src/tmp/legacy/usr/share/groff_font 
>  GROFF_TMAC_PATH=/usr/obj/usr/src/tmp/legacy/usr/share/tmac CC="cc "
> CXX="c++  "  CPP="cpp "  AS="as" AR="ar" LD="ld" NM=nm 
>  OBJDUMP=objdump OBJCOPY="objcopy"  RANLIB=ranlib STRINGS= 
>  SIZE="size"
> PATH=/usr/obj/usr/src/tmp/legacy/usr/sbin:/usr/obj/usr/src/tmp/legacy
> /usr/bin:/usr/obj/usr/src/tmp/legacy/bin:/usr/obj/usr/src/tmp/usr/sbi
> n:/usr/obj/usr/src/tmp/usr/bin:/tmp/install.XvgDGNcF 
>  LD_LIBRARY_PATH=/tmp/install.XvgDGNcF 
>  PATH_LOCALE=/tmp/install.XvgDGNcF/locale rm -rf
> /tmp/install.XvgDGNcF
> sh: cc: not found
> make[2]: "/usr/src/share/mk/bsd.compiler.mk" line 142: Unable to
> determine compiler type for CC=cc .  Consider setting COMPILER_TYPE.
> *** [installworld] Error code 1
> 
> Perhaps it's time to install the new kernel and reboot.
> 
> I also have two kernels that report uname -p as armv6, one pre-hf and
> one post-hf. Worth a try?
> 
> > As long as the build system uses uname (not sysctl
> > hw.machine_arch),
> > that should fool it enough to do the proper install.
> 
> Thanks for all your help!
> 
> bob prohaska
> 

Hmm, apparently my guess at the problem was wrong (although I think it
is true that the standard freebsd update sequence of booting to the new
kernel before installing world will fail here).

I'm out of ideas right now.

-- Ian