Re: port-amd64/58982: NetBSD 10.1 install image (amd64) failed to boot on J3455-ITX board (ASRock)

["wgstuken%grimme-reitenspezial.de@localhost via gnats" <gnats-admin%NetBSD.org@localhost>]

The following reply was made to PR port-amd64/58982; it has been noted by GNATS.

From: wgstuken%grimme-reitenspezial.de@localhost
To: Andrius V <vezhlys%gmail.com@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: port-amd64/58982: NetBSD 10.1 install image (amd64) failed to
 boot on J3455-ITX board (ASRock)
Date: Fri, 17 Jan 2025 10:44:17 +0100

 Hi Andrius,
 
 I have reinstalled source tree and had a try with './build.sh tools' 
 after everything else was failing before.
 
 But not even this is working !!!!
 
 Has anybody ever build the sources from the tgz files found in 
 /pub/NetBSD/NetBSD-10.1/source/sets ?
 
 It fails when it starts to setup includes with this error messages
 
 /usr/src/tools/host-mkdep/obj/host-mkdep -s .lo\ .ln\ .d -d -f .depend 
 date.d
 #   compile  date/date.lo
 cc -O  -I/usr/src/tools/compat/obj -I/usr/src/tools/compat/obj/include 
 -I/usr/src/tools/compat -DHAVE_NBTOOL_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 
 -I/usr/src/tools/date/../../bin/date -c -o date.lo.o 
 -Wno-format-nonliteral   /usr/src/tools/date/../../bin/date/date.c
 mv -f date.lo.o date.lo
 #      link  date/date
 cc -O -I/usr/src/tools/compat/obj -I/usr/src/tools/compat/obj/include 
 -I/usr/src/tools/compat -DHAVE_NBTOOL_CONFIG_H=1 -D_FILE_OFFSET_BITS=64 
 -I/usr/src/tools/date/../../bin/date   -o date date.lo 
 -L/usr/src/tools/compat/obj -lnbcompat -lrt -lz
 install-host-mkdep ===> .       (with: install-compat install-binstall 
 install-date)
 install ===> compat
 #   install  /usr/src/obj/tooldir.NetBSD-10.1-i386/lib/libnbcompat.a
 /usr/src/tools/binstall/obj/xinstall -d 
 /usr/src/obj/tooldir.NetBSD-10.1-i386/lib
 /usr/src/tools/binstall/obj/xinstall -c  -r -m 444 libnbcompat.a 
 /usr/src/obj/tooldir.NetBSD-10.1-i386/lib/libnbcompat.a
 #   install  
 /usr/src/obj/tooldir.NetBSD-10.1-i386/include/compat/nbtool_config.h
 /usr/src/tools/binstall/obj/xinstall -c  -r nbtool_config.h 
 /usr/src/obj/tooldir.NetBSD-10.1-i386/include/compat/nbtool_config.h
 cd: can't cd to include
 
 *** Failed target:  includes
 *** Failed command: (cd include && find . -name '*.h' -print | while 
 read f ; do /usr/src/tools/binstall/obj/xinstall -c -r $f 
 /usr/src/obj/tooldir.NetBSD-10.1-i386/include/compat/$f ; done)
 *** Error code 2
 
 Stop.
 nbmake[2]: stopped in /usr/src/tools/compat
 
 *** Failed target:  install-compat
 *** Failed command: _makedirtarget() { dir="$1"; shift; target="$1"; 
 shift; case "${dir}" in /*) this="${dir}/"; real="${dir}" ;; .) this=""; 
 real="/usr/src/tools" ;; *) this="${dir}/"; real="/usr/src/tools/${dir}" 
 ;; esac; show=${this:-.}; echo "${target} ===> ${show%/}${1:+ (with: 
 $@)}"; cd "${real}" && /usr/src/obj/tooldir.NetBSD-10.1-i386/bin/nbmake 
 _THISDIR_="${this}" "$@" ${target}; }; _makedirtarget compat install
 *** Error code 1
 
 Stop.
 nbmake[1]: stopped in /usr/src/tools
 
 *** Failed target:  build_install
 *** Failed command: _makedirtarget() { dir="$1"; shift; target="$1"; 
 shift; case "${dir}" in /*) this="${dir}/"; real="${dir}" ;; .) this=""; 
 real="/usr/src/tools" ;; *) this="${dir}/"; real="/usr/src/tools/${dir}" 
 ;; esac; show=${this:-.}; echo "${target} ===> ${show%/}${1:+ (with: 
 $@)}"; cd "${real}" && /usr/src/obj/tooldir.NetBSD-10.1-i386/bin/nbmake 
 _THISDIR_="${this}" "$@" ${target}; }; _makedirtarget . 
 install-host-mkdep install-compat install-binstall install-date
 *** Error code 1
 
 Stop.
 nbmake: stopped in /usr/src/tools
 
 ERROR: Failed to make build_install in "tools"
 
 *** BUILD ABORTED ***
 
 
 This looks like it happens when it tries to copy includes tree, but for 
 whatever reason the current working directory is not what it should be.
 I located the failing statement inside /usr/src/tools/compat/Makefile - 
 no real surprise.
 I add echo and pwd to see where we are.
 
 pwd command reports /usr/src/tools/compat/obj
 So this makefile seems to be broken ...... (or the upper level that 
 calls this one inside wrong directory)
 
 I checked .CURDIR variable - that one seems to be correct: 
 /usr/src/tools/compat
 When I change the cd statement to
 cd ${.CURDIR}/include
 it seems to work again. This kind of path creation is used at other 
 places too inside this makefile.
 
 I think this should be corrected asap. in the source tree.
 
 
 Regards,
 
 W. Stukenbrock
 
 Am 16.01.2025 21:06, schrieb wgstuken%grimme-reitenspezial.de@localhost:
 > Hi - system run into same error again ...
 > So just down load sources and build them failed on i386 system
 > 
 > I try to find some time to localte the problem - I do not think it
 > makes sense to start cross-build for amd64 in this state ..
 > 
 > Am 16.01.2025 10:30, schrieb wgstuken%grimme-reitenspezial.de@localhost:
 >> Hi again - update
 >> 
 >> I have setup old Celeron (32-Bit only) system in the meanwhile.
 >> Loaded complete source and had a try with 'make release'
 >> It build for long time on this slow CPU and then run into error during
 >> setup of ramdisk-big
 >> sysinstmsgs.fr not found
 >> /usr/src/deistrib/i396/i386/ramdisks/ramdisk-big/sysinstmsgs.fr
 >> 
 >> retry of 'make release' (without clearing everything) starts relinking
 >> lot of things from TOOL part - I forgot to disable the tools-build.
 >> Inside compilers also lot of object files are rebuild .... (looks like
 >> complete gcc build is done again)
 >> So it will take some time until I come to the point where it was
 >> failing before again ...
 >> Will send update if it runs into same error or not.
 >> 
 >> I have not started amd64 build on this system till now - so may need
 >> some more days to get to the point where I can test the 'fix'.
 >> 
 >> sorry
 >> 
 >> W. Stukenbrock
 >> 
 >> Am 14.01.2025 08:29, schrieb wgstuken%grimme-reitenspezial.de@localhost:
 >>> Hi,
 >>> 
 >>> my current plan is to setup an unused system with 10.1 because I have
 >>> always build the system from the sources.
 >>> I have some patches inside that has not reached the main source, but
 >>> some of then I need.
 >>> So I have to integrate them again. (Some kernel related, some 
 >>> user-level)
 >>> 
 >>> But I cannot give an eta for this atm. - lot of 'normal' work.
 >>> I try to do this in parallel now to get forward with this.
 >>> 
 >>> 
 >>> W. Stukenbrock
 >>> 
 >>> Am 13.01.2025 15:30, schrieb Andrius V:
 >>>> On Mon, Jan 13, 2025 at 12:01â?¯PM wgstuken%grimme-reitenspezial.de@localhost 
 >>>> via
 >>>> gnats <gnats-admin%netbsd.org@localhost> wrote:
 >>>>> 
 >>>>> The following reply was made to PR port-amd64/58982; it has been 
 >>>>> noted by GNATS.
 >>>>> 
 >>>>> From: wgstuken%grimme-reitenspezial.de@localhost
 >>>>> To: gnats-bugs%netbsd.org@localhost
 >>>>> Cc: port-amd64-maintainer%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
 >>>>>  netbsd-bugs%netbsd.org@localhost
 >>>>> Subject: Re: port-amd64/58982: NetBSD 10.1 install image (amd64) 
 >>>>> failed to
 >>>>>  boot on J3455-ITX board (ASRock)
 >>>>> Date: Mon, 13 Jan 2025 08:31:35 +0100
 >>>>> 
 >>>>>  Hi,
 >>>> Hi,
 >>>>> 
 >>>>>  I had attached them .....
 >>>> Sorry about that, maybe there were filtered out by mail server, not 
 >>>> sure.
 >>>> 
 >>>>> 
 >>>>>  now as inline copy again:
 >>>>> 
 >>>>>  output of cpuctl identify 0 ...
 >>>>> 
 >>>>>  cpu0: highest basic info 00000015
 >>>>>  cpu0: highest extended info 80000008
 >>>>>  cpu0: "Intel(R) Celeron(R) CPU J3455 @ 1.50GHz"
 >>>>>  cpu0: Intel Atom (Goldmont) (686-class), 1497.60 MHz
 >>>>>  cpu0: TSC freq CPUID 1497600000 Hz
 >>>>>  cpu0: family 0x6 model 0x5c stepping 0xa (id 0x506ca)
 >>>> ...
 >>>>>  cpu family     : 6
 >>>>>  model          : 92
 >>>>>  model name     : Intel(R) Celeron(R) CPU J3455 @ 1.50GHz
 >>>>>  stepping       : 10
 >>>> Thanks again. I can confirm that your CPU has a newer stepping 0xa
 >>>> (10), thus a workaround is not applied. I will prepare the patch to
 >>>> remove the stepping check.
 >>>> It will be possible to test either by building kernel from current
 >>>> source or taking current image from
 >>>> https://nycdn.netbsd.org/pub/NetBSD-daily/HEAD/latest/ around a day
 >>>> after commit.
 >>>> I will inform separately and let's see how it boots afterwards.
 >>>> 
 >>>> Regards,
 >>>> Andrius V