Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 5 Aug 2018 01:11:38 +0200
From:      Ulrich Grey <usenet@ulrich-grey.de>
To:        Mark Millard <marklmi@yahoo.com>
Cc:        Mark Millard via freebsd-arm <freebsd-arm@freebsd.org>, Marek Zarychta <zarychtam@plan-b.pwste.edu.pl>
Subject:   Re: Booting PINE64-LTS does not work
Message-ID:  <20180805011138.45fe945f2c3f79eb6c2fbedd@ulrich-grey.de>
In-Reply-To: <87BBB736-D555-4D55-9B36-85CB228CC9DC@yahoo.com>
References:  <20180730202020.472bbf8a1b785a12699703ed@ulrich-grey.de> <20180730203159.ec4e72ee641f6a13e05174f2@bidouilliste.com> <20180731014832.GH2884@funkthat.com> <20180731211925.420d4068a8447c8dcbe8c0f0@bidouilliste.com> <20180804203554.1143060d0a43f9b7777cba2e@ulrich-grey.de> <0efaadab-80b1-43bc-4e3b-28c71ffa0136@plan-b.pwste.edu.pl> <87BBB736-D555-4D55-9B36-85CB228CC9DC@yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 4 Aug 2018 14:22:17 -0700
Mark Millard via freebsd-arm <freebsd-arm@freebsd.org> wrote:

> On 2018-Aug-4, at 1:24 PM, Marek Zarychta <zarychtam at plan-b.pwste.edu.pl> wrote:
> 
> > W dniu 04.08.2018 o 20:35, Ulrich Grey pisze:
> >> Thank you for the hint and the commit.
> >> I successfully built an image (12.0-CURRENT r337040) for my board:
> >> PINE A64 LTS Version V 1.2 2018-01-29
> >> 
> >> After booting successfully I realized that one USB 2.0 host on the board does not
> >> work.
> >> 
> >> I tried to build FreeBSD 12.0 CURRENT r337282 buildworld on the board, using an
> >> external USB HD, connected via a D-Link Hub (DUB-H7).
> >> After some hours, the build failed:
> >> ##
> >> ===> lib/clang/liblldb (all)
> >> <jemalloc>: jemalloc_arena.c:647: Failed assertion: "nstime_compare(&decay->epoch,
> >> &time) <= 0" c++: error: unable to execute command: Abort trap (core dumped)
> >> c++: error: clang frontend command failed due to signal (use -v to see invocation)
> >> FreeBSD clang version 6.0.1 (tags/RELEASE_601/final 335540) (based on LLVM 6.0.1)
> >> Target: aarch64-unknown-freebsd12.0
> >> Thread model: posix
> >> InstalledDir: /usr/bin
> >> c++: note: diagnostic msg: PLEASE submit a bug report to
> >> https://bugs.freebsd.org/submit/ and include the crash backtrace, preprocessed
> >> source, and associated run script. c++: note: diagnostic msg: ********************
> >> 
> >> PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
> >> Preprocessed source(s) and associated run script(s) are located at:
> >> c++: note: diagnostic msg: /tmp/SBVariablesOptions-d0770a.cpp
> >> c++: note: diagnostic msg: /tmp/SBVariablesOptions-d0770a.sh
> >> c++: note: diagnostic msg: 
> >> 
> >> ********************
> >> --- API/SBVariablesOptions.o ---
> >> *** [API/SBVariablesOptions.o] Error code 254
> >> ##
> >> 
> >> On the console, I got this:
> >> 
> >> swap_pager: indefinite wait buffer: bufobj: 0, blkno: 693018, size: 4096
> >> pid 47471 (c++), uid 0: exited on signal 6 (core dumped)
> >> 
> >> Please see the build log and the diagnostic messages etc. here:
> >> 
> >> http://ulrich-grey.de/dl/FreeBSD/pineA64ltsErrMsg20180804.tgz
> >> --------------------------------------------------------------
> > 
> > 
> > Since I have subscribed to this list I see peoples' overwhelming desire
> > to build FreeBSD world on small boards consuming 1-3W of power designed
> > primarily for running lightweight embedded systems. I guess all the
> > builders are running make world with -j n >= 4 on these boards with
> > 1-2Gb of memory (these processors have usually 4 cores). The build
> > process is running for some time, let's say 24 hours than it fails due
> > to exhaustion of memory/swap or other issues. Running it for the next
> > time will probably lead to the same or different issue.
> > 
> > 
> > The waste of time is very frustrating, so people complain here. On the
> > other hand, the build process takes about 1-2 hours on not so modern
> > amd64 architecture hardware. So why not cross-build all on the faster
> > amd64 machine?
> > 
> >> . . .
> 
> I do both ways, rebuilding after installing a cross-built version.
> It is a form of checking for how well things are operating.
> 
> But I also normally restrict myself to contexts with 512 MiBytes or
> more per "cpu" put to use for buildworld buildkernel. So, for -j4, 
> 2 GiByte board (or more) for buildworld buildkernel tests. UFS,
> not ZFS. Swap partition, not swap file. "Disks" that are observed
> to be well behaved.
> 
> In the report:
> 
> <jemalloc>: jemalloc_arena.c:647: Failed assertion: "nstime_compare(&decay->epoch,
> &time) <= 0"
> 
> it is unlikely to be a out of RAM or swap/paging space issue
> and indicates that the environment is not stable (likely from
> time jumping backwards). So it is an example of discovering
> a problem. (Possibly
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=229644 .)

It wasn't an out of ram/swap issue. It was the 3rd try, I have eliminated such
options (to my best knowledge). It isn't the first time I build world or large ports on an
arm board. In the 1970ies I have learned to write programs using the COBOL language. I
know how to find and document errors. But I have never worked in IT (EDV in German)
realm, so I am a layman. Some errors remain hidden during cross compiling and arise
during a native build. My aim is to scrap my PC and use an arm board running FreeBSD (if
it is mature) for desktop/surfing. I don't like Linux.

> It may be fairly common that the folks playing with rpi3's and the like,
> including building FreeBSD on such, are not spending to invest in "modern
> amd64 architecture hardware". And, for some, if they did, it might mean
> then backing off on the use of rpi3's (or whatever).
> 
> 
> FreeBSD does not really target "self hosted builds" for these
> smaller-device contexts and so it can take effort to find a
> combination that works for self hosting (for a time).
> 
> ===
> Mark Millard
> marklmi at yahoo.com
> ( dsl-only.net went
> away in early 2018-Mar)
> 
> _______________________________________________
> freebsd-arm@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-arm
> To unsubscribe, send any mail to "freebsd-arm-unsubscribe@freebsd.org"



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20180805011138.45fe945f2c3f79eb6c2fbedd>