From: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
To: buildroot@busybox.net
Subject: [Buildroot] [arc-buildroot] [autobuild.buildroot.net] arc build results for 2016-06-02
Date: Fri, 3 Jun 2016 11:50:54 +0000 [thread overview]
Message-ID: <1464954612.8682.36.camel@synopsys.com> (raw)
In-Reply-To: <20160603063024.061EE102E94@stock.ovh.net>
Hi Thomas,
On Fri, 2016-06-03 at 08:30 +0200, Thomas Petazzoni wrote:
> Those results are limited to the arc architecture.
>
> Build statistics for 2016-06-02
> ===============================
>
> ????????success : 0??
> ???????failures : 15?
> ???????timeouts : 0??
> ??????????TOTAL : 15?
>
> Classification of failures by reason
> ====================================
>
> ?????host-binutils-arc-2016.03 | 8?
> ??????????binutils-arc-2016.03 | 2?
Here we're hitting what seems to be an upstream binutils
issue. If there's no prebuilt documentation (which is the case for
sources from git) binutils will fail to build documentation
if there's no "makeinfo" installed on a build host.
Note our previous binutils (from "arc-2015.12" release) even though being taken from
git repo did have prebuilt docs in place.
Now what happens with arc-2016.03 binutils or upstream sources from git...
Script "missing" returns code "127" if "makeinfo" "command not found" for real,
see?https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=blob;f=missing#l207
and its caller which is "bfd.info" in "bfd/doc/Makefile" exits with that same code,
see?https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=blob;f=bfd/doc/Makefile.in#l487
which "make" interprets as an error code, thus we see:
---------------------->8-------------------
make[4]: *** [bfd.info] Error 127
---------------------->8-------------------
On my machine I do have "makeinfo" installed as well as many real people have too that's why
nobody had this problem before.
So there're 2 solutions (one of which is a work-around which postpones the problem):
?1. Install "makeinfo" on your build server. That's obviously just a work-around.
?2. Fix bfd/doc/Makefile so it gracefully handles expected error codes returned by
? ? "missing". That looks like a proper fix.
Any thoughts are more than welcome. And indeed we'll try to fix this ASAP because nothing could
be built for target if host-binutils fails to build itself.
BTW our hack in BR when we set "ac_cv_prog_MAKEINFO=missing" to disable documentation building
turned out to be a bit naive. That's because "missing" script will anyways try to run target
tool (in our case "makeinfo").
> ?????????????bluez_utils-4.101 | 1?
> ? ? ? ? ? ?libmemcached-1.0.18 | 1?
That's binutils' segfault - we'll be looking at it.
> ? ? ? ? ? ? ? ? e2fsprogs-1.43 | 3?
This seems to be non-ARC specific problem.
-Alexey
parent reply other threads:[~2016-06-03 11:50 UTC|newest]
Thread overview: expand[flat|nested] mbox.gz Atom feed
[parent not found: <20160603063024.061EE102E94@stock.ovh.net>]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1464954612.8682.36.camel@synopsys.com \
--to=alexey.brodkin@synopsys.com \
--cc=buildroot@busybox.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox