All of lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot@busybox.net
Subject: [Buildroot] [autobuild.buildroot.net] Build results for 2014-03-28
Date: Mon, 31 Mar 2014 12:32:33 +0200	[thread overview]
Message-ID: <20140331123233.256f294c@skate> (raw)
In-Reply-To: <39A54937CC95F24AA2F794E2D2B66B1356C9AB43@de02wembxa.internal.synopsys.com>

Dear Anton Kolesov,

On Mon, 31 Mar 2014 10:24:47 +0000, Anton Kolesov wrote:

> > ARC compiler error. Anton?
> > 
> >   CC       magick/magick_libMagickCore_6_Q16_la-cipher.lo
> > magick/cipher.c: In function 'IncrementCipherNonce.constprop.2':
> > magick/cipher.c:535:1: internal compiler error: Segmentation fault
> >  }
> >  ^
> 
> [AK:] This one is fixed in our repository. We should have a new release on a two weeks that will resolve many of the compiler failures. I will bump versions in Buildroot after that.

Great, thanks for the feedback!


> > Anton ?
> > 
> > /home/test/test/2/output/host/usr/lib/gcc/arc-buildroot-linux-
> > uclibc/4.8.0/../../../../arc-buildroot-linux-uclibc/bin/ld: bayrad.so: hidden
> > symbol `__fini_array_end' isn't defined
> > /home/test/test/2/output/host/usr/lib/gcc/arc-buildroot-linux-
> > uclibc/4.8.0/../../../../arc-buildroot-linux-uclibc/bin/ld: final link failed: Bad
> > value
> > collect2: error: ld returned 1 exit status
> > 
> 
> [AK:] It looks like that this is caused by "-static -shared". In general it cannot work for ARC out of the box - because
> it is "-static" then shared library will be linked with libc.a (not
> libc.so), however because it is a shared library, then all of the code
> inside should be compiled -fPIC. But I believe that is not true for ARC
> tool chain: libc.a by default is compiled without -fPIC for the
> performance reasons. So the final shared object will have potions of
> position dependent code - that's not gonna work, previously that was
> causing an error at runtime in dynamic loader, due to invalid
> relocations, but since new release linker will abort on attempt to link
> non-PIC code into shared library. That means that for "-static -shared"
> to work libc.a should be compiled with -fPIC as well. In this
> particular case it results in some linker failure (maybe because it is
> C++, instead of C?). I've tried to link without -static and it worked
> fine. I presume the error will go away if libc.a will be compiled with
> -fPIC, but I haven't tried this myself yet. 

No, I don't think libc.a should be built with -fPIC. Just leave the ARC
build as it is. The real problem here is that lcdproc attempts to build
a shared library, even if the configuration says BR2_PREFER_STATIC_LIB.
So it should be fixed either by making lcdproc not available when
BR2_PREFER_STATIC_LIB=y, or by fixing lcdproc itself to generate a
shared library when needed.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

  reply	other threads:[~2014-03-31 10:32 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-29  7:30 [Buildroot] [autobuild.buildroot.net] Build results for 2014-03-28 Thomas Petazzoni
2014-03-29 16:54 ` Thomas Petazzoni
2014-03-29 17:06   ` Max Filippov
2014-03-30 15:27     ` Thomas Petazzoni
2014-03-30 22:09       ` Marcelo Gutierrez
2014-03-29 19:16   ` Gustavo Zacarias
2014-03-30  5:11   ` Baruch Siach
2014-04-03 13:13     ` Baruch Siach
2014-03-30 11:59   ` Yegor Yefremov
2014-04-05 12:48     ` Thomas Petazzoni
2014-04-05 14:44       ` Yegor Yefremov
2014-04-05 17:07         ` Thomas Petazzoni
2014-03-31 10:24   ` Anton Kolesov
2014-03-31 10:32     ` Thomas Petazzoni [this message]
2014-03-31 10:42       ` Anton Kolesov

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=20140331123233.256f294c@skate \
    --to=thomas.petazzoni@free-electrons.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.