All of lore.kernel.org
 help / color / mirror / Atom feed
From: Waldemar Brodkorb <wbx@openadk.org>
To: buildroot@busybox.net
Subject: [Buildroot] [autobuild.buildroot.net] Build results for 2015-02-13
Date: Sun, 15 Mar 2015 22:36:45 +0100	[thread overview]
Message-ID: <20150315213644.GC28815@waldemar-brodkorb.de> (raw)
In-Reply-To: <20150315142822.2e018813@free-electrons.com>

Hi Thomas,
Thomas Petazzoni wrote,

> Dear Waldemar Brodkorb,
> 
> On Sun, 15 Feb 2015 18:30:10 +0100, Waldemar Brodkorb wrote:
> 
> > I just build a sh4 toolchain with uClibc 0.9.33.2. No issues.
> 
> [...]
> 
> > What is the exact problem?
> 
> I just tried again today, on Buildroot 2015.02 official, with the
> following defconfig:
> 
> BR2_sh=y
> BR2_TOOLCHAIN_BUILDROOT_LARGEFILE=y
> BR2_TOOLCHAIN_BUILDROOT_INET_IPV6=y
> BR2_TOOLCHAIN_BUILDROOT_INET_RPC=y
> BR2_TOOLCHAIN_BUILDROOT_LOCALE=y
> BR2_TOOLCHAIN_BUILDROOT_CXX=y
> BR2_INIT_NONE=y
> # BR2_PACKAGE_BUSYBOX is not set
> # BR2_TARGET_ROOTFS_TAR is not set
> 
> The toolchain build fails with:
> if [ ! -e /opt/br-sh4-full-2015.02/usr/bin/sh4-buildroot-linux-uclibc-cc ]; then ln -snf sh4-buildroot-linux-uclibc-gcc /opt/br-sh4-full-2015.02/usr/bin/sh4-buildroot-linux-uclibc-cc; fi
> if [ ! -e /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/bin/cc ]; then ln -snf gcc /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/bin/cc; fi
> (cd /opt/br-sh4-full-2015.02/usr/bin; for i in sh4-buildroot-linux-uclibc-*; do ln -snf $i sh4-linux${i##sh4-buildroot-linux-uclibc}; done)
> cp -dpf /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/libgcc_s* /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/sysroot/lib/
> cp: cannot stat `/opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/libgcc_s*': No such file or directory
> make: [/opt/toolchain-build/build/host-gcc-final-4.8.4/.stamp_host_installed] Error 1 (ignored)
> cp -dpf /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/libgcc_s* /opt/toolchain-build/target/lib/
> cp: cannot stat `/opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/libgcc_s*': No such file or directory
> make: [/opt/toolchain-build/build/host-gcc-final-4.8.4/.stamp_host_installed] Error 1 (ignored)
> mkdir -p /opt/toolchain-build/target/usr/lib
> for i in  libstdc++ ; do cp -dpf /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/${i}.a /opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/sysroot/usr/lib/ ; done
> cp: cannot stat `/opt/br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib*/libstdc++.a': No such file or directory
> make: *** [/opt/toolchain-build/build/host-gcc-final-4.8.4/.stamp_host_installed] Error 1
> make: Leaving directory `/home/test/toolchains/build/buildroot'
> 
> It's because libgcc_s is in a directory named after the multilib
> variant, since sh4 toolchain builds are multilib, as far as I
> understand.
> 
> Any idea?

If you want to create a sh4 toolchain, which is usable to compile a
Linux Kernel you need to enable multilib support so that the
toolchain can be used to compile fpu/non-fpu code.
If you use make qemu_sh4_r2d_defconfig, then following option is
enabled:
BR2_EXTRA_GCC_CONFIG_OPTIONS="--with-multilib-list=m4,m4-nofpu"
This works fine and the result boots up fine in Qemu.

Your defconfig does not generate a multilib toolchain.
The compiler still generates
br-sh4-full-2015.02/usr/sh4-buildroot-linux-uclibc/lib/!m4/libgcc_s.so
which is not matched by your globbing in
package/gcc/gcc-final/gcc-final.mk ->
$(HOST_DIR)/usr/$(GNU_TARGET_NAME)/lib*/libgcc_s*

The sh4 stuff was already discussed last year by you:
http://lists.busybox.net/pipermail/buildroot/2014-March/091385.html

The patch will work fine.

I think defaulting to multilib toolchain for sh4 would be a good
thing. We do not need the nofpu libgcc on the target.

best regards
 Waldemar

  reply	other threads:[~2015-03-15 21:36 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-02-14  7:30 [Buildroot] [autobuild.buildroot.net] Build results for 2015-02-13 Thomas Petazzoni
2015-02-14 17:25 ` Fabio Porcedda
2015-02-14 17:49   ` Thomas Petazzoni
2015-02-14 21:42 ` Thomas Petazzoni
2015-02-15 12:32   ` Romain Naour
2015-02-15 13:25   ` Vicente Olivert Riera
2015-02-15 17:30   ` Waldemar Brodkorb
2015-03-15 13:28     ` Thomas Petazzoni
2015-03-15 21:36       ` Waldemar Brodkorb [this message]
2015-03-15 21:45         ` Thomas Petazzoni

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=20150315213644.GC28815@waldemar-brodkorb.de \
    --to=wbx@openadk.org \
    --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.