From: Gustavo Zacarias <gustavo@zacarias.com.ar>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH 04/10] binutils: make it a proper package
Date: Wed, 29 Dec 2010 17:41:58 -0300 [thread overview]
Message-ID: <4D1B9D16.1080902@zacarias.com.ar> (raw)
In-Reply-To: <874o9xldzj.fsf@macbook.be.48ers.dk>
On 12/28/10 19:38, Peter Korsgaard wrote:
> Seeing this I thought the build would be broken because GCC isn't moved
> yet, but I see you do move it later on in this patch, even if it isn't
> mentioned in the commit message.
Comment added.
> I find this quite complicated. I would prefer something like:
>
> config BR2_PACKAGE_BINUTILS
> bool "binutils"
> depends on BR2_TOOLCHAIN_BUILDROOT
> help
> ..
>
> config BR2_PACKAGE_BINUTILS_FULLINSTALL
> bool "perform full installation"
> depends on BR2_PACKAGE_BINUTILS
> help
> Select this to install all binary utilies and not just libbfd.
>
> And then move the MPC/MPFR/GMP selects under
> BR2_PACKAGE_GCC_TARGET. Also add selects for BR2_PACKAGE_BINUTILS +
> FULLINSTALL.
Easier to define BR2_PACKAGE_BINUTILS that builds all
of binutils and only installs libbfd + BR2_PACKAGE_BINUTILS_TARGET to
install the full version.
> You have trailing spaces here and elsewhere.
Hopefully fixed.
> Gustavo> -
> Gustavo> -# The --without-headers option stopped working with gcc 3.0 and has never been
> Gustavo> -# fixed, so we need to actually have working C library header files prior to
> Gustavo> -# the step or libgcc will not build...
>
> Gustavo> $(GCC_BUILD_DIR1)/.compiled: $(GCC_BUILD_DIR1)/.configured
> Gustavo> - # gcc >= 4.3.0 have to also build all-target-libgcc
>
> Why are you removing those comments? We still support gcc 4.2.x
Because these are in STAGE1 (DIR1) where it's completely irrelevant /
doesn't apply and just amounts to misinformation.
gcc-initial doesn't use nor install headers since the extra gcc build
stage was added. Same scenario for libgcc.
They're leftover comments. That's the reason i kept the ones in
gcc-intermediate.
> Gustavo> ifeq ($(BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE),y)
> Gustavo> $(GCC_CONF_ENV) $(MAKE) -C $(GCC_BUILD_DIR1) all-gcc
> Gustavo> else
> Gustavo> @@ -276,22 +262,16 @@ gcc_initial-dirclean:
> Gustavo> #############################################################
> Gustavo> GCC_BUILD_DIR2:=$(TOOLCHAIN_DIR)/gcc-$(GCC_VERSION)-intermediate
>
> Gustavo> -
> Gustavo> # The --without-headers option stopped working with gcc 3.0 and has never been
> Gustavo> # fixed, so we need to actually have working C library header files prior to
> Gustavo> # the step or libgcc will not build...
>
> Gustavo> $(GCC_BUILD_DIR2)/.configured: $(GCC_DIR)/.patched
> Gustavo> mkdir -p $(GCC_BUILD_DIR2)
> Gustavo> - -rmdir $(STAGING_DIR)/usr/$(REAL_GNU_TARGET_NAME)/lib
> Gustavo> - mkdir -p $(STAGING_DIR)/lib
> Gustavo> - ln -snf ../../lib $(STAGING_DIR)/usr/$(REAL_GNU_TARGET_NAME)/lib
> Gustavo> - $(if $(BR2_ARCH_IS_64),mkdir -p $(STAGING_DIR)/lib64)
> Gustavo> - $(if $(BR2_ARCH_IS_64),ln -snf ../../lib64 $(STAGING_DIR)/usr/$(REAL_GNU_TARGET_NAME)/lib64)
>
> Why remove this ARCH_IS_64 stuff?
This is from commit 3c77bab2eeace3ee675bd745ca335fa3dd1630bb which does
a symlink trick to make libstdc++ available. It's simpler to just copy
libstdc++ (and libgcj/libgcc while at it) to the sysroot which is fixed
in patch #5 from my series.
next prev parent reply other threads:[~2010-12-29 20:41 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-12-28 19:10 [Buildroot] [PATCH 00/10] Toolchain rework, take 3 Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 01/10] sstrip: make it a proper package Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 02/10] toolchain: move sysroot to host dir Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 03/10] uclibc: move tools " Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 04/10] binutils: make it a proper package Gustavo Zacarias
2010-12-28 22:38 ` Peter Korsgaard
2010-12-29 20:41 ` Gustavo Zacarias [this message]
2010-12-29 20:58 ` Peter Korsgaard
2011-01-03 9:14 ` Thomas Petazzoni
2011-01-03 10:13 ` Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 05/10] gcc: install copies of libgcc, libstdc++ and libgcj to the sysroot too Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 06/10] gdb: make it a proper package Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 07/10] package: drop sparc64 bits Gustavo Zacarias
2010-12-29 22:15 ` Peter Korsgaard
2010-12-28 19:10 ` [Buildroot] [PATCH 08/10] oprofile: use new libbfd option Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 09/10] binutils: needs libintl when locales are enabled Gustavo Zacarias
2010-12-28 19:10 ` [Buildroot] [PATCH 10/10] toolchain: drop BR2_CROSS_TOOLCHAIN_TARGET_UTILS option Gustavo Zacarias
2010-12-29 22:15 ` Peter Korsgaard
2010-12-29 8:17 ` [Buildroot] [PATCH 00/10] Toolchain rework, take 3 Peter Korsgaard
2010-12-29 20:43 ` Gustavo Zacarias
2010-12-29 21:04 ` Peter Korsgaard
2010-12-29 22:02 ` Gustavo Zacarias
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=4D1B9D16.1080902@zacarias.com.ar \
--to=gustavo@zacarias.com.ar \
--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