From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH v2 00/30] Removal of BR2_GCC_TARGET_TUNE and other arch/ updates
Date: Thu, 6 Nov 2014 08:24:32 +0100 [thread overview]
Message-ID: <20141106082432.11646f24@free-electrons.com> (raw)
In-Reply-To: <1413923237-2861-1-git-send-email-thomas.petazzoni@free-electrons.com>
Hello,
Ping about this series? The first 6 patches have been applied, but not
the remainder. Would it be possible to apply them for the 2014.11
release, in order to solve the conflicting cflags used on ARM, and
reported by many users?
Thanks,
Thomas
On Tue, 21 Oct 2014 22:26:47 +0200, Thomas Petazzoni wrote:
> Hello,
>
> As discussed during the Buildroot Developers meeting, this patch
> series implement a few changes to the handling of the
> BR2_GCC_TARGET_{ARCH,CPU,TUNE} options.
>
> In sequence, this patch series:
>
> - PATCH 1 introduces a number of ARM architecture blind options:
> BR2_ARM_CPU_ARMV4, BR2_ARM_CPU_ARMV5, BR2_ARM_CPU_ARMV6,
> BR2_ARM_CPU_ARMV7A. These options will allow to simplify testing
> against an ARM architecture (instead of using each individual ARM
> core), and also allow to remove tests against BR2_GCC_TARGET_ARCH
> on ARM, since we later get rid of all values for this variable on
> ARM as part of a later patch in this series.
>
> - PATCH 2 to 19 make use of those new blind options to replace a
> number of dependencies/conditions on individual ARM cores by
> dependencies/conditions on ARM architecture levels.
>
> - PATCH 20 fixes the dhcpcd package as it uses the
> BR2_GCC_TARGET_ARCH value, which may not always exists on certain
> architectures.
>
> - PATCH 21 and 22 remove the BR2_GCC_TARGET_ARCH values from the ARM
> architecture, since they are causing conflicts with the
> BR2_GCC_TARGET_CPU values. We now only specify the -mcpu value
> (through BR2_GCC_TARGET_CPU), and let gcc derive the corresponding
> ARM architecture. This fixes a long standing bug report where
> Buildroot would pass conflicting -march and -mcpu values for
> certain versions of gcc.
>
> - PATCH 23 prepares ffmpeg for the removal of BR2_GCC_TARGET_TUNE by
> using BR2_GCC_TARGET_ARCH or BR2_GCC_TARGET_CPU instead.
>
> - PATCH 24 to 27 modifies the architecture definitions to no longer
> use BR2_GCC_TARGET_TUNE, and instead use BR2_GCC_TARGET_CPU or
> BR2_GCC_TARGET_ARCH as appropriate.
>
> - PATCH 28 to 30 finally get rid of BR2_GCC_TARGET_TUNE entirely
> since it is no longer used.
>
> Changes since v1:
>
> * Instead of having multiple BR2_ARM_CPU_ARMV{4,4T,5TE,etc.} per
> architecture level, use only one. All the follow-up patches have
> been fixed up following this.
>
> * Add a patch to take into account the bug noticed by Arnout about
> VFP enabling in the ffmpeg package.
>
> * Remove the two patches that have been merged.
>
> * Add Reviewed-by tags from Yann on a certain number of patches, as
> he gave them on the list.
>
> Note that I haven't done any modification of the Thumb / Thumb-2
> handling suggested by Yann. This is orthogonal to this series, and can
> be done in a follow-up series.
>
> Best regards,
>
> Thomas
>
> Thomas Petazzoni (30):
> arch/arm: add blind options to know the ARM architecture
> ffmpeg: use the new BR2_ARM_CPU_ARM* options
> ffmpeg: fix ARM condition for --{enable,disable}-vfp
> gst-ffmpeg: use the new BR2_ARM_CPU_ARM* options
> gst-ffmpeg: allow usage of ARMv6 optimizations on ARMv7 platforms
> gst1-libav: use the new BR2_ARM_CPU_ARM* options
> gst1-libav: allow usage of ARMv6 optimizations on ARMv7 platforms
> mplayer: use BR2_ARM_CPU_ARM* options
> libplayer: use BR2_ARM_CPU_ARM* option
> nettle: use the BR2_ARM_CPU_ARM* options
> nodejs: use BR2_ARM_CPU_ARM* options
> qt5: use the BR2_ARM_CPU_ARM* options
> speex: use BR2_ARM_CPU_ARM* options
> vo-aacenc: use the BR2_ARM_CPU_ARM* options
> webkit: use BR2_ARM_CPU_ARM* options
> valgrind: use the BR2_ARM_CPU_ARM* options
> x264: use the BR2_ARM_CPU_ARMV7A option
> libnspr: use the BR2_ARM_CPU_ARMV7A option
> toolchain-external: use BR2_ARM_CPU_ARM* options
> dhcpcd: do not pass --target option
> arch/arm: do not distinguish revisions of ARM1136JF-S
> arch/arm: remove BR2_GCC_TARGET_ARCH definitions on ARM
> ffmpeg: use arch and cpu instead of tune
> arch/x86: get rid of BR2_GCC_TARGET_TUNE
> arch/sparc: get rid of BR2_GCC_TARGET_TUNE
> arch/powerpc: get rid of BR2_GCC_TARGET_TUNE
> arch/m68k: get rid of BR2_GCC_TARGET_TUNE
> gcc: do not use BR2_GCC_TARGET_TUNE anymore
> toolchain-external: do not use BR2_GCC_TARGET_TUNE anymore
> arch: remove the BR2_GCC_TARGET_TUNE option
>
> arch/Config.in | 3 --
> arch/Config.in.arm | 61 +++++++++++-----------
> arch/Config.in.m68k | 8 ---
> arch/Config.in.powerpc | 2 +-
> arch/Config.in.sparc | 4 +-
> arch/Config.in.x86 | 32 +-----------
> package/dhcpcd/dhcpcd.mk | 1 -
> package/ffmpeg/ffmpeg.mk | 13 +++--
> package/gcc/gcc.mk | 3 --
> package/gstreamer/gst-ffmpeg/gst-ffmpeg.mk | 4 +-
> package/gstreamer1/gst1-libav/gst1-libav.mk | 4 +-
> package/libnspr/libnspr.mk | 2 +-
> package/libplayer/Config.in | 3 +-
> package/mplayer/Config.in | 3 +-
> package/mplayer/mplayer.mk | 4 +-
> package/nettle/nettle.mk | 2 +-
> package/nodejs/Config.in | 4 +-
> package/qt5/Config.in | 2 +-
> package/speex/Config.in | 2 +-
> package/valgrind/Config.in | 7 ++-
> package/valgrind/valgrind.mk | 4 +-
> package/vo-aacenc/vo-aacenc.mk | 2 +-
> package/webkit/Config.in | 3 +-
> package/x264/x264.mk | 2 +-
> toolchain/toolchain-external/Config.in | 14 +++--
> .../toolchain-external/ext-toolchain-wrapper.c | 12 ++---
> toolchain/toolchain-external/toolchain-external.mk | 5 --
> 27 files changed, 74 insertions(+), 132 deletions(-)
>
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
next prev parent reply other threads:[~2014-11-06 7:24 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-21 20:26 [Buildroot] [PATCH v2 00/30] Removal of BR2_GCC_TARGET_TUNE and other arch/ updates Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 01/30] arch/arm: add blind options to know the ARM architecture Thomas Petazzoni
2014-10-21 21:06 ` Yann E. MORIN
2014-10-21 20:26 ` [Buildroot] [PATCH v2 02/30] ffmpeg: use the new BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-10-21 21:09 ` Yann E. MORIN
2014-10-21 20:26 ` [Buildroot] [PATCH v2 03/30] ffmpeg: fix ARM condition for --{enable, disable}-vfp Thomas Petazzoni
2014-10-21 21:15 ` Yann E. MORIN
2014-10-21 20:26 ` [Buildroot] [PATCH v2 04/30] gst-ffmpeg: use the new BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-10-21 21:18 ` Yann E. MORIN
2014-10-22 7:34 ` Thomas Petazzoni
2014-10-22 16:14 ` Arnout Vandecappelle
2014-10-23 7:49 ` Thomas Petazzoni
2014-10-23 8:27 ` Arnout Vandecappelle
2014-10-21 20:26 ` [Buildroot] [PATCH v2 05/30] gst-ffmpeg: allow usage of ARMv6 optimizations on ARMv7 platforms Thomas Petazzoni
2014-10-21 21:24 ` Yann E. MORIN
2014-10-21 20:26 ` [Buildroot] [PATCH v2 06/30] gst1-libav: use the new BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 07/30] gst1-libav: allow usage of ARMv6 optimizations on ARMv7 platforms Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 08/30] mplayer: use BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-11-06 22:33 ` Peter Korsgaard
2014-10-21 20:26 ` [Buildroot] [PATCH v2 09/30] libplayer: use BR2_ARM_CPU_ARM* option Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 10/30] nettle: use the BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 11/30] nodejs: use " Thomas Petazzoni
2014-10-21 20:26 ` [Buildroot] [PATCH v2 12/30] qt5: use the " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 13/30] speex: use " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 14/30] vo-aacenc: use the " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 15/30] webkit: use " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 16/30] valgrind: use the " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 17/30] x264: use the BR2_ARM_CPU_ARMV7A option Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 18/30] libnspr: " Thomas Petazzoni
2014-11-06 22:57 ` Peter Korsgaard
2014-10-21 20:27 ` [Buildroot] [PATCH v2 19/30] toolchain-external: use BR2_ARM_CPU_ARM* options Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 20/30] dhcpcd: do not pass --target option Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 21/30] arch/arm: do not distinguish revisions of ARM1136JF-S Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 22/30] arch/arm: remove BR2_GCC_TARGET_ARCH definitions on ARM Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 23/30] ffmpeg: use arch and cpu instead of tune Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 24/30] arch/x86: get rid of BR2_GCC_TARGET_TUNE Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 25/30] arch/sparc: " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 26/30] arch/powerpc: " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 27/30] arch/m68k: " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 28/30] gcc: do not use BR2_GCC_TARGET_TUNE anymore Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 29/30] toolchain-external: " Thomas Petazzoni
2014-10-21 20:27 ` [Buildroot] [PATCH v2 30/30] arch: remove the BR2_GCC_TARGET_TUNE option Thomas Petazzoni
2014-10-25 10:46 ` [Buildroot] [PATCH v2 00/30] Removal of BR2_GCC_TARGET_TUNE and other arch/ updates Thomas Petazzoni
2014-11-06 7:24 ` Thomas Petazzoni [this message]
2014-11-06 22:06 ` Peter Korsgaard
2014-11-06 23:20 ` Peter Korsgaard
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=20141106082432.11646f24@free-electrons.com \
--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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox