From mboxrd@z Thu Jan  1 00:00:00 1970
From: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Date: Sun, 20 Sep 2020 15:31:18 +0200
Subject: [Buildroot] [PATCH 1/1] package/dav1d: disable asm on i386
In-Reply-To: <20200920080930.864480-1-bernd.kuhls@t-online.de>
References: <20200920080930.864480-1-bernd.kuhls@t-online.de>
Message-ID: <20200920153118.6d58e6ba@windsurf>
List-Id: <buildroot.busybox.net>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: buildroot@busybox.net

Hello Bernd,

On Sun, 20 Sep 2020 10:09:30 +0200
Bernd Kuhls <bernd.kuhls@t-online.de> wrote:

> Fixes:
> http://autobuild.buildroot.net/results/19e/19ee2bdd4c8d14cc1554092594b3721117ca1005/
> http://autobuild.buildroot.net/results/72e/72e4089dafacd245e781214b72baeb81a1e3d3d7/
> 
> Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
> ---
>  package/dav1d/dav1d.mk | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/package/dav1d/dav1d.mk b/package/dav1d/dav1d.mk
> index 1e84c21b47..62f673b991 100644
> --- a/package/dav1d/dav1d.mk
> +++ b/package/dav1d/dav1d.mk
> @@ -14,12 +14,12 @@ DAV1D_CONF_OPTS = \
>  	-Denable_tests=false \
>  	-Denable_tools=false
>  
> -ifeq ($(BR2_i386)$(BR2_x86_64),y)
> +ifeq ($(BR2_x86_64),y)
>  DAV1D_DEPENDENCIES += host-nasm
>  endif
>  
> -# ARM assembly requires v6+ ISA
> -ifeq ($(BR2_ARM_CPU_ARMV4)$(BR2_ARM_CPU_ARMV5)$(BR2_ARM_CPU_ARMV7M),y)
> +# ARM assembly requires v6+ ISA, asm is broken in i386
> +ifeq ($(BR2_ARM_CPU_ARMV4)$(BR2_ARM_CPU_ARMV5)$(BR2_ARM_CPU_ARMV7M)$(BR2_i386),y)

Are you sure this is the right fix ? Indeed, when I look at:

  http://autobuild.buildroot.net/?reason=dav1d%

This issue started to appear with dav1d 0.7.1 on August 27, 2020.

However dav1d was bumped to 0.7.1 in June 2020, so it cannot be that.
But the nasm compiler was bumped on August 25, 2020, in commit
090a748aae1b9495362000ca6bf953ef8bc8154e. So it rather seems like a
change in nasm broke i386 assembly in dav1d.

And apparently, the x86 build has been fixed upstream, as there is an
issue in the upstream bug tracker asking for a new release with this
fix:

   https://code.videolan.org/videolan/dav1d/-/issues/347

Could you check if the latest dav1d master fixes this ?

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com