On Fri, Jan 23, 2026 at 5:00 AM Mikko Rapeli <[email protected]>
wrote:

> Hi,
>
> On Fri, Jan 09, 2026 at 11:35:01AM -0500, Bruce Ashfield wrote:
> > On Fri, Jan 9, 2026 at 10:35 AM Mikko Rapeli via lists.yoctoproject.org
> > <[email protected]> wrote:
> >
> > > And include sound_x86.cfg on i386 and x86_64 KARCH. These
> > > don't make sense on other platforms like arm/arm64 but cause
> > > warnings there, and the configs get disabled from effective kernel
> > > config due to dependencies.
> > >
> >
> > Another explicit design decision was to avoid arch checks
> > wherever possible. Yes, I created the ability to have conditional
> > parts of the fragments, since it can't be avoided sometimes.
> > So I am firmly to blame for it.
> >
> > But if I put conditionals in, it makes auditing and running the
> > fragments harder outside of a full build.
>
> Sorry I don't understand this.
>

If statements mean that the fragments can only be audited when
you are running within an include path that goes through a BSP
entry point that sets the variables (or some other fragment /
environment).

That's all I'm trying to say, it now means that the single
fragment now needs extra validation as it has multiple code
paths.



>
> Would be ok to add arm64 and arm specific configs to common
> or x86 specific fragments?
>
>
I'm confused. Did I say that ?

Bruce



> I don't think so.
>
> > In this case, it would be better to just go with sound_x86.scc
> > have it include sound.scc if it needs elements from there and
> > leave the if check out.
> >
> > If we audit OEcore and the kernel-cache itself and there are
> > no direct users of "sound.scc" across the architectures, then
> > users can just explicitly grab sound_x86 in their own
> > KERNEL_FEATURES or board .scc file if they know they need
> > the x86 options versus us doing a conditional and making it
> > happen behind the scenes.
>
> But then we add another API. Why not just keep the old sound.scc
> working as it is now and make it handle the arch specific configs
> correctly like this patch does?
>
> On genericarm64 side we have a huge sound fragment which could use
> this same approach, bsp/genericarm64/genericarm64-sound.cfg
>
> Current merge_config.sh and yocto do_kernel_configcheck ignore
> these issues but newer merge_config.sh detects them and shows
> a simple warning without too much clutter. And then these issues
> can easily be fixed once found. The full dependency tree and all
> possible usecases are not covered but at least the ones used
> in oe-core could be. I'm looking at genericarm64 and fixing these.
> I think this improves the quality of these fragments.
>
> Cheers,
>
> -Mikko
>
> > Bruce
> >
> >
> >
> > >
> > > Signed-off-by: Mikko Rapeli <[email protected]>
> > > ---
> > >  cfg/sound.cfg     | 20 --------------------
> > >  cfg/sound.scc     |  4 ++++
> > >  cfg/sound_x86.cfg | 21 +++++++++++++++++++++
> > >  3 files changed, 25 insertions(+), 20 deletions(-)
> > >  create mode 100644 cfg/sound_x86.cfg
> > >
> > > diff --git a/cfg/sound.cfg b/cfg/sound.cfg
> > > index af261bd0c06c..604f1da37e75 100644
> > > --- a/cfg/sound.cfg
> > > +++ b/cfg/sound.cfg
> > > @@ -44,15 +44,6 @@ CONFIG_SND_HDA_CODEC_VIA=m
> > >  CONFIG_SND_SOC_AC97_BUS=y
> > >  CONFIG_SND_SOC_COMPRESS=y
> > >  CONFIG_SND_SOC_AMD_ACP=m
> > > -CONFIG_SND_SST_ATOM_HIFI2_PLATFORM=m
> > > -CONFIG_SND_SOC_INTEL_HASWELL=m
> > > -CONFIG_SND_SOC_INTEL_HASWELL_MACH=m
> > > -CONFIG_SND_SOC_INTEL_BROADWELL_MACH=m
> > > -CONFIG_SND_SOC_INTEL_BYTCR_RT5640_MACH=m
> > > -CONFIG_SND_SOC_INTEL_BYTCR_RT5651_MACH=m
> > > -CONFIG_SND_SOC_INTEL_CHT_BSW_RT5672_MACH=m
> > > -CONFIG_SND_SOC_INTEL_CHT_BSW_RT5645_MACH=m
> > > -CONFIG_SND_SOC_INTEL_CHT_BSW_MAX98090_TI_MACH=m
> > >  CONFIG_SND_SOC_AC97_CODEC=m
> > >  CONFIG_SND_SOC_AK4104=m
> > >  CONFIG_SND_SOC_AK4554=m
> > > @@ -90,13 +81,10 @@ CONFIG_SND_SOC_PCM512x=m
> > >  CONFIG_SND_SOC_PCM512x_I2C=m
> > >  CONFIG_SND_SOC_PCM512x_SPI=m
> > >  CONFIG_SND_SOC_RL6231=m
> > > -CONFIG_SND_SOC_RT298=m
> > >  CONFIG_SND_SOC_RT5616=m
> > >  CONFIG_SND_SOC_RT5631=m
> > >  CONFIG_SND_SOC_RT5640=m
> > >  CONFIG_SND_SOC_RT5645=m
> > > -CONFIG_SND_SOC_RT5651=m
> > > -CONFIG_SND_SOC_RT5670=m
> > >  CONFIG_SND_SOC_SPDIF=m
> > >  CONFIG_SND_SOC_SSM4567=m
> > >  CONFIG_SND_SOC_TLV320AIC31XX=m
> > > @@ -129,12 +117,4 @@ CONFIG_SND_SIMPLE_CARD_UTILS=m
> > >  CONFIG_SND_SIMPLE_CARD=m
> > >  CONFIG_SND_USB=y
> > >  CONFIG_SND_USB_AUDIO=m
> > > -CONFIG_SND_USB_USX2Y=m
> > > -CONFIG_SND_USB_US122L=m
> > >  CONFIG_SND_USB_HIFACE=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_DA7219=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_MAX98357A=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_NAU8825=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_MAX98927=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_SSM4567=m
> > > -CONFIG_SND_SOC_INTEL_AVS_MACH_RT298=m
> > > diff --git a/cfg/sound.scc b/cfg/sound.scc
> > > index fef2dd3f4b52..8a5232a80e92 100644
> > > --- a/cfg/sound.scc
> > > +++ b/cfg/sound.scc
> > > @@ -3,3 +3,7 @@ define KFEATURE_DESCRIPTION "OSS sound support"
> > >  define KFEATURE_COMPATIBILITY board
> > >
> > >  kconf non-hardware sound.cfg
> > > +
> > > +if [ "$KARCH" = "i386" ] || [ "$KARCH" = "x86_64" ]; then
> > > +       kconf hardware sound_x86.cfg
> > > +fi
> > > diff --git a/cfg/sound_x86.cfg b/cfg/sound_x86.cfg
> > > new file mode 100644
> > > index 000000000000..14c9f68b9f19
> > > --- /dev/null
> > > +++ b/cfg/sound_x86.cfg
> > > @@ -0,0 +1,21 @@
> > > +# SPDX-License-Identifier: MIT
> > > +CONFIG_SND_SST_ATOM_HIFI2_PLATFORM=m
> > > +CONFIG_SND_SOC_INTEL_HASWELL=m
> > > +CONFIG_SND_SOC_INTEL_HASWELL_MACH=m
> > > +CONFIG_SND_SOC_INTEL_BROADWELL_MACH=m
> > > +CONFIG_SND_SOC_INTEL_BYTCR_RT5640_MACH=m
> > > +CONFIG_SND_SOC_INTEL_BYTCR_RT5651_MACH=m
> > > +CONFIG_SND_SOC_INTEL_CHT_BSW_RT5672_MACH=m
> > > +CONFIG_SND_SOC_INTEL_CHT_BSW_RT5645_MACH=m
> > > +CONFIG_SND_SOC_INTEL_CHT_BSW_MAX98090_TI_MACH=m
> > > +CONFIG_SND_SOC_RT298=m
> > > +CONFIG_SND_SOC_RT5651=m
> > > +CONFIG_SND_SOC_RT5670=m
> > > +CONFIG_SND_USB_USX2Y=m
> > > +CONFIG_SND_USB_US122L=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_DA7219=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_MAX98357A=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_NAU8825=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_MAX98927=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_SSM4567=m
> > > +CONFIG_SND_SOC_INTEL_AVS_MACH_RT298=m
> > > --
> > > 2.34.1
> > >
> > >
> > > 
> > >
> > >
> >
> > --
> > - Thou shalt not follow the NULL pointer, for chaos and madness await
> thee
> > at its end
> > - "Use the force Harry" - Gandalf, Star Trek II
>


-- 
- Thou shalt not follow the NULL pointer, for chaos and madness await thee
at its end
- "Use the force Harry" - Gandalf, Star Trek II
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#16205): 
https://lists.yoctoproject.org/g/linux-yocto/message/16205
Mute This Topic: https://lists.yoctoproject.org/mt/117176759/21656
Group Owner: [email protected]
Unsubscribe: https://lists.yoctoproject.org/g/linux-yocto/unsub 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to