[PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Khem Raj]

The module can not be compiled without patching kernel 5.10
This helps meta-ti to live in multi-BSP envs

Signed-off-by: Khem Raj <raj.khem@gmail.com>
---
 meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb | 1 +
 1 file changed, 1 insertion(+)

diff --git a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
index 4051c34..a4dd38c 100644
--- a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
+++ b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
@@ -27,6 +27,7 @@ do_install:append() {
 }
 
 COMPATIBLE_HOST = "(arm|aarch64).*-linux"
+COMPATIBLE_HOST:ti-soc = "null"
 
 # Kernel modules currently RDEPEND on the kernel, which is troublesome when you want to put a
 # kernel module into a initramfs without pulling the kernel into the initramfs, which would be
-- 
2.36.0

Re: [PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Jon Mason]

On Sun, May 01, 2022 at 11:47:31AM -0700, Khem Raj wrote:
> The module can not be compiled without patching kernel 5.10
> This helps meta-ti to live in multi-BSP envs
> 
> Signed-off-by: Khem Raj <raj.khem@gmail.com>
> ---
>  meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> index 4051c34..a4dd38c 100644
> --- a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> +++ b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> @@ -27,6 +27,7 @@ do_install:append() {
>  }
>  
>  COMPATIBLE_HOST = "(arm|aarch64).*-linux"
> +COMPATIBLE_HOST:ti-soc = "null"

Shouldn't this live in meta-ti?

>  
>  # Kernel modules currently RDEPEND on the kernel, which is troublesome when you want to put a
>  # kernel module into a initramfs without pulling the kernel into the initramfs, which would be
> -- 
> 2.36.0
> 
> 

Re: [PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Khem Raj]

On Tue, May 3, 2022 at 6:06 AM Jon Mason <jdmason@kudzu.us> wrote:
>
> On Sun, May 01, 2022 at 11:47:31AM -0700, Khem Raj wrote:
> > The module can not be compiled without patching kernel 5.10
> > This helps meta-ti to live in multi-BSP envs
> >
> > Signed-off-by: Khem Raj <raj.khem@gmail.com>
> > ---
> >  meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > index 4051c34..a4dd38c 100644
> > --- a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > +++ b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > @@ -27,6 +27,7 @@ do_install:append() {
> >  }
> >
> >  COMPATIBLE_HOST = "(arm|aarch64).*-linux"
> > +COMPATIBLE_HOST:ti-soc = "null"
>
> Shouldn't this live in meta-ti?

yeah perhaps better there.

>
> >
> >  # Kernel modules currently RDEPEND on the kernel, which is troublesome when you want to put a
> >  # kernel module into a initramfs without pulling the kernel into the initramfs, which would be
> > --
> > 2.36.0
> >
> >

Re: [meta-arm] [PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Denys Dmytriyenko]

On Tue, May 03, 2022 at 07:48:20AM -0700, Khem Raj wrote:
> On Tue, May 3, 2022 at 6:06 AM Jon Mason <jdmason@kudzu.us> wrote:
> >
> > On Sun, May 01, 2022 at 11:47:31AM -0700, Khem Raj wrote:
> > > The module can not be compiled without patching kernel 5.10
> > > This helps meta-ti to live in multi-BSP envs
> > >
> > > Signed-off-by: Khem Raj <raj.khem@gmail.com>
> > > ---
> > >  meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > index 4051c34..a4dd38c 100644
> > > --- a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > +++ b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > @@ -27,6 +27,7 @@ do_install:append() {
> > >  }
> > >
> > >  COMPATIBLE_HOST = "(arm|aarch64).*-linux"
> > > +COMPATIBLE_HOST:ti-soc = "null"
> >
> > Shouldn't this live in meta-ti?
> 
> yeah perhaps better there.

If the kernel requires patching to use this module from meta-arm, shouldn't 
meta-arm set a bit stricter COMPATIBLE_HOST? How are other BSPs supposed to 
work along with meta-arm?


> > >  # Kernel modules currently RDEPEND on the kernel, which is troublesome when you want to put a
> > >  # kernel module into a initramfs without pulling the kernel into the initramfs, which would be
> > > --
> > > 2.36.0
> > >
> > >

-- 
Regards,
Denys Dmytriyenko <denis@denix.org>
PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
Fingerprint: 25FC E4A5 8A72 2F69 1186  6D76 4209 0272 9A92 C964

Re: [meta-arm] [PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Khem Raj]

On Tue, May 3, 2022 at 8:21 AM Denys Dmytriyenko <denis@denix.org> wrote:
>
> On Tue, May 03, 2022 at 07:48:20AM -0700, Khem Raj wrote:
> > On Tue, May 3, 2022 at 6:06 AM Jon Mason <jdmason@kudzu.us> wrote:
> > >
> > > On Sun, May 01, 2022 at 11:47:31AM -0700, Khem Raj wrote:
> > > > The module can not be compiled without patching kernel 5.10
> > > > This helps meta-ti to live in multi-BSP envs
> > > >
> > > > Signed-off-by: Khem Raj <raj.khem@gmail.com>
> > > > ---
> > > >  meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb | 1 +
> > > >  1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > > index 4051c34..a4dd38c 100644
> > > > --- a/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > > +++ b/meta-arm/recipes-kernel/ffa-debugfs/ffa-debugfs-mod_2.1.0.bb
> > > > @@ -27,6 +27,7 @@ do_install:append() {
> > > >  }
> > > >
> > > >  COMPATIBLE_HOST = "(arm|aarch64).*-linux"
> > > > +COMPATIBLE_HOST:ti-soc = "null"
> > >
> > > Shouldn't this live in meta-ti?
> >
> > yeah perhaps better there.
>
> If the kernel requires patching to use this module from meta-arm, shouldn't
> meta-arm set a bit stricter COMPATIBLE_HOST? How are other BSPs supposed to
> work along with meta-arm?

I think meta-arm is a base dependency of meta-ti not other way around,
so in this case I think its better to fix
in meta-ti, since meta-ti is not using the kernel that meta-arm tests
so this layer has no control over things downstream
layers do.

>
>
> > > >  # Kernel modules currently RDEPEND on the kernel, which is troublesome when you want to put a
> > > >  # kernel module into a initramfs without pulling the kernel into the initramfs, which would be
> > > > --
> > > > 2.36.0
> > > >
> > > >
>
> --
> Regards,
> Denys Dmytriyenko <denis@denix.org>
> PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
> Fingerprint: 25FC E4A5 8A72 2F69 1186  6D76 4209 0272 9A92 C964

[PATCH] ffa-debugfs-mod: Do not build for TI SOCs

[Khem Raj]

The module can not be compiled without patching kernel 5.10
This helps meta-ti to live in multi-BSP envs

Signed-off-by: Khem Raj <raj.khem@gmail.com>
---
 .../linux/ffa-debugfs/ffa-debugfs-mod_%.bbappend           | 7 +++++++
 1 file changed, 7 insertions(+)
 create mode 100644 meta-ti-bsp/recipes-kernel/linux/ffa-debugfs/ffa-debugfs-mod_%.bbappend

diff --git a/meta-ti-bsp/recipes-kernel/linux/ffa-debugfs/ffa-debugfs-mod_%.bbappend b/meta-ti-bsp/recipes-kernel/linux/ffa-debugfs/ffa-debugfs-mod_%.bbappend
new file mode 100644
index 00000000..78b509c1
--- /dev/null
+++ b/meta-ti-bsp/recipes-kernel/linux/ffa-debugfs/ffa-debugfs-mod_%.bbappend
@@ -0,0 +1,7 @@
+FILESEXTRAPATHS:prepend := "${THISDIR}/${PN}:"
+
+# Added my meta-arm, this needs ceetain backports in 5.10
+# Disable until kernel upgrades to newer than 5.10 where it
+# is already upstreamed
+# Revisit when kernel is upgraded to 5.15 or newer e.g.
+COMPATIBLE_HOST:ti-soc = "null"
-- 
2.36.0