[v2] kernel-module-split: Allow for external conf files

Some recipes might provide conf files produced during build phase or
simply tracked in the VCS instead of generating them with Yocto.
In such cases those conf files wouldn't be assigned to correct packages.
With this change, if user wants to generate a conf file they still can,
but not generating them won't prevent assigning the file to proper
package given the file exists.

Signed-off-by: Michal Sieron <michalwsieron@gmail.com>
---

I am quite busy lately so didn't have time to go back to this topic.

This v2 differs from v1 by added comments and post install hook is now
treated the same way as (CONF)FILES:* variable modification.

 .../kernel-module-split.bbclass               | 27 +++++++++++++------
 1 file changed, 19 insertions(+), 8 deletions(-)

On Tue, Apr 29, 2025 at 9:21 AM Michal Sieron <michalwsieron@gmail.com>
wrote:

> Some recipes might provide conf files produced during build phase or
> simply tracked in the VCS instead of generating them with Yocto.
> In such cases those conf files wouldn't be assigned to correct packages.
> With this change, if user wants to generate a conf file they still can,
> but not generating them won't prevent assigning the file to proper
> package given the file exists.
>
> Signed-off-by: Michal Sieron <michalwsieron@gmail.com>
> ---
>
> I am quite busy lately so didn't have time to go back to this topic.
>
> This v2 differs from v1 by added comments and post install hook is now
> treated the same way as (CONF)FILES:* variable modification.
>

And sorry for my very slow reply.

This version of the patch was hidden below the original, so I missed it
when it arrived.



>  .../kernel-module-split.bbclass               | 27 +++++++++++++------
>  1 file changed, 19 insertions(+), 8 deletions(-)
>
> diff --git a/meta/classes-recipe/kernel-module-split.bbclass
> b/meta/classes-recipe/kernel-module-split.bbclass
> index 9487365eb7..a2d81f18e2 100644
> --- a/meta/classes-recipe/kernel-module-split.bbclass
> +++ b/meta/classes-recipe/kernel-module-split.bbclass
> @@ -99,9 +99,12 @@ python split_kernel_module_packages () {
>              bb.warn("module_autoload_%s was replaced by
> KERNEL_MODULE_AUTOLOAD for cases where basename == module name, please drop
> it" % basename)
>          if autoload and basename not in autoloadlist:
>              bb.warn("module_autoload_%s is defined but '%s' isn't
> included in KERNEL_MODULE_AUTOLOAD, please add it there" % (basename,
> basename))
> +
> +        # The .conf file can either be installed by a recipe or generated
> from module_autoload_*
> +        conf = '%s/%s.conf' % (d.getVar('modulesloaddir'), basename)
> +        name = '%s%s' % (dvar, conf)
> +        # If module name is in KERNEL_MODULE_AUTOLOAD, then generate the
> .conf file and write to `name`.
>          if basename in autoloadlist:
> -            conf = '%s/%s.conf' % (d.getVar('modulesloaddir'), basename)
> -            name = '%s%s' % (dvar, conf)
>              os.makedirs(os.path.dirname(name), exist_ok=True)
>              with open(name, 'w') as f:
>                  if autoload:
> @@ -109,6 +112,9 @@ python split_kernel_module_packages () {
>                          f.write('%s\n' % m)
>                  else:
>                      f.write('%s\n' % basename)
> +        # If the .conf file exits, then add it to FILES:* and CONFFILES:*
> and add postinstall hook.
>

Minor typo s/exits/exists/



> +        # It doesn't matter if it was generated from module_autoload_* or
> installed by the recipe.
> +        if os.path.exists(name):
>              conf2append = ' %s' % conf
>              d.appendVar('FILES:%s' % pkg, conf2append)
>              d.appendVar('CONFFILES:%s' % pkg, conf2append)
> @@ -121,18 +127,23 @@ python split_kernel_module_packages () {
>          # Write out any modconf fragment
>          modconflist = (d.getVar("KERNEL_MODULE_PROBECONF") or "").split()
>          modconf = d.getVar('module_conf_%s' % basename)
> +
> +        # The .conf file can either be installed by a recipe or generated
> from module_conf_*
> +        conf = '%s/%s.conf' % (d.getVar('modprobedir'), basename)
> +        name = '%s%s' % (dvar, conf)
> +        # If module name is in KERNEL_MODULE_PROBECONF, then generate the
> .conf file and write to `name`.
>          if modconf and basename in modconflist:
> -            conf = '%s/%s.conf' % (d.getVar('modprobedir'), basename)
> -            name = '%s%s' % (dvar, conf)
>              os.makedirs(os.path.dirname(name), exist_ok=True)
>              with open(name, 'w') as f:
>                  f.write("%s\n" % modconf)
> -            conf2append = ' %s' % conf
> -            d.appendVar('FILES:%s' % pkg, conf2append)
> -            d.appendVar('CONFFILES:%s' % pkg, conf2append)
> -
>          elif modconf:
>              bb.error("Please ensure module %s is listed in
> KERNEL_MODULE_PROBECONF since module_conf_%s is set" % (basename, basename))
> +        # If the .conf file exits, then add it to FILES:* and CONFFILES:*.
> +        # It doesn't matter if it was generated from module_conf_* or
> installed by the recipe.
> +        if os.path.exists(name):
> +            conf2append = ' %s' % conf
> +            d.appendVar('FILES:%s' % pkg, conf2append)
> +            d.appendVar('CONFFILES:%s' % pkg, conf2append)
>
>
This version looks good to me.

I would still be nice to have a test for this, but the code paths are
sufficiently
similar now that I don't see anything that could break.

Bruce



>          if "description" in vals:
>              old_desc = d.getVar('DESCRIPTION:' + pkg) or ""
> --
> 2.49.0
>
>