All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE
@ 2025-12-01 16:07 Jose Javier Rodriguez Barbarin
  2025-12-01 16:07 ` [PATCH v2 1/1] mcb: Add missing modpost build support Jose Javier Rodriguez Barbarin
  2025-12-01 18:04 ` [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Andy Shevchenko
  0 siblings, 2 replies; 5+ messages in thread
From: Jose Javier Rodriguez Barbarin @ 2025-12-01 16:07 UTC (permalink / raw)
  To: andy, krzk, linus.walleij, nathan, nsc, gregkh, bleung,
	heikki.krogerus, abhishekpandit, masahiroy, legion, hughd
  Cc: linux-kbuild, linux-kernel, Jose Javier Rodriguez Barbarin

During the process of update of one of the device drivers that are part of
mcb bus (gpio-menz127.c), Krzysztof from GPIO subsystem asked me
why I was adding new MODULE_ALIAS when I also added the same new
information on MODULE_DEVICE_TABLE.

You can find the messages here:

https://lore.kernel.org/linux-gpio/80a20b13-7c6a-4483-9741-568424f957ef@kernel.org/

After a deeper analysis, I came across that the mcb_table_id defined inside
MODULE_DEVICE_TABLE on all device drivers was being ignored as modpost was
not processing the mcb MODULE_DEVICE_TABLE entries. For this reason, former
contributors were using MODULE_ALIAS for enabling mcb to autoload the
device drivers.

My proposal with these changes is to complete the mcb bus by adding
modpost support for processing mcb MODULE_DEVICE_TABLE.

Once this patch is merged, I will send patches one by one for removing
MODULE_ALIAS from all device drivers as they are no longer needed
(as Andy Shevchenko suggested in v1 review).

---
Changes in v2:

- Place "mcb" before "mei" to follow the order.
- Send changes for mcb drivers in a per-driver basis.

Jose Javier Rodriguez Barbarin (1):
  mcb: Add missing modpost build support

 scripts/mod/devicetable-offsets.c | 3 +++
 scripts/mod/file2alias.c          | 9 +++++++++
 2 files changed, 12 insertions(+)

-- 
2.51.1

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PATCH v2 1/1] mcb: Add missing modpost build support
  2025-12-01 16:07 [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Jose Javier Rodriguez Barbarin
@ 2025-12-01 16:07 ` Jose Javier Rodriguez Barbarin
  2025-12-01 18:02   ` Andy Shevchenko
  2025-12-01 18:04 ` [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Andy Shevchenko
  1 sibling, 1 reply; 5+ messages in thread
From: Jose Javier Rodriguez Barbarin @ 2025-12-01 16:07 UTC (permalink / raw)
  To: andy, krzk, linus.walleij, nathan, nsc, gregkh, bleung,
	heikki.krogerus, abhishekpandit, masahiroy, legion, hughd
  Cc: linux-kbuild, linux-kernel, Jose Javier Rodriguez Barbarin,
	Jorge Sanjuan Garcia

mcb bus is not prepared to autoload client drivers with the data defined on
the drivers' MODULE_DEVICE_TABLE. modpost cannot access to mcb_table_id
inside MODULE_DEVICE_TABLE so the data declared inside is ignored.

Add modpost build support for accessing to the mcb_table_id coded on device
drivers' MODULE_DEVICE_TABLE.

Reviewed-by: Jorge Sanjuan Garcia <dev-jorge.sanjuangarcia@duagon.com>
Signed-off-by: Jose Javier Rodriguez Barbarin <dev-josejavier.rodriguez@duagon.com>
---
 scripts/mod/devicetable-offsets.c | 3 +++
 scripts/mod/file2alias.c          | 9 +++++++++
 2 files changed, 12 insertions(+)

diff --git a/scripts/mod/devicetable-offsets.c b/scripts/mod/devicetable-offsets.c
index d3d00e85edf7..0470ba7c796d 100644
--- a/scripts/mod/devicetable-offsets.c
+++ b/scripts/mod/devicetable-offsets.c
@@ -198,6 +198,9 @@ int main(void)
 	DEVID(cpu_feature);
 	DEVID_FIELD(cpu_feature, feature);
 
+	DEVID(mcb_device_id);
+	DEVID_FIELD(mcb_device_id, device);
+
 	DEVID(mei_cl_device_id);
 	DEVID_FIELD(mei_cl_device_id, name);
 	DEVID_FIELD(mei_cl_device_id, uuid);
diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
index b3333560b95e..4e99393a35f1 100644
--- a/scripts/mod/file2alias.c
+++ b/scripts/mod/file2alias.c
@@ -1110,6 +1110,14 @@ static void do_cpu_entry(struct module *mod, void *symval)
 	module_alias_printf(mod, false, "cpu:type:*:feature:*%04X*", feature);
 }
 
+/* Looks like: mcb:16zN */
+static void do_mcb_entry(struct module *mod, void *symval)
+{
+	DEF_FIELD(symval, mcb_device_id, device);
+
+	module_alias_printf(mod, false, "mcb:16z%03d", device);
+}
+
 /* Looks like: mei:S:uuid:N:* */
 static void do_mei_entry(struct module *mod, void *symval)
 {
@@ -1444,6 +1452,7 @@ static const struct devtable devtable[] = {
 	{"mipscdmm", SIZE_mips_cdmm_device_id, do_mips_cdmm_entry},
 	{"x86cpu", SIZE_x86_cpu_id, do_x86cpu_entry},
 	{"cpu", SIZE_cpu_feature, do_cpu_entry},
+	{"mcb", SIZE_mcb_device_id, do_mcb_entry},
 	{"mei", SIZE_mei_cl_device_id, do_mei_entry},
 	{"rapidio", SIZE_rio_device_id, do_rio_entry},
 	{"ulpi", SIZE_ulpi_device_id, do_ulpi_entry},
-- 
2.51.1

^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH v2 1/1] mcb: Add missing modpost build support
  2025-12-01 16:07 ` [PATCH v2 1/1] mcb: Add missing modpost build support Jose Javier Rodriguez Barbarin
@ 2025-12-01 18:02   ` Andy Shevchenko
  0 siblings, 0 replies; 5+ messages in thread
From: Andy Shevchenko @ 2025-12-01 18:02 UTC (permalink / raw)
  To: Jose Javier Rodriguez Barbarin
  Cc: andy, krzk, linus.walleij, nathan, nsc, gregkh, bleung,
	heikki.krogerus, abhishekpandit, masahiroy, legion, hughd,
	linux-kbuild, linux-kernel, Jorge Sanjuan Garcia

On Mon, Dec 01, 2025 at 05:07:20PM +0100, Jose Javier Rodriguez Barbarin wrote:
> mcb bus is not prepared to autoload client drivers with the data defined on
> the drivers' MODULE_DEVICE_TABLE. modpost cannot access to mcb_table_id
> inside MODULE_DEVICE_TABLE so the data declared inside is ignored.
> 
> Add modpost build support for accessing to the mcb_table_id coded on device
> drivers' MODULE_DEVICE_TABLE.

I believe the idea to add Fixes tag here and make sure it goes to v6.19-rcX.

-- 
With Best Regards,
Andy Shevchenko



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE
  2025-12-01 16:07 [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Jose Javier Rodriguez Barbarin
  2025-12-01 16:07 ` [PATCH v2 1/1] mcb: Add missing modpost build support Jose Javier Rodriguez Barbarin
@ 2025-12-01 18:04 ` Andy Shevchenko
  2025-12-01 18:26   ` Jose Javier Rodriguez Barbarin
  1 sibling, 1 reply; 5+ messages in thread
From: Andy Shevchenko @ 2025-12-01 18:04 UTC (permalink / raw)
  To: Jose Javier Rodriguez Barbarin
  Cc: andy, krzk, linus.walleij, nathan, nsc, gregkh, bleung,
	heikki.krogerus, abhishekpandit, masahiroy, legion, hughd,
	linux-kbuild, linux-kernel

On Mon, Dec 01, 2025 at 05:07:19PM +0100, Jose Javier Rodriguez Barbarin wrote:
> During the process of update of one of the device drivers that are part of
> mcb bus (gpio-menz127.c), Krzysztof from GPIO subsystem asked me
> why I was adding new MODULE_ALIAS when I also added the same new
> information on MODULE_DEVICE_TABLE.
> 
> You can find the messages here:
> 
> https://lore.kernel.org/linux-gpio/80a20b13-7c6a-4483-9741-568424f957ef@kernel.org/
> 
> After a deeper analysis, I came across that the mcb_table_id defined inside
> MODULE_DEVICE_TABLE on all device drivers was being ignored as modpost was
> not processing the mcb MODULE_DEVICE_TABLE entries. For this reason, former
> contributors were using MODULE_ALIAS for enabling mcb to autoload the
> device drivers.
> 
> My proposal with these changes is to complete the mcb bus by adding
> modpost support for processing mcb MODULE_DEVICE_TABLE.
> 
> Once this patch is merged, I will send patches one by one for removing
> MODULE_ALIAS from all device drivers as they are no longer needed
> (as Andy Shevchenko suggested in v1 review).

Not sure if we need a cover letter for a single change, but yes, this
what I think the best approach and code wise it's fine to me:

Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>

(but we still need a Fixes tag I assume).

-- 
With Best Regards,
Andy Shevchenko



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE
  2025-12-01 18:04 ` [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Andy Shevchenko
@ 2025-12-01 18:26   ` Jose Javier Rodriguez Barbarin
  0 siblings, 0 replies; 5+ messages in thread
From: Jose Javier Rodriguez Barbarin @ 2025-12-01 18:26 UTC (permalink / raw)
  To: Andy Shevchenko
  Cc: andy, krzk, linus.walleij, nathan, nsc, gregkh, bleung,
	heikki.krogerus, abhishekpandit, masahiroy, legion, hughd,
	linux-kbuild, linux-kernel

On Mon, Dec 01, 2025 at 08:04:45PM +0200, Andy Shevchenko wrote:
> On Mon, Dec 01, 2025 at 05:07:19PM +0100, Jose Javier Rodriguez Barbarin wrote:
> > During the process of update of one of the device drivers that are part of
> > mcb bus (gpio-menz127.c), Krzysztof from GPIO subsystem asked me
> > why I was adding new MODULE_ALIAS when I also added the same new
> > information on MODULE_DEVICE_TABLE.
> > 
> > You can find the messages here:
> > 
> > https://lore.kernel.org/linux-gpio/80a20b13-7c6a-4483-9741-568424f957ef@kernel.org/
> > 
> > After a deeper analysis, I came across that the mcb_table_id defined inside
> > MODULE_DEVICE_TABLE on all device drivers was being ignored as modpost was
> > not processing the mcb MODULE_DEVICE_TABLE entries. For this reason, former
> > contributors were using MODULE_ALIAS for enabling mcb to autoload the
> > device drivers.
> > 
> > My proposal with these changes is to complete the mcb bus by adding
> > modpost support for processing mcb MODULE_DEVICE_TABLE.
> > 
> > Once this patch is merged, I will send patches one by one for removing
> > MODULE_ALIAS from all device drivers as they are no longer needed
> > (as Andy Shevchenko suggested in v1 review).
> 
> Not sure if we need a cover letter for a single change, but yes, this
> what I think the best approach and code wise it's fine to me:
> 
> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
> 
> (but we still need a Fixes tag I assume).

Yes, I also though that a cover letter for a single patch could be a bit
useless but I wanted to explain myself the changes from v1 to v2.

I think now I understood why I should include the fixes tag in the
commit message. Let me send v3 with fixes tag and without a cover letter.

Best regards,

Javier R.

> 
> -- 
> With Best Regards,
> Andy Shevchenko
> 
> 

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2025-12-01 18:26 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-12-01 16:07 [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Jose Javier Rodriguez Barbarin
2025-12-01 16:07 ` [PATCH v2 1/1] mcb: Add missing modpost build support Jose Javier Rodriguez Barbarin
2025-12-01 18:02   ` Andy Shevchenko
2025-12-01 18:04 ` [PATCH v2 0/1] mcb: Add modpost support for processing MODULE_DEVICE_TABLE Andy Shevchenko
2025-12-01 18:26   ` Jose Javier Rodriguez Barbarin

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.