Re: [Mjpeg-users] [PATCH] media: pci: Use PCI_DEVICE macros and drop zeros from pci_device_id arrays

[Uwe Kleine-K. (T. C. H. <u.k...@ba...>]

Hello,

On Wed, May 06, 2026 at 07:44:12PM +0200, Uwe Kleine-König (The Capable Hub) wrote:
> Some pci_device_id arrays already used the PCI_DEVICE macros, others
> used named initializers or list initializers. Unify all to use the
> PCI_DEVICE macros and (if applicable) a named initializer for
> .driver_data. Also drop all zeros that the compiler can care for and
> unify indention and spacing.
> 
> There are no changes introduced by this commit to the compiled arrays for
> ARCH=x86 and ARCH=arm64 builds.
> 
> Signed-off-by: Uwe Kleine-König (The Capable Hub) <u.k...@ba...>
> ---
> Hello,
> 
> this is a preparing change for making struct pci_device_id::driver_data
> an anonymous union (similar to
> https://lore.kernel.org/all/cov...@ba.../).
> This requires named initializers for .driver_data. Not all changes here
> are required for that, but I expanded the quest to unify a bit more than
> necessary.
> 
> I put all changes in a single patch, if you prefer it being split I can do that.

I just noticed I failed to put the following hunk into the patch:

diff --git a/include/media/drv-intf/saa7146.h b/include/media/drv-intf/saa7146.h
index 71ce63c99cb4..b8f2fe5680f0 100644
--- a/include/media/drv-intf/saa7146.h
+++ b/include/media/drv-intf/saa7146.h
@@ -77,13 +77,10 @@ struct saa7146_pci_extension_data {
 	void *ext_priv;			/* most likely a name string */
 };
 
-#define MAKE_EXTENSION_PCI(x_var, x_vendor, x_device)		\
-	{							\
-		.vendor    = PCI_VENDOR_ID_PHILIPS,		\
-		.device	   = PCI_DEVICE_ID_PHILIPS_SAA7146,	\
-		.subvendor = x_vendor,				\
-		.subdevice = x_device,				\
-		.driver_data = (unsigned long)& x_var,		\
+#define MAKE_EXTENSION_PCI(x_var, x_vendor, x_device)						\
+	{											\
+		PCI_VDEVICE_SUB(PHILIPS, PCI_DEVICE_ID_PHILIPS_SAA7146, (x_vendor), (x_device)),\
+		.driver_data = (unsigned long)&(x_var),						\
 	}
 
 struct saa7146_extension

but that's not urgent as the current definition of MAKE_EXTENSION_PCI()
is good enough to not block the next step of my quest. I'll address that
in a v2 (if it comes to that) or later together with the next step
making the driver_data line:

	.driver_data_ptr = &(x_var),

(or if you apply this version, feel free to squash the diff into the
commit).

Best regards
Uwe