[PATCH] v3: per-device PCI subsystem IDs

[Carl-Daniel Hailfinger]

Prepare for per-device subsystem IDs.
Fix the bus location for Qemu IDE.

This patch only provides the needed infrastructure for per-device
subsystem IDs, it does not hook them up to the PCI core yet, so this
patch is a no-op.
By the way, the on_mainboard property is activating lots of completely
untested code paths in v3, so someone might want to audit them.

Signed-off-by: Carl-Daniel Hailfinger

Index: corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts
===================================================================
--- corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts	(Revision 842)
+++ corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts	(Arbeitskopie)
@@ -30,8 +30,11 @@
    	bus@0 {
    		pci@0,0 {
    		};
-			pci@1,0 {
+			pci@1,1 {
    			/config/("southbridge/intel/i82371eb/ide");
+				subsystem_vendor = "0x15ad";
+				subsystem_device = "0x1976";
+				on_mainboard;
    		};
    	};
    };
Index: corebootv3-subsystem_id_per_device/util/dtc/flattree.c
===================================================================
--- corebootv3-subsystem_id_per_device/util/dtc/flattree.c	(Revision 842)
+++ corebootv3-subsystem_id_per_device/util/dtc/flattree.c	(Arbeitskopie)
@@ -622,6 +622,12 @@
    	if (streq(prop->name, "on_mainboard")){
    		fprintf(f, "\t.on_mainboard = 1,\n");
    	}
+		if (streq(prop->name, "subsystem_vendor")){
+			fprintf(f, "\t.subsystem_vendor = %s,\n", prop->val.val);
+		}
+		if (streq(prop->name, "subsystem_device")){
+			fprintf(f, "\t.subsystem_device = %s,\n", prop->val.val);
+		}
    	if (streq(prop->name, "enabled")){
    		enabled = 1;
    	}

[ron minnich]

Acked-by: Ronald G. Minnich

[Stefan Reinauer]

Carl-Daniel Hailfinger wrote:
> Prepare for per-device subsystem IDs.
> Fix the bus location for Qemu IDE.
>
> This patch only provides the needed infrastructure for per-device
> subsystem IDs, it does not hook them up to the PCI core yet, so this
> patch is a no-op.
>
> Signed-off-by: Carl-Daniel Hailfinger

Acked-by: Stefan Reinauer
> Index: corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts
>
> --- corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts	(Revision 842)
> +++ corebootv3-subsystem_id_per_device/mainboard/emulation/qemu-x86/dts	(Arbeitskopie)
> @@ -30,8 +30,11 @@
>    	bus@0 {
>    		pci@0,0 {
>    		};
> 	pci@1,0 {
> 	pci@1,1 {
> 	/config/("southbridge/intel/i82371eb/ide");
> 		subsystem_vendor = "0x15ad";
> 		subsystem_device = "0x1976";
> 		on_mainboard;
> };
> };
> };
> Index: corebootv3-subsystem_id_per_device/util/dtc/flattree.c
>
> --- corebootv3-subsystem_id_per_device/util/dtc/flattree.c	(Revision 842)
> +++ corebootv3-subsystem_id_per_device/util/dtc/flattree.c	(Arbeitskopie)
> @@ -622,6 +622,12 @@
>    	if (streq(prop->name, "on_mainboard")){
>    		fprintf(f, "\t.on_mainboard = 1,\n");
>    	}
> if (streq(prop->name, "subsystem_vendor")){
> 	fprintf(f, "\t.subsystem_vendor = %s,\n", prop->val.val);
> }
> if (streq(prop->name, "subsystem_device")){
> 	fprintf(f, "\t.subsystem_device = %s,\n", prop->val.val);
> }
> if (streq(prop->name, "enabled")){
> 	enabled = 1;
> }

[Carl-Daniel Hailfinger]

On 29.08.2008 09:02, Stefan Reinauer wrote:
> Carl-Daniel Hailfinger wrote:
> > Prepare for per-device subsystem IDs.
> > Fix the bus location for Qemu IDE.
> >
> > This patch only provides the needed infrastructure for per-device
> > subsystem IDs, it does not hook them up to the PCI core yet, so this
> > patch is a no-op.
> >
> > Signed-off-by: Carl-Daniel Hailfinger
>
> Acked-by: Stefan Reinauer

Acked-by: Ronald G. Minnich

Thanks, r844.

Regards,
Carl-Daniel