adds AMD Northbridge PCI Id
Signed-off-by: Stephane Eranian <[email protected]>
Signed-off-by: Robert Richter <[email protected]>
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -497,6 +497,8 @@
#define PCI_DEVICE_ID_AMD_K8_NB_ADDRMAP 0x1101
#define PCI_DEVICE_ID_AMD_K8_NB_MEMCTL 0x1102
#define PCI_DEVICE_ID_AMD_K8_NB_MISC 0x1103
+#define PCI_DEVICE_ID_AMD_10H_NB 0x1200
+#define PCI_DEVICE_ID_AMD_10H_NB_MISC 0x1203
#define PCI_DEVICE_ID_AMD_LANCE 0x2000
#define PCI_DEVICE_ID_AMD_LANCE_HOME 0x2001
#define PCI_DEVICE_ID_AMD_SCSI 0x2020
Greg,
Ingo suggested I send ou the following patch which adds a PCI ID
for AMD Barcelona Northbridge. This is used by the perfmon initialization
code on this platform.
Please apply to your tree.
Thanks.
---------------------------------------------------
adds AMD Northbridge PCI Id
Signed-off-by: Stephane Eranian <[email protected]>
Signed-off-by: Robert Richter <[email protected]>
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -497,6 +497,8 @@
#define PCI_DEVICE_ID_AMD_K8_NB_ADDRMAP 0x1101
#define PCI_DEVICE_ID_AMD_K8_NB_MEMCTL 0x1102
#define PCI_DEVICE_ID_AMD_K8_NB_MISC 0x1103
+#define PCI_DEVICE_ID_AMD_10H_NB 0x1200
+#define PCI_DEVICE_ID_AMD_10H_NB_MISC 0x1203
#define PCI_DEVICE_ID_AMD_LANCE 0x2000
#define PCI_DEVICE_ID_AMD_LANCE_HOME 0x2001
#define PCI_DEVICE_ID_AMD_SCSI 0x2020
On Tue, Mar 11, 2008 at 09:50:36AM -0700, stephane eranian wrote:
> Greg,
>
> Ingo suggested I send ou the following patch which adds a PCI ID
> for AMD Barcelona Northbridge. This is used by the perfmon initialization
> code on this platform.
>
> Please apply to your tree.
We are trying to not add pci ids that are only used in one place within
the kernel. Why can't the perfomon init code just add this value to
their code?
Or make this part of the perfmon patch itself, if you really must have
it. No need to add it to the kernel if there are no in-tree users,
right?
thanks,
greg k-h
Greg,
On Tue, Mar 11, 2008 at 10:12 AM, Greg KH <[email protected]> wrote:
> On Tue, Mar 11, 2008 at 09:50:36AM -0700, stephane eranian wrote:
> > Greg,
> >
> > Ingo suggested I send ou the following patch which adds a PCI ID
> > for AMD Barcelona Northbridge. This is used by the perfmon initialization
> > code on this platform.
> >
> > Please apply to your tree.
>
> We are trying to not add pci ids that are only used in one place within
> the kernel. Why can't the perfomon init code just add this value to
> their code?
>
I thought you would want a central place where all PCI ids are defined.
But if you prefer not, then so be it. I'll put it into the perfmon code.
> Or make this part of the perfmon patch itself, if you really must have
> it. No need to add it to the kernel if there are no in-tree users,
> right?
>
I am starting to push perfmon patches upstream. So you should expect
patches which *currently* have no in-tree users. You have to start from
somewhere, right?
On Tue, Mar 11, 2008 at 04:10:41PM -0700, stephane eranian wrote:
> Greg,
>
>
> On Tue, Mar 11, 2008 at 10:12 AM, Greg KH <[email protected]> wrote:
> > On Tue, Mar 11, 2008 at 09:50:36AM -0700, stephane eranian wrote:
> > > Greg,
> > >
> > > Ingo suggested I send ou the following patch which adds a PCI ID
> > > for AMD Barcelona Northbridge. This is used by the perfmon initialization
> > > code on this platform.
> > >
> > > Please apply to your tree.
> >
> > We are trying to not add pci ids that are only used in one place within
> > the kernel. Why can't the perfomon init code just add this value to
> > their code?
> >
> I thought you would want a central place where all PCI ids are defined.
> But if you prefer not, then so be it. I'll put it into the perfmon code.
>
> > Or make this part of the perfmon patch itself, if you really must have
> > it. No need to add it to the kernel if there are no in-tree users,
> > right?
> >
> I am starting to push perfmon patches upstream. So you should expect
> patches which *currently* have no in-tree users. You have to start from
> somewhere, right?
Submit them at the same time, is that a problem? We don't want to be
accepting changes for an out-of-tree user that doesn't make it in :)
thanks,
greg k-h