Hi Greg and Andrew,

I'm attaching updated patches for adding pcibiod_disable_device()
hook based on the feedback from Ashok (Thank you, Ashok!).

I made two patches, one of them is against 2.6.9-rc2-mm1 and the
another is against 2.6.9-rc2-mm4 to which the previous version of
the patch has already been applyed. Please use the one convenient
for you.

Thanks,
Kenji Kaneshige


Greg KH wrote:

> On Fri, Sep 24, 2004 at 01:02:52PM -0700, Ashok Raj wrote:
>> On Thu, Sep 16, 2004 at 05:54:14AM -0700, Kenji Kaneshige wrote:
>> > 
>> >    Hi,
>> > 
>> >    This patch adds a hook 'pcibios_disable_device()' into
>> >    pci_disable_device() to call architecture specific PCI resource
>> >    deallocation code. It's a opposite part of pcibios_enable_device().
>> >    We need this hook to deallocate architecture specific PCI resource
>> >    such as IRQ resource, etc.. This patch is just for adding the hook,
>> >    so pcibios_disable_device() is defined as a null function on all
>> >    architecture so far.
>> > 
>> >    I tested this patch on i386, x86_64 and ia64. But it has not been
>> >    tested on other architectures because I don't have these machines.
>> > 
>> >    Signed-off-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
>> > 
>> 
>> Hi Kenji
>> 
>> I think instead of modifying all the arch specific code, you could use the __attribute__(weak)
>> and define a default dummy funcion in 	drivers/pci/pci.c
>> 
>> void __attribute__((weak)) pcibios_disable_device(struct pci_dev *dev)	{ }
>> 
>> 
>> each arch that really needs this can define the override function.
>> That way you dont need to put the dummy function in several places,
>> containing your changes to a very few set of files.
> 
> Ohhh, nice.  I like that option better.  Kenji, care to respin your
> patches based on this change?
> 
> thanks,
> 
> greg k-h
>