CONFIG_PCI_LEGACY removal

CONFIG_PCI_LEGACY removal

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 868 bytes --]

Hi all,

Now that it is more than a year since CONFIG_PCI_LEGACY was added (Oct
2007) and longer since pci_find_slot() (May 2007) and pci_find_device()
(Nov 2006) were deprecated, is there any way that we can get rid of the
last vestiges of them?

Users of pci_find_device():

drivers/isdn/hisax/avm_pci.c
drivers/isdn/hisax/bkm_a4t.c
drivers/isdn/hisax/bkm_a8.c
drivers/isdn/hisax/diva.c
drivers/isdn/hisax/elsa.c
drivers/isdn/hisax/enternow_pci.c
drivers/isdn/hisax/gazel.c
drivers/isdn/hisax/hfc_pci.c
drivers/isdn/hisax/niccy.c
drivers/isdn/hisax/nj_s.c
drivers/isdn/hisax/nj_u.c
drivers/isdn/hisax/sedlbauer.c
drivers/isdn/hisax/telespci.c
drivers/isdn/hisax/w6692.c

User of pci_find_slot():

drivers/pci/hotplug/cpqphp_pci.c

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: CONFIG_PCI_LEGACY removal

[Matthew Wilcox]

On Fri, Nov 21, 2008 at 06:14:26PM +1100, Stephen Rothwell wrote:
> Users of pci_find_device():
> 
> drivers/isdn/hisax/avm_pci.c
> drivers/isdn/hisax/bkm_a4t.c
> drivers/isdn/hisax/bkm_a8.c
> drivers/isdn/hisax/diva.c
> drivers/isdn/hisax/elsa.c
> drivers/isdn/hisax/enternow_pci.c
> drivers/isdn/hisax/gazel.c
> drivers/isdn/hisax/hfc_pci.c
> drivers/isdn/hisax/niccy.c
> drivers/isdn/hisax/nj_s.c
> drivers/isdn/hisax/nj_u.c
> drivers/isdn/hisax/sedlbauer.c
> drivers/isdn/hisax/telespci.c
> drivers/isdn/hisax/w6692.c

I think the plan is to delete all these drivers once the new ISDN layer
is a bit more battle-hardened.

> User of pci_find_slot():
> 
> drivers/pci/hotplug/cpqphp_pci.c

The cpqphp driver is basically unmaintained.  Last time I spoke to Greg
about it (which was probably three years ago) he said it wasn't possible
to find a machine which had one of these any more.

I suspect the driver should simply be deleted, but if anyone steps up
and claims to have one of these to test the resulting patch, I'm willing
to remove pci_find_slot() from it.

-- 
Matthew Wilcox				Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."

Re: CONFIG_PCI_LEGACY removal

[Alan Cox]

On Fri, 21 Nov 2008 18:14:26 +1100
Stephen Rothwell <sfr@canb.auug.org.au> wrote:

> Hi all,
> 
> Now that it is more than a year since CONFIG_PCI_LEGACY was added (Oct
> 2007) and longer since pci_find_slot() (May 2007) and pci_find_device()
> (Nov 2006) were deprecated, is there any way that we can get rid of the
> last vestiges of them?

Jeff Garzik I believe it was posted patches to do this. They were ignored
and the old ISDN code appears unmaintained. These drivers should IMHO
therefore be marked as BROKEN in 2.6.28 or Jeff's patch applied. Ditto
for the old cpq driver.

If Jeff's patch is wrong we'll find out if anyone even uses that code.

Re: CONFIG_PCI_LEGACY removal

[Greg KH]

On Fri, Nov 21, 2008 at 04:08:19AM -0700, Matthew Wilcox wrote:
> > drivers/pci/hotplug/cpqphp_pci.c
> 
> The cpqphp driver is basically unmaintained.  Last time I spoke to Greg
> about it (which was probably three years ago) he said it wasn't possible
> to find a machine which had one of these any more.
> 
> I suspect the driver should simply be deleted, but if anyone steps up
> and claims to have one of these to test the resulting patch, I'm willing
> to remove pci_find_slot() from it.

I say mark it CONFIG_BROKEN for now.

thanks,

greg k-h

Re: CONFIG_PCI_LEGACY removal

[Alex Chiang]

* Matthew Wilcox <matthew@wil.cx>:
> On Fri, Nov 21, 2008 at 06:14:26PM +1100, Stephen Rothwell wrote:
> 
> I think the plan is to delete all these drivers once the new ISDN layer
> is a bit more battle-hardened.
> 
> > User of pci_find_slot():
> > 
> > drivers/pci/hotplug/cpqphp_pci.c
> 
> The cpqphp driver is basically unmaintained.  Last time I spoke to Greg
> about it (which was probably three years ago) he said it wasn't possible
> to find a machine which had one of these any more.

I've managed to track some down.

> I suspect the driver should simply be deleted, but if anyone steps up
> and claims to have one of these to test the resulting patch, I'm willing
> to remove pci_find_slot() from it.

Converting cpqphp to use the modern pci_get_slot() API has been
on my TODO list for a while. It was languishing because I
couldn't find hardware nor time.

I've got hardware now. Just need to free up a few cycles. :-/

But if it's not worth it, I'm fine with marking as CONFIG_BROKEN
and scheduling for eventual deletion.

/ac