This patch adds the following changes into generic PCI code especially
for PCI legacy I/O port free drivers.
- Moved the following two things from pci_enable_device() into
pci_enable_device_bars(). Already some drivers are using
pci_enable_device_bars() to enable only the specific regions,
but they had no chance to be fixup and to up is_enabled flag.
o Call pci_fixup_device() on the device
o Checking & Setting dev->is_enabled
- Trivial cleanup
Signed-off-by: Kenji Kaneshige <[email protected]>
Signed-off-by: Hidetoshi Seto <[email protected]>
---
drivers/pci/pci.c | 20 ++++++++------------
1 files changed, 8 insertions(+), 12 deletions(-)
Index: linux-2.6.19-rc6/drivers/pci/pci.c
===================================================================
--- linux-2.6.19-rc6.orig/drivers/pci/pci.c
+++ linux-2.6.19-rc6/drivers/pci/pci.c
@@ -558,12 +558,18 @@
{
int err;
+ if (dev->is_enabled)
+ return 0;
+
err = pci_set_power_state(dev, PCI_D0);
if (err < 0 && err != -EIO)
return err;
err = pcibios_enable_device(dev, bars);
if (err < 0)
return err;
+
+ pci_fixup_device(pci_fixup_enable, dev);
+ dev->is_enabled = 1;
return 0;
}
@@ -578,17 +584,7 @@
int
pci_enable_device(struct pci_dev *dev)
{
- int err;
-
- if (dev->is_enabled)
- return 0;
-
- err = pci_enable_device_bars(dev, (1 << PCI_NUM_RESOURCES) - 1);
- if (err)
- return err;
- pci_fixup_device(pci_fixup_enable, dev);
- dev->is_enabled = 1;
- return 0;
+ return pci_enable_device_bars(dev, (1 << PCI_NUM_RESOURCES) - 1);
}
/**
@@ -737,7 +733,7 @@
{
if (pci_resource_len(pdev, bar) == 0)
return 0;
-
+
if (pci_resource_flags(pdev, bar) & IORESOURCE_IO) {
if (!request_region(pci_resource_start(pdev, bar),
pci_resource_len(pdev, bar), res_name))
On Wed, Nov 22, 2006 at 05:06:07PM +0900, Hidetoshi Seto wrote:
> --- linux-2.6.19-rc6.orig/drivers/pci/pci.c
> +++ linux-2.6.19-rc6/drivers/pci/pci.c
> @@ -558,12 +558,18 @@
> {
> int err;
>
> + if (dev->is_enabled)
> + return 0;
This is unfortunately going to collide with the previous
patch posted by [email protected]:
Subject: [patch 0/2] pci: make pci_{enable,disable}_device() be nested
grant
On Wed, Nov 22, 2006 at 11:09:01AM -0700, Grant Grundler wrote:
> On Wed, Nov 22, 2006 at 05:06:07PM +0900, Hidetoshi Seto wrote:
> > --- linux-2.6.19-rc6.orig/drivers/pci/pci.c
> > +++ linux-2.6.19-rc6/drivers/pci/pci.c
> > @@ -558,12 +558,18 @@
> > {
> > int err;
> >
> > + if (dev->is_enabled)
> > + return 0;
>
> This is unfortunately going to collide with the previous
> patch posted by [email protected]:
>
> Subject: [patch 0/2] pci: make pci_{enable,disable}_device() be nested
Not a problem, that's what I'm here for :)
thanks,
greg k-h
On Wed, Nov 22, 2006 at 11:09:01AM -0700, Grant Grundler wrote:
> On Wed, Nov 22, 2006 at 05:06:07PM +0900, Hidetoshi Seto wrote:
> > --- linux-2.6.19-rc6.orig/drivers/pci/pci.c
> > +++ linux-2.6.19-rc6/drivers/pci/pci.c
> > @@ -558,12 +558,18 @@
> > {
> > int err;
> >
> > + if (dev->is_enabled)
> > + return 0;
>
> This is unfortunately going to collide with the previous
> patch posted by [email protected]:
>
> Subject: [patch 0/2] pci: make pci_{enable,disable}_device() be nested
Grant, you were right. This has changed the logic around this area, and
the pci_enable_device() stuff conflicts with this.
Hidetoshi, I tried to merge things together, but I think I got it wrong,
as the logic is different now. Can you please respin this patch and
resend all of them?
Sorry for the delay,
thanks,
greg k-h