Re: [patch 2.5] PCI: kill pdev_enable_device()

Jeff Garzik (jgarzik@pobox.com)
Tue, 17 Dec 2002 19:52:50 -0500


On Wed, Dec 18, 2002 at 12:42:26AM +0000, Dave Jones wrote:
> On Tue, Dec 17, 2002 at 05:31:57PM -0500, Jeff Garzik wrote:
> > Ivan Kokshaysky wrote:
> > >- So, if we don't touch the PCI command registers, there is no point in
> > > using pdev_enable_device(). Most drivers properly use
> > > pci_enable_device() anyway.
> > Not only that, a driver _should_ be calling pci-enable-device, it's an
> > API requirement. J Random Driver should have a good reason _not_ to
> > call pci_enable_device() ...
>
> What about the xircom issue that was discussed in the last days ?
> Sounds like the solution isn't a full on pci_enable_device() as
> pcmcia 'knows better than us' at that stage aparently.

The solution in the driver is almost always pci_enable_device().

That recent issue was related to subsystem code not driver code;
for that specific situation, you are absolutely right:
pci_enable_device is not the right thing to do.

Jeff

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/