Re: [PATCH] i2c-scmi: Quirk to work on IBM machines with broken BIOSes

[Jean Delvare <khali@linux-fr.org>]

On Thu, 22 Oct 2009 10:43:04 -0700, Darrick J. Wong wrote:
> On Thu, Oct 22, 2009 at 03:17:53PM +0800, Crane Cai wrote:
> 
> > This patch below represents my meanings:
> > *) add a new HID for IBM SMBus CMI devices
> > *) add methods for IBM SMBus CMI devices as you did
> > *) hook different HID with different control methods set
> > It may be more smooth for i2c-scmi, please consider.
> 
> Looks fine to me.
> 
> I still need the changes to drivers/acpi/scan.c, but if you push this patch
> upstream then I'll reroll my patch as a follow-on to yours...
> 
> >  static const struct acpi_device_id acpi_smbus_cmi_ids[] = {
> > -	{"SMBUS01", 0},
> > +	{"SMBUS01", (kernel_ulong_t)&smbus_methods},
> > +	{"SMBUSIBM", (kernel_ulong_t)&ibm_smbus_methods},
> 
> ...with the custom HID string #define'd in a header file someplace.
> 
> Actually, if you'll add a Signed-off-by line to your patch, I'll submit both of
> them as a patchset and save you a little work. :)

As soon as there is an agreement on how the problem would better be
addressed, I'll be happy to pick the patches.

> > +	for (; id->id[0]; id++)
> 
> Stylistic nit--would it be clearer to initialize id in the for loop instead of
> at the beginning of the function?

Definitely, yes.

-- 
Jean Delvare