lkml.org 
[lkml]   [2012]   [Feb]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH 1/3 v2] mfd: Add LPC driver for Intel ICH chipsets
From
Date
On Tue, 2012-02-07 at 16:00 -0500, Aaron Sierra wrote:
> > > +static int __devinit lpc_ich_probe(struct pci_dev *dev,
> > > + const struct pci_device_id *id)
> > > +{
> > > + u32 base_addr_cfg;
> > > + u32 base_addr;
> > > + u8 reg_save;
> > > + int ret;
> > > + int cells = 0;
> > > + int acpi_conflict = 0;
> > > +
> > You can use bool for acpi_conflict (and cells, but I don't think that
> > is needed anyway).
>
> I agree that bool is a better type choice. See my comment at the end
> regarding my reason for using cells.
>
>
> > > +pm_done:
> > > + /* Setup GPIO base register */
> > > + pci_read_config_dword(dev, GPIOBASE, &base_addr_cfg);
> > > + base_addr = base_addr_cfg & 0x0000ff80;
> > > + if (!base_addr) {
> > > + dev_err(&dev->dev, "I/O space for GPIO
> > > uninitialized\n");
> > > + /* GPIO in power-management space may still be
> > > available */
> > > + goto gpio_reg;
> > > + }
> > > +
> > > + gpio_ich_res[ICH_RES_GPIO].start = base_addr;
> > > + gpio_ich_res[ICH_RES_GPIO].end = base_addr +
> > > GPIOBASE_IO_SIZE - 1;
> > > + ret =
> > > acpi_check_resource_conflict(&gpio_ich_res[ICH_RES_GPIO]);
> > > + if (ret) {
> > > + /* this isn't necessarily fatal for the GPIO */
> > > + gpio_ich_res[ICH_RES_GPIO].start = 0;
> > > + gpio_ich_res[ICH_RES_GPIO].end = 0;
> > > + acpi_conflict++;
> >
> > acpi_conflict = true;
> >
> > After all, it does not really matter how many conflicts were
> > detected.
>
> I agree.
>
> > > +gpio_reg:
> > > + lpc_ich_finalize_cell(&lpc_ich_cells[LPC_GPIO], id);
> > > + ret = mfd_add_devices(&dev->dev, 0,
> > > &lpc_ich_cells[LPC_GPIO],
> > > + 1, NULL, 0);
> > > + if (!ret)
> > > + cells++;
> > > +
> > So if there is an error, ret < 0, correct ?
> >
> > > + if (acpi_conflict)
> > > + dev_info(&dev->dev, "ACPI resource conflicts found;
> > > "
> > > + "consider using
> > > acpi_enforce_resources=lax?\n");
> > > +
> > > + if (cells)
> > > + return 0;
> > > + else
> > > + return -ENODEV;
> >
> > If the above is true, you can just return ret, and you don't need the
> > cells variable. Or, even better, move the acpi warning above the call
> > to mfd_add_devices().
>
> The cells variable isn't strictly necessary when we're only dealing
> with one cell registration, as we have if only looking at the lpc_ich
> and gpio-ich patches. The iTCO_wdt patch adds a second call to
> mfd_add_devices, so when we return we're interested if either of the
> calls succeeded. This was intended to be a forward thinking
> implementation, but I have no qualms about simplifying it in the
> initial lpc_ich patch.

No, just keep it - I realized that when I looked into the iTCO patch.

Thanks,
Guenter




\
 
 \ /
  Last update: 2012-02-07 22:13    [W:0.124 / U:3.808 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site