2012-11-19 20:04:17

by Paul Bolle

Subject: [PATCH] mfd: lpc_ich: one uninitialized cell is no error

At every boot of an (outdated) laptop lpc_ich prints an error:
lpc_ich 0000:00:1f.0: I/O space for GPIO uninitialized

But if one looks at lpc_ich's probe function one notices that the code
only cares if both lpc_ich_init_wdt() and lpc_ich_init_gpio() fail to
add any cells. So stop treating the failure to add a single cell as an
error. Those messages can be printed at notice level. And then only warn
if no cells were added.

Signed-off-by: Paul Bolle <[email protected]>
---
0) This patch was written on top of v3.7-rc4 and tested on v3.6.7.

1) Please note that this patch prints a warning when no MFD cells are
added at all. I didn't make that an error because I'm not sure there's
anything one can do if this happens. Is there?

drivers/mfd/lpc_ich.c | 11 ++++++-----
1 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/drivers/mfd/lpc_ich.c b/drivers/mfd/lpc_ich.c
index a22544f..a102f7a 100644
--- a/drivers/mfd/lpc_ich.c
+++ b/drivers/mfd/lpc_ich.c
@@ -734,7 +734,7 @@ static int __devinit lpc_ich_init_gpio(struct pci_dev *dev,
pci_read_config_dword(dev, ACPIBASE, &base_addr_cfg);
base_addr = base_addr_cfg & 0x0000ff80;
if (!base_addr) {
- dev_err(&dev->dev, "I/O space for ACPI uninitialized\n");
+ dev_notice(&dev->dev, "I/O space for ACPI uninitialized\n");
lpc_ich_cells[LPC_GPIO].num_resources--;
goto gpe0_done;
}
@@ -760,7 +760,7 @@ gpe0_done:
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");
+ dev_notice(&dev->dev, "I/O space for GPIO uninitialized\n");
ret = -ENODEV;
goto gpio_done;
}
@@ -810,7 +810,7 @@ static int __devinit lpc_ich_init_wdt(struct pci_dev *dev,
pci_read_config_dword(dev, ACPIBASE, &base_addr_cfg);
base_addr = base_addr_cfg & 0x0000ff80;
if (!base_addr) {
- dev_err(&dev->dev, "I/O space for ACPI uninitialized\n");
+ dev_notice(&dev->dev, "I/O space for ACPI uninitialized\n");
ret = -ENODEV;
goto wdt_done;
}
@@ -834,8 +834,8 @@ static int __devinit lpc_ich_init_wdt(struct pci_dev *dev,
pci_read_config_dword(dev, RCBABASE, &base_addr_cfg);
base_addr = base_addr_cfg & 0xffffc000;
if (!(base_addr_cfg & 1)) {
- pr_err("RCBA is disabled by hardware/BIOS, "
- "device disabled\n");
+ dev_notice(&dev->dev, "RCBA is disabled by "
+ "hardware/BIOS, device disabled\n");
ret = -ENODEV;
goto wdt_done;
}
@@ -871,6 +871,7 @@ static int __devinit lpc_ich_probe(struct pci_dev *dev,
* successfully.
*/
if (!cell_added) {
+ dev_warn(&dev->dev, "No MFD cells added\n");
lpc_ich_restore_config_space(dev);
return -ENODEV;
}
--
1.7.7.6

2012-11-21 16:35:07

by Samuel Ortiz

Subject: Re: [PATCH] mfd: lpc_ich: one uninitialized cell is no error

Hi Paul,

On Mon, Nov 19, 2012 at 09:04:11PM +0100, Paul Bolle wrote:
> At every boot of an (outdated) laptop lpc_ich prints an error:
> lpc_ich 0000:00:1f.0: I/O space for GPIO uninitialized
>
> But if one looks at lpc_ich's probe function one notices that the code
> only cares if both lpc_ich_init_wdt() and lpc_ich_init_gpio() fail to
> add any cells. So stop treating the failure to add a single cell as an
> error. Those messages can be printed at notice level. And then only warn
> if no cells were added.
>
> Signed-off-by: Paul Bolle <[email protected]>
> ---
> 0) This patch was written on top of v3.7-rc4 and tested on v3.6.7.
>
> 1) Please note that this patch prints a warning when no MFD cells are
> added at all. I didn't make that an error because I'm not sure there's
> anything one can do if this happens. Is there?
>
> drivers/mfd/lpc_ich.c | 11 ++++++-----
> 1 files changed, 6 insertions(+), 5 deletions(-)
Applied, thanks.

Cheers,
Samuel.

--
Intel Open Source Technology Centre
http://oss.intel.com/