sony-vaio-z-series team mailing list archive

[Adam Hill <sidepipeuk@xxxxxxxxx>]

> 	If I use sony-laptop--zseries-0.9np7 to switch to stamina mode at boot
> time, I get a load of:
> 
> hda-intel: spurious response 0x0:0x0, last cmd=0x0f0000

	Ok, to follow up my own e-mail, I've fixed this, though it has wider 
kernel implications and so I'm not sure whether this is the best way to solve 
it.

	Turns out that the messages are not about the Intel card at all, but the 
NVidia sound card which deals with HDMI audio, and also happens to be a "hda-
intel". The problem happens because, as with everything else, even though it 
is disabled using ACPI, and the kernel detects that it won't wake up again, it 
fails to communicate that fact back to the driver, which then tries to 
initialise it anyway ( despite it being powered off. )

	MY fix is:

diff -rupN linux-2.6.35-rc6.orig/drivers/pci/pci.c linux-2.6.35-
rc6.new/drivers/pci/pci.c
--- linux-2.6.35-rc6.orig/drivers/pci/pci.c     2010-07-22 20:13:38.000000000 
+0100
+++ linux-2.6.35-rc6.new/drivers/pci/pci.c      2010-07-27 12:27:02.219517502 
+0100
@@ -594,8 +594,12 @@ static int pci_raw_set_power_state(struc
        pci_read_config_word(dev, dev->pm_cap + PCI_PM_CTRL, &pmcsr);
        dev->current_state = (pmcsr & PCI_PM_CTRL_STATE_MASK);
        if (dev->current_state != state && printk_ratelimit())
+       {
                dev_info(&dev->dev, "Refused to change power state, "
                        "currently in D%d\n", dev->current_state);
+               if(dev->current_state == 3)
+                       return -ENXIO;
+       }
 
        /* According to section 5.4.1 of the "PCI BUS POWER MANAGEMENT
         * INTERFACE SPECIFICATION, REV. 1.2", a device transitioning


	This would have global implications, though I'm not sure what the problem 
would  be in returning an error if ANY device fails to come out of D3 state. 
I'm not sure what error should really be returned though... maybe ENODEV 
instead. Off to find a kernel PCI developer.....

Adam.