Re: ACPI Error: AE_NO_HARDWARE_RESPONSE

On Wed, 2 Jun 2004 j.e.drews_at_att.net wrote:
> I made the polling adjustment that you suggested:
> sysctl hw.acpi.thermal.polling_rate=1
>
> and then ran the script while compiling qt33 and Mozilla at the same time.
> the initial reading was at:  hw.acpi.thermal.tz0.temperature: 3112
> the high temps were:
> hw.acpi.thermal.tz0.temperature: 3242
> hw.acpi.thermal.tz0.temperature: 3252
>
> the fan kicked on at 3232. I could not get the error again, even though
> I ran the computer under load for quite a while.

Ok, thanks for all the debugging info.  There are two issues here:

---
GEOM: Configure ad0s1f, start 1851564032 length 22418914816 end 24270478847
acpi_ec0: info: new max delay is 75 us
acpi_ec0: info: new max delay is 90 us
acpi_ec0: info: new max delay is 210 us
acpi_ec0: info: new max delay is 805 us
acpi_ec0: info: new max delay is 900 us
system power profile changed to 'economy'
---

1. Embedded controller read times out during boot.
As seen above, during the probe of your flash key, the EC hits higher and
higher delays until finally timing out.  The issue is that ACPI acquires
Giant and so does CAM (which is probing your flash drive).  Try booting
without your flash key inserted and see if the error goes away.  I am
working on a patch for ACPI to run Giant-free that should address this.
It's almost done.

---
POWERRES-0257 [20] acpi_pwr_register_cons: registered power consumer
\\_SB_.PCI0.USB7
POWERRES-0350 [19] acpi_pwr_switch_consum: setup to switch
\\_SB_.PCI0.USB7 D255 -> D3
POWERRES-0479 [19] acpi_pwr_switch_consum: attempt to set unsupported state D3
pci0: Failed to set ACPI power state D3 on (null): AE_BAD_PARAMETER
POWERRES-0257 [20] acpi_pwr_register_cons: registered power consumer
\\_SB_.PCI0.MODM
POWERRES-0350 [19] acpi_pwr_switch_consum: setup to switch
\\_SB_.PCI0.MODM D255 -> D3
POWERRES-0479 [19] acpi_pwr_switch_consum: attempt to set unsupported state D3
pci0: Failed to set ACPI power state D3 on (null): AE_BAD_PARAMETER
---

2. Error messages for power switching consumer.
These are harmless but I'll remove them.  The power state routines try to
power down your USB7 and built-in modem since no driver attached them.
However, they don't offer power control via ACPI so the attempt returns a
harmless error message.  I'll add checks for this.

Thanks for the help,
Nate