[PATCH] Please test: new ACPI release (20041105) import

[M. Warner Losh]

In message: <419C3AC2.5010505 at root.org>
            Nate Lawson <nate at root.org> writes:
: Warner Losh wrote:
: >>My serial port.  It's named UART in the AML, that's all.  Setting it to 
: >>D3 hangs the system.
: > 
: > Is it a PCI device too?  If not, then maybe you need to disable
: > interrupts in the UART before setting the power state.  Chances are
: > its interrupt cause register is returning 0xff, which hangs sio, since
: > it never breaks out of the loop when this happens.
: 
: It is not, it's a legacy device:
:      sio0 pnpinfo _HID=PNP0501 _UID=0 at handle=\_SB_.PCI0.LPC_.UART
: 
: The power methods for it are:
:                      Method (_PS0, 0, NotSerialized)
:                      {
:                          Store (0x03, LDN)
:                          Store (0x01, LDA)
:                          Store (0x01, SSBY)
:                      }
: 
:                      Method (_PS3, 0, NotSerialized)
:                      {
:                          Store (0x03, LDN)
:                          Store (0x00, LDA)
:                          Store (0x00, SSBY)
:                      }
: 
: The regions accessed by these Store ops are as follows, and all appear 
: to be mapped to the ICH:
:                  OperationRegion (IMGA, SystemIO, 0x15E0, 0x10)
:                  OperationRegion (NCFG, SystemIO, 0x2E, 0x02)
: 
: Is there a standard for how to do power transitions for non-PCI devices? 
:   I thought you just call the acpi methods if present.

There might be some driver stuff that will need to be done on top of
the calling these things.  I don't know what the behavior of the
device is, but many pci devices read 0xffffffff when in d3 mode.

Warner