cvs commit: src/sys/dev/acpica acpi.c

[Eric Anderson]

Nate Lawson wrote:
> Eric Anderson wrote:
> 
>>>>>
>>>>> Ok - I've narrowed it down.  A GENERIC kernel will go into S3 just 
>>>>> fine
>>>>> on this laptop.  Removing apic from the kernel will break that.
> 
> 
> It is interesting that the suspend path without the apic support causes 
> a hang for you.  This should be investigated.  Does it work if you have 
> apic compiled in but boot with hint.apic.0.disabled="1" ? Any ideas 
> where to look John?

Setting this to 1 does a funny thing.  When I ask the machine to go to 
s3, it acts like it will, the power light dims, the machine clicks like 
it's going into s3, but then immediately the power light comes back on, 
the hard drive spins back up, and the machine starts a normal power up.


>>>>> I've also run into some bugs with having smp in the kernel and certain
>>>>> modules, so beware.
>>>>>
>>>>> Now, I can successfully go into S3, but coming back out (using the lid
>>>>> switch button or the power button are the only ways I know of) 
>>>>> seems to
>>>>> reboot the machine.  Maybe I should say that I can't tell if it's
>>>>> rebooting the machine, or if the machine just 'powers up' as if it was
>>>>> off.
>>>>
>>>>
>>>>
>>>> Oh - and here's what it looks like when it goes to sleep:
>>>>
>>>> acpi_lid0: Lid closed
>>>> acpi_lid0: wake_prep enabled for \_SB_.LID_ (S3)
>>>> acpi_button0: wake_prep enabled for \_SB_.PBTN (S3)
>>>> uhci0: wake_prep disabled wake for \_SB_.PCI0.USB0 (S3)
>>>> uhci1: wake_prep disabled wake for \_SB_.PCI0.USB1 (S3)
>>>> uhci2: wake_prep disabled wake for \_SB_.PCI0.USB2 (S3)
>>>> uhci3: wake_prep disabled wake for \_SB_.PCI0.USB4 (S3)
>>>> ehci0: wake_prep disabled wake for \_SB_.PCI0.USB3 (S3)
> 
> 
> This is acpi enabling devices to wake the system from S3 and disabling 
> others.  This is done before running all DEVICE_SUSPEND methods since 
> those could potentially power down the device, and then it won't be able 
> to do the wake prep.  (Unlikely since it is the chipset which does the 
> wake enable function but still just being careful.)
> 
> You could disable this by commenting out the acpi_wake_prep_walk(state) 
> call in acpi_SetSleepState().  I doubt this will change anything.
> 
>>>> pci2:0:0: Transition from D0 to D3
>>>> vga0: saving 68 bytes of video state
>>>> pci0:31:2: Transition from D0 to D3
> 
> 
> You could disable PCI power state support with hw.pci.do_power_state="0"
> 
>>>> ioapic_suspend: not implemented!
>>>> ioapic_suspend: not implemented!
> 
> 
> I still think this needs to be implemented although it's not likely to 
> be your problem.
> 
>>>> ======== acpi_printcpu() debug dump ========
>>>> gdt[0097:c09c7380] idt[07ff:c09c7ea0] ldt[0050] tr[0048] efl[00080086]
>>>> eax[00000001] ebx[c23ccc80] ecx[00000001] edx[00000004]
>>>> esi[00000003] edi[00000000] ebp[e3618c5c] esp[e3618c40]
>>>> cr0[8005003b] cr2[2813d704] cr3[00c1e000] cr4[00000691]
>>>> cs[0020] ds[0028] es[0028] fs[0008] gs[003b] ss[0028]
>>>
>>>
>>
>> No - it boots up like I had it powered off.  Anything I can debug to 
>> figure that out?
> 
> 
> When it's suspended, is the sleep light on or are all lights off like it 
> is already powered off?  If it's on, it is likely actually suspended. On 

The power light is throbbing, same thing it does when in WinXP, so it 
appears to be actually suspended.

> power up, it's possible your bios doesn't even call the wake vector. 
>  There was a beep patch I never committed from takawata that can be used 
> to see if your bios even calls the asm wake code.  imp has it 
> somewhere.  Try disabling the calls to video reset in the wake code 
> (hw.acpi.reset_video=0).  Basically, you see if you're even executing 
> the wake code.

I'll check that out.

For now, setting these two:
hw.acpi.reset_video=0
hw.pci.do_powerstate=0

Makes it act like it's going to come back - the system power light turns 
to steady, and things like the caps lock/numlock work, but nothing else, 
and the video stays off.  I have to power it off and back on.  Closer!

However, setting just this:
hw.pci.do_powerstate=0
makes it do the same thing as it does when it is set to 1 - that is it 
reboots like a normal power on, instead of resume.

Eric



-- 
------------------------------------------------------------------------
Eric Anderson        Sr. Systems Administrator        Centaur Technology
A lost ounce of gold may be found, a lost moment of time never.
------------------------------------------------------------------------