Jump to content

DSDT fixes for Gigabyte boards

iSoprano

By iSoprano,
in Guides

 Share
1908 posts in this topic

Recommended Posts

Johnny V
Johnny V
Posted
Posted
So you have never even been using vanilla powermanagement. Probably you havent fixed LPC.

 

I have it according to lspci -nn via the Terminal:

 

00:1f.0 ISA bridge [0601]: Intel Corporation 82801IR (ICH9R) LPC Interface Controller [8086:2916] (rev 02)

 

Here's my kexts living in the Cham/Extra/Extensions folder:

 

  • fakesmc.kext
  • LegacyHDA.kext
  • NullCPUPowerManagement.kext
  • OrangeIconFix.kext
  • OSXRestart.kext
  • SleepEnabler.kext
  • VoodooPowerMini.kext

 

Could the VoodooPowerMini.kext be the problem?

 

Also I wasn't sure what data goes under Name (CST, Package (0x02) Here's my Scope data:

Scope (_PR) {
       Name (PSS, Package (0x05) {
           Package (0x06) { 3200, Zero, 10, 10, 0x0A24, Zero },
           Package (0x06) { 2880, Zero, 10, 10, 0x0924, One },
           Package (0x06) { 2560, Zero, 10, 10, 0x0824, 0x02 },
           Package (0x06) { 2240, Zero, 10, 10, 0x0724, 0x03 },
           Package (0x06) { 1920, Zero, 10, 10, 0x0624, 0x04 }
       })
       Name (PSD, Package (0x05) {
           0x05,Zero,Zero,0xFC,0x04
        })
       Name (CST, Package (0x02) {
           One,
           Package (0x04){ResourceTemplate (){Register (FFixedHW,0x01,0x02,0x0000000000000000,0x00,)},One,One,0x03E8}, 
       })
       Processor (CPU0, 0x00, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU1, 0x01, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU2, 0x02, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU3, 0x03, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
   }

What do you guys think?

 

Many Thanks!

mm67
mm67
Posted
Posted
I have it according to lspci -nn via the Terminal:

 

00:1f.0 ISA bridge [0601]: Intel Corporation 82801IR (ICH9R) LPC Interface Controller [8086:2916] (rev 02)

 

Here's my kexts living in the Cham/Extra/Extensions folder:

 

  • fakesmc.kext
  • LegacyHDA.kext
  • NullCPUPowerManagement.kext
  • OrangeIconFix.kext
  • OSXRestart.kext
  • SleepEnabler.kext
  • VoodooPowerMini.kext

 

Could the VoodooPowerMini.kext be the problem?

 

Also I wasn't sure what data goes under Name (CST, Package (0x02) Here's my Scope data:

Scope (_PR) {
       Name (PSS, Package (0x05) {
           Package (0x06) { 3200, Zero, 10, 10, 0x0A24, Zero },
           Package (0x06) { 2880, Zero, 10, 10, 0x0924, One },
           Package (0x06) { 2560, Zero, 10, 10, 0x0824, 0x02 },
           Package (0x06) { 2240, Zero, 10, 10, 0x0724, 0x03 },
           Package (0x06) { 1920, Zero, 10, 10, 0x0624, 0x04 }
       })
       Name (PSD, Package (0x05) {
           0x05,Zero,Zero,0xFC,0x04
        })
       Name (CST, Package (0x02) {
           One,
           Package (0x04){ResourceTemplate (){Register (FFixedHW,0x01,0x02,0x0000000000000000,0x00,)},One,One,0x03E8}, 
       })
       Processor (CPU0, 0x00, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU1, 0x01, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU2, 0x02, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
       Processor (CPU3, 0x03, 0x00000410, 0x06){
           Alias (PSS, _PSS)
           Alias (PSD, _PSD)
           Alias (CST, _CST)
       }
   }

What do you guys think?

 

Many Thanks!

 

All of these have to go:

[*]NullCPUPowerManagement.kext

[*]SleepEnabler.kext

[*]VoodooPowerMini.kext

 

And your PSS table looks very strange. Somehow I don't believe that all your P-states have same VID value

keeza
keeza
Posted
Posted
Do you think it could even remotely be related to the fact you have a Q9550 that supports more C states than mine? If I had hair I'd have pulled it all out by now. Now I need to know why they work for you but not me. What's your DSDT look like these days? What bootloader? Are you using any bootloader device enablers? Which non-vanilla kexts are you using?

 

As soon as I install either of those apps on the virgin fresh install I loose sleep.

I've also started to get sleep issues.

This has happened since I:

 

Updated Chameleon to RC4

Installed Istat Menu

Changed my version of FakeSMC to 2.5 (for istat menu temp support, using 2.1 before)

Used the RestartFix=Yes in .com.plist (deleted Evoreboot)

Used Weaksauce ifconfig for networking

 

So one of these is the culprit for me.

(Auto sleep doesn't shut off completely, get a black screen and fans run. Manual sleep is fine)

 

I don't use the PCI=1 fix as I have graphics in DSDT.

Johnny V
Johnny V
Posted
Posted
All of these have to go:

[*]NullCPUPowerManagement.kext

[*]SleepEnabler.kext

[*]VoodooPowerMini.kext

 

And your PSS table looks very strange. Somehow I don't believe that all your P-states have same VID value

 

I'll start over but what's weird I just launched PStateChanger and now all the VID values are different. The VIDs were all the same... not sure what happened.

 

Thanks again.

kdawg
kdawg
Posted
Posted
I've also started to get sleep issues.

This has happened since I:

 

Updated Chameleon to RC4

Installed Istat Menu

Changed my version of FakeSMC to 2.5 (for istat menu temp support, using 2.1 before)

Used the RestartFix=Yes in .com.plist (deleted Evoreboot)

Used Weaksauce ifconfig for networking

 

So one of these is the culprit for me.

(Auto sleep doesn't shut off completely, get a black screen and fans run. Manual sleep is fine)

 

I don't use the PCI=1 fix as I have graphics in DSDT.

I would uninstall (use uninstaller and check logs for errors) iStat, and the ipconfig script then see if auto-sleep comes back.
keeza
keeza
Posted
Posted
And your cpu is E0 revision, not C1 ? Those support only C1E. You can also verify that C4 works at all by booting into Linux with C4 enabled and disabled from bios. Run that same setpci command and see if result changes with different bios settings, for me it does.

 

OK, can I start by verifying my BIOS settings?

I've attached them all, they may be handy as a reference to someone starting out...

Keeza_BIOS.zip

Johnny V
Johnny V
Posted
Posted

http://www.insanelymac.com/forum/index.php?showtopic=181631

 

OK I followed the above link but still have temps about 8 degrees higher. At idle it's a steady 59C. Took out from the Extra folder:

 

NullCPUPowerManagement.kext

SleepEnabler.kext

VoodooPowerMini.kext

 

I've launched both PStateChanger and VoodooMonitor and all VID values are at 24 and all mV values are at 1275. So not sure what's going on.

 

Attached is the dsdt.aml file... maybe you brainiacs can figure it out! I think I did the P-State right this time!?

Thanks!

 

 

 

I would uninstall (use uninstaller and check logs for errors) iStat, and the ipconfig script then see if auto-sleep comes back.

I'm running iStat and sleep and all works fine.

dsdt.aml.zip

keeza
keeza
Posted
Posted
I would uninstall (use uninstaller and check logs for errors) iStat, and the ipconfig script then see if auto-sleep comes back.

thanks, not too worried about it at this stage as it shouldn't be too hard to isolate the cause. Getting working CStates is my priority at the moment....

waldo_pepperjack
waldo_pepperjack
Posted
Posted
This is a followup on my trek to find out why my machine wasn't sleeping. ... But for those who want auto-sleep iStat menus and Dropbox seem to be out.

 

kdawg, I also notice that a virgin vanilla install had good Auto/IdleSleep, but after I added in my favorite apps, I lost that.

 

This seems to be true on non-Hacked Macs sometimes, as well (like my wife's MacBookPro).

 

My solution for it is an off-the-shelf application called PleaseSleep.

rayap
rayap
Posted
Posted

Hi All,

With only FFixedHW in CST per mm67 method, able to get "80". But my cpu runs a couple of degrees hotter and my rig consumes several extra Watts on idle; compared with second case when SystemIO is included in CST method - thus no C3/C4.

On Ubuntu 9.10 with ('80") or without("00") C4/ C4E enabled in BIOS, cpu temperatures are similar to the second case, however consumes some Watts less.

 

In SL temp with VoodooMonitor, in Ubuntu "sensors" Power reading with Zalman ZM-MFC2.

greythorne
greythorne
Posted
Posted

ok guys i am having this very annoying issue. Basically what happens is that my system will have random lockup whatever on the screen can be seen only thing is that the mouse cursor will not move and there will be no power to the keyboard. This happens after a certain period of time like couple of minutes into using OS X. I got a new gfx card which is ASUS 9600GT 512mb. I am using graphics enabler in my com.boot.plist so i do not have gfx fix in the DSDT i am using now.

 

I even changed to a new PSU which is Antec truepower 550W with 80+ bronze rating.

 

Any idea if this issue is somehow related to the DSDT i am using?

 

Thank you

 

Attached the DSDT i am using now.

 

mm67
mm67
Posted
Posted
OK, can I start by verifying my BIOS settings?

I've attached them all, they may be handy as a reference to someone starting out...

 

Mostly our settings look similar, you just seem to have the JMicron controller enabled in bios, so I assume that you still have code for that in DSDT ?

 

 

Hi All,

With only FFixedHW in CST per mm67 method, able to get "80". But my cpu runs a couple of degrees hotter and my rig consumes several extra Watts on idle; compared with second case when SystemIO is included in CST method - thus no C3/C4.

On Ubuntu 9.10 with ('80") or without("00") C4/ C4E enabled in BIOS, cpu temperatures are similar to the second case, however consumes some Watts less.

 

In SL temp with VoodooMonitor, in Ubuntu "sensors" Power reading with Zalman ZM-MFC2.

 

I do think that we get C4, but the voltages and multis are really jumpy with these settings, seems like the system is changing states all the time.

mm67
mm67
Posted
Posted
mm67

 

May i check with you...

 

when i issue the command 

sudo setpci -s 0:1f.0 0xa6.b

the output was 00 so does this means that my C States are not working??

 

Just want to confirm.

 

Thank you.

 

That only tells that your cpu is not going to C3 or C4. It says nothing about C1 and C2

Aargh-a-Knot
Aargh-a-Knot
Posted
Posted
This is a followup on my trek to find out why my machine wasn't sleeping. Everything was fine as far as DSDT, BIOS etc. I went and did a fresh install and it sleeps just fine, Windows 7 Was sleeping fine as well. But, I wanted to find out why my old install wasn't sleeping. The answer is iStat menus. Using the uninstaller that came with the program didn't fully remove everything. I realized this when looking through the logs. I kept getting error messages looking for unavailable resources. So I reinstalled iStat then uninstalled again and BAM! Auto-sleep works again. This on top of Dropbox that also doesn't allow sleep. It had nothing to do with any of the networking kexts as I thought previously. All the standard kexts and DSDT patches all work fine.

 

As I reinstall everything one at at time and see how it effect sleep I'll report back. But for those who want auto-sleep iStat menus and Dropbox seem to be out.

 

kdawg, not to throw a wrench in the gears of your theory, but my Gigabyte system with working sleep has iStat menus installed.

 

On my new p55 build, sleep problems are directly related to the optical drive. What a P.I.T.A this whole sleep thing is!

keeza
keeza
Posted
Posted
And your cpu is E0 revision, not C1 ? Those support only C1E. You can also verify that C4 works at all by booting into Linux with C4 enabled and disabled from bios. Run that same setpci command and see if result changes with different bios settings, for me it does.

 

In Ubuntu (using LiveCD) I get ff. And yes I have E0 version.

 

tried enabling and disabling C1E C2E C4E and Eist and I still get 05 on my hack.

 

I'm out of ideas

mm67
mm67
Posted
Posted
In Ubuntu (using LiveCD) I get ff. And yes I have E0 version.

 

tried enabling and disabling C1E C2E C4E and Eist and I still get 05 on my hack.

 

I'm out of ideas

 

I just did some testing on Ubuntu using different C - state settings. These are my results :

C1 only

    PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.0%)         2.84 Ghz     0.0%
polling           0.0ms ( 0.0%)         2.00 Ghz   100.0%
C1               47.4ms (100.3%)



Wakeups-from-idle per second : 21.1     interval: 20.0s
no ACPI power usage estimate available

Top causes for wakeups:
 28.8% ( 37.5)     <kernel core> : hrtimer_start (tick_sched_timer) 
 15.3% ( 19.9)     <kernel core> : timer_action (ehci_watchdog) 
  9.8% ( 12.8)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
  8.6% ( 11.2)       <interrupt> : ehci_hcd:usb1, uhci_hcd:usb5, uhci_hcd:usb8,
  7.8% ( 10.2)       <interrupt> : ahci 
  7.7% ( 10.0)     <kernel core> : ath_ani_calibrate (ath_ani_calibrate) 

C1 & C2 

  PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.2%)         2.84 Ghz     0.0%
polling           0.0ms ( 0.0%)         2.00 Ghz   100.0%
C1 halt           0.0ms ( 0.0%)
C2                8.7ms (99.8%)


Wakeups-from-idle per second : 114.2    interval: 15.0s
no ACPI power usage estimate available

Top causes for wakeups:
 31.4% ( 68.1)       <interrupt> : extra timer interrupt 
 20.1% ( 43.5)     <kernel core> : hrtimer_start (tick_sched_timer) 
 10.3% ( 22.3)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
  9.2% ( 20.0)     <kernel core> : timer_action (ehci_watchdog) 
  5.8% ( 12.6)       <interrupt> : ehci_hcd:usb1, uhci_hcd:usb5, uhci_hcd:usb8,
  4.6% ( 10.0)     <kernel core> : ath_ani_calibrate (ath_ani_calibrate) 

C1 & C2 & C4

    PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.2%)         2.84 Ghz     0.0%
polling           0.0ms ( 0.0%)         2.00 Ghz   100.0%
C1 halt           0.0ms ( 0.0%)
C2                9.1ms (98.2%)
C3                8.0ms ( 1.6%)

Wakeups-from-idle per second : 110.0    interval: 15.0s
no ACPI power usage estimate available

Top causes for wakeups:
 32.2% ( 68.2)       <interrupt> : extra timer interrupt 
 15.9% ( 33.7)     <kernel core> : hrtimer_start (tick_sched_timer) 
 14.3% ( 30.3)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
  9.4% ( 20.0)     <kernel core> : timer_action (ehci_watchdog) 
  6.1% ( 12.9)       <interrupt> : ehci_hcd:usb1, uhci_hcd:usb5, uhci_hcd:usb8,
  4.7% ( 10.0)     <kernel core> : ath_ani_calibrate (ath_ani_calibrate)

 

If I understand those results correctly using C4 is totally pointless since cpu spends 98.2 % of it's time in C2 anyway even if C4 is enabled. What do you think ?

leslieking
leslieking
Posted
Posted

Hi,

I'm reading page5 of "DSDT Fixes for Gigabyte" thread.. This is my PICM:

 

Name (PICM, Package (0x16)
{
   Package (0x04)
   {
       0x001BFFFF, 
       Zero, 
       LNK0, 
       Zero
   }, 

   Package (0x04)
   {
       0x0001FFFF, 
       Zero, 
       LNKA, 
       Zero
   }, 

   Package (0x04)
   {
       0x0001FFFF, 
       One, 
       LNKB, 
       Zero
   },
..
..
..

 

I have two questions:

 

1. I need to rename PICM to _PRT, right?

2. I have LNK0, LNKA, LNKB. mm67 and kdawg uses Zero instead of LNK0, LNKA, LNKB... Why ? I need to do the same?

FKA
FKA
Posted
Posted
Do you think it could even remotely be related to the fact you have a Q9550 that supports more C states than mine? If I had hair I'd have pulled it all out by now. Now I need to know why they work for you but not me. What's your DSDT look like these days? What bootloader? Are you using any bootloader device enablers? Which non-vanilla kexts are you using?

 

As soon as I install either of those apps on the virgin fresh install I loose sleep.

 

If you find anything out please let me know ..

Still using RIP script here for auto sleep. even without istats installed and with disk in my TSSTCROPS drive i have no auto sleep!

 

D.

rayap
rayap
Posted
Posted
I just did some testing on Ubuntu using different C - state settings. These are my results :

 

If I understand those results correctly using C4 is totally pointless since cpu spends 98.2 % of it's time in C2 anyway even if C4 is enabled. What do you think ?

 

These are my results. C2:C4 is about 50:50.

 

Only C1E enabled

    PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.8%)         2.67 Ghz   100.0%
polling           0.0ms ( 0.0%)         2.00 Ghz     0.0%
C1               14.3ms (99.2%)



Wakeups-from-idle per second : 69.2     interval: 5.0s
no ACPI power usage estimate available

Top causes for wakeups:
 29.1% ( 60.0)       <interrupt> : fglrx[0]@PCI:1:0:0 
 27.1% ( 55.8)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
 12.3% ( 25.4)      <kernel IPI> : Rescheduling interrupts 
  8.3% ( 17.2)     <kernel core> : hrtimer_start (tick_sched_timer) 
  5.3% ( 11.0)     <kernel core> : timer_action (ehci_watchdog) 
  4.1% (  8.4)      <kernel IPI> : TLB shootdowns

Suggestion: Enable SATA ALPM link power management via:
 echo min_power > /sys/class/scsi_host/host0/link_power_management_policy
or press the S key.
Q - Quit   R - Refresh   S - SATA Link Power Management 

C1E & C2E enabled

    PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.6%)         2.67 Ghz     0.0%
polling           0.0ms ( 0.0%)         2.00 Ghz   100.0%
C1 halt           0.0ms ( 0.0%)
C2                5.7ms (99.4%)


Wakeups-from-idle per second : 175.5    interval: 4.1s
no ACPI power usage estimate available

Top causes for wakeups:
 18.9% ( 69.8)   USB device 1-5.3 : Apple Optical USB Mouse (Fujitsu Takamisawa
 17.2% ( 63.5)      <kernel IPI> : Rescheduling interrupts 
 16.5% ( 61.0)       <interrupt> : fglrx[0]@PCI:1:0:0 
 16.3% ( 60.2)       <interrupt> : extra timer interrupt 
  8.1% ( 30.0)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
  5.5% ( 20.5)       <interrupt> : ehci_hcd:usb1, uhci_hcd:usb5, uhci_hcd:usb8,
ohci1394 



Q - Quit   R - Refresh  

C1E, C2E & C4E enabled.

    PowerTOP version 1.11      © 2007 Intel Corporation

Cn                Avg residency       P-states (frequencies)
C0 (cpu running)        ( 0.3%)         2.67 Ghz     0.0%
polling           0.0ms ( 0.0%)         2.00 Ghz   100.0%
C1 halt           0.0ms ( 0.0%)
C2                4.9ms (46.7%)
C3                7.8ms (53.0%)

Wakeups-from-idle per second : 163.4    interval: 1.1s
no ACPI power usage estimate available

Top causes for wakeups:
 31.2% (140.0)   USB device 1-5.3 : Apple Optical USB Mouse (Fujitsu Takamisawa
 15.4% ( 69.0)       <interrupt> : fglrx[0]@PCI:1:0:0 
 14.1% ( 63.0)       <interrupt> : extra timer interrupt 
  9.2% ( 41.0)      <kernel IPI> : Rescheduling interrupts 
  8.5% ( 38.0)     <kernel core> : hrtimer_start_range_ns (tick_sched_timer) 
  6.0% ( 27.0)     <kernel core> : hrtimer_start (tick_sched_timer) 




Q - Quit   R - Refresh

kdawg
kdawg
Posted
Posted
If you find anything out please let me know ..

Still using RIP script here for auto sleep. even without istats installed and with disk in my TSSTCROPS drive i have no auto sleep!

 

D.

Did you previously have iStats installed? If so try reinstalling it then using the uninstall script again. There were remnants that the uninstaller script didn't get the firt time around for me.

 

Use 

pmset -g log

to view pmset entries.

 

Entries like this are problems. They may not tell you exactly what the problem is but it might help you track them down.

 * Domain: applicationresponse.slowresponse
- Message: PMConnection IPConfiguration com.apple.powermanagement.applicationresponse.slowresponse 2990 ms
- Time: 1/3/10 10:15:17 AM EST 
- Signature: IPConfiguration
- UUID: FBD38D98-62A3-4B17-84D7-0F63A1991621
- Result: Noop
- Response time (ms): 2990

 Share
Go to topic listing
×
×
  • Create New...