G73SW ML 10.8.2 Install

jeffc2009 · December 20, 2012

Can u send me a copy of your ioreg?

ellnic · December 20, 2012

Yes of course. See attached. I dumped 2 of them, one had significantly more info than the other.

I can get you information from AIDA64 in win as well, but I was just about to boot into Linux, I will attach the IOreg from there as well, incase they differ.

EDIT: Ok, really disappointed in a way to find out that ubuntu "just worked" OTB with the backlit keys...booted from the live 12.10 disc and they lit up. It also supported the fn+F4 and F5 and it displays a little level indicator like the brightness. Didn't realise that Linux didn't have IOreg, so haven't attached that - but will grab the ACPI tables from AIDA64 and attach shortly.

ellnic_macos.zip

jeffc2009 · December 20, 2012

My NVDA, Display-A@0

   | | | +-o NVDA,Display-A@0  <class IONDRVDevice, id 0x1000002ba, registered, matched, active, busy 0 (36 ms), retain 8>
   | | | |   {
   | | | |	 "nativeNDRV" = <6164766e>
   | | | |	 "IODeviceMemory" = (({"address"=0xf2000000,"length"=0x1000000}),({"address"=0xf80000000,"length"=0x8000000}),({"address"=0xf88000000,"length"=0x2000000}),"IOSubMemoryDescriptor is not serializable",({"address"=0xf4000000,"length"=0x80000}))
   | | | |	 "AAPL,gray-page" = <01000000>
   | | | |	 "AAPL,ndrv-interrupt-set" = "IONDRVInterruptSet is not serializable"
   | | | |	 "built-in" = <01>
   | | | |	 "noNVRAM" = <01000000>
   | | | |	 "IOFBDependentID" = 0xffffff8011af3000
   | | | |	 "display-cfg" = <03010300>
   | | | |	 "driver-ist" = <80d14b1380ffffff010000000000000080d14b1380ffffff020000000000000080d14b1380ffffff0300000000000000>
   | | | |	 "name" = <4e5644412c446973706c61792d4100>
   | | | |	 "NVDA,SubDeviceClass" = <80200000>
   | | | |	 "AAPL,ndrv-dev" = Yes
   | | | |	 "compatible" = <4e5644412c4e564d616300>
   | | | |	 "display-connect-flags" = <00000000>
   | | | |	 "av-signal-type" = <00000000>
   | | | |	 "AAPL,iokit-ndrv" = <20cd90807fffffff>
   | | | |	 "IOFBDependentIndex" = 0x0
   | | | |	 "NVDA,DeviceClass" = <80000000>
   | | | |	 "AAPL,gray-value" = <c38c6400>
   | | | |	 "port-number" = <00000000>
   | | | |	 "AAPL,RegEntryID" = <00dc641280ffffffff239bed7f000000656d6953000000007a6c755200000000>
   | | | |	 "device_type" = <646973706c617900>
   | | | |	 "AAPL,boot-display" = <01000000>
   | | | |   }
   | | | |  
   | | | +-o NVDA,Display-B@1  <class IONDRVDevice, id 0x1000002bb, registered, matched, active, busy 0 (1 ms), retain 8>

Ellnic, NVDA, Display-A@0

   | | | +-o NVDA,Display-A@0  <class IONDRVDevice, id 0x1000002e7, registered, matched, active, busy 0 (1268 ms), retain 8>
   | | | |   {
   | | | |	 "AAPL,ndrv-interrupt-set" = "IONDRVInterruptSet is not serializable"
   | | | |	 "av-signal-type" = <10000000>
   | | | |	 "driver-ist" = <c0fa111e80ffffff0100000000000000c0fa111e80ffffff0200000000000000c0fa111e80ffffff0300000000000000>
   | | | |	 "IOFBDependentIndex" = 0
   | | | |	 "noNVRAM" = <01000000>
   | | | |	 "link8257-signal-amplitude" = <08000000>
   | | | |	 "link4161-signal-amplitude" = <08000000>
   | | | |	 "link8257-pre-emphasis-level" = <00000000>
   | | | |	 "edid" = <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>
   | | | |	 "display-connect-flags" = <00000000>
   | | | |	 "AAPL,gray-value" = <c38c6400>
   | | | |	 "display-cfg" = <03010300>
   | | | |	 "AAPL,boot-display" = <01000000>
   | | | |	 "link4161-pre-emphasis-level" = <00000000>
   | | | |	 "IODeviceMemory" = (({"address"=4060086272,"length"=16777216}),({"address"=66571993088,"length"=134217728}),({"address"=66706210816,"length"=33554432}),"IOSubMemoryDescriptor is not serializable",({"address"=4093640704,"length"=524288}))
   | | | |	 "AAPL,RegEntryID" = <0054331e80ffffffffabcce17f000000656d6953000000007a6c755200000000>
   | | | |	 "dp-device-id" = <41000000>
   | | | |	 "name" = <"NVDA,Display-A">
   | | | |	 "IOFBDependentID" = 18446743524384146944
   | | | |	 "AAPL,ndrv-dev" = Yes
   | | | |	 "AAPL,gray-page" = <01000000>
   | | | |	 "link8257-bit-rate" = <01000000>
   | | | |	 "NVDA,DeviceClass" = <80000000>
   | | | |	 "link4161-bit-rate" = <01000000>
   | | | |	 "link8257-num-lanes" = <02000000>
   | | | |	 "device_type" = <"display">
   | | | |	 "compatible" = <"NVDA,NVMac">
   | | | |	 "NVDA,SubDeviceClass" = <80200000>
   | | | |	 "link4161-downspread-amplitude" = <01000000>
   | | | |	 "nativeNDRV" = <"advn">
   | | | |	 "link8257-downspread-amplitude" = <01000000>
   | | | |	 "link4161-num-lanes" = <02000000>
   | | | |	 "port-number" = <01000000>
   | | | |	 "AAPL,iokit-ndrv" = <206de2807fffffff>
   | | | |   }
   | | | |  
   | | | +-o NVDA,Display-B@1  <class IONDRVDevice, id 0x1000002e8, registered, matched, active, busy 0 (0 ms), retain 8>

ellnic · December 20, 2012

Correct me if I am wrong, but does this mean my GPU is different to yours in minor ways? Cos a lot of that looks completely different to one another. A lot of this goes over my head.

jeffc2009 · December 20, 2012

are you injecting anything in your dsdt for GFX@0 like this? Bootloader?

{
						 "AAPL,slot-name",
						 "PCI x16",
						 "@0,compatible",
						 Buffer (0x0B)
						 {
							 "NVDA,NVMac"
						 },

						 "@0,connector-type",
						 Buffer (0x04)
						 {
							 0x00, 0x08, 0x00, 0x00
						 },

						 "@0,device_type",
						 Buffer (0x08)
						 {
							 "display"
						 },
						 "@0,name",
						 Buffer (0x0F)
						 {
							 "NVDA,Display-A"
						 },
						 "@1,compatible",
						 Buffer (0x0B)
						 {
							 "NVDA,NVMac"
						 },

						 "@1,connector-type",
						 Buffer (0x04)
						 {
							 0x00, 0x08, 0x00, 0x00
						 },								
						 "@1,device_type",
						 Buffer (0x08)
						 {
							 "display"
						 },
						 "@1,name",
						 Buffer (0x0F)
						 {
							 "NVDA,Display-B"
						 },
						 "NVCAP",
						 Buffer (0x18)
						 {
							 /* 0000 */ 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x03, 0x00,
							 /* 0008 */ 0x1c, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0A,
							 /* 0010 */ 0x00, 0x00, 0x00, 0x00
						 },
						 "VRAM,totalsize",
						 Buffer (0x04)
						 {
							 0x00, 0x00, 0x00, 0x40
						 },
						 "device_type",
						 Buffer (0x0C)
						 {
							 "NVDA,Parent"
						 },
						 "model",
						 Buffer (0x10)
						 {
							 "GeForce GTX 460"
						 },
						 "rom-revision",
						 Buffer (0x0e)
						 {
							 "70.08.29.00.30"
						 },

Correct me if I am wrong, but does this mean my GPU is different to yours in minor ways? Cos a lot of that looks completely different to one another. A lot of this goes over my head.

We got the same exact chip.

yours:

"subsystem-vendor-id" = <43100000>
| | | | "vendor-id" = <de100000>

My:

| | | | "subsystem-vendor-id" = <43100000>
| | | | "vendor-id" = <de100000>

After some DSDT edit's: look for backlight-control

| | | +-o NVDA,Display-A@0 <class IONDRVDevice, id 0x1000002d1, registered, matched, active, busy 0 (1490 ms), retain 8>
| | | | {
| | | |	 "backlight-control" = <01000000>
| | | |	 "nativeNDRV" = <6164766e>
| | | |	 "IODeviceMemory" = (({"address"=0xf2000000,"length"=0x1000000}),({"address"=0xf80000000,"length"=0x8000000}),({"address"=0xf88000000,"length"=0x2000000}),"IOSubMemoryDescriptor is not serializable",({"address"=0xf4000000,"length"=0x80000}))
| | | |	 "AAPL,gray-page" = <01000000>
| | | |	 "AAPL,ndrv-interrupt-set" = "IONDRVInterruptSet is not serializable"
| | | |	 "built-in" = <01>
| | | |	 "connector-type" = <00080000>
| | | |	 "noNVRAM" = <01000000>
| | | |	 "IOFBDependentID" = 0xffffff801210c300
| | | |	 "display-cfg" = <03010300>
| | | |	 "driver-ist" = <c03e7e1780ffffff0100000000000000c03e7e1780ffffff0200000000000000c03e7e1780ffffff0300000000000000>
| | | |	 "name" = <4e5644412c446973706c61792d4100>
| | | |	 "NVDA,SubDeviceClass" = <80200000>
| | | |	 "AAPL,ndrv-dev" = Yes
| | | |	 "compatible" = <4e5644412c4e564d616300>
| | | |	 "display-connect-flags" = <00000000>
| | | |	 "av-signal-type" = <00000000>
| | | |	 "AAPL,iokit-ndrv" = <20bdd5807fffffff>
| | | |	 "IOFBDependentIndex" = 0x0
| | | |	 "AAPL,gray-value" = <c38c6400>
| | | |	 "NVDA,DeviceClass" = <80000000>
| | | |	 "port-number" = <00000000>
| | | |	 "AAPL,RegEntryID" = <0038731680ffffffffc78ce97f000000656d6953000000007a6c755200000000>
| | | |	 "device_type" = <646973706c617900>
| | | |	 "AAPL,boot-display" = <01000000>
| | | | }

noticed my port number and av-signal-type is different from yours!

ellnic · December 21, 2012

I'm not doing anything that you haven't described in the first post. I have been following your instructions. However, I have some weird news... Since booting into Ubuntu 12.10, my backlit keys have started working...... All I did was, boot into Ubuntu, play around with fn+F4 and fn+F5, set to max, boot back to ML, then sleep and wake. This really perpelexes me... Do you think Ubuntu has saved something to CMOS? Have you ever booted from a Linux disc? I've only done it in a VM on this machine prior to just now... Might I suggest all those who aren't getting backlit keys try this

Now we just need to fix the issue with sleep on low battery so our hard drives don't die on us... found this: http://www.insanelymac.com/forum/topic/213310-shutdown-on-low-battery/

And that thread gives this: http://www.jinx.de/SmartSleep.html

jeffc2009 · December 21, 2012

have you tried to pull your battery and boot back into osx to check backlit?

Download this and mount efi. Let me know if there's anything in there.

http://www.mediafire...qhcvoj29ll935xr

Getting closer. All without

AsusNBWMI.kext.

Brightness.tiff

ellnic · December 21, 2012

Getting closer. All without

AsusNBWMI.kext.

Brightness.tiff

NICE

Nothing in my EFI partition. Pulling the battery didn't seem to work, it's only worked since Ubuntu. Now it works all the time after sleep.

EDIT: Do you have fan speed control?

EDIT2: Sorry, I get what you mean, pull the battery and see if it still works after. Will report back in a mo.

EDIT3: Pulling the battery had no effect. The backlit keys still work after sleep. Thinking about it, it can't be OS dependant - because backlit keys now work in both my installs (and didn't before) so I know it must be something to do with the CMOS/BIOS. Maybe a glitch yours didn't have? Asus probably don't know about it because they only test Win/Linux.

Hugo_bee · December 21, 2012

You don't need to stay with 8, 2 to have sleep/wake. I think you are staying with 8, 2 to have graphics powermanagement right? my new setup is on 9, 1. Once I have it tested out, I am going to change OP.

I know sleep/wake was a issue before ML ( i think it had started with Lion/that's why I had left it ), but now it's fixed with the latest apple/nvidia kext's. You have so many patches on your dsdt, have no idea where to start! Do you still have a fresh copy of your windows extracted dsdt? Kernel is from MacBook Air and MacBook Pro Update 2.0.

i tested a lot of smbios.. but 8,2 matches and invoke the sleep. if I change it.. the cpu still run

9,1 in original apple macbookpro is an i7 IVy bridge,, not sandy bridge

confirm ubuntu enables the keyboard backlight. so at reboot is still turned on!

edit: i'm trying to enable brightness.. I have the control bar but it's not working.

in the log

Dec 22 00:26:15 localhost kernel[0]: ACPIBacklightPanel: ACPI Methods _DOS _BCL _BCM _BQC found. Device path: _SB.PNLF

Dec 22 00:26:15 localhost kernel[0]: ACPIBacklightPanel: getIndexForLevel(-246) not found in _BCL table !

still.. working

Hugo_bee · December 22, 2012

I can bypass the indexing error but I can't engage the brightness bar

my opinion is that the fn+f3 or f4 are so well described in the dsdt.. and I think that the system passes those commands.

but I suppose there's not a real substrate that passes the command through the IO as happen in other fn commands.

What happens in ubuntu or windows depends on the kernel module wmi.

but i'm not sure.. I tried to manage the source of iowmi.. but I have error on the compile.. I think it miss completely the part about backlight keybiard..

that's because all the external forum tell a map of the ability of the iowmi kext.. and all miss the backlight kbd. !

so .. maybe.. the real work is to add the missing code in the iowmi xcode project.

AsusNBWMI-master.zip

jeffc2009 · December 22, 2012

i tested a lot of smbios.. but 8,2 matches and invoke the sleep. if I change it.. the cpu still run

9,1 in original apple macbookpro is an i7 IVy bridge,, not sandy bridge

confirm ubuntu enables the keyboard backlight. so at reboot is still turned on!

edit: i'm trying to enable brightness.. I have the control bar but it's not working.

in the log

Dec 22 00:26:15 localhost kernel[0]: ACPIBacklightPanel: ACPI Methods _DOS _BCL _BCM _BQC found. Device path: _SB.PNLF

Dec 22 00:26:15 localhost kernel[0]: ACPIBacklightPanel: getIndexForLevel(-246) not found in _BCL table !

still.. working

On our system is structured different from hotKoffy's system. We need to change DSDT to match ours. I have been reading a lot today trying to figure out how DSDT on G73 works. I think Method GLKB to control Keyboard Light level. Looking for new method for backlight.

Hugo_bee · December 22, 2012

indeed.

sleeping brang me the same idea.

the INI/REG acpi method in the EC registare can force some behaviour at boot.

If the soft reboot from ubuntu to mac makes the light untouched .. and if the wake did the same.. is for the same reason:

the EC register has a bit with the info of the value of the keyboard light. And the DSDT tell to BIOS to RESTORE the value that EC has before the sleep.

In fact the WAKE procedure reads all the prvious states and restore them.. and it invoke the ATKD backlight.

so: my idea is to manually create the bit in the EC INI/REG section of the DSDT.

if so.. at boot you'll always have the light turned on.

manmal · December 22, 2012

nice news for the backlit keyboard! So it's enough i run linux ubuntu (downloadable for free) from virtualbox (free) and after that keyboard works on ML after resuming from sleep?

about batterysave / smartsleep , with other laptops i had there was nothing to do when battery is near to zero. please try it and let us know if it works as it should (even if you reach 0% level on your battery...here usually when it is near to zero, it begins to startup again, then power off, then poweron 4-5 times until the battery discharges completely: quite dangerous for harddrives!!

FOR SD cards, perhaps in ubuntu we may check if sd card works (booting from true ubuntu cd, not vm...or in vm too? dunno) and investigating about the driver ... the source may be ported to osx! does anyone have installed xcode here to import / adapt the source and try?

ellnic · December 22, 2012

indeed.

sleeping brang me the same idea.

the INI/REG acpi method in the EC registare can force some behaviour at boot.

If the soft reboot from ubuntu to mac makes the light untouched .. and if the wake did the same.. is for the same reason:

the EC register has a bit with the info of the value of the keyboard light. And the DSDT tell to BIOS to RESTORE the value that EC has before the sleep.

so: my idea is to manually create the bit in the EC INI/REG section of the DSDT.

if so.. at boot you'll always have the light turned on.

You have a point - since I have got my backlit working, if I sleep then wake (so it comes on) I can reboot and it stays lit through the boot process. Only shutting down entirely turns it off. Or, I guess, loading another OS that calls a 'proper control'.

nice news for the backlit keyboard! So it's enough i run linux ubuntu (downloadable for free) from virtualbox (free) and after that keyboard works on ML after resuming from sleep?

FOR SD cards, perhaps in ubuntu we may check if sd card works (booting from true ubuntu cd, not vm...or in vm too? dunno) and investigating about the driver ... the source may be ported to osx! does anyone have installed xcode here to import / adapt the source and try?

It won't work in a VM, because the host OS needs to be completely aware of the hardware in order to pass it to the guest OS. Boot natively into Ubuntu 12.10 and your keyboard should start working after sleep.

The SD card is an entirely different matter, I think we should put that to one side for the time being.

Hugo_bee · December 22, 2012

ok

with RW everything I read the

9C cell in the EC register

it changes from 00/55/aa/ff

in the 4 states

in the dsdt it's the var TSTP

and it's in the RAMB register (volatile 1) and in the EC register (volatile 0)

TSTP, 8,

Offset (0x9C)

in the LCDD (under gfx0) device there is an external entry of EC for that specific control:

 Method (_QAA, 0, NotSerialized)
 {
	 Store (One, ^^^^TSTP)
	 ISMI (0xA7)
	 Store (One, OEMP)
	 PNOT ()
 }

 Method (_QA6, 0, NotSerialized)
 {
	 Store (0x02, ^^^^TSTP)
	 ISMI (0xA7)
 }

 Method (_QA8, 0, NotSerialized)
 {
	 Store (0x03, ^^^^TSTP)
	 ISMI (0xA7)
	 Store (One, OEMP)
	 PNOT ()
 }

these (the _Q methods) are theQUERY : what happens if you press some in the direct connection with bios.. or what happens when you detach a battery.. etc..

the ISMI is locate under the root.. and point to another bank of ram/register

OperationRegion (GPSC, SystemIO, 0xB2, 0x02)
    Field (GPSC, ByteAcc, NoLock, Preserve)
    {
	    SMCM,   8,
	    SMST,   8
    }
    Method (ISMI, 1, Serialized)
    {
	    Store (Arg0, SMCM)
    }

in the WAKE.. the 1st operation is to write ISMI xxxx.. so I suppose that if the value of the ISMI is the same in the 3 query.. is not important or is a flag to tell that the same family of query has changed it's value

the 3 query stores 1, 2, 3 in the TSTP register. and the 4?

I suppose this is not the command but the only a query. because the 4 values 00/55/aa/ff are quite different from 1,2,3.

infact I discovered the under the DEVICE ATKD

Method (GLKB, 1, NotSerialized)

with the 1,2,3 and it STORE that value in the KBLV var.

and

	    Name (PWKB, Buffer (0x04)
	    {
		    0x00, 0x55, 0xAA, 0xFF
	    })
	    Method (SLKB, 1, NotSerialized)
	    {
		    Store (And (Arg0, 0x7F), KBLV)
		    If (And (Arg0, 0x80))
		    {
			    Store (DerefOf (Index (PWKB, KBLV)), Local0)
		    }
		    Else
		    {
			    Store (Zero, Local0)
		    }

		    ^^PCI0.LPCB.EC.WRAM (0x044B, Local0)
		    Return (One)
	    }

this is the core!

in fact it use the INDEX of KBLV (1,2,3) and get the relative value from the array.

the comand is the last row: it write in the EC.WRAM the proper value in the reg 044B

windows and ubuntu can translate the Fn+key with that command.

If we can change the xproj of the wmi.. we can add a command for the missing f3/f4 unmanaged buttons.. and for the actual needing.. we can invoke that command in the boot.

how?

the Sleep (Method (ECS, 1, NotSerialized)) set to zero the light.. WRAM (0x044B, Zero) after writing the last value Store (RRAM (0x0449), Local0)

and the wake read the value in the RRAM

Store (RRAM (0x0449), Local0)

and fix the value only if we are in a wake:

If (LOr (LEqual (Arg0, 0x03), LEqual (Arg0, 0x04)))
	    {
		    Store (RRAM (0x0449), Local0)
		    And (Local0, 0x02, Local1)
		    If (Local1)
		    {
			    Or (Local0, One, Local0)
			    WRAM (0x0449, Local0)
		    }
	    }

so we're sure.. the WRAM 0x0449 fixes the value in a separate bank.

WRAM 044B fix the value in the live system

i'll try

WRAM (0x044B, 0xFF)

i'ss test the ini and reg EC

see ya

yeah

here you are

this is for g53

search for the text

//**

grab the text i've added in the 3 positions ad put in your dsdt.dsl in the same position

now we've to mod the kext!!

latestdsdt.zip

jeffc2009 · December 22, 2012

ok

with RW everything I read the

9C cell in the EC register

it changes from 00/55/aa/ff

in the 4 states

in the dsdt it's the var TSTP

and it's in the RAMB register (volatile 1) and in the EC register (volatile 0)

TSTP, 8,

Offset (0x9C)

in the LCDD (under gfx0) device there is an external entry of EC for that specific control:
 Method (_QAA, 0, NotSerialized)
 {
	 Store (One, ^^^^TSTP)
	 ISMI (0xA7)
	 Store (One, OEMP)
	 PNOT ()
 }

 Method (_QA6, 0, NotSerialized)
 {
	 Store (0x02, ^^^^TSTP)
	 ISMI (0xA7)
 }

 Method (_QA8, 0, NotSerialized)
 {
	 Store (0x03, ^^^^TSTP)
	 ISMI (0xA7)
	 Store (One, OEMP)
	 PNOT ()
 }
these (the _Q methods) are theQUERY : what happens if you press some in the direct connection with bios.. or what happens when you detach a battery.. etc..

the ISMI is locate under the root.. and point to another bank of ram/register
OperationRegion (GPSC, SystemIO, 0xB2, 0x02)
 Field (GPSC, ByteAcc, NoLock, Preserve)
 {
	 SMCM, 8,
	 SMST, 8
 }
 Method (ISMI, 1, Serialized)
 {
	 Store (Arg0, SMCM)
 }
in the WAKE.. the 1st operation is to write ISMI xxxx.. so I suppose that if the value of the ISMI is the same in the 3 query.. is not important or is a flag to tell that the same family of query has changed it's value

the 3 query stores 1, 2, 3 in the TSTP register. and the 4?

I suppose this is not the command but the only a query. because the 4 values 00/55/aa/ff are quite different from 1,2,3.

infact I discovered the under the DEVICE ATKD

Method (GLKB, 1, NotSerialized)

with the 1,2,3 and it STORE that value in the KBLV var.

and
	 Name (PWKB, Buffer (0x04)
	 {
		 0x00, 0x55, 0xAA, 0xFF
	 })
	 Method (SLKB, 1, NotSerialized)
	 {
		 Store (And (Arg0, 0x7F), KBLV)
		 If (And (Arg0, 0x80))
		 {
			 Store (DerefOf (Index (PWKB, KBLV)), Local0)
		 }
		 Else
		 {
			 Store (Zero, Local0)
		 }

		 ^^PCI0.LPCB.EC.WRAM (0x044B, Local0)
		 Return (One)
	 }
this is the core!

in fact it use the INDEX of KBLV (1,2,3) and get the relative value from the array.

the comand is the last row: it write in the EC.WRAM the proper value in the reg 044B

windows and ubuntu can translate the Fn+key with that command.

If we can change the xproj of the wmi.. we can add a command for the missing f3/f4 unmanaged buttons.. and for the actual needing.. we can invoke that command in the boot.

how?

the Sleep (Method (ECS, 1, NotSerialized)) set to zero the light.. WRAM (0x044B, Zero) after writing the last value Store (RRAM (0x0449), Local0)

and the wake read the value in the RRAM

Store (RRAM (0x0449), Local0)

and fix the value only if we are in a wake:
If (LOr (LEqual (Arg0, 0x03), LEqual (Arg0, 0x04)))
	 {
		 Store (RRAM (0x0449), Local0)
		 And (Local0, 0x02, Local1)
		 If (Local1)
		 {
			 Or (Local0, One, Local0)
			 WRAM (0x0449, Local0)
		 }
	 }
so we're sure.. the WRAM 0x0449 fixes the value in a separate bank.

WRAM 044B fix the value in the live system

i'll try

WRAM (0x044B, 0xFF)

i'ss test the ini and reg EC

see ya

yeah

here you are

this is for g53

search for the text

//**

grab the text i've added in the 3 positions ad put in your dsdt.dsl in the same position

now we've to mod the kext!!

Good find Ugokind, I will test it when i get back. For PNLF, Method GCBL and it's Storing to ECO.STBR.

For Backlit Keyboard Why no store that value under SLKB? This is the last argument before it is invoked by ATK. Since we don't have ATK software to trigger it, just leave it on.

Method (SLKB, 1, NotSerialized)

{

Store (And (Arg0, 0x7F), KBLV)

If (And (Arg0, 0x80))

{

Store (DerefOf (Index (PWKB, KBLV)), Local0)

}

Else

{

Store (Zero, Local0)

}

^^PCI0.SBRG.EC0.WRAM (0x044B, Local0)

Return (One)

}

I think Method _Q0E is how example: (Fn-F5) down backlight stores to EC and so on...

Hugo_bee · December 22, 2012

My fixes works.

But the light is always turned on

Why not to engage a team to mod the wmikext source that i posted?

jeffc2009 · December 22, 2012

My fixes works.

But the light is always turned on

Why not to engage a team to mod the wmikext source that i posted?

We can do that too but I don't know anything about Xcode. Didi you get that from EMlyDinEsH? Let me do more reading!

We can do that too but I don't know anything about Xcode. Didi you get that from EMlyDinEsH? Let me do more reading!

Just need to find the right code for FN-F5/FN-F6 for back light

FN-F3/FN-F4 for backlit keys.

#include "AsusWMIController.h"

#define DEBUG_START 0

#if DEBUG_START

#define DEBUG_LOG(fmt, args...) IOLog(fmt, ## args)

#else

#define DEBUG_LOG(fmt, args...)

#endif

#define ASUS_WMI_MGMT_GUID "97845ED0-4E6D-11DE-8A39-0800200C9A66"

#define ASUS_NB_WMI_EVENT_GUID "0B3CBB35-E3C2-45ED-91C2-4C5A6D195D1C"

/* WMI Methods */

#define ASUS_WMI_METHODID_SPEC 0x43455053 /* BIOS SPECification */

#define ASUS_WMI_METHODID_SFBD 0x44424653 /* Set First Boot Device */

#define ASUS_WMI_METHODID_GLCD 0x44434C47 /* Get LCD status */

#define ASUS_WMI_METHODID_GPID 0x44495047 /* Get Panel ID?? (Resol) */

#define ASUS_WMI_METHODID_QMOD 0x444F4D51 /* Quiet MODe */

#define ASUS_WMI_METHODID_SPLV 0x4C425053 /* Set Panel Light Value */

#define ASUS_WMI_METHODID_SFUN 0x4E554653 /* FUNCtionalities */

#define ASUS_WMI_METHODID_SDSP 0x50534453 /* Set DiSPlay output */

#define ASUS_WMI_METHODID_GDSP 0x50534447 /* Get DiSPlay output */

#define ASUS_WMI_METHODID_DEVP 0x50564544 /* DEVice Policy */

#define ASUS_WMI_METHODID_OSVR 0x5256534F /* OS VeRsion */

#define ASUS_WMI_METHODID_DSTS 0x53544344 /* Device STatuS */

#define ASUS_WMI_METHODID_DSTS2 0x53545344 /* Device STatuS #2*/

#define ASUS_WMI_METHODID_BSTS 0x53545342 /* Bios STatuS ? */

#define ASUS_WMI_METHODID_DEVS 0x53564544 /* DEVice Set */

#define ASUS_WMI_METHODID_CFVS 0x53564643 /* CPU Frequency Volt Set */

#define ASUS_WMI_METHODID_KBFT 0x5446424B /* KeyBoard FilTer */

#define ASUS_WMI_METHODID_INIT 0x54494E49 /* INITialize */

#define ASUS_WMI_METHODID_HKEY 0x59454B48 /* Hot KEY ?? */

#define ASUS_WMI_UNSUPPORTED_METHOD 0xFFFFFFFE

/* Wireless */

#define ASUS_WMI_DEVID_HW_SWITCH 0x00010001

#define ASUS_WMI_DEVID_WIRELESS_LED 0x00010002

#define ASUS_WMI_DEVID_CWAP 0x00010003

#define ASUS_WMI_DEVID_WLAN 0x00010011

#define ASUS_WMI_DEVID_BLUETOOTH 0x00010013

#define ASUS_WMI_DEVID_GPS 0x00010015

#define ASUS_WMI_DEVID_WIMAX 0x00010017

#define ASUS_WMI_DEVID_WWAN3G 0x00010019

#define ASUS_WMI_DEVID_UWB 0x00010021

/* Leds */

/* 0x000200XX and 0x000400XX */

#define ASUS_WMI_DEVID_LED1 0x00020011

#define ASUS_WMI_DEVID_LED2 0x00020012

#define ASUS_WMI_DEVID_LED3 0x00020013

#define ASUS_WMI_DEVID_LED4 0x00020014

#define ASUS_WMI_DEVID_LED5 0x00020015

#define ASUS_WMI_DEVID_LED6 0x00020016

/* Backlight and Brightness */

#define ASUS_WMI_DEVID_BACKLIGHT 0x00050011

#define ASUS_WMI_DEVID_BRIGHTNESS 0x00050012

#define ASUS_WMI_DEVID_KBD_BACKLIGHT 0x00050021

#define ASUS_WMI_DEVID_LIGHT_SENSOR 0x00050022 /* ?? */ (we don't have this)

/* Misc */

#define ASUS_WMI_DEVID_CAMERA 0x00060013

/* Storage */

#define ASUS_WMI_DEVID_CARDREADER 0x00080013

/* Input */

#define ASUS_WMI_DEVID_TOUCHPAD 0x00100011

#define ASUS_WMI_DEVID_TOUCHPAD_LED 0x00100012

/* Fan, Thermal */

#define ASUS_WMI_DEVID_THERMAL_CTRL 0x00110011

#define ASUS_WMI_DEVID_FAN_CTRL 0x00110012

/* Power */

#define ASUS_WMI_DEVID_PROCESSOR_STATE 0x00120012

/* DSTS masks */

#define ASUS_WMI_DSTS_STATUS_BIT 0x00000001

#define ASUS_WMI_DSTS_UNKNOWN_BIT 0x00000002

#define ASUS_WMI_DSTS_PRESENCE_BIT 0x00010000

#define ASUS_WMI_DSTS_USER_BIT 0x00020000

#define ASUS_WMI_DSTS_BIOS_BIT 0x00040000

#define ASUS_WMI_DSTS_BRIGHTNESS_MASK 0x000000FF

#define ASUS_WMI_DSTS_MAX_BRIGTH_MASK 0x0000FF00

/*

* <platform>/ - debugfs root directory

* dev_id - current dev_id

* ctrl_param - current ctrl_param

* method_id - current method_id

* devs - call DEVS(dev_id, ctrl_param) and print result

* dsts - call DSTS(dev_id) and print result

* call - call method_id(dev_id, ctrl_param) and print result

*/

#define EEEPC_WMI_METHODID_SPEC 0x43455053

#define EEEPC_WMI_METHODID_DEVP 0x50564544

#define EEEPC_WMI_METHODID_DEVS 0x53564544

#define EEEPC_WMI_METHODID_DSTS 0x53544344

#define EEEPC_WMI_METHODID_CFVS 0x53564643

#define EEEPC_WMI_DEVID_BACKLIGHT 0x00050011

#define EEEPC_WMI_DEVID_BACKLIGHT2 0x00050012

#define EEEPC_WMI_DEVID_BLUETOOTH 0x00010013

#define EEEPC_WMI_DEVID_WIRELESS 0x00010011

#define EEEPC_WMI_DEVID_TRACKPAD 0x00100011

#define super AsusNBWMI

OSDefineMetaClassAndStructors(AsusWMIController, AsusNBWMI)

bool AsusWMIController::init(OSDictionary *dictionary)

{

return super::init(dictionary);

}

bool AsusWMIController::start(IOService *provider)

{

return super::start(provider);

}

void AsusWMIController::stop(IOService *provider)

{

super::stop(provider);

}

void AsusWMIController::free(void)

{

super::free();

}

IOService * AsusWMIController::probe(IOService *provider, SInt32 *score )

{

IOService * ret = NULL;

OSObject * obj;

OSString * name;

IOACPIPlatformDevice *dev;

do

{

if (!super::probe(provider, score))

continue;

dev = OSDynamicCast(IOACPIPlatformDevice, provider);

if (NULL == dev)

continue;

dev->evaluateObject("_UID", &obj);

name = OSDynamicCast(OSString, obj);

if (NULL == name)

continue;

if (name->isEqualTo("ATK"))

{

*score +=20;

ret = this;

}

name->release();

}

while (false);

return (ret);

}

[b]const wmiKeyMap AsusWMIController::keyMap[] = {

{0x30, NX_KEYTYPE_SOUND_UP, "NX_KEYTYPE_SOUND_UP"},

{0x31, NX_KEYTYPE_SOUND_DOWN, "NX_KEYTYPE_SOUND_DOWN"},

{0x32, NX_KEYTYPE_MUTE, "NX_KEYTYPE_MUTE"},

{0xCC, NX_KEYTYPE_VIDMIRROR, "NX_KEYTYPE_VIDMIRROR"},

{0x40, NX_KEYTYPE_PREVIOUS, "NX_KEYTYPE_PREVIOUS"},

{0x41, NX_KEYTYPE_NEXT, "NX_KEYTYPE_NEXT"},

{0x45, NX_KEYTYPE_PLAY, "NX_KEYTYPE_PLAY"},

{0,0xFF,NULL}

};[/b]

Hugo_bee · December 22, 2012

this is what i got from a blog

yes is it

https://github.com/EMlyDinEsHMG/AsusNBWMI

but i got pne error in compile.. so this is exactly what we have to do

we have to replicate a behaviour for other keypress.. and hook the internal dsdt function

jeffc2009 · December 22, 2012

maybe we can get some guidance from EMlyDinEsH. Looks like the controls is there, but key press is not enabled. I am currently spending some quality time with family at six flags for the light show tonight. When I get done ill try to look into this later tonight.

this is what i got from a blog

yes is it

https://github.com/EMlyDinEsHMG/AsusNBWMI

but i got pne error in compile.. so this is exactly what we have to do

we have to replicate a behaviour for other keypress.. and hook the internal dsdt function

Can u post what u have done so far?

Hugo_bee · December 22, 2012

take it as it comes.. we've a spring .. but the world will be also tomorrow.

I've xcode for ml.. and i'm not a developer.. but a little bit.. yes.

so i'll check the code.. because recent xcode give errors about obsolete code.

then.. i'll check the events-keypress

jeffc2009 · December 23, 2012

I had recompile both ACPIBacklight.kext and AsusNBWMI.kext to newest codes. Give me sometime to add new features to it.

12/22/12 11:53:51.000 PM kernel[0]: AsusWMIController: Found WMI Device ATKD

12/22/12 11:53:51.000 PM kernel[0]: AsusWMIController: Hotkey Events Enabled

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::probe()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::findDevices()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::hasBacklightMethods()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: ACPI device PNLF has _BCL

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: ACPI device PNLF has _BCM

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: ACPI device PNLF has _BQC

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: PNLF has backlight Methods

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: ACPI Methods _DOS _BCL _BCM _BQC found. Device path: _SB.PNLF

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: probe(devices found : true)

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: PNLF has backlight Methods

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::start()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::findDevices()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::getDeviceControl()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: BIOS control disabled: _DOS

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::hasSAVEMethod()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::setupIndexedLevels()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::queryACPISupportedBrightnessLevels()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: PNLF _BCL 11

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getIndexForLevel(100) is 10

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getIndexForLevel(80) is 9

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::queryACPICurentBrightnessLevel()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: queryACPICurentBrightnessLevel _BQC = 10

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: queryACPICurentBrightnessLevel returning 10

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getIndexForLevel(10) not found in _BCL table !

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::getACStatus()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel::getBatteryDevice()

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getBatteryDevice() serviceMatching OK

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getBatteryDevice() iter OK

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getBatteryDevice() bat is of class AppleSmartBattery

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: getACStatus() AC is 1

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: min = 0, max = 10, value = 10

12/22/12 11:53:52.000 PM kernel[0]: ACPIBacklightPanel: Version 1.2

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel::setDisplay()

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel::doUpdate()

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel: ACPILevel min 0, max 10, value 10

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel::doUpdate()

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel: ACPILevel min 0, max 10, value 10

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel::doIntegerSet(brightness)

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel::setACPIBrightnessLevel()

12/22/12 11:53:55.000 PM kernel[0]: ACPIBacklightPanel: setACPIBrightnessLevel _BCM(100)