[GUIDE] Making a DSDT.aml for Dell XPS M1330, XPS M1530, and XPS M1730

[Anjan Tek]

Its right the DSDT_M1530_NVIDIA_ANYCPU_20111029.zip did not inject anything but i have modified one just for testing DSDT Graphics Injection and Dual Monitor Support.

 

@luggi - Just to clarify: Are you saying that you are not using the DSDT_M1530_NVIDIA_ANYCPU_20111029.zip posted by immo, but using a modified version to support the external monitor? Then Chimera is not to blame?

[luggi]

No, Sorry for confusing you. I use Chameleon to inject the Graphics Card. I just tested Multi-Monitor support with the DSDT Injection just of Curiosity.

[LatinMcG]

No, Sorry for confusing you. I use Chameleon to inject the Graphics Card. I just tested Multi-Monitor support with the DSDT Injection just of Curiosity.

talkin about injecting.. dont forget to inject this to the PXS# where your wireless AR5B91 is in ioregistry (search for 168c)

under the line _ADR

				   Method (_DSM, 4, NotSerialized)
				{
					Store (Package ()
						{
							"AAPL,slot-name", Buffer () {"AirPort"},
							"device_type", Buffer () {"AirPort"},
							"subsystem-id", Buffer () {0x8F, 0x00, 0x00, 0x00},
							"subsystem-vendor-id", Buffer () {0x6B, 0x10, 0x00, 0x00}
						}, Local0)
					DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
					Return (Local0)
				}

[immo]

I mean the latter. The laptop screen + an external monitor connected via an HDMI-to-DVI cable. Very similar to luggi's screenshot in Post 964.

 

I haven't actually tested this with my M1330. Unfortunately my external monitor is HDMI so it won't work at all unless I use the VGA port. It crashes OS-X if I plug it into HDMI. You can try the VGA port just for kicks. It would at least prove it works as well as my M1330.

[mageus]

Been away for a while. Wife kept noting the constant request to update to 10.6.8, so I decided to make the DSDT fixes and update to 10.6.8.

 

M1530, 8600M GT, on 10.6.7, Chimera 1.4

 

Immo,

I applied all your fixes. I diff'd my DSDT against your latest DSDT. The only differences were Nvidia injection (which I had left in), and some extra USB code, likely because you have a 1330.

 

It booted, but was very laggy with no wifi. I used Talisman's recommendation to revert to the old Airport Monitor Bundle, and things seem to work perfectly. I'll update if I find any new problems. Immo, you are the bestest!

 

Now the only thing missing is HDMI-HDMI output.

[DarwinX]

talkin about injecting.. dont forget to inject this to the PXS# where your wireless AR5B91 is in ioregistry (search for 168c)

under the line _ADR

				   Method (_DSM, 4, NotSerialized)
				 {
					 Store (Package ()
						 {
							 "AAPL,slot-name", Buffer () {"AirPort"},
							 "device_type", Buffer () {"AirPort"},
							 "subsystem-id", Buffer () {0x8F, 0x00, 0x00, 0x00},
							 "subsystem-vendor-id", Buffer () {0x6B, 0x10, 0x00, 0x00}
						 }, Local0)
					 DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
					 Return (Local0)
				 }

 

Is there a way to inject the Dell TrueMobile 1505 into the XPS M1530 DSDT.aml?

[LatinMcG]

Is there a way to inject the Dell TrueMobile 1505 into the XPS M1530 DSDT.aml?

Brett tried with this method and found he had to do it in registers (like the ioata panic fix).

slice has an example on other forum.

 

whats the devce id 4328? u need to change subsystem id

VEN_14E4&DEV_4328&SUBSYS_000A1028 1028 is dell

 

to 106b apple subsystem 0087

or 0x8D

my hp bcm94321mcg p3 works fine with 8D

 

http://www.insanelymac.com/forum/index.php...t&p=1394464

 

change 0x8F to 0x87 or 0x8D then find the PXS# where your 4328 is. add under _ADR

 

if nothing works. rebrand.. or get different card. i read somewhere depending on revision of card it had some issues.

[DarwinX]

Brett tried with this method and found he had to do it in registers (like the ioata panic fix).

slice has an example on other forum.

 

whats the devce id 4328? u need to change subsystem id

VEN_14E4&DEV_4328&SUBSYS_000A1028 1028 is dell

 

to 106b apple subsystem 0087

or 0x8D

my hp bcm94321mcg p3 works fine with 8D

 

http://www.insanelymac.com/forum/index.php...t&p=1394464

 

change 0x8F to 0x87 or 0x8D then find the PXS# where your 4328 is. add under _ADR

 

if nothing works. rebrand.. or get different card. i read somewhere depending on revision of card it had some issues.

 

I patched my DSDT.aml according to you advice and it compiled without errors, yet it also yielded no results likely due to my own ineptitude in this matter.

Would you mind properly editing my current DSDT.aml to include the appropriate injection? Thank You.

 

iDarwin.aml.zip

iDarwin.ioreg.zip

[LatinMcG]

I patched my DSDT.aml according to you advice and it compiled without errors, yet it also yielded no results likely due to my own ineptitude in this matter.

Would you mind properly editing my current DSDT.aml to include the appropriate injection? Thank You.

 

iDarwin.aml.zip

iDarwin.ioreg.zip

 

i had same issue bfore. i have a device id 4315 and tried patch a while bak and didnt work either.. think it has to be done in registers to change subsystem.. but it should still work without subsystem..so something else might be wrong.

 

 

by the way youre using macbookPRO not MacBook5,1 or Pro6,1

 

but driver is loading fine.

 

PXS3 is fine. shows 4228 sub 8F (atheros oops u need 8D not F) is for the atheros 2a apple 8F .. bcm94321mc needs 8D (i rebranded mine as u saw in link for rebrand: think dsdt fake didnt work properly with bcm.. does with atheros)

Device (PXS3)
			{
				Name (_ADR, Zero)
			  Method (_DSM, 4, NotSerialized)
				 {
					 Store (Package ()
						 {
							 "AAPL,slot-name", Buffer () {"AirPort"},
							 "device_type", Buffer () {"AirPort"},
							 "subsystem-id", Buffer () {0x8D, 0x00, 0x00, 0x00},
							 "subsystem-vendor-id", Buffer () {0x6B, 0x10, 0x00, 0x00}
						 }, Local0)
					 DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
					 Return (Local0)
				 }
			}

 

any feedback ?

 

i see this in ioregistry

vendorid: 0x14e4
deviceid: 0x4328
radiorev: 0x42055000
chipnum: 0x4321
chiprev: 0x3
corerev: 0xc
boardid: 0xa
boardvendor: 0x1028
boardrev: 0x20
driverrev: 0x50a8324
ucoderev: 0x1fc00a2
bus: 0x1

boardvendor 1028 = dell.. not sure if that would matter much

give me 20 minutes and i will try my hp bcm94321mcp1 P3 ver 6.7r HP SPS: 436253-001 and compare. ioregistry hardwareversion

iDarwin4228.aml.zip

[DarwinX]

i had same issue bfore. i have a device id 4315 and tried patch a while bak and didnt work either.. think it has to be done in registers to change subsystem.. but it should still work without subsystem..so something else might be wrong.

 

 

by the way youre using macbookPRO not MacBook5,1 or Pro6,1

 

but driver is loading fine.

 

PXS3 is fine. shows 4228 sub 8F (atheros oops u need 8D not F) is for the atheros 2a apple 8F .. bcm94321mc needs 8D (i rebranded mine as u saw in link for rebrand: think dsdt fake didnt work properly with bcm.. does with atheros)

Device (PXS3)
			 {
				 Name (_ADR, Zero)
			   Method (_DSM, 4, NotSerialized)
				  {
					  Store (Package ()
						  {
							  "AAPL,slot-name", Buffer () {"AirPort"},
							  "device_type", Buffer () {"AirPort"},
							  "subsystem-id", Buffer () {0x8D, 0x00, 0x00, 0x00},
							  "subsystem-vendor-id", Buffer () {0x6B, 0x10, 0x00, 0x00}
						  }, Local0)
					  DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
					  Return (Local0)
				  }
			 }

 

any feedback ?

 

i see this in ioregistry

vendorid: 0x14e4
deviceid: 0x4328
radiorev: 0x42055000
chipnum: 0x4321
chiprev: 0x3
corerev: 0xc
boardid: 0xa
boardvendor: 0x1028
boardrev: 0x20
driverrev: 0x50a8324
ucoderev: 0x1fc00a2
bus: 0x1

boardvendor 1028 = dell.. not sure if that would matter much

give me 20 minutes and i will try my hp bcm94321mcp1 P3 ver 6.7r HP SPS: 436253-001 and compare. ioregistry hardwareversion

 

Much like my very own attempt to inject this Broadcom card into my DSDT, yours still yields no results. As you mentioned, this has to be done in registers or otherwise. Thank You for giving this matter your time and attention.

 

On a different note, per your advice, I have previously experimented with the various model identifiers, including the ones that you suggested, and the MacBookPro5,1 has always remained the best choice for my particular M1530. Is it problematic and if so, then why is it so? Thanks again!

[LatinMcG]

interesting.. i found minor issues with pro 5,1 temperature was not normal due to chipset being nforce on mobos of pro5,1 so i believe intelcpupm,agpm were slightly affected.

 

i have wifi switch set to off in bios. (os controls it fine)

 

rebrand the bcm94321 u have in linux live.

--subv 0x106b --subp 0x008D

(seems broadcom in mac works directly to rom thats why register dsdt works not _DSM injection )

 

does your wireless work? shows as 3rd party ?

[DarwinX]

interesting.. i found minor issues with pro 5,1 temperature was not normal due to chipset being nforce on mobos of pro5,1 so i believe intelcpupm,agpm were slightly affected.

 

i have wifi switch set to off in bios. (os controls it fine)

 

rebrand the bcm94321 u have in linux live.

--subv 0x106b --subp 0x008D

(seems broadcom in mac works directly to rom thats why register dsdt works not _DSM injection )

 

does your wireless work? shows as 3rd party ?

 

I believe that other model identifiers I experimented with have failed to successfully manage the GPU power profiles resulting in a significantly reduced graphics performance.

I also disabled the wireless switch, per your advice. My Broadcom BCM4328 wireless card does operate properly and is seen as a "Third-Party Wireless Card" by the OS. Unfortunately, I do not have the Linux Live installed.

[LatinMcG]

linuxmint.com.. burn to cd.. or use yumi usb or unetbootin to make usb live.

 

i had a feeling m1530 was slightly different in the way it handles cpu/gpu profiles.

 

other way would be to find the registers for the subvendor subsystem and dsdt inject them

[DarwinX]

i had a feeling m1530 was slightly different in the way it handles cpu/gpu profiles.

 

other way would be to find the registers for the subvendor subsystem and dsdt inject them

 

Unless someone else is willing to continue the research, I will seize any further investigation of this matter as it appears to require a significant amount of effort and time for simply cosmetic reasons with no other gain which is simply unreasonable.

[DougJoseph]

In the first post, in the tutorial, under the topic "Optional: Clamshell Fix (added 2010/04/12)"

 

It says to add under "_SB" ...

 

However, I see two different "_SB." One is "Scope (_SB)" at about line 64, and it is a shorter section, and the other is "Scope (_SB)" at about line 1203, and it has a lot under it, including Device (PCI0), Device (ISAB), and the newly added Device (SBUS) that we were told to add above Device (ISAB).

 

Am I supposed to have two of "Scope (_SB)"? And if so, does it matter which one I add the "Device (PNLF)" section to?

 

Sorry to be such a newb.

 

The content to be added is:

Device (PNLF)
{
Name (_HID, EisaId ("APP0002"))
Name (_CID, "backlight")
Name (_UID, 0x0A)
Name (_STA, 0x0B)
}

[DougJoseph]

A section in the tutorial gives me trouble with my M1530 DSDT editing. Whenever I try to follow the step below, I instantly go from a file that compiles with 0 errors to one that compiles with over 200 errors. The errors all start with "Object does not exist...."

 

The section in focus says:

Optional: Adding SBUS Device (Added 2009/10/22) (Edit 2010/04/12 -> SMBI Address was Wrong)

jkbuha came up with this fix using a guide made by Master Chief (thank you). This allows the detection of the SBUS device, and it can be confirmed as detected using IORegistry Explorer by searching for SBUS. This device is required to get C-states working, however we still do not have working C-states even with this patch.

 

Add this under the first curly brace under Scope (\_SB):

		Method (DBEN, 0, NotSerialized)
		   {
		   }

[LatinMcG]

A section in the tutorial gives me trouble with my M1530 DSDT editing. Whenever I try to follow the step below, I instantly go from a file that compiles with 0 errors to one that compiles with over 200 errors. The errors all start with "Object does not exist...."

 

The section in focus says:

 

Add this under the first curly brace under

 

either _SB is same ..just separated

[DougJoseph]

Add this under the first curly brace under

 

either _SB is same ..just separated

 

Thanks! Very helpful.

 

On my second post above, note that the intended location is not "Scope (_SB)" but rather "Scope (\_SB)"

 

There is a slash. When I place the suggested tweak there, it gives about 200 errors. Any idea what I might be doing wrong?

[LatinMcG]

Thanks! Very helpful.

 

On my second post above, note that the intended location is not "Scope (_SB)" but rather "Scope (\_SB)"

 

There is a slash. When I place the suggested tweak there, it gives about 200 errors. Any idea what I might be doing wrong?

 

probably missing a }

[raimdal]

sorry, can you tell me why are you using macbookpro 5,1 in smbios plist? is better than 5,3? I'm using 5,3 , the scaling of cpu is working (0x0 in ioregistry) but I read that many of you are using 5,1...why? It seems to me that every things works fine but I'm not an expert...there is a big difference ? my cpu is t7500..

 

however last week I replaced my intel4965 wireless with bcm94311 and it works great ootb... well this is not really important but thanks this now I'm really happy )))) and if you are looking for to kill intel4965 this is a perfect solution...I've paid it 10euro. thanks for the answer at the question on the smbios.plist.

[LatinMcG]

sorry, can you tell me why are you using macbookpro 5,1 in smbios plist? is better than 5,3? I'm using 5,3 , the scaling of cpu is working (0x0 in ioregistry) but I read that many of you are using 5,1...why? It seems to me that every things works fine but I'm not an expert...there is a big difference ? my cpu is t7500..

 

however last week I replaced my intel4965 wireless with bcm94311 and it works great ootb... well this is not really important but thanks this now I'm really happy )))) and if you are looking for to kill intel4965 this is a perfect solution...I've paid it 10euro. thanks for the answer at the question on the smbios.plist.

 

i dont use pro 5,1 i use pro6,1 .. but plain not pro 5,1 works fine too.. not sure why pro5,1 was picky for me.

 

94311 is G .. N is bcm94321

[DougJoseph]

i dont use pro 5,1 i use pro6,1 .. but plain not pro 5,1 works fine too.. not sure why pro5,1 was picky for me.

 

94311 is G .. N is bcm94321

Pro5,1 would most closely match my hardware. But when I used Lizard to create an SMbios.plist that was accurate for the right model, I got full halt on boot with some message about memory tables not being allocated right or some such. Probably something in my DSDT.aml was not right for that selection. I went back to using plain 5,1 and it boots again. I have everything working on the Dell M1530 (8300 processor) except for sleep. Well, of course AirDrop is not available due to the selection plain 5,1.

 

I am not sure that the voltage states on the power management are ideal. They seem to work, as I am watching the temp, and it is behaving like voltage is being reduced when possible. But the temp is running a little high always, and the fans run almost constantly. That was why I tried to switch to a MacBook Pro model that used the same processor as I have.

 

probably missing a }

 

I know for a fact it was not a missing brace, at least not there. I tried many times. I think there must be some error somewhere else. That part below is the only part I have not been able to add. Not sure what that might be costing me in problems or lost functionality.

 

		Method (DBEN, 0, NotSerialized)
			{
			}

[DarwinX]

Zprood has recently developed the most complete and accurate battery implementation for our rigs to date.

 

AppleACPIBatteryManager/AppleSmartBatteryManager For Lion, New battery driver for lion.

 

 

In your DSDT.aml, under the Scope _SB, create the "Method (B1B2, 2, NotSerialized)" as global:

 Method (B1B2, 2, NotSerialized)
	{
		 Or (ShiftLeft(Arg1,0x08), Arg0, Local0)
		 Return (Local0)
	 } //for 16bit

 

DSDT.aml.zip

AppleSmartBatteryManager.kext.zip

Do not forget to completely remove any other battery related kexts (AppleACPIBatteryManager, VoodooBattery, etc.) from the System/Library/Extensions and/or the /Extra/Extensions directories.

 

[DougJoseph]

Zprood has recently developed the most complete and accurate battery implementation for our rigs to date.

 

DarwinX

 

Thanks! My M1530's processor is T8300 (2.4 gHz) -- slightly slower than yours. Do you think I might be able to use your DSDT file and SMbios.plist, etc, even with that difference? My DSDT file is functional, yet somewhat flawed. I am seeing some minor issues with it. If yours would work, would you be willing to post your current SMbios.plist file?

 

EDITED:

 

Well, I just noticed there is also a slight difference in our video cards. There may be other differences, such as wifi card. Not sure it would work. Might be worth a try.

[DarwinX]

DarwinX

 

Thanks! My M1530's processor is T8300 (2.4 gHz) -- slightly slower than yours. Do you think I might be able to use your DSDT file and SMbios.plist, etc, even with that difference? My DSDT file is functional, yet somewhat flawed. I am seeing some minor issues with it. If yours would work, would you be willing to post your current SMbios.plist file?

 

EDITED:

 

Well, I just noticed there is also a slight difference in our video cards. There may be other differences, such as wifi card. Not sure it would work. Might be worth a try.

 

My current DSDT.aml is not generic to my bios but is based on the front page non specific CPU DSDT.aml. So, you should be able to use it without any problems. As far as the video, since we are both using Chameleon's native graphics injection, it isn't an issue either. Here are my smbios and boot plists:

 

smbios.plist.ziporg.chameleon.Boot.plist.zip