1350 replies to this topic

[jeepguy_1980]

I am unable to create the .aml tables that contain any useful information, through Ubuntu.

The following is a copy of my terminal session. Do you see anything that I may have done wrong? I just get the following error:

acpi_os_map_memory: cannot open /dev/mem

To run a command as administrator (user "root"), use "sudo ".
See "man sudo_root" for details.

ubuntu@ubuntu:~$ sudo nano /etc/apt/sources.list
ubuntu@ubuntu:~$ sudo apt-get update
Ign cdrom://Ubuntu 9.04 _Jaunty Jackalope_ - Release amd64 (20090420.1) jaunty/main Translation-en_US
Ign cdrom://Ubuntu 9.04 _Jaunty Jackalope_ - Release amd64 (20090420.1) jaunty/restricted Translation-en_US
Get:1 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security Release.gpg [189B]						   
Ign [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/main Translation-en_US	 
Get:2 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty Release.gpg [189B]
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/main Translation-en_US
Ign [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/restricted Translation-en_US
Ign [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/universe Translation-en_US
Get:3 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security Release [57.9kB]
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/restricted Translation-en_US
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/universe Translation-en_US
Get:4 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates Release.gpg [189B]
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/main Translation-en_US
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/restricted Translation-en_US
Ign [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/universe Translation-en_US
Get:5 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty Release [74.6kB]
Get:6 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates Release [57.9kB]
Get:7 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/main Packages [116kB]	 
Get:8 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/main Packages [1251kB]				
Get:9 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/restricted Packages [2594B]
Get:10 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/main Sources [30.3kB]	
Get:11 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/restricted Sources [623B]	  
Get:12 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/universe Packages [53.6kB]
Get:13 [url="http://security.ubuntu.com"]http://security.ubuntu.com[/url] jaunty-security/universe Sources [12.6kB]		  
Get:14 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/restricted Packages [8858B]																						 
Get:15 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/main Sources [555kB]																								
Get:16 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/restricted Sources [3156B]																						  
Get:17 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/universe Packages [4732kB]																						  
Get:18 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/universe Sources [2375kB]																						   
Get:19 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/main Packages [193kB]																					   
Get:20 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/restricted Packages [2594B]																				 
Get:21 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/main Sources [54.8kB]																					   
Get:22 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/restricted Sources [623B]																				   
Get:23 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/universe Packages [74.4kB]																				  
Get:24 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty-updates/universe Sources [21.2kB]																				   
Fetched 9678kB in 2min 7s (76.0kB/s)																														
Reading package lists... Done
ubuntu@ubuntu:~$ sudo apt-get install acpidump
Reading package lists... Done
Building dependency tree	   
Reading state information... Done
The following NEW packages will be installed:
  acpidump
0 upgraded, 1 newly installed, 0 to remove and 253 not upgraded.
Need to get 16.5kB of archives.
After this operation, 94.2kB of additional disk space will be used.
Get:1 [url="http://archive.ubuntu.com"]http://archive.ubuntu.com[/url] jaunty/universe acpidump 20071116-1 [16.5kB]
Fetched 16.5kB in 0s (19.2kB/s)
Selecting previously deselected package acpidump.
(Reading database ... 103503 files and directories currently installed.)
Unpacking acpidump (from .../acpidump_20071116-1_amd64.deb) ...
Processing triggers for man-db ...
Setting up acpidump (20071116-1) ...
ubuntu@ubuntu:~$ mkdir ACPI && dmesg | perl -we '$n=0; while (<>) { if (($t,$a,$l,$o) = (/^[^a-zA-Z]*ACPI: ([-._A-Z0-9]{4,4}) +([0-9A-F]{8,8}), ([0-9A-F]{4,4})+(?:\s*\(([^)]+))?/)) { $o && $o=~s/[^-._a-zA-Z0-9]+/-/g; ($cmd="acpidump -a $a -l $l > \"ACPI/${t}".($o?"_$o":"").".aml\""); print "Running command: \"$cmd\"\n"; system($cmd); ++$n; } } die("No match") unless $n;' && zip -r ACPI-Tables.zip ACPI
Running command: "acpidump -a 000FBC00 -l 0024 > "ACPI/RSDP_r2-DELL-.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF66F200 -l 005C > "ACPI/XSDT_r1-DELL-M08-27D80B13-ASL-61.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF66F09C -l 00F4 > "ACPI/FACP_r4-DELL-M08-27D80B13-ASL-61.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF66F800 -l 5733 > "ACPI/DSDT_r2-INT430-SYSFexxx-1001-INTL-20050624.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF67E000 -l 0040 > "ACPI/FACS.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF66F300 -l 0038 > "ACPI/HPET_r1-DELL-M08-1-ASL-61.aml""
acpi_os_map_memory: cannot open /dev/mem
Running command: "acpidump -a DF66F400 -l 0068 > "ACPI/APIC_r1-DELL-M08-27D80B13-ASL-47.aml""
acpi_os_map_memory: cannot open /dev/mem

[cyberbuddhah]

Great job. I think we can, together, develop this thread into "SL for Dell XPS M1330 & M1530". There are many other threads on this, but some outdated and too long too convoluted to search for info. DSDT is the crucial part, then come video, sound, networking.

[Chrysaor]

jeepguy_1980, on Oct 14 2009, 09:52 AM, said:

I am unable to create the .aml tables that contain any useful information, through Ubuntu.

The following is a copy of my terminal session. Do you see anything that I may have done wrong? I just get the following error:

acpi_os_map_memory: cannot open /dev/mem

[jeepguy_1980]

Chrysaor, on Oct 14 2009, 10:37 AM, said:

I'm not sure if that was meant to insult me or not, but it was helpful. After I Googled that picture and learned its meaning, I realized what I was doing wrong. Thanks.

I'm new to the OS X project and have very limited experience with Linux, so I apologize for my stupid question.

[immo]

Chrysaor, on Oct 14 2009, 03:37 PM, said:

Sudo make me a sandwich

Well played Chrysaor!

[immo]

jkbuha, on Oct 14 2009, 11:28 AM, said:

Thanks to everyone for all the support and helpful hints/tips!

I've uploaded my latest version of the DSL file for the M1330 (T8300 undervolted but just add your own _PSS values for flavour/flavor). This version also includes SMBusPCI functionality derived from lspci.txt as per immo/FMulder's excellent tips - to verify if correctly loaded search in IORegistry for SMBus and a key with SBus@xx,x (mine is 1f,3) should appear.

The SBUS code was adapted from Master Chief's excellent P5K PRO DSDT guide here

Worth a look for any aspiring dsl gurus.
Let's try and encourage that mindset into evolving our Dell DSDT's similarly!

Cheers
jkbuha

Hi Jkbuha.  What is the effect of adding the SBUS code?

Thanks
Immo

[jkbuha]

immo, on Oct 15 2009, 05:20 PM, said:

Hi Jkbuha.  What is the effect of adding the SBUS code?

Thanks
Immo

In conjunction with LPC (Device (LPCB) in DSDT or modded AppleLPC kext to support your own LPC device from lspci.txt) this should provide proper C-state functionality. Once c-states are properly working, which is not my case unfortunately. I tend to be a stickler for proper device declaration

On that note, any hints and tips about how to get C-states working properly for my T8300, with MacBookPro5,1 (or should I change this??)

Cheers
jkbuha

[immo]

jkbuha, on Oct 15 2009, 06:14 PM, said:

In conjunction with LPC (Device (LPCB) in DSDT or modded AppleLPC kext to support your own LPC device from lspci.txt) this should provide proper C-state functionality. Once c-states are properly working, which is not my case unfortunately. I tend to be a stickler for proper device declaration

On that note, any hints and tips about how to get C-states working properly for my T8300, with MacBookPro5,1 (or should I change this??)

Cheers
jkbuha

Sorry not working for me yet either.

P.S. I was thinking about the whole MacBookPro5,1 thing, and did some research, and it seems that our computers are closer to a MacBookPro3,1, as it has the same north and south bridge as the M1330 (PM965 and ICH8).  My guess that the reason we get better functionality with MacBookPro5,1 is because there may be special case handling in OS-X for the MacBookPro3,1 and MacBookPro4,1 that goofs us up.  OS-X is probably designed so that it will work with computers newer than it is familiar with using generic methods.  This is purely a guess.
Edit: I'm beginning to doubt this theory because I really can't get speed stepping to work with anything but MacBookPro5,1.  It looks more like the reverse is true and our computers need the MacBookPro5,1 special case handling.

[jkbuha]

immo, on Oct 15 2009, 09:47 PM, said:

Edit: I'm beginning to doubt this theory because I really can't get speed stepping to work with anything but MacBookPro5,1.  It looks more like the reverse is true and our computers need the MacBookPro5,1 special case handling.

From my own experimentation it appears that speedstep will only work natively with those cpus defined in PLimitDict=0 and StepDataDict=1 section in the ACPI_SMC_Plugin kext (MacPro4,1 is by default excluded, ergo no speedstep). More info here.

What I'd like to do (but cannot atm because I don't have a livecd) is to boot into Ubuntu live and see if the SSDT table locations change on runtime as detailed here. If this is true then we would need to do one or more of the following:

1. Change the final destination of the SSDT mapping for CPU0IST, CPU1IST etc
2. Change the mac designation from MacBookPro5,1 to MacPro4,1 in smbios.plist and insert the new designation under PLimitDict and StepDataDict. This worked partially for me (ie: no more cst errors) but the t8300 became very unresponsive once it loaded. At least some progress...
3. Add a kext in E/E to enable vanilla stepping if 2. works but speedstepping doesn't.

However it seems that at least one T8300 user out there has managed to implement this in some form: The King. I haven't tried to ping him/her yet but maybe we could get some advice?

Well that's my 0.02...!

Cheers
jkbuha

[immo]

jkbuha, on Oct 16 2009, 12:04 AM, said:

From my own experimentation it appears that speedstep will only work natively with those cpus defined in PLimitDict=0 and StepDataDict=1 section in the ACPI_SMC_Plugin kext (MacPro4,1 is by default excluded, ergo no speedstep). More info here.

What I'd like to do (but cannot atm because I don't have a livecd) is to boot into Ubuntu live and see if the SSDT table locations change on runtime as detailed here. If this is true then we would need to do one or more of the following:

1. Change the final destination of the SSDT mapping for CPU0IST, CPU1IST etc
2. Change the mac designation from MacBookPro5,1 to MacPro4,1 in smbios.plist and insert the new designation under PLimitDict and StepDataDict. This worked partially for me (ie: no more cst errors) but the t8300 became very unresponsive once it loaded. At least some progress...
3. Add a kext in E/E to enable vanilla stepping if 2. works but speedstepping doesn't.

However it seems that at least one T8300 user out there has managed to implement this in some form: The King. I haven't tried to ping him/her yet but maybe we could get some advice?

Well that's my 0.02...!

Cheers
jkbuha

Would moving SSDT addresses even matter to us if we put our SSDT data in the DSDT.aml file and use the "DropSSDT=y" flag?  From what I understand, you'd only see those addresses if you did your dump from OS-X, and from the linux dump you actually get the real content that is in the memory rather than the addresses.  So by putting the real data dumped from linux into the DSDT.aml and ignoring the SSDT data in memory you work around the problem.  At least that's the way I understand it.

Immo

[jkbuha]

immo, on Oct 16 2009, 02:14 AM, said:

Would moving SSDT addresses even matter to us if we put our SSDT data in the DSDT.aml file and use the "DropSSDT=y" flag?

Ah now the DropSSDT flag explains itself That said, I don't think the SSDT tables are the only factor - there's a lengthy _CST method after that which may or may not need to be changed as well depending on whether we're selecting MBP5,1 MP4,1 MP3,1 or otherwise. From the look of things the ACPI_SMC_Plugin may also need to be patched depending on which is chosen. At least that's my reading of the various posts. What do you think?

Cheers
jkbuha

[immo]

jkbuha, on Oct 16 2009, 09:57 AM, said:

Ah now the DropSSDT flag explains itself That said, I don't think the SSDT tables are the only factor - there's a lengthy _CST method after that which may or may not need to be changed as well depending on whether we're selecting MBP5,1 MP4,1 MP3,1 or otherwise. From the look of things the ACPI_SMC_Plugin may also need to be patched depending on which is chosen. At least that's my reading of the various posts. What do you think?

Cheers
jkbuha

Could very well be.  BTW something else I read somewhere is IRQ conflicts can cause issues, and if I recall, our HPET and something else both are set to IRQ 2.  I've read of someone fixing their issues by freeing up IRQ's 2 and 8 for the HPET.  I'll look into it when I have more time.

Immo

[smartmobili]

immo, on Oct 16 2009, 02:25 PM, said:

Could very well be.  BTW something else I read somewhere is IRQ conflicts can cause issues, and if I recall, our HPET and something else both are set to IRQ 2.  I've read of someone fixing their issues by freeing up IRQ's 2 and 8 for the HPET.  I'll look into it when I have more time.

Immo

Hi,

I have dumped my acpi tables from a Dell XPSM1530 with the following characteristics :

Core2Duo T8300 2.4 GHz
3 GB DDR2
GForce 8600MGT (1680x1050)
120 GB SSD OCZ Vertex
Marvell Yukon 88E8040 + Wifi Intel 4965AGN

but I didn't manage to create a dsdt (actually I think I am not very good in copy/paste). So if someone would be interested to make a dsdt for people using my configuration it would help the community.
You can find my ACPI-tables.zip here

[immo]

smartmobili, on Oct 17 2009, 01:34 PM, said:

Hi,

I have dumped my acpi tables from a Dell XPSM1530 with the following characteristics :

Core2Duo T8300 2.4 GHz
3 GB DDR2
GForce 8600MGT (1680x1050)
120 GB SSD OCZ Vertex
Marvell Yukon 88E8040 + Wifi Intel 4965AGN

but I didn't manage to create a dsdt (actually I think I am not very good in copy/paste). So if someone would be interested to make a dsdt for people using my configuration it would help the community.
You can find my ACPI-tables.zip here

Try this one by Chrysaor:
http://www.insanelym...p...t&p=1242766

[smartmobili]

immo, on Oct 17 2009, 10:09 PM, said:

Try this one by Chrysaor:
http://www.insanelym...p...t&p=1242766

So I have modified superhai preboot cd (v0.4) by using Chrysaor DSDT and I managed to install SL but my resolution is in 1280x800 while it should be 1680x1050 ... However I changed resolution in plist file.
And it seems there is no 2D/3D acceleration because welcome animation at first boot is very slow.
How can I fix that ?
I have also noticed my cpu fan is getting crazy sometimes but maybe this is normal.

Does a DSDT specify display resolution ?
I suppose I need to create custom dsdt for my machine ...

UPDATE : I have started to merge dsdt and ssdt as described in tutorial. Result is here. (see myDSDT.dsl)
Now I am trying to make a diff between your versions and mine and at first glance it seems we don't use the same acpi decompiler because for instance instead of putting 0x00 or 0x01, version on this thread use String Zero and One. Of course it's not important but it make diff more complicated. So I have replaced string by their value (Zero->0x00 and One->0x01) and now I have some slight differences but I don't know what it means.

[immo]

Hi smartmobili.

I do not believe that the DSDT specifies the resolution at all.  My guess is that you have not installed your DSDT.aml correctly.  Double-check the last section of installing the DSDT, make sure that you've editted and installed the smbios.plist in the /Extra directory, and that the DSDT is called /Extra/DSDT.aml, and that you've set permissions of your /Extra folder.  Also make sure that you've set up the boot flag "DropSSDT=y".

About the decompiler, it makes no difference.  When you compile and decompile with iasl you get those Zeros and Ones.
Immo

smartmobili, on Oct 18 2009, 03:29 PM, said:

So I have modified superhai preboot cd (v0.4) by using Chrysaor DSDT and I managed to install SL but my resolution is in 1280x800 while it should be 1680x1050 ... However I changed resolution in plist file.
And it seems there is no 2D/3D acceleration because welcome animation at first boot is very slow.
How can I fix that ?
I have also noticed my cpu fan is getting crazy sometimes but maybe this is normal.

Does a DSDT specify display resolution ?
I suppose I need to create custom dsdt for my machine ...

UPDATE : I have started to merge dsdt and ssdt as described in tutorial. Result is here. (see myDSDT.dsl)
Now I am trying to make a diff between your versions and mine and at first glance it seems we don't use the same acpi decompiler because for instance instead of putting 0x00 or 0x01, version on this thread use String Zero and One. Of course it's not important but it make diff more complicated.

[smartmobili]

immo, on Oct 18 2009, 08:24 PM, said:

Hi smartmobili.

I do not believe that the DSDT specifies the resolution at all.  My guess is that you have not installed your DSDT.aml correctly.  Double-check the last section of installing the DSDT, make sure that you've editted and installed the smbios.plist in the /Extra directory, and that the DSDT is called /Extra/DSDT.aml, and that you've set permissions of your /Extra folder.  Also make sure that you've set up the boot flag "DropSSDT=y".

About the decompiler, it makes no difference.  When you compile and decompile with iasl you get those Zeros and Ones.
Immo

yes I forgot to add DropSSDT to kernel flags, that might be the reason.
By the way the tutorial is great but why don't you use DSDT Patcher utility from Koala, it seems to do everything.

UPDATE : So I did what you said and now I have good resolution but  USB works randomly and sleep/hibernate is broken.

[immo]

smartmobili, on Oct 19 2009, 03:06 PM, said:

yes I forgot to add DropSSDT to kernel flags, that might be the reason.
By the way the tutorial is great but why don't you use DSDT Patcher utility from Koala, it seems to do everything.

UPDATE : So I did what you said and now I have good resolution but I don't have USB nor sleep/hibernate ...

Read the guide through.  Your solution is there.

If you or anyone has tried the Koala patcher and it works as well/better please post your results.

[smartmobili]

immo, on Oct 19 2009, 08:52 PM, said:

Read the guide through.  Your solution is there.

If you or anyone has tried the Koala patcher and it works as well/better please post your results.

Arghh I thought Chrysaor's version was already including this fix but it wasn't the case.
However other fixes are available and from what I understand it should work with nothing more.
I will check again very carefully if there is not mistake in the DSDT.dsl/

UPDATE : I have checked and here is the results

DTGP Method is not under but above _WAK method but I imagine this is not important
EHC2 and EHCI are available
NVidia Injection available
HD Audio Patching available

Last thing to check is what kext is included and I will do it when back home.

UPDATE2: Here is my kext
drwxr-xr-x  3 vincent  staff  102  4 sep 16:04 AppleACPIPS2Nub.kext
drwxr-xr-x  3 vincent  staff  102  4 sep 22:30 ApplePS2Controller.kext
drwxr-xr-x@ 3 vincent  staff  102 17 sep 17:55 IOATAFamily.kext
drwxr-xr-x  3 vincent  staff  102  4 sep 22:49 NullCPUPowerManagement.kext
drwxr-xr-x  3 vincent  staff  102  5 sep 00:31 VoodooBattery.kext
drwxr-xr-x  3 vincent  staff  102  4 sep 22:34 VoodooHDA.kext
drwxr-xr-x  3 vincent  staff  102 21 sep 23:43 VoodooTSCSync.kext
drwxr-xr-x  3 vincent  staff  102  4 sep 16:04 fakesmc.kext

If you have any idea why shutdown/sleep/hibernate doesn't work please let me know

[immo]

smartmobili, on Oct 20 2009, 08:24 AM, said:

UPDATE2: Here is my kext
drwxr-xr-x 3 vincent staff 102 4 sep 16:04 AppleACPIPS2Nub.kext
drwxr-xr-x 3 vincent staff 102 4 sep 22:30 ApplePS2Controller.kext
drwxr-xr-x@ 3 vincent staff 102 17 sep 17:55 IOATAFamily.kext
drwxr-xr-x 3 vincent staff 102 4 sep 22:49 NullCPUPowerManagement.kext
drwxr-xr-x 3 vincent staff 102 5 sep 00:31 VoodooBattery.kext
drwxr-xr-x 3 vincent staff 102 4 sep 22:34 VoodooHDA.kext
drwxr-xr-x 3 vincent staff 102 21 sep 23:43 VoodooTSCSync.kext
drwxr-xr-x 3 vincent staff 102 4 sep 16:04 fakesmc.kext

If you have any idea why shutdown/sleep/hibernate doesn't work please let me know

Remove NullCPUPowerManagement.kext.  This kext blocks the vanilla speed stepping and power management functions.  I'll update the first post to include this (the first post mentions another kext that does the same thing).  I think you'll find that speed stepping will start to work once you remove this kext.

Shutdown has not worked for anyone on these machines to the best of my knowledge.  It'll be big news once someone gets that to work.