Jump to content
371 posts in this topic

Recommended Posts

"I can has model" (in your System Profiler screenshot) is from Diabolik's old injector NVKush.

This is the original site: http://diabolik1605.com/DHF/

NVKush is from August 2008 and the target nvidia cards were 7000 and 8000 series. It's obsolete.

 

To fix your issues, upgrade to Chameleon 2.0 RC5, set GraphicsEnabler=y in /Extra/com.apple.Boot.plist and delete NVKush.kext.

 

This should take care of the CPU as well. Your CPU will be faked/identified as a Core 2 Duo or Xeon Quad Core depending on how many cores it has. This is done on purpose for compatibility reasons, Apple don't use AMD CPUs.

 

There's a fairly recent SVN build of Chameleon 2.0 RC5 attached to this thread a couple of pages back:

http://www.insanelymac.com/forum/index.php...t&p=1703409

 

thanks for the tutorial mate..this time the screen goes into blue color once boot at -v ,-f ,arch=i386 .

but goes through once boot at -x any suggestion about this error?

 

thanks mate

npci=0x2000 works perfectly here. GTX 460, June 24 Quadro drivers, Intel P45/ICH10R chipset.

the screen goes into blue color once boot at -v ,-f ,arch=i386 but goes through once boot at -x any suggestion about this error?

The standard fix for that is using the other display output on the video card. Use a DVI-VGA adapter if you have to. The screen is blue because OS X thinks its the secondary display.

 

If you absolutely need to use the output that is currently showing a blue screen you need to look for a custom NVCAP.

NVCAP determines the order of your display outputs, if you're using Chameleon's GraphicsEnabler you can set a custom NVCAP via com.apple.Boot.plist.

For more information on this, go look for Krazubu's nvidia/nvenabler threads over at the Project OS X forums.

I think I have a partial clue (maybe I'm wrong) to the problem of black screen after sleep on my eeepc:

I compiled to 10.6 compatible SMBiosEnabler and AppleACPIPS2nub from meklort source (thanks to him) and I noticed after sleeping chameleon is complaining of having secure virtual memory while in fact I'm not using any secure memory!!!!

 

I think I have a partial clue (maybe I'm wrong) to the problem of black screen after sleep on my eeepc:

I compiled to 10.6 compatible SMBiosEnabler and AppleACPIPS2nub from meklort source (thanks to him) and I noticed after sleeping chameleon is complaining of having secure virtual memory while in fact I'm not using any secure memory!!!!

Sorry ,wrong clue ,solving the problem didn't change anything!

Dear JBraddock, do you have any idea what this npci value does to the kernel? You also seem to have an NVidia basic graphics adapter... Any idea where to start looking for this npci thing? Could it be related to a memory address of our graphics card?

 

It's 2 bytes long... Perhaps boot without it first, and then open up IORegistryExplorer and in our GFX0 section (which I inject via DSDT method) find and our own value for npci?

 

Really, anyone?

Sorry, I really have no idea what it does.

I'll have a look at what you'd asked for and let you know. It renders OS X unusable so I want to do it when I have a proper backup.

Also, I replied to bcc9.

But what does npci do? Could not find a reference in Help (GUI) at all... Thank you all !

0x2000 is the kIOPCIConfiguratorPFM64 flag, as seen in the IOPCIFamily source code. (PFM64 probably stands for prefetch memory 64 bit).

 

npci=0x2000 turns off this flag, as npci negates whatever flag bits you specify.

 

With this flag turned off, the logic that sets fConsoleRange is disabled, and so the new graphics console relocation code is not triggered.

 

However, the rest of the other new relocation code is still executed, and therein probably lies a problem for some, depending upon your PCI memory config.

 

AFAIK, npci=0x2000 is working fine without side effects for most users.

Ok I updated to 10.6.8, but like most everybody else I had to revert back to my copy of AppleHDA.kext that I was using under 10.6.7. Also, SleepEnabler.kext doesn't work & causes a kernel panic. Setting pmVersion to 23 didn't help either. ALso, FYI I had to disable GlimmerBlocker for the update to work. I got a KP in the middle of the update, and luckly I was able to reboot to turn off GlimmerBlocker before trying again.

 

Is there a new SleepEnabler that will work under 10.6.8 ??

 

Edit: Thanks for the SleepEnabler links. I found this SleepEnabler before reading the postings below, and this one does work for me with pmVersion=23. The SleepEnabler link that I used is below:

 

http://blog.nawcom.com/?p=791

npci=0x2000 works perfectly here. GTX 460, June 24 Quadro drivers, Intel P45/ICH10R chipset.

 

The standard fix for that is using the other display output on the video card. Use a DVI-VGA adapter if you have to. The screen is blue because OS X thinks its the secondary display.

 

If you absolutely need to use the output that is currently showing a blue screen you need to look for a custom NVCAP.

NVCAP determines the order of your display outputs, if you're using Chameleon's GraphicsEnabler you can set a custom NVCAP via com.apple.Boot.plist.

For more information on this, go look for Krazubu's nvidia/nvenabler threads over at the Project OS X forums.

 

 

Thanks Mate for the info ..works well now.

Hi guys, I updated to 10.6.8 and had to delete the SleepEnabler.kext cause I was getting a KP. Now it boots fine but for some reason none of the main drivers load. No mouse or keyboard function, graphics are default, as if I'm in safe mode. This only happened after I deleted the SleepEnabler.kext file. Before which worked fine except I had to run pmVersion=0 to bypass KP. Any ideas?

 

Thanks

Man thanks!! That did it. Everything works perfect now.. I only have one small annoyance though, but it's done this for awhile now even before updating to 10.6.8

 

SL just takes forever on first boot. Mouse works and everything else works except all menus and desktop are frozen for like 10 mins. Then it catches up and everything runs fast and smooth. I don't think it's hardware related as that would cause it to be slow no matter what, I think it's a driver at startup causing it to not respond for a time then finally stopping. Any ideas or has anyone else had this problem?

 

Thanks

i get stuck in the message "dsmos was arrived"

i even deleted all my extra kexts , just leaved the "fakesmc.kext" and the problem remains...

Try deleting your fakesmc plugins and/or upgrade fakesmc. The latest version is attached in an earlier post in this thread.

"dsmos was arrived" is normally one of the last log entries shown before it goes into graphics mode. FakeSMC should be Ok, if it booted to this point, however upgrading it never hurts. FWIW, Any FakeSMC that works in 10.6.3 (used with a normal retail install), will work in all releases of Snow Leopard, and even works with Lion (to GM at least).

 

To me, it would appear to be a video issue. If you boot up with GraphicsEnabler=Yes (or No) in chameleon, does it work any better?

 

I have no experience with ATI cards in SL, so the help I can offer may be limited.

Confirming this works for me too (god knows what i did last night :) ). Would love to know what the deal is!

I tried Hoolock with the 10.6.8 ATI kext and it KP'd. didn't try any others.

 

I've just rolled back ATI Kexts and AppleHDA

I've not touched IOPCIFamily or AppleACPIPlatform ..

 

D

 

can you upload the old 10.6.7 ati drivers please ?

 

looks like it is the problem i have...

 

in verbose mode i cant see no errors but i get stuck in the last line... i listen the computer working (looks like everything is working fine) but i cant see the desktop i get stuck in the command line

 

i have one ATI 4850 1GB + ASUS PQ5

 

i removed all extra kexts and the problem remains, must be related to the graphic card for sure

 

the netkas ATI fix is in .PGK installer , i need the files in .kext ... (i use macdrive in windows to get access to OSX drives)

 

thanks in advance.

the ATI drivers of 10.6.7 dont work on 10.6.8 ; you need to use Netkas's qe_ci_exotic patch 10.6.8 for your HD 4850 !

 

for open a .pkg , you can to use "pacifist" and then extract ATI4800Controller.kext and ATIRadeonX2000.kext ! :)

 

use kext helper to install them ! :)

the ATI drivers of 10.6.7 dont work on 10.6.8 ; you need to use Netkas's qe_ci_exotic patch 10.6.8 for your HD 4850 !

 

for open a .pkg , you can to use "pacifist" and then extract ATI4800Controller.kext and ATIRadeonX2000.kext ! :blink:

 

use kext helper to install them ! ;)

 

 

"pacifist" for windows? , i found a solution 7-zip for windows can extract .pkg just fine (i say in the other post i use macdrive to manage osx partitions under windows, the only osx i have is this one)

 

anyway now its too late... i deleted all ati*.* files from extensions folder...

 

i can boot just fine now but i lost qe/ci

 

where can i found all 10.6.8 ati kexts now? crazy situation...

pff , i'm sorry , i dont see you talk about Windows ! :blink:

 

yes , only need ATI4800Controller.kext and ATIRadeonX200.kext patched by Netkas for your HD 4850 ! ;)

 

this is all my ATI drivers from 10.6.8 , ATI4800Controller.kext and ATIRadeonX200.kext patched by Netkas are there too ! :)

 

with this , your HD 4850 will work , i hope ! :D

 

http://www.mediafire.com/?c7goh2514v5a51a

finally its solved !

 

once again i was late to see your post fantomas1 , but your file will help other ATI users for sure!

 

i found the original ati 10.6.8 kexts in another post (16mb)

 

then i apply the netkas 10.6.8 fix

 

 

thanks for your posts fantomas1 ;)

0x2000 is the kIOPCIConfiguratorPFM64 flag, as seen in the IOPCIFamily source code. (PFM64 probably stands for prefetch memory 64 bit).

 

npci=0x2000 turns off this flag, as npci negates whatever flag bits you specify.

 

With this flag turned off, the logic that sets fConsoleRange is disabled, and so the new graphics console relocation code is not triggered.

 

However, the rest of the other new relocation code is still executed, and therein probably lies a problem for some, depending upon your PCI memory config.

 

AFAIK, npci=0x2000 is working fine without side effects for most users.

Finally! Someone to technically explain this flag! Thanks bcc9!

 

I would like to report that, on my mobo with Gigabyte GA-P55M-UD4 with Core i7 set as iMac11,1 there is no need for this flag as the "PCI Configuration Begins" doesn't stop/block. It's working without any further editing!

 

Also, when using this flag npci=0x2000 on my other Core2Quad ICH7 mobo, when running lspci -nn I do not get the "pcilib: 0000:01:00.0 64-bit device address ignored" error mentioned a few pages ago (this mobo is set as MacPro3,1). That's why you are again confirmed, bcc9.

 

Both my mobos have proper SMboardproduct set which seems important.

 

Should there be some code modification in Chameleon, now that you guys found the cause/effect of npci=0x2000? I am curious as to why this flag doesn't work for everyone, as we all have the SAME 10.6.8 update...

 

Thanks again to everyone.

finally its solved !

 

once again i was late to see your post fantomas1 , but your file will help other ATI users for sure!

 

i found the original ati 10.6.8 kexts in another post (16mb)

 

then i apply the netkas 10.6.8 fix

 

 

thanks for your posts fantomas1 ;)

 

 

happy for you ! ^_^

×
×
  • Create New...