Jump to content

[How to] Asus P5K-VM guide

delish

By delish,
in Installation Guides

 Share
1,244 posts in this topic

Recommended Posts

BlackCH
BlackCH
Posted
Posted
- No more Firewire 800 card

- Speedstep is only working on the first two CPUs out of four. The rest are stuck at the lowest speed.

 

For the second problem, I think I need to edit the equivalent of:

 

0x02 // This needs to be 0x04 if you have quad core CPU

 

switch, but where is it? Thanks to both of you, really appreciated.

 

			Method (_PSD, 0, NotSerialized)
		{
			Return (Package (0x05)
			{
				0x05,
				0x00,
				0x00,
				0xFC,
				0x02 // This needs to be 0x04 if you have quad core CPU
			})
		}

 

What about your firewire card? is gone completely from the IOReg?

 

@DB1

Do you have that value set to x04?

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted
What about your firewire card? is gone completely from the IOReg?

The closest thing for CPU count that I can find in the 3.2 is the following code, and it appears to be set to 4 CPUs. But not really sure.

			Name (_PSD, Package (0x05)
		{
			0x05,
			0x00,
			0x00,
			0xFC,
			0x04 
		})

 

FW800 card is not completely gone from the ioregistry. But not doing a whole lot. See attached pix.

IORegistry.zip

Link to comment
Share on other sites
DB1
DB1
Posted
Posted 

@DB1
Do you have that value set to x04?[/quote]

I previously had it set x02 for my core 2 but when updated I left it at 4. 4 is ok if you have core 2 but 4 is needed for quad - apparently.

@Bikinifarm

Just noticed this change in Chiefs v3.2 which I took into our v3.2

[codebox]                    Method (_DSM, 4, NotSerialized)
                    {
                        Store (Package (0x02)
                        {
                            "fwports",
                            Buffer (0x04)
                            {
                                0x02, 0x00, 0x00, 0x00
                            }
                        }, Local0)

                        MCDP (Arg2, RefOf (Local0))
                        Return (Local0)
                    }
                }
            }
[/codebox]

 

Try this instead it is what was in v3.1 and worked for you.

 

[codebox]Method (_DSM, 4, NotSerialized)
{
Store (Package (0x02)
{
"fwhub",
Buffer (0x02)
{
0x00, 0x00, 0x00, 0x00
}
}, Local0)

MCDP (Arg2, RefOf (Local0))
Return (Local0)
}
}
}
[/codebox]

 

I dont have firewire devices so cannot check. Let me know if it resolves so I can update the dsl.

 

which program must use to edit dsdt file?

 

[u][b]TextWrangler[/b][/u] linked in post #784

Link to comment
Share on other sites
BlackCH
BlackCH
Posted
Posted
The closest thing for CPU count that I can find in the 3.2 is the following code, and it appears to be set to 4 CPUs. But not really sure.

Yes, thats it. I was using a little different Scope (_PR) section

 

FW800 card is not completely gone from the ioregistry. But not doing a whole lot. See attached pix.

A bit of a guess but to me it looks like when HPET was conflicting with the onboard firewire (now solved issue). Check the IRQs assigned to PCI card and see if any devices has the same IRQ

post-51637-1259101889_thumb.jpg

This is how it look my airport PCI card.

See this as reference:

http://www.projectosx.com/forum/index.php?showtopic=564

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted

@DB1;

 

I tested out the fwhub thing, and did not make any difference. FW800 still missing in action.

 

@BlackCH;

 

I looked at the IRQ thing, but that looks like it is pretty involved.

 

I don't know what good it'll do, but I will try to compare 3.1 and 3.2, and see anything glaringly obvious. Otherwise, I guess I will have to stick with 3.1. The restart after sleep is not bothering that much... Thank you both.

 

PS. Speedstep is still only for the first two cores (v3.2) so that's a mystery too.

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
@DB1;

 

I tested out the fwhub thing, and did not make any difference. FW800 still missing in action.

 

@BlackCH;

 

I looked at the IRQ thing, but that looks like it is pretty involved.

 

I don't know what good it'll do, but I will try to compare 3.1 and 3.2, and see anything glaringly obvious. Otherwise, I guess I will have to stick with 3.1. The restart after sleep is not bothering that much... Thank you both.

 

PS. Speedstep is still only for the first two cores (v3.2) so that's a mystery too.

 

Try this one, I built it of the v3.1 i made for you previous that worked ok so in theory this should work (JMB0 taken out also). I left the FRWR as I originally posted, so if that's the only thing not working with this change as previous post.

 

Q6600bikinifarm3.2_dsdt.zip

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted
Try this one, I built it of the v3.1 i made for you previous that worked ok so in theory this should work (JMB0 taken out also). I left the FRWR as I originally posted, so if that's the only thing not working with this change as previous post.

 

DB1; The dsdt you made for me does work fully, including firewire and speedstep. Comparing your file and the one I did, I see that I made two wrong assumptions which likely caused the processor issue, and may be even the firewire issue. Apologies for creating additional work, and many thanks!

Link to comment
Share on other sites
maxk
maxk
Posted
Posted
Check post #733 for BIOS settings. Double check everything because it should be working.

I think the problem might be here:

//	OperationRegion (BIOS, SystemMemory, 0x7FF8E064, 0x2D) // This code needs to match the one from your original dsdt extract.
	OperationRegion (BIOS, SystemMemory, 0xCFF8E064, 0xFF)

Clean it up (delete upper line); try using '0x7FF8E064, 0x2D' value to see if it make a change

Also you left out this part of the code:

 

Thanks BlackCH! Seems like the problem was caused USBBusFix parameter in com.apple.Boot.plist. Once it was removed sleep started to work. Well at least partially. Keyboad/mouse wake up still not functioning - power supply just flashes briefly the first time any keyboard key is pressed, but not fully waking up. Anyway, it's better than nothing.

Oh, and I had to take out CPU3/4 lines from dsdt as my E2160 only has 2 cores.

 

Upd: seems like VoodooMonitor is also crashing after wake-up.

Link to comment
Share on other sites
Freaky Chokra
Freaky Chokra
Posted
Posted

Ok.. let's see..

 

The ATI issue.. that's why it also has troubles within (Any) Windows too... In the first place before AMD acquired ATI, it did have problems... with their Catalyst drivers... and did continue for half a decade or so.

That's why nVidia gained lead then... but now the scene has changed but not much from then... the drivers are not so "Finer" then the drivers of nVidia... for whichever model they make...

That's why nVidia still works Wonders on Windows os Mac OS X.... ATI Drivers are still rogue and often are reported for many bugs like crashing / quitting of games in midst of an intense shoot out. ;) tee hee

:offtopic: Now the problem about my friend's card.... :poster_oops:

I referred to the older XFX 9600GT... NOT the 9800GT... I bought it in March '08; He got it in Aug '08

So, we both experienced problems with the same model product from XFX, (more or less) exactly 18 months down the line.

But I "was forced" to get that replacement as 9600GT model was discontinued and I couldn't be given a new piece. So I shelled out $18.5.. as the difference in the 9800GT piece market price and my card's market price according to their product price sheet.... what Bull$h1t... right? XFX SUCKS for customer support! :rolleyes::blink::thumbsdown_anim:

Same thing was told to him too...

Actually my friend's card still works... but Games crash out when Anti-Aliasing is turned on or increased to higher levels... differs between games... like Wolfenstein.. NFS Shift.... Prototype... COD6:MW2, etc...

As he just wants a perfectly working card to enjoy games to the max...

So He is yet to decide.... nVidia or ATI...... OS I hope this thing is now clear to you..

Basically... it all means that the XFX product sucks.....big time....!!! end of discussion.... for XFX....(losers!) :angry2:

 

Now about Benchmarks...etc.

I have had an older 17 inch TFT-LCD monitor ... initially plugged in to the 9600GT ..then I got another 20 inch (both Samsung) TFT-LCD monitors on the 9600GT.... when correctly configured, says EFi or even DSDT (I wud always say... DSDT is safer....better, but tougher :weight_lift: ....I know!, ok?), dual output works... and amazingly... with different resolution options.... TV out etc...

 

That card had twin DVI + S-Video tv output jack..

Now after that card blew, my newer 9800GT also has the same dual DVI + TV (S-Video jack) output... and it still works..

Worked... coz I sold off my 17 inch Samsung TFT-LCD.... so I haven't changed anything anywhere.... you see..

 

It is B-E-A-U-T-I-F-U-L-L-Y :thumbsup_anim::trumpet: plug and play.. not plug and pray like WinBlows :censored2: $h1t!

 

Lastly, I can't comment on any benchmarks availability for GPU on Hackintosh machines or even native Mac models... lappies or deskpros... coz I think there aren't benhmarking software like Futuremark products available for Mac architecture.

 

So... the only thing that people are doing is running some OpenCL benchmarks like QJulia and Galaxy.. etc... to get a fair idea about GPU FLOPS.

 

There exists XBench and GeekBench... as benchmarking software.

 

Though, out of which, only XBench gives gives a fair report of your display card's performance in terms of refreshes and triangles drawn per second... etc... not full scale reports like Futuremark products... if you have ever used any of them... I've followed since 3DMark 98... till date... a tester you see...

 

So I would finally say... rest assured.... if anyone is going in for nVidia.. 9800GT or GTX is the best bet... cost wise and performance wise as I had said earlier... and it can easily be configured in Mac... Don't worry...there is me and Slice and who can even help build :moil: your own DSDT using this post's guides and references... it is really simple actually :guitar: ...ok? :angel:

 

And then you can enjoy a spanking Hackintosh :plane: ... which otherwise would have cost as a Mac Pro.. (desktop) about $2500+

 

 

-Regards,

 

:superman: Freaky Chokra :robot:

Link to comment
Share on other sites
maxk
maxk
Posted
Posted
Thanks BlackCH! Seems like the problem was caused USBBusFix parameter in com.apple.Boot.plist.

 

Turns out it wasn't USBBusFix, but rather "arch=i386" 32-bit mode kernel flag that was removed and which helped with the sleep. I put it back for Paragon NTFS support (not yet working in 64bit) and it can't go to sleep again. Have any of you guys tried 32bit mode with your dsdt? Is sleep working?

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
DB1; The dsdt you made for me does work fully, including firewire and speedstep. Comparing your file and the one I did, I see that I made two wrong assumptions which likely caused the processor issue, and may be even the firewire issue. Apologies for creating additional work, and many thanks!

 

No worries, at least you tried and know what to look out for next time. It don't take much to mess up a dsdt.

 

Turns out it wasn't USBBusFix, but rather "arch=i386" 32-bit mode kernel flag that was removed and which helped with the sleep. I put it back for Paragon NTFS support (not yet working in 64bit) and it can't go to sleep again. Have any of you guys tried 32bit mode with your dsdt? Is sleep working?

 

Everything works 32 or 64bit. It may be the Paragon NTFS support causing the problem. Post your dsdt.dsl and I'll check there's nothing there causing the problem.

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted

DB1;

 

Looks like I spoke too soon. I did not realize that the dsdt.aml that you included was the 3.1 version, and in order to get the 3.2, I needed to recompile dsdt_3.2_with_Q6600_CPU.dsl you included. Anyway, I did that this morning, and retested. The good news is that, the CPU thing has been resolved per the corrections you've made. However, the Firewire is not working, so may need to revisit the fwhub thing, per your earlier posting. Thanks again.

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
DB1;

 

Looks like I spoke too soon. I did not realize that the dsdt.aml that you included was the 3.1 version, and in order to get the 3.2, I needed to recompile dsdt_3.2_with_Q6600_CPU.dsl you included. Anyway, I did that this morning, and retested. The good news is that, the CPU thing has been resolved per the corrections you've made. However, the Firewire is not working, so may need to revisit the fwhub thing, per your earlier posting. Thanks again.

 

dsdt.aml is v3.2 but with the firewire section from v3.1. If you just compile the dsdt_3.2_with_Q6600-CPU.dsl that has the v3.2 firewire code then rename file to dsdt.aml. My mistake I intended to give you both options but it was late last night when I did it.

 

Where was my head last night! That has JMB0 in it.

 

Heres the v3.2 aml & dsl with the proper new firewire code & JMB0 removed.

Q6600_3.2_dsdt.zip

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted
dsdt.aml is v3.2 but with the firewire section from v3.1. If you just compile the dsdt_3.2_with_Q6600-CPU.dsl that has the v3.2 firewire code then rename file to dsdt.aml. My mistake I intended to give you both options but it was late last night when I did it.

 

The dsdt.aml has a modification date of Nov 16, and when you decompile it, it matches version 3.1 that you had kindly provided me with before. Same size as well. So, it must be pure version 3.1.

 

But no sweat, I understood where you were heading, and compiled dsdt_3.2_with_Q6600-CPU.dsl as is (with the 3.2 FW code), and as I reported before, that did not work for FW, but was fine otherwise. Then, I made the modification you mentioned in post #803, so reverting back to the 3.1 FW code essentially, but that also did not work. So, there is some other thing in there that is goofing the FW800 card. Thanks.

 

PS. Attaching the modded file I tested.

dsdt_3.2_with_Q6600_CPU_fwhub.dsl.zip

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
The dsdt.aml has a modification date of Nov 16, and when you decompile it, it matches version 3.1 that you had kindly provided me with before. Same size as well. So, it must be pure version 3.1.

 

But no sweat, I understood where you were heading, and compiled dsdt_3.2_with_Q6600-CPU.dsl as is (with the 3.2 FW code), and as I reported before, that did not work for FW, but was fine otherwise. Then, I made the modification you mentioned in post #803, so reverting back to the 3.1 FW code essentially, but that also did not work. So, there is some other thing in there that is goofing the FW800 card. Thanks.

 

PS. Attaching the modded file I tested.

 

The dsl & aml just posted in the update previous post is slightly different to the one you just compiled. Try it.

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted
The dsl & aml just posted in the update previous post is slightly different to the one you just compiled. Try it.

 

Tested the updated versions from post #814. Same as before, everything but the FW800 works.

 

You might wanna try deleting (or commenting out) the:

						IRQNoFlags ()
						{8}

from RTC, TIMR and IPIC devices. Maybe helps

 

Tested with TIMR and IPIC IRQNoFlags removed using DB1's latest. (RTC does not have the flag) Unfortunately, FW800 is still missing in action.

 

Thanks to both of you...

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
Tested the updated versions from post #814. Same as before, everything but the FW800 works.

 

Tested with TIMR and IPIC IRQNoFlags removed using DB1's latest. (RTC does not have the flag) Unfortunately, FW800 is still missing in action.

 

Thanks to both of you...

 

Found a slight difference in the PSD section of the last one I posted so have modified and checked against my own. Also there is a copy of the same v3.2 but with the firewire from v3.1, please test both

 

Q6600.zip

Link to comment
Share on other sites
maxk
maxk
Posted
Posted
Everything works 32 or 64bit. It may be the Paragon NTFS support causing the problem. Post your dsdt.dsl and I'll check there's nothing there causing the problem.

 

I took out all Paragon stuff along with VMWare. Same deal - sleep works in 64bit but not in 32. Even in 64bit mode keyboard/mouse can't wake up the system and cause only brief power supply blip. VoodooMonitor not working after wake-up.

I found no major differences in my dsdt file and the ones posted here recently. Dunno if there are any other components that may be causing the problem like old kext. I'm planning to set a new SL partition and see if virgin system is doing the same for me.

e2160_dsdt.zip

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
I took out all Paragon stuff along with VMWare. Same deal - sleep works in 64bit but not in 32. Even in 64bit mode keyboard/mouse can't wake up the system and cause only brief power supply blip. VoodooMonitor not working after wake-up.

I found no major differences in my dsdt file and the ones posted here recently. Dunno if there are any other components that may be causing the problem like old kext. I'm planning to set a new SL partition and see if virgin system is doing the same for me.

 

Your right nothing wrong with the dsdt you made. There may be left over configuration files or kexts following removal of Paragon or VMWare unless you used something like AppCleaner which removes everything.

 

A clean install may be the quickest way to sort out problems.

Link to comment
Share on other sites
maxk
maxk
Posted
Posted
Your right nothing wrong with the dsdt you made. There may be left over configuration files or kexts following removal of Paragon or VMWare unless you used something like AppCleaner which removes everything.

 

A clean install may be the quickest way to sort out problems.

Unfortunately even with clean install the system behaves exactly the same: 64bit sleep works, 32bit doesn't. Weird.

Link to comment
Share on other sites
DB1
DB1
Posted
Posted
Unfortunately even with clean install the system behaves exactly the same: 64bit sleep works, 32bit doesn't. Weird.

 

Are you using Bios version 1001? And is the Bios call number in dsdt as per your original extracted dsdt? Have you checked your Bios settings set as in previous posts?

 

Anyone else not able to sleep/wake from sleep in 32bit?

Link to comment
Share on other sites
bikinifarm
bikinifarm
Posted
Posted
Found a slight difference in the PSD section of the last one I posted so have modified and checked against my own. Also there is a copy of the same v3.2 but with the firewire from v3.1, please test both

 

Q6600.zip

 

Tested both of them. Neither one shows the FW800 card. Appear to be working fine otherwise. Interestingly, the built-in FireWire 400 does not seem to work in any of the cases including the 3.1 version. I am including the IOReg output from 3.2, 3.2 w/3.1 FW, and 3.1. Many thanks.

IOReg_etc.zip

Link to comment
Share on other sites
BlackCH
BlackCH
Posted
Posted
Tested both of them. Neither one shows the FW800 card. Appear to be working fine otherwise. Interestingly, the built-in FireWire 400 does not seem to work in any of the cases including the 3.1 version. I am including the IOReg output from 3.2, 3.2 w/3.1 FW, and 3.1. Many thanks.

 

did you try disabling the onboard firewire?

 

BTW, those ioreg are truncated. You need to use:

 

ioreg -lw0

 

or better yet; an IORegistryExplorer file. Is much easier to look at

Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...