Slow Hard Drive Performance ICH10

[orionquest]

THE STEPS

First, you would need to copy the iasl tool to your /usr/bin directory so it can be run from anywhere.

 

 

 

CSharpMac

 

Hi CSharpeMac, thanks, just what I was looking for. I'll report back with my findings.

 

My system is actually 10.5.8 on a P6T 4 core i7 with sata HD and DVD (dvd is the issue). From what as been stated in the thread the slow drive issue is with both 10.5.x and 10.6.

 

Thanks again.

 

oq

 

Is not the king is THe KiNG, pay attention on the details...if you do the same with your DSDT you may get trouble..

 

THe KiNG! Thanks for the reminder.

 

oq

[orionquest]

Hey Guys, well I tried the .iasl in terminal and got a bunch of errors "no such file or directory". Terminal is not my thing. I wish I could master it. I even trid to expand the zip file on my desktop and do a cd ~/ but for some reason I could even get to this directory without issue. I moved on and used iASLMe =) much better got the file and was able to edit and convert back to .aml file.

 

So far I tried out CSharpeMac edits. It works better. Seems the DVD drive each time I access it needs to "warm" up then starts to copy files. My test disk has a 30meg folder burned on a real mac and copies in 5 sec. On the Hack about a couple mins. Plus when I launch toast it lags and lags... So there is still and issue there.

 

I will now try THe KiNG's edits and let you know.

 

Thanks!

 

oq

[csharpmac]

Hey Guys, well I tried the .iasl in terminal and got a bunch of errors "no such file or directory". Terminal is not my thing. I wish I could master it. I even trid to expand the zip file on my desktop and do a cd ~/ but for some reason I could even get to this directory without issue. I moved on and used iASLMe =) much better got the file and was able to edit and convert back to .aml file.

 

So far I tried out CSharpeMac edits. It works better. Seems the DVD drive each time I access it needs to "warm" up then starts to copy files. My test disk has a 30meg folder burned on a real mac and copies in 5 sec. On the Hack about a couple mins. Plus when I launch toast it lags and lags... So there is still and issue there.

 

I will now try THe KiNG's edits and let you know.

 

Thanks!

 

oq

 

Hi,

I have just created a fast DSDT for your motherboard.

[csharpmac.aml.zip]Test it and inform me about your DVD speed.

CSharpMac

[orionquest]

Hi,

I have just created a fast DSDT for your motherboard (ASUS P6T LGA 1366).

 

Test it and inform me about your DVD speed.

CSharpMac

 

Hi CSharpeMac, it didn't work for me....

It got stuck at the gray apple screen no spinning circle

I booted with -v

It got to a point where it said "skipping sycronization cpo1..etc" then two lines PIC version 0x20 version 64.87 then second line with the last value at 88111

 

I had to boot off a second drive to get the DSDT back to my original.

 

Here is my DSDT file if you can take a look.

oqDSDT.aml.zip

 

Thanks

 

oq

[csharpmac]

Hi CSharpeMac,

 

Thanks

 

oq

Try this:

[csharpmac1.aml.zip]

Should work best on vanilla kernel and kexts

[Extra.zip]

CsharpMac

[orionquest]

Try this:

 

Should work best on vanilla kernel and kexts

 

CsharpMac

 

Awesome! I see you have included Chameleon 2 RC3 and some extenstions, no problem there. But quick question. How do I setup the vanilla kernal to startup by default.? I know there is a prompt command for when you boot, but it will be temporary. Sorry more noobness....

 

I'll let you know the results when I get home from work.

 

oq

[VultUx]

Here my results with my ASUS P6T motherboard:

 

CSharpMac solution

• solve Slow SATA issue
  
• but dont solve my problem with my SATA DVD drive (see below the erro msg)
  

26/09/09 10:05:48 kernel SerialATAPI device reconfiguration did not complete successfully. (failedCommandInfo = 0x1)

The King solution

• solve Slow SATA issue
  
• solve my SATA DVD issue too
  

I prefer CSharpMac solution because change less code, but with DVD SATA problem Im using the two patches just for sure.

 

Thank you guys!

 

[]s

VultU

Curitiba - Brasil

[orionquest]

SUCCESS!!!

 

It's all working now. I used THeKiNG's notes and patched my DSDT. Burning/Mounting/Copying works like it should.

 

Thanks for all the input and education!

 

 

Cheers!

 

oq

[THe KiNG]

I prefer CSharpMac solution because change less code...

If you are worry about changing code, you shouldn't use OS X on a hack, stick with Windows, will work with all that {censored} from DSDT, no need to change anything...or use linux with 9998 fixes in the kernel...

There is a "word" around here: "If you didn't speak you were a philosopher"...

SUCCESS!!!

Great, I'm happy that you learned to do it...

[VultUx]

If you are worry about changing code, you shouldn't use OS X on a hack, stick with Windows, will work with all that {censored} from DSDT, no need to change anything...or use linux with 9998 fixes in the kernel...

There is a "word" around here: "If you didn't speak you were a philosopher"...

You are right THe KiNG!

 

Windows no way I prefer live with all that codes stuff.

 

Thank you again guys!

 

[]s

VultU

Curitiba - Brasil

[John.Appleseed]

NEW PROBLEM: Sound stuttering when CPU usage.

 

REASON: SpeedStep keeps on switching too frequently.

 

After testing and chatting with CSharpMac about my DSDT and SpeedStep here is the SOLUTION:

 

Removing all the IRQs from the following three devices:

(i) Device RTC / RTC0,

(ii) Device TMR / TIMR, and

(iii) Device PIC / IPIC.

 

 

 

Now I have SpeedStep and all I had to do was the following

(i) add the Mac Pro CST to my DSDT, (ii) delete the IRQs above, and (iii) create a LPC kext.

I have also modified the LPC device names in the DSDT as well, including the LPC itself which is named SBRG in my DSDT- on a real iMac/MacPro its name is LPCB.

Also, I have noticed that KEXTs use the "device-id" of devices attached to the PCI bus to decide whether to start and to tweak their behavior depending on the specific device. This allows KEXTs to be tricked into believing that you have hardware that is used by Apple!

 

Thanks CSharpMac!

[csharpmac]

NEW PROBLEM: Sound stuttering when CPU usage.

REASON: SpeedStep keeps on switching too frequently.

After testing and chatting with CSharpMac about my DSDT and SpeedStep here is the SOLUTION:

 

Removing all the IRQs from the following three devices:

(i) Device RTC / RTC0,

(ii) Device TMR / TIMR, and

(iii) Device PIC / IPIC.

 

Now I have SpeedStep and all I had to do was the following:

(i) add the Mac Pro CST to my DSDT, (ii) delete the IRQs above, and (iii) create a LPC kext.

I have also modified the LPC device names in the DSDT as well, including the LPC itself which is named SBRG in my DSDT- on a real iMac/MacPro its name is LPCB.

Also, I have noticed that KEXTs use the "device-id" of devices attached to the PCI bus to decide whether to start and to tweak their behavior depending on the specific device. This allows KEXTs to be tricked into believing that you have hardware that is used by Apple!

 

Thanks CSharpMac!

 

John as we have discussed:

 

a. Search and Replace SBRG with LPCB

 

Will need to add dev id to AppleLPC.kext

<string>pci10de,aac</string>

 

This Device (SBRG) in our new DSDT, renamed and with the device-id patching code in place.

Rename all instances of SBRG to LPCB using search and replace.

This makes the LPC device appear correctly in the PCI device tree - as LPCB@1F under PCI0@0.

b. LPC Devices

 

i. Search and Replace DMAD with DMAC

ii. Search and Replace RTC0 with RTC

iii. Search and Replace OMSC with LDRC

iv. Search and Replace COPR with MATH

v. Search and Replace (PIC) with (IPIC)

vi. Search and Replace TMR with TIMR

 

c.

 

i. Search and Replace USB0 with OHC1

ii. Search and Replace US15 with OHC2

iii. Search and Replace USB2 with EHC1

iv. Search and Replace US12 with EHC2

 

d. Search and Replace SMB0 with SBUS

 

e. Search and Replace TRM0 with TRIM

[John.Appleseed]

John as we have discussed:

a. Search and Replace SBRG with LPCB

 

Will need to add dev id to AppleLPC.kext

<string>pci10de,aac</string>

b. LPC Devices

 

a. Search and Replace DMAD with DMAC

b. Search and Replace RTC0 with RTC

c. Search and Replace OMSC with LDRC

d. Search and Replace COPR with MATH

e. Search and Replace (PIC) with (IPIC)

f. Search and Replace TMR with TIMR

 

It is now working great!

I have now P-state switching and I fixed the skipping sound.

Thanks CSharpMac!

[VultUx]

Ok, I did some test and know I found a intermediate solution that I dont need to change too much code.

 

I only changed the DEVICE (PIC) and DEVICE (TMR) removing only the IRQ thing like suggested by Csharp (thanks man), with this two changes the SLOW SATA issue are solved and my DVD SATA drive works too.

 

[]s

VultU

 

 

ps.: sorry about the bad english.

[THe KiNG]

I can't believe how some don't even bother to read or use their brain...

Anyway here is the code that will fix everything for some people:

        Name (BRN, Zero)
       Name (GGL, One)
       Method (N00B, 0, NotSerialized)
       {
           If (LNotEqual (BRN, Zero))
           {
               Return (GGL)
           }
           Else
           {
               Return (Zero)
           }
       }

       Method (RTFM, 0, NotSerialized)
       {
           If (LGreater (BRN, Zero))
           {
               Return (One)
           }
           Else
           {
               Return (Zero)
           }
       }

       Method (STFU, 0, NotSerialized)
       {
           If (LEqual (BRN, Zero))
           {
               Return (One)
           }
           Else
           {
               Return (Zero)
           }
       }

 

In case some still don't get what does that code:

BRN=brain

GGL=google

 

Lemme know if I have to explain the rest of the code...

[Apple.User]

Looking at the suggestions here and following the recommendations from the users, I am perplexed while observing the dsdt file that was supplied to jimmydigital00

On that dsdt file the Device (TIMR) is as follows:

				{
				Name (_HID, EisaId ("PNP0100"))
				Name (_CRS, ResourceTemplate ()
				{
					IO (Decode16,
						0x0040,			 // Range Minimum
						0x0040,			 // Range Maximum
						0x00,			   // Alignment
						0x04,			   // Length
						)
				})
			}

 

But on the suggestion presented says clearly to replace the Device (TIMR) as:

			  {
				Name (_HID, EisaId ("PNP0100"))
				Name (_CRS, ResourceTemplate ()
				{
					IO (Decode16,
						0x0040,			 // Range Minimum
						0x0040,			 // Range Maximum
						0x01,			   // Alignment
						0x04,			   // Length
						)
						IO (Decode16,
						0x0050,			 // Range Minimum
						0x0050,			 // Range Maximum
						0x01,			   // Alignment
						0x04,			   // Length
						)
				})
			}

 

Why that proposed modification was not performed on the dsdt? Was it actually required?

Was as previously noted in the end enough just the simple removal of IRQs on (RTC) (PIC) and (TIMR) ?

[THe KiNG]

Read my previous post^^^

[orionquest]

NEW PROBLEM: Sound stuttering when CPU usage.

 

REASON: SpeedStep keeps on switching too frequently.

 

After testing and chatting with CSharpMac about my DSDT and SpeedStep here is the SOLUTION:

 

Removing all the IRQs from the following three devices:

(i) Device RTC / RTC0,

(ii) Device TMR / TIMR, and

(iii) Device PIC / IPIC.

 

 

 

Now I have SpeedStep and all I had to do was the following

(i) add the Mac Pro CST to my DSDT, (ii) delete the IRQs above, and (iii) create a LPC kext.

I have also modified the LPC device names in the DSDT as well, including the LPC itself which is named SBRG in my DSDT- on a real iMac/MacPro its name is LPCB.

Also, I have noticed that KEXTs use the "device-id" of devices attached to the PCI bus to decide whether to start and to tweak their behavior depending on the specific device. This allows KEXTs to be tricked into believing that you have hardware that is used by Apple!

 

Thanks CSharpMac!

 

What are you doing when you get the sound shuttering? iTunes, QT movie? Web page? I haven't noticed any sound shuttering, and I use itunes sometimes all day streaming music?

 

Just my 2 cents.

 

oq

[Apple.User]

Read my previous post^^^

 

My question to THe KiNG:

Why the modification proposed by you on the Device (TIMR) was not performed on the dsdt file that you provided?

I want to make sure that is required or is just cosmetic.

In my system I see no difference.

 

 

Thanks

 

 

							0x01,			   // Alignment
						0x04,			   // Length
						)
						IO (Decode16,
						0x0050,			 // Range Minimum
						0x0050,			 // Range Maximum
						0x01,			   // Alignment
						0x04,			   // Length

[THe KiNG]

Intel DX58SO:

                Device (TMR)
               {
                   Name (_HID, EisaId ("PNP0100"))
                   Name (_CRS, ResourceTemplate ()
                   {
                       IO (Decode16,
                           0x0040,             // Range Minimum
                           0x0040,             // Range Maximum
                           0x01,               // Alignment
                           0x04,               // Length
                           )
                       IO (Decode16,
                           0x0050,             // Range Minimum
                           0x0050,             // Range Maximum
                           0x01,               // Alignment
                           0x04,               // Length
                           )
                       IRQNoFlags ()
                           {0}
                   })
               }

 

Apple MacPro4,1:

                Device (TIMR)
               {
                   Name (_HID, EisaId ("PNP0100"))
                   Name (_CRS, ResourceTemplate ()
                   {
                       IO (Decode16,
                           0x0040,             // Range Minimum
                           0x0040,             // Range Maximum
                           0x01,               // Alignment
                           0x04,               // Length
                           )
                       IO (Decode16,
                           0x0050,             // Range Minimum
                           0x0050,             // Range Maximum
                           0x01,               // Alignment
                           0x04,               // Length
                           )
                   })
               }

Do I have to explain more?

[Apple.User]

Intel DX58SO:

Device (TMR)...

 

Apple MacPro4,1:

Device (TIMR)...

 

 

Great thanks!

I am also comparing the dsdt file from the Apple MacPro4,1

THe KiNG's dsdt has the typical Asus board format

   (_PR)
   {
       Processor (P001, 0x01, 0x00000810, 0x06) {}
       Processor (P002, 0x02, 0x00000000, 0x00) {}
       Processor (P003, 0x03, 0x00000000, 0x00) {}
       Processor (P004, 0x04, 0x00000000, 0x00) {}
       Processor (P005, 0x05, 0x00000000, 0x00) {}
       Processor (P006, 0x06, 0x00000000, 0x00) {}
       Processor (P007, 0x07, 0x00000000, 0x00) {}
       Processor (P008, 0x08, 0x00000000, 0x00) {}
       Processor (P009, 0x09, 0x00000000, 0x00) {}
       Processor (P010, 0x0A, 0x00000000, 0x00) {}
       Processor (P011, 0x0B, 0x00000000, 0x00) {}
       Processor (P012, 0x0C, 0x00000000, 0x00) {}
       Processor (P013, 0x0D, 0x00000000, 0x00) {}
       Processor (P014, 0x0E, 0x00000000, 0x00) {}
       Processor (P015, 0x0F, 0x00000000, 0x00) {}
       Processor (P016, 0x10, 0x00000000, 0x00) {}
   }

...But comparing the above dsdt with the Apple MacPro4,1

that is clearly different

   (\_PR)
   {
       Processor (CPU0, 0x00, 0x00000410, 0x06) {}
       Processor (CPU1, 0x01, 0x00000410, 0x06) {}
       Processor (CPU2, 0x02, 0x00000410, 0x06) {}
       Processor (CPU3, 0x03, 0x00000410, 0x06) {}
       Processor (CPU4, 0x04, 0x00000410, 0x06) {}
       Processor (CPU5, 0x05, 0x00000410, 0x06) {}
       Processor (CPU6, 0x06, 0x00000410, 0x06) {}
       Processor (CPU7, 0x07, 0x00000410, 0x06) {}
       Processor (CPU8, 0x08, 0x00000410, 0x06) {}
       Processor (CPU9, 0x09, 0x00000410, 0x06) {}
       Processor (CPUA, 0x0A, 0x00000410, 0x06) {}
       Processor (CPUB, 0x0B, 0x00000410, 0x06) {}
       Processor (CPUC, 0x0C, 0x00000410, 0x06) {}
       Processor (CPUD, 0x0D, 0x00000410, 0x06) {}
       Processor (CPUE, 0x0E, 0x00000410, 0x06) {}
       Processor (CPUF, 0x0F, 0x00000410, 0x06) {}
   }

Observing the two dsdts above I have replaced all P00x to CPUx and added to each line the 0x00000810, 0x06

  
Scope (_PR)
   {
       Processor (CPU0, 0x01, 0x00000810, 0x06) {}
       Processor (CPU1, 0x02, 0x00000810, 0x06) {}
       Processor (CPU2, 0x03, 0x00000810, 0x06) {}
       Processor (CPU3, 0x04, 0x00000810, 0x06) {}
       Processor (CPU4, 0x05, 0x00000810, 0x06) {}
       Processor (CPU5, 0x06, 0x00000810, 0x06) {}
       Processor (CPU6, 0x07, 0x00000810, 0x06) {}
       Processor (CPU7, 0x08, 0x00000810, 0x06) {}
       Processor (CPU8, 0x09, 0x00000810, 0x06) {}
       Processor (CPU9, 0x0A, 0x00000810, 0x06) {}
       Processor (CPUA, 0x0B, 0x00000810, 0x06) {}
       Processor (CPUB, 0x0C, 0x00000810, 0x06) {}
       Processor (CPUC, 0x0D, 0x00000810, 0x06) {}
       Processor (CPUD, 0x0E, 0x00000810, 0x06) {}
       Processor (CPUE, 0x0F, 0x00000810, 0x06) {}
       Processor (CPUF, 0x10, 0x00000810, 0x06) {}
   }

 

As I have guessed all the CPUs are listed correctly as my MacPro on the IORegistryExplorer both at 10.5.8 and at 10.6.

I have also tried all the other combinations of the above format by replacing for example only the CPUx and let all the other with their initial values:

So Processor (P001, 0x01, 0x00000810, 0x06) {}...becomes

1st combination : Processor (CPU0, 0x01, 0x00000810, 0x06) {}

having as 2nd line: Processor (CPU1, 0x02, 0x00000000, 0x00) {}... etc

 

2nd combination : Processor (CPU0, 0x00, 0x00000810, 0x06) {}

having as 2nd line: Processor (CPU1, 0x01, 0x00000000, 0x00) {}... etc

 

Then I repeat the above, and I add to each line the 0x00000810, 0x06

3rd combination : Processor (CPU0, 0x01, 0x00000810, 0x06) {}

having as 2nd line: Processor (CPU1, 0x02, 0x00000810, 0x06) {}... etc

 

4th combination : Processor (CPU0, 0x00, 0x00000810, 0x06) {}

having as 2nd line: Processor (CPU1, 0x01, 0x00000810, 0x06) {}... etc

 

All the possible CPUx combinations are having the same result.

But looking closer at the IORegistryExplorer using 10.5.8 I did have noticed something very interesting... there was one tiny difference: the ACPI_SMC_PlatformPlugin is having the PerformaceStateArray (with 10 values) only while the dsdt is on P00x format(!) Any ideas why is this so? Why the P00x vs CPUx etc notation makes such difference ? Nevertheless, the system works at its best and the IORegistryExplorer looks now the same with the macpro4,1.

[eneris]

Hello everyone! I have the GA-EX58-UD4P, before I used the iDeneb 10.5.8, and the icons of the HDD were in orange like the icon of the green drive TIMEMACHINE now with SnowLeopard the desktop icon has the appearance of an HDD, and be VERY slow, even to throw the WORD, iMovie also when I try to access the HDD pra download a video. I tried the procedure of THE KING, but without success, please find attached my DSDT without alterations. HELP ME. Thank you.

 

http://www.mediafire.com/download.php?l2akyinjnbj

[THe KiNG]

@Apple.User see Method N00B and RTFM since in your case BRN is not 0

A hint for you look on APIC table on Processor ID & Local Apic ID...

[XPbIM3]

i am keep reading this thread and one big question remains about such a "mysterios" issue:

 

Why on my old good p5q with 10.5.8 the HDD's and DVD'd performance was awesome(without ANY DSDT used, just vanila+ additional kexts, including ich10r fix), on my current p6t SE + ideneb 1.3 + sse kernel without DSDT the performance still awesome....

 

BUT! with my current p6tse+totaly vanila 10.6.1 +cham rc3 + one(yeah!) fakeSMC2.kext my system shows insane performance, but getting problems accessing secondary HHDs and DVDs.

 

(offcourse i can add an proper DSDT with deleted aliases+HPET patch + add some AHCI injectors, but problem still persists....)

 

How we guys can explain this?

 

Maybe chasing for the irq conflicts inside DSDT is the wrong path? as i can understand ressolving IRQ/IO conflicts is the OS job....or not?

 

 

 

 

anyway.... i am still keep playing with my system....looking for results....

 

 

 

oh.... one little thing:

is there anybody with X58 NON asus MoBO experiencing such a problems?

is there anybody with x58 ASUS MoBOs , working fine out-of-the-box? (i can give a step-by-step manual for x58 MB's how to instal vanilla 10.6.1 and check HDD performance....anyone needs?)

[Pluskat5000Schiffe]

i am keep reading this thread and one big question remains about such a "mysterios" issue:...

 

....anyone needs?)

 

Exactly my thoughts, why is our Asus {censored} causing these delays in OSX Snow. I can install 10.6.1 within 30 min up and running. I have Video - Lan - Sound all running in 64 bit mode. The thing i noticed is that when I use the EvoEnabler together with the QE-CL patch from netkas for exotic cards (Ati 4890) the system complains about SerialAtapa Reconfiguration error (0x1) sometimes (0x2) after the system reboots - shutdown's on the end of the kill processes etc.

 

I have a {censored} rampage Extreme II with Ati 4890 Graphics and plenty of disks. Maybe somebody can look to my dsdt.

DSDTFilesRampageExtremeII.rar