HD3000 freezing in Sierra

[pandaDS3]

HD3000 works with QE/CI, but dock glitching and Sierra freezing, when i folding windows

 

 

 

 

 

 

 

     

 

[vrishcons]

With Intel HD3000, you're supposed to inject Intel graphics in Sierra. You do not.

 

You would also normally inject HD3000 IGPU properties in your DSDT device @0x00020000; something like this:

           Method (_DSM, 4, NotSerialized)  // _DSM: Device-Specific Method
           {
                If (LEqual (Arg2, Zero))
                {
                    Return (Buffer (One)
                    {
                         0x03                                           
                    })
                }
                Return (Package (0x0A)
                {
                    "AAPL00,DualLink", 
                    Buffer (0x04)
                    {
                         0x00, 0x00, 0x00, 0x00                         
                    }, 
                    "device-id", 
                    Buffer (0x04)
                    {
                         0x26, 0x01, 0x00, 0x00                         
                    }, 
                    "model", 
                    Buffer (0x18)
                    {
                        "Intel HD Graphics 3000"
                    }, 
                    "AAPL,snb-platform-id", 
                    Buffer (0x04)
                    {
                         0x00, 0x00, 0x01, 0x00                         
                    }, 
                    "hda-gfx", 
                    Buffer (0x0A)
                    {
                        "onboard-1"
                    }
                })
            }

Obviously DualLink would be set to 1 if your built-in LCD has a resolution higher than 1600x900                                                                                                                                                                                                                 

IGPU properties in your DSDT device @0x00020000, but where to enter. I didnt find mentioned above. Is this entry in config.plist?

[Dr. Hurt]

I found out that the single most influential workaround for HD3000 artifacts is "slide=0" as a boot-arg.

 

Ever since updating to Sierra, I do get occasional freezes, for which I have yet to find a fix.

It seems to be related to memory pressure though. Using RAM intensive apps, I frequently get artifacts then system freezes after a few minutes.

[febriyan1302]

try boot args : "slide=0" 

it will be reduce artifact, but can't solve this problem.

[vrishcons]

try boot args : "slide=0" 

it will be reduce artifact, but can't solve this problem.

If resolution is native 1366x768, I do get freezing especially when minimizing and Maximizing browser.  I started having this problem from El Capitan on wards. I started using SwitchResx and added custom 1366x768. I changed Back Porch Vertical from 1 to 2. and saved it. Restart and select from switchresx custom 1366x768. The problem solved. 

[Dr. Hurt]

UPDATE: I've done a lot of experimenting and researching and here are my findings:

 

- Freezes are caused by a mismatch between the ports defined in the driver's frame buffer table and the physical ports in the laptop.

I fixed them by loading the second frame buffer table which limits the driver to 1 LVDS port. I don't use HDMI or VGA so I don't mind.

I think the same could be done by patching the first table to match BOTH port number and type. I prefer DSDT since it's cleaner.

"AAPL,snb-platform-id", 
Buffer (0x04)
{
0x00, 0x00, 0x02, 0x00    //load the 2nd frame buffer table (0x02)                   
}, 

- Artifacts are caused by IGPU memory corruption due a clover bug.

Clover somehow messes with the BIOS-reserved memory for IGPU.

apianti and Slice know about the issue and couldn't fix it (very long ago)

 

Slide=0 as boot-arg reduces the artifacts significantly.

Installing 8GB of RAM also seems to fix it. I haven't tried that.

 

Hope this helps

[gujiangjiang]

I have also have this problem.

 

Before 10.11.4 wo can use slide=0 to reduce it but after 10.11.4 the slide=0 have no use to reduce this problem.

[Sherlocks]

I have also have this problem.

 

Before 10.11.4 wo can use slide=0 to reduce it but after 10.11.4 the slide=0 have no use to reduce this problem.

 

try this

{ 0x00, 0x00, 0x00, 0x00 }, //0 "AAPL00,PixelFormat"

{ 0x00, 0x00, 0x00, 0x00 }, //1 "AAPL00,T1"

{ 0x14, 0x00, 0x00, 0x00 }, //2 "AAPL00,T2"

{ 0xfa, 0x00, 0x00, 0x00 }, //3 "AAPL00,T3"

{ 0x2c, 0x01, 0x00, 0x00 }, //4 "AAPL00,T4"

{ 0x00, 0x00, 0x00, 0x00 }, //5 "AAPL00,T5"

{ 0x14, 0x00, 0x00, 0x00 }, //6 "AAPL00,T6"

{ 0xf4, 0x01, 0x00, 0x00 }, //7 "AAPL00,T7"

{ 0x00, 0x00, 0x00, 0x00 }, //8 "AAPL00,LinkType"

{ 0x00, 0x00, 0x00, 0x00 }, //9 "AAPL00,LinkFormat"

{ 0x00, 0x00, 0x00, 0x00 }, //10 "AAPL00,DualLink"         -> if you have high resolution you have to { 0x01, 0x00, 0x00, 0x00 }

{ 0x00, 0x00, 0x00, 0x00 }, //11 "AAPL00,Dither"

{ 0x00, 0x00, 0x00, 0x00 }, //12 "AAPL00,DataJustify"

{ 0x00, 0x00, 0x00, 0x00 }, //13 "graphic-options"

[gujiangjiang]

try this

{ 0x00, 0x00, 0x00, 0x00 }, //0 "AAPL00,PixelFormat"

{ 0x00, 0x00, 0x00, 0x00 }, //1 "AAPL00,T1"

{ 0x14, 0x00, 0x00, 0x00 }, //2 "AAPL00,T2"

{ 0xfa, 0x00, 0x00, 0x00 }, //3 "AAPL00,T3"

{ 0x2c, 0x01, 0x00, 0x00 }, //4 "AAPL00,T4"

{ 0x00, 0x00, 0x00, 0x00 }, //5 "AAPL00,T5"

{ 0x14, 0x00, 0x00, 0x00 }, //6 "AAPL00,T6"

{ 0xf4, 0x01, 0x00, 0x00 }, //7 "AAPL00,T7"

{ 0x00, 0x00, 0x00, 0x00 }, //8 "AAPL00,LinkType"

{ 0x00, 0x00, 0x00, 0x00 }, //9 "AAPL00,LinkFormat"

{ 0x00, 0x00, 0x00, 0x00 }, //10 "AAPL00,DualLink"

{ 0x00, 0x00, 0x00, 0x00 }, //11 "AAPL00,Dither"

{ 0x00, 0x00, 0x00, 0x00 }, //12 "AAPL00,DataJustify"

{ 0x00, 0x00, 0x00, 0x00 }, //13 "graphic-options"

I have test this prop 3 months ago but have no use.

 

I download a vanilla MacBookPro8,1 ioreg file from web and i find all the prop from the ioreg and i try to inject it through the clover.

 

However the gitch cant be solved.

 

The only way to fix this is downgrade the osx to 10.11.3 and use slide=0,after the 10.11.4,the slide=0 have nothing to do with gitch.

 

 

从我的 iPhone 发送，使用 Tapatalk

[Sherlocks]

I have test this prop 3 months ago but have no use.

 

I download a vanilla MacBookPro8,1 ioreg file from web and i find all the prop from the ioreg and i try to inject it through the clover.

 

However the gitch cant be solved.

 

The only way to fix this is downgrade the osx to 10.11.3 and use slide=0,after the 10.11.4,the slide=0 have nothing to do with gitch.

 

 

从我的 iPhone 发送，使用 Tapatalk

What is your ram?

 

I had hd3000 laptop 4months ago.

 

 

 

In ML, i exprienced glitch with 4g ram.

 

After added 4g ram(total 8g), resolved.

 

Ofc, l always use slide=0

 

In sirrea, Vere rare exprience this problem. Like you said, slide=0 is not work properly in sierra.

 

If someone give ioreg file of latest MBP8,1 in Sierra, we could research.

 

나의 LG-F410S 의 Tapatalk에서 보냄

[gujiangjiang]

What is your ram?

 

I had hd3000 laptop 4months ago.

 

 

 

In ML, i exprienced glitch with 4g ram.

 

After added 4g ram(total 8g), resolved.

 

Ofc, l always use slide=0

 

In sirrea, Vere rare exprience this problem. Like you said, slide=0 is not work properly in sierra.

 

 

My ram is 2G+4G

 

The gitch from 10.8 till now,in 10.8-10.11.3 it can be easy fixed my add slide=0 and have no problem.

 

However from 10.11.4 till now i cant find way to solve this.

 

I have tried add properties,edit edid information,inject vbios,add ram,add some boot-flag,etc but still have gitch.

 

The problem is not gitch but freezes and must force shut down the osx.

 

PS:rehabman mentioned it may be related with DMVT,but does HD3000 have DMVT?and how or where to patch it?

 

 

从我的 iPhone 发送，使用 Tapatalk

[Sherlocks]

My ram is 2G+4G

 

The gitch from 10.8 till now,in 10.8-10.11.3 it can be easy fixed my add slide=0 and have no problem.

 

However from 10.11.4 till now i cant find way to solve this.

 

I have tried add properties,edit edid information,inject vbios,add ram,add some boot-flag,etc but still have gitch.

 

The problem is not gitch but freezes and must force shut down the osx.

 

PS:rehabman mentioned it may be related with DMVT,but does HD3000 have DMVT?and how or where to patch it?

 

 

从我的 iPhone 发送，使用 Tapatalk

Maybe.. related in DVMT.

 

You did test all cases except DVMT change in bios(we can't common DVMT) or hex patch(snbbuffer.kext or kernel)

 

1. If desktop user with HD 3000 can test changed higher dvmt value, it will be cleared.

 

2. We should find difference kernel source related slide=0 between 10.11.3 and 10.11.4.

 

3. Search hex of dvmt in SNBframebuffer

 

 

 

 

 

나의 LG-F410S 의 Tapatalk에서 보냄

[gujiangjiang]

Maybe.. related in DVMT.

 

You did test all cases except DVMT change in bios(we can't common DVMT) or hex patch(snbbuffer.kext or kernel)

 

1. If desktop user with HD 3000 can test changed higher dvmt value, it will be cleared.

 

2. We should find difference kernel source related slide=0 between 10.11.3 and 10.11.4.

 

3. Search hex of dvmt in SNBframebuffer

 

 

 

 

 

나의 LG-F410S 의 Tapatalk에서 보냄

My laptop have no DMVT select in my bios.

 

I have searched the web but cant find any information about snbframebuffer dmvt patch.

 

Ps:my hd3000 dmvt is 32M.

 

 

从我的 iPhone 发送，使用 Tapatalk

[Sherlocks]

My laptop have no DMVT select in my bios.

 

I have searched the web but cant find any information about snbframebuffer dmvt patch.

 

Ps:my hd3000 dmvt is 32M.

 

 

从我的 iPhone 发送，使用 Tapatalk

Okay i will research difference between 10.11.3 source and 10.11.4

 

Anyway did you surely see that slide=0 works or not?

 

나의 LG-F410S 의 Tapatalk에서 보냄

[gujiangjiang]

Okay i will research difference between 10.11.3 source and 10.11.4

 

Anyway did you surely see that slide=0 works or not?

 

나의 LG-F410S 의 Tapatalk에서 보냄

Yes I'm pretty sure the slide=0 can solve the gitch and freeze in HD3000 from 10.8.x-10.11.3.

I use HD3000 from 10.6.3 and in 10.8 it had gitch and freeze problem and fixed well by slide=0.

 

I noticed that 10.11.4 enforced the SIP so i dont know if this is the problem cause the slide=0 failed.

 

 

从我的 iPhone 发送，使用 Tapatalk

[Sherlocks]

Yes I'm pretty sure the slide=0 can solve the gitch and freeze in HD3000 from 10.8.x-10.11.3.

I use HD3000 from 10.6.3 and in 10.8 it had gitch and freeze problem and fixed well by slide=0.

 

I noticed that 10.11.4 enforced the SIP so i dont know if this is the problem cause the slide=0 failed.

 

 

从我的 iPhone 发送，使用 Tapatalk

Thanks.

 

Hope someone support latest ioreg file mbp81(like skylake glitches fix by supported ioreg in sierra) or desktop user's change dvmt result or you add 8g(4+4) ram and result.

 

I will research source and report result you.

 

Thank you.

 

나의 LG-F410S 의 Tapatalk에서 보냄

[gujiangjiang]

Thanks.

 

Hope someone support latest ioreg file mbp81(like skylake glitches fix by supported ioreg in sierra) or desktop user's change dvmt result or you add 8g(4+4) ram and result.

 

I will research source and report result you.

 

Thank you.

 

나의 LG-F410S 의 Tapatalk에서 보냄

I only have ioreg 10.9.x but the properties have no use to solve this problem.

 

I search the web cant see any MacBookPro8,x ioreg use 10.12.x.

 

 

从我的 iPhone 发送，使用 Tapatalk

[Sherlocks]

I only have ioreg 10.9.x but the properties have no use to solve this problem.

 

I search the web cant see any MacBookPro8,x ioreg use 10.12.x.

 

 

从我的 iPhone 发送，使用 Tapatalk

Me too. So we need latest ioreg.

This problem can resolve simply or not like slide=0 and skylake case.

 

 

 

나의 LG-F410S 의 Tapatalk에서 보냄

[Sergey_Galan]

Me too. So we need latest ioreg.

This problem can resolve simply or not like slide=0 and skylake case.

 

 

 

나의 LG-F410S 의 Tapatalk에서 보냄

HD3000 on the empty connector. In case of video playback by means of the hardware decoding (IQSV) there are black and/or green bands. To escape from it it was possible to register slide=0 key, but it any more does not work.

[makk]

On 1/22/2017 at 3:00 AM, Dr. Hurt said:

UPDATE: I've done a lot of experimenting and researching and here are my findings:

 

- Freezes are caused by a mismatch between the ports defined in the driver's frame buffer table and the physical ports in the laptop.

I fixed them by loading the second frame buffer table which limits the driver to 1 LVDS port. I don't use HDMI or VGA so I don't mind.

I think the same could be done by patching the first table to match BOTH port number and type. I prefer DSDT since it's cleaner.

"AAPL,snb-platform-id", 
Buffer (0x04)
{
0x00, 0x00, 0x02, 0x00    //load the 2nd frame buffer table (0x02)                   
}, 

- Artifacts are caused by IGPU memory corruption due a clover bug.

Clover somehow messes with the BIOS-reserved memory for IGPU.

apianti and Slice know about the issue and couldn't fix it (very long ago)

 

Slide=0 as boot-arg reduces the artifacts significantly.

Installing 8GB of RAM also seems to fix it. I haven't tried that.

 

Hope this helps

 

Dr. Hurt

 

Thank you for your answer.

 

I've just switched to the second Framebuffer to see if it works with High Sierra HD3000 kexts in Mojave.  

 

I used to use HDMI to watch movies and use the TV screen to do work.  But, I no longer have TV. 

 

Do you have a loss of the CL/QE?

 

Also need to get the ports to change from 4000 to 8000 ?

 

The port change.

 

Have to reconfigure.

 

This laptop cannot upgrade beyond 8GB men.  

Need 16GB.  There is some kind of software that emulates RAM.  I used in in Windows.

Ram drive.

But these SSD drives should have the ability to use Swap like in Solaris/HP

 

Virtual memory, swap, RAMDrive.

 

I hope to figure this out.

 

Thank you Dr.

[makk]

Intel HD Graphics 2000/3000 (Sandy Bridge processors)

Supported from Mac OS X 10.7.x to macOS 10.13.6. The instructions are for OS X 10.8.x - macOS 10.13.6. On older operating systems follow the "ancient ways". On newer operating systems these are not supported. But if you really want to - read this.Metal support is absent.
SNB framebuffer list:
— 0x00010000 (mobile, 4 connectors, no fbmem)
— 0x00020000 (mobile, 1 connectors, no fbmem)
— 0x00030010 (desktop, 3 connectors, no fbmem)
— 0x00030020 (desktop, 3 connectors, no fbmem)
— 0x00030030 (desktop, 0 connectors, no fbmem)
— 0x00040000 (mobile, 3 connectors, no fbmem)
— 0x00050000 (desktop, 0 connectors, no fbmem)

 

For the above this is from the Acidanthera GitHub.  This configuration goes to config.plist

or I got lucky one day and used GPU thingy in Windows and created a pcidevices.plist, pcidevices.dsl> made into SSDT-PCI.aml.  in this lists the PCIRoots where you can edit and add what is necessary to convert to MacOS. So in the PCI devices list the .dsl one, I put 

AAPL,snb-platform-id to 000002000 and created a .aml file.  so this loads via boot.

Plus this plist has the all the root PCI's so it gets loaded.  looking at the boot log of clover states success.  this then eliminates DSDT edits I would think. Haven't tried this yet.

but I am loading the SSDT-PCI.aml. it is loading. an override at the least I would think.  Logic is not sure here.

 

in this Readme file from GitHub, states, empty buffer of 0x0050000 if not inserted. But, whatever green.kext loads according to what it has stored concerning SNB buffers.  This can be changed in config.plist at Devices>Properties down on the bottom of Clover Config.

If using some other won't bring up these extras and you have to manually do this.

So using Clover Configurator use it to populate, then save and exit then use your choice of configurator.  then edit as needed.

 

Notice this is for Sandybridge.  They recommend the first one. Obviously.

But your research comes to use the second one. 1 connector. 

 

there is a third for mobile. 3 connector.  Now on this laptop there are two visible ports on the side.  DVI & HDMI outs. then the VGA or LCDD > Builtin Display.

I'm wondering if I change to the 00000400 which states 3 connectors, this would solve the  HDMI issue and DVI while keeping the artifacts to 0 and have a Happy HDMI.

 

I've had to switch to AppleALC.kext, Lilu.kext, VirtualSMC.kext, Whatevergreen.kext to eliminate much of the patching in config.plist.  Kind of redoing all this patching to be a bit more less hectic and according to Acidanthera, a bit more like 'real Macs' is the theme.

Seems to work so far.  The only problem is this is an ancient Asus laptop running a weird BIOS.  has UEFI but it don't work. some sort of hype-bred.

 

but it works.  the DSDT is so unordinary that it is hard to make out the mess. no wonder windows runs slow.  machine to OS handoff is gibberish.

 

Framebuffering is convenient but is a problem.  Too much flex.  RAM DRIVE, Virtual memory with flexibility works better.  A pool with options. 

 

[makk]

Also when using Clover NVRAM.plist, have to delete so the new config takes places.

 

Otherwise have the mismatch and all kinds of errors.

 

 

[makk]

INVOKE: pmset -g rtc

Error: Read RTC power bytes error = -536870207

RTC bytes unavaiable/invalid: Error 0xe00002bc.

raw bytes: 0x00000000 0x00000000

PM Byte = On (0x00)

LW Byte = OK (0x00)

 

this is invoking in Terminal:  peseta -g everything

 

RTC no talking here.

 

USB                         2          2  USEABLE

ELECOM                      2          2  USEABLE

AppleLPC                    1          1  USEABLE

AppleIntelPchSeriesAHCI     2          2  USEABLE

IOAHCISerialATAPI           1          1  USEABLE

AppleSMBusControllerICH     1          1  USEABLE

AsusNBFnKeys                1          1  USEABLE

AppleRTC                    2          2  None.  <<<<. RTC 

AppleSMC                    1          1  USEABLE

[oldman20]

Still no solution for HD3000 now?

[makk]

On 12/24/2022 at 7:26 AM, oldman20 said:

Still no solution for HD3000 now?

for Sierra and El Captain worked out better after switching to OpenCore older version of OpenCore such as 0.5.x. for quite a few years until the laptop died of overdose of movies.

I don't own this system any longer so, all I can say is I switched from Clover to OpenCore. This seemed to have helped.

 

With Clover 3xxx

It ran pretty normal without the bugs using the older ones 3xxx. 4xxx no go.
I don't have the config.plist and the files any longer.  

did have to use plenty of boot arguments. DSDT.aml. SSDT PreGen for power management

all kinds of patches on the laptop to where the boot glitch stopped, and could run movies for weeks on end without a crash or anomaly. 

 

Device | Properties>Properties and added all the PCI Roots.

in the PCI for the Graphics had all necessaries.  

 

slide=0

dart=0 or 1

boot glitch patch

plus more..

 

It really depends on the computer: Desktop or Laptop. 

Desktop: I think chris1111 had it working but did not get full details.  He may have retired it.

 

Laptops depends entirely on the Maker and BIOS. 

There was something I had done that fooled the Apple system. I don't recall what it was at this point.

If the computer uses the Apple kext, this could cause a problem, what do I mean? 

Need a sort of like a buffer where the kext is working and being used, but the hardware is not taking damage.

Remember this is not a true Apple made to be hardware. There are certain parameters, algorithms written or code I should say

into each Apple designated hardware like fingerprints. HD3000 for Apple specifically. And HD3000 for Windows specifically.
They both had different base structure.

 

HD3000 for Windows based they are both the same in basic construct, but the differences are the way each OS works with

the hardware. Apple is UNIX Based.  Windows is a Hybrid.

Instructions and code is different though they produce an end result.

Like telling a cow to chill so one can get milk. How each farmer comes to getting milk is not the same.

One uses machines, the other uses his hands.

 

The one that uses machines uses artificial construct, not his own hands.  <-- kind of like a buffer.

 

I don't recall exactly what I had to do so the Apple OS did not fully immerse into the hardware. 

At the core is where the problem occurs and the anomalies such as boot glitches are symptoms.

Compromised state that leads to a dead graphics in the end.

 

HD3000 as I recall due to being Sandybridge required the SSDT PreGen to produce Power management SSDT's.

This in itself produces ills later on as time with the usage. Constant updating does not help either in this case.

Have to do it once and leave it be. Then find the buffer zone to where the graphics works.

 

the Patch statements the PCI Root Device Properties is needed.

SSDT instead of DSDT.

 

 

 

Edited December 25, 2022 by makk