Sharing my Experience on ATI Radeon HD 6470M (Asus A42JZ Notebook)

[k540kkk]

are we getting close to get lvds working on mobility 6470?

Stuck with LVDS puzzle. Find VGA or HDMI monitor to use for awhile.

We maybe need to wait until 6470M is more popular and hope that there will be some experts to solve this problem.

[pnegry]

@k540kkk: good effort... you lasted longer than me

 

just wondering if you took a look at a dsdt dump from new macbooks? it would seem to me to copy and change the connector, height and res in there would be the right approach.

[dickhouse]

guys, is there any progress regarding lvds puzzle?

[pnegry]

guys, is there any progress regarding lvds puzzle?

 

None by me.

[dertachyon]

Hi k540kk,

 

I have a Toshiba laptop with an ATI 6450m, quite similar to your graphics card. I'm still stuck in 1024x768 with no QE/CI. The radeondump of my bios is similar to yours, and I think your method should work on my laptop. I attach the connector info below. I tried Chimera and Chameleon, but the only way to reach the system is using GraphicsEnabler=No. If I use GraphicsEnabler=Yes, the loading stuck in the line ATI card posted.

 

Currently, I'm running 10.6.8. What framebuffer would you suggest for the ATI6000 kexts?

 

Thanks in advance,

Tachyon

 

ATOM BIOS Rom:

SubsystemVendorID: 0x1179 SubsystemID: 0x0001

IOBaseAddress: 0x3000

Filename: AALT1SM.400

BIOS Bootup Message:

TOSHIBA Video BIOS v4.00

 

PCI ID: 1002:6760

Connector at index 0

Type [@offset 45010]: LVDS (7)

Encoder [@offset 45014]: INTERNAL_UNIPHY (0x1e)

Connector at index 1

Type [@offset 45020]: DisplayPort (10)

Encoder [@offset 45024]: INTERNAL_UNIPHY1 (0x20)

i2cid [@offset 45121]: 0x90, OSX senseid: 0x1

Connector at index 2

Type [@offset 45030]: HDMI-A (11)

Encoder [@offset 45034]: INTERNAL_UNIPHY1 (0x20)

i2cid [@offset 45148]: 0x91, OSX senseid: 0x2

Connector at index 3

Type [@offset 45040]: VGA (1)

Encoder [@offset 45044]: INTERNAL_KLDSCP_DAC1 (0x15)

i2cid [@offset 45175]: 0x97, OSX senseid: 0x8

[k540kkk]

GraphicsEnabler=Yes, the loading stuck in the line ATI card posted.

 

For above problem, you need to (1)checkout chamelon kalby branch and (2)modify ati.c to add your DeviceID, SubsystemID and SubsystemVendorID and (3)recompile to get boot binary file that can detect your card correctly.

 

ATOM BIOS Rom:

SubsystemVendorID: 0x1179 SubsystemID: 0x0001

IOBaseAddress: 0x3000

Filename: AALT1SM.400

BIOS Bootup Message:

TOSHIBA Video BIOS v4.00

 

PCI ID: 1002:6760 <- DeviceID

 

You can find reference and my ati.c file in first post.

Example :

line 483:	{ 0x6760,	0x1CB21043, CHIP_FAMILY_SEYMOUR, "ATI Radeon HD 6470M Graphics", kPithecia }
For DeviceID:6760 SubsystemVendorID: 0x1043 SubsystemID: 0x1cb2

 

 

Remarks : I still stuck with LVDS problem and current status as shown in first post (no new update).

[dertachyon]

Man, you are a genious. The HDMI output worked!

 

I downloaded the azimutz bootloader from chameleon website, added my device-id to the ati.c file and then compiled it. I plugged the HDMI, turned on the laptop and the graphics card was finally recognized. I attach a screen capture of my ATI 6450m on 10.6.8 fully working.

 

The internal LCD still doesn't work and I'm just getting a black screen as you. I think a I need to modify the closest framebuffers, Cattail and Fanwort in my case. Let you know any success with the internal LCD.

 

Cheers,

Tachyon

 

 

For above problem, you need to (1)checkout chamelon kalby branch and (2)modify ati.c to add your DeviceID, SubsystemID and SubsystemVendorID and (3)recompile to get boot binary file that can detect your card correctly.

 

ATOM BIOS Rom:

SubsystemVendorID: 0x1179 SubsystemID: 0x0001

IOBaseAddress: 0x3000

Filename: AALT1SM.400

BIOS Bootup Message:

TOSHIBA Video BIOS v4.00

 

PCI ID: 1002:6760 <- DeviceID

 

You can find reference and my ati.c file in first post.

Example :

line 483:	{ 0x6760,	0x1CB21043, CHIP_FAMILY_SEYMOUR, "ATI Radeon HD 6470M Graphics", kPithecia }
For DeviceID:6760 SubsystemVendorID: 0x1043 SubsystemID: 0x1cb2

 

 

Remarks : I still stuck with LVDS problem and current status as shown in first post (no new update).

[dickhouse]

can someone with heavy brain juice please i beg u to crack lvds nuts!

[dertachyon]

Hi guys,

 

Any news about a working LVDS for 6470m? I have a similar chipset, the 6450m, and only HDMI and VGA works with full QE/CI.

 

I have played with the ATI6000controller framebuffer, and it seems that ControlFlag is the key. I have used several values for ControlFlag and have obtained backlight turned on and grey background. Still no good news from my side.

 

Cheers,

Tachyon

[d@le]

If u look at this

 

APPLE LVDS default

02 00 00 00 40 00 00 00 09 01 00 00 02 01 00 03

and your's LVDS

02 00 00 00 40 00 00 00 09 01 00 00 20 01 00 07

 

divide it to 8 parts:

02 00 00 00 - 40 00 00 00 - 09 01 - 00 00 - 02 - 01 - 00 - 03

 

Then byte swapped

 

1- ConnectorType : 02 00 00 00 > 0x00000002 - - -

2- ATY,ControlFlags : 40 00 00 00 > 0x00000040 +++

3- Features : 09 01 > 0x0109 +++

4- Unkown : 00 00 > 0x0000 - - -

5- Transmitter : 02 > 0x02 - - -

6- Encoder : 01 > 0x01 - - -

7- HotplugID : 00 > 0x00 - - -

8- SenseID : 03 > 0x03 ++++++

 

- - - > Value not important, but need to be accurate for better result

+++ > Important value, incorrect value == not working personality

 

check mucha's post here for more detail's if U didn't

http://www.insanelymac.com/forum/index.php?showtopic=252061

 

lp

[son01490517]

have you try edid injection ?

[tomuxzzz]

Any process to get 6470m LVDS working ?

[ham4ever]

i have same problem with my ati 6650m no LVDS ( blackscreen or white lines) tried almost everything ,,

 

i am almost to give up

[HackMan85]

Hi,I have a asus k54ly and I have a ati radeon hd 6470M and from what I read you were able to make it work well,I do not even understand these changes and do not even know where to start, I'm looking on the Italian forum but I can not answers, can you tell mewhere to start for any changes?

[applehacker]

Trying to attempt with 6990m, same deal as everybody, blackscreen nomatter what values I use. VNC gives me a beautiful screen though, no corruption or anything at all. ATYactive points to active DVI (which I'm not using), but not active LVDS. I am very confused.

[Gorbush]

Guys, i got some news!

 

I patched the Hydrilla(LVDS) and Pithecia (LVDS + VGA) frame buffers for my HP Probook 6560b - ATI 6470M

 

And now i got the LVDM and external VGA monitor running with acceleration!

 

But with one bad thing - internal monitor seems like swapped pixel columns - dos anybody stuck with this?

 

 

ati6000Controller.tar.gz - is a patched file

 

To install patched file (Make a backup first!!!!):

 

sudo cp ATI6000Controller /System/Library/Extensions/ATI6000Controller.kext/Contents/MacOS/ATI6000Controller

sudo chown root:wheel /System/Library/Extensions/ATI6000Controller.kext/Contents/MacOS/ATI6000Controller

 

Patch sequence for LVDS:

0200 0000 4000 0000 0901 0000 2001 0001

 

Patch sequence for VGA:

1000 0000 1000 0000 0001 0000 0000 0001

 

VGA TO TEST - cause it works in Pithecia, but it has not my senceID

1000 0000 1000 0000 0001 0000 0000 0008

 

My connectors:

 

ATOM BIOS Rom:

SubsystemVendorID: 0x103c SubsystemID: 0x161a

IOBaseAddress: 0x4000

Filename: BR41780.001

BIOS Bootup Message:

HP / Foxconn Cure SEYMOUR XT DDR3 64Mx16 512MB

 

PCI ID: 1002:6760

Connector at index 0

Type [@offset 45000]: LVDS (7)

Encoder [@offset 45004]: INTERNAL_UNIPHY (0x1e)

i2cid [@offset 45110]: 0x90, OSX senseid: 0x1

Connector at index 1

Type [@offset 45010]: DisplayPort (10)

Encoder [@offset 45014]: INTERNAL_UNIPHY1 (0x20)

i2cid [@offset 45133]: 0x93, OSX senseid: 0x4

Connector at index 2

Type [@offset 45020]: DisplayPort (10)

Encoder [@offset 45024]: INTERNAL_UNIPHY1 (0x20)

i2cid [@offset 45160]: 0x91, OSX senseid: 0x2

Connector at index 3

Type [@offset 45030]: DisplayPort (10)

Encoder [@offset 45034]: INTERNAL_UNIPHY2 (0x21)

i2cid [@offset 45187]: 0x92, OSX senseid: 0x3

Connector at index 4

Type [@offset 45040]: VGA (1)

Encoder [@offset 45044]: INTERNAL_KLDSCP_DAC1 (0x15)

i2cid [@offset 45214]: 0x97, OSX senseid: 0x8

 

 

 

 

Detected:

ati6000Controller.tar.gz

[Gorbush]

Really interesting behaviour:

 

It changes from boot - to - boot : every time there is a different distance between swapped columns - some times it is 1 pixel, sometimes it is three pixels, and once -

ITS NORMAL .... but it's seldom event...

 

Now i got it work well - and in fear about next reboot....

 

By the way - if connected to the doc station - it disables notebook VGA output - and moves it to the docking VGA... loosing the acceleration...

[Kvashir]

Hi, great work!

 

I've the same card but with 1GB Ram + 6620G 512MB card (AMD LLANO) Do you think could it works?

[Gorbush]

Who knowns...

Now i'm trying to get it work stable - replacing some bits.... but without any success..

At least i can log in and use another desktop on secondary monitor with acceleration... :-)

[HackMan85]

hello guys, these days I installed on my asus Mountain lion retail x54hy with ati 6470m and in the path in S / L / And we are dedicated to the series new kext hd 6000 and also to the framebuffer, but also I can not see the internal display how it should work, plugging the vga or the hdmi and enabling GraphicsEnabler external monitors work fine but the inside is black, but if i disable the boot graphicsEnabler recognize my video card but the same as "AGPM controller", since I am not an expert on kext I can not change how I should but I am sure that with these new kext someone with some experience will be able to resolve this situation, I am not an expert in January we lose hours, not a programmer but I have made ​​several attempts with the file of my monitor with my inject edid but I think that both the internal display.

[HackMan85]

Hello, I have an idea, but since i have a mac that activates gmux swith and video cards such as ours are secondary because in case of intensive use are activated is not that they must find a way to switch it? I without any modifications whatsoever to any active file GraphicsEnabler and connect the monitor turns off completely and if you deadlift, you should not work on this? I gave my id 0x67601002 20011043 subsystemid Ermac and added to my data and Enoch did not work, I think that is the problem.

[Jacksoft]

Hi all! Same issue here with an HD5650M on a Vaio VPCEB4C5E, but...

 

In these days I've tryied many and many personality to try to fix my LVDS that not show. I've fixed the VGA and the HDMI, but sometimes if I modify some LVDS bits VGA does not work anymore...

 

Well, usually I connect througt the HDMI a Sansui TV. With this, when work, I can repatch the AtiController5000.kext and retry again...

Today I had to move my notebook in another place, where I have an LG M2362D. Connected to HDMI, powered on and... TADAAAAANN!!! Internal LVDS works!!!

Rebooted two-three times and it works! But HDMI not.

In monitor settings OS X recognize the LG, but on it there's "No Signal".

 

Isn't that bad, but I can't trasport with me the LG everytime

 

Here's the binpatch!

 

Original Eulemur Framebuffer:

04 00 00 00 14 00 00 00 00 01 00 00 01 02 01 04

00 08 00 00 00 02 00 00 00 71 00 00 12 04 04 02

10 00 00 00 10 00 00 00 00 00 00 00 00 10 00 01

 

Patched:

02 00 00 00 40 00 00 00 90 01 00 00 10 00 00 07

00 08 00 00 04 02 00 00 00 01 00 00 11 02 04 05

10 00 00 00 10 00 00 00 00 01 00 00 00 10 01 01

 

 

More info:

- I'm using Chameleon by Enoch rev.2042.

- By decoding the VBIOS my LVDS does not have a SenseID.

- I dumped the EDID information of the LVDS with a tool on Windows 7 and injected into the right DisplayProduct, but didn't helped.

- Firstly, with the Sansui TV, in IORegistryExplorer, The Framebuffer0 was of the LVDS, Framebuffer1 of the HDMI and Framebuffer2 of the VGA. Now, with the LG connected on the HDMI, Framebuffer0 is the HDMI, Framebuffer1 is LVDS (but still not recognized???) and Framebuffer3 is the VGA.

- If I unplug the HDMI the LVDS goes black and won't turn back if I replug, but the screen appear "blocky" on the LG. I have only to reboot.

 

EDIT:

- By swapping the LVCD personality with the HDMI I have the same result, but by unplugging and replugging the HDMI I have this affects:

Good screen -> Screen with some vertical lines -> Screen seem a 3D TV (image doubled) -> Black Screen -> Screen seem a 3D TV (image doubled) -> Screen with some vertical lines -> Good Screen

I'm making other tests... Uff!!!

- With moninfo, on Windows 7, I've found somethin strange. Take a look at this screenshot: http://i45.tinypic.com/2hxsltv.png

The GSM5758 is the LG M2362D connected though the HDMI, but is OFF! He read it has active, and the LVDS (MS_0025) not, but works because I'm writing from it.

I think the VBIOS was written from some monkeys... I have to try to use a VBIOS from an equivalent card, using AtiROM flag active... °_°

 

EDIT2: Hot news! On mine machine (I think also for others) the EDID is stored in the VBIOS!!!

I've dumped the VBIOS with Linux, and have some bytes different from the dump made with Everest/Aida64. So I've found a tool that can decompile totally the Ati/ATOM VBIOS (AtomDis), and by analizyng the results I've found a "LVDS_Info", and tadaaan! In that pack of bytes there's the EDID found thougt Windows! But is much longer and I'll try to test with this!

Yeah... One day lost to dump the EDID from the monitor EEPROM's, and I've found that is in the VBIOS... -__-

 

Check the screenshot: http://i50.tinypic.com/auljm.png

 

(Sorry for bad english... )

[Tauio111]

Hello,

I'm using an ASUS K54LY laptop (i3 2330m, 4gb ddr3, 6470m) and somehow managed to get my 6470m to run with full acceleration, full resolution(1440x900@75hz) and QE/CI, but the problem is that it only outputs onto the VGA screen, the LVDS is dead.

I have no way of extracting the ATI rom, all the tools fail at this.

[DavidRocha]

How you managed to get the ATOM BIOS Rom? I have the same board, but my notebook is Dell.

[macsane123]

hey all

i found the if ati 6 series and the lvds senseid is 7,we got a black screen,is it right?

and any good news for this problem?