Jump to content

Editing custom personalities for ATI Radeon HD[45]xx

bcc9

By bcc9,
in ATi

 Share
775 posts in this topic

Recommended Posts

382jsl881
382jsl881
Posted
Posted

On my Sapphire 4670 (Shrike framebuffer), I have tried changing from

0000000	02  00  00  00  40  00  00  00  09  00  00  00  02  01  00  03
0000010	02  00  00  00  00  01  00  00  09  01  00  00  20  01  02  02
0000020	00  04  00  00  04  07  00  00  00  01  00  00  10  00  01  01
0000030	00  00  00  00  00  00  00  00  00  00  00  00  00  00  00  00

to

0000000	02  00  00  00  40  00  00  00  09  00  00  00  02  01  00  03
0000010	08  00  00  00  00  01  00  00  09  01  00  00  20  01  02  05 (HDMI)
0000020	00  04  00  00  04  07  00  00  00  01  00  00  10  00  01  01 (DVI)
0000030	00  00  00  00  00  00  00  00  00  00  00  00  00  00  00  00

to enable HDMI on Port 1, but but i can't get hdmi output (dvi works). http://pastebin.com/1tUc8SHj

 

What did i make wrong?

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
byte senseid; /* Sense line is bits 0-3

Use hw i2c flag is bit 4 */

/* i2cid = (senseid & 0xf-1)+0x90 */

/* senseid = (i2cid & 0xf) +1*/[/code]

For instance, in earlier post you told me to use senseid 01

But if I want to set hw i2c flag, should I not use in byte stream

xx xx xx 11 ?

In the static ConnectorInfo tables, I have not seen one that has this "use hw i2c" flag bit set. Have you?

I would first look for similar hardware to yours in a genuine mac and see what bits it is setting.

I actually haven't seen this bit get used other than to pretty print that it is called the "use h2 i2c" flag bit.

But I have not done an exhaustive search.

I kind of stopped doing exhaustive searches of the code once I found what was sufficient for the typical desktop case. That being setting the ConnectorType and senseid values properly, and leaving the other bytes as is (but taken from a genuine entry whose connectortype matches what you're trying to get working).

And can you add any more info in the 1st 3 bytes?
I haven't had time to look further, you could try reading the ATISupport code to see the usages, following thru to ATIFramebuffer if necessary (that's what I did).

 

Thanks again for bcc9 !

You have saved my time and money !

Can you provide more guide for Nvidia GT-260M card which is internal display for Asus G51J/60J laptop.

Updated to 10.6.2 ~ 10.6.6 its HDMI port became unable to connect, only VGA is available for external monitor.

I think your info for 9400M can be applied to it.

High five!

Sounds like the NVCAP ioregistry entry isn't set up right, but I'm not sure. Is there some other thread discussing that?

For my nvidia 9400, I'm having a lot of trouble figuring out how to control av-signal-type properly. I get av-signal-type=2 for hdmi and 16 for displayport instead of 8 for hdmi like with genuine mac hardware (mac mini).

If someone would solve that for me I'd have more time to look at ConnectorInfo :)

 

to
0000000	02  00  00  00  40  00  00  00  09  00  00  00  02  01  00  03
0000010	08  00  00  00  00  01  00  00  09  01  00  00  20  01  02  05 (HDMI)
0000020	00  04  00  00  04  07  00  00  00  01  00  00  10  00  01  01 (DVI)
0000030	00  00  00  00  00  00  00  00  00  00  00  00  00  00  00  00

to enable HDMI on Port 1, but but i can't get hdmi output (dvi works). http://pastebin.com/1tUc8SHj

 

What did i make wrong?

Well for starters you got the byte order wrong for hdmi.

 

Well for starters you got the byte order wrong for hdmi.
Secondly you're modifying an existing LVDS entry instead of working with an HDMI entry, so some of the other bytes are wrong such as the features byte (you have the use_internal flag set for example which must be wrong for hdmi).
Link to comment
Share on other sites
jsl
jsl
Posted
Posted
On my Sapphire 4670 (Shrike framebuffer), I have tried changing from
0000000	02  00  00  00  40  00  00  00  09  00  00  00  02  01  00  03
0000010	02  00  00  00  00  01  00  00  09  01  00  00  20  01  02  02
0000020	00  04  00  00  04  07  00  00  00  01  00  00  10  00  01  01
0000030	00  00  00  00  00  00  00  00  00  00  00  00  00  00  00  00

to

0000000	02  00  00  00  40  00  00  00  09  00  00  00  02  01  00  03
0000010	08  00  00  00  00  01  00  00  09  01  00  00  20  01  02  05 (HDMI)
0000020	00  04  00  00  04  07  00  00  00  01  00  00  10  00  01  01 (DVI)
0000030	00  00  00  00  00  00  00  00  00  00  00  00  00  00  00  00

to enable HDMI on Port 1, but but i can't get hdmi output (dvi works). http://pastebin.com/1tUc8SHj

 

What did i make wrong?

Please change from this :

0000010 08 00 00 00 00 01 00 00 09 01 00 00 20 01 02 05 (HDMI)

To:

0000010 00 08 00 00 00 01 00 00 09 01 00 00 20 01 02 05 (HDMI)

And try again !

 

If it does not work, try to copy from other framebuffer in which starting with 00 08 xx xx .....

My experience showed HDMI audio need both of the following:

(1) starting at 00 08

(20 end with correct senseID for your HDMI (using radeondump to get it)

Link to comment
Share on other sites
jsl
jsl
Posted
Posted
High five!

Sounds like the NVCAP ioregistry entry isn't set up right, but I'm not sure. Is there some other thread discussing that?

For my nvidia 9400, I'm having a lot of trouble figuring out how to control av-signal-type properly. I get av-signal-type=2 for hdmi and 16 for displayport instead of 8 for hdmi like with genuine mac hardware (mac mini).

If someone would solve that for me I'd have more time to look at ConnectorInfo ;)

Thanks for your prompt reply.

After patched Connector-type and working HDMI audio I met just one minor problem again.

Before this I can easily Sleep and Wake my Hackintosh by mouse or keyboard, but now I can Sleep & Unable to Wake by keyboard or mouse any more.

Have you the same problem after patching ?

 

Solved only for MSI Hawk 5770 in Asus P6T SE MB (arch=x86_64) by myself now.

Need Sleep at DVI dispaly only without mirror in Vervet, Uakari, Eulemur FB, then I can Wake by keyboard or mouse now.

One exception for Baboon FB: need Sleep at HDMI only display, then I can Wake by KB or Mouse.

 

Still suffering the same Sleep/Wake Problems for Asus EAH 5770 in Asus M3A MB (arch=i386) & Asus EAH 5850 in Asus P5Q PRO (arch=x86_64)

 

Very tricky !

Another {censored} phenomenon:

Different ATI 5xxx cards have their unique behavor.

For MSI Hawk 5770 in Asus P6T SE MB (arch=x86_64):

Even I can activate Zonalis FB in DVI and HDMI (including its audio) dispaly, they behave normally only by mirror.

Without mirror both display shows only background without Apple Menu for selection at all.

And its Baboon FB sometimes behave {censored} too.

For Asus EAH 5770 in Asus M3A MB (arch=i386)

Both Baboon and Zonalis seem much better without such problems because it has VGA port instead of DP port.

 

So my conclusion:

Each ATI 5xxx card need individual modification and test by your specific OSX and MB.

You need refer other's successful ATIFramebuffer and re-adjust to your specific evironment for trial & error.

Otherwise you may waste too much time, energy, and even money !

bcc9's guide is the best up to now.

Link to comment
Share on other sites
wmarsh
wmarsh
Posted
Posted
In the static ConnectorInfo tables, I have not seen one that has this "use hw i2c" flag bit set. Have you?

I would first look for similar hardware to yours in a genuine mac and see what bits it is setting.

Well the whole problem for Mobility Radeons is there is not similiar genuine macs, at least until these new ones came out. We could give up -- dongs FB works great -- but with no QE usability is limited.

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
Well the whole problem for Mobility Radeons is there is not similiar genuine macs, at least until these new ones came out. We could give up -- dongs FB works great -- but with no QE usability is limited.

You could give up if you don't want to do any homework on this... Look at the personality used by the genuine macs, look at their ioregistry, their bios dump. This should tell you what values can work in practice. Then work backwards from there.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
@ kizwan

 

Did you add Device(GFX0) & Device (HDAU) into your DSDT?

 

           Device (PEGP)  //adding this device
           {
               Name (_ADR, 0x00010000)   //using lspci tool for your graphic address
               Name (_PRW, Package (0x02)  //based on Gigabyte DSDT
               {
                   0x09, 
                   0x05
               })
               Device (GFX0)
               {
                   Name (_ADR, Zero)
                   Method (_DSM, 4, NotSerialized)
                   {
                       Store (Package (0x02)
                           {
                               "hda-gfx", 
                               Buffer (0x0A)
                               {
                                   "onboard-1"
                               }
                           }, Local0)
                       DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
                       Return (Local0)
                   }
               }

               Device (HDAU)
               {
                   Name (_ADR, One)
                   Method (_DSM, 4, NotSerialized)
                   {
                        Store (Package (0x02)
                            {
                                "hda-gfx", 
                                Buffer (0x0A)
                                {
                                    "onboard-1"
                                }
                            }, Local0)
                       DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
                       Return (Local0)
                   }
               }
           }

Hi tmongkol,

 

I already inject HDAU device in DSDT. System Profiler now reported HDMI audio but there is no HDMI audio in audio/sound preferences. I checked & found the device ID 0x1002AA01 already exist in AppleHDA binary. What did I miss?

Link to comment
Share on other sites
VCH888
VCH888
Posted
Posted
Hi tmongkol,

 

I already inject HDAU device in DSDT. System Profiler now reported HDMI audio but there is no HDMI audio in audio/sound preferences. I checked & found the device ID 0x1002AA01 already exist in AppleHDA binary. What did I miss?

post-93383-1299723792_thumb.jpg

Do you have this in IORegistryExplorer? Did you connect HDMI cable? Which framebuffer? HDMI sense-id?

Sorry for asking many questions.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
post-93383-1299723792_thumb.jpg

Do you have this in IORegistryExplorer? Did you connect HDMI cable? Which framebuffer? HDMI sense-id?

Sorry for asking many questions.

Yes, I have that in my IORegistryExplorer. Yes I did connect HDMI cable. I have "High Speed HDMI Cable with Ethernet" if that matters. :D Currently I'm using Vervet framebuffer. HDMI sense ID is 0x02. I'm using 10.7's kexts though.

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
Yes, I have that in my IORegistryExplorer. Yes I did connect HDMI cable. I have "High Speed HDMI Cable with Ethernet" if that matters. :) Currently I'm using Vervet framebuffer. HDMI sense ID is 0x02. I'm using 10.7's kexts though.

connector-id in your ioregistry for your hdmi connector=2048?

Link to comment
Share on other sites
VCH888
VCH888
Posted
Posted
No, there is no connector-id found in my ioregistry. I looked everywhere.

 

I think bcc9 means connector-type. I may be wrong about this. See the below picture from my HD5770;

post-93383-1299767222_thumb.jpg

 

Hex 0x800 = Dec 2048

 

Did you get video output thru HDMI? If yes, can you post a picture of av-signal & connector-type.

 

Did you try to change senseid of Vervet?

 

$ od -Ax -tx1 /tmp/Vervet

0000000 00 04 00 00 00 04 00 00 00 01 00 00 12 04 04 02 --> 01

0000010 04 00 00 00 14 00 00 00 00 01 00 00 01 12 01 04

0000020 00 02 00 00 14 00 00 00 00 01 00 00 00 00 06 03

0000030 00 08 00 00 00 02 00 00 00 01 00 00 22 05 05 01 --> 02

 

if not working for you, swap rows or copy the fourth row to the first row.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
Did you get video output thru HDMI? If yes, can you post a picture of av-signal & connector-type.

Yes, I get video output through HDMI.

post-125711-1299771154_thumb.jpg

Did you try to change senseid of Vervet?

 

$ od -Ax -tx1 /tmp/Vervet

0000000 00 04 00 00 00 04 00 00 00 01 00 00 12 04 04 02 --> 01

0000010 04 00 00 00 14 00 00 00 00 01 00 00 01 12 01 04

0000020 00 02 00 00 14 00 00 00 00 01 00 00 00 00 06 03

0000030 00 08 00 00 00 02 00 00 00 01 00 00 22 05 05 01 --> 02

 

if not working for you, swap rows or copy the fourth row to the first row.

I'm using 10.7's kext. The 10.7's ATIFramebuffer is different than 10.6's Framebuffer. So I'm unable to test the senseid.

 

With 10.6's kexts & ATIFramebuffer patched, no output on external monitor (connected to HDMI). The display just sleep. However, monitor does detected in IORegistry with correct EDID & correct connector-type (0x800). When checked via VNC, I found QE/CI are enabled.

 

These are screenshots of the HDAU device:-

post-125711-1299771671_thumb.jpgpost-125711-1299771695_thumb.jpgpost-125711-1299771715_thumb.jpg

Link to comment
Share on other sites
VCH888
VCH888
Posted
Posted
Yes, I get video output through HDMI.

post-125711-1299771154_thumb.jpg

 

I'm using 10.7's kext. The 10.7's ATIFramebuffer is different than 10.6's Framebuffer. So I'm unable to test the senseid.

 

With 10.6's kexts & ATIFramebuffer patched, no output on external monitor (connected to HDMI). The display just sleep. However, monitor does detected in IORegistry with correct EDID & correct connector-type (0x800). When checked via VNC, I found QE/CI are enabled.

 

These are screenshots of the HDAU device:-

post-125711-1299771671_thumb.jpgpost-125711-1299771695_thumb.jpgpost-125711-1299771715_thumb.jpg

 

You should remove IOKitPersonalities of BuiltInHDA from LegacyHDA.kext as I saw the middle picture. my ati & nvidia;

post-93383-1299773979_thumb.jpg post-93383-1299773998_thumb.jpg

 

I would like to tell you from my experience. I updated OS X to 10.6.6 and tried to used old AppleHDA.kext (10.6.4) and then I could not get HDMI audio for my HD5770. After that I got to use AppleHDA.kext v1.9.9f12 and then I got HDMI audio again. I was thinking that ATIframebuffer might be related to AppleHDA.kext in some way but I had no idea how. In contrast, I got to use old AppleHDA.kext (from 10.6.2-.4) for my GT240 to get HDMI audio after updating to 10.6.5-.6. I hope bbc9 can find out something.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
You should remove IOKitPersonalities of BuiltInHDA from LegacyHDA.kext as I saw the middle picture.

I'll remove it. Thanks.

I would like to tell you from my experience. I updated OS X to 10.6.6 and tried to used old AppleHDA.kext (10.6.4) and then I could not get HDMI audio for my HD5770. After that I got to use AppleHDA.kext v1.9.9f12 and then I got HDMI audio again. I was thinking that ATIframebuffer might be related to AppleHDA.kext in some way but I had no idea how. In contrast, I got to use old AppleHDA.kext (from 10.6.2-.4) for my GT240 to get HDMI audio after updating to 10.6.5-.6. I hope bbc9 can find out something.

I'll grab an older AppleHDA.kext from 10.6.2-4 & test it.

 

I will post here later after I have done some more test on this. Thank you very much.

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
I think bcc9 means connector-type. I may be wrong about this. See the below picture from my HD5770;
Yes, sorry that's what I meant. I see that he does have that value matching up just fine (as well as av-signal-type).

kizwan,

We've seen the hda-gfx under the HDAU node, but is the other matching one injecting correctly under GFX0?

If you use 

ioreg -l -t -w0 > file

and send in the HDAU and GFX0 subtrees from that file then there would be no need for 20 questions about your ioregistry & screen captures :thumbsup_anim:

 

Looks to me like Iago and Hypoprion do
Ah yes, I had looked at the wrong bytes when I first checked Hypoprion.

Good to see something uses that bit.

 

Have you seen this thread with working LVDS for an ati mobility system?

http://www.insanelymac.com/forum/index.php...=250405&hl=

Bit 4 is 0 in this case.

 

So which personality is the closest match for your hardware? (Ie which generation of ATI hardware does apple run these personalities on)?

 

I would like to tell you from my experience. I updated OS X to 10.6.6 and tried to used old AppleHDA.kext (10.6.4) and then I could not get HDMI audio for my HD5770. After that I got to use AppleHDA.kext v1.9.9f12 and then I got HDMI audio again. I was thinking that ATIframebuffer might be related to AppleHDA.kext in some way but I had no idea how. In contrast, I got to use old AppleHDA.kext (from 10.6.2-.4) for my GT240 to get HDMI audio after updating to 10.6.5-.6. I hope bbc9 can find out something.

Newer applehdacontroller has specific PCI-id checks that affect nvidia (it checks for 10de:0ac0 and 10de:0d94 in framebuffererEventHandler()). Older code has no such checks.

Since the gt240 is 10de:0be4, I suspect this check is breaking your support. I think the result is that it won't find the right codec-id # for your HDMI audio.

With 10.6.6 applehda, you could try either injecting audio-codec-index into the IGPU device, for example the following if your hdmi codec is #3:i

"@1,audio-codec-index",
 Buffer (0x04)
  {
  0x03, 0x00, 0x00, 0x00
 },

or edit applehdacontroller changing one of those pci id matches to match your hardware.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
Yes, sorry that's what I meant. I see that he does have that value matching up just fine (as well as av-signal-type).

kizwan,

We've seen the hda-gfx under the HDAU node, but is the other matching one injecting correctly under GFX0?

If you use 

ioreg -l -t -w0 > file

and send in the HDAU and GFX0 subtrees from that file then there would be no need for 20 questions about your ioregistry & screen captures :)

There is a matching set ("hda-gfx=onboard=1") under GFX0 (or to be precised M86). This is my ioreg dump:-

kizwan___s_1557.ioreg.zip

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
There is a matching set ("hda-gfx=onboard=1") under GFX0 (or to be precised M86). This is my ioreg dump:-

The hda-gfx set looks fine. I don't see anything obviously wrong from that dump. You are in a bit of unchartered ground with 10.7 video on a 10.6 system. (10.6 AppleHDA I presume).

Link to comment
Share on other sites
VCH888
VCH888
Posted
Posted
Newer applehdacontroller has specific PCI-id checks that affect nvidia (it checks for 10de:0ac0 and 10de:0d94 in framebuffererEventHandler()). Older code has no such checks.

Since the gt240 is 10de:0be4, I suspect this check is breaking your support. I think the result is that it won't find the right codec-id # for your HDMI audio.

With 10.6.6 applehda, you could try either injecting audio-codec-index into the IGPU device, for example the following if your hdmi codec is #3:i

"@1,audio-codec-index",
 Buffer (0x04)
  {
  0x03, 0x00, 0x00, 0x00
 },

or edit applehdacontroller changing one of those pci id matches to match your hardware.

 

Thank you. I will try this method.

 

Edit:

still couldn't get it. will try this weekend again.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
I would like to tell you from my experience. I updated OS X to 10.6.6 and tried to used old AppleHDA.kext (10.6.4) and then I could not get HDMI audio for my HD5770. After that I got to use AppleHDA.kext v1.9.9f12 and then I got HDMI audio again. I was thinking that ATIframebuffer might be related to AppleHDA.kext in some way but I had no idea how. In contrast, I got to use old AppleHDA.kext (from 10.6.2-.4) for my GT240 to get HDMI audio after updating to 10.6.5-.6. I hope bbc9 can find out something.

I tried AppleHDA.kext from 10.6.2 - 10.6.6, still failed to get HDMI audio to work. I always get:-

Sound assertion "0 == hdaGfxCandidate" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDAController/AppleHDAController.cpp" at line 889 goto Exit

I think you probably right on the relation between ATIFramebuffer & AppleHDA. I already tried 10.7 AppleHDA.kext. It just failed to load. I guess I have to wait 10.7 (official) to get HDMI audio working.

Link to comment
Share on other sites
bcc9
bcc9
Posted
  • Author
Posted
I tried AppleHDA.kext from 10.6.2 - 10.6.6, still failed to get HDMI audio to work. I always get:-

Sound assertion "0 == hdaGfxCandidate" failed in "/SourceCache/AppleHDA/AppleHDA-179.1.4/AppleHDAController/AppleHDAController.cpp" at line 889 goto Exit

That means the hda-gfx pair are not matching up between the audio&video drivers... You souldn't be getting that assertion failure if you have things configured right.

Link to comment
Share on other sites
kizwan
kizwan
Posted
Posted
That means the hda-gfx pair are not matching up between the audio&video drivers... You souldn't be getting that assertion failure if you have things configured right.

hda-gfx does matched up between audio & video. You did checked my ioreg earlier.

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