Jump to content

ATILead - natit-like injector for Radeon.


  • Please log in to reply
106 replies to this topic

#81
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow

Slice, can you post a version of Callisto wich work for dual screen with ATI Lead

No such version exists

and one version of callisto which work with ATINDVR.kext?

Callisto is not work with ATINDRV. It is replacement.
You probable means ATILead?
Take info.plist from Lebedou (previous posts)

thanks, because i don't under stand very well the modify to info.plist!

You must have more then enough experience to modify info.plist

#82
Lostgame

Lostgame

    Lostgame

  • Members
  • PipPipPipPipPipPipPip
  • 591 posts
  • Interests:Programming, game design, OSX, music design, photo editing, movie making.<br /><br /...
Would this be an option to get CI and QE enabled on my x300 mobility in leopard?

#83
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow

Would this be an option to get CI and QE enabled on my x300 mobility in leopard?

Yes, but
1. You must remember that ATIRadeon9700GLDriver.bundle in Leo is PowerPC only. Needs to replaced.
2. May be you need to do binary patch for your devID in ATIRadeon9700.kext. (for example koverg patch)
3. Remember about IOProvider in GA.plugin.info.plist

#84
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

Slice, can you post a version of Callisto wich work for dual screen with ATI Lead and one version of callisto which work with ATINDVR.kext?
thanks, because i don't under stand very well the modify to info.plist!


with such achievements done by Slice and confirmed by lebidou, I'd like to state, Callisto time is over, time has passed for new success story about framebuffer alignments.
Our new hero is Slice from this moment, now on, solutions will be referenced in wiki on his famous ATILead and laptop users with ATI cards will have their future in osx86 back again.

#85
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

Hello, I just tried ATILead with ATI Mobility X1900 and it's almost working. But the VRAM size in System Profiler is shown as 4Mo instead of 256, and I can't have 32 bits color depth with resolutions higher than 1024x640.

Moreover, 16 bits color-depth is crappy :
Posted Image
I attached the ioreg and the ATILead Info.plist file I used, in case you find some info in it.

Thanx for your help!


hello lebidou

following elements of Info.plist were found different from Slice ATILead.kext Info.plist internal one (original one, version 122 taken from 1st page of this Topic)

could you be so kind to explain/comment/describe your decision taken for each field:
- why you decided to add following field into original Info.plist (if it's new)
- what was your decision/source of information, based on which following field value was provided
thanks

new element:
<key>@0,fboffset</key>
<integer>0</integer>
new element:
<key>@0,ATY,EFIDisplay</key>
<string>LVDS</string>
different in value:
<key>@0,compatible</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>

missing element:
<key>@0,depth</key>
<integer>32</integer>

different in value:
<key>@0,display-connect-flags</key>
new set to:
<integer>1024</integer>
original was:
<integer>0</integer>

missing elements:
<key>@0,height</key>
<integer>768</integer>
<key>@0,linebytes</key>
<integer>4096</integer>

different in value:
<key>@0,name</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>

missing elements:
<key>@0,width</key>
<integer>1024</integer>

new elements:
<key>ATY,FrameBufferOffset</key>
<string>80</string>
<key>ATY,VRAM,MemSize</key>
<string>256</string>

different in value:
<key>CRT,EDID</key>
new set to:
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
original was:
<data>
AP///////wBN2QBcAQEBASgPAQMIAAB4Ctmto1VGmCQQ
R0otyABhTwEBAQEBAQEBAQEBAQEBZBkAQEEAJjAdiDYA
AAAAAAAYDh8AgFEAHjBAgDcAAAAAAAAcAAAA/QA6Ph4y
CAAKICAgICAgAAAA/ABLREwtUzMyQTEyVQogAMs=
</data>

different in value:
<key>IOName</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>

alternative new element:
<key>LVDS,EDID</key>
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
alternative old element:
<key>LCD,EDID</key>
<data>
AP///////wAGEGCcAAAAAAQQAQOAIhZ4Chz1l1hQjicn
UFQAAAABAQEBAQEBAQEBAQEBAQEBnyWgQFGEDDBAIDMA
S88QAAAZAAAAAQAGEDAAAAAAAAAAAAogAAAA/gBCMTU0
UFcwMSBWMAogAAAA/ABDb2xvciBMQ0QKICAgALw=
</data>

different in value:
<key>device_type</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>

different in value:
<key>name</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>


new element - such element not exist in original ATILead configuration
different in value - such element mismatch between ATILead configuration and yours
new set to - value set by you
original was - original value from ATILead package
alternative old/new element - new element very similar to old one, but yet not exactly same

#86
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France
here is my own following guess about which field which value require and from where it may be taken.
some of they are already aligned with Slice vision, so soon will become more useful for other testers.
let's assume, by today, all of those settings has to be added/applied to default Info.plist of ATILead.kext package.

<key>@0,fboffset</key>
<integer>0</integer>
this value probably must be set for R600/R700 chipsets according to following link (R700 should be set to 0x2024, R600 to 0x2180), but where do we taking value for R300 chipsets?
alternatively this value has to be set to 0.
yet from another source of information for ATIPC framebuffer it's populated by Slice ATI.kext driver, and described here
value to set not yet clarified

<key>@0,ATY,EFIDisplay</key>
<string>LVDS</string>
this value mean, LVDS is using for reading EFI information? can't understand
value to set not yet clarified

<key>@0,compatible</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>
this value is set to Wormy in order to use standard ATIRadeonX1000 driver, is set to ATIPC if you willing to use Slice custom framebuffer developed here
alternatively this could be your driver personal name, yet declared compatible with one of previously mentioned (as it's compatibility option)

<key>@0,depth</key>
<integer>32</integer>
this value probably mean color depth definition, 32 bit per pixel is it's default setting, however 16 bit per pixel working mode confirmed possible to operate (in investigation reason only of course) by lebidou here

<key>@0,display-connect-flags</key>
new set to:
<integer>1024</integer>
original was:
<integer>0</integer>
what is connect-flags, what is it for real, based on what decision to take a value?
value to set not yet clarified

<key>@0,height</key>
<integer>768</integer>
this value define number of display vertical lines (for VESA 1024x768, the one Apple fake display resolution it's 768, but for modern laptops it must be set at least for 800 = default 16:9 matrixes 1200x800)

<key>@0,linebytes</key>
<integer>4096</integer>
this value represent number of bytes allocated per each line of display picture and correspond to 4 * horisontal display resolution
in such case, it's 4096, it could be only in case if 1024*4 bytes/pixel = 32 bits (for VESA 1024x768, where one Apple fake display horisontal resolution is 1024, but for modern laptops it is 1200, so it must be set to 1200*4 = 4800)

<key>@0,name</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>
this value is set to Wormy in order to use standard ATIRadeonX1000 driver, is set to ATIPC if you willing to use Slice custom framebuffer developed here

<key>@0,width</key>
<integer>1024</integer>
this value define number of pixels per display line (for VESA 1024x768, the one Apple fake display resolution it's 1024, but for modern laptops it must be set at least for 1200 = default 16:9 matrixes 1200x800)

<key>ATY,FrameBufferOffset</key>
<string>80</string>
this key reduplicate same value as @0,fboffset key?
according to lebidou it must be corrected to
<integer>80</integer>
according to Slice, information for ATIPC FrameBufferOffset is populated by Slice ATI.kext driver, and described here, it's not clarified for Wormy neither value 80 has to be default, nor if it's applied automatically bo ATIRadeonX1000

<key>ATY,VRAM,MemSize</key>
<string>256</string>
this value represent amount of video memory in megabytes
according to Slice, information for ATIPC FrameBufferOffset is populated by Slice ATI.kext driver, and described here, however it may be set manually (in case of Wormy for example)

different in value:
<key>CRT,EDID</key>
new set to:
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
original was:
<data>
AP///////wBN2QBcAQEBASgPAQMIAAB4Ctmto1VGmCQQ
R0otyABhTwEBAQEBAQEBAQEBAQEBZBkAQEEAJjAdiDYA
AAAAAAAYDh8AgFEAHjBAgDcAAAAAAAAcAAAA/QA6Ph4y
CAAKICAgICAgAAAA/ABLREwtUzMyQTEyVQogAMs=
</data>
this is famous EDID value, describing your display capability, what is it and what we need it for, you may read here
unfortunately ATILead doesn't support EDID extraction yet, it will be implemented later, so try to extract yours from some alternative application.
CRT key mean, this EDID is for your external display

<key>IOName</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>
this value is set to Wormy in order to use standard ATIRadeonX1000 driver, however ATIPC is which driver name?
it should match @0,compatible key, maybe or it's free to be set to anything? just adding necessary "Parent" subword?
value to set not yet clarified

<key>LVDS,EDID</key>
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
alternative old element:
<key>LCD,EDID</key>
<data>
AP///////wAGEGCcAAAAAAQQAQOAIhZ4Chz1l1hQjicn
UFQAAAABAQEBAQEBAQEBAQEBAQEBnyWgQFGEDDBAIDMA
S88QAAAZAAAAAQAGEDAAAAAAAAAAAAogAAAA/gBCMTU0
UFcwMSBWMAogAAAA/ABDb2xvciBMQ0QKICAgALw=
</data>
this is famous EDID value, describing your display capability, what is it and what we need it for, you may read here
unfortunately ATILead doesn't support EDID extraction yet, it will be implemented later, so try to extract yours from some alternative application
LCD key mean, this EDID is for your internal display

<key>device_type</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>
this value is set to Wormy in order to use standard ATIRadeonX1000 driver, however ATIPC is which driver name?
it should match @0,compatible key, maybe or it's free to be set to anything? just adding necessary "Parent" subword?
value to set not yet clarified

<key>name</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>
this value is set to Wormy in order to use standard ATIRadeonX1000 driver, however ATIPC is which driver name?
it should match @0,compatible key, maybe or it's free to be set to anything? just adding necessary "Parent" subword?
value to set not yet clarified

#87
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow
OK. It is useful for all if I comment these changes

new element:
<key>@0,fboffset</key>
<integer>0</integer>

This key is implemented in Ati.kext
http://forum.insanel...o...6664&st=364
It give me screen shift down if non-zero.
But I don't know if it influence on ATINDRV:ATY,Wormy

new element:
<key>@0,ATY,EFIDisplay</key>
<string>LVDS</string>

May be it is a key? I dunno.

different in value:
<key>@0,compatible</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>
different in value:
<key>@0,name</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>

different in value:
<key>IOName</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>
different in value:
<key>device_type</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>

different in value:
<key>name</key>
new set to:
<string>ATY,WormyParent</string>
original was:
<string>MyRadeon</string>

To call ATY,Wormy instead of Ati.kext

different in value:
<key>@0,display-connect-flags</key>
new set to:
<integer>1024</integer>
original was:
<integer>0</integer>

Another interesting find of Lebidou. I dunno.

missing element:
<key>@0,depth</key>
<integer>32</integer>
missing elements:
<key>@0,height</key>
<integer>768</integer>
<key>@0,linebytes</key>
<integer>4096</integer>
missing elements:
<key>@0,width</key>
<integer>1024</integer>

It is custom resolution settings used in Ati.kext. Dunno about ATY,Wormy

new elements:
<key>ATY,FrameBufferOffset</key>
<string>80</string>
<key>ATY,VRAM,MemSize</key>
<string>256</string>

These key don't used by Ati.kext, but used by ATY,Wormy. Dunno how.

different in value:
<key>CRT,EDID</key>
new set to:
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
original was:
<data>
AP///////wBN2QBcAQEBASgPAQMIAAB4Ctmto1VGmCQQ
R0otyABhTwEBAQEBAQEBAQEBAQEBZBkAQEEAJjAdiDYA
AAAAAAAYDh8AgFEAHjBAgDcAAAAAAAAcAAAA/QA6Ph4y
CAAKICAgICAgAAAA/ABLREwtUzMyQTEyVQogAMs=
</data>

alternative new element:
<key>LVDS,EDID</key>
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
alternative old element:
<key>LCD,EDID</key>
<data>
AP///////wAGEGCcAAAAAAQQAQOAIhZ4Chz1l1hQjicn
UFQAAAABAQEBAQEBAQEBAQEBAQEBnyWgQFGEDDBAIDMA
S88QAAAZAAAAAQAGEDAAAAAAAAAAAAogAAAA/gBCMTU0
UFcwMSBWMAogAAAA/ABDb2xvciBMQ0QKICAgALw=
</data>

Yes! Lebidou is quite right writing his own EDID for CRT and LCD monitors. Grates!!!
I think all other testers must find own EDIDs, for example from Windows, to inject into these field.

Other questions?

#88
Eddie94

Eddie94

    iDeneb Team Member

  • Members
  • PipPipPipPipPipPipPipPip
  • 1,067 posts
  • Gender:Male
  • Location:Italy
can you (ole2) post the version of atilead with yuor fix?
thanks!I will try it!

#89
lebidou

lebidou

    InsanelyMac Geek

  • Members
  • PipPipPip
  • 153 posts
  • Gender:Male
  • Location:Fr
Hello,

Here are informations about how I managed to make it works.
First I tried in March to make it works with ATIinject, but my screen was splitted and streched in two like that. Some values I used come from that first attempt.

As I had more important things to do at that moment I gave up till now. But I was following Slice's alternative frambuffer with interest. So I first tried to make it working with his framebuffer with no succes, always blackscreen.
However, in single user mode I had good hardware detection and good EDID parsing. So I decided to see if ATILead could have success where ATIinject failed : using the ATI's framebuffer. After some reading on this forum I found some important things :


new element:
<key>@0,fboffset</key>
<integer>0</integer>
According to Omni (Callisto developer) the @0,fboffset=0 is a value that may work with most cards : link


new element:
<key>@0,ATY,EFIDisplay</key>
<string>LVDS</string>
This is a value I found on most ATIinject plist files, I grabbed over the forum. "LVDS" is a laptop specific value, to tell the framebuffer to use the LVDS port which is use for internal display in most laptops. This is the reason why I choosed the Wormy framebuffer, this is the only one with LVDS support on MacOS. It is compatible with x1600, x1700(?), x1800, and x1900 cards.


different in value:
<key>@0,compatible</key>
new set to:
<string>ATY,Wormy</string>
original was:
<string>ATI,ATIPC</string>
I choosed the Wormy framebuffer for the reason above. So all ATI,ATIPC values had to be replaced by ATY,Wormy and MyRadeon by ATY,WormyParent. For example to call the Iago framebuffer replace them by ATY,Iago and ATY,IagoParent.


missing element:
<key>@0,depth</key>
<integer>32</integer>

missing elements:
<key>@0,height</key>
<integer>768</integer>
<key>@0,linebytes</key>
<integer>4096</integer>

missing elements:
<key>@0,width</key>
<integer>1024</integer>
After the second attempt, I had a color issue (read a few posts above), that's why I removed those keys. However, they do not seem to be requiered by the ATI frambuffer.


different in value:
<key>@0,display-connect-flags</key>
new set to:
<integer>1024</integer>
original was:
<integer>0</integer>
Like for the "ATY,EFIDisplay" value, this is somthing I found on most ATIinject plists. I have no idea of what it means.


new elements:
<key>ATY,FrameBufferOffset</key>
<integer>80</integer>
I don't remember where I found this value, all I know is it has to be the half of VRAM (0x80 = 128, I have 256 Mo VRAM). In Natit plists it often come with IOSpaceOffset,I read they are useless, but as I was trying to make it work, I didn't tried only one way.


<key>ATY,VRAM,MemSize</key>
<string>256</string>
This value have to be removed, it can't have effect since it is written as a string, should be integer. Plus it has no effects, the VRAM is defined by "F8,size" actually.


different in value:
<key>CRT,EDID</key>
new set to:
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
original was:
<data>
AP///////wBN2QBcAQEBASgPAQMIAAB4Ctmto1VGmCQQ
R0otyABhTwEBAQEBAQEBAQEBAQEBZBkAQEEAJjAdiDYA
AAAAAAAYDh8AgFEAHjBAgDcAAAAAAAAcAAAA/QA6Ph4y
CAAKICAgICAgAAAA/ABLREwtUzMyQTEyVQogAMs=
</data>
This is the EDID, it has to be change because it is a display specific value, as I don't have the same display as Slice I replaced it by mine. You can get yours with the Pheonix utility on Windows. I don't know on Linux but I heard about a getedid function.
There is CRT,EDID and LVDS,EDID because I didn't know which one will be used by the framebuffer. After looking the ioreg it use the LVDS one (obviously, I'm on a laptop and the EFIDisplay is LVDS).
After having the card working I noticed there five entries under VGA@0 in the IODeviceTree: "CRT@100", "DVI@120", "DVIA@210", "LCD@110", and "ATY,Wormy@0". I didn't tried, but I bet I'd find LVDS@x instead of ATY,Wormy@0 if ATILead were not loaded.


Other important information : I'm stuck with an ATINDRV.kext from Tiger. Most of people with working ATI Mobility are. Trying to use the Leopard one give black screen. But replacing the ATY_Wormy.kext by a newer one works, but I still use ATINDRV.kext 10.4.38.

For QE/CI I use patched drivers from there. I have to patch them myself replacing 7280 by 7284 in plists and 8072 by 8472 in ATIRadeonX1000.kext binaries.

I guess I could make dual screen work setting @1,... values but I don't have an external screen to test it.

Finally, I used this working plist to make an efistring. It works but had to add some other values I found on working efistring for desktop X1900, such as "MVAD", "saved-config", "ATY,MCLK", "ATY,SCLK"...

I hope I've been clear and this is usefull. (Sorry for late answer)

PS. Sorry for redundancy

#90
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow
2 Lebidou
Thank you for your explanations. Hope it would be useful.
<key>ATY,FrameBufferOffset</key>
<integer>80</integer>
It seems to be not good as 80 != 0x80 = 128.
As I remember it is IODeviceMemory property as well as ATY,IOSpaceOffset. I am not sure that these value used by any drivers.
Only information.

Caution!
ATILead can't be full replaced by EFI strings. Trust me!
ATILead is not only injector as Natit or ATIinject, is is part of ATI driver making some internal tuning. Namely "F8,size". You can't inject this value as it used only by ATILead (or ATISlice in past).

I can rewrite ATILead so that it write ATY,MCLK ATY,SCLK into registry from VGABIOS but I don't know if it is useful.

#91
Eddie94

Eddie94

    iDeneb Team Member

  • Members
  • PipPipPipPipPipPipPipPip
  • 1,067 posts
  • Gender:Male
  • Location:Italy
slice, i don't under stand the Modifications of ole2 into the info.plist of atilead....can you post a good info.plist with ole2's Modifications?

#92
lebidou

lebidou

    InsanelyMac Geek

  • Members
  • PipPipPip
  • 153 posts
  • Gender:Male
  • Location:Fr

Caution!
ATILead can't be full replaced by EFI strings. Trust me!

I do, I'll give ATILead another try.


I can rewrite ATILead so that it write ATY,MCLK ATY,SCLK into registry from VGABIOS but I don't know if it is useful.

I don't think this is useful. It was in order to make a working efistring, but I got those values thanks to ATILead.


I attached a cleaned ATILead Info.plist file. I removed ATY,FrameBufferOffset and ATY,VRAM,MemSize. As I can't test it for dual display I removed all @1 values and CRT,EDID.

Important : This file is useful only if you want to use ATINDRV.kext (Official ATI Framebuffer). Keep the one given by Slice if your aim is to use his framebuffer.

Little "Howto modify this file" :
Values you'll have to change :
<key>@0,ATY,EFIDisplay</key>
<string>LVDS</string>
LVDS is for laptops, try with CRT, DVI, DVIA or LCD. One of those may work. It implies to change "LVDS,EDID" according to it, I mean replace "<key>LVDS,EDID</key>" by "<key>CRT,EDID</key>" or "<key>LVDS,EDID</key>" ...

<key>@0,compatible</key>
<string>ATY,Wormy</string>

<key>@0,name</key>
<string>ATY,Wormy</string>

<key>IOName</key>
<string>ATY,WormyParent</string>

<key>device_type</key>
<string>ATY,WormyParent</string>
<key>name</key>
<string>ATY,WormyParent</string>
Replace all "Wormy" by the name of the framebuffer you want to use.
AFAIK :
X1900 --> Alopias
X1300 --> Caretta
HD2600 --> Hypoprion
HD2400 --> Iago
X1600, X1800 and Mobilities --> Wormy
Browse the forum to know what is the framebuffer that fits your card.

<key>F8,size</key>
<integer>3145728</integer>
Is your VRAM in octets. Your amount of VRAM by 1024*1024. For instance 256 Mo => 256*1024*1024 = 268435456.

<key>LVDS,EDID</key>
<data>AP///////wAyDAC5AAAAAAAQAQKAIRV4Cg8Ql1hSiCgjUFQAAAABAQEBAQEBAQEBAQEBAQEB1RsAoFAgFzAwICYAI
UQAAAZAAAAAAAAAAAAAAAAAAAAAAAAAAAA/gBMR1BoaWxpcHNMQ0QKAAAA/gBMUDE1NFcwMS1UTEFKADY=</data>
As said before, your EDID, you can get it from Windows with Phoenix. I recommend to have PlistEditor to add it in the file because Phoenix will give you the EDID in hex while TextEdit prints it as characters.

After that your ready to test.

PS.: If you still get Black Screen try with this ATINDRV.kext (works with 10.5.5, don't with 10.5.2)
If you have KP because ATIRadeonX1000.kext, backup it, remove it from the extension folder.
It gives you QE/CI but you may need to patch it.

Attached Files



#93
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

can you (ole2) post the version of atilead with yuor fix?
thanks!I will try it!


Eddie94, you definitely addressed right person, I have x200 chipset laptop and would like to enable it's video capabilities from VESA standard (fake 1024x768) to natural 1200x800 also.

my issues so far from past research are following:
- we may need to apply alternative parameters discovery (ATILead business)
- we have to try (it may not help, but suggested by Slice) to use ATIRadeon9700 driver for x200 purpose.
- standard ATINDRV framebuffer may not work with x200 chipset due to NUMA architecture and/so require to support/test/help in development of ATIFramebuffer driver Slice project from here

currently I'm reinstalling your configuration (thank's to your team iDeneb) in order to reproduce your level on my laptop. unfortunately iDenev v1.3 missing AppleATIATA driver, so I was unable to install/boot system "right away" and have to fix installations manually, so you'll have to wait until success.

I'll post here results of ATILead execution on my x200 as soon as resolve IDE drives access problem.

#94
ridgeline

ridgeline

    InsanelyMac Legend

  • Members
  • PipPipPipPipPipPipPip
  • 532 posts
  • Gender:Male
  • Location:Cape Coral, FL
Are there any updates ole2?

#95
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

Are there any updates ole2?

last night was working on mine x200 trying new Info.plist from lebidou.
EDID was aquired from windows with a help of Phoenix EDID editor.
ATINDRV from lebidou was applied as well.

9700 was patched on 0x5b60 (pci_ids 1002 5b60 'RV370 5B60 [Radeon X300 (PCIE)]')
replaced by 0x5955 (pci_ids 1002 5955 'Radeon XPRESS 200M 5955 (PCIE)')
with a help of HT utility (thanks to Slice) as well as story, where to look for Intel execution binary from here

of cource it's didn't helped yet, as Wormy isn't 9700 nickname (by the way, which one it is?)

current situation:
- with ATILead running, it's hanging (trying to load Wormy on x200 probably?)
- without ATILead text booting terminal remain

#96
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow
ATY,Wormy is intended for X1000 (R500 and up) so it can't work for X200.
I can propose you to use Callisto or Ati.kext injected by ATILead.

#97
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

ATY,Wormy is intended for X1000 (R500 and up) so it can't work for X200.
I can propose you to use Callisto or Ati.kext injected by ATILead.

this is the very last option, however these days might be only way to go
If I'm not mistaken, ATINDRV v etom sluchae ne trebuetsya takge?

#98
Slice

Slice

    InsanelyMacaholic

  • Local Moderators
  • 2,529 posts
  • Gender:Male
  • Location:Moscow

this is the very last option, however these days might be only way to go
If I'm not mistaken, ATINDRV v etom sluchae ne trebuetsya takge?


Let it present!

#99
ridgeline

ridgeline

    InsanelyMac Legend

  • Members
  • PipPipPipPipPipPipPip
  • 532 posts
  • Gender:Male
  • Location:Cape Coral, FL
Ole2 is this the 34XX card you are talking about or another card all together>

#100
ole2

ole2

    InsanelyMac Geek

  • Members
  • PipPipPipPip
  • 180 posts
  • Gender:Male
  • Location:Grenoble, France

Ole2 is this the 34XX card you are talking about or another card all together>


I do have both, actually I do have x200, x1250, x1270 and HD3470 (last one has standalone memory, rest share common one) in access.
I can investigate x200 and HD3470 actively, while AMD690g only as test drive reference (i.e. sometimes)

yesterday was x200 against ATIRadeon9700 investigation with binary patching and PCIID application based on Slice manual/suggestion/insistence.
it showd lack of success with Tiger's ATINDRV (the one provided by lebidou), now it's time to move to SliceFB framebuffer (aka ATI.kext - old name)

some feedbacks about SliceFB may appear under it's own topic





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

© 2014 InsanelyMac  |   News  |   Forum  |   Downloads  |   OSx86 Wiki  |   Mac Netbook  |   Web hosting by CatN  |   Designed by Ed Gain  |   Logo by irfan  |   Privacy Policy