overdue-scratch

Author Topic: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!  (Read 225256 times)

0 Members and 2 Guests are viewing this topic.

akbar102

  • Observer
  • Posts: 20
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #120 on: August 16, 2011, 09:46:04 AM »
This is similar to what I face for my card as well which is 0x95531002 (ATI4530). I am using a special boot file which allows me to have proper display but once I put my laptop to sleep and wake it up the display goes all scrambled like yours. I believe there is something happening within chameleon which is causing this issue. Another user Akbar also had the same issue which has been mentioned before in this topic.
...
Unfortunately I cannot get a Bios Dump as the reading of the Bios is prevented somehow, stupid Compaq.

Any ideas on what to do now as I suspect that if I use a different FB I will still get the same output.

Not even with Everest?.. it works on my HP, while e.g. GPU-Z doesn't.

Did you checked AniV's posts?

OK tried Everest and got the bios dump.   Read all of AniV's posts and set AtiPorts=2 with no change.

Ran bbc9's radeon bios decoder and got the following output
Code: [Select]
ATOM BIOS Rom:
SubsystemVendorID: 0x103c SubsystemID: 0x308b
IOBaseAddress: 0x6000
Filename: BR33975.001
BIOS Bootup Message:
BR33975-001 M92 DDR2 32bit 450e/500m                                       

PCI ID: 1002:9552
Connector at index 0
Type [@offset 45650]: VGA (1)
Encoder [@offset 45654]: INTERNAL_KLDSCP_DAC1 (0x15)
i2cid [@offset 45694]: 0x90, OSX senseid: 0x1
Connector at index 1
Type [@offset 45660]: LVDS (7)
Encoder [@offset 45664]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 45717]: 0x96, OSX senseid: 0x7
Here is the output of ioreg | grep ATY
Code: [Select]
    | |   |     +-o ATY,ATY,RadeonFramebuffer@0  <class AtiFbStub, id 0x1000002bb, registered, matched, active, busy 0 (4517 ms), retain 7>
    | |   |     | +-o ATY_ATY,RadeonFramebuffer  <class ATIFramebuffer, id 0x1000002c0, registered, matched, active, busy 0 (245 ms), retain 17>
    | |   |     +-o ATY,ATY,RadeonFramebuffer@1  <class AtiFbStub, id 0x1000002bc, registered, matched, active, busy 0 (8 ms), retain 7>
    | |   |     | +-o ATY_ATY,RadeonFramebuffer  <class ATIFramebuffer, id 0x1000002c8, registered, matched, active, busy 0 (6 ms), retain 13>

I can confirm that the VGA output is OK and showing me that QE/CI is working. The LVDS output still has the colors scrambled like in the picture. I am just wondering if it is due to the EDID not being detected properly?

Click to view full size!

your scrambled screen is from ATYradeonframbuffer .
what is the framebuffer that you moded it personality?
--------------------------------------------------------------------------------------------------------
isoperano our problem is deffer with this guy we can get our card work in first step but when goes to sleep or display sleep after wake up we get scrambled screeen .at this time i am fixe it with ATY_Init .as azimutz mention before may the problem is with differrent between kext and bootloader injection behavior .i try evrything that exist in the web and try to add missing part from real mac dsdt to my dsdt but nothing .seems that this problem is with 4000 mobile user only.

Azimutz

  • VoodooLabs
  • Posts: 420
  • Paranoid Android
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #121 on: August 16, 2011, 06:57:19 PM »
OK tried Everest and got the bios dump.   Read all of AniV's posts and set AtiPorts=2 with no change.
...
I can confirm that the VGA output is OK and showing me that QE/CI is working. The LVDS output still has the colors scrambled like in the picture. I am just wondering if it is due to the EDID not being detected properly?
I suggested AniV because his card has also dev id 9552, but it looks like you have one port less,
so ATIPorts=2 doesn't make sense anymore. Maybe ATIPorts=1 ;D
One thing i was missing with these dev ids (9552, 9553, etc...), is that they belong now to 4500Controller (10.6.8 and above)
which declares no fb personality; Shrike it's on 4600Controller so i don't even know how AniV get's to use Shrike??
By the logic, he should get ATY,RadeonFramebuffer like you, unless he added his dev id to 4600Controller?!
Also, he did patched the fb personality; did you try that?... here's the post with the patched drivers he shared.

This is similar to what I face for my card as well which is 0x95531002 (ATI4530). I am using a special boot file which allows me to have proper display but once I put my laptop to sleep and wake it up the display goes all scrambled like yours. I believe there is something happening within chameleon which is causing this issue. Another user Akbar also had the same issue which has been mentioned before in this topic.
Well, the sources for that booter would certainly help ;) if it's a booter your talking about..?

Remember that GE only takes advantage of the support already present on Apples drivers,
which you might know is very model limited and worse, hardware limited; they support their implementations and that's it.
All we do is give it a pull by injecting a few properties to ioreg; if the card's implementation is a perfect match,
close enough, or easily patched, bingo!.. else it's a shitty world :P
« Last Edit: August 16, 2011, 07:08:47 PM by Azimutz »
 System & Patches: http://goo.gl/i961
 Chameleon:
- trunk builds: http://goo.gl/9G1Hq
- pref pane: http://goo.gl/OL2UT

Beej

  • Entrant
  • Posts: 5
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #122 on: August 16, 2011, 08:01:01 PM »
As a side note for everyone, i'm placing this graphics stuff in stand by for a week or two;
need to focus on other stuff and get some rest :) summer is almost gone.

Just wanted to say, yes PLEASE go enjoy your summer!  Kick back with your beverage of choice and bask in the warm glow of knowing that you have helped a lot of desperate folks, including me!  We all thank you Azimutz.  You give far too much of yourself.

akbar102

  • Observer
  • Posts: 20
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #123 on: August 16, 2011, 09:15:35 PM »
Code: [Select]
Quote from: MadMungo on Today at 07:43:48 AM
OK tried Everest and got the bios dump.   Read all of AniV's posts and set AtiPorts=2 with no change.
...
I can confirm that the VGA output is OK and showing me that QE/CI is working. The LVDS output still has the colors scrambled like in the picture. I am just wondering if it is due to the EDID not being detected properly?
I suggested AniV because his card has also dev id 9552, but it looks like you have one port less,
so ATIPorts=2 doesn't make sense anymore. Maybe ATIPorts=1 
One thing i was missing with these dev ids (9552, 9553, etc...), is that they belong now to 4500Controller (10.6.8 and above)
which declares no fb personality; Shrike it's on 4600Controller so i don't even know how AniV get's to use Shrike??
By the logic, he should get ATY,RadeonFramebuffer like you, unless he added his dev id to 4600Controller?!
Also, he did patched the fb personality; did you try that?... here's the post with the patched drivers he shared.

if you add your device id to 4500 you are in wrong place .you must patch ati4600 and add your device id to it.so far i see that all of 4500 user get thier crad working with 4600controller and shrike

mckibble

  • Entrant
  • Posts: 8
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #124 on: August 17, 2011, 01:25:04 AM »
Hi Mckibble, sorry i didn't answered your sooner; didn't had much more to add than i already told Beej.
Well, can you try to get an ioreg?.. i just do it blindly :P
Also, a radeon_bios_decode dump or the rom it self, might help.
Something is not matching on the connector side, i guess..?!
Just did some tests on 10.6.7 and i still can use Megalodon, Alopias, Lamna and Triakis, just like way back on 10.6
when i first tested this with DSDT. Unfortunately we are stuck with the FB present on the respective Controller on 10.6.8+...

Hi Azimutz, no worries! Chances are we are in different timezones so the replies will come at different times  :)

The decoded ROM info:
Code: [Select]
ATOM BIOS Rom:
SubsystemVendorID: 0x1028 SubsystemID: 0x0402
IOBaseAddress: 0xdc00
Filename: S3A77109.110
BIOS Bootup Message:
A77109 RV516 LE DDR2 BIOS 450e/400m Channel A                               

PCI ID: 1002:7187
Connector at index 0 type: DVI-I
Connector at index 2 type: Composite
Connector at index 3 type: DVI-I
Connector at index 8 type: eDP

relevant parts of ioreg without EVOenabler:
Code: [Select]

    | |   +-o PCI1@1  <class IOPCIDevice, id 0x100000152, registered, matched, active, busy 0 (1436 ms), retain 12>
    | |   | +-o IOPCI2PCIBridge  <class IOPCI2PCIBridge, id 0x100000179, registered, matched, active, busy 0 (1434 ms), retain 9>
    | |   |   +-o display@0  <class IOPCIDevice, id 0x10000016b, registered, matched, active, busy 0 (1432 ms), retain 11>
    | |   |   | +-o .Display_boot  <class IONDRVFramebuffer, id 0x100000299, registered, matched, active, busy 0 (472 ms), retain 11>
    | |   |   | | +-o AppleMCCSControlModule  <class AppleMCCSControlModule, id 0x1000002b6, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |   | | | +-o AppleMCCSParameterHandler  <class AppleMCCSParameterHandler, id 0x1000002b8, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |   | | +-o AppleUpstreamUserClientDriver  <class AppleUpstreamUserClientDriver, id 0x1000002b7, registered, matched, active, busy 0 (3 ms), retain 4>
    | |   |   | | +-o display0  <class IODisplayConnect, id 0x10000030b, registered, matched, active, busy 0 (2 ms), retain 5>
    | |   |   | |   +-o AppleDisplay  <class AppleDisplay, id 0x10000030d, registered, matched, active, busy 0 (1 ms), retain 7>
    | |   |   | +-o ATISupport  <class ATISupport, id 0x100000298, registered, matched, active, busy 0 (1 ms), retain 4>
    | |   |   | +-o ATIRadeonX1000  <class ATIRadeonX1000, id 0x10000029a, registered, matched, active, busy 0 (190 ms), retain 5>
    | |   |   |   +-o IOService  <class IOService, id 0x1000002e4, registered, matched, active, busy 0 (190 ms), retain 6>
    | |   |   |     +-o IOHWSensor  <class IOHWSensor, id 0x1000002e7, registered, matched, active, busy 0 (0 ms), retain 7>
    | |   |   +-o display@0,1  <class IOPCIDevice, id 0x10000016c, registered, matched, active, busy 0 (1112 ms), retain 10>
    | |   |     +-o ATISupport  <class ATISupport, id 0x10000027c, registered, matched, active, busy 0 (1 ms), retain 4>
    | |   |     +-o IONDRVFramebuffer  <class IONDRVFramebuffer, id 0x10000027d, registered, matched, active, busy 0 (608 ms), retain 7>
    | |   |       +-o AppleMCCSControlModule  <class AppleMCCSControlModule, id 0x1000002b9, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |       | +-o AppleMCCSParameterHandler  <class AppleMCCSParameterHandler, id 0x1000002bb, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |       +-o AppleUpstreamUserClientDriver  <class AppleUpstreamUserClientDriver, id 0x1000002ba, registered, matched, active, busy 0 (1 ms), retain 4>

relevant parts of ioreg with EVOenabler:
Code: [Select]
    | |   +-o PCI1@1  <class IOPCIDevice, id 0x100000152, registered, matched, active, busy 0 (1324 ms), retain 12>
    | |   | +-o IOPCI2PCIBridge  <class IOPCI2PCIBridge, id 0x10000018d, registered, matched, active, busy 0 (1316 ms), retain 9>
    | |   |   +-o display@0  <class IOPCIDevice, id 0x10000016b, registered, matched, active, busy 0 (1314 ms), retain 13>
    | |   |   | +-o ATI1300Controller@1  <class ATI1300Controller, id 0x10000028d, registered, matched, active, busy 0 (1 ms), retain 4>
    | |   |   | +-o ATY,Caretta@0  <class AtiFbStub, id 0x100000291, registered, matched, active, busy 0 (457 ms), retain 7>
    | |   |   | | +-o ATY_Caretta  <class ATIFramebuffer, id 0x1000002b5, registered, matched, active, busy 0 (97 ms), retain 15>
    | |   |   | |   +-o AppleMCCSControlModule  <class AppleMCCSControlModule, id 0x1000002bc, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |   | |   | +-o AppleMCCSParameterHandler  <class AppleMCCSParameterHandler, id 0x1000002be, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |   | |   +-o AppleUpstreamUserClientDriver  <class AppleUpstreamUserClientDriver, id 0x1000002bd, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |   | |   +-o AMDNDRVService  <class AtiAppServices, id 0x100000318, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |   | |   +-o IOFramebufferI2CInterface  <class IOFramebufferI2CInterface, id 0x100000319, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |   | |   +-o display0  <class IODisplayConnect, id 0x10000031a, registered, matched, active, busy 0 (3 ms), retain 5>
    | |   |   | |   | +-o AppleDisplay  <class AppleDisplay, id 0x10000031c, registered, matched, active, busy 0 (1 ms), retain 7>
    | |   |   | |   +-o IOFramebufferUserClient  <class IOFramebufferUserClient, id 0x10000031e, !registered, !matched, active, busy 0, retain 5>
    | |   |   | +-o ATISupport  <class ATISupport, id 0x10000028e, registered, matched, active, busy 0 (1 ms), retain 4>
    | |   |   | +-o ATIRadeonX1000  <class ATIRadeonX1000, id 0x10000028f, registered, matched, active, busy 0 (193 ms), retain 13>
    | |   |   |   +-o IOService  <class IOService, id 0x1000002ef, registered, matched, active, busy 0 (193 ms), retain 6>
    | |   |   |   | +-o IOHWSensor  <class IOHWSensor, id 0x1000002f3, registered, matched, active, busy 0 (1 ms), retain 7>
    | |   |   |   +-o ATIR5002DContext  <class ATIR5002DContext, id 0x10000031f, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500Device  <class ATIR500Device, id 0x10000035b, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500GLContext  <class ATIR500GLContext, id 0x10000035c, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500Surface  <class ATIR500Surface, id 0x100000368, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500Surface  <class ATIR500Surface, id 0x10000036a, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500GLContext  <class ATIR500GLContext, id 0x10000036b, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500Surface  <class ATIR500Surface, id 0x10000036e, !registered, !matched, active, busy 0, retain 5>
    | |   |   |   +-o ATIR500Surface  <class ATIR500Surface, id 0x10000036f, !registered, !matched, active, busy 0, retain 5>
    | |   |   +-o display@0,1  <class IOPCIDevice, id 0x10000016c, registered, matched, active, busy 0 (1008 ms), retain 14>
    | |   |     +-o ATISupport  <class ATISupport, id 0x10000027f, registered, matched, active, busy 0 (1 ms), retain 4>
    | |   |     +-o ATY,Caretta@0  <class IONDRVDevice, id 0x100000281, registered, matched, active, busy 0 (486 ms), retain 7>
    | |   |     | +-o IONDRVFramebuffer  <class IONDRVFramebuffer, id 0x100000283, registered, matched, active, busy 0 (484 ms), retain 7>
    | |   |     |   +-o AppleMCCSControlModule  <class AppleMCCSControlModule, id 0x1000002bf, registered, matched, active, busy 0 (2 ms), retain 5>
    | |   |     |   | +-o AppleMCCSParameterHandler  <class AppleMCCSParameterHandler, id 0x1000002c1, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |     |   +-o AppleUpstreamUserClientDriver  <class AppleUpstreamUserClientDriver, id 0x1000002c0, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |     +-o ATY,Caretta@1  <class IONDRVDevice, id 0x100000282, registered, matched, active, busy 0 (7 ms), retain 7>
    | |   |       +-o IONDRVFramebuffer  <class IONDRVFramebuffer, id 0x1000002c2, registered, matched, active, busy 0 (6 ms), retain 7>
    | |   |         +-o AppleMCCSControlModule  <class AppleMCCSControlModule, id 0x1000002c4, registered, matched, active, busy 0 (1 ms), retain 5>
    | |   |         | +-o AppleMCCSParameterHandler  <class AppleMCCSParameterHandler, id 0x1000002c6, registered, matched, active, busy 0 (0 ms), retain 4>
    | |   |         +-o AppleUpstreamUserClientDriver  <class AppleUpstreamUserClientDriver, id 0x1000002c5, registered, matched, active, busy 0 (0 ms), retain 4>

So without EVOenabler, the framebuffer isn't loaded?

Furthermore, looking at the output of kextstat without EVOenabler:
Code: [Select]
   52    7 0x59041000 0x22000    0x21000    com.apple.iokit.IOGraphicsFamily (2.2) <14 7 5 4 3>
   53    1 0x594ff000 0x12e000   0x12d000   com.apple.kext.ATISupport (6.3.6) <52 14 13 7 5 4 3 1>
   54    0 0x590d0000 0x21000    0x20000    com.apple.kext.ATI1300Controller (6.3.6) <53 52 14 13 5 4 3 1>
   69    1 0x59425000 0xd000     0xc000     com.apple.iokit.IONDRVSupport (2.2) <52 14 7 5 4 3>
   75    0 0x59885000 0x5c000    0x5b000    com.apple.ATIRadeonX1000 (6.3.6) <69 52 14 7 6 5 4 3 1>
Here com.apple.kext.ATIFramebuffer is not loaded.

Hope this gives more clues on what's going on.

Enjoy your holiday, you have been a great help!  8)

Azimutz

  • VoodooLabs
  • Posts: 420
  • Paranoid Android
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #125 on: August 18, 2011, 01:32:09 AM »
Just wanted to say, yes PLEASE go enjoy your summer!  Kick back with your beverage of choice and bask in the warm glow of knowing that you have helped a lot of desperate folks, including me!  We all thank you Azimutz.  You give far too much of yourself.
Thanks for the kind words, Beej :) have your self a good time too, even if you're not on vacations, which is my case...
just taking some days, not too far away from the computer ;D

if you add your device id to 4500 you are in wrong place .you must patch ati4600 and add your device id to it.so far i see that all of 4500 user get thier crad working with 4600controller and shrike
Correct! Only now i connected the dots with these cards and the 4500Controller "no framebuffer" thing.


Hi Azimutz, no worries! Chances are we are in different timezones so the replies will come at different times...
Not really... i could have answered you earlier, but had little or nothing to add... and still don't!
Also, in my mind all the talk with Beej was also covering your problem.
I don't really know what's going on... your card should work.

Hope this gives more clues on what's going on.
A bit, but a full ioreg (or just the full graphics related) would help more; this only shows 1% of the big picture :P
Yep, the framebuffer seems to not load, but at least it seems you're not getting ATY,Framebuffer...
Pity there's so little support for these old cards... the radeon_bios_decode could be a lot more helpful; mine shows like yours:
Code: [Select]
ATOM BIOS Rom:
    SubsystemVendorID: 0x1462 SubsystemID: 0x0400
    IOBaseAddress: 0x0000
    Filename: SAA99000.101
    BIOS Bootup Message:
113-MSV040 HP RV516XT Infineon 32Mx16 128bit 512MB                         

PCI ID: 1002:7181
Connector at index 0 type: DVI-I
Connector at index 2 type: Composite
Connector at index 3 type: DVI-I
Connector at index 4 type: DVI-I
Connector at index 8 type: eDP
Connector at index 9 type: DVI-I
Try the latest package; there were some fixes on the booter that could cause problems with graphics, in specific situations e.g. check this. I already spent some time scratching my head... will give more time one of these days.

Enjoy your holiday, you have been a great help!  8)
Thanks :) my pleasure to help.
« Last Edit: August 18, 2011, 01:46:14 PM by Azimutz »
 System & Patches: http://goo.gl/i961
 Chameleon:
- trunk builds: http://goo.gl/9G1Hq
- pref pane: http://goo.gl/OL2UT

iSoprano

  • Entrant
  • Posts: 2
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #126 on: August 18, 2011, 07:06:39 AM »
if you add your device id to 4500 you are in wrong place .you must patch ati4600 and add your device id to it.so far i see that all of 4500 user get thier crad working with 4600controller and shrike
Correct! Only now i connected the dots with these cards and the 4500Controller "no framebuffer" thing.
Well my case is when I use the old module viz ATI enabler, I get full QE/CI but once it goes to sleep and wakes up then the screen is scrambled.
If I use the new AMD module enabler, I get scrambled screen.
So looks to me like there is something different between the old ATI module and the New AMD module in this chameleon branch. And if we figure this out then we can even fix the wake from sleep issue as well.

I hope I was clear in my explanation  :)


Motive for edit: just adding a proper quote, while thinking of the answer :)
 
« Last Edit: August 18, 2011, 01:03:32 PM by Azimutz »

huyhung411991

  • Entrant
  • Posts: 1
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #127 on: August 18, 2011, 07:27:49 AM »
In my case, the scrambled screen with new GraphicsEnabler (not the scrambled screen after sleep/display sleep) only appears if I use AMDGraphicsEnabler module, it doesn't if I use directly from trunk (rev > 784).

MadMungo

  • Entrant
  • Posts: 6
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #128 on: August 18, 2011, 07:45:17 AM »
OK tried Everest and got the bios dump.   Read all of AniV's posts and set AtiPorts=2 with no change.
...
I can confirm that the VGA output is OK and showing me that QE/CI is working. The LVDS output still has the colors scrambled like in the picture. I am just wondering if it is due to the EDID not being detected properly?
I suggested AniV because his card has also dev id 9552, but it looks like you have one port less,
so ATIPorts=2 doesn't make sense anymore. Maybe ATIPorts=1 ;D

Tried AtiPorts=1 and still get the scrambled screen. But now the CRT monitor only has resolutions of 800x600@60Hz & 800x600@59Hz. So forcing port 1 disables the external VGA output.

By the logic, he should get ATY,RadeonFramebuffer like you, unless he added his dev id to 4600Controller?!
Also, he did patched the fb personality; did you try that?... here's the post with the patched drivers he shared.

Am using his patched FB for lion, but still get the scrambled screen with either an origional or patched kext.

In my case, the scrambled screen with new GraphicsEnabler (not the scrambled screen after sleep/display sleep) only appears if I use AMDGraphicsEnabler module, it doesn't if I use directly from trunk (rev > 784).

I will try doing that and see what happens.

Azimutz

  • VoodooLabs
  • Posts: 420
  • Paranoid Android
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #129 on: August 18, 2011, 01:45:15 PM »
Well my case is when I use the old module viz ATI enabler, I get full QE/CI but once it goes to sleep and wakes up then the screen is scrambled.
If I use the new AMD module enabler, I get scrambled screen.
So looks to me like there is something different between the old ATI module and the New AMD module in this chameleon branch. And if we figure this out then we can even fix the wake from sleep issue as well.

I hope I was clear in my explanation  :)
I still didn't understood what booter are you using??
You mention a "special boot file" on the previous post... what boot is this?
What i can tell i already explained; on the old code, which is on the ATiGraphicsEnabler module on the 1st post package,
you will get Peregrine as framebuffer; on the new code (AMDGraphicsEnabler and trunk) the default for dev id 9553
is Shrike added by me recently. That's the major difference, when it comes to HD cards.
Also, on 10.6.8 and above, the framebuffer personalities were moved to the Controller kexts, while previously they were
on ATIFramebuffer.kext and Peregrine is gone.
If there's something we can fix on the code, i still didn't had time to find it :(

In my case, the scrambled screen with new GraphicsEnabler (not the scrambled screen after sleep/display sleep) only appears if I use AMDGraphicsEnabler module, it doesn't if I use directly from trunk (rev > 784).
Haleluia, some info about modules ;D
This is the reason why i add the devices both to AMDGraphicsEnabler and trunk, so you guys can compare;
if you get problems with the module, try the trunk before reporting.
I don't do the same with the old code because that would mean keep another branch and
i barely can keep the ones i have up to date.
The only module i can keep an eye on is the ATiGraphicsEnabler one; no HD card to test the other and
i still didn't managed to fix merging both codes. It's done partly but need time and lot's of testing for the rest.

Thanks for the info :)


Tried AtiPorts=1 and still get the scrambled screen. But now the CRT monitor only has resolutions of 800x600@60Hz & 800x600@59Hz. So forcing port 1 disables the external VGA output.
Makes sense! You have one of those "weird" roms:
Code: [Select]
ATOM BIOS Rom:
SubsystemVendorID: 0x103c SubsystemID: 0x308b
IOBaseAddress: 0x6000
Filename: BR33975.001
BIOS Bootup Message:
BR33975-001 M92 DDR2 32bit 450e/500m                                       

PCI ID: 1002:9552
Connector at index 0
Type [@offset 45650]: VGA (1)
Encoder [@offset 45654]: INTERNAL_KLDSCP_DAC1 (0x15)
i2cid [@offset 45694]: 0x90, OSX senseid: 0x1
Connector at index 1
Type [@offset 45660]: LVDS (7)
Encoder [@offset 45664]: INTERNAL_UNIPHY2 (0x21)
i2cid [@offset 45717]: 0x96, OSX senseid: 0x7
as you can see the VGA port shows in the first place; we always assume that we have to match the ports order
on the fb personality, but i think there are cases were this doesn't apply. In fact i've seen this in some Mac dump,
can't remember now what it was... will try to find it again.
Just for the sake of it, i'd try patching a personality with the ports inverted, lvds first, just to see what happens :P

Am using his patched FB for lion, but still get the scrambled screen with either an original or patched kext.
Try huyhung411991's tip too and see what happens.
Don't know what more to say atm...
« Last Edit: August 18, 2011, 02:10:41 PM by Azimutz »
 System & Patches: http://goo.gl/i961
 Chameleon:
- trunk builds: http://goo.gl/9G1Hq
- pref pane: http://goo.gl/OL2UT

mckibble

  • Entrant
  • Posts: 8
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #130 on: August 18, 2011, 06:14:30 PM »
Not really... i could have answered you earlier, but had little or nothing to add... and still don't!
Also, in my mind all the talk with Beej was also covering your problem.
I don't really know what's going on... your card should work.
...
A bit, but a full ioreg (or just the full graphics related) would help more; this only shows 1% of the big picture :P
Yep, the framebuffer seems to not load, but at least it seems you're not getting ATY,Framebuffer...

Oops, don't know why I didn't attached the whole ioreg earlier... attached now, along with kextstat.

It's no coincidence that my problems are the same as Beej's. I inherited the parts used to build my hackingtosh, and it happened to have the X1300, which led me to Beej's blog and eventually this thread.

Tried with r1421: the same failure, GUI still won't load.

Thanks!
« Last Edit: August 18, 2011, 08:06:45 PM by mckibble »

Azimutz

  • VoodooLabs
  • Posts: 420
  • Paranoid Android
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #131 on: August 18, 2011, 08:24:00 PM »
Mackibble, use
Code: [Select]
ioreg -lfw0to get the ioreg; these are not what we call "full" :)
But these are interesting to, i see now...
Do you have XCode installed?
« Last Edit: August 18, 2011, 08:27:00 PM by Azimutz »
 System & Patches: http://goo.gl/i961
 Chameleon:
- trunk builds: http://goo.gl/9G1Hq
- pref pane: http://goo.gl/OL2UT

robertx

  • Entrant
  • Posts: 5
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #132 on: August 18, 2011, 09:29:40 PM »
Hi Az...joining you here from the other place(Insanely) i have the HIS4670 IceQ 1gb AGP...hardware ids from win7 device manager PCI\VEN_1002&DEV_9495&SUBSYS_00281002&REV_00
PCI\VEN_1002&DEV_9495&SUBSYS_00281002
PCI\VEN_1002&DEV_9495&CC_030000
PCI\VEN_1002&DEV_9495&CC_0300

10.6.8... Ati4600Controller.kext has ATY,Gliff specified...

added bin(bios) for my card ....can you add this to the list....also having issues...(ati4600controller.kext) added id but can't reach desktop if loaded...depending on boot file i can get(black screen but active system)(apple logo and spinner spinning...but can still login blind)(or boot with 4600controller.kext disabled reaching desktop and then when i load it through KextWizard i get ATY,ATY Framebuffer under pci devices in system profiler)...i'm close i think...hope this is enough info to start...thx in advance


Mod notes: AGP.

Motive for edit: add info.
« Last Edit: August 18, 2011, 10:43:10 PM by Azimutz »

Azimutz

  • VoodooLabs
  • Posts: 420
  • Paranoid Android
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #133 on: August 18, 2011, 10:23:51 PM »
Hi Az... i have the HIS4670 IceQ 1gb AGP...
Oh man, you have AGP bus :-\ not good; your dev id is not even present on Apples drivers.
Let me check something...

I talked with someone recently that had his ATI AGP working, with some Netkas kext i think,
but can't find/remember who it was atm. I think it was an older card legacy, there are records of some of them work..
don't know about yours. All i found was your topic on Insanely.
I'll post or pm you if i remember who it was... getting old for this shit :P

Sorry mate...
« Last Edit: August 18, 2011, 10:42:07 PM by Azimutz »
 System & Patches: http://goo.gl/i961
 Chameleon:
- trunk builds: http://goo.gl/9G1Hq
- pref pane: http://goo.gl/OL2UT

robertx

  • Entrant
  • Posts: 5
Re: ATI/AMD GraphicsEnabler support (legacy, modules & missing devices)!
« Reply #134 on: August 19, 2011, 02:13:54 AM »
...perhaps i could load an alternate rom?(which other cards are R730?) my understanding is that my card is a PCI-E with a bridge to the AGP built on...AGPgart seems to work still...i'll send along my ioregistryexplorer output from when  i had the x1650 pro agp installed...