Contribute
Register

macOS Sierra PB: Need testers for new AMD Radeon drivers!

Status
Not open for further replies.
You can dump it yourself using this script: https://www.tonymacx86.com/threads/...amd-radeon-drivers.197273/page-3#post-1285769

Can't do it myself at the moment since the (OS X) SSD in my Hackintosh has failed, and my old MBP is still running El Capitan.

Following up on this in the new 10.12.4 PB v3, with a Gigabyte RX 460 4GB card (Dual-link DVI, DP and HDMI connectors). I did not expect to have miracles occur, but it was worth a shot. I found:

Baladi:
No discernible difference between this and inject ATI = false. Card listed as AMD R9 xxx 4096 MB. The DP and HDMI connectors work.

BerbiceA:
Lost DP connector, as well as HDMI connector. Card listed as AMD Radeon RX 460 4096 MB.

Berbice:
Lost DP connector, as well as HDMI connector. Card listed as AMD Radeon RX 460 4096 MB.

Exmoor:
No discernible difference between this and inject ATI = false. Card listed as AMD R9 xxx 4096 MB. The DP and HDMI connectors work.

So it looks like, at the moment, using inject ATI=true in 10.12.4 is going to cause some problems for some of the RX cards.
 
Following up on this in the new 10.12.4 PB v3, with a Gigabyte RX 460 4GB card (Dual-link DVI, DP and HDMI connectors). I did not expect to have miracles occur, but it was worth a shot. I found:

Baladi:
No discernible difference between this and inject ATI = false. Card listed as AMD R9 xxx 4096 MB. The DP and HDMI connectors work.

BerbiceA:
Lost DP connector, as well as HDMI connector. Card listed as AMD Radeon RX 460 4096 MB.

Berbice:
Lost DP connector, as well as HDMI connector. Card listed as AMD Radeon RX 460 4096 MB.

Exmoor:
No discernible difference between this and inject ATI = false. Card listed as AMD R9 xxx 4096 MB. The DP and HDMI connectors work.

So it looks like, at the moment, using inject ATI=true in 10.12.4 is going to cause some problems for some of the RX cards.

How are you loading these? Clover configurator with Framebuffer name inserted, or other type of injection?
 
NOT TRUE,

cause we have another Framebuffer within AMD9510Controller.kext:

-------------------------AMD9510Controller.kext-------------------------

Exmoor (6) @ 0x106070
LVDS, LVDS, DP, DP, DP, DP
020000000001000000010151000000002205020400000000
020000000001000000010261010000001204010300000000
000400000403000000010343000000001102030100000000
000400000001000000010431000000002103050500000000
000400000403000000010523000000001000040200000000
000400000001000000010611000000002001050500000000

Berbice (5) @ 0x106100
LVDS, DP, DP, DP, DP
020000000001000039050108000000002001010100000000
000400000001000000010243000000001000020200000000
000400000403000000010313000000002103030300000000
000400000403000000010453000000001102040400000000
000400000403000000010533000000001204050500000000

BerbiceA (5) @ 0x1062b0
LVDS, DP, DP, DP, DP
020000000001000039050108000000002001010100000000
000400000001000000010243000000001000020200000000
000400000403000000010313000000002103030300000000
000400000403000000010453000000001102040400000000
000400000403000000010533000000001204050500000000

Baladi (6) @ 0x106370
DP, DP, DP, DP, DP, DP
000400000403000000010300000000001204030300000000
000400000403000000010100000000001102010100000000
000400000403000000010200000000002103020200000000
000400000403000000010400000000002205040400000000
000400000403000000010500000000001000050500000000
000400000403000000010600000000002001060600000000

-----------------------------------------------------------------------------

and we have a new kext for AMD: AMDLegacySupport.kext
with some interesting codesnippet:

<key>IOPCIMatch</key>
<string>0x94C81002 0x95831002 0x95881002 0x95011002 0x95531002 0x95401002 0x94901002 0x94881002 0x94401002 0x944a1002 0x68981002 0x68991002 0x68E01002 0x68E11002 0x68D81002 0x68C01002 0x68C11002 0x68D91002 0x68B81002 0x68B01002 0x68B11002 0x68A01002 0x68A11002 0x67681002 0x67701002 0x67791002 0x67601002 0x67611002 0x67501002 0x67581002 0x67591002 0x67401002 0x67411002 0x67451002 0x67381002 0x67391002 0x67201002 0x67221002 0x67181002 0x67191002 0x68401002 0x68411002 0x67041002</string>

Of cause i allready tried to insert RX480 ID "0x67DF1002" into it... and YES, it gets loaded than. But all it does is just this:
Bildschirmfoto_2017_02_21_um_00_15_55.png

So i thought, i just disable IGPU, set CSM to boot legacy and restartetd...
but SORRY, nothing has changed, connected screens still stay black. But was worth a try.

Mork, when you loaded the Legacy Framebuffer did you undo your 4100kext fixes? It looks like the 4100kext took over and didn't give the legacy kext a chance.
 
How are you loading these? Clover configurator with Framebuffer name inserted, or other type of injection?

The usual Clover way; for example:
Code:
    <key>Graphics</key>
    <dict>
        <key>Inject</key>
        <dict>
            <key>ATI</key>
            <true/>
            <key>Intel</key>
            <false/>
            <key>NVidia</key>
            <false/>
        </dict>
        <key>FBName</key>
        <string>Baladi</string>
    </dict>

Repeat with each of the options.

I tend to not use Configurator, just edit using Text Wrangler or Text Edit.
 
The usual Clover way; for example:
Code:
    <key>Graphics</key>
    <dict>
        <key>Inject</key>
        <dict>
            <key>ATI</key>
            <true/>
            <key>Intel</key>
            <false/>
            <key>NVidia</key>
            <false/>
        </dict>
        <key>FBName</key>
        <string>Baladi</string>
    </dict>

Repeat with each of the options.

I tend to not use Configurator, just edit using Text Wrangler or Text Edit.

It's the same I guess either way, did it pick it up in IORegistry?
 
Mork, when you loaded the Legacy Framebuffer did you undo your 4100kext fixes? It looks like the 4100kext took over and didn't give the legacy kext a chance.
No, just inserted the RX480 ID (0x67DF1002) to both kexts: AMDLegacySupport.kext and 4100.kext.
AMDLegacySupport.kext loaded as you can see in my screenshot above.
 
No, just inserted the RX480 ID (0x67DF1002) to both kexts: AMDLegacySupport.kext and 4100.kext.
AMDLegacySupport.kext loaded as you can see in my screenshot above.

I'll give it a try later and try deactivating the 4100 info.plist edit so it can't load the 4100kext.
 
No, just inserted the RX480 ID (0x67DF1002) to both kexts: AMDLegacySupport.kext and 4100.kext.
AMDLegacySupport.kext loaded as you can see in my screenshot above.
Hello,

To be clear, with the new beta, we're now to add 0x67DF1002 to AMDLegacySupport.kext, in addition to AMDController9510.kext, and AMDRadeonX4100.kext for the RX480 to work?

Thanks.
 
Hello,

To be clear, with the new beta, we're now to add 0x67DF1002 to AMDLegacySupport.kext, in addition to AMDController9510.kext, and AMDRadeonX4100.kext for the RX480 to work?

Thanks.


No, he was just testing the legacy kext to see if he could get dedicated graphics mode to work. Just add 67DF1002 to info.plist of 9510 and 4100 kexts.
 
It's the same I guess either way, did it pick it up in IORegistry?

I only checked when trying the Baladi framebuffer name. I did not find anything beyond the generic AMDFrameBuffer listing. The two Berbice names did not just give black screens but no-signal screens, so there was not much point to looking at those. I figured the Exmoor would be the same as Baladi.
 
Status
Not open for further replies.
Back
Top