Slice Posted January 2, 2019 Share Posted January 2, 2019 I have no own experience with such card. I may propose to fake your card to be Hawaii family, for example 0x67B9. Link to comment Share on other sites More sharing options...
Steve Evans Posted January 3, 2019 Share Posted January 3, 2019 Thanks @Slice, that's prompted some changed behaviour. I can now booth into the Mojave installer, but there's no acceleration. Booting only works if I have SSDT-AMD-Acre.aml, posted earlier in this thread in place. This is shown in the attached log file FakeATI.0x67B91002.AcreSSDT.10.14.2.txt where we get the following of relevance to the GPU: [AGPM Controller] build gpuDict by GPU PSX1. GTrace synchronization point 0 [AGPM Controller] build gpuDict by GPU PSX1. GTrace synchronization point 1 [AGPM Controller] build gpuDict by GPU PSX1. GTrace synchronization point 2 [AGPM Controller] build gpuDict by GPU PSX1. DSMOS has arrived GIOSce synchllowiatioloboi loo ps (with warnings) Unsupported CPU Unsupported CPU Unsupported PCH IOConsoleUsers: time(0) 0->0, lin 0, llk 1, IOConsoleUsers: gIOScreenLockState 3, hs 0, bs 0, now 0, sm 0x0 GTracekPynisaoleSatien po nt 9 ATY,Acre: Not usable kPEDisableScreen -1 kPEEnableScreen -1 PSX1-B: newUserClient failed: 0 0xe00002be 0x100000602 IONDRVFramebuffer ATY,Acre: Not usable PSX1-C: newUserClient failed: 0 0xe00002be 0x100000606 IONDRVFramebuffer ATY,Radeon: Not usable PSX1-D: newUserClient failed: 0 0xe00002be 0x10000060a IONDRVFramebuffer I grabbed a copy of the output of ioreg from the Mojave installer, see ioreg.10.14.2.txt. Without the SSDT patch in place I got FakeATI.0x67B91002.10.14.2.txt with the familiar failure: Invalid Register Offset: 0xc0500198 >= 0x40000. Invalid Register Offset: 0xc0500198 >= 0x40000. Invalid Register Offset: 0xc0500198 >= 0x40000. --> [2:0:0] !!! Failed to read register 0xc0500198.. FATAL ERROR : ATIController failed to access PCI device [2:0:0]! --> [2:0:0] !!! Failed to read register 0x12480.. I'm not sure why defining ATY,Acre in the SSDT causes acceleration to fail to initialise, or is it that the AMD9500 KEXT is then failing to load? Your input would, again, be much appreciated, and in the meantime I shall look at what other device IDs I could try, Polaris compatible ones being best I would guess. Thanks, Steve Link to comment Share on other sites More sharing options...
Steve Evans Posted January 3, 2019 Share Posted January 3, 2019 67df provides accelerated GPU operations, but whilst the HDMI output works, DVI does not. Mojave fails to boot as normal. 67ef is better, but only in so far as DVI works. Steve Link to comment Share on other sites More sharing options...
Steve Evans Posted March 3, 2019 Share Posted March 3, 2019 In desperation I purchased a Sapphire Pulse RX 580 8G in the hope it would fix the problem, but alas it was the same. I figured that the issue may be that the AMD driver doesn't expect to be the sole driver and so if I had another primary display used by the BIOS, things might work better. I installed an old Nvidia Quadro 2000 card which MacOS will only use as a dumb frame buffer with no acceleration. I configured the BIOS to use this instead of the RX580 to boot, and voila, the High Sierra 2019-001 security patch installed without issue.Both RX580 displays work perfectly. The only problem is that I now have a third, unwanted display. How can I disable the Nvidia driver? Removing the IONDRVSupport.kext and all the NVDA*.kext drivers out of /System/Library/Extensions works, but is hardly elegant. How can I disable the Quadro 2000 display using Clover or SSDT etc.?Thanks,Steve Link to comment Share on other sites More sharing options...
Recommended Posts