Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iM 4,1-14,4 2011 iMac Graphics Card Upgrade
- Thread starter MichaelDT
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
That build doesn't crash - but still fails to detect my RX480, although its device-id is listed in ModelArray.py ...
This crashes still on iMac12,2 in the iGPU loop
Code:
Traceback (most recent call last):
File "/Users/me/Downloads/Archive/OpenCore-Patcher.command", line 190, in <module>
OpenCoreLegacyPatcher().main_menu()
File "/Users/me/Downloads/Archive/OpenCore-Patcher.command", line 29, in __init__
Build.BuildOpenCore(self.constants.custom_model or self.current_model, self.constants).check_pciid(False)
File "/Users/me/Downloads/Archive/Resources/Build.py", line 50, in check_pciid
self.constants.igpu_vendor = self.hexswap(binascii.hexlify(self.constants.igpu_devices[0]["vendor-id"]).decode()[:4])
IndexError: list index out of rangeMaybe the iGPU detection doesn't work when it's running on a system with iGPU disabled by OCLP?
Correct, the issue is that we force class code FFFFFFFF to ensure macOS doesn't treat the iGPU as a real GPU and instead treats as a generic PCI device allowing easy sleep wake support
Need to add some error checking to avoid issues when incorrect class code is present
This one runs, discover my dGPU, but uses the iMac15,1 - I would call it a near miss
When selecting the Patcher Options -> Force iMac Metal Patch -> AMD manually everything is fine.
Working here as well!
You might also want to revise lines 412 and 416, and the code around those lines.
Code:
- Adding OpenCore v0.6.8 RELEASE
- Adding config.plist for OpenCore
- Adding Lilu.kext 1.5.2
- Adding WhateverGreen.kext 1.4.9
- Adding NightShiftEnabler.kext 1.1.0
- Adding SMC-Spoof.kext 1.0.0
- Adding CPUFriend.kext 1.2.3
- Adding CatalinaBCM5701Ethernet.kext 1.0.1
- Adding AppleALC.kext 1.6.0
- Detected Wifi Card: 168C:002A
- Adding IO80211HighSierra.kext 1.0.0
- Adding SSDT-CPBG.aml
- Adding USB-Map.kext
- Adding audio properties
- No iGPU detected
- Detected dGPU: 1002:67DF
- Adding AMD DRM patches
- Adding OpenCanopy GUI
- Adding ExFatDxeLegacy.efi
- Adding SATA Hibernation Patch
- Enabling ShowPicker
- Disabling SecureBootModel
- Spoofing to iMacPro1,1
- Using Minimal SMBIOS patching
- Cleaning up filesPerfect! Seeing as the current 0.0.23 build breaks for most iMac users with upgraded GPUs I will do a new release shortly. Gonna fix that issue in the SMBIOS section first, thanks for pointing that out @internetzel!
You got the iMacPro1,1 without selecting
Patcher Options -> Force iMac Metal Patch -> AMD
manually?
Yes, the iMac 11,1 is not eligible for iMac15,1 spoofing because it cannot have an iGPU.
I got iMacPro 1,1 after selecting Patcher Options -> Force iMac Metal Patch -> Nvidia. Otherwise, I got iMac 15,1.
Will be fixed, too. At least we can make it work for the next hours until a new version comes up.
This is a Geekbench 5 screenshot of someone here recently using a dGPU K1100M:
iMac 11,2, using an i7-860S, notice the spoofed Board-ID. It looks like a iMacPro 1,1.
iMac 11,2, using an i7-860S, notice the spoofed Board-ID. It looks like a iMacPro 1,1.
Last edited:
@internetzel Could you try the below build and see if it correctly selects iMacPro1,1 over iMac15,1?
Nvm forgot to save, sending a new build in a sec
Nvm forgot to save, sending a new build in a sec
@internetzel Sorry about the double ping, this build should be good now with correct SMBIOS being used
@internetzel got the iMacPro1,1, it was me with my iMac12,2...
This latest build generates this
Code:
Traceback (most recent call last):
File "/Users/wolf/Downloads/Archive-2/OpenCore-Patcher.command", line 11, in <module>
from Resources import Build, ModelArray, Constants, SysPatch, Utilities, CliMenu
File "/Users/wolf/Downloads/Archive-2/Resources/Build.py", line 412
elif (self.constants.current_gpuv == "AMD (0x1002)") & (self.constants.current_gpud in ModelArray.AMDMXMGPUs) & (not self.constants.custom_model:
SyntaxError: invalid syntax@Ausdauersportler Haha yeah that was the issue with the earlier build I sent. Forgot to press save ;p
Please try the latest one I sent:
forums.macrumors.com
Please try the latest one I sent:
2011 iMac Graphics Card Upgrade
@internetzel Sorry about the double ping, this build should be good now with correct SMBIOS being used
forums.macrumors.com
@Ausdauersportler Haha yeah that was the issue with the earlier build I sent. Forgot to press save ;p
Please try the latest one I sent:
2011 iMac Graphics Card Upgrade
@internetzel Sorry about the double ping, this build should be good now with correct SMBIOS being used
looks perfect to me!
Code:
Created build folder
- Adding OpenCore v0.6.8 RELEASE
- Adding config.plist for OpenCore
- Adding Lilu.kext 1.5.2
- Adding WhateverGreen.kext 1.4.9
- Adding NightShiftEnabler.kext 1.1.0
- Adding SMC-Spoof.kext 1.0.0
- Adding CPUFriend.kext 1.2.3
- Adding CatalinaBCM5701Ethernet.kext 1.0.1
- Adding AppleALC.kext 1.6.0
- Detected Wifi Card: 14E4:43BA
- Found supported WiFi card, skipping wifi patches
- Adding USB-Map.kext
- Adding dual GPU patch
- Adding audio properties
- Detected iGPU: 8086:0102
- Detected dGPU: 1002:67EF
- Adding AMD DRM patches
- Disabling unsupported iGPU
- Adding OpenCanopy GUI
- Adding SATA Hibernation Patch
- Enabling ShowPicker
- Disabling SIP
- Disabling SecureBootModel
- Spoofing to iMacPro1,1
- Using Minimal SMBIOS patching
- Cleaning up files
Your OpenCore EFI has been built at:
/Users/me/Downloads/Archive/Build-Folder/OpenCore-RELEASE
Press [Enter] to go back.Great to hear, did a new release of 0.1.0:
Was kinda hoping for 0.1.0 to be our GUI release but I guess that can be 0.1.1 ;p
Release 0.1.0 · dortania/OpenCore-Legacy-Patcher
0.1.0 Fix crash on iMacs with Metal GPUs 0.0.23 Fix MacBookPro4,1 15" and 17" audio support Fix iMac7,1 24" and iMac9,1 24" audio support Fix Macmini4,1 audio support Increment binaries AppleAL...
github.com
Was kinda hoping for 0.1.0 to be our GUI release but I guess that can be 0.1.1 ;p
I installed the 0.10 from Github, it worked as expected. Thanks and thanked!Great to hear, did a new release of 0.1.0:
Release 0.1.0 · dortania/OpenCore-Legacy-Patcher
0.1.0 Fix crash on iMacs with Metal GPUs 0.0.23 Fix MacBookPro4,1 15" and 17" audio support Fix iMac7,1 24" and iMac9,1 24" audio support Fix Macmini4,1 audio support Increment binaries AppleAL...github.com
Was kinda hoping for 0.1.0 to be our GUI release but I guess that can be 0.1.1 ;p
PS Even though I am in Montreal, I am originally from AB. Cheers!
I have helped yesterday a friend to upgrade his 27'' 2011 from K1100M to K3100M. It is a BFR variant, flashed this rom and all is working fine. Tested on Catalina and later upgraded to Big Sur using OCLP, no problem at all. Also installed UEFI windows 10 with updated Nvidia drivers, all is good and brightness control works (after adding registry entries).
Installation was not easy:
It's been the second time I have the problem that, after attaching display, display gets black screen (backlight works) and iMac enters a boot loop and does not even respond to PRAM resets. Only solution seems to remove GPU and then do the PRAM reset without gpu installed. Without display attached iMac booted fine and card could be seen and flashed from linux.
At first I thought card was broken or incompatible, but having done so in the past, I decided to try it without heatsink and bracket (just a quick boot, no more that 30 to 45 seconds without heatsink). This time card booted fine and boot selector was shown... what the heck! Attached heatsink, bracket and.... black screen and boot loop, but just after attaching display, without display iMac booted fine on linux. Removed once again heatsink and bracket and mac booted fine with display attached. It was clearly some kind of hardware install problem I could not spot looking at bracket isolation or heatsink attachment, so I decided to add some isolation to bracket with tape and do a more careful install of heatsink without over-tightening screws... and this time it finally worked!
I guess that for people having similar black screen / boot loop issues when attaching display it's worth doing the quick boot test without heatsink and bracket (just use the side screws that attach the card to the motherboard). This way you can easily discard or confirm hardware install problems. DO NOT have the card powered more than 30 to 45 seconds without heatsink, specially Nvidia cards heat very quickly.
Last edited:
This
Giving the card more air to breath made the problem go away.
Can confirm having had such a physical stress situation where I had over tightened the screws with a (new Dell) K4100M back in summer 2020.
Giving the card more air to breath made the problem go away.