Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
That framebuffer thing may actually valid for PULSE RX580 (if that's not purely cosmetic with that injection).

Apple do apply Orinoco framebuffer for the PULSE RX580 8GB card that use in the BlackMagic eGPU kit.

Nowadays, I don't think there is any difference anymore. But in the old days, having Orinoco framebuffer can avoid few bugs (e.g. the GPU can't display after wake from sleep).

The VBIOS mod to achieve that was discuss at here
Also with Orinoco I don't have the artefacts at the last boot stage just before the login screen.
 
I know but you should not use it with other cards. Orinoco has a fixed output configuration which depending on the card you are using might be incompatible. I will publish RadeonBoost 1.6 today. Maybe try it to see if it gives you the same results. You can of course also grab my parameters and inject them via bootloader if you want.

Code:
<dict>
    <key>PP,PP_Falcon_QuickTransition_Enable</key>
    <integer>1</integer>
    <key>CFG,CFG_NVV</key>
    <integer>2</integer>
    <key>CFG,CFG_USE_CP2</key>
    <true/>
    <key>PP,PP_EnableLoadFalconSmcFirmware</key>
    <integer>1</integer>
    <key>CFG,CFG_PTPL2_TBL</key>
    <data>ggAAAHwAAAB2AAAAcAAAAGoAAABkAAAAXgAAAFgAAABSAAAATAAAAEYAAABAAAAAOgAAADQAAAAuAAAAKAAAAA==</data>
    <key>ATY,EFIVersionB</key>
    <string>RadeonBoost</string>
</dict>

Those parameters are reduced to the absolute minimum necessary and aim at compatibility with all Polaris cards. Also fixes issues of some cards where the clocks are wrong after sleep or OpenGL performance is limited after sleep.

As usual, needs to be used with WEG.
 
  • Like
Reactions: h9826790
I know but you should not use it with other cards. Orinoco has a fixed output configuration which depending on the card you are using might be incompatible. I will publish RadeonBoost 1.6 today. Maybe try it to see if it gives you the same results. You can of course also grab my parameters and inject them via bootloader if you want.

Code:
<dict>
    <key>PP,PP_Falcon_QuickTransition_Enable</key>
    <integer>1</integer>
    <key>CFG,CFG_NVV</key>
    <integer>2</integer>
    <key>CFG,CFG_USE_CP2</key>
    <true/>
    <key>PP,PP_EnableLoadFalconSmcFirmware</key>
    <integer>1</integer>
    <key>CFG,CFG_PTPL2_TBL</key>
    <data>ggAAAHwAAAB2AAAAcAAAAGoAAABkAAAAXgAAAFgAAABSAAAATAAAAEYAAABAAAAAOgAAADQAAAAuAAAAKAAAAA==</data>
    <key>ATY,EFIVersionB</key>
    <string>RadeonBoost</string>
</dict>

Those parameters are reduced to the absolute minimum necessary and aim at compatibility with all Polaris cards. Also fixes issues of some cards where the clocks are wrong after sleep or OpenGL performance is limited after sleep.

As usual, needs to be used with WEG.
I just tested 2 identical cards (but different Vbios) with the Orinoco:
AMD Radeon RX 580:
Name: ATY,Orinoco
Type: Display Controller
Driver Installed: Yes
MSI: Yes
Bus: PCI
Slot: Slot-1
Vendor ID: 0x1002
Device ID: 0x67df
Subsystem Vendor ID: 0x1028
Subsystem ID: 0x1701
Revision ID: 0x00c7
Link Width: x16
Link Speed: 5.0 GT/s
AMD Radeon RX 580:
Name: ATY,Orinoco
Type: Display Controller
Driver Installed: Yes
MSI: Yes
Bus: PCI
Slot: Slot-1
Vendor ID: 0x1002
Device ID: 0x67df
Subsystem Vendor ID: 0x1002
Subsystem ID: 0x0b37
Revision ID: 0x00c7
Link Width: x16
Link Speed: 5.0 GT/s
On one of them one of the display ports does not output. Otherwise everything else does. That one is actually flashed by MVC.
Also from a regshot:
Code:
"CFG_APER_MODE" = 1
  |   "ATY,TileInfo" = <03000000000000000200000002000000030000000c000000>
  |   "IOInterruptSpecifiers" = (<1000000007000000>,<0c00000000000100>)
  |   "CFG_USE_FBC" = No
  |   "compatible" = <"pci1002,b37","pci1002,67df","pciclass,030000","GFX1">
  |   "CFG_USE_FBWRKLP" = Yes
  |   "IOPCIExpressLinkStatus" = 4354
  |   "CFG_UFL_CHK" = No
  |   "CFG_USE_AGDC" = Yes
  |   "DALReadDelayStutterOff" = 4
  |   "IOPMIsPowerManaged" = Yes
  |   "acpi-pmcap-offset" = 80
  |   "LoadController" = Yes
  |   "VRAM,totalMB" = 8192
  |   "CAIL_EnableLBPWSupport" = 0
  |   "PP_StaticMappingInfo6" = <00202000>
  |   "rebuild-device-tree" = <00>
  |   "PP_SclkDpmTuning1" = 1991680
  |   "IOPCIResourced" = Yes
  |   "PP_MclkActivityTarget" = 5
  |   "PP_SclkDpmTuning6" = 1991680
  |   "built-in" = <00>
  |   "PP_StaticMappingInfo1" = <ceffffff>
  |   "PP_DisableFFC" = 1
  |   "PP_DisableVoltageIsland" = 1
  |   "pcidebug" = "11:0:0"
  |   "CFG_NO_MST" = No
  |   "subsystem-id" = <370b0000>
  |   "PP_MclkDpmTuning0" = 2057216
  |   "CFG_USE_SCANOUT" = Yes
  |   "PP_Falcon_QuickTransition_Enable" = 1
  |   "IOName" = "display"
  |   "iofb_no_regptr" = Yes
  |   "saved-config" = <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>
  |   "CFG_FORCE_MAX_DPS" = No
  |   "revision-id" = <c7000000>
  |   "ATY,SCLK" = 30000
  |   "MM_EnableHEVCEncode" = Yes
  |   "LoadHWServices" = Yes
  |   "CFG_NO_MSI" = No
  |   "ATY,MCLK" = 30000
  |   "CFG_USE_CP2" = Yes
  |   "attached-gpu-control-path" = "IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/IOU1@7/IOPP/PXS2@0/IOPP/pci-bridge@4/IOPP/GFX1@0/ATY,Orinoco@0/AMDFramebufferVIB/AGPM"
  |   "CFG_USE_SRRB" = No
  |   "ATY,EFIVersion" = <"01.01.183">
  |   "LoadAMDSupport" = Yes
  |   "PP_SclkDpmTuning2" = 1991680
  |   "PP_StaticMappingInfo2" = <0070a730>
  |   "PP_SclkDpmTuning7" = 1991680
  |   "IOPCIMSIMode" = Yes
  |   "pci-bus-number" = 11
  |   "CFG_PAA" = 0
  |   "MVC,Version" = <"V1">
  |   "CFG_USE_STUTTER" = Yes
  |   "CFG_USE_CPSTATUS" = Yes
  |   "PP_PhmUseDummyBackEnd" = 0
  |   "IOBusyInterest" = "IOCommand is not serializable"
  |   "pci-device-number" = 0
  |   "shikigva" = <90>
  |   "PP_MclkDpmTuning1" = 2057216
  |   "model" = <"AMD Radeon RX 580">
  |   "HDCPLIB_VERSION" = <"1.500">
  |   "IOServiceDEXTEntitlements" = (("com.apple.developer.driverkit.transport.pci"))
  |   "PP_DisableDIDT" = 0
  |   "CFG_USE_HDMI20" = Yes
  |   "CFG_TRANS_WSRV" = Yes
  |   "assigned-addresses" = <10000bc200000000000000a0000000000000001018000bc200000000000000b0000000000000200020000b810000000000300000000000000001000024000b8200000000000030b1000000000000040030000b8200000000000034b10000000000000200>
  |   "CFG_USE_FEDS" = Yes
  |   "AAPL,slot-name" = <"Slot-1">
  |   "DALUseUrgencyWaterMarkOffset" = 0
  |   "ATY,memsize" = 8589934592
  |   "PP_SclkDpmTuning3" = 1991680
  |   "PP_StaticMappingInfo3" = <00000000>
  |   "CFG_NODM" = Yes
  |   "CFG_FB_LIMIT" = <04>
  |   "ATY,vram_base" = <0000000000000000>
  |   "name" = <"display">
  |   "ATY,PageFlipEnable" = <01000000>
  |   "CFG_PULSE_INT" = Yes
  |   "IOChildIndex" = 1
  |   "PP,PP_WorkLoadPolicyMask" = <20>
  |   "CAIL_EnableMaxPayloadSizeSync" = 0
  |   "PP_DisableCAC" = 0
  |   "PP_MclkDpmDisabled" = 0
  |   "PP_EnableBAPM" = 0
  |   "LoadAccelerator" = Yes
  |   "CFG_NO_SLS" = No
  |   "agdpmod" = <"pikera">
  |   "CFG_INT_SSPC" = 25
  |   "vendor-id" = <02100000>
  |   "CFG_UFL_STP" = No
  |   "IOInterruptControllers" = ("io-apic-0","IOPCIMessagedInterruptController")
  |   "CFG_GEN_FLAGS" = 0
  |   "reg" = <00000b000000000000000000000000000000000010000b420000000000000000000000000000001018000b420000000000000000000000000000200020000b010000000000000000000000000001000024000b020000000000000000000000000000040030000b0200000000000000000000000000000200>
  |   "PP_StaticMappingInfo4" = <00002000>
  |   "PP_SclkDpmTuning4" = 1991680
  |   "CFG_USE_DPT" = Yes
  |   "CFG,CFG_FB_LIMIT" = <04>
  |   "ATY,remap-size" = 1073741824
  |   "CFG_PTPL2_TBL" = <820000007c00000076000000700000006a000000640000005e00000058000000520000004c00000046000000400000003a000000340000002e00000028000000>
  |   "CFG_USE_LPT" = No
  |   "AsicOnLowPower" = <00000000>
  |   "PP_EnablePerDPM" = 1
  |   "MM_EnableHEVCDecode" = Yes
  |   "IODeviceMemory" = (({"address"=2684354560,"length"=268435456}),({"address"=2952790016,"length"=2097152}),"IOSubMemoryDescriptor is not serializable",({"address"=2972712960,"length"=262144}),({"address"=2972975104,"length"=131072}))
  |   "CFG_NO_HDCP" = No
  |   "PP,PP_PowerPlayEnabled" = <01000000>
  |   "device-id" = <df670000>
  |   "CFG_USE_TCON" = Yes
  |   "PP_DisableULV" = 0
  |   "IOPCIExpressCapabilities" = 18
  |   "subsystem-vendor-id" = <02100000>
  |   "IOPowerManagement" = {"ChildrenPowerState"=2,"CurrentPowerState"=2,"CapabilityFlags"=258,"ChildProxyPowerState"=2,"MaxPowerState"=3}
  |   "CFG_CAA" = 0
  |   "CFG_FORCEMAXDPM" = No
  |   "CFG,CFG_PTPL2_TBL" = <820000007c00000076000000700000006a000000640000005e00000058000000520000004c00000046000000400000003a000000340000002e00000028000000>
  |   "class-code" = <00000300>
  |   "PP_DisablePowerContainment" = 0
  |   "IOPCIExpressLinkCapabilities" = 71305475
  |   "PP_PowerPlayEnabled" = <01000000>
  |   "pci-function-number" = 0
  |   "CFG_NO_PP" = No
  |   "PP_CNEscapeInput" = <0c0000000000000002000000>
So all these properties already exist without injection:
Code:
<dict>
    <key>PP,PP_Falcon_QuickTransition_Enable</key>
    <integer>1</integer>
    <key>CFG,CFG_NVV</key>
    <integer>2</integer>
    <key>CFG,CFG_USE_CP2</key>
    <true/>
    <key>PP,PP_EnableLoadFalconSmcFirmware</key>
    <integer>1</integer>
</dict>
 
Hello everyone!

Together with the help of @DSM2.Hackintosh I built a Kext aiming at removing limitations put in place by Apple for PC graphics cards that reduce performance in some benchmarks like Geekbench and some scenarios where memory bandwidth is crucial.

Likely most of you already know the issue that some GPUs don't perform as expected in Geekbench. For example, a Radeon VII only scores around 50k to 60k points while on Windows it can easily reach the 82k range - with overclocking even more. This issue is not only affecting the Radeon VII but numerous other graphics cards as well.

This problem started back when macOS 10.14.5 Mojave was released, introducing the fix for fan control on Vega 56 / 64 cards. As it turns out, Apple never fixed the fan issues but instead put a dirty workaround in place. Modern AMD GPUs utilize a so-called SMU firmware to manage some of the cards functions. This SMU firmware needs to be uploaded by the graphics driver. It looks like the Vega 56 / 64 fan control bug was caused by a faulty SMU firmware implementation by Apple. Instead of fixing it, all they did was to disable loading of SMU firmware on PC graphics cards. This can be verified by force loading the SMU firmware again on Vega 56 / 64 cards. The fan control bug will then be back. Not loading SMU firmware causes a number of issues and one of them is memory bandwidth being cut in half. Since Geekbench is very sensitive to memory bandwidth, this is what causes the low scores.

So essentially, Apple scammed their customers here, putting a dirty fix in place that now affects a large amount of graphics cards. And of course this not only affects hackintosh users but also Mac Pro users or eGPU users, utilizing PC graphics cards on their Mac.

Now this is where my RadeonBoost Kext comes in. It makes sure that SMU firmware is loaded again, removing the memory bandwidth limitations, making sure the GPU can be used to its fullest potential. The results really speak for themselves.

Radeon VII Geekbench Metal (1st with Radeon Boost, 2nd without)
View attachment 906795View attachment 906796
Radeon VII Geekbench OpenCL (1st with Radeon Boost, 2nd without)
View attachment 906797View attachment 906798

RadeonBoost.kext has been built with Hackintosh users in mind, but of course it does work on a real Mac as well. That includes both Mac Pros as well as Macs with an eGPU. You can either inject the kext via OpenCore or put it into /Library/Extensions, repair permissions and rebuild kext cache. Putting it inside /L/E of course requires SIP (System Integrity Protection) to be disabled since the Kext is not signed.

Of course you can go ahead and edit the info.plist of RadeonBoost.kext to your likings. For example you can add your own properties. This comes handy if you want to additionally overclock or undervolt your GPU by using a PowerPlay Table, which of course can be added to the Kext as well.

List of supported GPUs:
  • RX480, RX580, RX590 (!!! USE ALONG WHATEVERGREEN !!!)
  • Radeon VII, RX 5500, RX 5500 XT, RX 5600, RX 5700, RX 5700 XT (WHATEVERGREEN NOT NECESSARY)

Download & Changelog
  • May 28 2020: V1.6 - Disabled SMU Firmware Loading for Navi10/14 due to broken SMU firmware starting from macOS 10.15.5, improved Polaris compatibility
  • April 24 2020: V1.4 - Added target SMBIOSes for AGPM injector to prevent overriding of stock AGPM configuration on SMBIOSes that don't need an AGPM injector (e.g. iMac19,1 etc)
  • April 20 2020: V1.3 Replaced "ATY,EFIVersion" with "ForceLoadFalconSmuFirmware" (thanks to Whatevergreen developers for discovery of this parameter), changed Polaris properties (RX480, 580, 590 users please re-test), removed Apple framebuffer properties (performance difference is minor)
  • April 19 2020: V1.2 Removed support for Vega 56, Vega 64 and Vega Frontier (REASON)
  • April 19 2020: V1.1 Removed unnecessary device properties, added integrated AGPM injector for SMBIOSes where it makes sense (iMacPro1,1; MacPro7,1; MacPro5,1; MacPro4,1)
  • April 17 2020: V1.0 Initial Public Release
For download of RadeonBoost.kext please check this posts attachements.

Enjoy!

Is there any repository where we can see all releases? Some github maybe?
Thanks for your work.
 
CMMChris,

I want to see if I understand the functioning of a Radeon VII with RadeonBoost v1.6 and Catalina 15.5.

With the latest release of Catalina, RadeonBoost v1.6 disables SMU Firmware loading for the Radeon VII . This means that the Radeon VII will have poorer performance under Catalina 15.5 + RadeonBoost v1.6, then if the Radeon VII were run under Mojave with RadeonBoost v1.4. If true, this implies that Catalina 15.5 is a step down in performance for a Radeon VII?

And no matter whether a Radeon VII is used under Mojave or Catalina, WEG is not necessary? Is WEG discouraged under these conditions?

Thanks for your input.
 
With the latest release of Catalina, RadeonBoost v1.6 disables SMU Firmware loading for the Radeon VII . This means that the Radeon VII will have poorer performance under Catalina 15.5 + RadeonBoost v1.6, then if the Radeon VII were run under Mojave with RadeonBoost v1.4. If true, this implies that Catalina 15.5 is a step down in performance for a Radeon VII?

No. Where did you read that?!
 
CMMChris,

I want to see if I understand the functioning of a Radeon VII with RadeonBoost v1.6 and Catalina 15.5.

With the latest release of Catalina, RadeonBoost v1.6 disables SMU Firmware loading for the Radeon VII . This means that the Radeon VII will have poorer performance under Catalina 15.5 + RadeonBoost v1.6, then if the Radeon VII were run under Mojave with RadeonBoost v1.4. If true, this implies that Catalina 15.5 is a step down in performance for a Radeon VII?

And no matter whether a Radeon VII is used under Mojave or Catalina, WEG is not necessary? Is WEG discouraged under these conditions?

Thanks for your input.
Radeon VII belongs to Vega20, not Navi10/14
Screenshot 2020-06-02 at 3.32.40 PM.png


WEG is not necessary for Radeon VII to use RadeonBoost, but not "will have adverse effect for Radeon VII".
 
I use 1.6 for my 5700XT with "WHATEVERGREEN" lastet version.

According to you, "RadeonBoost" should lose effect in my 5700XT, but I found the geekbench record is no promote, but my graphics card can stop fan, crazy!
 
Hi @CMMChris, i have a Radeon RX 5500 XT in my Hackintosh setup, for motion graphics with After Effects. In the first post it shows the name striked. Is it not supported? I'm curious in the performance gain with this Kext. Thanks for the work liberating the full potential anyway! Stay safe.
 
How's the performance for the Apple W5700X looking after the latest update?
 
Hello everyone!

Together with the help of @DSM2.Hackintosh I built a Kext aiming at removing limitations put in place by Apple for PC graphics cards that reduce performance in some benchmarks like Geekbench and some scenarios where memory bandwidth is crucial.

Likely most of you already know the issue that some GPUs don't perform as expected in Geekbench. For example, a Radeon VII only scores around 50k to 60k points while on Windows it can easily reach the 82k range - with overclocking even more. This issue is not only affecting the Radeon VII but numerous other graphics cards as well.

This problem started back when macOS 10.14.5 Mojave was released, introducing the fix for fan control on Vega 56 / 64 cards. As it turns out, Apple never fixed the fan issues but instead put a dirty workaround in place. Modern AMD GPUs utilize a so-called SMU firmware to manage some of the cards functions. This SMU firmware needs to be uploaded by the graphics driver. It looks like the Vega 56 / 64 fan control bug was caused by a faulty SMU firmware implementation by Apple. Instead of fixing it, all they did was to disable loading of SMU firmware on PC graphics cards. This can be verified by force loading the SMU firmware again on Vega 56 / 64 cards. The fan control bug will then be back. Not loading SMU firmware causes a number of issues and one of them is memory bandwidth being cut in half. Since Geekbench is very sensitive to memory bandwidth, this is what causes the low scores.

So essentially, Apple scammed their customers here, putting a dirty fix in place that now affects a large amount of graphics cards. And of course this not only affects hackintosh users but also Mac Pro users or eGPU users, utilizing PC graphics cards on their Mac.

Now this is where my RadeonBoost Kext comes in. It makes sure that SMU firmware is loaded again, removing the memory bandwidth limitations, making sure the GPU can be used to its fullest potential. The results really speak for themselves.

Radeon VII Geekbench Metal (1st with Radeon Boost, 2nd without)
View attachment 906795View attachment 906796
Radeon VII Geekbench OpenCL (1st with Radeon Boost, 2nd without)
View attachment 906797View attachment 906798

RadeonBoost.kext has been built with Hackintosh users in mind, but of course it does work on a real Mac as well. That includes both Mac Pros as well as Macs with an eGPU. You can either inject the kext via OpenCore or put it into /Library/Extensions, repair permissions and rebuild kext cache. Putting it inside /L/E of course requires SIP (System Integrity Protection) to be disabled since the Kext is not signed.

Of course you can go ahead and edit the info.plist of RadeonBoost.kext to your likings. For example you can add your own properties. This comes handy if you want to additionally overclock or undervolt your GPU by using a PowerPlay Table, which of course can be added to the Kext as well.

List of supported GPUs:
  • RX480, RX580, RX590 (!!! USE ALONG WHATEVERGREEN !!!)
  • Radeon VII, RX 5500, RX 5500 XT, RX 5600, RX 5700, RX 5700 XT (WHATEVERGREEN NOT NECESSARY)

Download & Changelog
  • May 28 2020: V1.6 - Disabled SMU Firmware Loading for Navi10/14 due to broken SMU firmware starting from macOS 10.15.5, improved Polaris compatibility
  • April 24 2020: V1.4 - Added target SMBIOSes for AGPM injector to prevent overriding of stock AGPM configuration on SMBIOSes that don't need an AGPM injector (e.g. iMac19,1 etc)
  • April 20 2020: V1.3 Replaced "ATY,EFIVersion" with "ForceLoadFalconSmuFirmware" (thanks to Whatevergreen developers for discovery of this parameter), changed Polaris properties (RX480, 580, 590 users please re-test), removed Apple framebuffer properties (performance difference is minor)
  • April 19 2020: V1.2 Removed support for Vega 56, Vega 64 and Vega Frontier (REASON)
  • April 19 2020: V1.1 Removed unnecessary device properties, added integrated AGPM injector for SMBIOSes where it makes sense (iMacPro1,1; MacPro7,1; MacPro5,1; MacPro4,1)
  • April 17 2020: V1.0 Initial Public Release
For download of RadeonBoost.kext please check this posts attachements.

Enjoy!

Hi Chris,

I have a question. When I shared my result with the radeonboost.kext in the tonymac forum, the TO couldn't explain why my geekbench metal scores are so high. I score 82000 with the radeonboost.kext enabled and 57000 without it. They said that 57000 with a vega 56 card was good anyway, but 82000 is really astonishing. Maybe you can help me to clear things up? After all, it's just a synthetic benchmark. (By the way, when I boot with OC and the radeonboost.kext enabled, there is very little difference to if I disable it)


ATB Mark
 

Attachments

  • Bildschirmfoto 2020-06-17 um 13.27.22.png
    Bildschirmfoto 2020-06-17 um 13.27.22.png
    742.5 KB · Views: 201
  • Bildschirmfoto 2020-06-17 um 14.00.57.png
    Bildschirmfoto 2020-06-17 um 14.00.57.png
    623.9 KB · Views: 191
it's just sad because I bought the mac to do mac things, when I could have just built a windows box to get a better optimized system. lol.

I do appreciate everyone's work here though. They've made owning the cmp that much more enjoyable.

Why is that sad? Use the best tools for the job.

I have Windows and Ubuntu sessions active now - but no Apple OSX, it's never the right tool.
 
Confirmed the codeless kext still works in Big Sur beta 1. Without RadeonBoost, the GB GPU performance still drop by 40%. In other words, we still need this kext in macOS 11 to remove the Apple's artificial lock.
RadeonBoost.png
 
  • Like
Reactions: Ausdauersportler
I've noticed some strange things sometimes when resuming from sleep. I'm assuming that this might be related to this mod. But as I have multiple monitors connected, i will sometimes get a flashing screen across my monitors. I haven't figured out an exact cause, except when attempting to resume from sleep and turning on more than one monitor on at a time prior to resuming. I did notice the following in console:

Code:
fault    00:29:49.498450 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398879) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.498476 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398880) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.529520 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398881) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.557060 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784593) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.558590 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398882) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.599394 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784594) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.601266 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398883) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
error    00:29:49.615565 -0700    Synesthesia    CMIOHardware.cpp:748:CMIOObjectRemovePropertyListener no object with given ID 105
error    00:29:49.615665 -0700    Synesthesia    CMIOHardware.cpp:748:CMIOObjectRemovePropertyListener no object with given ID 105
error    00:29:49.615727 -0700    Synesthesia    CMIOHardware.cpp:748:CMIOObjectRemovePropertyListener no object with given ID 105
error    00:29:49.615785 -0700    Synesthesia    CMIOHardware.cpp:748:CMIOObjectRemovePropertyListener no object with given ID 105
fault    00:29:49.642259 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398885) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.699351 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784597) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.701024 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398886) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.725591 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784598) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.727337 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398887) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
error    00:29:49.753733 -0700    deleted    unable to create CacheDeleteDaemonVolume for <private>
fault    00:29:49.778916 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784599) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.827492 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784600) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.827932 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398889) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.868604 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784601) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.869535 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398890) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.898100 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784602) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.899069 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398891) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.923992 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784603) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.924050 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398892) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.940574 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784604) generated error (0xe00002d5). dirtyBits=0x1, options=0x1
fault    00:29:49.941605 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398893) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.943193 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398894) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.997626 -0700    kernel    AMDRadeonAccelerator [3]: transaction ID (6784605) generated error (0xe00002d5). dirtyBits=0x81, options=0x801
fault    00:29:49.997697 -0700    kernel    AMDRadeonAccelerator [1]: transaction ID (2398895) generated error (0xe00002d5). dirtyBits=0x1, options=0x1

I'm assuming this is related? Or is this an issue with WEG?
 
I've experimented a little bit with enabling Vega64 in 1.4 and achieved even better results in OpenCL, now it's 66124, while Metal is a hair slower — 80757.

I don't have a fan issue, never had it. But as soon as I bought my card I flashed it with Sapphire BIOS as per this thread. There are a lot of info of this problem, and _some_ Vega cards apparently don't have it. But I full expect it can still reveal itself under some heavy load with new kext.
Hi DanSilov, can you share the kext? I would like to try with my VEGA64.
Or if you can show how to enable VEGA64 support manually?
Thanks
 
  • Like
Reactions: DanSilov
Hi DanSilov, can you share the kext? I would like to try with my VEGA64.
Or if you can show how to enable VEGA64 support manually?
Thanks
Thanks to DanSilov, I've got the RadeonBoost.kext 1.4 from him.
However the fan problem is not really fixed, adding below setting into the plist:
<key>PP_DisableULV</key>
<integer>1</integer>
<key>PP_Falcon_QuickTransition_Enable</key>
<integer>1</integer>
<key>PP_FuzzyFanControl</key>
<integer>0</integer>

Will make my fan stay at ZERO rpm when I started my Mac, however after I execute machine learning benchmark the fan start running into full blast and stay there (also whe I run the GB5).

However running Geekbench 5 Metal I've got below score which is about ±25% above the score (50K) without the RadeonBoost.kext.

If I ever find a solution for the fan, I'll share my finding.


1594572334430.png
 
Last edited:
sudo codesign --deep -fs - /Library/Extensions/RadeonBoost.kext

Codesign fails with:

/Library/Extensions/RadeonBoost.kext: resource fork, Finder information, or similar detritus not allowed
 
sudo codesign --deep -fs - /Library/Extensions/RadeonBoost.kext

Codesign fails with:

/Library/Extensions/RadeonBoost.kext: resource fork, Finder information, or similar detritus not allowed
Use xattr to find resource fork, finder information, etc. Use it to delete such stuff also.
 
I did. Deleted unnecessary stuff with: xattr -cr /Library/Extensions/RadeonBoost.kext, Then resigned. it went through, no errors but still won’t load?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.