苹果系统之家

v1.5.8

• Inverted logic for GVA support, which is now disabled by default and can be enabled by enable-gva-support.

v1.5.7

• Fixed maximum backlight level on Ice Lake IGPUs

v1.5.6

• Fixed deprecated code in unfairgva

v1.5.5

• Changed the default delay of optimizing display data buffer allocations from 0 to 1 second to fix the issue that both internal and external displays flicker on some Ice Lake-based laptops. (Thanks @m0d16l14n1)
• Disabled the black screen fix on Ice Lake platforms as it is only applicable to SKL/KBL/CFL/CML platforms.
• Disabled the force complete modeset submodule on Ice Lake platforms as HDMI/DVI connections are not supported by the driver.
• Added AMD Radeon RX 5000 series PWM backlight control support. (Thanks to @kingo132)

v1.5.4

• Added the fix for the short period garbled screen after the system boots on Ice Lake platforms. (by @0xFireWolf, also thanks @m0d16l14n1 and @kingo132)

v1.5.3

• Added no-gfx-spoof to avoid forcing device-id values from PCI I/O.
• Added the backlight smoother submodule that makes brightness transitions smoother on Intel IVB+ platforms. (by @0xFireWolf)
• MMIO Register Access submodules are now available on Intel IVB+ platforms. (by @0xFireWolf)
• Improved ASUS-made AMD R9 380 GPU identification
• Fixed applbkl property with <00 00 00 00> value failing to disable backlight patches
• Note: This release requires Lilu v1.5.6 or later.

v1.5.2

• Added device-id spoofing support for AMD graphics

v1.5.1

• Added constants required for macOS 12 update
• Added Intel Arrandale graphics support on 10.6 and 10.7 64-bit

v1.5.0

• Fixed AMD WX-4170 name for 67E0 device id
• Added NVIDIA driver error logging with -ngfxdbg

v1.4.9

• Added per-GPU disabling API: inject disable-gpu to disable
• Added per-GPU disabling kernel version specification: inject disable-gpu-min / disable-gpu-max to select kernel version to disable (inclusive range)
• Added IGPU disabling API: inject disable-gpu to disable or use -wegnoigpu boot argument
• Optimised Rocket Lake startup as IGPU is unsupported

v1.4.8

• Fixed debug messages from cursor manipulation with NVIDIA GPUs on macOS 11

v1.4.7

• Implemented unfairgva device property (use <01 00 00 00> value for MP5,1 to enable streaming DRM)

v1.4.6

• Backlight registers fix replaces the previous Coffee Lake backlight fix and is now available on Intel Ice Lake platforms.
• Boot argument igfxcflbklt=1 as well as device property enable-cfl-backlight-fix are deprecated and replaced by -igfxblr and enable-backlight-registers-fix.
• Add max pixel clock override through -igfxmpc boot argument or enable-max-pixel-clock-override and max-pixel-clock-frequency device properties
• Moved PNLF samples to OpenCore

v1.4.5

• Enabled loading in safe mode (mainly for AGDP fixes)
• Resolved an issue that the maximum link rate fix is not working properly on Intel Comet Lake platforms. (Thanks @CoronaHack)
• Allowed enabling igfxrpsc on Comet Lake
• Fixed failed to route IsTypeCOnlySystem warning from Skylake to Ice Lake

v1.4.4

• Extended the maximum link rate fix: Now probe the rate from DPCD automatically and support Intel ICL platforms. (by @0xFireWolf)
• Fixed an issue that LSPCON driver causes a page fault if the maximum link rate fix is not enabled. (by @0xFireWolf)

v1.4.3

• Added CFL and CML P630
• Added MacKernelSDK with Xcode 12 compatibility
• Fixed loading on macOS 10.11 and earlier

v1.4.2

• Fixed disable-external-gpu (-wegnoegpu) on some systems
• Disabled RPS control patch by default due to a bug in 10.15.6 IGPU drivers
• Replaced igfxnorpsc=1 with igfxrpsc=1 to opt-in RPS control patch
• Support all valid Core Display Clock (CDCLK) frequencies to avoid the kernel panic of “Unsupported CD clock decimal frequency” on Intel ICL platforms. (by @0xFireWolf)
• Fix the kernel panic caused by an incorrectly calculated amount of DVMT pre-allocated memory on Intel ICL platforms. (by @0xFireWolf)

v1.4.1

• Added igfxmetal=1 boot argument (and enable-metal property) to enable Metal on offline IGPU
• Fixed applying patches on CometLake IGPUs, thx @apocolipse
• Added constants required for 11.0 update
• Added the use of RPS control for all the command streamers on IGPU (disabled via igfxnorpsc=1)
• Add -igfxvesa to disable Intel Graphics acceleration.
• Fix black screen on igfx since 10.15.5
• Add workaround for rare force wake timeout panics on Intel KBL and CFL.
• Add Intel Westmere graphics support.

v1.4.0

• Added 0x3EA6, 0x8A53, 0x9BC4, 0x9BC5, 0x9BC8 IGPU device-id
• Fixed framebuffer-conX-alldata patching regression
• Added disable-hdmi-patches device property alias to -igfxnohdmi

v1.3.9

• Added igfxdumpdelay boot argument to delay -igfxdump in ms
• Partially fix ICL framebuffer patching
• Add support to injecting Force_Load_FalconSMUFW from OpenCore
• Disabled automatic enabling of GVA for Polaris on 10.13 and lower
• Replaced -radnogva argument with radgva=0/1 to force GVA for Polaris
• Added wegtree=1 boot argument (rebuild-device-tree property) to force device renaming on Apple FW

v1.3.8

• Added igfxfw=2 boot argument and igfxfw IGPU property to load Apple GuC firmware
• Added igfxpavp=1 boot argument (and igfxpavp property) to force enable PAVP output
• Added igfxfcms=1 boot argument (and complete-modeset property) on Skylake and Apple
• Improved performance with Lilu 1.4.3 APIs
• Added -igfxfbdbg boot argument to debug IGPU framebuffer (debug builds only)
• Added igfxagdc=0 boot argument and disable-agdc IGPU property to disable AGDC
• Added igfxonln=1 boot argument and force-online IGPU property force online status for all displays
• Added igfxonlnfbs=MASK boot argument and force-online-framebuffers IGPU property to override display status

v1.3.7

• Improved the maximum link rate fix: Now correct the value read from extended DPCD as well. (by @0xFireWolf)
• Improved firmware loading handling on 10.15.4 (may fix booting issues on KBL+)
• Improved support for Comet Lake IGPUs (thx @stormbirds)

v1.3.6

• Enabled CoreLSKD streaming patches by default for AMD hardware DRM on Ivy Bridge
• Repurposed 64 bit for FP 2.x streaming hardware accelerated streaming patches (can be used as shikigva=80)
• Fixed accelerator name update logic for X4xxx kexts
• Fixed Verde IOGVACodec injection to make hardware video decoder work
• Enable software TV+ decoding on all CPUs without IGPU (shikigva=256)
• Added HEVC capabilities to AMD6 decoders for all GPUs (disabled by -radnogva or disable-gva-support)
• Added HW decoder device-id spoofing via -radcodec boot-arg, by @osy86

v1.3.5

• Added Lilu 1.4.0 support, which is now the minimum supported version
• Dropped legacy boot arguments (-shikigva, -shikifps)
• Fixed handling agdpmod GPU property (in IGPUs and in conjunction with boot-arg)
• Added -wegtree boot argument to force device renaming
• Fixed FairPlay DRM playback patches on 10.15
• Added shikigva and shiki-id aliases in IORegistry
• Added applbkl aliases to IORegistry (data, 32-bit)
• Added applbkl-name and applbkl-data IORegistry data keys to provide custom backlight data
• Fixed applying CoreFP patches on Apple firmware, when they are not needed
• Added shikigva=16 (repurposed) property to use AMD hardware DRM decoder in select apps
• Added shikigva=128 (repurposed) property to use hardware decoder for FairPlay 1.0 (can be used as shikigva=144)
• Do not disable DRM patches when shikigva is used even on Apple hardware for MacPro5,1 support

v1.3.4

• Added support for disabled AppleGraphicsDevicePolicy in AMD drivers on 10.15.1
• Added basic support for -radcfg and -radgl on AMD Navi GPUs

v1.3.3

• Rework backlight panel info injection to fix Mac issues

v1.3.2

• Added more GPUs for detection
• Enable IGPU graphics kernel panic workaround on 10.14.4+ on SKL

v1.3.1 更新

• Fixed an issue that LSPCON driver fails to set the mode after the adapter power is off, i.e. sleep/wake up cycle.
• Unified release archive names
• Enforce complete IGPU modeset on Kaby Lake and newer (overridable by igfxfcmsfbs bootarg or
  complete-modeset-framebuffers device property)
• Disable VRAM testing on AMD GPUs on 10.14.4+ (based on vladie’s patch)