Slice Posted September 26, 2017 Author Share Posted September 26, 2017 Rev 4222 It is my own wish The task was follow. My laptop Dell Latitude E6430 can be powered with Optimus enabled or disabled in BIOS. They are two different computers and I want to see in which mode I started before timeout is ended because sometimes it switched despite on my wish. Now I wrote in config.plist <key>GUI</key> <dict> <key>ShowOptimus</key> <true/> and see Clover GUI as The criteria is a number of videocard found. 2 = Embedded Intel + Discrete == "Optimus" "Intel" 1 = Discrete only == "Discrete" I am not sure if someone else needed this feature so default behaviour to be not show. 6 Link to comment Share on other sites More sharing options...
Slice Posted September 28, 2017 Author Share Posted September 28, 2017 Rev 4223It is accumulated update because I can't commit changes before.update FakeSMCmove BrandString to be calculated before useallow StrictHibernate when Fixup usedrename Optimus->Intel as it is more correctblock Dell specific SetVariable In config.plist you may set <key>Boot</key> <dict> <key>HibernationFixup</key> <true/> if you are using Lilu+HibernationFixup.kext It will allow you to use mode 25. EDITED. Don't bother! It is calculated automatically. 6 Link to comment Share on other sites More sharing options...
Slice Posted September 30, 2017 Author Share Posted September 30, 2017 Rev 4231 P-State generator in Clover will be compatible with some DSDT. I found in my Dell E6430 follow line ^CPU0._PPC = Local0 But it was not possible with current clover generated SSDT because the _PPC was a method. Now I make it to be the Name for CPU0 and link for others. Scope (\_PR.CPU0) { Name (_PPC, Zero) // _PPC: Performance Present Capabilities .... Scope (\_PR.CPU1) { Method (_PPC, 0, NotSerialized) // _PPC: Performance Present Capabilities { Return (\_PR.CPU0._PPC) } and so on. 2 Link to comment Share on other sites More sharing options...
Slice Posted October 1, 2017 Author Share Posted October 1, 2017 Rev 4233 Make menu for kext disabling inside Details menu to disable some kexts immediatly before OS started See screens enter menu But if you enter Option before choosing an OS then you can see Rev 4237 Kext disabling excluded from Option menu. It will be only in Details menu 7 Link to comment Share on other sites More sharing options...
Zenith432 Posted October 15, 2017 Share Posted October 15, 2017 Rev 4247 There was an oversight in handling hidden volumes. Volumes hidden with the GUI->HIDE mechanism in config.plist did not show up when pressing F3 to show them. This is fixed. 7 Link to comment Share on other sites More sharing options...
Slice Posted October 17, 2017 Author Share Posted October 17, 2017 Rev 4248 Fix for undefined OS version in High Sierra installer. Explanation here 4 Link to comment Share on other sites More sharing options...
Slice Posted October 17, 2017 Author Share Posted October 17, 2017 Rev 4249 Corrected two bugs. 1. Menu separation lines was clickable. 2. Entering kexts management menu may cause boot without kexts after return. Both bugs corrected here, more tests needed. 4 Link to comment Share on other sites More sharing options...
Slice Posted October 17, 2017 Author Share Posted October 17, 2017 Rev 4250 XCPM patch will no more be unconditional. <key>KernelXCPM</key> <true/> Default value is . It can arise some problems because old revision has default to true. But many users can't boot at all because of this. Let it be conditional setting from config or from Clover GUI. 10 Link to comment Share on other sites More sharing options...
Slice Posted November 2, 2017 Author Share Posted November 2, 2017 Rev 4272 - 4289 Accept updates by Rehabman #2590 3 Link to comment Share on other sites More sharing options...
Slice Posted November 4, 2017 Author Share Posted November 4, 2017 Rev 4268correct MachineSignature before wake from hibernation. It is attempt to cure an issue at some configurations that boot.efi failed to check // // check machine signature // UINT32 machineSignature = 0; if(!EFI_ERROR(AcpiGetMachineSignature(&machineSignature)) && machineSignature != localHeader.MachineSignature) try_leave(NOTHING); The patch will write the MachineSignature from hibernation image into FACS table. But it is wrong for good machines Rev 4270 condition for signature fixup Now this Signature Fixup will be conditional through config.plist <key>SignatureFixup</key> <false/> Rev 4291 if not SignatureFixup, it will be zeroed. Anyway, the Signature must be fixed If true then the value is from hibernate image if false then the value is zero and it will be copied by kernel into the image. Rev 4293 detect 10.13 installer version, by fusion71au Explanation here #15351 9 Link to comment Share on other sites More sharing options...
Slice Posted November 7, 2017 Author Share Posted November 7, 2017 Rev 4294 New Shell is able to correctly edit plists. Most important is config.plist. TAB characters converted to double spaces on the fly Enter Shell Shell> fs0: fs0:> cd EFI\CLOVER fs0:\EFI\CLOVER> edit config.plist ... Press F2 to save the result Press F3 to exit from editing fs0:> exit you will return to Clover Enjoy! 16 Link to comment Share on other sites More sharing options...
Popular Post Slice Posted November 9, 2017 Author Popular Post Share Posted November 9, 2017 Rev 4295Show kexts versions in menu to disable them. Rev 4296A possibility to de-init Radeon cards.It was very old problem since ElCapitan that AMD Radeon HD6xxx and up will not wake after sleep if used as first card.If the card is secondary and CSM enabled then the card works fine. So the differences is initialisation.The solution was found by vit9696 and used in private WhateverGreen.kext. Then Mieze proposed the solution as ACPI patch.Now I included this into Clover.You have to set in config.plist <key>Graphics</key> <dict> <key>RadeonDeInit</key> <true/> Related threads No graphics / USB / Audio after wake Tracing back the AMD GPU wakeup issue to its origin 25 Link to comment Share on other sites More sharing options...
Sherlocks Posted November 12, 2017 Share Posted November 12, 2017 Rev 4299Support Reset NVRAM with F11 in Clover GUI.operationsnative NVRAM system- directly cleanup variablesnot support NVRAM system(EmuVariableUefi-64.efi + rc script) and Legacy(+rc script) for only ESP. link- search for nvram.plist and will delete itafter reset NVRAM, will be reboot.if you want to know support NVRAM on your system, in terminaltype sudo nvram boot-args="-v"then reboot, check verbose log, if it doesn't work, not support native NVRAM on your system.Rev 4301Support Fusion Drive for 10.12+.Fix empty in block kext inject in GUI if os version is null. link linkRev 4303Improve Reset NVRAMvariables list to reset NVRAMnative NVRAM system- boot-args- nvda_drvnot support NVRAM system- detect nvram.plist and delete it on only ESP partition- boot-args- nvda_drv- hibernate keys if there isabout hibernate keys. link- after you are using Lilu + HibernationFixup without knowing for sure whether hibernation worksif you experienced instant reboot when using any macos(recovery, usbinstaller, partition),reset NVRAM with F11 help you can boot. 15 Link to comment Share on other sites More sharing options...
Sherlocks Posted November 19, 2017 Share Posted November 19, 2017 Rev 4307Cleanup, fix and support more CPU for Generate option process for power management in old/new Clover -for sandy-, default - disabled P/C States generate option -for ivy+, default - enabled P/C States generate option Do not mix the custom power management file(SSDT.aml) and the P/C States option if you want to use power management 1. only use P/C States option without SSDT.aml 2. turn off P/C States option. then put SSDT.aml in ACPI/patched folder <key>SSDT</key> <dict> <key>Generate</key> <dict> <key>CStates</key> <false/> <key>PStates</key> <false/> </dict> </dict> Rev 4313Fix without cache option(-f) -it's useful for snow leopard installation to avoid AppleIntelCPUPowerManagement KP or patched AppleHDA Rev 4316Fix potential uncertainty of without caches option(-f) - without option(-f) and SystemParameters/NoCaches option work on 10.6 to 10.9. 11 Link to comment Share on other sites More sharing options...
Slice Posted November 23, 2017 Author Share Posted November 23, 2017 Rev 4314Limit scope of binary DSDT patch by "Device" section with specified name in additional "TgtBridge" tag[/size]by goodwin_c Config.plist sample <dict> <key>Comment</key> <string>Rename PXSX to SSD0</string> <key>Disabled</key> <false/> <key>Find</key> <data>UFhTWA==</data> <key>Replace</key> <data>U1NEMA==</data> <key>TgtBridge</key> <data>UlAwOQ==</data> </dict> <dict> <key>Comment</key> <string>Rename PXSX to SSD0 2</string> <key>Disabled</key> <false/> <key>Find</key> <data>UlAwOS5QWFNY</data> <key>Replace</key> <data>UlAwOS5TU0Qw</data> </dict> 7 Link to comment Share on other sites More sharing options...
Slice Posted December 1, 2017 Author Share Posted December 1, 2017 Rev 4331 Fix EDK2 bug notice at #15797 by nms (same as nms42 at Clover credits). Thunk.S in Baselib is wrong. Good one is Thunk.nasm and we want use it. The difference in compilation Thunk.S 0000000000000077 662E0F0197F8FFFFFF lgdt fword [cs:rdi+0xfffffffffffffff8] Thunk.nasm 0000000000000077 2E660F0157EE lgdt fword [cs:rdi+0xffffffffffffffee] 5 Link to comment Share on other sites More sharing options...
Sherlocks Posted December 5, 2017 Share Posted December 5, 2017 Rev 4334 Fix potential OSInstall.mpkg issue in High Sierra -no more remove EmuVariableUefi-64.efi and nvram.plist to install/update High Sierra if still happen OSInstall.mpkg issue, delete macOS Install Data folder in root. download update/beta from appstore again. 15 Link to comment Share on other sites More sharing options...
Slice Posted December 9, 2017 Author Share Posted December 9, 2017 Rev 4337 Play Paper-Rock-Scissors game. Paper covers Rock. Rock beats Scissors. Scissors cut Paper. Discussion here rev 4339 make this compatible with gcc. 3 Link to comment Share on other sites More sharing options...
Slice Posted December 15, 2017 Author Share Posted December 15, 2017 Rev 4351, 4355 Additional patches for EDID. Now config.plist may contain follow EDID section <key>Graphics</key> <dict> <key>EDID</key> <dict> <key>#Custom</key> <data>AP///////wAGECGSAAAAAAASAQOAIRV4CunVmVlTjigmUFQAAAABAQEBAQEBAQEBAQEBAQEB3iGgcFCEHzAgIFYAS88QAAAY3iGgcFCEHzAgIFYAS88QAAAAAAAA/gBXNjU3RwAxNTRXUDEKAAAA/gAjMz1IZYSq/wIBCiAgAJo=</data> <key>#Inject</key> <true/> <key>#ProductID</key> <string>0x9221</string> <key>#VendorID</key> <string>0x1006</string> <key>#HorizontalSyncPulseWidth</key> <integer>100</integer> <key>#VideoInputSignal</key> <string>0xA5</string> </dict> </dict> Custom is EDID obtained or created manually. You mostly don't need do it as Clover usually can obtain it by himself. Inject=true informs Clover to inject Custom or Discovered EDID into video device properties. You have to set it to true is you do patches. ProductID and VendorID - may be you want to mimic your display to be Apple manufactured. But you should know what you do. Making this patch I found that Brightness control stop working. Remove this patch and brightness control is working again! HorizontalSyncPulseWidth - this patch helps to resolve garbled screen at boot with Intel graphics. Known problem of 8 apples. I see that the patch is still needed <dict> <key>Comment</key> <string>Patch 8 apples</string> <key>Disabled</key> <false/> <key>Find</key> <data>AQAAdRc=</data> <key>MatchOS</key> <string>10.11</string> <key>Name</key> <string>IOGraphicsFamily</string> <key>Replace</key> <data>AQAA6xc=</data> </dict> but this patch alone is not full solution. The apple is broken for a several seconds and then restored. With new Clover setting HorizontalSyncPulseWidth = 100 there is a full solution. The image is good and not jumped. VideoInputSignal - this is additional patch for the same problem but I found no inluence. Definition of the value is follow Bit 7=1 Digital input. If set, the following bit definitions apply: Bits 6–4 Bit depth: 000=undefined, 001=6, 010=8, 011=10, 100=12, 101=14, 110=16 bits per color, 111=reserved Bits 3–0 Video interface: 0000=undefined, 0001=HDMIa, 0010=HDMIb, 0100=MDDI, 0101=DisplayPort My default values for desktop 0x80 (Digital, all undefined), for laptop 0x90 -6bits per color?! - not good. It seems value 0xA0 will be better if used. And I don't know interface for intenal screen. Credits. I was informed by Derek about this patch but initially it was wrong (byte 64 while it must be 63+65). https://github.com/1and1get2/Clover/commit/ee301efdcce8ae1e47be21267490753a0c145c1d The discussion was at tmx site as well wrong and without result. And I don't know who is initial inventor because in those matter it can't work properly. 7 Link to comment Share on other sites More sharing options...
apianti Posted January 7, 2018 Share Posted January 7, 2018 Rev 4368, 4369 NVRAM and other runtime fixes for AptioFix. Runtime code areas are no longer allowed to be relocated by boot.efi like runtime data areas. Runtime methods for NVRAM variables are shimmed to enable writing into runtime code regions during the methods, then disabled upon return. Native NVRAM should work for everyone now. AptioFix2 should be preferred over AptioFix, but you may need to calculate a slide value for now. 14 Link to comment Share on other sites More sharing options...
Popular Post Slice Posted January 17, 2018 Author Popular Post Share Posted January 17, 2018 Rev 4378-4382 New set of memory fix drivers. OsxAptioFixDrv and OsxAptioFix2Drv reverted to rev before 4368 so without NVRAM support. Let them stay here for history. And sometimes OsxAptioFixDrv is only working version for high warped memory map. OsxAptioFix2Drv is a reduced version without relocation blocks and may not works on some hardware or with older OS version. But this version supported Hibernation while first one no. OsxAptioFix3Drv is a OsxAptioFix2Drv rev 4369 renamed to be different. This version supported NVRAM. More runtime fixes in 4379. This version can be improved by clover crew in future. AptioMemoryFix included in the package is third-party memory fix driver used as is. It is probably better then OsxAptioFix*. Choose it for a first attempt. OsxLowMemFixDrv is also memory fix driver, not improved since rev608. SIC! These 5 drivers should not be used simultaneously! Choose one of them! Clover will be updated for mutually exclusion. AptioInputFix is included in the package now. It is keyboard driver to support FileVault2 for those who have AMI UEFI BIOS. 36 2 Link to comment Share on other sites More sharing options...
Zenith432 Posted February 6, 2018 Share Posted February 6, 2018 Rev 4406 Updated built_gcc7.sh to create GCC 7.3 toolchain for use with './ebuild.sh -t GCC53'. Note that latest binutils is 2.30, but due to some unresolved bug, all versions of binutils later than 2.27 cause build failures. 1 Link to comment Share on other sites More sharing options...
Sherlocks Posted March 31, 2018 Share Posted March 31, 2018 Rev 4420 Fix 10.13.4 USB Installer issue with OS Version detection 5 Link to comment Share on other sites More sharing options...
Slice Posted April 6, 2018 Author Share Posted April 6, 2018 Rev 4427-4429 FixHeaders should be in common ACPI section, it is not DSDT-only fix. <key>ACPI</key> <dict> <key>FixHeaders</key> <true/> The FIX is recommended to all users even if you are not going to fix DSDT. Anyway the fix is safe. Old setting inside DSDT fixes remains for backward compatibility but I recommend to exclude it from those section. 11 1 Link to comment Share on other sites More sharing options...
Sherlocks Posted April 12, 2018 Share Posted April 12, 2018 Rev 4435 Fix a rare case that osversion cannot be read and cleanup for maintenance and new macOS ready - checked all macOS installation with test from 10.6 to 10.13.4 for osversion detection and improved osversion read 6 Link to comment Share on other sites More sharing options...
Recommended Posts