Jump to content
Welcome to InsanelyMac Forum

Register now to gain access to all of our features. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.


  • Content count

  • Joined

  • Last visited

About mcsmart

  • Rank
    InsanelyMac Geek

Contact Methods

  • Website URL

Profile Information

  • Gender
  • Location
  • Interests
    Un*x-like systems, PCs, Scripting and Programming
  1. Clover General discussion

    Wow, Clover's development is moving forward very quickly. I've now installed Clover r2474, but it still fails to inject a HDAU device. This is my Darwin Dumper output. Thanks for looking into this again.
  2. Clover General discussion

    Alright, this is my Darwin Dumper output (full ZIP file, 7,7MB). Thanks for looking into this. - mcsmart
  3. Clover General discussion

    Okay, I found your post about AddProperties. With this I got "hda-gfx" in my GFX0 section, yay! But I still don't get HDMI audio from my Radeon 7850, although the connector types are correct in the IOReg. When looking at my DSDT I don't see any HDAU device (I've only got HDEF), which is probably the reason why HDMI audio still fails. I have set the FIX_DISPLAY DSDT fix, which is supposed to add that, but HDAU is simply not there. Anything else I could do about that? This is my Clover DSDT patch log: 13:394 0:001 ========= Auto patch DSDT Starting ======== 13:397 0:003 VideoCard devID=0x68191002 13:399 0:001 DisplayADR1[0] = 0x10000, DisplayADR2[0] = 0x0 13:400 0:001 Display 0 is notPCIE 13:402 0:001 Audio HDA (addr:0x10000) setting specified layout-id=1 (0x1) 13:403 0:001 USBADR[0] = 0x140000 and PCIe = 0xFFFE 13:405 0:001 USBADR[1] = 0x1A0000 and PCIe = 0xFFFE 13:656 0:251 Audio HDA (addr:0x1B0000) setting specified layout-id=1 (0x1) 13:658 0:001 USBADR[2] = 0x1D0000 and PCIe = 0xFFFE 13:661 0:003 Found ACPI CPU: CPU0 And CPU1 And CPU2 And CPU3 And CPU4 And CPU5 And CPU6 And CPU7 13:675 0:014 Found PCIROOTUID = 0 13:677 0:001 patch RTC in DSDT 13:678 0:001 Start RTC Fix 13:680 0:001 found RTC Length not match, Maybe will case CMOS reset will patch it. 13:681 0:001 found RTC had IRQNoFlag will move -3 bytes 13:683 0:001 ...len=B380 13:685 0:001 new size written to 2302 shift=0 len=B380 13:686 0:001 len after correct outers B380 13:688 0:001 patch TMR in DSDT 13:691 0:003 patch IPIC in DSDT 13:692 0:001 PIC size=A5 at 20CB 13:694 0:001 found CRS at 20DF size 8D 13:696 0:001 found PIC had IRQNoFlag will move -3 bytes 13:697 0:001 Fix Device PIC size -3 13:699 0:001 patch HPET in DSDT 13:701 0:001 patch LPC in DSDT 13:702 0:001 Start LPCB Fix 13:704 0:001 found LPCB device NAME(_ADR,0x001F0000) at 19AD And Name is LPCB 13:705 0:001 NewName LPCB already present, renaming impossibble 13:707 0:001 Start Display0 Fix 13:709 0:001 add device GFX0 13:713 0:004 Creating DSM for ATI card 13:715 0:001 now inserting Video device 13:718 0:003 ... into existing bridge 13:721 0:003 patch Display #0 of Vendor=0x1002 in DSDT old way 13:724 0:002 Start NetWork Fix 13:726 0:002 found NetWork device [0x001C0002:0] at 3C8C and Name is PXSX 13:729 0:002 Name PXSX present at 0x259, renaming to GIGE 13:731 0:002 Name PXSX present at 0x2FA, renaming to GIGE 13:734 0:002 NetworkADR1=1C0002 NetworkADR2=0 13:737 0:002 network DSM created, size=74 13:741 0:004 Start SBUS Fix PCI=1081 len=B47B 13:743 0:001 device (SBUS) found at 70DA 13:745 0:002 SBUS address 70DA code size = 0x0000005D 13:748 0:003 SBUS code size fix = 0x0000005D 13:751 0:002 patch HDEF in DSDT 13:753 0:002 Start HDA Fix 13:756 0:002 found HDA device NAME(_ADR,0x001B0000) And Name is HDEF 13:759 0:002 NewName HDEF already present, renaming impossibble 13:761 0:002 Start USB Fix 13:764 0:002 USB bridge[140000] at 5E94, size = FB0 13:769 0:004 USB bridge[1A0000] at 5AD7, size = 3BB 13:770 0:001 Device definition before adr=16F not found 13:773 0:003 found slave _DSM, ignore 13:776 0:002 USB bridge[1D0000] at 554D, size = 588 13:779 0:002 Device definition before adr=23A not found 13:781 0:002 found slave _DSM, ignore 13:784 0:002 Start _WAK Return Fix 13:787 0:003 len = 0x0000B7AF 13:790 0:002 _WAK Method find return data, don't need to patch. 13:792 0:002 return len=B7AF 13:794 0:002 deleting device CRT_ 13:798 0:004 deleting device DVI_ 13:800 0:001 deleting device SPKR 13:802 0:002 deleting device ECP_ 13:805 0:002 deleting device LPT_ 13:807 0:002 deleting device FDC0 13:809 0:002 deleting device ECP1 13:811 0:002 deleting device LPT1 13:814 0:002 Start PNLF Fix 13:816 0:002 found PWRB at 4E89 13:819 0:002 Start _S3D Fix 13:821 0:002 Start ADP1 fix 13:823 0:002 no device(AC) exists 13:826 0:002 Start Darwin Fix 13:828 0:002 Name _OSI present at 0x172D, renaming to OOSI 13:831 0:003 Name _OSI present at 0x8F90, renaming to OOSI 13:833 0:001 Name _OSI present at 0x8F97, renaming to OOSI 13:836 0:003 Name _OSI present at 0x8FAC, renaming to OOSI 13:838 0:002 Name _OSI present at 0x8FC8, renaming to OOSI 13:840 0:002 Name _OSI present at 0x8FE8, renaming to OOSI 13:843 0:002 Name _OSI present at 0x9008, renaming to OOSI 13:845 0:002 Name _OSI present at 0x9026, renaming to OOSI 13:847 0:002 Name _OSI present at 0x9042, renaming to OOSI 13:851 0:003 Name _OSI present at 0x905E, renaming to OOSI 13:852 0:001 Name _OSI present at 0x96DA, renaming to OOSI 13:855 0:002 Start SHUTDOWN Fix len=B884 13:857 0:002 ========= Auto patch DSDT Finished ========
  4. Clover General discussion

    Is there a chance we can have Clover automatically add a "hda-gfx" entry in GPU0 so we can get HDMI audio to work with an auto-patched DSDT?
  5. Realtek ALC AppleHDA Audio Injection

    Okay, I got it working. I simply thought this would work with a vanilla AppleHDA.kext, but patching is still required. I guess Clover's binary kext patching method is also of no use here, as you need to replace entire files within the kext Just two suggestions for your patching script: audio_alc892-91_patch.command is not executable your patch scripts assume that the files were downloaded to ~/Downloads/audio_ALC892-master/. Maybe some dirname magic or relative paths would be better, as they would allow other download locations Thanks for your efforts. - mcsmart
  6. Realtek ALC AppleHDA Audio Injection

    I've sent you the IOReg through pm. What do you mean by "patched AppleHDA.kext"? I am using Clover and I followed the instructions in [Guide]-Add_HDEF-Clover.pdf. In there it says: "no kext, dsdt or ssdt enabler required". This why I am using the vanilla AppleHDA.kext, but after reading through your first post carefully I figured that I still need to patch the kext. {censored}, I forgot. I am gonna try that ASAP. You probably do not need the IOReg dump then, as layout injection is already working...
  7. Realtek ALC AppleHDA Audio Injection

    Hi toleda, thanks for your work. I'd like to report a problem with audio injection on my newest build (see my signature for basic HW specs). Description of audio problem: No audio device recognized/usable, although AppleHDA is loaded. System Profiler shows the correct (injected) layout. As it says in your guide I am not using any custom AppleHDA.kext or enabler. OS X version/motherboard model/BIOS version/processor/graphics OS X 10.9.0 (vanilla install), GA-H87M-D3H with ALC892, Radeon HD 7850 Procedure/Guide Used/AppleHDA.kext version I tried the following methods: Clover with Audio Inject=1 and DSDT/FIX_HDA=False Clover with Audio Inject=3 and DSDT/FIX_HDA=False Clover without Audio Inject and DSDT/FIX_HDA=True Extra/dsdt.aml (if installed): none Copy of IOReg: I'm going to send you this one through a PM, as it contains my serial number Console/All Messages/kernel Sound Assertions selected: 22.12.13 19:13:01,000 kernel[0]: Sound assertion in AppleHDADriver at line 1525 22.12.13 19:13:01,000 kernel[0]: Sound assertion in AppleHDADriver at line 2193 22.12.13 19:13:01,000 kernel[0]: Sound assertion in AppleHDADriver at line 207 22.12.13 19:13:01,000 kernel[0]: Sound assertion in AppleHDADriver at line 1084 Screenshots: attached Edit: Sorry, I cannot send you a PM, as your inbox is full clover-config.plist.zip
  8. [UEFIPatch] UEFI patching utility

    Alright, thanks for your help. Since the Mavericks kernel really does not work with a locked 0xE2 MSR (further reading) I guess I am going to try to swap my board.
  9. [UEFIPatch] UEFI patching utility

    Okay, I guessed that. I am already using Clover (with autopatching enabled) but the kernel (Darwin 13.0) only loads if my firmware is PMPatched. Without a PMPatched firmware my system reboots instantly either after boot.efi hands off control to the kernel (when using Clover) or when Chimera loads mach_kernel. That's kinda hard to debug, because I am not getting any backtraces... Could this reboot/reset be triggered by the firmware when it detects an invalid MSR access? Does PMPatch really only affect 0xE2's state? I would guess that 0xE2 should only affect AppleIntelCPUPowerManagement.kext and not the kernel itself (at least not in such an early state), but maybe this has changed with Mavericks? Thanks & schönes Wochenende
  10. [UEFIPatch] UEFI patching utility

    Oh nein, I guess I spoke too soon. While the patching tool itself does, what it's supposed to do, I experienced one major drawback: It completely messed up the UEFI firmware on my MSI-H87M-G43. Short story: I guess the NVRAM or (maybe?) other parts of the firmware might get corrupted Long story: The board worked just as normal for a few days with the PMPatched firmware After rebooting from Windows I noticed that my "Linux Boot Manager" entry was missing in the F11 boot selection menu I thought: not a big deal, let's add it again with efibootmgr on Linux, which I did. I rebooted again, but it still was not there. This is where the odd part starts. I entered the UEFI firmare menu through the <DEL> key and I was shocked: almost all of the firmware options (system time, options for on-board devices, boot options, factory reset, etc.) were missing. Oh Nein! I disconnected the power, let it sit there for a while and turned it back on. Now I could not even get into the UEFI firmware with <DEL>. It was just stuck at "entering BIOS...". Now I was stuck with a non-bootable system. {censored}! I did a CMOS reset (disconnect power, set CLR_CMOS jumper, let it sit there for a while, removed the jumper). This still did not help. After turning the thing back on I got a message "CMOS cleared, press <DEL> to enter setup". I pressed <DEL>. Nothing happened. {censored}! Sh*t! Oh Nein! I disconnected the power, removed the CMOS battery, set the CLR_CMOS and disconnected all SATA drives. I let it sit there for 30 minutes. I put back the CMOS battery, removed the CLR jumper, started up without any drives and I could finally get to the UEFI firmware again. All options were where they were supposed to be. I immediately flashed the unpatched BIOS. Everything was fine. Then I felt adventurous and flashed the patched BIOS again. I did a factory reset. Everything was working as expected, register 0xE2 was unlocked. Just after one day this sh*t started again (missing bootloader entries, corrupt firmware setup). I went through this torture one more time and managed to flash the official BIOS again, which is running for over a week now without any issues (except that 0xE2 is write-protected). I am sorry that I can not really provide any more technical details. But maybe you have an idea of what could have gone wrong here. Thanks for making PMPatch.
  11. OK, it turns out that Clover's PCI-ID injection does not work on the vanilla install media. I have created another install media with myHack and I can now at least boot into the installer GUI (with my Radeon HD 7850 - I could not get the Intel Graphics to work yet). The installer segfaults, but that's a different story -> at least the "stuck in text mode issue" is solved for now.
  12. Hi, I managed to install and configure Clover on my new build (see sig for specs). I've also created a vanilla USB install media from the App Store version of OSX Mavericks. The install media is loaded by Clover in UEFI mode and it boots without any kernel panics or warnings. The only issue is that the installer GUI is never displayed - it's stuck in text mode. The last thing I am seeing is: [iOBluetoothHCIController][start] -- completed Right now I am only using the iGUP of my i5-4670, but when installing a Radeon HD 7850 I don't get any further either. When I let it sit there in text mode for like 5 minutes the screen turns grey and the dreaded beachball appears (I can even move it with my mouse), but that is it. This is what I have tried so far: Just booting with "-v" -> no changes booting with "-v npci=0x2000" -> no changes booting with "-v npci=0x3000" -> I am getting a Kernel panic with somewthing related to AppleIntelFramebufferAzul.kext I have also tried many settings in Clovers config.plist: Tried GraphicsInject for Intel: true | false Set custom ig-platform-id: 0x0c260000 | 0x0c260005 Set FakeID for my ATI card (when I tried it with the ATI card) Can anyone help me with this? My Clover configuration: <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>Boot</key> <dict> <key>Timeout</key> <integer>5</integer> <key>Arguments</key> <string>-v slide=0</string> </dict> <key>KernelAndKextPatches</key> <dict> <key>Debug</key> <false/> <key>KernelCpu</key> <false/> <key>AsusAICPUPM</key> <false/> <key>AppleRTC</key> <false/> <key>KextsToPatch</key> <array> <dict> <key>Name</key> <string>AppleAHCIPort</string> <key>Comment</key> <string>External icons patch</string> <key>Find</key> <data>RXh0ZXJuYWw=</data> <key>Replace</key> <data>SW50ZXJuYWw=</data> </dict> </array> </dict> <key>ACPI</key> <dict> <key>HaltEnabler</key> <false/> <key>DSDT</key> <dict> <key>Debug</key> <false/> <key>Fixes</key> <dict> <key>AddDTGP_0001</key> <true/> <key>FixDisplay_0100</key> <true/> </dict> </dict> <key>DropTables</key> <array> <dict> <key>Signature</key> <string>DMAR</string> </dict> <dict> <key>Signature</key> <string>SSDT</string> <key>TableId</key> <string>CpuPm</string> </dict> </array> <key>SSDT</key> <dict> <key>Generate</key> <dict> <key>PStates</key> <true/> <key>CStates</key> <true/> </dict> </dict> </dict> <key>RtVariables</key> <dict> <key>MountEFI</key> <true/> <key>LogLineCount</key> <integer>3000</integer> <key>LogEveryBoot</key> <string>10</string> </dict> <key>GUI</key> <dict> <key>TextOnly</key> <false/> <key>Language</key> <string>en:0</string> <key>Theme</key> <string>metal</string> </dict> <key>Devices</key> <dict> <key>Inject</key> <false/> <key>Audio</key> <dict> <key>Inject</key> <string>No</string> </dict> <key>FakeID</key> <dict> <key>ATI</key> <string>0x68181002</string> <key>IntelGFX</key> <string>0x0</string> <key>NVidia</key> <string>0x0</string> <key>SATA</key> <string>0x0</string> <key>WIFI</key> <string>0x0</string> <key>LAN</key> <string>0x0</string> <key>XHCI</key> <string>0x0</string> </dict> <key>USB</key> <dict> <key>Inject</key> <true/> <key>FixOwnership</key> <true/> <key>AddClockID</key> <true/> </dict> </dict> <key>Graphics</key> <dict> <key>Inject</key> <dict> <key>Intel</key> <true/> <key>ATI</key> <false/> <key>NVidia</key> <false/> </dict> <key>ig-platform-id</key> <string>0x0c260000</string> </dict> </dict> </plist>
  13. [UEFIPatch] UEFI patching utility

    Success on MSI-H87M-G43 with BIOS 1.4! Now I can finally boot Mavericks through Clover! For the record, this is my patchlog: 3IMS_patched.140 PMPatch 0.5.13 PowerManagement modules not found. PowerMgmtDxe/PowerManagement2.efi module at 00AC4770 patched. AMI nest modules not found. Phoenix nest modules not found. CpuPei module at 00FB7080 not patched: Patch pattern not found. Output file generated. Oh, and this is the output of rdmsr 0xe2 on linux, before and after the BIOS patch: before: 1e008405 after: 1e000405 Thanks a lot for your work!
  14. Hello everyone. I just got my SL disc I ordered and wanted to install SL on my GA-EP45-DS4. I already have a working Leopard Vanilla Installation and I have a custom DSDT which also includes common SL fixes (Ethernet, CMOS Reset). I even have vanilla speedstep working on Leopard. So I copied the SL Install disc to a USB flash drive, put Chameleon 2.0-rc4 on it and successfully installed SL on my internal SATA HDD using this pen drive. I also installed Chameleon 2.0-rc4 on my internal SATA HDD and placed these kexts in the /E/E folder: fakesmc (newest release) IOAHCIBlockInjector EvoXReboot Disabler.kext (just to be safe) Now whenever I want to boot to save graphics mode (-v -x), I get the following messages: The system simply does not boot. Of course full graphics mode isn't working either - I am however able to get into single user mode. I also tried switching between 32 and 64 bit - no difference. Is there anything I am missing? Any cool DSDT-fix I need? Just give me a hint - I can figure out the rest by myself - mcsmart
  15. Hello everyone! After spending hours of reading, patching and asking around I thought it is time to share my experience. First of all: This is not an installation guide! There are plenty of EP45 guides out there and I only want to cover the following DS4-related things: DSDT Sound Native speedstepping Sleep and maybe other issues 1. Prerequisites: a vanilla installation of 10.5.8 (10.6 should also work, but untested) Chameleon 2.0-rc3 + PC EFI 10.5 installed some understanding of the basic system layout of OS X (or at least another UNIX-like system) some commandline experience If you do not have any of the above - stop right here! 2. BIOS Settings Important settings are: Limit CPUID Max. to 3: Disabled No-Execute Memory Protect: Enabled CPU EIST Function: Enabled CPU Enhanced Halt: Enabled C2/C2E: Enabled C4/C4E: Enabled SATA RAID/AHCI Mode: AHCI SATA Port 0-3 Native Mode: Enabled Azalia Codec: Auto LAN1: Enabled LAN2: Disabled Green LAN: Disabled HPET Support: Enabled HPET Mode: 64-bit 3. Kexts and files I am using fakesmc.kext OpenHaltRestart.kext LegacyHDA.kext IOAHCIBlockStorageInjector.kext (part of Chameleon) And before somebody asks: YES, these kexts are everything I need. They all go in /Extra/Extensions, so the original /S/L/E can stay untouched. To get my graphics card to work I simply needed to set GraphicsEnabler=y in Chameleon! 4. DSDT Patching Why? Because by patching the DSDT we can use all vanilla kexts that come with the operating system - no need to add device ids. You do not even need any PowerManagementDisabler any more! Before you start hacking your DSDT I suggest you check the following threads about it: http://www.insanelymac.com/forum/index.php?showtopic=133683 http://www.insanelymac.com/forum/index.php?showtopic=168014 BTW: This is how I started getting into DSDT pathcing: How_to_patch_DSDT_with_LAN_and_HDEF_functions_including_LegacyHDA.zip I have created a custom DSDT for this board, using the latest BIOS (F10b) and a QuadCore CPU. It has everything you need, including: HPET fix RTC fix TM + Network fix CMOS reset fix for Snow Leopard Complete sound pinconfig (ALC889A) specific to this board + HDEF fix AppleLPC loading using native kext How did I do it? I used fassl's DSDT patcher and made my own additions to the DSDT (Sound, Net, LPC, CMOS). DSDT for EP45-DS4 with F10b BIOS: dsdt_nopstate.zip 5. Sound Sound is an easy one after DSDT patching. Unfortunately you have to use an extra kext for it. Simply put the LegacyHDA.kext in your /Extra/Extensions folder. Working: Line out (Back) Line out/Headphone (Front) SPDIF out (Back) Microphone in (Back) Line in (Back) SPDIF in (Back) Headphone plug-in autodetection (if you are using an HD I/O frontpanel) ALC889A kext: LegacyHDA.kext.zip 6. Native speedstepping With the patched DSDT native speedstepping works almost out of the box. You simply need to add the P-states of your CPU. To get those I'd recommend temporarily loading VodooPower.kext. If speedstep is working for you with VodooPower then you can use the P-States Calculator to display the current P-State configuration that VodooPower uses. Just replace the values at the end of this DSDT with the values from P-State Calculator: DSDT with P-States for Core2Quad Q9650: dsdt_pstate_q9650.zip You really don't need to add any C-States to your DSDT on this board, because the EP45-DS4 natively supports them (with the BIOS settings from 2)! For reference I also recommend reading this thread. 7. Sleep Sleep is working pretty well on my DS4 with a Radeon HD 3870. In order to get sleep to work you need to open System Preferences -> Energy saver -> Options and check "Restart automatically after a power failure" (AppleLPC needs to be loaded to get this option, which should be the case if you are using my DSDT.) Appendix A: Sound Pinconfig Sound Pinconfig for this board. Note that in Windows I got a "0" instead of a "2" as the first character of each word. But in in the linux codec dump it says "2". 21471c10 21471d44 21471e01 21471f01 21571c12 21571d14 21571e01 21571f01 21671c11 21671d64 21671e01 21671f01 21771c14 21771d24 21771e01 21771f01 21871c40 21871d9c 21871ea1 21871f01 21971c50 21971d9c 21971ea1 21971f02 21a71c4f 21a71d34 21a71e81 21a71f01 21b71c20 21b71d4c 21b71e21 21b71f02 21c71cf0 21c71d01 21c71e33 21c71f59 21d71c01 21d71de6 21d71e05 21d71f40 21e71c30 21e71d61 21e71e4b 21e71f01 21f71c60 21f71d71 21f71ecb 21f71f01