Jump to content


  • Content count

  • Joined

  • Last visited

About landmark

  • Rank
    InsanelyMac Protégé
  1. Yes, I have 10.6.3 installed on my two internal HDDs using a software RAID0. This installation works flawlessly. The 10.6.4 one is on an external 2.5" USB drive, so I can easily try out different approaches. What doesn't go in my head is that the Intel 82574L is the LAN adapter that Apple itself uses in the Mac Pro - hence the Intel82574L.kext in the IONetworkingFamily.kext. My feeling is that it might have something to do with the release of the new unibody Mac Mini - it utilizes the 82574L, got released on the very same day as 10.6.4 and there are known compatibility issues. Apple even released, for the first time ever, a 10.6.4 update tailor-made for those devices which includes an Intel82574L.kext! However, both read version 2.1.0b1 (haven't compared them binary-wise) and overwriting the 10.6.4 standard kext with the one from the Mac Mini upgrade doesn't help either - both behave exactly the same. Moreover, tweaks specific to the Mac Mini shouldn't matter at all since my system's defined as a MacPro3,1 in SMBIOS.plist. I've attached those three Intel82574L.kexts - the one from 10.6.3 perfectly working, the 10.6.4 ones not. Maybe there's a smart mind out there to bring them to work... EDIT: The Driver Parameters section in the kexts Info.plist differs from 10.6.3 to 10.6.4. I've pasted those 10.6.3 settings into the 10.6.4 plist with no effort - misbehavior's exactly the same. Intel_82574L.kexts_from_10.6.3_and_10.6.4.7z.txt
  2. Well, I ruled out some possible reasons in the meantime and as you predicted, the bootloader isn't accountable at all. This is what has no influence on the error: 1) The bootloader. I checked RC4 r684, RC5 r193, r257 and r300. Regarding the ethernet issue, all behave exactly the same. 2) Having the Realtek onboard LAN enabled or disabled at BIOS setup - this surely results in en1 being present or not, but en0 (the i82574L one) remains unstable. The Realtek onboard LAN works fine, however. 3) dsdt.aml: Even renaming /E/dsdt.aml (and thus falling back to the dsdt embedded in the cartri BIOS) makes no difference. Pressing the Power button to enable sleep and waking up the system restores the power-on condition: Network works for pinging, but fails even under the slightest load. Very dissatisfactory...
  3. I'm experiencing some weird networking issues with my Intel 82574L card since the update to 10.6.4 - anyone having similar problems? I checked against both Chameleon RC5 r193 and r257, Speedstep enabled and disabled - same issue everytime. Any hints? Thanks in advance!
  4. Hi there, I'm experiencing some weird behaviour with a Snow Leopard installation from scratch; my source is a vanilla 10.6.3 build 10D575 retail DVD. The system works perfectly with 10.6.3, the only thing I have to edit to make my Intel 82574L work is the VID:PID as explained inter alia here. However after the update to 10.6.4 (and of course having patched the Info.plist again), the madness begins. The driver gets loaded properly: The system aquires an IP via DHCP, I can ping internal and external IPs, DNS resolution works properly, too. However, as soon as the NIC gets utilized noteworthy (firing up Apple Safari or the Software update is sufficient), the connection fails with no hint at all in the logs. OS X claims to be still connected (green icon in network preferences), but any attempt to connect anything except for fails. When I try to renew the DHCP adress at this point of time, OS X assigns itself an internal IP. The bottom line: 82574L networking appears to work with 10.6.4, but fails with the slightest load. Any hints? Thanks in advance! EDIT: Some more information, including download for the working 10.6.3 and the non-working 10.6.4 kexts, here.
  5. Here's another minor issue: My GA-EP43-DS3L based build accesses its DVD+-R drive (TSSTcorp CDDVDW SH-S223F) without any reason. The drive is a tray one, no CD or DVD is inserted and there's no application running except for the finder. Despite of that, every few minutes the DVD drive access LED lits and I can hear the laser pickup carriage move. This doesn't happen with Windows 7. I don't want the DVD drive lifespan being degraded by the unnecessary accesses. Any hints?
  6. @landmark - did titan4's suggestion of using MacPro3,1 in SMBIOS help with your audio issue? I haven't used the digital audio output on my hack so haven't been able to test that. That did it in fact. The sound output stays on now. Thanks a lot, titan4!
  7. Yes. I have it too. In my case, there's a 5.1 surround system connected via TOSLINK (e. g. digital optical out). The problem is that OS X shuts down the output after a minute or so without sound. In the TOSLINK case, it even deactivates the transmitter. You can see that the red light goes off. What's worse is that my surround system has an auto shut down if no signal is received for five minutes. That means a noisy relay dropout, and of course the first two seconds of the next audio output get muted due to the amplifier restart and the speaker protection circuit. Overall, the behavior is VERY annoying and I'd REALLY appreciate a hack to disable the output shutdown completely.
  8. Here you are! I assume that the RAID device has already been created by the diskutil GUI. Afterwards do the following in a terminal session: At first, become root: sudo sh Find out diskX from which the BIOS actually boots: diskutil list Install MBR code: ./fdisk440 -f boot0 -u -y /dev/rdiskX Install further boot stages: dd if=boot1h of=/dev/rdiskXs3 diskutil mount diskXs3 cp boot /Volumes/Boot\ OSX/ Copy the usual needed stuff to /Volumes/Boot\ OSX/Extra /Volumes/Boot\ OSX/Extra/Extensions OS X creates a virtual RAID disk Y. Find out its number via… diskutil list … and grep its volume UUID: diskutil info diskY | grep "Volume UUID" Now instruct the bootloader to boot exactly that volume: nano /Volumes/Boot\ OSX/Extra/com.apple.Boot.plist Add to Kernel Flags: rd=uuid boot-uuid=[UUID from above] Save with CTRL+X, Y and return. Now unmount the raid helper partition: diskutil unmount diskXs3 Done! @landmark - Good question and to be honest I don't know the answer to extracting the Intel82574L.kext from IONetworkingFamily.kext. For this install, like you want, I have kept the retail installation completely untouched. So you could try extracting the Intel82574L.kext and putting it in to /E/E and see if it works, but it would be simpler to just copy the revised IONetworkingFamily.kext to /E/E and see if it loads instead of the one in /S/L/E. If the modified one doesn't load then you can try changing the CFBundleVersion key/string in the info.plist of IONetworkingFamily.kext as I think OS X will choose to load a newer version kext over an older one. You could also try trashing your /System/Library/Caches folder. blackosx, I owe you some observation results: It's actually perfectly sufficient to only copy the Intel82574L.kext to /E/E/, modify its Info.plist to ones needs and to boot with '-f -v' once afterwards. Neither IONetworkingFamily.kext nor the comprised Intel82574L.kext on the OS X volume have to be changed. This, however, applies to Snow Leopard. On Snow Leopard Server, I do experience a weird issue: The NIC gets recognized only with the boot parameters '-f -v' on every boot. To be honest, I haven't got the faintest idea why this happens. I also tried to delete /System/Library/Caches/com.apple.kext.caches/Startup/Extensions.mkext with no effect at all. EDIT: I was wrong. Both Snow Leopard and SL Server recognise the kext in /E/E/ only when booted with -f -v. Deleting or renaming /System/Library/Extensions/IONetworkingFamily.kext does resolve the issue. But that's not exactly the plain vanilla approach. I'll check out the CFBundleVersion suggestion.
  9. I don't use the onboard NIC because of speed issues. Instead, there's a PCI-E Intel Pro/1000 installed. It is running nearly out of the box; the only thing to be changed was PID:VID in /System/Library/Extensions/IONetworkingFamily.kext/Contents/PlugIns/Intel82574L.kext/Contents/Info.plist. I'd like to keep my OS X partition as vanilla as possible - because of that I want to move that modification to the Cham partition. However, I'm not familiar with those nested kexts. What exactly do I have to copy to /Volumes/Cham/Extra/Extensions? Only Intel82574L.kext or the whole IONetworkingFamily.kext? What about the ones at /S/L/E - do I have to delete them afterwards or do the ones at /Volumes/Cham/E/E supersede them automatically? Thanks for some wise words!
  10. I came to different findings. Resume from sleep is broken in my vanilla 10.6.2 installation; the corresponding entry on the Bluetooth settings page is greyed out. Also, the DBT-120 continues to work even with Windows after the firmware update - all you need is to tweak it into the HCI mode. You might want to read the in-depth explanation and download the Windows utility from this post.
  11. Hey antipop, sorry, I took some time to check your suggestion out just to be sure that I'm not doing it wrong. I'm sorry say - still no luck for me. Even with pairing the devices via the keyboard and mouse panels in the OS X settings, only my keyboard remains paired in Windows 7. I did recognize something interesting nevertheless: When I boot up in Windows 7 and the mouse isn't in its list of known devices yet, pressing the mouse button leads to an incoming connection request from the mouse. Yes, the mouse tries to establish a connection to the computer! It also requests a pairing code (0000 works) what it doesn't do when the connection is initiated from the computers side. However, the result remains the same: As soon as the mouse works in Windows 7, it gets unpaired in OS X and vice versa.
  12. The most recent firmware (Apple Firmware Update 1.2) for the DBT-120 enables a so called HID mode in which the Bluetooth keyboard and mouse are presented as plain USB devices. Any PC with 'legacy support for USB devices' can use them at BIOS level and even in applications that are not aware of bluetooth input devices at all, e. g. Acronis boot CDs. As soon as OS X has finished its boot, it switches the DBT-120 from the aforementioned HID mode into the known HCI mode in which the keyboard and mouse are seen as what they really are - Bluetooth devices. Windows doesn't know about HID and HCI - you have to use a 3rd party tool there called HCItray (remove the .txt suffix from the attached file, unpack with 7z) to let Windows see the Bluetooth device at all. BluetoothFWUpdate1.2.dmg HciTray_Setup.7z.txt
  13. I've got a pairing issue between Snow Leopard and Windows 7. My setup is a GA-EP43-DS3L based Hackintosh with a D-Link DBT-120 hardware revision B4 attached. The DBT-120 has been upgraded to firmware 1586 already. An Apple Wireless keyboard (2009 model) and an Apple Magic Mouse are attached. Both keyboard and mouse are working flawlessly either with Windows 7 (generic Microsoft Bluetooth stack, Boot Camp 3.1 keyboard and mouse drivers installed) or Snow Leopard (10.6.2). I can even use them at BIOS level thanks to the 1586 firmware HID mode. However, everytime when I start Windows after having used OS X and vice versa, I have to repair the mouse - ONLY the mouse! OS X claims that there's no mouse attached. I have to switch it off and on again, let it get discovered and paired. In Windows, I have to delete the mouse from the list of known devices, also switch it off and on again and pair it again. The keyboard however doesn't need to be re-paired and is available in both installations flawlessly. How do I manage to get my Magic Mouse permanently paired with OS X AND Windows 7? Thanks in advance for any hints!
  14. defaults write com.insanelymac.forum.users.lord_muad_dib output verbose
  15. Hi there, though I'm using Titan for quite a long time now without any problems, I thought I'd give Natit a try. Using DiaboliKs installer, I found out that it (at least) offers me rotation support which Titan doesn't. It seems to work fine, too. However, there's an issue: There's no video output at all for 25 seconds while booting up the system. I do see the grey screen with the spinning circle as usual and in the very moment when it's supposed to switch to the blue screen, my monitor reports "no signal" and gets to sleep. Exactly 25 seconds later, it awakes and the boot process continues. Any hints on this? This is what the system profiler says about my graphics (however with Titan installed): NVIDIA NV43: Chipsatz-Modell: NVIDIA NV43 Typ: Monitor Bus: PCIe VRAM (gesamt): 128 MB Hersteller: NVIDIA (0x10de) Geräte-ID: 0x0147 Versions-ID: 0x00a2 Monitore: NEC LCD71VM: Auflösung: 1280 x 1024 @ 75 Hz Farbtiefe: 32-Bit Farbe Core Image: Unterstützt Hauptmonitor: Ja Synchronisierung: Aus Eingeschaltet: Ja Quartz Extreme: Unterstützt Monitor: Status: Kein Monitor angeschlossen Thanks in advance!