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.

Mr T

  • Content count

  • Joined

  • Last visited

About Mr T

  • Rank
    InsanelyMac Protégé
  1. I tried both the previous and your most recent shell script with the ESD image mounted -- seems to be having some trouble. Also tried with & without the ATY_init kext (even though I need that myself, I saw that it had worked for someone above). I've attached a screen shot of my console output for reference. -- T
  2. A Simple Guide to the 4830/4850/4870/4870x2/4890

    The new Natit kext has got my 4870 working with 2 monitors! Thank you so much! Sleep quit working and I now have some new "AppleHDA Sound assertion" errors in my logs; but everything else appears fine. The only real problem I notice is that I'm dropping a fair number of frames when activating Expose' with my second 24" monitor attached; (Spaces and other window effects don't appear to be affected). Seems on par with my old Nvidia 7100GS, but far behind my 8600GT; so I assume this is just a driver/timings quirk for now. I'll be sure to ask on IRC if there is anything that can be done; but for the moment, this is great! A million thanks once again! Side note: The previous Natit would allow me to launch Leopard with one or the other, but resulted in a blue screen with both. Specs: Video card: Diamond 4870 1GB w/ reference cooler Proc: i7 920 Mobo: GA-EX58-UD5 Kernel: tested and working fine in 10.5.6 with both Andy's 9.6 and the new vanilla 9.7 beta kernel. One other thing to note, Expose performance seems smoother with the new kernel, and the window drop shadows don't vanish when the effect is activated; however, it still appears choppy compared to my 8600GT.
  3. If you have a 9-series Nvidia 92/92b-based video card and are experiencing strange issues regarding lag when using 2+ monitors with your UD5 board, you're not alone. Below is a drawn-out story of my post-install board-related problems as well as a (hopefully temporary and rather ghetto) solution. ---------------------------- I've been trying to get a 2-card/4-monitor setup working for the last 2 weeks on an EFI install of Leopard Server 10.5.6 (installed using Wolfenuke's method & some of DD's recommended kexts). After banging my head against the wall, I have finally (partially) succeeded, but it has been a royal PITA. My card of choice for the setup are the Gigabyte 9800GTX+ 1GB series. I performed the install on my SSD drive using an already-working hack. After double and triple checking the install, I transplanted the SSD into the new machine and fired it up. Just about everything worked perfectly the very first boot, including QE/CI. I then added the 9-series_v2 drivers (i.e. the modified MBP driver pkg, hat tip DD for posting the link). Through testing with xbench and opengl extension viewer, I found the pkg seems to significantly improve graphic performance across the board for 9-series cards as mentioned earlier in this thread, (hat tip JaE-V for the advice). But, when I connected a second monitor to the first 9800gtx+ card (before even worrying about getting the second card working), I ran into my first problem. With 2 monitors connected, I began to drop frames with Expose, Spaces, and when activating the Dashboard. With 15+ windows open, Expose and Spaces would slow to a crawl. About 1/3rd of the time, the animation would simply be dropped all together. This seems to be a common problem with real Macs too with the DVI link, and the recent QT7.6 and 10.5.6 updates. So far, Apple has been unable to provide a working solution. [Please note that my problem does not involve the QT 7.6 upgrade - I disabled the AppleUpstreamUserClient.kext during the initial install process (after running into the QT laggy-mouse issue on my other hacks).] To test the hardware, I installed one of the 9800gtx+ cards in my fully-working vanilla q6600 hack. I experienced no Expose lag with the card in the q6600 machine, which was puzzling. I decided to move on to getting the dual graphics cards working, before I lost my mind with trying to figure out the lag issue. After reading Aquamac's extremely thorough tutorial on getting two cards working using GFXutil to single out the cards' individual PCI addresses, I built a custom EFI string and attempted to fire-up both cards at runtime per his instruction. (Thanks again to JaE-V for providing the link in this thread!) Unfortunately, this process seemed to have no effect (while it appeared to be quite a reliable trick according to the majority of the replies on his forum). When I tested this same method on my q6600, it worked the very first time and both cards were immediately recognized w/ QE/CI. I then started to think the problem with Expose/Spaces/Dashboard and the dual-card issue might be related. Going back to the lag issue, I tried the 7100GS card from my q6600 on the i7/UD5. Once I corrected the EFI string, I found Expose ran almost flawlessly! Seeing this gave me hope. I popped the 9800gtx+ back in and went back to reading / testing... and reading / testing... and even f*ck-tons more reading / testing... I decided at first it must be a simple driver problem. During this time, I tried: - Voodoo 9.5 and 9.6 (Andy's) Kernel - 10.5.2 gfx kexts and injector (various versions) - 10.5.3 gfx kexts and injector (" ") - 10.5.4 gfx kexts and injector (" ") - 10.5.5 kexts (various revisions) + EFI string (w/ updated device ids) - 10.5.6 kexts + 10 different EFI strings - Recalled 2009 Apple Graphics Update kexts (w/ updated device ids) - Various versions of QT Not one single thing seemed to help. Not to mention, mounting and unmounting the EFI partition; editing and swapping out various kexts; and rebooting over and over was no fun at all. At this point, I figured I could rule out problems with graphics driver and possible OS/driver version conflicts/incompatibilities. The only thing I hadn't screwed with yet was the video card bios. Starting with cyclonefr's 8800 bios flashing guide, I was disappointed when I could not get any OS to boot with his available 8800GT ROMs. I read further on the mac rom and supplemented my research with dozens of other sources, including helpful irc folks, local overclocking gurus and mac/bsd/linux forums. Starting with a copy of my 9800gtx+ rom, the revision 1 Mac 8800GT rom, and a rom from a GTS250 engineering sample; I managed to cobble together one hell of a video frankenBIOS. I used the 8800GT rom as the base rom, and the EFI portion was removed so it would fit on my 9800gtx's tiny bios chip. I then started pulling clocks, voltages, memory timings, and other information from the 9800gtx+ and GTS250 roms. I was able to use the GUI in NiBiTor (running in XP32) for most of the edits, but there were a couple that had to be done directly to the hex (thanks DarwinFreak #BSD / Undernet2). In the end, I ended up with a video card that works in XP32, Vista64, and Mac OS, showing up in XP and Vista as a GTS250 1GB, and in Mac OS as an 8800GT 1GB. Furthermore, my OpenGL scores are within 5% of their highest speeds. And best of all... the Expose/Spaces/Dashboard LAG IS GONE! Upon further examination, I now have access to two video cards using AquaMacs method! (Well, one 9800GTX+ and my 7100GS tested so far.) Lastly, I was able to re-enable the disabled AppleUpstream...kext as well. I'm still not sure why this is happened or really even exactly what I did that fixed it , but I'm guessing it has something to do with an incompatibility in the 9000-series drivers, some core OS files, and the Voodoo kernel itself. Perhaps the x58 hardware as well. I did notice a quantifiable improvement in graphic performance with the recalled 2009 Apple Graphics update; it's almost like Apple had the problem half-way solved. I ended up with 4 different Mac 8800GT factory BIOS files, so I believe minor differences in revisions are playing a part in the similar problems some real Mac owners are complaining about recently. I've read hundreds of pages and I feel I haven't even scratched the surface, but for now, I'm satisfied. Too bad the solution required such a complicated and somewhat dangerous fix. I just noticed the 10.5.7 beta update is now available on Apple's dev site. I'll try to download it tomorrow and flash my card back to stock to retest. Hopefully Apple will have solved the problem; but if not, this is at least a semi-viable solution for now. BTW - Sorry for the long write up. This laggy BS has been bugging me for a while! I'm drinking a glass of Crown XR as I type this, happy that my install is finally usable. To DD, Wolfenuke and all the other OSXperts -- if I missed an easier solution for this strange problem, please share! Best regards, Mr. T P.S. I'm reluctant to post my personal BIOS file, since near the end of the process, I bricked one of the 9800gtx+ cards. I'm not sure if it was my handy work that caused it or if I had simply reflashed the EPROM one too many times. However, I'm happy to provide the bios source files and reference sources to anyone experiencing the same issue. Processor / OS: i7 920 @ 3.4 GHz (F5 BIOS slowdown, ugh...) running Mac OS X 10.5.6 Leopard Server Motherboard / ram: GA-EX58-UD5 w/ 12GB OCZ Platinum 1600 MHz DDR3 (@ 1400 MHz) Main system disk: 60GB OCZ Solid State MLC Disk Storage disks: 2x 150GB Raptor 10k, 4x 1.5TB SATAII 7.2k Graphics: 2x Gigabyte 9800GTX+ 1GB DDR3 (well, just 1 for now until the RMA fairy visits me)
  4. Thank you gentleman for putting so much time and hard work into this process. Many folks, including myself, have benefited greatly from your efforts. Using the resources available in this post, I was able to get my machine setup in less than 2 hours with minimal hassle. My only remaining question is has anyone managed to get multiple video cards working on the UD5? I'm using EFI strings for two identical 9800GTX+ cards, (located at Pci(0x3) & Pci(0x7)), yet only one card works at a time depending on which is selected to boot with in the BIOS. Does anyone have any thoughts or experience with this? I can post my string info in xml format if that would help. Thanks for your time, folks! Best regards, Tyler Processor / OS: i7 920 @ 4.04Ghz (Mac OS X 10.5.6 Leopard Server) Motherboard / ram: GA-EX58-UD5 w/ 12GB OCZ Platinum 1600Mhz DDR3 Main system disk: 60GB OCZ Solid State MLC Disk Storage disks: 2x 150GB Raptor 10k RPM, 4x 1.5TB SATAII 7.2k RPM Graphics: 2x Gigabyte 9800GTX+ 1GB DDR3
  5. nForce LAN driver

    I have the very same issue and I remain puzzled regarding a solution. I tried the 60-second reset method as well with no luck. My ethernet card is visible to the OS and it can see my network, but for some reason I can't access the internet. Version: Mac OS X v10.5.6 Distro: iPC w/ PP2 patch Kernel: Voodoo 9.5.5 Notes: added Slashack's AppleNForceATA.kext System Info: Processor: Intel Core2 Quad Q6600 @ 2.4Ghz Mainboard: Asus P5N-D SLI 750i Other: 4GB DDR2-800, 320GB HD (SATA), LiteOn DVD-RW (IDE), GeForce 7100GS 256MB With the NForceEthernet.kext included in the iPC installer, I get full speed network transfer and reliable internet service for about 4 hours at a time. Once the network connection quits, the rest of the system becomes unstable. Unfortunately, with the NForceEthernet kext I receive a kernel panic everytime I try to access the internet without the cpus=1 boot flag set. I'd like to offer a quick word of thanks to Eno, as well as other OSX86 scene contributors, and not to forget the InsanelyMac forum staff. Your hard work is very much appreciated! After months of lurking, this is the first time I've ever had the need to post a question, simply due to the shear volume of information contained here. Truly amazing! Best regards, Tyler *FIXED* Throughout the installation and setup process, I've been connected directly to my cable modem (only because I thought it might limit the number sources for possible trouble). Adding my router into the loop instantly solved the connection issue and a correct local IP was assigned instantly. At the moment, everything works! (I've really been dying to say that.) Eno, you are the man! Do you have a PayPal donation link?