Jump to content

HOW TO: Get the Sapphire X1950XT(id7244) and X1900XT (id7249) to work NOT THE PRO (Jccool Method)


jccool
 Share

804 posts in this topic

Recommended Posts

The important component in Jccool package is Natit.kext so you will need to remove that one also.

 

in terminal as sudo:

rm -R /System/Library/Extensions/Natit.kext

 

when i put that in terminal i get this:

 

override rwxr-xr-x root/wheel for /System/Library/Extensions/Natit.kext/Contents/Info.plist?

 

not exactely sure what to do...i looked in the links you gave me, and i didnt see anything to do with rwxr-xr-x...

 

or just drag the Natit.kext to trash, and type your pw when prompted.

 

to remove the ATI files like jnolla posted you need to type

cd /System/Library/Extensions/

type this to tag for cache rebuild on next boot

sudo touch /System/Library/Extensions

Link to comment
Share on other sites

Hi Guys,

 

I need help. I'm using ASROCK 939 Dual Sata II and my graphic card is extactly the same card as Jccool's test card: Sapphire X1950XT(id7244).

 

Now the weird thing is, whenever I use Jccool's driver package, I lost display right after the end of boot sequence and before entering the GUI. Let me explain this little more: I boot in Verbose mode. I can see all the boot info. Right at the point that login window started and GUI suppose to kick in, the screen stays at boot display. In another word, it can not go into GUI. I can boot into GUI using generic driver (no driver selected) or other ati driver. Some ATI driver can give me CI and QE support and resolution change, but system profiler can not recognize my graphic card, just say unknown ATI device. Generic driver obviously doesn't support CI and QE nor resolution change.

 

What is happening? Does Jccool method have a known conflict with this ASROCK board?

 

Help please!!! Thanks.

Link to comment
Share on other sites

It's very strange that I have to use Jccool package together with natitx1800.kext to get display working. Jccool package alone won't work. No signal to monitor in DUI mode.

 

After I installed Jccool's package, when I boot up in verbose mode, I noticed that my device id shows 7264. But in my system profiler it's actually 7244. What's wrong? Maybe this is the reason that Jccool package didn't work for me?

 

And I'm very certain my card is Sapphire X1950xt 256 mb PCIE.

Link to comment
Share on other sites

whoa FINALLY...after months i got a fully working x1950xt. yay! thanks to wigworm for the idea. i was having the same problem as him. jccool method alone would not work for me. and the method that he linked to didnt work either, but in a PM he suggested that i boot up from v1.3 uphuck and install ALL of the ATI's drivers and reboot. weird, but that worked. now to test this sucker out some more.

 

BTW would ati accelerator work?

Link to comment
Share on other sites

Hello,

 

could one of you 2 tell me which plug in you used so I can creat an alternate build for ppl like you 2,

 

thanks,

 

im not really sure, although wigworm thinks it might be the natit x1800.kext.

Link to comment
Share on other sites

Ok got some great news. I've got the ATI Display Utilities fully supported on the X1950XT.picture2ik4.pngHere are the instructions:Download the following:http://ati.amd.com/support/drivers/mac/Mac...0G5Updater.htmlOpen Terminal
sudo su

Make a backup of sysctl

cp /usr/sbin/sysctl /usr/sbin/sysctl.real

Remove sysctl

rm /usr/sbin/sysctl

Create a new sysctl file

nano /usr/sbin/sysctl

Copy and paste the code bellow:

#!/bin/shecho "PowerMac11,2"

Hit Control + XAnd enter to save the new fileMake it executable:

$ sudo chmod +x /usr/sbin/sysctl

Run the installer just like anyone would. Do not restart your computer before completeing the next step.Put the real sysctl back.

mv /usr/sbin/sysctl.real /usr/sbin/sysctl

Everything should Just Work. You’ll now have an ATI Displays preference pane where you can tweak all of the “secret” options.Now I'm still trying to get the AVIVO (HDTV/S-VIDEO) out to work! I've found that the kext needed for this to work is the following:ATITVOut.kextThe video out does work on Darwin, and it actually lets you work with the RCA out a Darwin's boot prompt, but once the kext's are loaded, I loose the image.Best Regards

THANK YOU, now that my card works, i did this, and BAM, works flawlessly! thanks again!
Hi,Do you just use my method and use his ATINDRV?
actually at first i tried the atindrv but it didnt do nothing, so i just did a blanket install of all the ati (except callisto) drivers that were in the v1.3 install dvd.
Link to comment
Share on other sites

Hey guys, great work all around! I have an ASUS x1900xt and I'm having an interesting situation on my hands.

 

jccool your driver is awesome, when it's working for me my system flies. However, more often than not when I boot up/reboot whatever, I get stuck at 'loading apple.login.window' or whatever. I boot in Verbose mode since I have a dual opteron rig. I've tried to figure out a work around, originally it seemed if I pulled the power cord it would work right away, lately it seems if I make some type of simple hardware change like removing or inserting my USB bluetooth adapter. It's almost like the old days when Plug and Pray first came out and you had a resource conflict and if you changed something it would clear up.

 

Nevertheless eventually, sometimes after several attempts, sometimes after just one reboot it'll boot properly. I can tell when it's going to boot properly or not, after it loads the login.window app it'll either sit there quietly if it's not going to work or the screen will flicker and the fan on my x1900xt will whir up a second and then it'll go into the GUI.

 

I have a spare GeForce 6600GT and the problem doesn't exist with that card.

 

Anyone else have this problem or have any ideas on what to do to fix it?

 

I do notice while booting it seems to be trying different device IDs and it'll say initially something like family match but then near the bottom it'll say something mismatch. I'll have to keep a closer eye between when it does or doesn't work when I get home tonight.

 

Maybe if I delete all the dev id's other than mine it'll help? Any ideas anyone? If I could just solve this problem it'll be a huge step in demoting XP from primary to secondary OS.

 

EDIT: Just realized my card is an XTX, my head is still spinning from reading through the entire thread so I don't remember what the status of XTX support is, someone help me out? I mean obviously it works on some level since it works on my machine for the most part... HELP

 

EDIT 2: Ok so apparently you need to actually have a proper monitor setup on both connections. Before I tried just connecting it to my KVM but not actually setting it up, this time I actually fully set it up and mirrored it, ran it via DVI-VGA adapter to the VGA on my monitor (the other to the DVI) and so far so good! Keeping fingers crossed!

Link to comment
Share on other sites

Hi Jccool, Here's what I gathered so far after experiementing with this whole issue and reading on this board. Using your method get system recognize the device during boot, but it seems that if I only have one dvi connected, that connection was always put to sleep right around login window GUI started, and the actual display signal sent to the empty dvi port. Therefore, it appear that system hang at login window app. but in fact it didn't hang, it just that display signal is not coming to the monitor. To resolve this issue without having two monitor hooked up at the same time, I have to install Diabolik's NatitX1800.kext. Now with this kext, the catch is that in system profiler, it says: unknown ATI (doesn't recognize it X1950). I've been trying to see what's in that kext and try to modify it, with no success.Why does your method works for some but not others? Here's my theory (it's only a theory). Some motherboards have more than 2 display port (AGP/PCIe). I think issues mostly occur on this kind of MOBO. I have a ASROCK 939 Dual SATA board, which has both AGP port and PCIe port. drho2004 has been experiencing similar problem, and his mobo has a integrated AGP graphic chip and also a PCIe port in which he pluged his X1950 XT in. For some reason, when there are 2 port present on motherboard, the device ID got confusing. So when OS X boot up, it's been trying assigning different device ID to the same card.

Hi,Do you just use my method and use his ATINDRV?
ATINDRV with the plugins doesn't work either. It's the NatitX1800.kext got it working with full CI/QE/Rotation support. But like I said, the catch is that system profiler identifies it as unknown ATI. If you can get it recognize the device ID, that would be perfect!
Link to comment
Share on other sites

Hi Jccool, Here's what I gathered so far after experiementing with this whole issue and reading on this board. Using your method get system recognize the device during boot, but it seems that if I only have one dvi connected, that connection was always put to sleep right around login window GUI started, and the actual display signal sent to the empty dvi port. Therefore, it appear that system hang at login window app. but in fact it didn't hang, it just that display signal is not coming to the monitor. To resolve this issue without having two monitor hooked up at the same time, I have to install Diabolik's NatitX1800.kext. Now with this kext, the catch is that in system profiler, it says: unknown ATI (doesn't recognize it X1950). I've been trying to see what's in that kext and try to modify it, with no success.Why does your method works for some but not others? Here's my theory (it's only a theory). Some motherboards have more than 2 display port (AGP/PCIe). I think issues mostly occur on this kind of MOBO. I have a ASROCK 939 Dual SATA board, which has both AGP port and PCIe port. drho2004 has been experiencing similar problem, and his mobo has a integrated AGP graphic chip and also a PCIe port in which he pluged his X1950 XT in. For some reason, when there are 2 port present on motherboard, the device ID got confusing. So when OS X boot up, it's been trying assigning different device ID to the same card.ATINDRV with the plugins doesn't work either. It's the NatitX1800.kext got it working with full CI/QE/Rotation support. But like I said, the catch is that system profiler identifies it as unknown ATI. If you can get it recognize the device ID, that would be perfect!

 

To clarify so we can troubleshoot all the bugs/caveats, while my post may lead one to think otherwise, this wasn't the case for me. I could tell when it would boot properly because my X1900XTX would whir up it's fan briefly (and at full speed, so loudly). When it didn't work, it didn't spin up the fan. I did think the same thing you did, and when it wouldn't work I'd switch my monitor to the other DVI port but that didn't work.

 

I more carefully watched the boot sequence with -v on last night and noticed that the card seems to have multiple DevIDs, apparently one for each DVI port. I realized that when I connected the second port via VGA adapter to the VGA port on my monitor, switched my KVM so it would take input from it, it would no longer say mismatch for one of the DevID, it became clear that it was looking for another display when it wasn't working and if it didn't find it it would not spin up. If this is the case, I have NO idea how or why it would work with just one display the times it did and more importantly why it didn't always work with just one display.

 

Also, my machine definitely did NOT immediately go into sleep mode because I could tell it was continuing to boot into the OS by the HD activity and once the activity stopped, if I hit the power button (which normally puts the computer into sleep mode) it would then go into sleep mode. I could also tell because OS X would still play sounds etc.

 

As far as my issues, I *think* plugging something "real" into the second port has solved the problem. After I connected it properly, each time I reboot.... er shutdown then booted (can't reboot my machine, NForce4 w/USB = can't reboot?) it didn't hang at the login app. (Knock on wood....)

 

It's a bit annoying but thankfully the other computer on the KVM is my server and I don't need to access that too often and I can always get to it remotely from my workstation.

 

I think, since JCCool managed to get both ports working this is the tradeoff. Obviously an official driver would probably not have this issue, but we're far from official! :(

Link to comment
Share on other sites

To clarify so we can troubleshoot all the bugs/caveats, while my post may lead one to think otherwise, this wasn't the case for me. I could tell when it would boot properly because my X1900XTX would whir up it's fan briefly (and at full speed, so loudly). When it didn't work, it didn't spin up the fan. I did think the same thing you did, and when it wouldn't work I'd switch my monitor to the other DVI port but that didn't work.

 

I more carefully watched the boot sequence with -v on last night and noticed that the card seems to have multiple DevIDs, apparently one for each DVI port. I realized that when I connected the second port via VGA adapter to the VGA port on my monitor, switched my KVM so it would take input from it, it would no longer say mismatch for one of the DevID, it became clear that it was looking for another display when it wasn't working and if it didn't find it it would not spin up. If this is the case, I have NO idea how or why it would work with just one display the times it did and more importantly why it didn't always work with just one display.

 

Also, my machine definitely did NOT immediately go into sleep mode because I could tell it was continuing to boot into the OS by the HD activity and once the activity stopped, if I hit the power button (which normally puts the computer into sleep mode) it would then go into sleep mode. I could also tell because OS X would still play sounds etc.

 

As far as my issues, I *think* plugging something "real" into the second port has solved the problem. After I connected it properly, each time I reboot.... er shutdown then booted (can't reboot my machine, NForce4 w/USB = can't reboot?) it didn't hang at the login app. (Knock on wood....)

 

It's a bit annoying but thankfully the other computer on the KVM is my server and I don't need to access that too often and I can always get to it remotely from my workstation.

 

I think, since JCCool managed to get both ports working this is the tradeoff. Obviously an official driver would probably not have this issue, but we're far from official! :P

 

aint that the truth! 'we're far from official!'

 

heh heh

 

:)

 

BTW: this x1950xt is the fastest 'fully' working card we can get into our osx86 system at this moment right? i know there are faster PCIe cards (8800gtx) but they dont work yet?

Link to comment
Share on other sites

AFAIK, x1950xt is it. I almost gave up and put my GF6600 back in and was going to look for a GF7950 or whatever the fastest GF card was. Seems to me GF cards have an easier time getting going, I couldn't get an ATI X1300 working on my work computer so dumped it back to GF5200 and it worked immediately without any issues, same for my home computer no issues with my GF6600

 

Hopefully this current workaround keeps working.

Link to comment
Share on other sites

GimpyViper, I think I'm essentially saying the same thing as you are describing here. The computer didn't go to sleep, I meant the display looks as if it went to sleep but in fact, it's just not receiving any signals. You are right, the computer is still going as sound is still working and it's responding to keyboard strokes.

 

Also, I've noticed the same issue as you in verbose mode: it is assigning different device ID to the same card. At the time I can get the display working, the device id is 7244. Other times I can't, the device id is showing 7264. I checked the vendor profile, both ids are assigned to the Same X1950 series.

 

When I used Jccool's method, during boot in verbose mode, it recognized that the device id is 7264.

 

Jccool, is it possible to add device id 7264 in your kext file? I'm wondering if that can resolve the problem we are experiencing here.

 

Thanks.

Link to comment
Share on other sites

wigworm - We're definitely having the same issue then. As I said above, so far it appears connecting the second output port has solved the problem (knock on wood) because when it does it doesn't say mismatch for the second DevID, I don't believe it's missing from JCCool's driver, it's actually there and recognizes but freaks out when it can't sense a monitor there. Do you have a spare monitor or one with a second input to try?

Link to comment
Share on other sites

wigworm - We're definitely having the same issue then. As I said above, so far it appears connecting the second output port has solved the problem (knock on wood) because when it does it doesn't say mismatch for the second DevID, I don't believe it's missing from JCCool's driver, it's actually there and recognizes but freaks out when it can't sense a monitor there. Do you have a spare monitor or one with a second input to try?

 

When I get home later today, I will try connecting to two monitors and see what happens. I will also try connecting both ports to the same monitor. My monitor has both dvi and vga connector. I will just connect both port into the same LCD and see what happens. I will report back the findings.

Link to comment
Share on other sites

 Share

×
×
  • Create New...