Jump to content

ATILead - natit-like injector for Radeon.

Slice

By Slice,
in ATI Framebuffer

 Share
107 posts in this topic

Recommended Posts

Soaa
Soaa
Posted
Posted

I see, we both have the same problem then... that of not knowing where to go! I tried adding some more values from the Mac's ioreg into ATILead's Info.plist, but I'm still getting a crazy fuzz like before.

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
I see, we both have the same problem then... that of not knowing where to go! I tried adding some more values from the Mac's ioreg into ATILead's Info.plist, but I'm still getting a crazy fuzz like before.

Digital photo or your screen?

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
Fuzzy pics. Hope this helps?

:unsure:

Yes! It is clear!

These pictures arisez when ATIRadeon8500.kext sets Display_Base_address to zero so you see RAM on the screen. Or AGP_BASE_ADDRESS=0 so AGP_transaction send data to zero.

Check once again all info.plists about memory bases, sizes and addressing.

 

AGPGart 265 permit you to change AGP_Base. Please, prepare detailed report, what is happen with different settings. Dmesg when possible.

Link to comment
Share on other sites
Soaa
Soaa
Posted
Posted

I've tried AGP_Base=0, 0xf0000000, 0xe0000000 before posting these pics. All had the same result, I'll post more test results as they come.

 

Edit: I tried fiddling with ATILead's IOSpaceOffset a bit, but no difference. I don't see any other settings to play with. Do you have any suggestions for me?

Link to comment
Share on other sites
Soaa
Soaa
Posted
Posted

Thanks! I'm at work now, so I can't test it right away. I'll test it once I get home.

 

May I ask, what did you patch, and how? I'm curious, and if this doesn't work, maybe I can do some patching of my own. :)

 

Also, I test this with ATILead and AGPGart, right? With what settings?

Link to comment
Share on other sites
Soaa
Soaa
Posted
Posted

Okay, tested your new driver.

 

Without AGPGart nor ATILead, it loads, and the bottom quarter of my screen goes fuzzy, then my computer reboots itself instantly.

 

With ATILead, with or without AGPGart, it does the same fuzzy lines as I've shown.

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted

I make new patch for Radeon8500. (other version 1.4.40.4). I replaced commands from addresses 15ea8 to 15f1d and later.

Now I can boot to desktop with such settings

I have 1Gb total RAM = 0x40000000

VRAM 128Mb is shared with RAM so

VRAM_addr = 0x38000000 VRAM_size=0x08000000

I patch ATIRadeon8500.kext to make this addresses.

AGP_aperture set to 0x04000000 = 64Mb.

In AGPGart.info.plist I write

			<key>AGP_Base</key>
		<integer>0x3c000000</integer>

That is VRAM_addr+AGP_aper=0x38000000+0x04000000

Then in ATI1Slice I set F8,size=0x00300000. It is too bad. I have very slow system. But I can enter desktop!

With Callisto I can't enter desktop with any settings!!!!!!

So, ATILead require corrections, or Callisto don't work with AGP at all.

I think a good way is to launch ATINDRV. But how?

Real Mac with Radeon9000 uses AppleNDRV which is PowerPC only, so useless.

One more probable way is to revive RadeonFB project, but no one programmer is successful.

The most real way is to use IONDRVSupport without switch resolutions.

ATIRadeon8500.kext_3.zip

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
Slightly off-topic... I use JaS 10.4.7, should I use Software Update to update to 10.4.11 before testing?

 

I am using JaS 10.4.6 so... at your will. The kext will work in any system.

Link to comment
Share on other sites
Soaa
Soaa
Posted
Posted

I see. I'll test it once I get home. I'd also like to upgrade to Leo, would you think that's a good idea or a bad one? =P

 

For testing purposes though, I'll stick to Tiger for now. I might jump to Leo once things work.

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
I see. I'll test it once I get home. I'd also like to upgrade to Leo, would you think that's a good idea or a bad one? =P

 

For testing purposes though, I'll stick to Tiger for now. I might jump to Leo once things work.

AFAIK drivers for Tiger work better then for Leo.

Link to comment
Share on other sites
Soaa
Soaa
Posted
Posted

I see. Tiger it is then.

 

I tried playing with AGPGart's and ATILead's values with your latest ATIRadeon8500 and I'm still getting a fuzzy screen. If you have anything for me to test in particular, please tell me.

 

Also, this might be useful to you: I have 768MB of system RAM and 32MB of dedicated VRAM, and I can't share. That probably affects some of the values.

Link to comment
Share on other sites
someguy429
someguy429
Posted
Posted

Hello, I've been trying some of your drivers with x200 mobility (dev id 5955), I can load the ATILead with Calisto and with 8500 kext. I can get to my desktop and I have two screens on the same screen, I can change my screen resolution (on both screens). Thou I have no QE/CI and I only have 8mb VRam. I just thought I'd tell you. Anything else you want to know?

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
Hello, I've been trying some of your drivers with x200 mobility (dev id 5955), I can load the ATILead with Calisto and with 8500 kext. I can get to my desktop and I have two screens on the same screen, I can change my screen resolution (on both screens). Thou I have no QE/CI and I only have 8mb VRam. I just thought I'd tell you. Anything else you want to know?

Your kext is not 8500 but 9700. I am doubt that 8500 is loaded in your case.

Look thread about Radeon 200M

If your test the patch report in those thread. About ATILead you got all possible result. What do you say about two monitors? You can use it with different desktops?

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted

I think you have two desktop but similar on two monitors...

QE disabled if VRAM 

				<key>AAPL,gart-width</key>
			<integer>32</integer>
			<key>AAPL,vram-memory</key>
			<integer>67108864</integer>
			<key>ATY,Copyright</key>
			<string>Copyright ATI Technologies Inc. 2005</string>
			<key>ATY,DeviceID</key>
			<integer>22581</integer>
			<key>ATY,EFIVersion</key>
			<string>01.00.068</string>
			<key>ATY,FrameBufferOffset</key>
			<integer>64</integer>
			<key>ATY,VRAM,MemSize</key>
			<integer>64</integer>

Better value for gart-width = 64 for your Radeon

ATY,VRAM,MemSize - I dunno. Erase it at all!

Link to comment
Share on other sites
someguy429
someguy429
Posted
Posted

I tried changing those values but it didn't seem to change anything.

And when I say two displays on one monitor, i mean that when I move my mouse out of the left side of screen it comes back on the right side. If I highlight things the screen will redraw the background.

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted

That's what I found

	prop = OSDynamicCast( OSData, nub->getProperty("AAPL,vram-memory"));
if (prop && (prop->getLength() >= (2 * sizeof(IOByteCount))))
{
	IOByteCount * lengths;

	lengths = (IOByteCount *) prop->getBytesNoCopy();
	length = lengths[1];
	vramBase = lengths[0];
}

It means AAPL,vram-property must be and contains two value

vramBase and length

For example for me

vramBase=0x38000000

length=0x08000000

Using Property List Editor I write the values and got

				<key>AAPL,vram-memory</key>
			<data>
			OAAAAAgAAAA=
			</data>

Testers, where are you?

Developers... nobody here :D

Link to comment
Share on other sites
Krazubu
Krazubu
Posted
Posted

Hi Slice,

I managed to get the EFI module "Radeon.efi", it could give some precious infos, tell me if you're interested.

However, it needs to be decompressed/reverse engineered, it seems there are tools for this that will extract all data in a rather "clear" form, but I didn't look further yet.

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
Hi Slice,

I managed to get the EFI module "Radeon.efi", it could give some precious infos, tell me if you're interested.

However, it needs to be decompressed/reverse engineered, it seems there are tools for this that will ectract all data in a rather "clear" form, but I didn't look further yet.

Yes, it is interesting. I can reverse any module but I don't know what to expect. Give me it to try!

Link to comment
Share on other sites
Slice
Slice
Posted
  • Author
Posted
well, could you explain what vramBase and length are. For example on my "5975" with 128MB dedicated and 128 shared?

I use leo for the moment, any chance of getting it working?

 

Tell me exactly what you want me to test that I will do.

Now I have 10.5.2 running and my ID i 5975.

 

I can't show you detailed instruction to make it working. I can only propose you to search new situation what is working and what no. Any information may be useful.

VRAM-address (if shared) = RAM - VRAM_size.

I have 1Gb RAM = 0x40000000

and 128Mb VRAM = 0x08000000

thus VRAM_address=0x40000000-0x08000000=0x38000000 - it is hexadecimal system.

Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...