106 replies to this topic
#81
Posted 02 August 2012 - 10:44 PM
Wow... thanx for the long explanation, it seems that there is a huge progress about it, i will try the Chiken fix, and the kext from g62, i hope to get same results as you, (better would be awesome), and then wait to G62 to his miracle kext or fix!!!.../
Thanx again to all of you for all your effort!!!
Thanx again to all of you for all your effort!!!
#82
Posted 04 August 2012 - 12:11 PM
Guys, i need your input!
I'm trying to get Intel HD Graphics to work with a DSDT injection.
I tried to inject it and now with original kexts i got a KP, so it does not hang on boot.
I need someone with a MBP6,1 or MBP6,2 to get this infos: (about HD Graphics)
Thanks to all in advance!
I'm trying to get Intel HD Graphics to work with a DSDT injection.
I tried to inject it and now with original kexts i got a KP, so it does not hang on boot.
I need someone with a MBP6,1 or MBP6,2 to get this infos: (about HD Graphics)
- Device ID
- Vendor ID
- Subsystem Vendor ID
- Subsystem Device ID
- Revision
Thanks to all in advance!
#83
Posted 04 August 2012 - 01:28 PM
giofrida, on 04 August 2012 - 12:11 PM, said:
Guys, i need your input!
I'm trying to get Intel HD Graphics to work with a DSDT injection.
I tried to inject it and now with original kexts i got a KP, so it does not hang on boot.
I need someone with a MBP6,1 or MBP6,2 to get this infos: (about HD Graphics)
I'm trying to get Intel HD Graphics to work with a DSDT injection.
I tried to inject it and now with original kexts i got a KP, so it does not hang on boot.
I need someone with a MBP6,1 or MBP6,2 to get this infos: (about HD Graphics)
- Device ID
- Vendor ID
- Subsystem Vendor ID
- Subsystem Device ID
- Revision
Whoa!! Slow down. First, how are you trying to get DSDT injection to enable the Intel HD Graphics? What is your reasoning, and what are you changing? Second, is the KP early in the boot stage like "Could not find driver for this platform: ACPI" or is it something else? Because in my earlier attempts to use a DSDT, I got that kernel panic. I eventually gave up using it because I found out I was only using a version that would be read from my computer directly, anyway, and unless I had something to change for something to work (which I didn't), there was no need for a DSDT.aml file.
#84
Posted 04 August 2012 - 03:46 PM
I tried to inject the GFX0 Device with _DSM Method (changing infos of GFX). I can boot with original kexts with "GE=Yes -v" but the screen is:

Strange
. I need more info about injecting the display into DSDT.
These values are picked up from IORegistryExplorer (on SL 10.6.7 w/ Partial QE/CI)
Other useful Info:
Chipset Model: Intel HD Graphics
Type: GPU
Bus: Built-In
VRAM (Total): 288 MB
Vendor: Intel (0x8086)
Device ID: 0x0046
Revision ID: 0x0012
gMux Version: 1.9.21
Displays:
Display Connector:
Status: No Display Connected
So, the new DSDT Code will be like: (add in Device (GFX0))
I think if we can grab the right values, we can have fully QE/CI.

Strange
These values are picked up from IORegistryExplorer (on SL 10.6.7 w/ Partial QE/CI)
Other useful Info:
Chipset Model: Intel HD Graphics
Type: GPU
Bus: Built-In
VRAM (Total): 288 MB
Vendor: Intel (0x8086)
Device ID: 0x0046
Revision ID: 0x0012
gMux Version: 1.9.21
Displays:
Display Connector:
Status: No Display Connected
So, the new DSDT Code will be like: (add in Device (GFX0))
Method (_DSM, 4, NotSerialized)
{
Store (Package (0x12)
{
"model",
Buffer (0x11)
{
"Intel HD Graphics"
},
"built-in",
Buffer (One)
{
0x01
},
"device-type",
Buffer (0x08)
{
"display"
},
"device-id",
Unicode ("F"),
"vendor-id",
Buffer (0x04)
{
0x86, 0x80, 0x00, 0x00
},
"subsystem-vendor-id",
Buffer (0x04)
{
0x6B, 0x10, 0x00, 0x00
},
"subsystem-id",
Buffer (0x04)
{
0x3A, 0x14, 0x00, 0x00 <--- THIS IS WRONG!
},
"revision-id",
Buffer (0x04)
{
0x12, 0x00, 0x00, 0x00
},
"VRAM,totalsize",
Buffer (0x04)
{
0x00, 0x00, 0x00, 0x12
}
}, Local0)
DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
Return (Local0)
}
I think if we can grab the right values, we can have fully QE/CI.
#85
Posted 06 August 2012 - 03:22 PM
I have been trying to get the 10.7.4 lion kexts to work for a while now. From what I can see the frame buffer is needed as it looks like this does the VRAM allocation and many other things. On mine with only the AppleIntelHDGraphics kext loaded I get the screen scrambled, via screen sharing there are ways I can get it to show the correct Graphics type but I think these are only cosmetic. I can get the Lion frame buffer to load on mine but have to add the following to the graphics card section of the DSDT using the above method in giofrida's post.
I can see a section in the kext which looks for this value, without the os-info addition and the frame buffer I get the spinning thingy continually at startup, with the os-info addition I get a grey screen eventually, and a lot more promising entries in the registry when viewed via screen sharing and it shows 768Mb of video memory. When I tried injecting the VRAM size as above this caused me problems, one of the kexts would not load, can't remember which, but with the frame buffer loaded it shows this value anyway.
I think the problem I have is it is now not detecting the screen at all to get the size and timing information, display information is blank which without the intel frame buffer shows at least some information even if wrong. I tried many ways to try and inject display EDID information without success. What I did wonder is if the default NDRIV framebuffer on a real Mac gets this information, on our systems with a single graphics card because the Intel frame buffer has grabbed the card, the NDriv framebuffer can't get any information out of the display. I know without the Intel FB the Ndriv FB loads and shows in ioregistryexplorer against the Intel graphics. If this is the case perhaps someone could write a fake ndriv frame buffer that just reads the information, I believe the source code is available for this frame buffer?
Hope this helps someone get a little further, unfortunately I have made many changes to my DSDT to get it nearer to a real MacBookPro 6.2 so not sure if there are any other changes that have helped my intel framebuffer kext to load.
"AAPL,os-info",
Buffer (0x14)
{
/* 0000 */ 0x30, 0x49, 0x01, 0x11, 0x01, 0x10, 0x08, 0x00,
/* 0008 */ 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
/* 0010 */ 0xFF, 0xFF, 0xFF, 0xFF
},
I can see a section in the kext which looks for this value, without the os-info addition and the frame buffer I get the spinning thingy continually at startup, with the os-info addition I get a grey screen eventually, and a lot more promising entries in the registry when viewed via screen sharing and it shows 768Mb of video memory. When I tried injecting the VRAM size as above this caused me problems, one of the kexts would not load, can't remember which, but with the frame buffer loaded it shows this value anyway.
I think the problem I have is it is now not detecting the screen at all to get the size and timing information, display information is blank which without the intel frame buffer shows at least some information even if wrong. I tried many ways to try and inject display EDID information without success. What I did wonder is if the default NDRIV framebuffer on a real Mac gets this information, on our systems with a single graphics card because the Intel frame buffer has grabbed the card, the NDriv framebuffer can't get any information out of the display. I know without the Intel FB the Ndriv FB loads and shows in ioregistryexplorer against the Intel graphics. If this is the case perhaps someone could write a fake ndriv frame buffer that just reads the information, I believe the source code is available for this frame buffer?
Hope this helps someone get a little further, unfortunately I have made many changes to my DSDT to get it nearer to a real MacBookPro 6.2 so not sure if there are any other changes that have helped my intel framebuffer kext to load.
#86
Posted 06 August 2012 - 03:49 PM
#87
Posted 06 August 2012 - 04:33 PM
iWin32, on 06 August 2012 - 03:49 PM, said:
Just a quick question. Did you try this with Snow Leopard? I would try Lion, but until I get graphics working, I only have an old version of iMovie that requires Rosetta to run.
No unfortunately only have Lion around these days, bit of a pain to install on my laptop and get running.
#88
Posted 08 August 2012 - 03:10 PM
I decided to cave in and give Lion a shot, but:
I'm confused as to WHERE in the DSDT this should go. You mention giofordia's post, but can't find exactly where to put it. Googling around didn't help at all. Can you help me?
ebmesnow, on 06 August 2012 - 03:22 PM, said:
I can get the Lion frame buffer to load on mine but have to add the following to the graphics card section of the DSDT using the above method in giofrida's post.
"AAPL,os-info",
Buffer (0x14)
{
/* 0000 */ 0x30, 0x49, 0x01, 0x11, 0x01, 0x10, 0x08, 0x00,
/* 0008 */ 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
/* 0010 */ 0xFF, 0xFF, 0xFF, 0xFF
},
#89
Posted 09 August 2012 - 01:02 AM
You need to find the Graphics section of your DSDT, probably called GFX0 or IGPU then add this code under it
You will also need to add the DTGP method in not already included in your DSDT, this is a fairly common method of injecting stuff, used to trick kexts into think the hardware is something other than it is and used a lot in fixes, if you look at some of the standard hacks in DSDTSE you will see it frequently used.
Method (_DSM, 4, NotSerialized)
{
Store (Package (0x0A)
{
"hda-gfx",
Buffer (0x0A)
{
"onboard-1"
},
"device_type",
Buffer (0x08)
{
"display"
},
"model",
Buffer (0x12)
{
"Intel HD Graphics"
},
"built-in",
Buffer (One)
{
0x01
},
"AAPL,os-info",
Buffer (0x14)
{
/* 0000 */ 0x30, 0x49, 0x01, 0x11, 0x01, 0x10, 0x08, 0x00,
/* 0008 */ 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
/* 0010 */ 0xFF, 0xFF, 0xFF, 0xFF
}
}, Local0)
DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
Return (Local0)
}
You will also need to add the DTGP method in not already included in your DSDT, this is a fairly common method of injecting stuff, used to trick kexts into think the hardware is something other than it is and used a lot in fixes, if you look at some of the standard hacks in DSDTSE you will see it frequently used.
#90
Posted 09 August 2012 - 04:07 AM
Thanks for the help, ebmesnow, and now I'm getting these errors in compiling the DSDT.aml.
Intel ACPI Component Architecture ASL Optimizing Compiler version 20100331 [Mar 31 2010] Copyright (c) 2000 - 2010 Intel Corporation Supports ACPI Specification Revision 4.0 dsdt.dsl 5284: Name (NBTT, Package (0x08) Remark 5048 - ^ Initializer list shorter than declared package length dsdt.dsl 6327: Method (BINI, 0, NotSerialized) Warning 1088 - ^ Not all control paths return a value (BINI) dsdt.dsl 7056: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7060: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7119: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7123: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7188: Name (_PLD, Buffer (0x10) Error 4080 - Invalid object type for reserved name ^ (found BUFFER, requires Package) dsdt.dsl 7205: Name (_PLD, Buffer (0x10) Error 4080 - Invalid object type for reserved name ^ (found BUFFER, requires Package) dsdt.dsl 7275: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7279: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7338: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7342: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7453: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7457: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7516: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7520: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7631: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7635: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7694: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7698: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7809: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7813: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7872: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7876: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 7987: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 7991: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8050: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8054: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8165: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8169: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8228: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8232: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8343: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8347: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8406: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8410: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8521: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8525: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 8584: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 8588: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 9671: Method (_WED, 1, NotSerialized) Warning 1099 - Unknown reserved name ^ (_WED) dsdt.dsl 9671: Method (_WED, 1, NotSerialized) Warning 1099 - Unknown reserved name ^ (_WED) dsdt.dsl 9671: Method (_WED, 1, NotSerialized) Warning 1099 - Unknown reserved name ^ (_WED) dsdt.dsl 9671: Method (_WED, 1, NotSerialized) Warning 1099 - Unknown reserved name ^ (_WED) dsdt.dsl 9671: Method (_WED, 1, NotSerialized) Warning 1099 - Unknown reserved name ^ (_WED) dsdt.dsl 9731: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 9944: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 10505: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 10601: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 10669: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 10728: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 11192: Name (_T_0, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_0) dsdt.dsl 11196: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 11220: Name (_T_2, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_2) dsdt.dsl 11242: Name (_T_3, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_3) dsdt.dsl 11267: Name (_T_4, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_4) dsdt.dsl 11306: Name (_T_5, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_5) dsdt.dsl 11348: Name (_T_6, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_6) dsdt.dsl 11352: Name (_T_7, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_7) dsdt.dsl 11372: Name (_T_8, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_8) dsdt.dsl 11394: Name (_T_9, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_9) dsdt.dsl 11419: Name (_T_A, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_A) dsdt.dsl 11454: Name (_T_B, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_B) dsdt.dsl 11503: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 11508: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 11562: Name (_T_0, Zero) Remark 5111 - ^ Use of compiler reserved name (_T_0) dsdt.dsl 11618: Name (_T_1, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_1) dsdt.dsl 11658: Name (_T_2, Zero) Remark 5111 - Use of compiler reserved name ^ (_T_2) dsdt.dsl 11720: Name (_WDG, Buffer (0xF0) Warning 1099 - Unknown reserved name ^ (_WDG) dsdt.dsl 13491: Subtract (Local0, 0x03) Warning 1106 - ^ Result is not used, operator has no effect ASL Input: dsdt.dsl - 14562 lines, 538932 bytes, 6032 keywords Compilation complete. 2 Errors, 8 Warnings, 60 Remarks, 8 Optimizations [Completed]Can you help me, please? Thanks in advanced!
#91
Posted 09 August 2012 - 09:11 PM
Quote
Thanks for the help, ebmesnow, and now I'm getting these errors in compiling the DSDT.aml.
#92
Posted 10 August 2012 - 06:35 AM
I have a question...
What is AAPL,os-info? What does it do?
Also the FB detects my screen properly just that the screen would have a backlight with no image...
I also get a screwed up graphics if I boot with the display on VGA. My laptop is Lenovo T410. The DSDT I can try to edit... Any way for DSDT to get the image displayed?
My screen is 1280x800 btw
What is AAPL,os-info? What does it do?
Also the FB detects my screen properly just that the screen would have a backlight with no image...
I also get a screwed up graphics if I boot with the display on VGA. My laptop is Lenovo T410. The DSDT I can try to edit... Any way for DSDT to get the image displayed?
My screen is 1280x800 btw
#93
Posted 13 August 2012 - 05:54 PM
ebmesnow, on 09 August 2012 - 09:11 PM, said:
Can you attach your DSDT and I will have a look, have you done this before? Some of the DSDT compilations are known to be full of errors before, does not look like anything we have introduced.
P1070478.JPG 1.1MB
37 downloadsI also have attached my compiled DSDT.aml as well as my error-filled DSDT source code. Please take a look at it. And yes, this is the first time I am using DSDT fixes.
my-dsdt.zip 82.15K
10 downloads
#94
Posted 19 August 2012 - 02:48 PM
Guys.. I don't think this works...
I did everything, os-info and all, the FB detected my screen fine, but still black screen
I did everything, os-info and all, the FB detected my screen fine, but still black screen
#95
Posted 23 August 2012 - 11:04 PM
I recently found a link to this useful open source tool on Apple's User forums (of all places), where a user of a real Macbook Pro (I think with the same Graphics card as us) wanted to use the integrated Graphics. Here is what one user said to use:
http://codykrieger.com/gfxCardStatus
If we use this tool (starting it by screen sharing), we may be able to use Vanilla Kexts (with or without DSDT?) to switch to the Intel HD Graphics card without kext modding, etc.
http://codykrieger.com/gfxCardStatus
If we use this tool (starting it by screen sharing), we may be able to use Vanilla Kexts (with or without DSDT?) to switch to the Intel HD Graphics card without kext modding, etc.
#96
Posted 24 August 2012 - 06:05 PM
iWin32, on 23 August 2012 - 11:04 PM, said:
I recently found a link to this useful open source tool on Apple's User forums (of all places), where a user of a real Macbook Pro (I think with the same Graphics card as us) wanted to use the integrated Graphics. Here is what one user said to use:
http://codykrieger.com/gfxCardStatus
If we use this tool (starting it by screen sharing), we may be able to use Vanilla Kexts (with or without DSDT?) to switch to the Intel HD Graphics card without kext modding, etc.
http://codykrieger.com/gfxCardStatus
If we use this tool (starting it by screen sharing), we may be able to use Vanilla Kexts (with or without DSDT?) to switch to the Intel HD Graphics card without kext modding, etc.
dont waste Time on this application its a nogo
tryed this 2 years ago
Sorry for the bad news
read the original thread of GMA 5700
kind regards
fmac
#97
Posted 24 August 2012 - 06:25 PM
fmac, on 24 August 2012 - 06:05 PM, said:
dont waste Time on this application its a nogo
tryed this 2 years ago
Sorry for the bad news
read the original thread of GMA 5700
kind regards
fmac
tryed this 2 years ago
Sorry for the bad news
read the original thread of GMA 5700
kind regards
fmac
I read the original thread, and the reason it DIDN'T Work was because of 4 screens without being able to see the menu bar. I suggested trying to enable screen sharing, but I can't test this out (as of now) because of network issues. Specifically, my built in wi-fi is unsupported, my built in ethernet is supported, but I can't get the kexts to enable it for the life of me. And I have the infamous Netgear WG111v3, and as many of you know, it doesn't save profiles (where WiFi default network configuration is stored) in lion, so I'll go start a thread to see if I can get Ethernet to work. I will say, though, for those that CAN get networking to work, try enabling the vanilla HD Graphics kexts, and start the program through screen sharing. We'll see what we can do if this program works.
P.S. If you are so convinced that it won't work, than why not modify the code to help us? After all, it's open source!!
#98
Posted 25 August 2012 - 10:37 AM
I used Snow Leopard...
Lion better? I will try it
Lion better? I will try it
#99
Posted 26 August 2012 - 04:40 AM
iWin32, on 24 August 2012 - 06:25 PM, said:
I read the original thread, and the reason it DIDN'T Work was because of 4 screens without being able to see the menu bar. I suggested trying to enable screen sharing, but I can't test this out (as of now) because of network issues. Specifically, my built in wi-fi is unsupported, my built in ethernet is supported, but I can't get the kexts to enable it for the life of me. And I have the infamous Netgear WG111v3, and as many of you know, it doesn't save profiles (where WiFi default network configuration is stored) in lion, so I'll go start a thread to see if I can get Ethernet to work. I will say, though, for those that CAN get networking to work, try enabling the vanilla HD Graphics kexts, and start the program through screen sharing. We'll see what we can do if this program works.
P.S. If you are so convinced that it won't work, than why not modify the code to help us? After all, it's open source!!
P.S. If you are so convinced that it won't work, than why not modify the code to help us? After all, it's open source!!
I had email the owner of the gfxcardstatus and the owner himself does not even know how the mechanism works. it is a work from another people.
#100
Posted 30 August 2012 - 01:34 PM
I have a developer-based idea on how to fix resolution on Intel HD Graphics. The problem: I'm only a beginner in developing. So, here is my idea:
If I understand G62's kext correctly, it is enabling the graphics card and extending the display to our EDID's resolution. However, OS X is still outputing 1024x768 because it doesn't have a framebuffer. However, if my idea is right, a framebuffer won't be needed.
I tried to use the VBIOS dump earlier in this post, but I couldn't get it to work. According to Slice (the developer of Clover on Project OS X forums), the reason is Chameleon has poor support of Intel VBIOS ROM files. Clover has better support for the use of Intel ROMs with its GraphicsInjector. However, the problem is that Clover is based off of UEFI firmware, so without a graphics driver for the firmware, it can only output three resolutions: 640x480, 800x600, and 1024x768. This is true for those who are even lucky that their VBIOS has their custom resolution and Chameleon picks it up: if those lucky people boot with Clover, OS X will only output 1024x768; switch back to Chameleon, and their custom resolution can be used.
I have two ideas: one Chameleon-based (which may or may not work), and one for Clover. First, the Clover one (because it needs to be based heavily off of G62's AppleSamplePCI.kext). Before Clover loads the boot menu, it loads EFI drivers. My idea is to port G62's kext to an EFI driver. Any driver, in theory, can be ported to EFI, but without knowing a good deal on how the driver works, it can be difficult. That's where G62's kext comes in. He knows how to code the driver. I think that if the kext is used before the OS is loaded, OS X may load our correct resolution (especially since the driver would enable the Graphics Card and have it output our correct resolution in EFI).
And as for a Chameleon fix, I'm not sure how we would start this one (other than looking at how 915resolution loads VBIOS info). We would need to have an external tool (perhaps a Grub2 module?) copy a patched VBIOS (or the VBIOS in the forum) into memory, and then chainload Chameleon from there. The idea here is to override the VBIOS by loading a different one into memory before Chameleon loads. This would enable Mac OS X to load a different resolution.
Anyways, those are my ideas. Like I said up above, I'm not a skilled developer, so if those are concepts that you think you can use, than feel free to try them out. I'd also like feedback: would these ideas even work at a first glance?
If I understand G62's kext correctly, it is enabling the graphics card and extending the display to our EDID's resolution. However, OS X is still outputing 1024x768 because it doesn't have a framebuffer. However, if my idea is right, a framebuffer won't be needed.
I tried to use the VBIOS dump earlier in this post, but I couldn't get it to work. According to Slice (the developer of Clover on Project OS X forums), the reason is Chameleon has poor support of Intel VBIOS ROM files. Clover has better support for the use of Intel ROMs with its GraphicsInjector. However, the problem is that Clover is based off of UEFI firmware, so without a graphics driver for the firmware, it can only output three resolutions: 640x480, 800x600, and 1024x768. This is true for those who are even lucky that their VBIOS has their custom resolution and Chameleon picks it up: if those lucky people boot with Clover, OS X will only output 1024x768; switch back to Chameleon, and their custom resolution can be used.
I have two ideas: one Chameleon-based (which may or may not work), and one for Clover. First, the Clover one (because it needs to be based heavily off of G62's AppleSamplePCI.kext). Before Clover loads the boot menu, it loads EFI drivers. My idea is to port G62's kext to an EFI driver. Any driver, in theory, can be ported to EFI, but without knowing a good deal on how the driver works, it can be difficult. That's where G62's kext comes in. He knows how to code the driver. I think that if the kext is used before the OS is loaded, OS X may load our correct resolution (especially since the driver would enable the Graphics Card and have it output our correct resolution in EFI).
And as for a Chameleon fix, I'm not sure how we would start this one (other than looking at how 915resolution loads VBIOS info). We would need to have an external tool (perhaps a Grub2 module?) copy a patched VBIOS (or the VBIOS in the forum) into memory, and then chainload Chameleon from there. The idea here is to override the VBIOS by loading a different one into memory before Chameleon loads. This would enable Mac OS X to load a different resolution.
Anyways, those are my ideas. Like I said up above, I'm not a skilled developer, so if those are concepts that you think you can use, than feel free to try them out. I'd also like feedback: would these ideas even work at a first glance?
Also tagged with one or more of these keywords: gma, clarkdale, arrandale, 5700, intel, hd, fixed
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users



Sign In
Create Account








