#761
Posted 31 December 2010 - 01:59 PM
#762
Posted 01 January 2011 - 11:51 AM
-Fakesmc 3.0 together with x3100 plugin from here. I´ve tryed other versions but I was getting KP on the x3100 plugin. Try at your own risk and have always a rescue pendrive at hand
-IntelCPUmonitor and ACPImonitor from here
Then you need to insert these two codes on your DSDT (as explained a couple of post ago).
Just before Device (PDRC):
Device (SPIO)
{
Name (_HID, EisaId ("PNP0C02"))
Name (_CID, "monitor")
Name (_STA, 0x0F)
Method (FTN0, 0, NotSerialized)
{
Store (\_SB.PCI0.LPCB.EC.SFAN, Local1)
Return (Local1)
}
}And within Device (EC):
OperationRegion (ECOR, EmbeddedControl, Zero, 0xFF)
Field (ECOR, ByteAcc, NoLock, Preserve)
{
Offset (0xD7),
SFAN, 8
}and lastly you will need to hack the iStatMenusTemps.menu file (Library/ApplicationSupport/iStatLocal/extras) to show "%" instead of "rpm". This can be done with a hex editor, finding the "rpm" string and replacing it with " % " (take care on the spaces)...
#763
Posted 03 January 2011 - 12:24 AM
maybe only one restart was needed to reinstall the kexts after updating to 16.6.5 and 10.6.6.... i only tested this with my umts-expresscard (huawei e870)... maybe someone can test this with an other expresscard...
#764
Posted 03 January 2011 - 03:00 AM
Recently I have disassembled AppleIntelGMAX3100FB (already patched for 2a12), and I found interesting function:
AppleIntelGMAX3100FB::hwSetBacklight(unsigned long)In this function I have replaced this code:
+90 000016cc 899154120600 movl %edx,0x00061254(%ecx)by this(register %eax contains current brightness):
+90 000016cc 898154120600 movl %eax,0x00061254(%ecx))
(To patch simply replace this opcode: 899154120600 by this: 898154120600 in 0xED)
And now, my FB doesn't set uncorrect brightness values, but everithing else works as previous(I have slider for brightness and hot kyes works).
This patched driver works realy good with GenericBrightness by Ivik(from first post)
Also I have changed _BCM method in PNLF device(this gives more smooth brightness change for me):
Method (_BCM, 1, Serialized)
{
Store (Arg0, C1B1)
If (LEqual (Arg0, 0x01EE)) { Store (0xFFFFF800, LEVL) }
If (LEqual (Arg0, 0x01CF)) { Store (0xFFFFF000, LEVL) }
If (LEqual (Arg0, 0x01B0)) { Store (0xFFFFE000, LEVL) }
If (LEqual (Arg0, 0x0191)) { Store (0xFFFFD000, LEVL) }
If (LEqual (Arg0, 0x0173)) { Store (0xFFFFC000, LEVL) }
If (LEqual (Arg0, 0x0154)) { Store (0xFFFFB000, LEVL) }
If (LEqual (Arg0, 0x0135)) { Store (0xFFFFA000, LEVL) }
If (LEqual (Arg0, 0x0116)) { Store (0xFFFF9000, LEVL) }
If (LEqual (Arg0, 0x00F7)) { Store (0xFFFF8000, LEVL) }
If (LEqual (Arg0, 0x00D8)) { Store (0xFFFF7000, LEVL) }
If (LEqual (Arg0, 0x00B9)) { Store (0xFFFF6000, LEVL) }
If (LEqual (Arg0, 0x009A)) { Store (0xFFFF5000, LEVL) }
If (LEqual (Arg0, 0x007C)) { Store (0xFFFF4000, LEVL) }
If (LEqual (Arg0, 0x009A)) { Store (0xFFFF3000, LEVL) }
If (LEqual (Arg0, 0x007C)) { Store (0xFFFF2800, LEVL) }
If (LEqual (Arg0, 0x005D)) { Store (0xFFFF2000, LEVL) }
If (LEqual (Arg0, 0x003E)) { Store (0xFFFF1800, LEVL) }
If (LEqual (Arg0, 0x001F)) { Store (0xFFFF1000, LEVL) }
If (LEqual (Arg0, 0x0000)) { Store (0xFFFF0000, LEVL) }
}About keys Fn+F7 , Fn+F8 ... - they simply doesn't work, and I can not find-out why:
- I have tried to get scancodes using modified VoodooPS2Keyboard - no result
- I have looked for keyboard port(0x60) in Reggie SE - no result(no changes in port when pressing this keys).
Any ideas about this?
Also I had have one problem with Fn+F11, Fn+F12 (Volume Up/Down) - slider moves for two points(by one key press)...:
By this reason, I have modified VoodooPS2Keyboard from this topic,:
- fix for Volume Up/Down (warning if haven't this problem DON'T USE this)
- Replaced "Application" key by "Right Command"
- fixed key "`~"
VoodooPS2Keyboard.zip 23.52KB
14 downloads
#765
Posted 03 January 2011 - 10:34 AM
I think you have a typo mistake (the original and replaced values are the same?) but,To patch simply replace this opcode: 899154120600 by this: 899154120600 in 0xED)
Sounds great! Please, could you post your GMA modified binary so I can diff it? Maybe would be nice to do an auto-patcher script
Did you try without ivik driver?
As for the keyboard FN+F7/F8 seems to be blocked by the GMA driver brightness control function. Maybe Im wrong but I did some testing some time ago and if you bybass the brightness control routine in the GMA framebuffer (patching the binary) the FN+F7/F8 would change the brightness (by ACPI calls). I dont know if they would produce scancodes in this situation.
I will try your PS2keyboard kext as I have the same bug on the volume control.
Would be possible to assign some F keys to control functions (like volume, etc) without having to press the FN key?
Thanks!
#766
Posted 03 January 2011 - 11:18 AM
Sorry ... my mistake...I think you have a typo mistake (the original and replaced values are the same?)
Should be this:
replace this opcode: 899154120600 by this: 898154120600 in 0xED
Yes, brightness slider moves but without brightness change.Did you try without ivik driver?
Probably, this can be useful:
VoodooPS2Keyboard(compiled kext) :
VoodooPS2Keyboard.kext.zip 25.95KB
36 downloadsAppleIntelGMAX3100FB patched & original files:
AppleIntelGMAX3100FB_files.zip 107.9KB
96 downloadsI have this:Would be possible to assign some F keys to control functions (like volume, etc) without having to press the FN key?
#767
Posted 03 January 2011 - 05:57 PM
Sorry ... my mistake...
Should be this:
replace this opcode: 899154120600 by this: 898154120600 in 0xED
Works great! No more wierd screens. I will try to do a patcher script (in the fashion of the 2a12 patcher) so we dont have to patch the kext manually after every update...
About your DSDT code, I didnt notice much difference in the brightess change, but I´ve commited it anyways to the "official" DSDT beacause it looks a lot cleaner. Thanks!
#768
Posted 03 January 2011 - 09:48 PM
I have a big problem with my laptop. I want to install snow leopard on hp 6720s with Intel® Celeron® CPU 550 @ 2.00GHz. Can you give me some hints for proper installation step by step? I've tried few times and always someting was wrong. When I used HP 6720s_BootCD there was an error message "CPU not supported"... Help please...:-)
#769
Posted 03 January 2011 - 09:58 PM
Thats new to me. As far as I know your CPU should work....could you take a picture of the screen?When I used HP 6720s_BootCD there was an error message "CPU not supported"... Help please...:-)
Which install disc (not the boot CD) are you using? You need Snow Leopard 10.6 retail
#770
Posted 04 January 2011 - 12:15 AM
Black, "tal vez" su celeron no es soportado porque a partir de tu nuevo boot cd todo el proceso se realiza con speedstep activado, lo cual "en teoría" requiere un C2D o dual core. Con tu anterior cd esto no era posible (speedstep durante la instalación) porque había incoherencia entre el perfil (MB3,2) y el legacy (MB3,1) si no me equivoco, habiéndolo comprobado con el system.log en el perfil del sistema desde el DVD retail. Posible solución: usar el v0.2 temporalmente. Saludos.Thats new to me. As far as I know your CPU should work....
#771
Posted 04 January 2011 - 01:14 AM
To get work of native brightness control, replace in AppleIntelGMAX3100FB
this bytecode:
8b81541206008b932402000083e0fe39c27413
by this one:
8b932402000069d28400000081ca0000ffff90
this would replace original code:
movl 0x00061254(%ecx),%eax #give current value to EAX movl 0x00000224(%ebx),%edx #give value to set to EDX andl $0xfe,%eax #remove all gabredge from current value cmpl %eax,%edx #compare current and "value to set" je 0x000016df #if EDX=EAX jump to end of function(else change it)by this:
movl 0x00000224(%ebx),%edx #give value to set to EDX imull $0x00000084,%edx,%edx #multiplie it by 0x84 orl $0xffff0000,%edx #give 0xffff<value> format nop #do nothing and set patched brightness value =)
About constant 0x84 ...
values to set are:
0xffff01ee,0xffff0170,0xffff0154,0xffff014c,0xffff013c, 0xffff012a,0xffff0118,0xffff0108,0xffff00f6,0xffff00e4,0xffff00d2, 0xffff00c0,0xffff00b0,0xffff009e,0xffff008c,0xffff007c,0xffff0000after patch them are:
0xfffffeb8,0xffffbdc0,0xffffaf50,0xffffab30,0xffffa2f0, 0xffff99a8,0xffff9060,0xffff8820,0xffff7ed8,0xffff7590,0xffff6c48, 0xffff6300,0xffff5ac0,0xffff5178,0xffff4830,0xffff3ff0,0xffff0000
And them are almost the same as we need =)
Patched files from by me(original already patched for 2a12):
AppleIntelGMAX3100FB_brightness_fix_patched_files.zip 107.58KB
529 downloads
#772
Posted 04 January 2011 - 10:52 AM
Hooray!!!! I have done work on patch to forget about GenericBrightness.kext for us...
Amazing mate! Is working great!
Thanks a lot!
BTW: I´ve PMed you about the FN keys
No lo se; hace algunas semanas tube acceso a una 6720s con un celeron 1.7 ghz. Hice una copia exacta de mi disco duro (10.6.4 en ese momento), lo conecte y booteo sin problemas (tuve que pachear la x3100 2a12 solamente y borrar la informacion del cpu en el dsdt)...obviamente speedstep no funcionaba, pero todo lo demas si.Black, "tal vez" su celeron no es soportado porque a partir de tu nuevo boot cd todo el proceso se realiza con speedstep activado, lo cual "en teoría" requiere un C2D o dual core. Con tu anterior cd esto no era posible (speedstep durante la instalación) porque había incoherencia entre el perfil (MB3,2) y el legacy (MB3,1) si no me equivoco, habiéndolo comprobado con el system.log en el perfil del sistema desde el DVD retail. Posible solución: usar el v0.2 temporalmente. Saludos.
Cuando se carga la informacion de speedstep desde las tablas SSDT, los datos son dinamicos y cambian segun el CPU que este instalado; en el caso de un celeron supongo que no deberia cargar ningun P-state
#773
Posted 04 January 2011 - 06:18 PM
#774
Posted 04 January 2011 - 06:32 PM
http://www.mediafire.com/?unhhwkjhzzyCan you show mi where can I find v02 installer for Hp 6720s? On BlackCH's Blog?
#775
Posted 04 January 2011 - 06:52 PM
#776
Posted 05 January 2011 - 08:39 AM
What's now...? :-(
#777
Posted 05 January 2011 - 09:55 AM
In method "_SB.PCI0.LPCB.EC.C22C"
Removed this:
Store (Local0, Index (DerefOf (Index (C22F, Arg0)), One))
(because at this moment Local0 contains "Last Full Charge Capacity" but not "Design Capacity" which must be )
And changed default value of "Design Capacity" by correct for my notebook(probably we can read it somewhere ... but i didn't know where.. )
[codebox]Name (C22F, Package (0x02)
{
Package (0x0D)
{
One,
0x1130, // 4400 mAh
Ones,
One,
Ones,
Zero,
Zero,
0x64,
0x64,
"Primary",
"100000",
"LIon",
"Hewlett-Packard"
},
Package (0x0D)
{
One,
0x1130, // 4400 mAh
Ones,
One,
Ones,
Zero,
Zero,
0x64,
0x64,
"Travel",
"100000",
"LIon",
"Hewlett-Packard"
}
})[/codebox]
Result of patching:
Screen_shot_2011_01_05_at_11.46.26_AM.png 43.76KB
141 downloads
#778
Posted 05 January 2011 - 11:24 AM
I have modified dsdt to get more correct battery info:
In method "_SB.PCI0.LPCB.EC.C22C"
Removed this:
Store (Local0, Index (DerefOf (Index (C22F, Arg0)), One))
Is great that you are fixing this. I´ve been trying for a while but I have very little code experience.
The big probelm I noticed is that the "Full Charge Capacity" fluctuates all the time (in OSX) thus reducing the charge capacity a lot. If I go to Windows and discharge/charge the battery a couple of times it recovers its full capacity (around 2600 mAh for mine), but as soon as I use it in OSX this value starts dropping... I think the dsdt battery section is not very compatible with OSX. Maybe you want to have a look to the macbook 3,1 dsdt to see how is done there:
dsdt_mb3_1.dsl.zip 14.04KB
22 downloadsI´ve been scanning the (EC) registers with RW Tool and this is what I found so far:
offset 0x84 (8 bits) holds the AC/Battery flag
offset 0xA8 (16 bits) holds the current charge value
Maybe this registers could be usefull to implement native battery reading via fakesmc.kext and ACPImonitor.kext
#779
Posted 05 January 2011 - 12:26 PM
#780
Posted 05 January 2011 - 04:01 PM
Is great that you are fixing this. I´ve been trying for a while but I have very little code experience.
The big probelm I noticed is that the "Full Charge Capacity" fluctuates all the time (in OSX) thus reducing the charge capacity a lot. If I go to Windows and discharge/charge the battery a couple of times it recovers its full capacity (around 2600 mAh for mine), but as soon as I use it in OSX this value starts dropping... I think the dsdt battery section is not very compatible with OSX. Maybe you want to have a look to the macbook 3,1 dsdt to see how is done there:dsdt_mb3_1.dsl.zip 14.04KB 22 downloads
About battery, recently I have similar problem ...:
I have used notebook most of time from AC adapter and battery begin loosing "maximum capacity"...
After reading some Li-Ion battery exploitation recommendations,(like this) I has found solution - simply charge/discharge battery more often.
Unfortunately, I am too non ACPI programmer(I am web programmer.. ) ...
But, I have litle bit cleaned up management section and tried to implement Smart battery system (simply coping it definition from mb31 dsdt, because i didn't know how to implement it correctly) - but without result ...
Probably you can find-out some my changes in dsdt useful for future editing, my current version:
dsdt_1.zip 21.46KB
15 downloads
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users



Sign In
Create Account





