AnV XNU Kernel V1.4

Andy Vandijck · February 23, 2009

Version 1.2 works fine for me, with every dyld I've tested (yours and vanilla). I'll be waiting for that version...
Thanks for all your efforts.

OK work for me thus...

It is really too bad since it is way more responsive...

I have also noticed it tends to vary from user to user...

Guess it will be making non-optimized dyld built-in to the core.

Intel users who experience problems can allways use vanilla dyld...

mitch_de · February 23, 2009

OK work for me thus...
It is really too bad since it is way more responsive...

I have also noticed it tends to vary from user to user...

Guess it will be making non-optimized dyld built-in to the core.

Intel users who experience problems can allways use vanilla dyld...

I deleted all dyld caches (seems to be some .db files) before i reboot the V1.3 version.

Perhaps there are only some conflikts (like installing new .kext, old .mkext ?) ?

Two Apps crashed with V1.3 , but after second App start it runs without app crash!!

Also some system caches maybe the reason and not the V1.3 .pkg itself ?

Would it be possible to include an clean.sh in the .pkg which cleans all needed caches ?

My only problem was sleep, but that maybe because ov pcefi V9 - chameleon V11 has not that problem.

Azimutz · February 23, 2009

I think mitch_de has point..

Process: Xbench [252]
Path: /Applications/Utilities/Xbench.app/Contents/MacOS/Xbench

Identifier: com.spiny.xbench

Version: 1.3 (1.3)

Code Type: X86 (Native)

Parent Process: launchd [156]

Date/Time: 2009-02-23 05:01:18.779 +0000

OS Version: Mac OS X 10.5.6 (9G66)

Report Version: 6

Exception Type: EXC_BAD_ACCESS (SIGBUS)

Exception Codes: KERN_PROTECTION_FAILURE at 0x0000000000000004

Crashed Thread: Unknown

Error Formulating Crash Report:

Called memoryAtAddress: 0x85d8b14, which is in an unmappable portion of [0x0 -> 0xffffffffffffffff] in PID# 252.

0x0089f10b

0x00bdce3b

0x000b933b

0x000b2f76

0x000b2683

0x000b25ee

0x000c7b96

0x000c78aa

0x000c2f95

0x00002e74

0x000094a2

0x0000b5d4

0x0000b121

0x00a3b03b

0x0000a828

0x00a0f095

0x00a0ef52

Backtrace not available

Unknown thread crashed with X86 Thread State (32-bit):

eax: 0x0327a490 ebx: 0x031b80de ecx: 0x1ce81f98 edx: 0x0036e30a

edi: 0x00000000 esi: 0x1ce070a0 ebp: 0xb04096d8 esp: 0xb0409680

ss: 0x0000001f efl: 0x00010246 eip: 0x031b80fc cs: 0x00000017

ds: 0x0000001f es: 0x0000001f fs: 0x0000001f gs: 0x00000037

cr2: 0x00000004

Binary images description not available

..i got these crashes with XBench, always after installing kernels.

Crashes 2 or 3 times at first launches and then, no more crashes. Happened to me 2 times.

Don't know if it has something to do with the matter..?



Andy.. finally got some results at loading HPET at start up.

I forgot to try load the old HPET with DSDT.. it loads and doesn't break Ethernet. I was conviced that the problem with network was caused by patching DSDT with -newHPET but, it's not! It's the new HPET kext that breaks it.

I still can't load it with "useoldhpet=1", though.

Naunaud · February 23, 2009

I would like to help you but then I need info on this error/bug.
Also I would then like to know what kind of system you have (CPU, etc...) plus the type of system in case of an AMD (retail/patched).

And when exactly did it forever loop something/what did it exactly run in an endless loop?

Cannot provide you any log unfortunately, but here's some infos:

It's a laptop: Pentium M 1.7Ghz and 1Go DDR, installed on an USB drive

I'm running Leopard 10.5.6 (XxX version made by BlackCH)

The problem was after install, the computer booted up fine until the blue screen before entering the session and it looped there, with the circle running every 30 secondes, then a very fast message from console (I was in verbose) never had the chance to read it... (just saw something like "run loop")

Couldn't even boot in safe mode (actually as I needed the lappy, I quickly reinstalled from scratch)

Don't know if it'll help you, but I tried

Andy Vandijck · February 23, 2009

Cannot provide you any log unfortunately, but here's some infos:
It's a laptop: Pentium M 1.7Ghz and 1Go DDR, installed on an USB drive

I'm running Leopard 10.5.6 (XxX version made by BlackCH)

The problem was after install, the computer booted up fine until the blue screen before entering the session and it looped there, with the circle running every 30 secondes, then a very fast message from console (I was in verbose) never had the chance to read it... (just saw something like "run loop")

Couldn't even boot in safe mode (actually as I needed the lappy, I quickly reinstalled from scratch)

Don't know if it'll help you, but I tried

Hmmm that is pretty similar to my own laptop (Celeron M, the same amount + type of RAM).

Did any older version of my kernel work fine + which version of my core.

Soon I'll post a rebuilt version...

Did you try with vanilla dyld + my kernel?

It seems the new dylinker gives problems for some...

I'll continue working on the issues though...

Andy Vandijck · February 23, 2009

I deleted all dyld caches (seems to be some .db files) before i reboot the V1.3 version.
Perhaps there are only some conflikts (like installing new .kext, old .mkext ?) ?

Two Apps crashed with V1.3 , but after second App start it runs without app crash!!

Also some system caches maybe the reason and not the V1.3 .pkg itself ?

Would it be possible to include an clean.sh in the .pkg which cleans all needed caches ?

My only problem was sleep, but that maybe because ov pcefi V9 - chameleon V11 has not that problem.

Very interesting...

If this would solve every ones problems it would be great...

Now only figuring out where those caches are kept (NOTE: mkext file is deleted after kernel install...).

How about rebuilding using update_dyld_shared_cache binary?

That should auto rebuild these caches, right?

Maybe I should include in the installer script a method for deleting + updating dyld cache...

That way we would be sure it thoroughly rebuilds this...

Andy Vandijck · February 24, 2009

Got it.. fassl's patcher applies HPET patch by default, but not with -newHPET option. At least that's what i got from his instructions.
Lately i've been patching DSDT mainly with ACPI Patcher (great tool, by the way) and manualy. Usualy i don't patch HPET and RTC. I'm using DSDT for Graphics, injecting Audio and some readings in Sys Profiler.

Ok.. installed System.kext and AppleHPET.kext.. used useoldhpet=1 and the kext doesn't load by it self. But, now i don't get the error i was getting when loading it manually.

I did reinstalled my system.. was a bit messed up of all the testing.

Will give it a rest for now. AFAIK, HPET it's not required by any SO.

Starting to get the feeling this is one of those things that "has to be there", like EFI partition.. it's there but Apple doesn't use it. Don't know if Mac's use HPET?

As i read (http://en.wikipedia.org/wiki/High_Precision_Event_Timer) XP also contains a driver for HPET but it's not functional. I do have it listed as a device under Windows XP and i can disable/enable it.. it's enabled by default. Also shows on Everest report as sys device, PnP device, Windows device and in ACPI. In fact, Everest shows a HPET table under ACPI, apart from DSDT table...

I have Pentium D 925 and i can't find any reference of HPET on it Googlin' (or any other CPU) and i read that HPET is part of the chipset (http://software.intel.com/en-us/forums/watercooler-catchall/topic/52108/), not CPU!?

As for OS X all i know is, i can load the kext.. if it's doing something or used by something, i don't know ..just trying to know my machine better.. it's hard to be noob

Thanks for all.

I have also done some reading on this...

It is very interesting the way HPET is built up.

It is apparently meant as a replacement for the 8254 PIT, 8259 PIC, RTC + a 32/64bit timer.

It is controlled by BIOS (ACPI...).

It is comprised of 3 main parts.

I think it would be possible to patch this up to the old timers as well + construction of an HPET table would also be possible for older machines by mapping...

I have also included the specs for those who are interested in knowing more about these...

Block diagrams are included in all... (Hell we could even duplicate it in a logic simulator...).

ACPIspec30b.part1.rar

ACPIspec30b.part2.rar

alien2xx · February 24, 2009

Hi,

I installed it on Ideneb 10.5.6, updated from 9.4.0 StageXNU kernel, it works great but I have these error messages, not sure what the mean, I would like to fix the "macmodel" error at the end too, any ideas?:

new-host:~$ sudo dmesg

Password:

npvhash=4095

hi mem tramps at 0xffe00000

PAE enabled

64 bit mode enabled

rtclock_init: Taking bus ratio path 4 (Intel / Apple)

TSC: Verification of clock speed PASSED.

TSC: Frequency = 3391.461597MHz, FSB frequency = 199.497741MHz, bus ratio = 17

Darwin Kernel Version 9.6.0: Wed Feb 18 15:39:36 CET 2009; AnV XNU Kernel Release 1.3 :xnu-1228.9.59/BUILD/obj/RELEASE_I386

standard timeslicing quantum is 10000 us

vm_page_bootstrap: 511611 free pages and 12677 wired pages

mig_table_max_displ = 79

AppleDecrypt: Starting...

AppleACPICPU: ProcessorApicId=1 LocalApicId=0 Enabled

AppleACPICPU: ProcessorApicId=2 LocalApicId=1 Enabled

AppleACPICPU: ProcessorApicId=3 LocalApicId=6 Disabled

AppleACPICPU: ProcessorApicId=4 LocalApicId=7 Disabled

Loading security extension com.apple.security.TMSafetyNet

calling mpo_policy_init for TMSafetyNet

Security policy loaded: Safety net for Time Machine (TMSafetyNet)

Loading security extension com.apple.nke.applicationfirewall

Loading security extension com.apple.security.seatbelt

calling mpo_policy_init for mb

Seatbelt MACF policy initialized

.

PRB0_TAIL: 00000108

PRB0_HEAD: 00000108

PRB0_STRT: 02120000

PRB0_LEN: 00000003

SRB0_TAIL: 00000000

SRB0_HEAD: 00000000

SRB0_STRT: 02121000

SRB0_LEN: 00000003

PGTBL_CTL: 7ffc0001

PGTBL_ERR: 00000000

EXCC: 00000000

IPEIR: 00000000

IPEHR: 7d9d003f

INSTDONE: 7fffffc0

NOPID: 00000000

HWSTAM: ffffffff

IIR: 00000000

IMR: fffdff1f

ISR: 00000000

EIR: 00000000

EMR: ffffffff

ESR: 00000000

INSTPM: 00000000

INSTPS: 800f04ce

ACTHD: 00000108

MEM_MODE: 00000000

DMA_FADD_P 02120108

DMA_FADD_S 00000000

MI_STATE 00000040

PR0_STR: 00000000

PR0_END: 00000000

CCID0: 00000000

CTX_SIZE_PAL: 00000009

CTX_SIZE_NOPAL: 00000003

3D_CTL: 00000307

CHUNK_CTL: 00000000

HWB_TAIL: 00000000

HWB_HEAD: 00000000

HWB_START: 00000000

HWB_CTL: 003ff000

BIN_CTL: 00000000

BIN_SCENE: 00000000

BMP_BUFER: 00000000

BMP_GET: 00000000

BMP_PUT: 00000000

ICS: 00000003

ISP: 00000000

FSI: 00000000

SSI: 00000000

FST: 00000000

SST: 00000000

INTC: 00000000

IDLC: 00000000

SVDebug 000000ff

SHDebug 00080000

SFDebug 018781e0

SEDebug c0020080

WMDebug 04400040

IZDebug0 20000020

IZDebug0 00000000

PIDebug0 01807f80

PIDebug1 00000000

PIDebug2 01807f80

DIPDebug 05ef7f13

PLDebug 00000180

DGDebug 00000050

QCDebug 00000000

FTDebug 00000000

DMDebug 00000050

SCDebug 003c0000

FLDebug 3ef67040

BYDebug 00055440

PSDebug 00080000

CCDebug 882c1200

status[iSR]: 00000000

status[PRB0_HEAD]: 00000000

status[sRB0_HEAD]: 00000000

status[sRB1_HEAD]: 00000000

status[HWB_HEAD]: 00000000

status[sTAMP]: 00000000

AppleBCM5751Ethernet - en0 link active, 100-Mbit, full duplex, symmetric flow control enabled

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

new-host:~

gx620 usff

P4 3.4ghz

2gb DDR memory

Andy Vandijck · February 24, 2009

Hi,
I installed it on Ideneb 10.5.6, updated from 9.4.0 StageXNU kernel, it works great but I have these error messages, not sure what the mean, I would like to fix the "macmodel" error at the end too, any ideas?:

new-host:~$ sudo dmesg

Password:

npvhash=4095

hi mem tramps at 0xffe00000

PAE enabled

64 bit mode enabled

rtclock_init: Taking bus ratio path 4 (Intel / Apple)

TSC: Verification of clock speed PASSED.

TSC: Frequency = 3391.461597MHz, FSB frequency = 199.497741MHz, bus ratio = 17

Darwin Kernel Version 9.6.0: Wed Feb 18 15:39:36 CET 2009; AnV XNU Kernel Release 1.3 :xnu-1228.9.59/BUILD/obj/RELEASE_I386

standard timeslicing quantum is 10000 us

vm_page_bootstrap: 511611 free pages and 12677 wired pages

mig_table_max_displ = 79

AppleDecrypt: Copyright © 2008 Dale Walsh

AppleDecrypt: Starting...

AppleACPICPU: ProcessorApicId=1 LocalApicId=0 Enabled

AppleACPICPU: ProcessorApicId=2 LocalApicId=1 Enabled

AppleACPICPU: ProcessorApicId=3 LocalApicId=6 Disabled

AppleACPICPU: ProcessorApicId=4 LocalApicId=7 Disabled

Loading security extension com.apple.security.TMSafetyNet

calling mpo_policy_init for TMSafetyNet

Security policy loaded: Safety net for Time Machine (TMSafetyNet)

Loading security extension com.apple.nke.applicationfirewall

Loading security extension com.apple.security.seatbelt

calling mpo_policy_init for mb

Seatbelt MACF policy initialized

.

.

.

.

.

PRB0_TAIL: 00000108

PRB0_HEAD: 00000108

PRB0_STRT: 02120000

PRB0_LEN: 00000003

SRB0_TAIL: 00000000

SRB0_HEAD: 00000000

SRB0_STRT: 02121000

SRB0_LEN: 00000003

PGTBL_CTL: 7ffc0001

PGTBL_ERR: 00000000

EXCC: 00000000

IPEIR: 00000000

IPEHR: 7d9d003f

INSTDONE: 7fffffc0

NOPID: 00000000

HWSTAM: ffffffff

IIR: 00000000

IMR: fffdff1f

ISR: 00000000

EIR: 00000000

EMR: ffffffff

ESR: 00000000

INSTPM: 00000000

INSTPS: 800f04ce

ACTHD: 00000108

MEM_MODE: 00000000

DMA_FADD_P 02120108

DMA_FADD_S 00000000

MI_STATE 00000040

PR0_STR: 00000000

PR0_END: 00000000

CCID0: 00000000

CTX_SIZE_PAL: 00000009

CTX_SIZE_NOPAL: 00000003

3D_CTL: 00000307

CHUNK_CTL: 00000000

HWB_TAIL: 00000000

HWB_HEAD: 00000000

HWB_START: 00000000

HWB_CTL: 003ff000

BIN_CTL: 00000000

BIN_SCENE: 00000000

BMP_BUFER: 00000000

BMP_GET: 00000000

BMP_PUT: 00000000

ICS: 00000003

ISP: 00000000

FSI: 00000000

SSI: 00000000

FST: 00000000

SST: 00000000

INTC: 00000000

IDLC: 00000000

SVDebug 000000ff

SHDebug 00080000

SFDebug 018781e0

SEDebug c0020080

WMDebug 04400040

IZDebug0 20000020

IZDebug0 00000000

PIDebug0 01807f80

PIDebug1 00000000

PIDebug2 01807f80

DIPDebug 05ef7f13

PLDebug 00000180

DGDebug 00000050

QCDebug 00000000

FTDebug 00000000

DMDebug 00000050

SCDebug 003c0000

FLDebug 3ef67040

BYDebug 00055440

PSDebug 00080000

CCDebug 882c1200

status[iSR]: 00000000

status[PRB0_HEAD]: 00000000

status[sRB0_HEAD]: 00000000

status[sRB1_HEAD]: 00000000

status[HWB_HEAD]: 00000000

status[sTAMP]: 00000000

AppleBCM5751Ethernet - en0 link active, 100-Mbit, full duplex, symmetric flow control enabled

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

"macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

new-host:~

gx620 usff

P4 3.4ghz

2gb DDR memory

hmm... check seatbelt.kext to confirm it is the right version.

Remove /System/Library/Extensions.mkext to clear extensions cache.

Boot with -f

To fix that Mac model.

Open /Library/Prefences/SystemConfiguration/com.apple.Boot.plist

Under boot arguments set:

macmodel=MacPro2,1.

Reboot...

Onixs · February 24, 2009

Andy Vandijck

Nicer if you can just drop/take out the ""macmodel" boot argument not found, Mac model set as "MacBookPro5,1"

So it will be cleaner when it boot up.

AppleIIGuy · February 24, 2009

Am currently using Voodoo 10.5.5 just fine

Tried the XNU 10.5.6

Never got to the desktop or login window with to XNU 10.5.6 kernel

Feb 24 00:13:40 EX58-UD3R authorizationhost[519]: MechanismInvoke 0x1d15710 retainCount 2

Feb 24 00:13:40 EX58-UD3R SecurityAgent[520]: MechanismInvoke 0x5b5c270 retainCount 1

Feb 24 00:13:41 EX58-UD3R SecurityAgent[520]: NSSecureTextFieldCell detected a field editor ((null)) that is not a NSTextView subclass designed to work with the cell. Ignoring...

Feb 24 00:13:41 EX58-UD3R SecurityAgent[520]: MechanismDestroy 0x5b5c270 retainCount 1

Feb 24 00:13:41 EX58-UD3R loginwindow[46]: Login Window - Returned from Security Agent

Feb 24 00:13:41 EX58-UD3R authorizationhost[519]: MechanismDestroy 0x1d15710 retainCount 2

Feb 24 00:13:41 EX58-UD3R loginwindow[46]: USER_PROCESS: 46 console

Feb 24 00:13:41 EX58-UD3R sudo[527]: root : TTY=unknown ; PWD=/ ; USER=root ; COMMAND=/usr/libexec/StartupItemContext /usr/bin/open -a /Library/Application Support/iStat Server/DiskTool.app

Feb 24 00:13:41 EX58-UD3R com.apple.launchd[1] (com.apple.UserEventAgent-LoginWindow[516]): Exited: Terminated

Feb 24 00:13:42 EX58-UD3R com.apple.coreservicesd[76]: NOTE: Using non-mach-based version of client -> server communication, via direct function calls.

So had to go back to 10.5.5 Voodoo

mitch_de · February 24, 2009

Very interesting...
If this would solve every ones problems it would be great...

Now only figuring out where those caches are kept (NOTE: mkext file is deleted after kernel install...).

How about rebuilding using update_dyld_shared_cache binary?

That should auto rebuild these caches, right?

Maybe I should include in the installer script a method for deleting + updating dyld cache...

That way we would be sure it thoroughly rebuilds this...

Yeah, even it would not harm the system, if its only one of the last problems.

Those caches are really huge, so maybe really be an problem if some of the dyld code changed.

alien2xx · February 24, 2009

hmm... check seatbelt.kext to confirm it is the right version.
Remove /System/Library/Extensions.mkext to clear extensions cache.

Boot with -f

To fix that Mac model.

Open /Library/Prefences/SystemConfiguration/com.apple.Boot.plist

Under boot arguments set:

macmodel=MacPro2,1.

Reboot...

Thanks Andy, I will try it tonight, other than that it seems to be working really nice.

Thanks.

Naunaud · February 24, 2009

Hmmm that is pretty similar to my own laptop (Celeron M, the same amount + type of RAM).
Did any older version of my kernel work fine + which version of my core.

Soon I'll post a rebuilt version...

Did you try with vanilla dyld + my kernel?

It seems the new dylinker gives problems for some...

I'll continue working on the issues though...

I'll try your new rebuild when you'll put it online :wacko:

Azimutz · February 24, 2009

Ok.. i found most of the answers i was looking for.

-Remembered to run "TSC stability" test, from Voodoo's beta Test Plan, on "old" HPET and got the "old" 1 sec difference. No difference with the kext unloaded.

-The "new" HPET breaks my network.

Given this, i only find reasons to not use HPET. Until someone gives one good reason to use it, will leave it alone. Plus, Voodoo team already had droped this.



So Andy, as "kind of" a conclusion i have two suggestions:

-Make us a "real" port of the Voodoo kernel, i mean, without your mods :thumbsdown_anim: and keep your version, of course. That would be really helping the community and at the same time, do things your own way.

-If not in the mood, at least can you remove the "macmodel" thing!? It's the only thing that really bugs me. We can set it with your SMBIOS, Chameleonsm and on the future Chameleon.

It overrides Chameleonsm so, will it override the future Chameleon?

I tested "MacPro2,1" with Chameleonsm (using Voodoo kernel) and it works fine.. still fools GeekBench.

I don't have what it takes to do it my self and i still have that error when trying to compile via voodoobuild.sh.

As for your work in general, polemics apart, you did a nice job at porting to 9.6.0!

Works fine in my system, installing only the kernel or with the rest of the files.

The installer needs some work, though the last one works fine. I think most of the problems people are having is because of installing problems!? I found i don't have problems if i extract the files and install them with e.g. OSx86Tools but, i saw that many people are installing with Pacifist.. so, i installed the kernel,dyld and system.kext with Pacifist and got a real mess!! Kp at normal boot and even my rescue DVD (INSTAHackintosh made) wouldn't boot the system (hang). Had to boot the DVD, repair permissions on the system and then managed to boot the system with the DVD (took forever).. installed everything with OSx86Tools and it was fixed.

A suggestion.. try to build the installer with old PackageMaker 2.1.. maybe it helps.

Andy Vandijck · February 24, 2009

Yeah, even it would not harm the system, if its only one of the last problems.
Those caches are really huge, so maybe really be an problem if some of the dyld code changed.

It is possible since then the symbols are present at different locations in the binary.

Even just recompile with different compiler ops would change this in the binary...

Next release I'll make sure to do this...

Ok.. i found most of the answers i was looking for.
-Remembered to run "TSC stability" test, from Voodoo's beta Test Plan, on "old" HPET and got the "old" 1 sec difference. No difference with the kext unloaded.

-The "new" HPET breaks my network.

Given this, i only find reasons to not use HPET. Until someone gives one good reason to use it, will leave it alone. Plus, Voodoo team already had droped this.



So Andy, as "kind of" a conclusion i have two suggestions:

-Make us a "real" port of the Voodoo kernel, i mean, without your mods and keep your version, of course. That would be really helping the community and at the same time, do things your own way.

-If not in the mood, at least can you remove the "macmodel" thing!? It's the only thing that really bugs me. We can set it with your SMBIOS, Chameleonsm and on the future Chameleon.

It overrides Chameleonsm so, will it override the future Chameleon?

I tested "MacPro2,1" with Chameleonsm (using Voodoo kernel) and it works fine.. still fools GeekBench.

I don't have what it takes to do it my self and i still have that error when trying to compile via voodoobuild.sh.

As for your work in general, polemics apart, you did a nice job at porting to 9.6.0!

Works fine in my system, installing only the kernel or with the rest of the files.

The installer needs some work, though the last one works fine. I think most of the problems people are having is because of installing problems!? I found i don't have problems if i extract the files and install them with e.g. OSx86Tools but, i saw that many people are installing with Pacifist.. so, i installed the kernel,dyld and system.kext with Pacifist and got a real mess!! Kp at normal boot and even my rescue DVD (INSTAHackintosh made) wouldn't boot the system (hang). Had to boot the DVD, repair permissions on the system and then managed to boot the system with the DVD (took forever).. installed everything with OSx86Tools and it was fixed.

A suggestion.. try to build the installer with old PackageMaker 2.1.. maybe it helps.

Very first release of my 10.5.6 kernel was a base Voodoo port.

You can find it somewhere on the "Voodoo kernel is now released" part in the forum...

It did have Mac model override though...

And yes, if Chameleon 2 comes out, it will override the model as the kernel controls the end part of this... (plan on using their bootloader rather than mine when it comes out as it will be MUCH better than mine...).

As for manual install... if you do it proper it shouldn't be problem.

Also did a manual install on a friends AMD (he had messed up his System.kext, Libkern.kext was missing, had to revert to older one and fix the rest up manually after boot...).

Old packagemaker... are you kidding me? this is Leopard so it isn't any issue, besides... you can use the new one too to create old packages... (Tiger compliant, doesn't use flat packages...).

Azimutz · February 25, 2009

Very first release of my 10.5.6 kernel was a base Voodoo port.
You can find it somewhere on the "Voodoo kernel is now released" part in the forum...

It did have Mac model override though...

And yes, if Chameleon 2 comes out, it will override the model as the kernel controls the end part of this... (plan on using their bootloader rather than mine when it comes out as it will be MUCH better than mine...).

As for manual install... if you do it proper it shouldn't be problem.

Also did a manual install on a friends AMD (he had messed up his System.kext, Libkern.kext was missing, had to revert to older one and fix the rest up manually after boot...).

Old packagemaker... are you kidding me? this is Leopard so it isn't any issue, besides... you can use the new one too to create old packages... (Tiger compliant, doesn't use flat packages...).

Yap.. i have all of them As you mention, it has Mac model override.

Well, old PackageMaker or new one, they are both a big :blink: ! I started using old PackageMaker because of INSTAHackintosh.. packages made with new one, caused errors in the build process. So i sticked with it.. find it more simple and less buggy. That's why i mentioned it, not because of compatibility. Anyways, i have to mess with it so, i will probably get to a conclusion shortly on who performs best.

Maniac10 · February 26, 2009

Andy, I'm getting several app crashes lately due to dyld, happen with kernel v1.2 and your latest dyld. Also tested vanilla dyld with the same results. (I've used mainmenu to rebuild prebindings but didn't help.

It seems that street view from google maps always crash my safari.

Process:		 Safari [551]
Path:			/Applications/Safari.app/Contents/MacOS/Safari
Identifier:	  com.apple.Safari
Version:		 3.2.1 (5525.27.1)
Build Info:	  WebBrowser-55252701~1
Code Type:	   X86 (Native)
Parent Process:  launchd [123]

Date/Time:	   2009-02-25 22:06:10.999 -0200
OS Version:	  Mac OS X 10.5.6 (9G55)
Report Version:  6

Exception Type:  EXC_BAD_ACCESS (SIGSEGV)
Exception Codes: KERN_INVALID_ADDRESS at 0x0000000040dd4f71
Crashed Thread:  0

Thread 0 Crashed:
0   ???						   	0x206e6fc1 0 + 544108481

Thread 1:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x00c7fd0d 0 + 13106445
2   ???						   	0x060088ff 0 + 100698367
3   ???						   	0x05fc1065 0 + 100405349
4   ???						   	0x00c7f095 0 + 13103253
5   ???						   	0x00c7ef52 0 + 13102930

Thread 2:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x010100ae 0 + 16842926
2   ???						   	0x01010cd8 0 + 16846040
3   ???						   	0x0450aebe 0 + 72396478
4   ???						   	0x00c7f095 0 + 13103253
5   ???						   	0x00c7ef52 0 + 13102930

Thread 3:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x010100ae 0 + 16842926
2   ???						   	0x01010cd8 0 + 16846040
3   ???						   	0x01f7cd40 0 + 33017152
4   ???						   	0x01f197ed 0 + 32610285
5   ???						   	0x01f19394 0 + 32609172
6   ???						   	0x00c7f095 0 + 13103253
7   ???						   	0x00c7ef52 0 + 13102930

Thread 4:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x00c7f095 0 + 13103253
2   ???						   	0x00c7ef52 0 + 13102930

Thread 5:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x206e5774 0 + 544102260
2   ???						   	0x00c7f095 0 + 13103253
3   ???						   	0x00c7ef52 0 + 13102930

Thread 6:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x00cc5539 0 + 13391161
2   ???						   	0x204bd14f 0 + 541839695
3   ???						   	0x204d691f 0 + 541944095
4   ???						   	0x204bd5ff 0 + 541840895
5   ???						   	0x00c7f095 0 + 13103253
6   ???						   	0x00c7ef52 0 + 13102930

Thread 7:
0   ???						   	0x00c4ea17 0 + 12904983
1   ???						   	0x010100ae 0 + 16842926
2   ???						   	0x01010cd8 0 + 16846040
3   ???						   	0x01f4dd75 0 + 32824693
4   ???						   	0x01f59e94 0 + 32874132
5   ???						   	0x0005b9b0 0 + 375216
6   ???						   	0x01f197ed 0 + 32610285
7   ???						   	0x01f19394 0 + 32609172
8   ???						   	0x00c7f095 0 + 13103253
9   ???						   	0x00c7ef52 0 + 13102930

Thread 0 crashed with X86 Thread State (32-bit):
 eax: 0x20d60000  ebx: 0x206e6fa4  ecx: 0x206ee040  edx: 0x00000000
 edi: 0x00000000  esi: 0x20dd5d90  ebp: 0xbfffe479  esp: 0xbfffe2c0
  ss: 0x0000001f  efl: 0x00010246  eip: 0x206e6fc1   cs: 0x00000017
  ds: 0x0000001f   es: 0x0000001f   fs: 0x00000000   gs: 0x00000037
 cr2: 0x40dd4f71

Binary Images:
0x8fe00000 - 0x8fe298a3  dyld 0.0 (???) <f0cc7d2176a94a8215b771badeb6923a> /usr/lib/dyld

EDIT: Safari and Firefox crashes were related to the latest flash update, not java. Anyway, dyld is crashing some other apps as well.

mercurysquad · February 26, 2009

The AppleHPET.kext is needed only if you tend to be one of the unlucky ones with no HPET in the CPU core, which will mean that probeHPET fails (useoldhpet=1 is only needed in case of the need of this...).

[...]

The HPET mods require the System.kext to be replaced since it will have extra symbols inside.

If you have an older Intel/AMD cpu you need this for the old AppleHPET driver to work.

Otherwise it's no really big deal...

ANDY! I feel like shouting at you. LOL.

9.6 kernel has NO hpet — everything got removed. WHY are you still shoving it back in the kernel and asking people to do all sorts of gymnastics to get going?

PHICO · February 26, 2009

Pacifist the 1.3pkg,and only install mach_kernel&system.kext

looks fine

Arial · February 26, 2009

Andy, I'm getting several app crashes lately due to dyld, happen with kernel v1.2 and your latest dyld. Also tested vanilla dyld with the same results. (I've used mainmenu to rebuild prebindings but didn't help.
It seems that street view from google maps always crash my safari.
Process:		 Safari [551]
  Path:			/Applications/Safari.app/Contents/MacOS/Safari
  Identifier:	  com.apple.Safari
  Version:		 3.2.1 (5525.27.1)
  Build Info:	  WebBrowser-55252701~1
  Code Type:	   X86 (Native)
  Parent Process:  launchd [123]

  Date/Time:	   2009-02-25 22:06:10.999 -0200
  OS Version:	  Mac OS X 10.5.6 (9G55)
  Report Version:  6

  Exception Type:  EXC_BAD_ACCESS (SIGSEGV)
  Exception Codes: KERN_INVALID_ADDRESS at 0x0000000040dd4f71
  Crashed Thread:  0

*****
  Binary Images:
  0x8fe00000 - 0x8fe298a3  dyld 0.0 (???) <f0cc7d2176a94a8215b771badeb6923a> /usr/lib/dyld

same happen with v1.3.

Crash Report:

MacSysInfo http://paste2.org/p/154344

PPF-O-Matic http://paste2.org/p/154342

PPFMaster http://paste2.org/p/154343

restore backuped dyld to resolve.

I think install kernel & system.kext only are safe for everyone.

Andy Vandijck · February 26, 2009

same happen with v1.3.

Crash Report:

MacSysInfo http://paste2.org/p/154344

PPF-O-Matic http://paste2.org/p/154342

PPFMaster http://paste2.org/p/154343

restore backuped dyld to resolve.

I think install kernel & system.kext only are safe for everyone.

Hmmm... all translate (PPC) initiated apps...

I should look into this.

Andy Vandijck · March 2, 2009

same happen with v1.3.

Crash Report:

MacSysInfo http://paste2.org/p/154344

PPF-O-Matic http://paste2.org/p/154342

PPFMaster http://paste2.org/p/154343

restore backuped dyld to resolve.

I think install kernel & system.kext only are safe for everyone.

I've done some thorough tests on Rosetta.

On my system everything works great.

On my friends AMD machine also.

Seems to be related only to certain computers...

Research continues...

Maniac10 · March 2, 2009

Some X86 apps crashes, so in my PC it's not related to PPC apps only.

Andy Vandijck · March 2, 2009

Some X86 apps crashes, so in my PC it's not related to PPC apps only.

I have had such reports from some people.

It is dyld related.

You can see it can't map the symbols down in some cases...

In your case it is the dyld built-in into the core.

Haven't had time to rebuild it with regular GCC 4 yet...

AnV XNU Kernel V1.4

AnV Kernel reliable 211 members have voted

1. Does the kernel work fine for you?

2. Did the installer work fine for you?

3. Which cpu do you use?

Recommended Posts

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

AnV Kernel reliable
211 members have voted