Mavericks kernel testing on AMD (formerly Mountain Lion kernel testing on AMD)

[spakk]

@Andy,

what can then the cause, be your Andy opinion ?
therefore , it is independent of which boot flags use, the system stops always at the same place .
why do not start the new kernel with your previous patches ?

 

@all

what do you think, what could have this cause ?. I will let the tests for some time. Since it had achieved no positive results.
Install a system that I'm blind .... give me some computer and I'll make in a short time an executable hackintosh ..
I'm pretty sure that is not a problem of the user!

[carlo_67]

guys this thing can not be 

you do a clean install using Pandora and retail 10.9 

then combined update 10.9.2 

using rc4 kernel of Bronya 

customized so

[MariusT]

I am very interested to know what systems work with which current kernel. Specifications with brief description of your hardware and a brief description of the measures you have taken so that your system can boot successfully. Thx

 

 

System hardware is listed in the signature.

 

What have done so far?

 

Installed 10.9.0 by m.y.H.a.c.k method using Bronya_rc8 kernel, no kext blacklist, no corecrypto, no patched system.kext,  touched by M.u.l.t.i.b.e.a.s.t with Nullcpupowermanagement, ACPIrollback to 10.6.x, Clover bootloader with npci=0x2000 bootflag, modified AMD kext for my graphics

 

updated to 10.9.2 by combo update, reboot using m.y.H.a.c.k. USB stick, copy the new kernel_fx_rc2 to root, repair permission, replace the IOFamily.kext and AppleAPCIxxxx.kext from 10.9.0, repair permission, reboot with -v -f npci=0x2000, and its done.

[Redbyte]

I don't know exactly the kind of logs you guys need to better the kernel development but I am just reporting all worked good with a fresh myHack install of 10.9.2 with bronya rc3/4 on my FX-8350 without any kext rollbacks. But for some reason my FX-8150 build is extremely slow and will give me a black screen saying its killing processes then return to the desktop, it was doing this with all other kernels from 10.9.0-10.9.1. I even tried swapping GPUs on the two but that made no difference.

Either way guys keep up the good work!

[Gismo]

Hi Redbyte,

 

try this on FXCPU http://www.insanelymac.com/forum/topic/281450-mavericks-kernel-testing-on-amd-formerly-mountain-lion-kernel-testing-on-amd/?p=2019324

 

and this http://www.insanelymac.com/forum/topic/281450-mavericks-kernel-testing-on-amd-formerly-mountain-lion-kernel-testing-on-amd/?p=2019475

 

Good Luck

[tragediana150]

So you telling that Bronya_kernels works on fresh installation of Mavericks?

It's soooo weird, I installed my Mavericks 2months ago and install only Logic Pro and Pro Tools and nothing (so it is fresh version of Mavericks).

Why these kernels doesn't works on our hackintosh? Why we need to install fresh OSX? HMM?

 

P.S. I do not grumble, just ask, because I don't understand it Anyone can explain me what is going on?

[Redbyte]

@Gismo Thanks, Replacing the kext fixed the speed problem on my 8150

 

@tragediana150 before the latest kernels from Bronya I needed to rollback a few kexts (Sandbox, System, and applemach) for it to boot on either of my rigs but the 8150 rig was unusably slow. Now with the new kernels from Bronya I only have to rollback on the one rig to get the speed back up.

 

Also the only reason I test with a fresh install is to keep from being any chance of a kext or something being out of place. Its just old habbit from all my years of working on Windows.

 

Thanks guys

[lpukraine]

Just reboot at the start with every kernel((

[carlo_67]

I then did another test 

I installed 10.9 mavericks upgrading to 10.9.2 using the old kernel bsa 

 

Then I did a test by replacing the kernel of the new rc4 Bronya and replacing the kext 

posted, 

the system has started arriving at the desk 

 

I do not know what the problem is for you

kext.zip

[YungRaj]

any luck on 10.9.3 on any amd cpus? just wondering since it hasn't been discussed yet in this topic

[Eweie]

people are probably busy working to spend time on this. I know I am but ironically i came to see people's results too lol.

Ill probably have time to update tonight after work.

 

however spakk said the current working kernels worked on the betas of 10.9.3 so it's probably going to work with not much issue.

[tragediana150]

I tried to install fresh mavericks and update to 10.9.3 and it's not working for me with the newest kernels. I forgot to try with older kernels

[MariusT]

Updated to 10.9.3 through Appstore, both Mach_fxrc2 n rc3 boots well

[spakk]

I tried to install fresh mavericks and update to 10.9.3 and it's not working for me with the newest kernels. I forgot to try with older kernels

Hi tragediana,

some here present means they have the absolute wisdom in itself. That is the fact that is even though the same hardware components to behave differently hackentoshs not understood. I am sure that something need still be adapted to the new kernel. What is exactly adjust, I do not know. It's me at the moment but do not care. I will continue to use the kernel based on 10.9.0 and make no other changes to my executable provisional Mavericks 10.9.3 . The development and test phase of the kernels runs my opinion, not like the previous kernel developments from constructive. This is my own opinion,

[eight]

Updated to 10.9.2 and then rolled back system and sandbox kexts so I could boot. Everything runs fine, except I get random opemu errors in the console and the cpu goes into full load or sometimes just freezes.

 

The error is the following  kernel[0]: OPEMU:  mov [rdx+0x10], rcx 

 

It happens with every kernel including Andy's (with his system.kext and without), Sinetek and Caasa in both 10.9.0 and 10.9.2. What do you guys think?

[spakk]

Fix the permissions then try to boot again

[carlo_67]

I tried to install fresh mavericks and update to 10.9.3 and it's not working for me with the newest kernels. I forgot to try with older kernels

be 'do other tests because the kernel Bronya rc4 acts well with 10.9.3 realase 

out yesterday and already installed on AMD

Below the kext replaced 

PS: After installed the kext, repair permissions from disk utility from the usb installer, kext wizard or kext utility does not repair them well 

 

Also try them on 10.9.2

kext.zip

[spakk]

This is only an information:
To all who have problems with the new Mavericks kernel 10.9.2 (xnu-2422.90.20) may continue to use the previous kernels based on Mavericks 10.9.0 (xnu-2422.1.72). the installation of the new Update 10.9.3 Build 13D65 went back without any problems.

 

Edit:

For the Update Build 13D65 must be installed Build 13D61 on the partition ! otherwise the installation will not work.

 

Edited May 16, 2014 by spakk

[Bronya]

hi , spakk , is it what kernel ? 

[spakk]

Hi Bronya,

 

There are all previous BSA kernel ( only with bsa_r9 I have problems) mach_kernel_beta_rc4_fix, mach_kernel_rc4, cavva2 and the kernel of sinitek. I have tested them all on 10.9.3. Build 13D65 all do good work!

[Kio​]

Hi Bronya,

 

There are all previous BSA kernel ( only with bsa_r9 I have problems) mach_kernel_beta_rc4_fix, mach_kernel_rc4, cavva2 and the kernel of sinitek. I have tested them all on 10.9.3. Build 13D65 all do good work!

Hi spakk! Do you had problems using these kernels with the latest version of flash player (13.x)? (artifacts, green lines, etc?)

[spakk]

My guess is Apple did made ​​a small change in the source code partly which caused the problems When booting from various CPUs the kernel panic comes with "extensions in backtrace, com.apple.kec.corecrypto. (1.0) ...." Unfortunately , I still get an error when i compile the kernel, unfortunately so far failed to stop the error of each attempt. Now I'm perplexed what It could be.

 

@ Bronya, i like to put the error message from my last kernel building and posting here, in the hope that you can did give me a hint where I have to start to fix this error. I am waiting for your answer.

 

Should it not be the kernel, of what would be a cause of the problem your opinion? Please see my last kernel panic error message, what I have posted here.

"Panic (cpu 0 calles 0xffffff80002f6ae4): kernel stack memory corruption detected ...........??????? What does that mean?

Hi spakk! Do you had problems using these kernels with the latest version of flash player (13.x)? (artifacts, green lines, etc?)

Hi kio,

No absolutely no problems! Everything is almost perfect! All tools and adobe tools work smoothly, no graphic problems, no flash problems absolutely nothing. I can prove through videos you like. The only thing not working is the internal sound card with AppleHDA.kext. but that is secondary because I use a very good external sound card.

[Kio​]

What kernel used by default? Some direct link? I have tried several kernels and they all give me the same problem with flash and somes adobe apps. (PSCS6, PSCC...)

[spakk]

I use my DSDT.aml (also tested with a customized Ssdt.aml) a efi-string and my graphic-rom file is possibly the reason why I have no graphics problems.
Now here are some of my, what does this spinner trying to say to us. Efi-string and rom file and a DSDT.aml ...... It is so

Edit:

 

@kio,
As I have already described above, I can use almost any kernel based on 10.9.0. all work almost perfectly on my amd hackintosh. I can not make one bad or highlight. The only kernel does not work with my hardware is bsa_r9.

 

That is the reason why I can not understand why the new kernel based on 10.9.2 requires only does his work. However, it is important to note major changes under s / l / e must be made with the new kernel! As with the older kernel.

I find it no use and no logic. 

I think it is not a user problem! As I have already described several times I have studied nothing left 

[spakk]

I would again like something important again to make it clear here. I want to with my previous criticisms of the current kernel in any case the great work of Bronya, andy, sinetek and others here. It is not my intention here represent the great work looking for or do bad. it's all about me here to make a constructive criticism which contribute very important for the next steps to create a good runnable amd kernel.