Jump to content

Lion kernel testing on AMD (don't ask help here: use the Help Topic)


ham4ever
 Share

613 posts in this topic

Recommended Posts

FSB detected correctly with RC12 and RC8 in this system

npvhash=4095
PAE enabled
64 bit mode enabled
CPU Model: model 0x2
Multiplayer detection: family 0x15, model 0x2, stepping 3, extFamily 0x1, extModel 0x0
FSB Detection: calculated Mult 0, cpuFreq 0rtclock_init: Phenom MSR 0xc0010071 returned: 0x8000012400240e
BUS: Frequency = 200.0000MHz, cvtt2n = 00000005.00000000, cvtn2t = 00000000.33333333, cvtInt = 004C4B3F.FFFE0DD1
TSC: Frequency = 3000.0000MHz, cvtt2n = 00000000.55555555, cvtn2t = 00000003.00000003, gran = 15
Darwin Kernel Version 11.4.2: \M-Q\M^G\M-P\M-5\M-Q\M^B\M-P\M-2\M-P\M-5\M-Q\M^@\M-P\M-3, 4 \M-P\M-0\M-P\M-?\M-Q\M^@\M-P\M-5\M-P\M-;\M-Q\M^O 2013 \M-P\M-3. 23:28:35 (MSK); BronzovkaAnVoodoo_Sinetek###:xnu-1699.32.7/BUILD/obj//RELEASE_I386
vm_page_bootstrap: 1017751 free pages and 22633 wired pages
zone leak detection enabled
standard timeslicing quantum is 10000 us
mig_table_max_displ = 73
CPU identification: AMD Athlon(tm) 7850 Dual-Core Processor
CPU features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH MMX FXSR SSE SSE2 HTT SSE3 MON CX16 POPCNT
HTT: 2 cores per package; 2 logical cpus per package
CPU extended features: SYSCALL XD 1GBPAGE EM64T LAHF RDTSCP TSCI

npvhash=4095
PAE enabled
64 bit mode enabled
CPU Model: model 0x2
Multiplayer detection: family 0x15, model 0x2, stepping 3, extFamily 0x1, extModel 0x0
rtclock_init: Phenom MSR 0xc0010071 returned: 0x8000012400240e
BUS: Frequency = 200.0000MHz, cvtt2n = 00000005.00000000, cvtn2t = 00000000.33333333, cvtInt = 004C4B3F.FFFE0DD1
TSC: Frequency = 3000.0000MHz, cvtt2n = 00000000.55555555, cvtn2t = 00000003.00000003, gran = 15
Darwin Kernel Version 11.4.2: Tue Feb 26 14:33:29 MSK 2013; BronzovkaAnVoodoo_Sinetek###:xnu-1699.32.7/BUILD/obj//RELEASE_I386
vm_page_bootstrap: 1017751 free pages and 22633 wired pages
zone leak detection enabled
standard timeslicing quantum is 10000 us
mig_table_max_displ = 73
CPU identification: AMD Athlon(tm) 7850 Dual-Core Processor
CPU features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH MMX FXSR SSE SSE2 HTT SSE3 MON CX16 POPCNT
HTT: 2 cores per package; 2 logical cpus per package
CPU extended features: SYSCALL XD 1GBPAGE EM64T LAHF RDTSCP TSCI

nowadays, looking at those output strings, seems that something changed between them

it would be nice to see the source code diffs maybe i can try to see something..

RC8 is the best build i have tested. with RC12 various programs crashes with EXC_BREAKPOINT exceptions

Edited by bcobco
Link to comment
Share on other sites

@bcobco and happel,

Am I right to say you are both using Athlon processors? ADHDMedia and I are both using Phenom II x6 processors,the detection may be different as I have audio problems using bron kernels rc8 and rc12 but with shane_kernel audio works flawlessly using flag busratio=14

Edited by clemj06
Link to comment
Share on other sites

I am running 10.6.7 and 10.7.5, but on separate drives, and I checked both plists and neither has any busratio or fsb flags (I've never had to use them with SL)

 

Is there anything else that would cause me to not be able to set the busratio with a flag when I boot with the Bronzovka kernels? Besides it being hardcoded into the kernel...?

I figured it out, go into your bios and disable CPU turbo...Mine is usually enabled, bringing my processor speed to 3.31GHz, as can be seen in the screenshot mine is currently on the stock speed of 2.81GHz, so bron_rc12 kernel is properly detecting the busratio now....

 

For those having audio stuttering problems due to incorrect bus ratio:

Disable AMD Turbo boost

post-1122366-0-13051400-1365701917_thumb.png

Edited by clemj06
  • Like 2
Link to comment
Share on other sites

 

I figured it out, go into your bios and disable CPU turbo...Mine is usually enabled, bringing my processor speed to 3.31GHz, as can be seen in the screenshot mine is currently on the stock speed of 2.81GHz, so bron_rc12 kernel is properly detecting the busratio now....

 

For those having audio stuttering problems due to incorrect bus ratio:

Disable AMD Turbo boost

 

I don't have that as an option in my BIOS. I am, however , manually overclocked from 3.01 to 3.51 GHz. Could that be my issue? It never has been under SL...

Link to comment
Share on other sites

I don't have that as an option in my BIOS. I am, however , manually overclocked from 3.01 to 3.51 GHz. Could that be my issue? It never has been under SL...

Remove the overclock, test it on the stock speed...should work properly...

You do have the ability to save your current bios settings right? If you can, save them and remove the overclock...

If you require any additional help, we should continue in the help topic

Edited by clemj06
  • Like 1
Link to comment
Share on other sites

 

Remove the overclock, test it on the stock speed...should work properly...

You do have the ability to save your current bios settings right? If you can, save them and remove the overclock...

If you require any additional help, we should continue in the help topic

 

Tested that theory, and its a no-go for me. Even under stock clock speeds, my bus ratio is still detected as 17, throwing everything off.

Link to comment
Share on other sites

nevew saw that.. maybe my bios does not have it, or amd kuma does not have it

 

i would like to suggest some performance tips:

disable all ASPM

disable C1E

disable everything related with power save

remove applegraphicspowermanagement and all applepowermanagement kexts

Link to comment
Share on other sites

:)

 

for CPU FX / Phenom / Atlon ? i386 / x86_64 ?

 

je vais le tester ;)

 

232333Capturede769cran20120310a768093626.png

 

808873Capturede769cran20120310a768094939.png

 

553186Capturede769cran20120310a768095344.png

 

397168Capturede769cran20120310a768095700.png

 

-chamelon 1820 trunck

-flag : i386

-graphics card : MSI GT 610 1024

x86_64 not work on Nvidia :(

Edited by gils83
Link to comment
Share on other sites

:)

 

for CPU FX / Phenom / Atlon ? i386 / x86_64 ?

 

je vais le tester ;)

 

232333Capturede769cran20120310a768093626.png

 

808873Capturede769cran20120310a768094939.png

 

553186Capturede769cran20120310a768095344.png

 

397168Capturede769cran20120310a768095700.png

 

-chamelon 1820 trunck

-flag : i386

-graphics card : MSI GT 610 1024

x86_64 not work on Nvidia :(

 

Hi, is restarting working for you?

Link to comment
Share on other sites

Download mach_kernel_rc13!

Fix cupid for old processor's .

I add flag "busratio=x" , x - figure.

It is you can write flag busratio !

Please try !

 

RC13 is giving me an error on launching Cinebench. Other previous versions would cause the red pixel error, but this is the first kernel that has caused it to refuse to open completely. RC8 seems to be running smooth as silk on my system now, however.

 

Thanks for all your hard work!! Much appreciated.

 

Not sure if this is related to the kernel, but some of my .mov files are badly corrupted, and show a green or black thumbnail. Others play fine. This is the same systemwide, be it thru Quicktime Player, FCPX, MPlayerX, iTunes, etc...

 

Is this a kernel issue? Or something I should be asking in the Help Topic...? Thanks guys.

Link to comment
Share on other sites

Download mach_kernel_rc13!

Fix cupid for old processor's .

I add flag "busratio=x" , x - figure.

It is you can write flag busratio !

Please try !

 

 

Thanks

 

I get no issue with fsb & busratio detection

 

 

but I get some different readouts of the SSSE3 & SSE4.2 detection

 

Here is the report:

 

 

 

 

sysctl: SSSE3 = 1 & SSE4.2 = 0

 

cpu_1a.png

 

 

 

dmesg: SSSE3 = 0 & SSE4.2 = 0

 

cpu_1b.png

 

 

 

sysctl: SSSE3 = 0 & SSE4.2 = 1

 

cpu_1c.png

 

 

CPU-X : SSSE3 = 0 & SSE4.2 = 1

 

cpu_1d.png

 

 

 

 

  • Like 1
Link to comment
Share on other sites

hi bronya

congrats! rc13 kernel is pretty stable

cpu: amd athlon 64 x2 7850 be

 

somewhere opemu has a bug!

i dont know how can i get more information of what is happening. nowadays everybody can reproduce the bug:

launching Maxima (Maxima, a Computer Algebra System. http://maxima.sourceforge.net/), or trying to compile it with HomeBrew (The missing package manager for OS X. http://mxcl.github.io/homebrew/). seems that the issue happens when it calls SBCL (Steel Bank Common Lisp. http://www.sbcl.org/)

the affected thread will not respond, and the kernel will enter a loop, printing "32 bit opemu wrong instruction" or "64 bit opemu wrong instruction".

 

ps: if there is a way to get more information of opemu please let me know so as we can obtain moar information

Edited by bcobco
  • Like 1
Link to comment
Share on other sites

 Share

×
×
  • Create New...