Jump to content

Private alpha test of Voodoo 9.5 Kernel

mercurysquad

By mercurysquad,
in New Releases and Updates

 Share
214 posts in this topic

Recommended Posts

xaiev
xaiev
Posted
Posted

Hi there,

 

Acer aspire 1620

 

Intel P4 HT northwood 3,06GHz

multiplier : 15x, bus speed: 200MHz, FSB: 800MHz

 

 

Really want to help, but i need some explanations:

 

1) which release do I need to download for this kernel to work, jas 10.5.4?

 

2) then how do I change for the voodoo kernel since my system will probably not boot with kernels provided on the DVD?

 

Cheers from France  ;)

Link to comment
Share on other sites
RobNyc
RobNyc
Posted
Posted

DFI LANPARTY UT NF4 SLI-DR EXPERT MOBO

DUAL CORE AMD OPTERON 165 @ 2.7ghz

But OSX shows as 1.75ghz =\ .. Stock speed are 1.8ghz

Without any flags. So far no browser crashes uptime 40mins

 

 

Even though usually my browsers crashes, so I use cpus=1 on 10.5.4 with 9.4.0 kernel.

CPUID_Log.txt

Link to comment
Share on other sites
ritalin
ritalin
Posted
Posted
Hi can you also post, when it panics, on screen you should see "Taking bus ratio path X" - can you post which path it is taking?

Are you sure this is with no bootflags except maxmem and -v?

The rtclock_init says "Taking bus ratio path 0 (Boot-time argument)" !

This will only appear if you specify busratio=xx option.

 

Also, your kernel timestamp says 03:22:36 IST -- for alpha 13 this should be 05:59:23 IST.

Check md5: a645413c43e6d315ff69e45279f7a9a3

 

Yes, I cut and pasted from logs but cut the wrong piece. I spotted the mistake and edited my post but you were too quick and read my unedited post.

 

Here is the correct info.

 

a11 With no boot flags

rtclock_init: Taking bus ratio path 3 (AMD Phenom)

 

 

 

>>> The real-time clock was not properly initialised on your system!

Use the busratio=xx flag or upgrade to a newer version of chameleon

 

TSC: Frequency = 98.223271MHz, bus frequency = 98.223271MHz, multiplier = 1

Darwin Kernel Version 9.5.0: Wed Sep 24 03:22:36 IST 2008; Voodoo; based on ToH:xnu-1228.7.58/BUILD/obj/RELEASE_I386

 

 

a11 With boot flags: busratio=11 fsb=200000000

rtclock_init: Taking bus ratio path 0 (Boot-time argument)

TSC: Frequency = 2200.0000MHz, bus frequency = 200.0000MHz, multiplier = 11

Darwin Kernel Version 9.5.0: Wed Sep 24 03:22:36 IST 2008; Voodoo; based on ToH:xnu-1228.7.58/BUILD/obj/RELEASE_I386

 

 

a13 With no boot flags

rtclock_init: Taking bus ratio path 3 (AMD Phenom)

rtclock_init: Phenom MSR 0xc0010071 returned: 0x30003806

TSC: Frequency = 2160.930508MHz, bus frequency = 98.224114MHz, multiplier = 22

Darwin Kernel Version 9.5.0: Thu Sep 25 05:59:23 IST 2008; Voodoo; based on ToH:xnu-1228.7.58/BUILD/obj/RELEASE_I386

Link to comment
Share on other sites
lemon
lemon
Posted
Posted
People getting kernel panic, can you please post the KP message also so I know *what* the panic is about :P thanks.

 

 

=) Hello mercurysquad,

I dont have KPs but I have massive problems with sound

I did the fsb=20000000 ratio=15 option at startup, but still there is massive shuttering, digital distortion and audio failing while in use with for, example Garageband.

 

It looks like the audio driver is getting kicked out by the system.

if I change the sample rate or the bufffersize it is coming back untill iI stop and start a song again.

 

It could be the dribver of that device Im using a Firewire TC Electronics 24D.

 

if you let me know how to get the info for you Id be glad to do so.

LeMoN =)

Link to comment
Share on other sites
twitterfire
twitterfire
Posted
Posted

Could you please look into raw sockets/mac spoofing issue? As my ISP expects the MAC address to match the registered value and I don't have the old nic, I will higly appreciate this. Thank you guys.

Link to comment
Share on other sites
mercurysquad
mercurysquad
Posted
  • Author
Posted
=) Hello mercurysquad,

I dont have KPs but I have massive problems with sound

I did the fsb=20000000 ratio=15 option at startup, but still there is massive shuttering, digital distortion and audio failing while in use with for, example Garageband.

 

It looks like the audio driver is getting kicked out by the system.

if I change the sample rate or the bufffersize it is coming back untill iI stop and start a song again.

 

It could be the dribver of that device Im using a Firewire TC Electronics 24D.

 

if you let me know how to get the info for you Id be glad to do so.

LeMoN =)

 

Does booting without any bootflag help?

 

The fsb is usually not a nicely rounded figure, so specifying 200000000 will boot the computer up but timing will still be very slightly off - and with audio/usb/firewire this causes issues.

 

The easy way to check your fsb is to boot with busratio=0 -- this will force a KP and you can check your busFreq in the message that is displayed.

 

Also please post your cpu log by downloading the cpuid tool from googlecode download section, and create a new Issue there attaching your logs so we can keep it in mind during beta bugfix phase.

 

Could you please look into raw sockets/mac spoofing issue? As my ISP expects the MAC address to match the registered value and I don't have the old nic, I will higly appreciate this. Thank you guys.

 

Most drivers allow changing MAC address directly: just try 

sudo ifconfig en0 ether x:x:x:x:x

replacing with whatever MAC address you want.

Link to comment
Share on other sites
lemon
lemon
Posted
Posted
Does booting without any bootflag help?

 

The fsb is usually not a nicely rounded figure, so specifying 200000000 will boot the computer up but timing will still be very slightly off - and with audio/usb/firewire this causes issues.

 

The easy way to check your fsb is to boot with busratio=0 -- this will force a KP and you can check your busFreq in the message that is displayed.

 

Also please post your cpu log by downloading the cpuid tool from googlecode download section, and create a new Issue there attaching your logs so we can keep it in mind during beta bugfix phase.

Most drivers allow changing MAC address directly: just try 

sudo ifconfig en0 ether x:x:x:x:x

replacing with whatever MAC address you want.

 

No flags produce the same problem.

 

I found a article about the Pentium D930

 

http://xtreview.com/review51.htm

there is a screenshot of CPU X which shows what you say about

the nicely round figure

 

it says :

 

Core speed = 3010.6

Multiplier = 15X

FSB = 200.7

Busspeed = 802.8

 

I guess thats what you re talking about.

Now, is that applying to any or only the one they testet ?

 

thanks for the replay

 

here my Cupid log

 

 

 

Edit:

 

In the log it says busfrequency is

========================

9/26/08 4:04:23 AM kernel 64 bit mode enabled

9/26/08 4:04:23 AM kernel

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

9/26/08 4:04:23 AM kernel TSC: Frequency = 2999.825941MHz, bus frequency = 199.988396MHz, multiplier = 15

========================

also is t says :

( TC Near is the Firewire Audio device)

========================

 

 

9/26/08 4:05:13 AM kernel TCNear::RestartStreaming

9/26/08 4:05:13 AM kernel TCNear: rg = 30, prg = 12, sg = 3, psg = 12

9/26/08 4:05:13 AM kernel TCNear: SafetyOffset = 96

9/26/08 4:05:13 AM kernel TCNear: SampleFramesPerBuffer = 8192

9/26/08 4:05:13 AM kernel TCNear: In Stream 00 Buffer = 0x6c64e000, Size = 524288

9/26/08 4:05:13 AM kernel TCNear: Out Stream 00 Buffer = 0x6c6ce000, Size = 524288

9/26/08 4:05:13 AM kernel TCNear::ReadFWBuffers dbc 2, received: 00a8, expected: 0000

 

and

 

9/26/08 4:05:15 AM kernel TCNear: Caliberation done, offsetToApply = 0000020a!

9/26/08 4:05:17 AM kextd[10] writing kernel link data to /var/run/mach.sym

9/26/08 4:05:23 AM pploader[138] Failed to get admin execute rights: 4294907290

 

9/26/08 4:05:43 AM kernel TCNear: Skipped, strm = 0, grp = 1, expected = 0000b098, actual = 0000b09c

9/26/08 4:05:43 AM kernel TCNear: Skipped, strm = 0, grp = 0, expected = 0000b438, actual = 0000b43b

9/26/08 4:05:43 AM kernel TCNear: Caliberation done, offsetToApply = 00000117!

9/26/08 4:05:52 AM kernel TCNear: Skipped, strm = 0, grp = 2, expected = 0000cea7, actual = 0000ceab

9/26/08 4:05:52 AM kernel TCNear: Caliberation done, offsetToApply = 0000005e!

=========================

 

Its doing the same also with the

busfrequency = 199988396

CPUID_Log.txt

Link to comment
Share on other sites
Alex HQuest
Alex HQuest
Posted
Posted

My OC'ed Pentium 4 HT 3GHz running on Voodoo 9.5 Alpha13:

 

9/25/08 10:52:23 PM kernel hi mem tramps at 0xffe00000

9/25/08 10:52:23 PM kernel PAE enabled

9/25/08 10:52:23 PM kernel rtclock_init: Taking bus ratio path 4 (Intel / Apple)

9/25/08 10:52:23 PM kernel TSC: Frequency = 3454.43010MHz, bus frequency = 230.269534MHz, multiplier = 15

9/25/08 10:52:23 PM kernel Darwin Kernel Version 9.5.0: Thu Sep 25 05:59:23 IST 2008; Voodoo; based on ToH:xnu-1228.7.58/BUILD/obj/RELEASE_I386

Link to comment
Share on other sites
mercurysquad
mercurysquad
Posted
  • Author
Posted
No flags produce the same problem.

<snip>

 

Hi,

It looks like it's just a poorly designed driver .. Your cpu is pentium D 9xx series, and so the kernel is taking the stock Intel/Apple path exactly like vanilla kernel. The fsb and multiplier are being detected properly.

 

I would suggest you try running garageband with the onboard audio card and see if you still get stuttering etc.

Link to comment
Share on other sites
lemon
lemon
Posted
Posted
Hi,

It looks like it's just a poorly designed driver .. Your cpu is pentium D 9xx series, and so the kernel is taking the stock Intel/Apple path exactly like vanilla kernel. The fsb and multiplier are being detected properly.

 

I would suggest you try running garageband with the onboard audio card and see if you still get stuttering etc.

 

I tryed that , but the onboard Audio driver would no work or load with the 9.5 voodoo kernel.

Link to comment
Share on other sites
moes
moes
Posted
Posted

Testing the kernel here with boot123.

Did some test with xbench....the outcomes vary a lot among the kernels....alpha scored 50% lower?!!

I cant see much difference btw....accept the spinners and and beachball running much faster(twice) when i didn't use -f .

 

Can the outcome of xbench1.3 be trusted ...using it with different kernels?

 

Will drop the file soon.....first finish testing.

 

Moes

Link to comment
Share on other sites
mercurysquad
mercurysquad
Posted
  • Author
Posted
hi there, still kernel panic here on dell vostro 400

Please, post what kernel panic you got. I have NO way to even identify the problem without any information.

 

I tryed that , but the onboard Audio driver would no work or load with the 9.5 voodoo kernel.

Maybe borrow a friend's USB soundcard or something? The idea is the test the app on another audio device which is known to work well in OS(x86) so we can pinpoint whether it's an audio driver issue or kernel issue.

 

Testing the kernel here with boot123.

Did some test with xbench....the outcomes vary a lot among the kernels....alpha scored 50% lower?!!

I cant see much difference btw....accept the spinners and and beachball running much faster(twice) when i didn't use -f .

 

Can the outcome of xbench1.3 be trusted ...using it with different kernels?

 

This can happen when there are timing issues, for example the kernel detects your fsb or bus ratio incorrectly. In this case system will have wrong notion of time so results can vary a lot. I've taken care of almost all types of CPUs right from original P4 up to latest AMDs. Older alphas might very well give wrong results but alpha13 onwards should have perfect timing.

 

Tested on Celeron D 320 sleep works and system is faster now, I have only one problem in system profiler Model Identifier is blank it was MacPro3.1 before that's all :)

Thanks

It's because older ToH (and other) kernels hardcoded the model identifier, but this kernel does not. I recommend usin an EFI bootloader along with vanilla AppleSMBIOS.kext and Chun-nan's AppleSMBIOSEFI.kext.

 

Workes perfectly on a P4HT 3.2ghz i think the name is 630 on boot-132

 

still cant get it working on a pentium m 9.4 does work when no using Stagexnu kernel

 

There's no way it won't work on a pentium M- I have a pentium M and I make damn sure it works not only well on it, but better than all kernels. Could you post a little more information about what problem you are having with it on the pentium M system?

Link to comment
Share on other sites
Nick14
Nick14
Posted
Posted

i get kernel panic i tried reinstalling it 7-8 times with 2 diffrent disc Deneb and JaS 10.5.4 but i've giving up on this computer it's probly my pc fault if it worked on your pc

 

 

i'm using this kext on my vostro c2d does it act like vanillia ??

Link to comment
Share on other sites
mercurysquad
mercurysquad
Posted
  • Author
Posted

Which kext?? :P On C2D or other Apple-supported CPUs this kernel is 99.99% vanilla.

 

Also, if no other kernel worked on your p-M then it's obviously not the kernel's fault. PentiumM is the best tested cpu because the kernel is mostly being developed on a pentium m.

 

And guys for the last time - PLEASE ALSO REPORT WHAT KERNEL PANIC YOU GET INSTEAD OF JUST SAYING KERNEL PANICC!!!11! :P

 

thanks.

Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...