Ryzen 7000 Testing

[CaseySJ]

... continued ...

After rebooting and checking the next OpenCore log, we get slightly different numbers:

00:079 00:013 OCCPU: MP services threads 16 (enabled 16) - Success
00:089 00:010 OCCPU: MP services Pkg 1 Cores 8 Threads 2 - Success
00:112 00:022 OCCPU: Found AMD Ryzen 7 7700X 8-Core Processor
00:135 00:023 OCCPU: Signature A60F12 Stepping 2 Model 61 Family F Type 0 ExtModel 6 ExtFamily A uCode 0 CPUID MAX (10/80000028)
00:250 00:115 OCCPU: FID 180 DID 8 Divisor 0 MaxBR 44
00:261 00:010 OCCPU: CPUFrequencyFromTSC 4500000914Hz 4500MHz
00:271 00:010 OCCPU: CPUFrequency 4500000914Hz 4500MHz
00:281 00:010 OCCPU: FSBFrequency 102272748Hz 102MHz

In the previous run these values were:

• CPUFrequencyFromTSC 4499999649Hz 4499MHz
• FSBFrequency 102272719Hz 102MHz

 

[fabiosun]

Loved boot loader

Last login: Thu Nov 17 08:18:02 on console
fabio@Mac-Pro ~ % sysctl -a | grep -e tsc -e freq
kern.hv.clock.tsc_base: 218667888965
kern.hv.clock.tsc_clock_last: 0
net.link.fake.switch_mode_frequency: 10
net.pktsched.netem.sched_output_ival_ms: 1
net.pktsched.netem.heap_size: 2048
net.pktsched.verbose: 0
hw.busfrequency: 408000000
hw.busfrequency_min: 408000000
hw.busfrequency_max: 408000000
hw.cpufrequency: 3700000000
hw.cpufrequency_min: 3700000000
hw.cpufrequency_max: 3700000000
hw.tbfrequency: 1000000000
machdep.tsc.nanotime.tsc_base: 187053828079
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 1160802062
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.tsc.frequency: 3699999712
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.at_boot: 0
machdep.tsc.rebase_abs_time: 50555092584
machdep.cpu.tsc_ccc.numerator: 0
machdep.cpu.tsc_ccc.denominator: 0

NOT loved Bootloader

Last login: Thu Nov 17 08:27:10 on console
fabio@Mac-Pro ~ % sysctl -a | grep -e tsc -e freq
kern.hv.clock.tsc_base: 236997679863
kern.hv.clock.tsc_clock_last: 0
net.link.fake.switch_mode_frequency: 10
net.pktsched.netem.sched_output_ival_ms: 1
net.pktsched.netem.heap_size: 2048
net.pktsched.verbose: 0
hw.busfrequency: 100000000
hw.busfrequency_min: 100000000
hw.busfrequency_max: 100000000
hw.cpufrequency: 3701000000
hw.cpufrequency_min: 3701000000
hw.cpufrequency_max: 3701000000
hw.tbfrequency: 1000000000
machdep.tsc.nanotime.tsc_base: 208073831146
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 1160488326
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.tsc.frequency: 3701000000
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.at_boot: 0
machdep.tsc.rebase_abs_time: 56220975692
machdep.cpu.tsc_ccc.numerator: 0
machdep.cpu.tsc_ccc.denominator: 0
fabio@Mac-Pro ~ %

 

[CaseySJ]

Loved boot loader

NOT loved Bootloader

All options are on the table for diagnosing the audio problem. Perhaps the focus today will be on the ‘NOT loved Bootloader’ — that which shall not be named…

 

[fabiosun]

Lord Voldemover!

 

[etorix]

And how's audio with the "NOT loved bootloader" compared with the "loved" one?

 

[CaseySJ]

And how's audio with the "NOT loved bootloader" compared with the "loved" one?

@mariettosun has TRX40, so the task falls on me to check on AM5. He has graciously provided a fully configured unname-able folder that I’ll try soon. Now 5:00am and so cozy by the fireplace in Hogwarts.

 

[fabiosun]

I would like to say I do not think that other bootloader solves this big audio problem
But, it could help to see another debug and to compare tsc data
In the past, famigerate bootloader had a big problem with data hub and tsc on amd cpu
And this caused weird behavior in audio and fast animation on you tube
If CaseySJ will be able to boot his system with this bootloader we could try to reproduce that weird condition
Data hub as I said time ago was the key to solve that problem

 

[CaseySJ]

… Besides this, is there anything else that you found and that's not related to the TSC problem?

I realize I didn’t answer this before, probably because the system is otherwise fully functional. In fact it operates remarkably well and feels more responsive than my Alder Lake mini-ITX built around the Asus ROG Strix Z690-i.

On an operating system that does not differentiate between E-cores and P-cores, I prefer to use a processor that does not have any E-cores. That is what drew me to the AM5 platform over the new Raptor Lake.

Intel has been tilting the P-to-E ratio increasing in favor of E-cores. This is especially alarming, to me at least, with rumors of the new Meteor Lake in which P cores will be reduced in number while E cores continue to multiply.

 

[craighazan]

On my 5700X, I always thought the processor or some hidden process decided. there are always 2 really active cores then 6 not so active. It's just not labelled like on M1/M2.

 

[drmeatball]

Casey you are a god among mortals, TYTYTY! Your guides got my Gigabyte Aero G up and running. Like you I've moved on (bonfilio124) and currently running a 7950X with 128GB, 6900XT. Going to start the hacking later this afternoon. If there's anything I can contribute I'm all in!

 

[ExtremeXT]

Casey you are a god among mortals, TYTYTY! Your guides got my Gigabyte Aero G up and running. Like you I've moved on (bonfilio124) and currently running a 7950X with 128GB, 6900XT. Going to start the hacking later this afternoon. If there's anything I can contribute I'm all in!

You can contribute by testing it long-term. If you have any problems make sure to ask here!

 

[backinblackx86]

I went with AM5 also for the traditional P core setup, like a V8 engine!

 

[CaseySJ]

Casey you are a god among mortals, TYTYTY! Your guides got my Gigabyte Aero G up and running. Like you I've moved on (bonfilio124) and currently running a 7950X with 128GB, 6900XT. Going to start the hacking later this afternoon. If there's anything I can contribute I'm all in!

Good to see another convert to AM5 platform!

Thank you for the kind words, but I hope this thread demonstrates the power of teamwork. I cannot do this alone because I don't have all of the requisite background. Each of us in this thread -- especially @ExtremeXT -- is bringing our backgrounds and skills to solve a common problem.

We demonstrated great teamwork when we built the first Alder Lake Hackintosh on TonyMac.

We demonstrated great teamwork when we flashed the Titan Ridge firmware (with Elias64Fr) and enabled full Thunderbolt Bus on Hackintosh for the first time.

We can do great things when we work together.

I hope that is the main lesson we all learn from this thread that is now 57 pages and growing!

 

[ExtremeXT]

@CaseySJ Can you try this version of VoodooTSCSync? Found it on an archived 2020 page.

 

[fabiosun]

@CaseySJ Can you try this version of VoodooTSCSync? Found it on an archived 2020 page.

<integer>#CORE</integer>

this is wrong (it seems wrong)
you have to use core number less 1 (if it works as in the past)
I suggest also to try 1
on xeon cpu (the similar kext give some good result)

 

[ExtremeXT]

<integer>#CORE</integer>

this is wrong (it seems wrong)
you have to use core number less 1 (if it works as in the past)
I suggest also to try 1
on xeon cpu (the similar kext give some good result)

The 2020 page says to do core - 1. AFAIK Monterey needs it set to 1 though.

 

[CaseySJ]

Although Lord-V is not yet booting, its log file shows something interesting:

0:156 0:000 === [ GetCPUProperties ] ========================
0:156 0:000 CPU Vendor = 68747541 Model=A60F12
0:156 0:000 The CPU supported SSE4.1
0:156 0:000 The CPU supported RDRAND
0:156 0:000 BrandString = AMD Ryzen 7 7700X 8-Core Processor
0:156 0:000 Looking to disable C1E if is already enabled by the BIOS:
0:156 0:000 CurrentSpeed: 4500
0:156 0:000 maxratio 88
0:156 0:000 cpudid_zen 72
0:156 0:000 FSBFrequency = 52 MHz, DMI FSBFrequency = 100 MHz, Corrected FSBFrequency = 100 MHz
0:156 0:000 MaxDiv/MinDiv: 44.0/30
0:156 0:000 Turbo: 0/0/0/0
0:156 0:000 Features: 0x76F8320B178BFBFF
0:156 0:000 Threads: 16
0:156 0:000 Cores: 8
0:156 0:000 FSB: 100 MHz
0:156 0:000 CPU: 4499 MHz
0:156 0:000 TSC: 4500 MHz
0:156 0:000 PIS: 100 MHz
0:156 0:000 ExternalClock: 25 MHz

These numbers are what we calculated manually:

• FSB 100 MHz
• TSC 4500 MHz
• MaxDiv 44

 

[fabiosun]

The 2020 page says to do core - 1. AFAIK Monterey needs it set to 1 though.

The 2020 page says to do core - 1. AFAIK Monterey needs it set to 1 though.

yes cores less 1 as I said

I would try also 1

but not confident it solves audio problem because in debug I have seen or in some kp I have seen there is no track of the original problem to use that kext

 

[CaseySJ]

** Not-Loved Bootloader Finally Booted **​

Finally managed to boot Ventura with the not-loved bootloader, and these are the tsc and freq values from macOS. However, audio static remains the same. Now I'll test VoodooTscSync.



Raw Data from Not-Loved Bootloader:

% sysctl -a | grep -e tsc -e freq

kern.hv.clock.tsc_base: 300124755090
kern.hv.clock.tsc_clock_last: 0
net.link.fake.switch_mode_frequency: 10
net.pktsched.netem.heap_size: 2048
net.pktsched.netem.sched_output_ival_ms: 1
net.pktsched.verbose: 0
hw.busfrequency: 100000000
hw.busfrequency_min: 100000000
hw.busfrequency_max: 100000000
hw.cpufrequency: 4500000000
hw.cpufrequency_min: 4500000000
hw.cpufrequency_max: 4500000000
hw.tbfrequency: 1000000000
machdep.tsc.nanotime.tsc_base: 272167334190
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 954437176
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.tsc.frequency: 4500000000
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.at_boot: 0
machdep.tsc.rebase_abs_time: 60481629683
machdep.cpu.tsc_ccc.numerator: 0
machdep.cpu.tsc_ccc.denominator: 0

And these are the values from OpenCore:

It seems that changing FSBFrequency and initialTSC in OpenCore does not change the values we see here. These are after changing FSB to 200,000,000 and TSC to 6,000,000,000.

% sysctl -a | grep -e tsc -e freq

kern.hv.clock.tsc_base: 234469918485
kern.hv.clock.tsc_clock_last: 0
net.link.fake.switch_mode_frequency: 10
net.pktsched.netem.heap_size: 2048
net.pktsched.netem.sched_output_ival_ms: 1
net.pktsched.verbose: 0
hw.busfrequency: 408000000
hw.busfrequency_min: 408000000
hw.busfrequency_max: 408000000
hw.cpufrequency: 4500000000
hw.cpufrequency_min: 4500000000
hw.cpufrequency_max: 4500000000
hw.tbfrequency: 1000000000
machdep.tsc.nanotime.tsc_base: 212930989155
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 954436207
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.tsc.frequency: 4500004571
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.at_boot: 0
machdep.tsc.rebase_abs_time: 47317949425
machdep.cpu.tsc_ccc.numerator: 0
machdep.cpu.tsc_ccc.denominator: 0

 

[CaseySJ]

The 2020 page says to do core - 1. AFAIK Monterey needs it set to 1 though.

Alas, VoodooTscSyncAMD with IOCPUNum set to either 1 or 15 still generates audio static. A setting of 15 leads to a system crash after about 2-3 seconds of music playback. A setting of 1 remains stable, but audio stutter is the same as before.

Update: Both 1 and 15 lead to system crash.