Request for verification of qemu-ppc problem on Asahi linux

Our Neighbors Asahi Linux Ask Asahi
1

Hi
I have a request related to QEMU. Unfortunately I don’t have the space to install Asahi Linux on my Mac M1. I have Linux fedora on macOS using qemu-system-aarch64 but I am not able to check that there.
There is a problem on macOS with different session speeds. Would anyone be so kind as to have qemu-system-ppc installed to check how it looks on Asahi linux ?
Thank you very much for your help
Here is the link to the qemu submission
https://lists.nongnu.org/archive/html/qemu-discuss/2023-12/msg00008.html
There is also a description on how to do the test.

Thank you for your help

2

Hi, I am running Fedora Asahi Remix on an M2 MacBook Air. I ran your script in the zip file and these are the results:

*** Running test 1
Loading, please wait...
*** Running lame ***
LAME 3.100 32bits (http://lame.sf.net)
Using polyphase lowpass filter, transition band: 18671 Hz - 19205 Hz
Encoding /file_example_WAV_5MG.wav to <stdout>
Encoding as 44.1 kHz j-stereo MPEG-1 Layer III VBR(q=2)
    Frame          |  CPU time/estim | REAL time/estim | play/CPU |    ETA 
  1149/1149  (100%)|    0:26/    0:26|    0:26/    0:26|   1.1452x|    0:00 
QEMU 8.1.1 monitor - type 'help' for more information
Accelerator settings:
one-insn-per-tb: off

Translation buffer state:
gen code size       28973796/976190464
TB count            52629
TB avg target size  24 max=2048 bytes
TB avg host size    311 bytes (expansion ratio: 12.8)
cross page TB count 0 (0%)
direct jump count   31853 (60%) (2 jumps=25829 49%)
TB hash buckets     24424/32768 (74.54% head buckets used)
TB hash occupancy   33.34% avg chain occ. Histogram: [0,10)%|▆ █  ▅▁▃▁▁|[90,100]%
TB hash avg chain   1.017 buckets. Histogram: 1|█▁|2

Statistics:
TB flush count      0
TB invalidate count 7823
TLB full flushes    0
TLB partial flushes 13234
TLB elided flushes  99662
[TCG profiler not compiled]
*** Done
*** Running test 2
Loading, please wait...
*** Running lame ***
LAME 3.100 32bits (http://lame.sf.net)
Using polyphase lowpass filter, transition band: 18671 Hz - 19205 Hz
Encoding /file_example_WAV_5MG.wav to <stdout>
Encoding as 44.1 kHz j-stereo MPEG-1 Layer III VBR(q=2)
    Frame          |  CPU time/estim | REAL time/estim | play/CPU |    ETA 
  1149/1149  (100%)|    0:26/    0:26|    0:26/    0:26|   1.1309x|    0:00 
QEMU 8.1.1 monitor - type 'help' for more information
Accelerator settings:
one-insn-per-tb: off

Translation buffer state:
gen code size       28977156/976190464
TB count            52641
TB avg target size  24 max=2048 bytes
TB avg host size    311 bytes (expansion ratio: 12.8)
cross page TB count 0 (0%)
direct jump count   31856 (60%) (2 jumps=25834 49%)
TB hash buckets     24476/32768 (74.69% head buckets used)
TB hash occupancy   33.35% avg chain occ. Histogram: [0,10)%|▆ █  ▅▁▃▁▁|[90,100]%
TB hash avg chain   1.017 buckets. Histogram: 1|█▁|2

Statistics:
TB flush count      0
TB invalidate count 7823
TLB full flushes    0
TLB partial flushes 13293
TLB elided flushes  100003
[TCG profiler not compiled]
*** Done
*** Running test 3
Loading, please wait...
*** Running lame ***
LAME 3.100 32bits (http://lame.sf.net)
Using polyphase lowpass filter, transition band: 18671 Hz - 19205 Hz
Encoding /file_example_WAV_5MG.wav to <stdout>
Encoding as 44.1 kHz j-stereo MPEG-1 Layer III VBR(q=2)
    Frame          |  CPU time/estim | REAL time/estim | play/CPU |    ETA 
  1149/1149  (100%)|    0:26/    0:26|    0:26/    0:26|   1.1392x|    0:00 
QEMU 8.1.1 monitor - type 'help' for more information
Accelerator settings:
one-insn-per-tb: off

Translation buffer state:
gen code size       28967012/976190464
TB count            52622
TB avg target size  24 max=2048 bytes
TB avg host size    311 bytes (expansion ratio: 12.8)
cross page TB count 0 (0%)
direct jump count   31846 (60%) (2 jumps=25825 49%)
TB hash buckets     24467/32768 (74.67% head buckets used)
TB hash occupancy   33.33% avg chain occ. Histogram: [0,10)%|▆ █  ▅▁▃▁▁|[90,100]%
TB hash avg chain   1.017 buckets. Histogram: 1|█▁|2

Statistics:
TB flush count      0
TB invalidate count 7818
TLB full flushes    0
TLB partial flushes 13204
TLB elided flushes  99388
[TCG profiler not compiled]
*** Done

It took about 1 minute and 25 seconds to complete according to time.

3

Thank you very much :slight_smile:

I use Fedora on MacOS under qemu-system-aarch64 -machine virt,accel=hvf -cpu host and virgl.
Here with this test it always comes out 0:56.
Here it is jabky all the time with a bad session. In this case I am not able to say whether it is supposed to be like this.
On qemu-system-pcc I am able to check this. Random fast sessions and bad sessions. I use this emulation for AmigaNG/pegasos 2/AmigaOS4 under macOS.
Here we noticed strange things with macOS (possibly with QEMU but on other systems it works fine)

Thank you again very much for your time and checking - I greet you warmly :slight_smile:

4

@ydalton

I have a request for you. Could you do a 10th sample of this test?
However, I need more data to prove that your 3 attempts were not just luck.
Thank you very much.

You can roll this by changing the lines “for i in seq 1 3; do” to “for i in seq 1 10; do”.
You can also run the test 3 times with 3 samples each if you want to change these lines.

@All

I also have a request for others to do the test. Maybe someone has a qemu on Asahi linux and MacOs on the same machine ?

Thank you