It's a real-time circus out there!

[autostatic]

Introducing rtcirqus, a Python script to automatically set real-time priorities of the IRQ threads of your onboard and USB audio devices.

https://codeberg.org/autostatic/rtcirqus

Code: Select all

Feb 16 22:34:10 lenovo kernel: usb 7-1: new high-speed USB device number 10 using xhci_hcd
Feb 16 22:34:11 lenovo kernel: usb 7-1: New USB device found, idVendor=0424, idProduct=3fb7, bcdDevice= 0.01
Feb 16 22:34:11 lenovo kernel: usb 7-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
Feb 16 22:34:11 lenovo kernel: usb 7-1: Product: Babyface (23596862)
Feb 16 22:34:11 lenovo kernel: usb 7-1: Manufacturer: RME
Feb 16 22:34:11 lenovo kernel: usb 7-1: SerialNumber: 18C0A73560538C8
Feb 16 22:34:11 lenovo kernel: usb 7-1: Quirk or no altest; falling back to MIDI 1.0
Feb 16 22:34:11 lenovo mtp-probe[15199]: checking bus 7, device 10: "/sys/devices/pci0000:00/0000:00:08.3/0000:66:00.4/usb7/7-1"
Feb 16 22:34:11 lenovo mtp-probe[15199]: bus: 7, device: 10 was not an MTP device
Feb 16 22:34:11 lenovo systemd[1]: Started run-r4c9e3d2d062846ecab03c6af31529349.service - /usr/local/bin/rtcirqus --action add --dev-path /devices/pci0000:00/0000:00:08.3/0000:66:00.4/usb7/7-1/7-1:1.0/sound/card0.
Feb 16 22:34:11 lenovo mtp-probe[15216]: checking bus 7, device 10: "/sys/devices/pci0000:00/0000:00:08.3/0000:66:00.4/usb7/7-1"
Feb 16 22:34:11 lenovo mtp-probe[15216]: bus: 7, device: 10 was not an MTP device
Feb 16 22:34:11 lenovo rtcirqus[15212]: WARNING: No configuration file found, continuing with defaults
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: Loaded kernel is using threaded IRQs and threaded IRQ processes detected
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: Onboard cards found: HD-Audio Generic, HD-Audio Generic, acp63
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: USB cards found: Babyface (23596862)
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: Setting RT priority 90 for USB card Babyface (23596862)
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: Setting RT priority 85 for onboard card HD-Audio Generic
Feb 16 22:34:11 lenovo rtcirqus[15212]: INFO: Setting RT priority 80 for onboard card HD-Audio Generic
Feb 16 22:34:11 lenovo systemd[1]: run-r4c9e3d2d062846ecab03c6af31529349.service: Deactivated successfully.

Hope it's useful!

Jeremy

[Audiojunkie]

This looks cool! I may be wrong, but I don't think Pipewire does any thread priorities. It makes use of IRQ timing when in Pro Audio mode, but I think prioritizing the threads may be a place where an extra boost can be achieved when using a RT kernel or a generic kernel using the /threadirqs kernel paramater--just like the rtirq script does. If I understand correctly, your script specifically focuses on the USB interface devices. Is this correct?

How is this script different from the rtirq script?

[autostatic]

This looks cool!

Thanks!

I may be wrong, but I don't think Pipewire does any thread priorities. It makes use of IRQ timing when in Pro Audio mode, but I think prioritizing the threads may be a place where an extra boost can be achieved when using a RT kernel or a generic kernel using the /threadirqs kernel paramater--just like the rtirq script does.

From my personal experience prioritizing IRQ threads is indispensable. It's the difference between running reliably at low latencies or having to cope with dozens of xruns.

If I understand correctly, your script specifically focuses on the USB interface devices. Is this correct?

Correct but rtcirqus handles both onboard and USB devices. The focus on USB is because USB devices can be plugged in and out. With onboard devices this is a bit harder

How is this script different from the rtirq script?

rtirq is not coupled to udev so it doesn't respond dynamically to newly added or removed audio interfaces. You can do so too though with rtirq, see https://discourse.ardour.org/t/jerky-pl ... 4/109844/7 That post actually got me started with creating rtcirqus. Another difference is that on modern machines rtirq does not allow to prioritize only the IRQ where the audio interface sits on. For USB, that mainly uses xhci_hcd related IRQ's these days, rtirq can only prioritize all xhci_hcd related IRQs, not just the one with your interface connected. rtirq is more generic, it allows you to prioritize any IRQ while rtcirqus is geared towards audio related IRQs. rtirq also allows for finer grained control, the control options for rtcirqus are still pretty basic. And then there are some meta differences (configuration file location standardized according to FHS and licensing).

[bluebell]

Introducing rtcirqus, a Python script to automatically set real-time priorities of the IRQ threads of your onboard and USB audio devices.

https://codeberg.org/autostatic/rtcirqus

Hi Jeremy,

looks promising. Unfortunately besides the missing chmod 755 of rtcirqus.py I have a severe problem:

Code: Select all

hm@bubu:~$ rtcirqus 
  File "/usr/local/bin/rtcirqus", line 53
    udev_context, 'sound', f'card{index}').device_path for index in
                                        ^
SyntaxError: invalid syntax

Code: Select all

hm@bubu:~$ python3 --version
Python 3.5.2

It seems that this Python version doesn't support f-strings. I managed - without really knowing what I am doing - to fix this line with

Code: Select all

udev_context, 'sound', fromstr('card{}'.format(index))).device_path for index in

but f-strings are heavily used in rtcirqus.py.

Yes, my well tuned system is a bit oldish but this way I'd be able to compare the effects of rtirq and rtcirqus.

[autostatic]

looks promising. Unfortunately besides the missing chmod 755 of rtcirqus.py I have a severe problem:

Thanks for the heads up, that's an easy fix.

Code: Select all

hm@bubu:~$ rtcirqus 
  File "/usr/local/bin/rtcirqus", line 53
    udev_context, 'sound', f'card{index}').device_path for index in
                                        ^
SyntaxError: invalid syntax

Code: Select all

hm@bubu:~$ python3 --version
Python 3.5.2

It seems that this Python version doesn't support f-strings.

Correct, formatted string literals were introduced in Python 3.6.

Yes, my system is a bit oldish but this way I could compare the effects of rtirq and rtcirqus.

Sorry to hear Can't promise anything but I could look into replacing the f-strings with str.format() ones so you could compare. I do prefer f-strings though so I'd like to keep those in the current script.

[bluebell]

There might be a solution with the package future_fstrings that needs a "coding:" directive but I wasn't successful. Not even future-fstrings-show worked which - as I understand - can convert Python scripts with f-strings.

There is a Python3.6 on my system, too, but without properly installed packages as psutil & co.

Let's see if I can tame that Python mess. I can't remember such problems with Perl but there must be a reason why people like Python. I know at least 2 reasons why I don't .

[sysrqer]

This is interesting to hear, I wouldn't think to consider that people have <3.6 of python. Could this error be caught be try/except?

Python is godly btw

[bluebell]

This is interesting to hear, I wouldn't think to consider that people have <3.6 of python. Could this error be caught be try/except?

Python is godly btw

Python is a sweet poison. It looks elegant (it isn't), readable (it isn't), sane (it isn't) and finally bites your ass with backwards or maybe upwards incompatibilities.

[rncbc]

Python is a sweet poison. It looks elegant (it isn't), readable (it isn't), sane (it isn't) and finally bites your ass with backwards or maybe upwards incompatibilities.

let alone that real programmers do their own indentation (you can quote me on that line:))

any programming language that makes whitespace a role in its syntax is plain ridiculous... in my own judgement I rather praise cobol instead (all is/was good after the 8th column XD).

anyway, back to the topic at hand

having rtirq is almost two decades old already, wouldn't it be enough to add just a few rules to it?

please try the attached script and tell which IRQ we're talking about...

ofc. this pledge applies to usb audio devices only, the more the better; try to plug them in and out and back again; also after rebooting; how many lines does it show?

eager to know, whether the (old) rtirq script might cope with this mess--usb audio devices in particular--and not having you all to depend on snake oil bloat&blunder

cheers

[bluebell]

please try the attached script and tell which IRQ we're talking about...

ofc. this pledge applies to usb audio devices only, the more the better; try to plug them in and out and back again; also after rebooting; how many lines does it show?

eager to know, whether the (old) rtirq script might cope with this mess--usb audio devices in particular--and not having you all to depend on snake oil bloat&blunder

Code: Select all

NAME2=<Focusrite Scarlett 2i4 USB>
UDEV2=<0000:00:14.0>
PRI1=<90>
IRQ=<127>
rtirq_start_irq snd-usb Focusrite Scarlett 2i4 USB 90 127

That's not bad but not perfect. My main USB interface MOTU UltraLite AVB which uses drumfix' ua101 driver is not recognized.

Code: Select all

hm@bubu:~$ cat /proc/asound/cards
 0 [MOTUAVB        ]: MOTU-AVB - MOTU-AVB
                      MOTU UltraLite AVB
 1 [USB            ]: USB-Audio - Scarlett 2i4 USB
                      Focusrite Scarlett 2i4 USB at usb-0000:00:14.0-5, high speed

Code: Select all

hm@bubu:/tmp$ lsusb
Bus 004 Device 003: ID 0bda:0411 Realtek Semiconductor Corp. 
Bus 004 Device 002: ID 0bda:0411 Realtek Semiconductor Corp. 
Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 003 Device 004: ID 30fa:0300  

Bus 003 Device 003: ID 0bda:5411 Realtek Semiconductor Corp. 
Bus 003 Device 002: ID 0bda:5411 Realtek Semiconductor Corp. 
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 001 Device 004: ID 04f2:b59e Chicony Electronics Co., Ltd 
Bus 001 Device 006: ID 1235:8200 Focusrite-Novation 
Bus 001 Device 005: ID 8087:0a2b Intel Corp. 
Bus 001 Device 002: ID 07fd:0005 Mark of the Unicorn 
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

Code: Select all

hm@bubu:/tmp$ lsusb -t
/:  Bus 04.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/2p, 10000M
    |__ Port 2: Dev 2, If 0, Class=Hub, Driver=hub/4p, 5000M
        |__ Port 4: Dev 3, If 0, Class=Hub, Driver=hub/4p, 5000M
/:  Bus 03.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/2p, 480M
    |__ Port 2: Dev 2, If 0, Class=Hub, Driver=hub/4p, 480M
        |__ Port 4: Dev 3, If 0, Class=Hub, Driver=hub/4p, 480M
            |__ Port 1: Dev 4, If 0, Class=Human Interface Device, Driver=usbhid, 1.5M
/:  Bus 02.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/8p, 5000M
/:  Bus 01.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/16p, 480M
    |__ Port 1: Dev 2, If 7, Class=Vendor Specific Class, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 5, Class=Vendor Specific Class, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 3, Class=Audio, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 1, Class=Audio, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 6, Class=Vendor Specific Class, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 4, Class=Audio, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 2, Class=Audio, Driver=snd-motu-avb, 480M
    |__ Port 1: Dev 2, If 0, Class=Audio, Driver=snd-motu-avb, 480M
    |__ Port 5: Dev 6, If 0, Class=Audio, Driver=snd-usb-audio, 480M
    |__ Port 5: Dev 6, If 1, Class=Audio, Driver=snd-usb-audio, 480M
    |__ Port 5: Dev 6, If 2, Class=Audio, Driver=snd-usb-audio, 480M
    |__ Port 5: Dev 6, If 3, Class=Audio, Driver=snd-usb-audio, 480M
    |__ Port 5: Dev 6, If 4, Class=Audio, Driver=snd-usb-audio, 480M
    |__ Port 5: Dev 6, If 5, Class=Vendor Specific Class, Driver=, 480M
    |__ Port 6: Dev 4, If 0, Class=Video, Driver=, 480M
    |__ Port 6: Dev 4, If 1, Class=Video, Driver=, 480M
    |__ Port 10: Dev 5, If 0, Class=Wireless, Driver=, 12M
    |__ Port 10: Dev 5, If 1, Class=Wireless, Driver=, 12M

Code: Select all

hm@bubu:/tmp$ ls -l /sys/devices/pci0000:00/*/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:00.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:33 /sys/devices/pci0000:00/0000:00:02.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:14.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:14.2/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:16.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:17.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1c.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1c.4/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1c.6/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1d.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1d.4/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1f.0/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1f.2/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1f.3/irq
-r--r--r-- 1 root root 4096 Feb 19 08:34 /sys/devices/pci0000:00/0000:00:1f.4/irq

Code: Select all

hm@bubu:/tmp$ cat /sys/devices/pci0000:00/*/irq
0
138
127
18
139
136
122
123
124
125
126
0
0
11
11

IRQ 127 is probably used for both interfaces when I look at /proc/interrupts

Code: Select all

hm@bubu:/tmp$ cat /proc/interrupts 
            CPU0       CPU1       CPU2       CPU3       CPU4       CPU5       CPU6       CPU7       

   0:         23          0          0          0          0          0          0          0  IR-IO-APIC    2-edge      timer
   1:          0          0          0      32975          0          0          0          0  IR-IO-APIC    1-edge      i8042
   8:          0          0          0          0          0          0          0          0  IR-IO-APIC    8-edge      rtc0
   9:          0        272          0          0          0          0          0          0  IR-IO-APIC    9-fasteoi   acpi
  12:          0          0      36310          0          0          0          0          0  IR-IO-APIC   12-edge      i8042
 120:          0          0          0          0          0          0          0          0  DMAR-MSI    0-edge      dmar0
 121:          0          0          0          0          0          0          0          0  DMAR-MSI    1-edge      dmar1
 127:          0          0          0          0          0          0          0  740673105  IR-PCI-MSI 327680-edge      xhci_hcd
 128:          0     270256          0          0          0          0          0          0  IR-PCI-MSI 1572864-edge      xhci_hcd
 129:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572865-edge      xhci_hcd
 130:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572866-edge      xhci_hcd
 131:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572867-edge      xhci_hcd
 132:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572868-edge      xhci_hcd
 133:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572869-edge      xhci_hcd
 134:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572870-edge      xhci_hcd
 135:          0          0          0          0          0          0          0          0  IR-PCI-MSI 1572871-edge      xhci_hcd
 136:          0          0          0          0          0     342137          0          0  IR-PCI-MSI 376832-edge      ahci[0000:00:17.0]
 137:    1734401          0          0          0          0          0          0          0  IR-PCI-MSI 2621440-edge      enp5s0
 138:          0          0          0          0          0          0    2835567          0  IR-PCI-MSI 32768-edge      i915
 139:          0          0          0          0          0          0          0         39  IR-PCI-MSI 360448-edge      mei_me
 NMI:          0          0          0          0          0          0          0          0   Non-maskable interrupts
 LOC:   46878557   46995424   47022404   46802751   46684791   47192215   47169442   46485698   Local timer interrupts
 SPU:          0          0          0          0          0          0          0          0   Spurious interrupts
 PMI:          0          0          0          0          0          0          0          0   Performance monitoring interrupts
 IWI:       7694       5731       5557       7344       4346      10472    1420597       9094   IRQ work interrupts
 RTR:          0          0          0          0          0          0          0          0   APIC ICR read retries
 RES:     282927     238517     293863     317499     406355     248732     181617      79114   Rescheduling interrupts
 CAL:    1209744     985576     967243    1435280     834423     876803     876404    1341983   Function call interrupts
 TLB:     631977     672675     690830     770550     676369     649612     588390     454202   TLB shootdowns
 TRM:        476        476        476        476        476        476        476        476   Thermal event interrupts
 THR:          0          0          0          0          0          0          0          0   Threshold APIC interrupts
 DFR:          0          0          0          0          0          0          0          0   Deferred Error APIC interrupts
 MCE:          0          0          0          0          0          0          0          0   Machine check exceptions
 MCP:        142        143        143        143        143        143        143        143   Machine check polls
 ERR:          0
 MIS:          0
 PIN:          0          0          0          0          0          0          0          0   Posted-interrupt notification event
 NPI:          0          0          0          0          0          0          0          0   Nested posted-interrupt event
 PIW:          0          0          0          0          0          0          0          0   Posted-interrupt wakeup event

I have no idea how I could get from the lsusb output to IRQ 127 in /sys/devices/pci0000:00/0000:00:14.0/irq

[rncbc]

ofc. this pledge applies to usb audio devices only...

I'll correct the above that it applies to snd-usb-audio driven devices only. iow. the ones which are usually said "usb audio class compliant"...

I could be wrong, but the motu avb doesn't seem to be one; does it appear on rtcirqus ?

[bluebell]

ofc. this pledge applies to usb audio devices only...

I'll correct the above that it applies to snd-usb-audio driven devices only. iow. the ones which are usually said "usb audio class compliant"...

I could be wrong, but the motu avb doesn't seem to be one; does it appear on rtcirqus ?

At the moment I can't run rtcircus because of Python's builtin incompatibilities.

Iam afraid that rtcircus relies on the sys-path info in /proc/alsa/cards as well.

I run the MOTU UltraLite AVB in Class Compliant mode.

[Loki Harfagr]

@rncbc

My card is a funny one and the script goes funny:

Code: Select all

 sh rtirq_snd-usb_test-1.sh.txt
cat: '/sys/devices/pci0000:00/0000:07:00.0/irq': No such file or directory

and actually, asound states:

Code: Select all

 grep usb /proc/asound/cards
                      TASCAM US-4x4HR at usb-0000:07:00.0-1, high speed

while lsusb shows:

Code: Select all

lsusb | grep US-4x4HR
Bus 003 Device 002: ID 0644:8071 TEAC Corp. US-4x4HR

and find gives:

Code: Select all

 find /sys/devices/pci* -iname '*usb*' -exec ls {} + | grep 00:07
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb4/4-0:1.0/usb4-port1/usb3_lpm_permit
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb4/4-0:1.0/usb4-port2/usb3_lpm_permit
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb3:
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb3/3-0:1.0/usb3-port1:
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb3/3-0:1.0/usb3-port2:
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb4:
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb4/4-0:1.0/usb4-port1:
/sys/devices/pci0000:00/0000:00:1c.4/0000:07:00.0/usb4/4-0:1.0/usb4-port2:

[bluebell]

I googled around and I'm afraid that there is no way to find out the irq for an usb audio card if the driver doesn't provide any hints, e,g, in /proc/asound/cards

I hope that I am wrong.

[rncbc]

My card is a funny one and the script goes funny: [...]

yeah, tascam and motu are infamous in this regard, as ever been: they are not driven by snd-usb-audio whatsoever...

question still applies: does the OP rtcirqus give you any better figures?

EDIT: btw. rtirq update - 2024 second edition is out now