On 03/15/2015 12:12 AM, Len Ovens wrote:
>
> There was a comment a while ago about not listing rtc in rtirq a while
> ago, that there are other better timers to use on modern systems anyway.
>
Hello Len,
There is barely any software left that uses rtc, most MIDI software uses
the ALSA timers (snd-timer or snd-hrtimer). Just installed Muse and the
version that got installed uses rtc, maybe Rosegarden uses rtc too but
afaik those are the only ones still relying on rtc for their timing.
Most other software that uses ALSA for MIDI get their timing from the
ALSA timers. You
> Does audio SW generally use these other timers by default then? Is it
> worth while removing rtc from my rtirq config?
Yes and yes, unless you're using software that relies on rtc. You can
check what timer software is using:
$ for i in $(find /proc/asound/seq/*);do echo $i && cat $i && echo -e
"---";done
/proc/asound/seq/clients
Client info
cur clients : 3
peak clients : 6
max clients : 192
Client 0 : "System" [Kernel]
Port 0 : "Timer" (Rwe-)
Port 1 : "Announce" (R-e-)
Connecting To: 128:17[t:0]
Client 14 : "Midi Through" [Kernel]
Port 0 : "Midi Through Port-0" (RWe-)
Client 128 : "seq24" [User]
Port 0 : "[1] seq24 1" (R-e-)
Port 1 : "[2] seq24 2" (R-e-)
Port 2 : "[3] seq24 3" (R-e-)
Port 3 : "[4] seq24 4" (R-e-)
Port 4 : "[5] seq24 5" (R-e-)
Port 5 : "[6] seq24 6" (R-e-)
Port 6 : "[7] seq24 7" (R-e-)
Port 7 : "[8] seq24 8" (R-e-)
Port 8 : "[9] seq24 9" (R-e-)
Port 9 : "[10] seq24 10" (R-e-)
Port 10 : "[11] seq24 11" (R-e-)
Port 11 : "[12] seq24 12" (R-e-)
Port 12 : "[13] seq24 13" (R-e-)
Port 13 : "[14] seq24 14" (R-e-)
Port 14 : "[15] seq24 15" (R-e-)
Port 15 : "[16] seq24 16" (R-e-)
Port 16 : "seq24 in" (-We-)
Port 17 : "seq24 in" (-w--)
Connected From: 0:1[t:0]
Input pool :
Pool size : 200
Cells in use : 0
Peak cells in use : 1
Alloc success : 1
Alloc failures : 0
---
/proc/asound/seq/drivers
snd-seq-midi,loaded,0
---
/proc/asound/seq/queues
queue 0: [Queue-0]
owned by client : 128
lock status : Locked
queued time events : 0
queued tick events : 0
timer state : Stopped
timer PPQ : 192
current tempo : 500000
current BPM : 120
current time : 0.000000000 s
current tick : 0
---
/proc/asound/seq/timer
Timer for queue 0 : HR timer
Period time : 0.000000000
Skew : 65536 / 65536
---
Also /proc/asound/timers might contain some useful info.
>
> The default rtirq list is:
> rtc snd usb i8042
>
> I have found that in the case of two sound cards it is best to be more
> specific:
>
> rtc snd_ice1712 snd i8042
>
> (I have removed usb because I don't use it for anything)
>
> The other question is why is i8042 included? Is it used for more than
> kb/mouse? It seems odd that the kb/mouse has higher priority than disk
> writes for example.
Maybe to ensure good performance when using the mouse or keyboard to
play MIDI notes? Not sure otherwise. I always omit it from
/etc/default/rtirq. Maybe it was a personal preference of Rui when he
initially concocted rtirq.
>
> For use of things like netjack (or other AoIP sw), should eth0 (or
> whatever eth is used for audio) be added to this list? (if one eth is
> used for audio and the other for Inet, then separating them into hi and
> lo priority makes sense)
>
I don't dare to answer this as my knowledge of using netjack is close to
zero ;) Seems like a reasonable idea though.
> Asking questions because I don't know, maybe disk writes are buffered
> well enough it doesn't matter.
Which disk writes? From the audio software? Aren't those performed in
the audio thread and if so don't they inherit the higher rtprio assigned
by JACK?
Jeremy
>
> I have not had problems with xruns with the setup I have, but would like
> to know anyway.
>
>
>
> --
> Len Ovens
> www.ovenwerks.net
Attachment:
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
