I've updated my image to hold back the bad kernel and i've ensure my apps are up to date.
I noticed that since then, the direwolf process is consuming an entire core 100%.
It never used to do that and I've not seen that with other non-udrc uses with direwolf on my other Pi's.
Curious if anyone has noticed this before or if there is a known issue/config I haven't found yet.
I checked my direwolf configuration and nothing out of the ordinary. one channel currently on ax25 vhf. (basic config here):
ADEVICE plughw:CARD=udrc,DEV=0 plughw:CARD=udrc,DEV=0
PTT GPIO 12
(hoping to go to split channel if i can figure out this cpu issue).
I noticed a rather old posting by WB02SZ:
Dire Wolf uses many threads for the various concurrent activities. Each audio input device has its own thread which reads data as quickly as the source can provide it. In this case, the null device provides zero bytes at an unlimited rate and one CPU core is completely consumedSO, to check this theory, I added back channel 2 into my direwolf config and stopped/restarted direwolf -- sure enough, the high cpu load went away and it's back down to approx 20% to 30% or so on one core -- which is normal.
So, with me trying to take direwolf out of using both channels and allocate one to HF with pulseaudio ... is this expected behaviour from direwolf? and are others seeing similar results? or am i missing some other configuration to get direwolf to stop trying to read null streams and becoming cpu bound.
73, Dan ve4drk