-
Notifications
You must be signed in to change notification settings - Fork 349
Closed
Labels
bugSomething isn't working as expectedSomething isn't working as expected
Milestone
Description
Follow up from here
Now that we have well defined behaviour for the reset handler we need to followup on cleaning up state in reset. Initial scan shows some components still in violation
- Codec adapter has number of variables set in prepare and params step that are not zeroed out @ranj063
- Crossover doesn't reset stream data or processing function pointer - @afq984
- Dcblock similar to above for crossover - @afq984
- DRC
cd->source_formatis not cleaned up - @johnylin76 - EQ FIR
source_formatisn't cleaned up but doesn't appear to be used? - @singalsu - EQ IIR same var again, could be a local var? Sink format as well - @singalsu
- IGO fails to clean up all
cddata stored inparamsstep - @fuyuntsuo - Multiband DRC same as DRC - @johnylin76
- Mux processing funcs not cleared - @akloniex
- Selector no state cleaned up - @eddy1021
- Smart Amp no state cleaned up - @ryans-lee
- SRC same as EQ IIR - @singalsu
- TDFB in/out buffer clear only appears to happen when new config appears on prepare, not on pipeline reset? - @singalsu
- Volume does not clear out old state set in prepare for processing - @singalsu
- Host appears to store a number of state variables in hd->config that are never cleared and many not used - @jsarha
- KPB number of variables set on prepare that are not cleared in reset - @andrula-song
- mixer fails to clear mixer function state also has a magic number to stop pipeline walk? - @kv2019i
Sorry for the large blast radius for messages just hoping to get some coherency for all components so we can stop propagating errors to new members joining the project through copy paste issues :)