How to reproduce: - Start PV, load dcf77_1800s.sr (and use libsigrokdecode commit 3f3c4614b, i.e. before the DCF77 PD bug was fixed which causes an error for dcf77_1800s.sr). - Add the DCF77 PD, let it run until it hits the error. At this point PV should abort the decode and the feeding of further data into srd. This appears to be the case (no further "srd: Decoding: abs start sample 1131151360, abs end sample 1131413504 (262144 samples, 262144 bytes, unitsize = 1), instance dcf77-1." and similar messages in an -l 5 log. However, something still seems to be going on, since after the PD error the CPU load is still relatively high for 10-20 seconds before it then goes to (near) zero. This is reproducible every time. I'm assuming PV might still be doing *some* operations over the rest of the sample data (after the error) or something like that?