Hi Sebastian, > Von: Sebastian Andrzej Siewior [mailto:bigeasy@xxxxxxxxxxxxx] > Gesendet: Dienstag, 17. September 2019 12:53 > > Due to the softirq rework we can't have multiple softirqs vectors > running in parallel _or_ have multiple local_bh_disable() section which > are executed in parallel on the same CPU. This made the softirq code not > only simpler but also allowed us to remove a little bit duct tape from > places which were expecting this kind of behaviour. Ok that explains why I haven’t seen this with older RT kernels, although the force-threading was always there... > In the long term we need to figure out if we want to go down that road > again and if so we need to figure out how to avoid that duct tape… Maybe not. But this means that a lot of drivers which have no requirement to be protected against softirq will have to be converted to make them fit for RT. But maybe I'm wrong and most RT applications are triggered by the timer and not by some io-card... Thanks again for all the work! Martin
