We are running a survey to help us improve the experience for all of our members. If you see the survey appear, please take the time to tell us about your experience if you can.
We have a processor cluster that is doing both real time event processing and background FP vector processing with NEON. Do the real time interrupts require storing all the NEON data on the stack - even if they don't involve NEON processing? Are the SIMD instructions interruptable or do they require completion before accepting the interrupt, thereby increasing average interrupt latency?