96Khz seems to hangup the Server

weefuzzy · October 26, 2024, 10:44pm

Issue here, should you wish to follow along

Crash in FluidBufStats

opened 10:44PM - 26 Oct 24 UTC

bug

This may end up moving to core, and hopefully will get a better title. Crash… in SC server reported by user here: https://discourse.flucoma.org/t/96khz-seems-to-hangup-the-server/2438 Can reproduce on my mac. Only have a release build stack trace so far, which isn't much use, but ``` 0 libsystem_kernel.dylib 0x1826a0764 __pthread_kill + 8 1 libsystem_pthread.dylib 0x1826d7c28 pthread_kill + 288 2 libsystem_c.dylib 0x1825e5ae8 abort + 180 3 libsystem_malloc.dylib 0x182506e28 malloc_vreport + 908 4 libsystem_malloc.dylib 0x18251d5d4 malloc_zone_error + 104 5 libsystem_malloc.dylib 0x182515148 nanov2_guard_corruption_detected + 44 6 libsystem_malloc.dylib 0x182514344 nanov2_allocate_outlined + 404 7 FluidBufStats.scx 0x105fac918 Eigen::PlainObjectBase<Eigen::Array<int, -1, 1, 0, -1, 1>>::resize(long, long) + 96 8 FluidBufStats.scx 0x105fa8e98 fluid::algorithm::MultiStats::process(fluid::FluidTensorView<double, 2ul>, fluid::FluidTensorView<double, 2ul>, double, fluid::FluidTensorView<double, 1ul>) + 172 9 FluidBufStats.scx 0x105fa8730 fluid::client::Result fluid::client::bufstats::BufferStatsClient::process<float>(fluid::client::FluidContext&) + 2436 ``` So, an attempt to resize something in `MultistatsProcess` is making badness happen. I think I remember similar issues before where something reasonable-seeming with Eigen causes woe.