On May 28, 2018, at 8:08 AM, Igor Kozhukhov <[email protected]> wrote: > > i have panic with these changes:
As I said, I don’t have an Illumos system (nor time to set one up for the next couple of weeks). However, I have no idea where those warning messages are coming from. I presume via memory management code that we in FreeBSD don’t have. That said: > panic[cpu8]/thread=fffffe0040c0fc40: > Deadlock: cycle in blocking chain > > > fffffe0040c0f0b0 genunix:process_type+1f275f () > fffffe0040c0f120 unix:mutex_vector_enter+3bf () > fffffe0040c0f170 zfs:vdev_queue_change_io_priority+62 () > fffffe0040c0f1d0 zfs:zio_change_priority+a5 () > fffffe0040c0f230 zfs:zio_change_priority+7a () > fffffe0040c0f290 zfs:zio_change_priority+7a () > fffffe0040c0f2f0 zfs:zio_change_priority+7a () that’s a very strange stack, since it looks recursive. Yay optimizer. I can simply revert the last patch, which added the priority changes, and that may be for the best if nobody knows why the panic happened. Thanks, Sean. ------------------------------------------ openzfs: openzfs-developer Permalink: https://openzfs.topicbox.com/groups/developer/discussions/T6b1e5de580958224-M057907f16cdde17ddddb8112 Delivery options: https://openzfs.topicbox.com/groups
