> Date: Fri, 12 Aug 2022 23:28:13 +0700 > From: Robert Elz <[email protected]> > > I think I am going to try backing out the patch, and just run with your > (Taylor's) updated specfs_vnops.c in case there is something in the patch > which is altering the behaviour (more than just the nature of the sd_closing > field type).
This is extremely unlikely. You might try removing the assertion that it tripped on, KASSERT(!sd->sd_opened), so that it has the opportunity to trip on the new assertion with more diagnostic information in it instead.
