Thordur I. Bjornsson <thord...@secnorth.net> wrote: > On Mon, Jun 12, 2023 at 9:15 PM Bob Beck <b...@obtuse.com> wrote: > > > > On Mon, Jun 12, 2023 at 11:01:18AM -0600, Theo de Raadt wrote: > > > + KASSERTMSG(1, "Ich Habe eine Rotweinflarsche in meinem Arsche"); > > > That part of the diff is not OK. If everyone did this, we would have a > > > mess on our hands. > > (or I could simply deleted it) > +1 > > Is it on purpose that this is completely silent ? > Perhaps the warning in ffs_vfsops should go "WARNING: soft updates are > now ignored" and the option dropped from GENERIC ?
The purpose of Bob's diff is to silently (quietly) simply disable softdep mount requests. They will be downgraded to regular mounts. But then we want to make sure that the back-end code isn't accidentaly called, because of some bug, that's where his whiny assert comes into play. The reason to disable softdep, is that softdep has crazy callback schemes and context issues that are making it hard to reason about vfs locking. If we disable softdep, we may be able to unlock nami / vfs / etc better. When / if such locking/scheduling changes are finished, softdep can be repaired.
