panic: vfs_getopt: caller passed 'opts' as NULL

[Ulrich Spoerlein]

Kris Kennaway wrote:
> > Nullfs seems more fragile than I initially thought ...
> 
> It's just that compiling in the extra debugging (it might be
> DEBUG_LOCKS or DEBUG_VFS_LOCKS, I forget which), causes the sizes of
> structures to change, so when the module tries to fondle the structure
> at a certain offset thinking it's accessing a certain field, it's
> really fondling something else entirely and the kernel gets a nasty
> surprise and panics.

It is DEBUG_LOCKS. The DEBUG_VFS_LOCKS macro only enables additional
code at runtime, it does not alter the ABI. Ironically, it is even
documented in conf/NOTES.

For the future, I have to remember that nullfs is a module.

Ulrich Spoerlein
-- 
A: Yes.
>Q: Are you sure?
> >A: Because it reverses the logical flow of conversation.
> >>Q: Why is top posting frowned upon?