Correct return values. Don't send pager errors back to vm_fault().

mea culpa. Can you commit to 12.x as well?

Fix a merge issue.

Decrease indent. Remove dead comment. Add missing KERN_RESTART define.

The mode == 0 issue is orthogonal. This guarantee was already broken for any filesystem that supported shared locks.

I agree with mjg. Inactive has been a half promise since shared locking went in. This is really my mess originally. Better to consistently not make the promise and clean up the interface.

This is a nice simplification

For the benefit of other readers; Ryan and I have discussed attempting a contig allocation with DMAP KVA first and then falling back to regular kva allocated sparse allocation. We could do this in the UMA multi-page allocator or natively in kmem_*. This would have the advantage that pmap_kextract() is very cheap and we use it to find the slab and also the domain for NUMA. It would have a disadvantage that it may increase memory fragmentation. If we take this approach we may want to only use power of two allocation sizes 'ppera' so that we don't negatively impact phys fragmentation.

else if ((keg->uk_flags & UMA_ZONE_CONTIG) != 0 && keg->uk_ppera > 1)

		keg->uk_allocf = contig_alloc;

Should we make something specific to this pattern? vcancel()? vabandon()? vabort()

I agree that this whole thing needs to be holistically reworked. I do not object to this patch but I would like kib to review.

I do not object to this patch but I would be more comfortable with kib reviewing.

I am happy to tweak parameters and possibly function names in a follow up patch.

I believe mjg is correct and this provides the same guarantee that exited before.

Review feedback. Mostly style and comments. Re-shuffle vm_fault_soft_fast() to properly handle the error case.

Add an assert and fix a number of cases missing pip. I do not believe they would have caused bugs before but they may after my collapse change.

Any volunteers to test on i386?