It occurred to me that Martijn van Buul wrote in gmane.os.netbsd.ports.x86-64:
> It occurred to me that Martijn van Buul wrote in gmane.os.netbsd.ports.x86-64:
>
>> ddb never kicks in
>
> This is *probably* related to the fact that i had disabled ddb. D-oh. I'll 
> build one with debugging symbols while I'm at it, and try to produce a more
> useful backtrace tonight.

Done. I built a kernel with DIAGNOSTICS and DEBUG. 4.99.1 as of yesterday.

The peculiar bit is that this *only* happens during a bulk-build (which I do
inside a pkg_comp chroot). Manually building it inside the chroot works, 
having mk/bulk/build.sh build it always crashes at exactly same point during 
the build of editors/emacs20.

panic: kernel diagnostic assertion "idx < VMEM_MAXORDER" failed: file "/usr/export/src/sys/kern/subr_vmem.c", line 267
stopped in pid 28858.1 (temacs) at      netbsd:cpu_Debugger+0x5:        leave
db> bt
cpu_Debugger() at netbsd:cpu_Debugger+0x5
panic() at netbsd:panic+0x1fc
__assert() at netbsd:__assert+0x21 
vmem_alloc() at netbsd:vmem_alloc+0x317
kmem_alloc() at netbsd:kmem_alloc+0x3a
amap_alloc() at netbsd:amap_alloc+0xf1
uvm_map_enter() at netbsd:uvm_map_enter+0x1f0
uvm_map() at netbsd:uvm_map+0x11f
sys_obreak() at netbsd:sys_obreak+0xe0
syscall_plain() at netbsd:syscall_plain+0x113
uvm_fault(0xffff80004b31a2e8, 0x100084e000, 1) -> e
kernel: page fault trap, code=0
Faulted in DDB; continuing...
db>

sync still fails.