Yo!

I'm not sure if I should send this i386 -current panic to this list
or to the ports-i386 list, so please tell me...

I had this panic a few months ago and forgot about it after hacking
around it, but I see it again with a recent -current kernel.  This
is on one particular machine I have two copies of:

BIOS MEMORY MAP (7 ENTRIES):
    addr 0x0  size 0x9fc00  type 0x1
    addr 0x9fc00  size 0x400  type 0x2
    addr 0xf50a2  size 0xaf5e  type 0x2
    addr 0x100000  size 0xf00000  type 0x1
    addr 0x100000  size 0xf00000  type 0x1
WARNING: CAN'T ALLOCATE MEMORY SEGMENT (0x100000/0xf00000/0x1) FROM IOMEM EXTENT MAP!
    addr 0x1000000  size 0x3800000  type 0x1
    addr 0xffff50a2  size 0xaf5e  type 0x2
loading 0x4000-0x9f000 (0x4-0x9f)
loading 0x2cd000-0x1000000 (0x2cd-0x1000)
loading 0x2cd000-0x1000000 (0x2cd-0x1000)
loading 0x1000000-0x1000000 (0x1000-0x1000)
panic: uvm_page_physload: start >= end


The WARNING: line is too long and overwrote itself so that all I was able to see was just the `` MAP!'' part of it.
Not sure if it's worth spitting out such a message in normal boots.

This was not a problem with 1.5.2, though there was another symbol
table problem for which I hope to build my own 1.5.2 kernel later
and see if I can figure out what's going on there.

I tried an obvious fix of changing
		if (seg_start <= (16 * 1024 * 1024) &&
to be just `<', which results in a kernel I can boot on these two
pieces of hardware, but usually panics sometime later.

Naturally the other i386-type machine is fine without any hacks.
The ones plaguing me are old Digital Venturis 575 boxen.


thanks
barry bouwsma