interrupt stack redzone -> page fault in ddb

[David Young <dyoung%pobox.com@localhost>]

I've applied this patch to restore the redzones above and below the
interrupt stack,
<ftp://cuw.ojctech.com/users/netbsd-1f0030b1/redzone.restore>.

Now, when I hit the magic keys to enter ddb, I see an endless page fault
loop.  Maybe there is something wrong with my patch that I am not seeing?
Does ddb take a lot more stack space than it used to?

# Stopped in pid 867.1 (sh) at    netbsd:breakpoint+0x4:  popl    %ebp 
breakpoint(0,5,c06d80b4,c06d810c,c07681f6) at netbsd:breakpoint+0x4
comintr(c06d8000,c6c90b60,aa55aa55,uvm_fault(0xc03ce4e0, 0xc622c000, 1) -> 0xe
kernel: supervisor trap page fault, code=0 
Faulted in DDB; continuing...
breakpoint(0,5,c06d80b4,c06d810c,c07681f6) at netbsd:breakpoint+0x4
comintr(c06d8000,c6c90b60,aa55aa55,uvm_fault(0xc03ce4e0, 0xc622c000, 1) -> 0xe
kernel: supervisor trap page fault, code=0
Faulted in DDB; continuing...
breakpoint(0,5,c06d80b4,c06d810c,c07681f6) at netbsd:breakpoint+0x4
comintr(c06d8000,c6c90b60,aa55aa55,uvm_fault(0xc03ce4e0, 0xc622c000, 1) -> 0xe
kernel: supervisor trap page fault, code=0
Faulted in DDB; continuing...

Dave

-- 
David Young             OJC Technologies
dyoung%ojctech.com@localhost      Urbana, IL * (217) 278-3933 ext 24