Subject: Re: uvm_trap during sysinst with nbsd1.5 on i386
To: Erich Meier <Erich.Meier@informatik.uni-erlangen.de>
From: Manuel Bouyer <bouyer@antioche.lip6.fr>
List: netbsd-bugs
Date: 01/30/2001 09:46:19 
On Mon, Jan 29, 2001 at 03:22:31PM +0100, Erich Meier wrote:
> Here we go:
> 
> uvm_fault(0xc9a542f0, 0x4d39e000, 0, 1) -> 1
> Kernel: page fault trap, code=0
> sysinst at: 0x....	movl 0x4(%eax), %edx
> 
> ( c9a55110 c9a73d00 c0792de0 bfefe000 bf800060 bf9fe000 0 ) @ 0xc025140b

Looks like it's in pmap_remove_ptes()

> (    "     bf800000 bf9fe000 0        c9a73d88 )            @ 0xc0251a6e
pmap_do_remove()

> (    "     bdbfe000 bf9fe000 )				    @ 0xc025175b
pmap_remove()

> ( c9a542f0 0        bfbfe000 c9a73da8 c9a542f0 )	    @ 0xc023e961
uvm_unmap_remove()

> (    "        "        "     bfbfdd84 c9a73e18 )	    @ 0xc023dd0e
uvm_unmap()

> ( c9a4fb0c c9a4fb0c c9a73f80 c036549c c9a73df8 )	    @ 0xc0240459
uvmspace_exec()

> (    "     c9a73f88    "     83200000 bfbfd7bc )	    @ 0xc017e408
sys_execve()

> ( bfbf001f 0832001f bfbfd7e8 bfbfd7bc bfbfd728 )	    @ 0xc02536f9
syscall()

Ok, looks like a problem in uvm. How much RAM do you have ?

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--