Subject: Re: yamt-km branch
To: YAMAMOTO Takashi <yamt@mwd.biglobe.ne.jp>
From: Chuck Silvers <chuq@chuq.com>
List: tech-kern
Date: 02/17/2005 09:21:38
On Thu, Feb 17, 2005 at 07:34:58AM +0900, YAMAMOTO Takashi wrote:
> hi,
> 
> > I tried sparc, alpha and mac68k this morning.  the first two are working,
> > but mac68k crashes while starting init.  I tried a vanilla -current from
> > a couple days ago and that works.
> 
> does m68k pmap rely on pm_ptab demand-paged?
> i assumed page table pages are always entered by pmap_enter_ptpage.

the problem is that pmap_remove_mapping() is removing the mapping in st_map
and freeing the page itself before calling uvm_km_free().  removing the calls
to pmap_remove() and uvm_pagefree() fixes it.  I noticed that the amiga copy
of the pmap code is inconsistent with UVM_KMF_WIRED vs. UVM_KMF_VAONLY
for the st_map mappings, I'll fix that too.

the problem that uvm_km_check_empty() noticed is from the call in
pmap_init_md().  here we try to claim the virtual space which is already
backed by PTEs set up back in pmap_bootstrap().  hopefully we can just
use the normal pmap interfaces to create those mappings later, but I haven't
tried it yet.


> > > please feel free to commit the fix to the branch.
> > 
> > I'll do that this evening.

tonight I'll really do it.  :-)

-Chuck