Subject: Re: inconsistency in MD mmmmap() implementations?
To: None <>
From: YAMAMOTO Takashi <>
List: tech-kern
Date: 10/25/2006 22:06:39
> YAMAMOTO Takashi wrote:
> > there is nothing wrong to be inconsistent between ports if they have
> > different memory layouts.
> > eg. "atop(off) >= physmem" can make sense if its memory is
> > contiguously mapped from physical address 0.
> > (well, maybe the "suser" part should be consistent.  but it isn't
> > what you are asking, right?)
> I'm interested in both knowing if these are not wrong (like you say,
> amd64 is wrong, maybe others are too?) and I also want to use kauth(9)
> there. So before I'm writing the code I'm verifying that the current
> behavior is correct.

i386 and alpha seems sane.  amd64 seems wrong.
i don't know others.

> > whether each versions are correct or not is a different question. :-)
> > at least amd64 version seems wrong.  it should be basically the same as i386.
> Do you want to fix it?

i don't think i'll fix it in a timely manner.

> or should we wait for the amd64 port-master?

no.  it shouldn't be too hard for anyone interested.