Subject: Re: uvm_km_valloc_align() and USPACE (Re: CVS commit: syssrc)
To: None <soda@sra.co.jp>
From: Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>
List: tech-kern
Date: 11/29/2000 21:02:33
In <200011290956.SAA02294@srapc342.sra.co.jp>
soda@sra.co.jp wrote:

> 	Jason R Thorpe <thorpej@zembu.com> said:
> 
> > Yah, it should just be 2 ... wonder why some of them are using 3 pages?!

> Is there any reason that UPAGES should be 2?

In general, there is no reason. But I guess
Jason says about m68k ports which have UPAGES == 3.

Currently, UPAGES is 3 on luna68k, mac68k, mvme68, news68k,
next68k and x68k. I think all these ports are based on hp300,
and according to cvsweb UPAGES on hp300 was changed from 3 to 2
on 4.4BSD-Lite merge. There is no reason it should be 3 neither.
(I just copied it from mvme68k on porting news68k ;-)

> For me, adding a symbol like USPACE_ALIGNMENT and use the symbol
> for uvm_km_valloc_align seems to be the right thing.
> 
> The USPACE_ALIGNMENT on mips should be USPACE,
> and the value on other architectures should be NBPG.

I agree with this point.
---
Izumi Tsutsui
tsutsui@ceres.dti.ne.jp