Subject: Re: cp(1) vs. UVM/UBC
To: Jason R Thorpe <>
From: Matthias Scheler <>
List: tech-userlevel
Date: 07/18/2001 17:47:41
On Wed, Jul 18, 2001 at 08:27:15AM -0700, Jason R. Thorpe wrote:
>  > our cp(1) has code which uses mmap(2) and write(2) instead of read(2) and
>  > write(2) to copy files. The mmap(2) implementation is not used because
>  > "VM_AND_BUFFER_CACHE_SYNCHRONIZED" is not defined. My question is if
>  > VM and buffer cache are synchronized after the integration of UBC?
> Yes, except for on LFS -- on LFS, mmap doesn't work right now because
> its UBC'ification hasn't been finished yet.

So turning that option on would break "cp" on LFS. How fatal would that
be? Is LFS usable at the moment despite the mmap() problem?

	Kind regards

Matthias Scheler