Subject: Re: Disk IO / UVM and Crypt FS..
To: enami tsugutomo <enami@but-b.or.jp>
From: Jorgen Lundman <lundman@lundman.net>
List: tech-kern
Date: 05/08/2002 14:23:17
I suspected that would be the issue.

Naturally, since it is in buffer cache it should be read only and 
nothing should be allowed to modify it.

However, in the interest of my small little cryptfs I naturally want to 
modify it, and I know I have to decrypt it once biodone is called. The 
presumption being that it is locked while IO is performed (disk_busy() 
?) and therefor not modified until I can decrypt it, and unbusy the disk.

Is there a solution for me? Would I have to produce a new "actual" 
buffer, which will be encrypted and put to disk and then dropped.

or a way to temporarily unmark it read-only, do my required 
modification. Will this work, or is it potential that something(?) can 
come and access the cache?

Lund


enami tsugutomo wrote:
>>genfs_gop_write(d7d7a180,d7d72cb8,3,11,c0821cfc) at genfs_gop_write+0x334
> 
> 
> uvm_pagermapin() maps the page cache pages with read-only protection.
> Since it is cached data itself, you can't modify it.
> 
> enami.
> 
> 


-- 
Jorgen "Lord" Lundman <lundman@lundman.net>
Technology Manager, Unix Administrator
Phone: +44 (0)20-86591860  Mobile: +44 (0)79-58642918
Pager: 07958642918@one2one.net
"Rare is the person who can weigh the faults of others
  without putting his thumb on the scales": Byron J. Langenfeld