lkml.org 
[lkml]   [1998]   [Jul]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Strange interrupt behaviour


On Mon, 13 Jul 1998 ralf@uni-koblenz.de wrote:
>
> One of the fundamental proporties of the buddy system is that it only
> allocations PAGE_SIZE << order aligned chunks. This is unnecessary for
> most allocations whith order > 0.

Actually, for the case we were talking of (fork()), the fact that we get
8kB-aligned 8kB chuks is _critical_, as the whole "current" thing depends
on it (getting to "current" by doing an 'and' of ~8191).

Not to mention the fact that it is almost impossible to come up with an
allocator that would return unaligned blocks anyway (apart from the very
simple "freelist" implementation, which was actually what Linux used to
use. However, if you think buddy fragments, think again. Buddy fragments a
lot less than freelist has ever done).

Note that we do have rather special allocation patterns, and the
allocation patterns may be the solution to the problem - do an allocator
that is not very good general-purpose, but works well with our patterns..

Linus


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html

\
 
 \ /
  Last update: 2005-03-22 13:43    [W:0.099 / U:0.388 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site