Messages in this thread | | | Date | Wed, 1 Sep 1999 09:10:19 +0200 (CEST) | From | Andrea Arcangeli <> | Subject | Re: vm kills processes in our 2.3.12 port of reiserfs - what was the story on the changes to mark_buffer_dirty() and the too many dirtybuffers issue? |
| |
On Wed, 1 Sep 1999, Hans Reiser wrote:
>whole file for small files. I don't understand why you guys think that >reiserfs is somehow harder to integrate with the page cache (except for
I didn't wanted to tell that. What I wanted to tell is that if you'll continue to duplicate the __data__ of the files in both the buffer cache and in the page-cache as always happens in 2.2.x, IMHO then you are going to be slower in real-world. This because IMHO in most of cases the filesystem performances are far less critical than the cache performances. And so if you waste memory by duplicating the cache, you'll be slower because you will be able to cache only half of the data in the worst case. That's a cohernecy issue.
NOTE: I am not talking about any filesystem in particular.
Andrea
- 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.tux.org/lkml/
| |