On Fri, 3 Sep 1999 18:50:31 +0200 (CEST), Andrea Arcangeli
<andrea@suse.de> said:
>> Nonsense! You do only one copy, inside ll_rw_block. You do it before
> If you don't show me the code it's hard to understand what are you doing
> exactly and what you mean with buffer-bounce.
Indeed, and I'm sorry if we seem to be confusing two separate issues.
There are two different ways forward here.
The quick and easy hack to get raw IO working is to do precisely what
the swap code does. That works fully synchronously, so it's easy to do.
> You said you was doing what my swap code does. But my code replaces
> the memory completly, it's not only an helper memory for the I/O.
Yes: the raw IO quick fix is just like swap. The harder problem is to
hide it inside ll_rw_block(), which we _do_ want to do if we move the
page cache to high memory. Only with ll_rw_block doing the hard work
can we do asynchronous kmap fixup, which we need for things like
readahead. Swap and raw IO are synchronous for now.
--Stephen
-
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/