> >i've just finished the patches that allow multiple (cached) writers to
> >write into the pagecache.
>
> I was talking about the write(2) performances of only one writer. `cp
> /dev/zero /tmp'. Probably they are better but the wait_on_buffer in
> flushpage make me not sure of that. Also considering that waiting for I/O
> completation for every flushpage means a very high scheduling rate during
> writes. I still have to verify in practice these things though.
i have just rewritten flushpage (to fix a bug, not necesserily for
performance) - could you wait for these initial things to settle down?
These changes should show up Really Soon :)
-- mingo
-
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/