Andrea Arcangeli wrote:
> >Ingo's "low latency" patch should fix the kernel-hogs-CPU shouldn't it?
>
> That's not the case because you never spin while waiting disk I/O
> completation.
Argh, I didn't quite catch whether Khimenko's comment referred to both
cases starting this thread or just the second. He said the kernel uses
up lots of CPU, which applies more to the first example than the second.
Of course, (second example) if the system is always waiting for I/O it's
not doing it in the kernel. That can be improved only be reordering
I/Os and changing the paging strategy.
For the first example, that system wasn't swapping at all and the data
set fit easily into memory. It should not be blocking at all, and I can
only assume the low performance is due to some slow processing in kernel
space.
Latency of slow processing is supposed to be fixed by the low latency
patch :-)
-- Jamie
-
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/
This archive was generated by hypermail 2b29 : Wed Mar 15 2000 - 21:00:14 EST