On Sun, 2011-12-04 at 18:22 +0200, Michael S. Tsirkin wrote:

On Sun, Dec 04, 2011 at 02:13:51PM +0200, Sasha Levin wrote:

quoted

On Sun, 2011-12-04 at 13:52 +0200, Avi Kivity wrote:

quoted

On 12/03/2011 01:50 PM, Sasha Levin wrote:

quoted

On Fri, 2011-12-02 at 11:16 +1030, Rusty Russell wrote:

quoted

On Thu, 1 Dec 2011 12:26:42 +0200, "Michael S. Tsirkin" [off-list ref] wrote:

quoted

On Thu, Dec 01, 2011 at 10:09:37AM +0200, Sasha Levin wrote:

quoted

On Thu, 2011-12-01 at 09:58 +0200, Michael S. Tsirkin wrote:

quoted

We'll presumably need some logic to increment is back,
to account for random workload changes.
Something like slow start?

We can increment it each time the queue was less than 10% full, it
should act like slow start, no?

No, we really shouldn't get an empty ring as long as things behave
well. What I meant is something like:

I was thinking of the network output case, but you're right.  We need to
distinguish between usually full (eg. virtio-net input) and usually
empty (eg. virtio-net output).

The signal for "we to pack more into the ring" is different.  We could
use some hacky heuristic like "out == 0" but I'd rather make it explicit
when we set up the virtqueue.

Our other alternative, moving the logic to the driver, is worse.

As to fading the effect over time, that's harder.  We have to deplete
the ring quite a few times before it turns into always-indirect.  We
could back off every time the ring is totally idle, but that may hurt
bursty traffic.  Let's try simple first?

I tried to take a different approach, and tried putting the indirect
descriptors in a kmem_cache as Michael suggested. The benchmarks showed
that this way virtio-net actually worked faster with indirect on even in
a single stream.

How much better?

host->guest was same with both indirect on and off, and guest->host went
up by 5% with indirect on.

This was just a simple 1 TCP stream test.

I'm confused. didn't you see a bigger benefit for guest->host by
switching indirect off?

The 5% improvement is over the 'regular' indirect on, not over indirect
off. Sorry for the confusion there.

I suggested this change regardless of the outcome of indirect descriptor
threshold discussion, since it would help anyways.

-- 

Sasha.