Re: [Qemu-devel] Re: [kvm-devel] [PATCH 0/4] Rework alarm timer infrastrucure - take2

[malc <av1474@comtv.ru>]

On Mon, 20 Aug 2007, Luca Tettamanti wrote:

> Il Sun, Aug 19, 2007 at 10:31:26PM +0300, Avi Kivity ha scritto:
>> Luca wrote:
>>> On 8/19/07, Luca Tettamanti <kronos.it@gmail.com> wrote:
>>>
>>>> +static uint64_t qemu_next_deadline(void) {
>>>> +    uint64_t nearest_delta_us = ULLONG_MAX;
>>>> +    uint64_t vmdelta_us;
>>>>
>>>
>>> Hum, I introduced a bug here... those vars should be signed.
>>>
>>> On the overhead introduced: how do you measure it?
>>>
>>>
>>
>> Run a 100Hz guest, measure cpu usage using something accurate like
>> cyclesoak, with and without dynticks, with and without kvm.
>
> Ok, here I've measured the CPU usage on the host when running an idle
> guest.
>

[..snip the numbers..]

After briefly looking at the cyclesoak it indeed looks like it does
the right thing, but considering the limitations of user-space only
approach it introduces some (sometimes really unwanted) variables
into the system, those can, and i guess will, influence things.

The upshot is this - if you have used any standard utility (iostat,
top - basically anything /proc/stat based) the accounting has a fair
chance of being inaccurate. If cyclesoak is what you have used then
the results should be better, but still i would be worried about
them.

In conclusion until kernel native accounting is fixed your best bet
is to use: http://www.boblycat.org/~malc/apc/

-- 
vale