cvs commit: src/sys/kern kern_poll.c

[Nate Lawson]

Robert Watson wrote:
> 
> On Tue, 6 Sep 2005, Luigi Rizzo wrote:
> 
>> On Tue, Sep 06, 2005 at 10:18:28AM +0400, Gleb Smirnoff wrote:
>>
>>>   Luigi,
>>
>> ...
>>
>>> The idlepoll thread is single.
>>
>>
>> ok this is very good. Re. netisr vs idlepoll, perhaps a way could be 
>> to bump the idlepoll priority very high upon a net soft interrupt, and 
>> drop it down to its normal value once done with the netisr cycle. so 
>> we don't have to arbitrate among the two.
> 
> Also, if we gradually move to a polling model that handles polling for 
> non-network devices, it would result in a rather mixed model.  One of 
> the challenges of moving to a mixed polling model (one that supports 
> non-network devices) is that network devices have a fairly well 
> understood currency for work: processing of packets.  Other devices may 
> have less well understood, or at least not easily comparable, workloads...

For the case of storage, you actually have a better model since all 
transactions are initiated from the host (as opposed to packet 
arrivals).  This gives an easy metric for a dynamic polling threshold -- 
if you have a deep queue of outstanding requests and one completes, you 
should poll a little more than normal.

-- 
Nate