Re: [PATCH 2/2] blk-stat: add a poll_size value to the request_queue struct

On 03/28/2017 01:38 PM, Stephen  Bates wrote:

quoted

quoted

 In order to bucket IO for the polling algorithm we use a sysfs entry
 to set the filter value. It is signed and we will use that as follows:

  0   : No filtering. All IO are considered in stat generation
  > 0 : Filtering based on IO of exactly this size only.
  < 0 : Filtering based on IO less than or equal to -1 time this value.>

I'd say that this is a fairly non-trivial semantic meanning to this...

Is there any use for the size exact match filter? If not then
I suggest we always make it (<=) in its semantics...

Thanks for the review Sagi. I�d be OK going with <=0 as the exact
match would normally be for minimal IO sizes (where <= and = are the
same thing). I will see what other feedback I get and aim to do a
respin soon�

No tunables for this, please. There's absolutely no reason why we should
need it.

-- 
Jens Axboe