The analysis of CFQ for Linux 2.6.28.5

thewayma

The
analysis of CFQ for Linux 2.6.28.5
       
       
       
       
How to add
request:


       
       
       
       
Instruction:
(1)(2)(5):
Every
process in CFQ has a correspond queue whose name is cfqq. Actually,
each cfqq has two queues, like DEADLINE, one queue is sorted by LBA
of requests, the other queue is sorted by arrival time. The queue
sorted by LBA is actually a red-black tree while the queue sorted by
arrival time is called FIFO queue.
(6)(7)(8):
        unsigned
long elapsed = jiffies - cic->last_end_request;
        unsigned
long ttime = min(elapsed, 2UL * cfqd->cfq_slice_idle);
        cic->ttime_samples
= (7*cic->ttime_samples + 256) / 8;
        cic->ttime_total
= (7*cic->ttime_total + 256*[color="#ff0000"]ttime)
/ 8;
        cic->[color="#ff0000"]ttime_mean
= (cic->ttime_total + 128) / cic->ttime_samples;
      
**********************************************************
        sector_t
sdist;
        u64
total;
       if
(cic->last_request_pos sector)
                      sdist
= rq->sector - cic->last_request_pos;
        else
                      sdist
= cic->last_request_pos - rq->sector;
        if
(cic->seek_samples
                      sdist
= min(sdist, (cic->seek_mean * 4) + 2*1024*1024);
        else
                      sdist
= min(sdist, (cic->seek_mean * 4) + 2*1024*64);
        cic->seek_samples
= (7*cic->seek_samples + 256) / 8;
        cic->seek_total
= (7*cic->seek_total + (u64)256*[color="#ff0000"]sdist)
/ 8;
        total
= cic->seek_total + (cic->seek_samples/2);
        do_div(total,
cic->seek_samples);
        cic->[color="#ff0000"]seek_mean
= (sector_t)total;
      
**********************************************************
Idle_window:
Every cfqq
has a flag to indicate whether this process is worth to do
anticipation. In a simple word, I use Idle_window to describe it. If
this parameter is set on, it means the I/O scheduler can do
anticipation(whether it should do still depend on other conditions,
however, if this parameter is set off, the I/O scheduler will never
do anticipation)
How to set
idle_window?

If
        this process is not sync, do not set;
       

If
        the priority of this process is idle, do not set;
       

If
        the process might have exited, do not set;
       

If
        the user has tuned off cfq_slice_idle(set it to 0), do not set;
       

If
        the process is seeky((cic)->seek_mean > (8 * 1024)), do not
        set;
       

If
        cic->ttime_mean > cfqd->cfq_slice_idle, do not set;
       

otherwise,
        set it;
How to
dispatch requests:
Suppose a
request queue has been selected:


       
       
       
       
The
condition to decide whether it is worth to wait:

there
        is idle_slice_timer or
       

there
        is req serving in the device driver from this cfqq and the
        idle_window of this cfqq is turned on
       
       
       
       
When a
request is served by the device driver, the I/O scheduler will check
whether the current active_queue is the same queue with this request,
if it is, the scheduler then will check whether this cfqq need to do
anticipation.

       
       
       
       
The
tunable parameters:
back_seek_max:
there is a
routine named cfq_choose_req which is used to select the next request
to be dispatched. Since the selection is occurred in the sorted
queue, it is just the same as AS which is based on DEADLINE. The next
request might be the nearest front request from the current disk head
or the nearest back request from the current disk head. However, the
I/O scheduler always want to select the request lied front of the
disk head, which means there is a limitation distance to choose the
request lied back. This is just the value.
back_seek_penalty:
The
distance between back request and current disk head should multi this
value as a penalty. In other word, this back request should be near
the current disk head enough so that it can be selected to be the
next request.
fifo_expire_async:
fifo_expire_sync:
Actually,
every process has two queues, one for sync, the other is for async.
And each sync actually has two queues, one for fifo, the other for
sorted. From this point of view, there are 4 queues for a process.
But generally, we use one queue to stand for a process. The two
parameters are the time expiration for fifo of sync and async.
Quantum:
there is a
limitation for the number of requests serving in the device driver.
Since most sata now support NCQ, the I/O scheduler can dispatch more
than one request to the device driver. Actually, I am not quite
understand the relationship between NCQ and the number of requests
serving in the device driver.
slice_sync:
slice_async:
time slice
for the queue
slice_async_rq:
I am not
sure about this parameter, since after dispatching the requests to
the device driver, the I/O scheduler will check this parameter and
decide whether the cfqq should be expired immediately. This parameter
is for async request.
slice_idle:
The
anticipation time for sync queue. There is no anticipation for async
queue.
               
               
               

本文来自ChinaUnix博客，如果查看原文请点：http://blog.chinaunix.net/u2/86301/showart_2178614.html