* Re: blktrace on my block device
@ 2011-07-07 9:53 Tao Ma
2011-07-07 9:53 ` Gangadhar Mylapuram
2011-07-07 12:57 ` Gangadhar Mylapuram
0 siblings, 2 replies; 3+ messages in thread
From: Tao Ma @ 2011-07-07 9:53 UTC (permalink / raw)
To: linux-btrace
On 07/07/2011 05:41 PM, Gangadhar Mylapuram wrote:
> Hi,
>
> As for my understanding I no need to handle (BLKRESTART, BLKTRACESTOP,
> BLKTRACESETUP ..) in my device ioctl handling. It is block l/O layer
> which intercepts the
> statistics and does the job.
>
> But for my device (not sda or hda), blktrace only show the Q events.
> It is not showing even C (completed) events. Do I need to add the
> support in my device ioctl handling code?
These trace pointers are only recorded if they are met.
So I guess the reason is that your device implements your own
make_request_fn(So you can see 'Q' event) and no other trace point is
met at all.
Thanks,
Tao
^ permalink raw reply [flat|nested] 3+ messages in thread
* blktrace on my block device
2011-07-07 9:53 blktrace on my block device Tao Ma
@ 2011-07-07 9:53 ` Gangadhar Mylapuram
2011-07-07 12:57 ` Gangadhar Mylapuram
1 sibling, 0 replies; 3+ messages in thread
From: Gangadhar Mylapuram @ 2011-07-07 9:53 UTC (permalink / raw)
To: linux-btrace
Hi,
As for my understanding I no need to handle (BLKRESTART, BLKTRACESTOP,
BLKTRACESETUP ..) in my device ioctl handling. It is block l/O layer
which intercepts the
statistics and does the job.
But for my device (not sda or hda), blktrace only show the Q events.
It is not showing even C (completed) events. Do I need to add the
support in my device ioctl handling code?
Thanks,
Gangadhar
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: blktrace on my block device
2011-07-07 9:53 blktrace on my block device Tao Ma
2011-07-07 9:53 ` Gangadhar Mylapuram
@ 2011-07-07 12:57 ` Gangadhar Mylapuram
1 sibling, 0 replies; 3+ messages in thread
From: Gangadhar Mylapuram @ 2011-07-07 12:57 UTC (permalink / raw)
To: linux-btrace
Thanks. yes, that is the reason. I just found that our block driver is
not using request queue because our device is flash memory.
-Gangadhar
On Thu, Jul 7, 2011 at 3:23 PM, Tao Ma <tm@tao.ma> wrote:
> On 07/07/2011 05:41 PM, Gangadhar Mylapuram wrote:
>> Hi,
>>
>> As for my understanding I no need to handle (BLKRESTART, BLKTRACESTOP,
>> BLKTRACESETUP ..) in my device ioctl handling. It is block l/O layer
>> which intercepts the
>> statistics and does the job.
>>
>> But for my device (not sda or hda), blktrace only show the Q events.
>> It is not showing even C (completed) events. Do I need to add the
>> support in my device ioctl handling code?
> These trace pointers are only recorded if they are met.
> So I guess the reason is that your device implements your own
> make_request_fn(So you can see 'Q' event) and no other trace point is
> met at all.
>
> Thanks,
> Tao
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2011-07-07 12:57 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-07-07 9:53 blktrace on my block device Tao Ma
2011-07-07 9:53 ` Gangadhar Mylapuram
2011-07-07 12:57 ` Gangadhar Mylapuram
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox