Re: correct implementation of FE_READ_UNCORRECTED_BLOCKS

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, 2009-08-02 at 20:56 +0300, Aleksandr V. Piskunov wrote:
> Oops, sent it way too fast. Anyway.
> 
> DVB API documentation says:
> "This ioctl call returns the number of uncorrected blocks detected by
> the device driver during its lifetime.... Note that the counter will
> wrap to zero after its maximum count has been reached."
> 
> Does it mean that correct implementation of frontend driver should
> keep its own counter of UNC blocks and increment it every time hardware
> reports such block?

No, but a frontend driver may wish to keep a software counter that is
wider than the hardware register counter, in case the hardware register
rolls over too frequently.


> >From what I see, a lot of current frontend drivers simply dump a value
> from some hardware register. For example zl10353 I got here reports 
> some N unc blocks and then gets back to reporting zero.

To support the use case of multiple user apps trying to collect UNC
block statistics, the driver should not zero out the UNC block counter
when read.  If the hardware zeros it automatically, then one probably
should maintain a software counter in the driver.

Regards,
Andy


--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux