[mythtv] [PATCH] RingBuffer - fsync thread

[Torbjörn Jansson]

mythtv-dev-bounces at mythtv.org <> wrote:
> On Sunday 12 December 2004 12:42 am, Hamish Moffatt wrote:
>> On Sat, Dec 11, 2004 at 08:57:37PM -0700, John Patrick Poet wrote:
>>> As Isaac suggested, this patch moves the fdatasync(fd) call to it's
>>> own thread.  The means that the ThreadedFileWriter does not have to
>>> wait for that call to return before processing data.
>> 
>> What is the need to force a sync anyway?
> 
> Kernel tends to store up unsynced writes for a while and then decide
> to commit them to disk all at once without it, which can cause a nice
> little cascading hiccup in the recording.
> 
> Isaac

Does anyone know if this happens with the linux 2.6 CFQ io scheduler too?
My guess is that it doesn't