[LAD] Ring buffers again

Will J Godfrey WillGodfrey at musically.me.uk
Fri May 23 18:14:43 UTC 2014


On Thu, 22 May 2014 17:03:52 -0400
Paul Davis <paul at linuxaudiosystems.com> wrote:

> On Thu, May 22, 2014 at 4:46 PM, Will Godfrey
> <willgodfrey at musically.me.uk>wrote:
> 
> >
> > As well as the jack ring buffer, I've looked at several others now, and
> > their
> > example code. The most significant thing that seems to be different about
> > them
> > (from a usage point of view) is the way they handle overflows. However, if
> > the
> > buffer size is defined as an exact multiple of the data type/structure and
> > only
> > complete structures are pushed or popped, would I be right in thinking
> > that you
> > would only need to check on an all/none basis?
> >
> > Have I missed something that could cause a partial data transfer?
> >
> 
> the jack ring buffers are byte-oriented, so you do have to be careful.
> however, if both the reader and writer only ever increment their respective
> pointer/index in multiples of the same basic byte count, then you should be
> OK.
> 
> if you use a C++ template ring buffer, then you necessarily cannot get
> partial transfers.

Thanks. It seems I'm on the right track then :)

-- 
It wasn't me! (Well actually, it probably was)

... the hard part is not dodging what life throws at you,
but trying to catch the good bits.


More information about the Linux-audio-dev mailing list