Messages in this thread | | | Subject | Re: [PATCH] [6/6] kfifo: Document everywhere that size has to be power of two | From | Stefani Seibold <> | Date | Sun, 27 Dec 2009 23:23:31 +0100 |
| |
Am Sonntag, den 27.12.2009, 14:14 -0800 schrieb Dmitry Torokhov: > On Sun, Dec 27, 2009 at 10:50:32PM +0100, Stefani Seibold wrote: > > Am Sonntag, den 27.12.2009, 22:03 +0100 schrieb Andi Kleen: > > > On my first try using them I missed that the fifos need to > > > be power of two, resulting in a runtime bug. Document that requirement > > > everywhere (and fix one grammar bug) > > > > > > Signed-off-by: Andi Kleen <ak@linux.intel.com> > > > > > > > Acked: Stefani Seibold <Stefani@Seibold.net> > > > > Now, does it really _have_ to be this way? For record-oriented > FIFO's it is customary to be 2*pow(n)*sizeof(rec) which is rarely > 2*pow(m). >
Right. This is why i wrote the new macro base kfifo API, where i am able to do this in that way.
Stefani
| |