Messages in this thread | | | Subject | Re: | From | Matt Mackall <> | Date | Thu, 26 Nov 2009 05:38:45 -0600 |
| |
On Thu, 2009-11-26 at 10:49 +0000, Ian Molton wrote: > Hi guys, > > This version uses a statically allocated buffer. I dont feel it is a > good idea not to pass the address and length of the buffer to the hardware > drivers, as they shouldnt have intimate knowledge of the core, IMO.
I agree, but let me quote myself: --- > + if (!data_avail) { > + bytes_read = rng_get_data(current_rng, rng_buffer, > + RNG_BUFFSIZE, !(filp->f_flags & O_NONBLOCK));
No need to pass rng_buffer to the helper as there's only one with global scope. ---
-- http://selenic.com : development and support for Mercurial and Linux
| |