From: Fred <fr...@sp...> - 2013-01-31 13:38:08
|
Hi Stuart, On 30/01/2013, at 8:43, Stuart Brady <sd...@zu...> wrote: > On Tue, Jan 29, 2013 at 11:44:19PM +1100, Fred wrote: > >> The aim of this change was not to produce a comprehensive solution to any of these issues, but I concede your point that we could be a bit better by having the reset functions for peripherals having the form: >> >> static void >> <peripheral>_reset( int hard_reset ) >> { >> <peripheral>_available = 0; >> <peripheral>_paged = 0; >> >> if( !periph_is_active( PERIPH_TYPE_<PERIPHERAL> ) ) >> return; >> >> I'll make that change when the opportunity arises over the next week or so. > > Thanks! I assume this means before the release, in that case. :-) Yes, indeed work is in progress and I think the release will still take a week or two. > I would perhaps be willing to have a go at this myself -- this would > effectively mean reverting r4864, updating spectranet_reset as described, > and then doing something about speccyboot and perhaps one or two of the > other interfaces, although I'm not sure which! That's cool, I'm happy to round out this change unless you fancy knocking it off tomorrow :) > Would the changes that I was suggesting regarding printer emulation be > acceptable for the release? I feel it's a fairly small set of changes > and I have a fairly good idea of what needs doing. If the code is done in time I think that's fine, but I wouldn't hold up a release for it as I think what we have is no worse and a little better than we've always had. > The scld_memory_map() change that was made seems odd to me as I could > have expected the function to return early at the start if the > PERIPH_TYPE_SCLD peripheral is *not* active. There seems to be a '!' > missing here, although maybe I've misunderstood something. Yes, I missed the ! Fred |