On 10-12-2011 11:09, Antti Palosaari wrote:
On 12/10/2011 02:47 PM, Mauro Carvalho Chehab wrote:
On 10-12-2011 09:47, Antti Palosaari wrote:
Hello Manu,
That patch looks now much acceptable than the older for my eyes, since
you removed that .set_state() (change from .set_params() to
.set_state()) I criticized. Thanks!
On 12/10/2011 06:44 AM, Manu Abraham wrote:
static int cxd2820r_set_frontend(struct dvb_frontend *fe,
[...]
+ switch (c->delivery_system) {
+ case SYS_DVBT:
+ ret = cxd2820r_init_t(fe);
+ ret = cxd2820r_set_frontend_t(fe, p);
Anyhow, I don't now like idea you have put .init() calls to
.set_frontend(). Could you move .init() happen in .init() callback as
it was earlier?
Changing .init() to happen at set_frontend() actually makes sense for me ;)
Why didn't you like this change?
Because there is already callback named .init() just for that as I guess. .init() differs from .set_frontend() as init is called only once when device is opened whilst .set_frontend() is called every time when tuning request is done.
Anyhow, it is up to decision remove .init() from the frontends callbacks as we can live without.
IMO, it makes sense to not do any init before set_frontend on most devices.
An obvious exception would be if is there any device where the capabilities need
to be discovered at runtime.
Regards,
Mauro
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
