12 May
2011
12 May
'11
3:48 p.m.
Hi Mark, Liam,
On Sun, May 8, 2011 at 18:42, Mark Brown broonie@opensource.wolfsonmicro.com wrote:
On Sun, May 08, 2011 at 08:44:41PM +0100, Liam Girdwood wrote:
On 08/05/11 16:56, Mark Brown wrote:
So it'll segfault or something instead of just not transferring data or whatever?
Yep, it will cause a CPU data abort exception in this case.
OK, that makes things clearer. It'd still be good to postpone the check for as long as possible, just in case things get racy - currently it feels too early in the process - but having the check makes sense.
I think if the check can be done in .hw_params. because this check makes sure the hw is available.