Messages in this thread | | | Date | Mon, 9 Apr 2012 17:09:22 +0800 | From | Dong Aisheng <> | Subject | Re: [PATCH 1/1] pinctrl: fix pinmux_check_ops error checking |
| |
On Mon, Apr 09, 2012 at 11:31:03AM +0800, Viresh Kumar wrote: > On 4/6/2012 9:02 PM, Stephen Warren wrote: > > On 04/06/2012 06:18 AM, Dong Aisheng wrote: > >> > From: Dong Aisheng <dong.aisheng@linaro.org> > >> > > >> > Do not use get_functions_count before checking. > >> > > >> > Signed-off-by: Dong Aisheng <dong.aisheng@linaro.org> > > Acked-by: Stephen Warren <swarren@wwwdotorg.org> > > > > IIRC, Viresh added some other get_*_count() functions in the same patch. > > Do we need any more fixes similar to the this for those too? > > Might not be required as this patch handles the pinmux_check_ops() routine. > At other places ops is guaranteed to have correct value. > Yes, core can guarantee it after pinmux_check_ops. But i noticed for user interfaces, we may still need add some error checking to avoid kernel oops like pinctrl_get_* and debugfs interfaces in case the mux or config function is not supported by driver. I'm going to send one more patch to fix it.
Regards Dong Aisheng
| |