On 3/1/20 8:08 AM, Tomer Maimon wrote: > Sorry Guenter probebly I didnt explain it well. > > > On Sun, 1 Mar 2020 at 12:48, Guenter Roeck <linux@xxxxxxxxxxxx <mailto:linux@xxxxxxxxxxxx>> wrote: > > On 3/1/20 1:40 AM, Tomer Maimon wrote: > > During probe NPCM watchdog sets the following bootstatus flags: > > - WDIOF_CARDRESET represent power and core reset. > > - WDIOF_EXTERN1 represent watchdog 0-2 reset. > > - WDIOF_EXTERN2 represent software 1-4 reset. > > > > Each flag is representing a group of bootstatus. > > The user can configure through the device treethe exact reset > > to each flag group. > > > > Sorry, this doesn't make sense to me. I could understand reporting > the above, but it looks to me like devicetree is used to associate > a reset bit from the controller with one of the above. > Devicetree only seems to be used to associate reset status bits > from the controller with WDIOF_CARDRESET, WDIOF_EXTERN1, or > WDIOF_EXTERN2. That adds a lot of complexity for little if any > gain. > > > > It would make sense to set the bootstatus bits as suggested above, > but that doesn't require devicetree properties. > > More comments inline. > > Guenter > > > > In the NPCM750 we have the following reset types: > > 1. board reset (Power on reset, Core reset) > 2. WD reset (0-2 WD reset). > 3. SW reset (1-4 SW reset). > > > Each board can use different reset types, because in the WD status bit there is not enough bits to represent the entire NPCM750 resets. > > The NPCM750 reset groups are represent as follow: > > - WDIOF_CARDRESET represent power and core reset. > - WDIOF_EXTERN1 represent watchdog 0-2 reset. > - WDIOF_EXTERN2 represent software 1-4 reset. > Exactly, and I don't see a need to be more specific than that. This can be implemented without all the DT complexity. Guenter
