Hi Ramiro, Am Montag, den 09.01.2017, 17:19 +0000 schrieb Ramiro Oliveira: > Hi Philipp > > On 1/9/2017 10:45 AM, Philipp Zabel wrote: > > Hi Ramiro, > > > > Am Dienstag, den 27.12.2016, 12:37 +0000 schrieb Ramiro Oliveira: > >> The optional functions weren't really optional so this patch makes them > >> really optional. > > > > Please add a bit of detail to the description. Since this changes the > > API, you should mention that the reset_control_get_optional variants now > > return NULL instead of an error if there is no matching reset phandle in > > the device tree and that the reset_control_* functions accept NULL rstc > > pointers. > > > > Would you be ok with something like this: > > "The *_get_optional_* functions weren't really optional so this patch makes > them > really optional. > > These *_get_optional_* functions will now return NULL instead of an error if > no > matching reset phandle is found in the DT, and all the reset_control_* > functions > now accept NULL rstc pointers
Yes, that looks fine to me. [...] > >> @@ -273,13 +295,13 @@ struct reset_control *__of_reset_control_get(struct > >> device_node *node, > >> index = of_property_match_string(node, > >> "reset-names", id); > >> if (index < 0) > >> - return ERR_PTR(-ENOENT); > >> + return optional ? NULL : ERR_PTR(-ENOENT); > > > > of_property_match_string can return -EINVAL, -ENODATA, or -EILSEQ. > > I think -EILSEQ should still be returned. > > > > I'll make the function return NULL, -ENOENT, or -EILSEQ. Ok. > >> } > >> > >> ret = of_parse_phandle_with_args(node, "resets", "#reset-cells", > >> index, &args); > >> if (ret) > >> - return ERR_PTR(ret); > >> + return optional ? NULL : ERR_PTR(ret); > > > > of_parse_phandle_with_args can return -ENOENT or -EINVAL. > > I think -EINVAL should still be returned. > > > > Same as above. I'll make the function return NULL, -ENOENT, or -EINVAL. thanks Philipp