Hi Pierrick, December 5, 2024 at 6:35 PM, "Pierrick Bouvier" wrote: > On 12/5/24 04:44, Julian Ganz wrote: > > > > > Hi Pierrick, > > December 4, 2024 at 11:45 PM, "Pierrick Bouvier" wrote: > > > > > > > > On 12/2/24 11:26, Julian Ganz wrote: > > > > > include/qemu/plugin.h | 1 + > > include/qemu/qemu-plugin.h | 43 ++++++++++++++++++++++++++++++++++++++ > > 2 files changed, 44 insertions(+) > > diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h > > index 9726a9ebf3..27a176b631 100644 > > --- a/include/qemu/plugin.h > > +++ b/include/qemu/plugin.h > > <snip> > > + > > +/** > > + * typedef qemu_plugin_vcpu_discon_cb_t - vcpu discontinuity callback > > + * @vcpu_index: the current vcpu context > > + * @type: the type of discontinuity > > + * @from_pc: the source of the discontinuity, e.g. the PC before the > > + * transition > > + * @to_pc: the PC pointing to the next instruction to be executed > > + * > > > > > > > > Missing those parameters when building doc. > > > include/qemu/qemu-plugin.h:198: warning: Function parameter or member > > > 'id' not described in 'qemu_plugin_vcpu_discon_cb_t' > > > include/qemu/qemu-plugin.h:289: warning: Function parameter or member > > > 'type' not described in 'qemu_plugin_register_vcpu_discon_cb' > > > 2 warnings as Errors > > > > > Yes, I forgot about id. But type is clearly documented. Maybe the tool > > is confused about the name and thinks it's a reserved word or something? > > In that case I better change that to something else. > > > The type (qemu_plugin_discon_type) is documented, but the parameter "type" is > not. Even if the name is correctly chosen, you still need to make it appear > in the doc.
Yes. I didn't realize that the two warnings were for different entities because I somehow failed to parse the entire line. Regards, Julian Ganz
