On Tue, May 19, 2026 at 01:10:54PM -0600, Tom Rini wrote: > On Tue, May 19, 2026 at 12:28:56PM -0600, Tom Rini wrote: > > > To make sure we cannot encounter the case where code is written to use > > the event framework, but EVENT is not enabled, guard most of the header > > with a build-time check. This will promote the case in question to a > > build time failure and not a runtime debug issue. > > > > Signed-off-by: Tom Rini <[email protected]> > > --- > > include/event.h | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --git a/include/event.h b/include/event.h > > index 3ce5f992b04e..71fb14391f25 100644 > > --- a/include/event.h > > +++ b/include/event.h > > @@ -308,6 +308,7 @@ static inline const char *event_spy_id(struct > > evspy_info *spy) > > #endif > > } > > > > +#ifdef CONFIG_EVENT > > /* > > * It seems that LTO will drop list entries if it decides they are not > > used, > > * although the conditions that cause this are unclear. > > @@ -348,6 +349,7 @@ static inline const char *event_spy_id(struct > > evspy_info *spy) > > __used ll_entry_declare(struct evspy_info_simple, \ > > _type ## _3_ ## _func, \ > > evspy_info) = _ESPY_REC_SIMPLE(_type, _func) > > +#endif > > > > /** > > * event_register - register a new spy > > This was really my v0 of the changes, I have made this more general and > will post a v2, of what I had intended, shortly.
On further testing, I'm going to stick with this version, trying to guard more of the file would require a larger restructure to avoid other compilation problems (the header being perhaps too broadly included, CONFIG_IS_ENABLED(...) checks, and so forth). -- Tom
signature.asc
Description: PGP signature
