On Fri, 16 Jan 2026, Kamila Szewczyk wrote:

> > So any patch proposed before the feature is in C2y should have 0 there,
> > not 1, and there should be corresponding tests of how -pedantic and
> > -pedantic-errors behave (in addition to such tests in C23 mode).
> 
> Could you elaborate on this?

(a) The feature is *not* currently in C2y, so a patch proposing it is 
proposing an extension, which -std=c2y -pedantic should diagnose.

(b) There should be tests that -std=c23 -pedantic gives a warning for 
this, -std=c23 -pedantic-errors gives an error; and similarly for 
-std=c2y, unless and until the feature gets into C2y.

-- 
Joseph S. Myers
[email protected]

Reply via email to