Daniel =?utf-8?q?Rodríguez?= <[email protected]>,
Daniel =?utf-8?q?Rodríguez?= <[email protected]>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/[email protected]>


================
@@ -10631,6 +10631,9 @@ def warn_printf_data_arg_not_used : Warning<
   "data argument not used by format string">, InGroup<FormatExtraArgs>;
 def warn_format_invalid_conversion : Warning<
   "invalid conversion specifier '%0'">, InGroup<FormatInvalidSpecifier>;
+def warn_format_conversion_specifier_requires_c23 : Warning<
+  "conversion specifier '%0' requires a C standard library compatible with "
+  "C23; data argument may not be used by format">, InGroup<Format>;
----------------
AaronBallman wrote:

I think we should add a new diagnostic group for this -- users shouldn't have 
to disable *all* format warnings just to silence this one. We have an existing 
diagnostic group that could be reasonable (`-Wformat-invalid-specifier`), but 
this could also warrant its own group because `%b` is only *sometime* invalid, 
depending on what versions of C the standard library supports as opposed to a 
wholly invalid specifier which is never supported. But maybe I'm splitting 
hairs?

That said, I want to rope in some llvm-libc folks for their opinions on the PR 
in general, so CC @michaelrj-google @frobtech 

https://github.com/llvm/llvm-project/pull/126694
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to