------- Additional Comments From mmitchel at gcc dot gnu dot org 2004-12-21
21:06 -------
This is not very easy to fix.
The first warning comes from the parser, during name-lookup, when it realizes
that the name resolves to a deprecated function. The second warning comes after
overload resolution, when we select a function to call. If func was really
overloaded, then the parser would not warn, so we need the warning after
overload resolution. However, if the function is never called, we want the
warning from the parser. Because the FUNCTION_DECL is doubling as a node in the
expression tree, we can't mark it as "warning issued" without losing any
subsequent warnings about other uses. Joseph and I both agree that we really
want some kind of separate parser-expression structure to use for exactly this
kind of thing, but we haven't got it yet.
This is not a high-priority item, so I'm downgrading it to P3. I'm also
removing the target milestone, despite being a regression, as this is a minor
annoyance, and would never block a release.
--
What |Removed |Added
----------------------------------------------------------------------------
Priority|P2 |P3
Target Milestone|3.4.4 |---
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17729
