[Bug c/16016] New: xfailed gcc.dg/20030612-1.c

["rth at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

This new warning was added to tree-ssa branch by Jason.  I seem to recall that
the object of this warning was control flow related, but that's certainly not
what the text of the warning itself implies.  We need to come to some resolution
as to exactly what we're intending so that it can be reimplemented or dropped.

Implementing the text of the warning (warn for any void type returned from a
statement expression) results in bootstrap failure, due to warnings generated
from obstack.h.

Possibly what we intend is for statement expressions to end in an expression?
E.g. it's legal to end in "(void)0;" because that's an expression, but not
"return 3;" because that isn't?  If that's true, we should warn for *any* 
statement that's not an expression -- if, while, for, and so forth.  I have no
data as to how that affects code in the wild.

-- 
           Summary: xfailed gcc.dg/20030612-1.c
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: rth at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org,jason at gcc dot gnu dot
                    org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16016