vaaarg3 failure

[Geoffrey Keating]

Hi Jason,

the regression tester was having trouble sending mail, but this failure:

> native g++.sum g++.old-deja/g++.pt/vaarg3.C

appeared between 2003-12-09T00:12:54Z and 2003-12-09T08:04:28Z, and 
during that time the only changes to the C++ compiler were:

  2003-12-08  Jason Merrill  <jason@redhat.com>

+       PR c++/11971
+       * tree.c (build_local_temp): Split out from build_cplus_new.
+       (force_target_expr): New fn.
+       * call.c (call_builtin_trap): Call it.  Take a type parm.
+       (convert_arg_to_ellipsis): Pass it.
+       (build_x_va_arg): Use call_builtin_trap.
+
+       PR c++/11929
+       * call.c (magic_varargs_p): New fn.
+       (build_over_call): Do no ellipsis conversions for arguments to
+       functions with magic varargs.
+

and it sure looks like the margic_varargs_p patch could have caused 
this regression.  Could you look at it?

The actual failure is:

FAIL: g++.old-deja/g++.pt/vaarg3.C  (test for errors, line 15)

and line 15 is:

va_start (argp, somearg); // { dg-error "" } cannot pass non-POD

I am not sure if the testsuite should be updated or if this really 
should be an error.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2361 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-regression/attachments/20031212/828f6747/attachment.p7s>