[ILUG] C++ and recursive va_start calls

[Brian Foster]

  | From: Vincent Cunniffe <vincent at cunniffe.net>
  | Date: Mon, 16 Feb 2004 22:28:28 +0000
  | 
  | anyone had any experience with this?  I have a situation
  | where a function with a va_start call in it is being called
  | from inside another va_start, and it's corrupting the hell
  | out of the arguments.
  | 
  | If I perform exactly the same operation, but in sequence,
  | it works fine, but it's a lot less elegant.  I can't find
  | any mention of it being illegal.

 I am assuming here both va_start()s are using
 the _same_ va_list:  that is 100% illegal in C,
 and I know of no reason for C++ to be different.

 from the ISO C99 standard (C89 is very similar):

    7.15.1.4  The va_start macro
  [ ...(N.b. `ap' is a va_list object) ]
    [#3] The va_start macro initializes ap for subsequent use by
    va_arg  and  va_end.   va_start  (or  va_copy)  shall not be
    invoked  again  for  the  same  ap  without  an  intervening
    invocation of va_end for the same ap.
  [ ... ]
   J.2  Undefined behavior

    [#1]   The   behavior   is   undefined   in   the  following
    circumstances:
  [ ... ]
      -- The va_copy or va_start macro is called to initialize a
         va_list that was previously initialized by either macro
         without an intervening invocation of the  va_end  macro
         for the same va_list (7.15.1.2, 7.15.1.4).

 if you are using GNU gcc(1), or a C99-conformant
 compiler, you should be able to use va_copy() to
 duplicate the va_list (_before_ the first va_start()),
 and then va_start() the copy when needed.

 hope this helps!
cheers!
	-blf-
--
«How many surrealists does it take to    |  Brian Foster      Montpellier,
 change a lightbulb?  Three.  One calms  |  blf at utvinternet.ie      France
 the warthog, and two fill the bathtub   |    Stop E$$o (ExxonMobile)!
 with brightly-colored machine tools.»   |        http://www.stopesso.com