https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68197
--- Comment #4 from Christophe Monat <christophe.monat at st dot com> --- (In reply to Jonathan Wakely from comment #3) > No, it seems underspecified. I have raised it with the C++ committee. Do you have feedback from the C++ committee ? I have only easily access to a 2011 standard, and reading 27.5.3.5 (3) : Returns : '(snip)... on failure, a valid long& initialized to 0'. I may be wrong, but I read this as an obligation to return 0 in the case described by Mickael.
