http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46821
--- Comment #4 from Jonathan Wakely 2010-12-06
14:04:58 UTC ---
I think Marc's right, tuple::swap is specified to call swap for each element,
and that is definitely something users can detect (even if there's rarely a
good reason to do so for emp
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46821
Paolo Carlini changed:
What|Removed |Added
CC||jwakely.gcc at gmail dot
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46821
--- Comment #2 from Marc Glisse 2010-12-06
13:50:45 UTC ---
Calling swap exactly like in the non-empty case? It seems expected to me that
swap be called for each element, and in the usual empty case the compiler
should still be able to optimize i
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46821
--- Comment #1 from Paolo Carlini 2010-12-06
13:25:19 UTC ---
In practice, which semantics would you suggest?
