__new_result definitely overflows, and this is *bad* (at least in principle ;)
when it's signed. I have a patch in testing which has small (probably bearable)
performance implications.
--
Summary: signed integer overflow in num_get<>::_M_extract_int
Product: gcc
Version: 4.0.0
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: libstdc++
AssignedTo: pcarlini at suse dot de
ReportedBy: pcarlini at suse dot de
CC: gcc-bugs at gcc dot gnu dot org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21209
