"new" statement and -fno-exceptions

[Paul Koning <pkoning at equallogic dot com>]

We recently ran into some problems with an application being killed by
unhandled exception processing -- even though it's compiled
-fno-exceptions. 

It turns out that a "new" statement will throw an exception on alloc
failure (if the "nothrow" variant isn't used) no matter whether
-fno-exceptions is used or not.  This is somewhat surprising; it would
make sense for the code generator to call the nothrow new operator in
this case.  The docs don't say one way or the other.

I suppose I can supply my own operator new(size_t) as a workaround...

     paul