Re: [PATCH] New fixes for hot/cold partitioning

[Dale Johannesen <dalej at apple dot com>]

On Apr 4, 2005, at 10:48 AM, Caroline M Tice wrote:
> I made the change because I discovered I could not free it until after
> dwarf2out_finish was called, after compiling ALL the functions in the
> compilation unit.  Since I couldn't free it, it seemed best not to
> malloc it (especially malloc'ing multiple times and having a memory
> leak).

If it has to remain live until end of the compilation unit, then it
won't get re-malloced, will it?  Or if it will for some reason I'm 
missing,
xrealloc should avoid the leak (although that's kind of ugly).