new valgrind results...

[Dan Kegel]

This is with an almost-unpatched trunk (the only patch
is my tweak to handle bogus environment variables),
so it's now using Alexandre's heap patches rather
than mine.
There are some interesting new warnings, not sure
quite what to make of them yet.

Logs at
http://kegel.com/wine/valgrind/logs/2010-01-29-20.52/
Stats at
http://kegel.com/wine/valgrind/logs/2010-01-29-20.52.errs.txt
      2 Source and destination overlap in memcpy
      2 Syscall param writev
      3 Use of uninitialised value of size 4
      7 Invalid read of size 2
     10 Invalid write of size 4
     20 Invalid write of size 1
     29 Invalid free/ delete / delete[]
     30 Invalid read of size 4
     46 Conditional jump or move depends on uninitialised value
     86 Invalid read of size 1
    279 XXX bytes in XXX blocks are definitely lost
total:  514