redzone1024 test failures b/c one leak is reported as a possible leak, non-deterministically

_From [bruen...@google.com](https://code.google.com/u/109494838902877177630/) on May 21, 2012 14:17:58_

I'm seeing a mismatch on leaks in the redzone1024 test but I can't repro outside of the suite. This is on xp64 (the pattern failures are issue #892 ):

drmemory-dbg-32: 42 tests passed, ***\* 4 tests failed: ****
redzone1024 => stderr failed to match ~~Dr.M~~ 3 unique, 3 total, 155 byte(s) of
leak(s), found ~~Dr.M~~ 2 unique, 2 total, 96 byte(s) of leak(s)
instead
free.pattern
registers-pattern
malloc.pattern
drmemory-rel-32: 42 tests passed, ***\* 4 tests failed: ****
redzone1024 => stderr failed to match ~~Dr.M~~ 3 unique, 3 total, 155 byte(s) of
leak(s), found ~~Dr.M~~ 2 unique, 2 total, 96 byte(s) of leak(s)
instead

Looks like the 42-byte leak is reported as a possible leak instead. Maybe garbage in a redzone happens to point at it?!?

Error `#8`: POSSIBLE LEAK 42 direct bytes 0x00fd0000-0x00fd002a + 17 indirect bytes
#0 main (0x0040144f )
e:\derek\drmemory\git\src\tests\malloc.c:219

Error `#9`: LEAK 16 direct bytes 0x00fd28a8-0x00fd28b8 + 48 indirect bytes
#0 main (0x004014b5 )
e:\derek\drmemory\git\src\tests\malloc.c:251

Error `#10`: POSSIBLE LEAK 16 direct bytes 0x00fd5120-0x00fd5130 + 0 indirect bytes
#0 main (0x004014f6 )
e:\derek\drmemory\git\src\tests\malloc.c:256

Error `#11`: LEAK 16 direct bytes 0x00fd5938-0x00fd5948 + 16 indirect bytes
#0 main (0x00401503 )
e:\derek\drmemory\git\src\tests\malloc.c:257

_Original issue: http://code.google.com/p/drmemory/issues/detail?id=895_