OK, here's something that seems to mostly work,
and even gives reasonable valgrind errors.
To use it, apply the patch and then do
export WINE_HEAP_REDZONE=16
or so. (Bigger values catch more problems but add more overhead.)
Only eleven tests generate warnings with that set:
runtest -q -P wine
On Wed, Nov 18, 2009 at 2:28 PM, Dan Kegel wrote:
> Here's a second try, this time only for large arenas, but
> with valgrind integration working. (It also has a workaround for
> a strange valgrind problem that leaves an empty variable in
> the environment that confuses wine.)
Found buglets alre
On Wed, Nov 18, 2009 at 12:47 PM, Eric Pouech wrote:
> interesting stuff
> however, you don't handle correctly HeapReAlloc(..., ZERO_MEMORY,...) calls
> when the already allocated block grows in place
Thanks for catching that.
I decided to start over and push the tail fill down into the heap
rout
Dan Kegel a écrit :
(Without a change like this, Valgrind can't properly
detect buffer overruns, but even without Valgrind,
this change will improve Wine's ability to detect
buffer overruns.)
With this change, users can enable buffer overrun
detection in Wine by setting WINE_GLOBAL_FLAGS=0x10.
T
(Without a change like this, Valgrind can't properly
detect buffer overruns, but even without Valgrind,
this change will improve Wine's ability to detect
buffer overruns.)
With this change, users can enable buffer overrun
detection in Wine by setting WINE_GLOBAL_FLAGS=0x10.
This is like the overru