On 5/14/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> + memset( ldtent, 0, sizeof(*ldtent) ); /* Keep valgrind happy */
I'd never do anything like that (i.e. doing something where you have to
put a comment besides that it's there to please lint, valgrind or what
else code checkers ther
On Sat, May 13, 2006 at 01:36:54PM -0700, Dan Kegel wrote:
> There's another valgrind warning that is much more understandable, and
> I'm sending an obvious fix for it to wine-patches. (My first patch
> missed one instance, so I'm resending.) I'd appreciate it if you could
> check that one, too. -
Dan Kegel wrote:
On 5/13/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
I know this kind of errors from working on GNU Parted with Valgrind, and
my policy is to leave the code alone when I can't spot why it should be
wrong.
I strongly advise against doing otherwise.
I'll buy that. That
On 5/13/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
I know this kind of errors from working on GNU Parted with Valgrind, and
my policy is to leave the code alone when I can't spot why it should be
wrong.
I strongly advise against doing otherwise.
I'll buy that. That's why I sent the no
On Sat, May 13, 2006 at 01:01:34PM -0700, Dan Kegel wrote:
> What do folks think - should we just make the allocation a little
> bigger to make valgrind happy?
I know this kind of errors from working on GNU Parted with Valgrind, and
my policy is to leave the code alone when I can't spot why it sho
I was trying to use valgrind-3.1.1 to track down
an OpenOffice crash, and I noticed that current
wine shows errors like this even running notepad:
==8986== Invalid read of size 4
==8986==at 0x20010C21: (within /lib/ld-2.3.6.so)
...
==8986==by 0x20189D10: dlopen (in /lib/tls/i686/cmov/libd
