Re: wined3d: Fix a memory leak in compile_state_table()

[Rico Schüller]

Allan Tong schrieb:
> Wrong pointer was being passed to HeapFree.
> - Allan

Why did you remove these two lines? Was this your intention?
-                    dev_multistate_funcs[cur[i].state][0] = 
multistate_funcs[cur[i].state][0];
-                    dev_multistate_funcs[cur[i].state][1] = 
multistate_funcs[cur[i].state][1];

Cheers
Rico

wined3d: Fix a memory leak in compile_state_table()

[Henri Verbeet]

2009/1/6 Rico Schüller :
> Allan Tong schrieb:
> > Wrong pointer was being passed to HeapFree.
> > - Allan
>
> Why did you remove these two lines? Was this your intention?
> -
>                dev_multistate_funcs[cur[i].state][0] =
>
> multistate_funcs[cur[i].state][0];
> -
>                dev_multistate_funcs[cur[i].state][1] =
>
> multistate_funcs[cur[i].state][1];

Because HeapReAlloc() should already take care of any copying, if
needed. Using HeapReAlloc() without checking the result first isn't
safe of course, but neither was the original code.

wined3d: Fix a memory leak in compile_state_table()

[Allan Tong]

On Tue, Jan 6, 2009 at 5:51 AM, Henri Verbeet  wrote:
> 2009/1/6 Rico Schüller :
> > Allan Tong schrieb:
> > > Wrong pointer was being passed to HeapFree.
> > > - Allan
> >
> > Why did you remove these two lines? Was this your intention?
> > -
> >                dev_multistate_funcs[cur[i].state][0] =
> >
> > multistate_funcs[cur[i].state][0];
> > -
> >                dev_multistate_funcs[cur[i].state][1] =
> >
> > multistate_funcs[cur[i].state][1];
>
> Because HeapReAlloc() should already take care of any copying, if
> needed. Using HeapReAlloc() without checking the result first isn't
> safe of course, but neither was the original code.

Oops, that was careless of me.  I'll send a revised patch.

- Allan