[PATCH v2 2/2] advapi32: Don't read the key name if the root key handle is invalid

[Alexandre Julliard]

Alex Henrie <alexhenrie24 at gmail.com> writes:

> On Wed, Nov 7, 2018 at 11:57 PM Alexandre Julliard <julliard at winehq.org> wrote:
>>
>> Alex Henrie <alexhenrie24 at gmail.com> writes:
>>
>> > On Wed, Nov 7, 2018 at 9:11 AM Alexandre Julliard <julliard at winehq.org> wrote:
>> >>
>> >> Alex Henrie <alexhenrie24 at gmail.com> writes:
>> >>
>> >> > Signed-off-by: Alex Henrie <alexhenrie24 at gmail.com>
>> >> > ---
>> >> >  dlls/advapi32/registry.c       | 2 ++
>> >> >  dlls/advapi32/tests/registry.c | 6 ++++++
>> >> >  2 files changed, 8 insertions(+)
>> >> >
>> >> > diff --git a/dlls/advapi32/registry.c b/dlls/advapi32/registry.c
>> >> > index 0a8ccfbf20..4b33ade88c 100644
>> >> > --- a/dlls/advapi32/registry.c
>> >> > +++ b/dlls/advapi32/registry.c
>> >> > @@ -313,6 +313,8 @@ static inline HKEY get_special_root_hkey( HKEY hkey, REGSAM access )
>> >> >  {
>> >> >      HKEY ret = hkey;
>> >> >
>> >> > +    if (hkey == INVALID_HANDLE_VALUE) return 0;
>> >>
>> >> There shouldn't be any reason to check for INVALID_HANDLE_VALUE, it
>> >> doesn't have a special meaning. I expect you'll get the same behavior
>> >> with any other invalid handle.
>> >
>> > Is there a straightforward way to check whether the handle is valid?
>>
>> You shouldn't need to add explicit checks, it should fall off from the
>> normal function behavior. That's probably true for your other patch too.
>
> Leaving aside the second patch for the moment, the tests show that
> RegCreateKeyEx[AW] checks the retkey parameter before checking the
> hkey parameter. If you don't want retkey to be explicitly checked at
> the beginning of RegCreateKeyEx[AW], how do you want the check to be
> done?

Is there an app that depends on this? Adding a lot of error checking
only to satisfy artificial test cases is usually not desirable.

-- 
Alexandre Julliard
julliard at winehq.org