Re: [PATCH] user32/tests: Fixed SetParent test on 64bit Vista

[Dmitry Timoshkov]

Austin Lund <austin.lund(a)gmail.com> wrote:
> --- a/dlls/user32/tests/msg.c
> +++ b/dlls/user32/tests/msg.c
> @@ -12681,6 +12681,7 @@ static const struct message WmSetParentSeq_2[] = {
>      { WM_WINDOWPOSCHANGING, sent|wparam, SWP_HIDEWINDOW|SWP_NOSIZE|SWP_NOMOVE },
>      { EVENT_OBJECT_HIDE, winevent_hook|wparam|lparam, 0, 0 },
>      { WM_WINDOWPOSCHANGED, sent|wparam, SWP_HIDEWINDOW|SWP_NOSIZE|SWP_NOMOVE|SWP_NOCLIENTSIZE|SWP_NOCLIENTMOVE },
> +    { HCBT_SETFOCUS, hook|optional },
>      { WM_NCACTIVATE, sent|wparam, 0 },
>      { WM_ACTIVATE, sent|wparam, 0 },
>      { WM_ACTIVATEAPP, sent|wparam, 0 },

This does look correct. It's pretty strange that other Windows versions
don't send HCBT_SETFOCUS notification before WM_KILLFOCUS (which is right
after WM_ACTIVATEAPP in the message sequence).

-- 
Dmitry.

[PATCH] user32/tests: Fixed SetParent test on 64bit Vista

[Austin Lund]

On 1 April 2011 14:03, Dmitry Timoshkov <dmitry(a)codeweavers.com> wrote:
> Austin Lund <austin.lund(a)gmail.com> wrote:
> > --- a/dlls/user32/tests/msg.c
> > +++ b/dlls/user32/tests/msg.c
> > @@ -12681,6 +12681,7 @@ static const struct message WmSetParentSeq_2[] = {
> >      { WM_WINDOWPOSCHANGING, sent|wparam, SWP_HIDEWINDOW|SWP_NOSIZE|SWP_NOMOVE },
> >      { EVENT_OBJECT_HIDE, winevent_hook|wparam|lparam, 0, 0 },
> >      { WM_WINDOWPOSCHANGED, sent|wparam, SWP_HIDEWINDOW|SWP_NOSIZE|SWP_NOMOVE|SWP_NOCLIENTSIZE|SWP_NOCLIENTMOVE },
> > +    { HCBT_SETFOCUS, hook|optional },
> >      { WM_NCACTIVATE, sent|wparam, 0 },
> >      { WM_ACTIVATE, sent|wparam, 0 },
> >      { WM_ACTIVATEAPP, sent|wparam, 0 },
>
> This does look correct. It's pretty strange that other Windows versions
> don't send HCBT_SETFOCUS notification before WM_KILLFOCUS (which is right
> after WM_ACTIVATEAPP in the message sequence).

Is there any obvious reason why this passes on the testbot when I sent
it myself, but the automated tests still fail?

http://test.winehq.org/data/123e4ad41d81bf4edc56eeab9f3cece9b511f690/vista_w...

The reported sequence seems to suggest that all the activate messages
are skipped. (?)

[PATCH] user32/tests: Fixed SetParent test on 64bit Vista

[Dmitry Timoshkov]

Austin Lund <austin.lund(a)gmail.com> wrote:
> Is there any obvious reason why this passes on the testbot when I sent
> it myself, but the automated tests still fail?
>
> http://test.winehq.org/data/123e4ad41d81bf4edc56eeab9f3cece9b511f690/vista_w...
>
> The reported sequence seems to suggest that all the activate messages
> are skipped. (?)

I should note that the message tests passed cleanly for me under Vista
64-bit on my laptop's dual boot even without your patch, so I don't know
what is going on on that system.

-- 
Dmitry.