[PATCH] user32/scroll: Fix tracking for non-client scrollbars

[Gabriel Ivăncescu]

Non-client scrollbars (SB_HORZ and SB_VERT) don't have their client origin at
(0,0) because they can have non-client borders. It is necessary to offset
by this origin just as it is done for LBUTTONDOWN.

Signed-off-by: Gabriel Ivăncescu <gabrielopcode(a)gmail.com>
---

This fixes, for example, SB_VERT scrollbars that have a non-client area
above them (like a title bar) so that when moving the mouse after clicking
(hot-tracking), they won't jump by the size of the title bar (which is
annoying).

 dlls/user32/scroll.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/dlls/user32/scroll.c b/dlls/user32/scroll.c
index 95d9761..d611cf0 100644
--- a/dlls/user32/scroll.c
+++ b/dlls/user32/scroll.c
@@ -1095,15 +1095,17 @@ static void SCROLL_HandleScrollEvent( HWND hwnd, INT nBar, UINT msg, POINT pt)
 void SCROLL_TrackScrollBar( HWND hwnd, INT scrollbar, POINT pt )
 {
     MSG msg;
+    RECT rect;
 
     if (scrollbar != SB_CTL)
     {
-        RECT rect;
         WIN_GetRectangles( hwnd, COORDS_CLIENT, &rect, NULL );
         ScreenToClient( hwnd, &pt );
         pt.x -= rect.left;
         pt.y -= rect.top;
     }
+    else
+        rect.left = rect.top = 0;
 
     SCROLL_HandleScrollEvent( hwnd, scrollbar, WM_LBUTTONDOWN, pt );
 
@@ -1115,8 +1117,8 @@ void SCROLL_TrackScrollBar( HWND hwnd, INT scrollbar, POINT pt )
             msg.message == WM_MOUSEMOVE ||
             (msg.message == WM_SYSTIMER && msg.wParam == SCROLL_TIMER))
         {
-            pt.x = (short)LOWORD(msg.lParam);
-            pt.y = (short)HIWORD(msg.lParam);
+            pt.x = (short)LOWORD(msg.lParam) - rect.left;
+            pt.y = (short)HIWORD(msg.lParam) - rect.top;
             SCROLL_HandleScrollEvent( hwnd, scrollbar, msg.message, pt );
         }
         else
-- 
2.19.1

[Marvin]

Hi,

While running your changed tests on Windows, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=45623

Your paranoid android.


=== debian9 (32 bit Chinese:China report) ===

user32:
menu.c:2354: Test failed: test 27
menu: Timeout

[Gabriel Ivăncescu]

On 12/11/18 14:16, Marvin wrote:
> Hi,
>
> While running your changed tests on Windows, I think I found new failures.
> Being a bot and all I'm not very good at pattern recognition, so I might be
> wrong, but could you please double-check?
>
> Full results can be found at:
> https://testbot.winehq.org/JobDetails.pl?Key=45623
>
> Your paranoid android.
>
> === debian9 (32 bit Chinese:China report) ===
>
> user32:
> menu.c:2354: Test failed: test 27
> menu: Timeout

user32 tests have too many false positives. Needless to say these are 
unrelated...