[NTUSER]
- Implement SPI_SETCURSORS case for SystemParametersInfoA/W. According to MSDN, it updates the system mouse cursors by the ones provided by user (can be set via main.cpl). It does not use any parameters from SystemParametersInfo:
1. First, get the cursor path from user defined values in registry via win32k!RegReadUserSetting.
2. Then load the cursor handle from the specified cursor via user32!LoadImageW called from the aprropriate win32k callback via KeUserModeCallback.
3. Set received handle for an appropriate resource ID via win32k!NtUserSetSystemCursor. Do this for each system defined cursor.
- NEW: Call an internal handler for SPI_SETCURSORS from win32k!SpiUpdatePerUserSystemParameters, to reload user-defined cursors from Registry after reboot. This is called from WinLogon at each startup.
- Implement co_IntLoadImage callback for user32!LoadImageW. Add an appropriate part in user32 also.
- Rewrite co_IntSetupDefaultCursors callback, responsible for default (system) cursors loading. Call a callback to user32!LoadImageW, to load each system cursor, and then use win32k!NtUserSetSystemCursor to set each of them as the current system cursor.
- Refactor some other several cursor/icon functions: NtUserFindExistingCursorIcon, NtUserSetSystemCursor and DefWndHandleSetCursor.
- Handle HTHELP case in win32k!GetNCHitEx and user32!CreateDialogIndirectA/W, which is responsible for help button class. Set an appropriate cursor for this case in DefWndHandleSetCursor (DefWindowProc!WM_SETCURSOR message).
- Remove bogus WM_SETCURSOR handing from win32k!DesktopWindowProc and user32!DesktopWndProcW, since it does not load a proper cursor at all, only default IDC_ARROW. It is already handled properly in win32k!IntDefWindowProc.
- Set correct GreSetPointerShape flags for animated mouse cursors.
- NEW: Add the system timer for animated mouse cursors where an each frame is enumerated separately and call it from win32k!UserSetCursor. This allows *.ani cursors to actually animate.
[USER32]
- Add/fix user mode parts of LoadImage and SetDefaultCursors callbacks. Don't try to load system cursor scheme, it should be done in main.cpl instead.
- Handle animated mouse cursors (*.ani). Load them correcly in user32!CURSORICON_LoadFromFileW. We already have CURSORICON_GetCursorDataFromANI, which handles it properly. Also set the correct flags for CURSORDATA structure and enable CURSORF_ACON flag on cursor creation in case cursor is animated.
- NEW: Load user-defined cursors from HKCU\Control Panel\Cursors Reigstry key. Call it from user32!ClientThreadSetup, to load a cursors at each startup.
- NEW: Add a small workaround to user32!LoadCursorW: try to find and load current cursor from Registry set by user first, in case it is set. Only in case it was not found, continue normal execution: load default system cursor, as the function should do. This allows to properly load the correct cursor for all UI elements.
Remaining bugs/issues:
- Animated cursors always have a bit wrong position compared to Windows. However it's absolutely correct for standart cursors (with a *.cur extension).
- Sometimes the animation becomes too fast (perhaps because of a recusrsive win32k!IntSetTimer calls, need it some another condition to kill the timer?).
- In case of changing *.cur -> *.ani, sometimes the animation is continuing infinitely on some UI elements (or in the window where the previous *.ani cursor was initially set), even after cursor is changed. Needs to restart an app/explorer/etc. to avoid the problem. However, this does not occur when changing *.cur -> *.cur, *.cur -> *.ani and *.ani -> *.ani. Again, it seems to require one more condition to kill the timer.
CORE-14165, CORE-14166
Based on the Doug Lyons' test in #7087, I found that my previous fix stopped working partially. Or rather, it would only work until the 32767 indexes were exhausted. It seems to me that the behavior of the bitfield has changed, because when I published the previous patch, it passed my tests.
- Bit array generates free ID cyclically, in the previous code after 32767 indexes expired the same index was returned, because of this the previous fix would stop working after expiration, so change the logic of calculating the next index.
- Change the index range to 256-32767 to match Windows, indexes 0-255 can theoretically be used as reserved for system purposes.
Addendum to fd327db20f. CORE-9141
The return value of RtlFindClearBitsAndSet is an ULONG, assigning it to an ULONG_PTR will not sign extend it. The error value will stay 0xFFFFFFFF. Comparing it to (UINT_PTR)-1 will sign extend and thus compare it to 0xFFFFFFFFFFFFFFFF on x64.
Also use NUM_WINDOW_LESS_TIMERS to initialize the bitmap, rather than the calculated size. This does not make a difference with the current value (32768), but if it was not the case, the bitmap would be larger than this, resulting in invalid bitmap indices being returned, which would cause bugs later on. Finally remove an ASSERT that can be triggered by tests.
* fix CORE-9141 - adding a change to IDEvent after each pass
If the first index is 0 the first returned ID will be 0x8000, which is fine.
Co-authored-by: Joachim Henze <joachim.henze@reactos.org>
- Change INIT_FUNCTION and INIT_SECTION to CODE_SEG("INIT") and DATA_SEG("INIT") respectively
- Remove INIT_FUNCTION from function prototypes
- Remove alloc_text pragma calls as they are not needed anymore
