April 2025 - Wine-GitLab - list.winehq.org

[PATCH 0/1] MR7865: windows.media.mediacontrol: Fix a memory leak.
by Mohamad Al-Jaf (＠maljaf) April 28, 2025

April 28, 2025

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/7865

3 2

[PATCH v23 0/6] MR7650: Add support for AF_UNIX sockets (followup)
by Ralf Habacker (＠rhabacker) April 28, 2025

April 28, 2025

Follow-up of !2786, which appears to have been abandoned. -- v23: ws2_32/tests: Add test for AF_UNIX sockets server: Fix getsockname() and accept() on AF_UNIX sockets. server: Introduce error when attempting to create a SOCK_DGRAM AF_UNIX socket. server: Allow for deletion of socket files. ws2_32: Add support for AF_UNIX sockets. ws2_32: Add afunix.h header. https://gitlab.winehq.org/wine/wine/-/merge_requests/7650

2 6

[PATCH v2 0/1] MR7857: winewayland: Don't update constraints without wl_pointer.
by Attila Fidan (＠atticf) April 28, 2025

April 28, 2025

If ClipCursor is called while the seat doesn't have a pointer and the call qualifies for locking the pointer, it would have tried to lock a null wl_pointer. For example, I'm launching a fullscreen application (which causes ClipCursor calls on startup) in a headless compositor before I launch my virtual keyboard/pointer client. While the pointer is deinitialized and there cannot be a focused hwnd, the cursor wl_surface will not be set, making it qualify for pointer locking. -- v2: winewayland: Require wl_pointer for pointer constraints. https://gitlab.winehq.org/wine/wine/-/merge_requests/7857

3 3

[PATCH 0/2] MR7924: windowscodecs/tiff: Remove unnecessary image data byte-swaping.
by Nikolay Sivov (＠nsivov) April 28, 2025

April 28, 2025

Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7924

4 4

[PATCH 0/1] MR7926: dsound/tests: Allocate right amount of memory in test_secondary8 (ASan).
by Alex Henrie (＠alexhenrie) April 28, 2025

April 28, 2025

The wave format is passed to DirectSoundDevice_CreateSoundBuffer, which duplicates it including the 1 byte of extra data after the end of the struct. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7926

2 1

[PATCH 0/1] MR7917: windowscodecs/png: Fix byte-swapping mode usage in the encoder.
by Nikolay Sivov (＠nsivov) April 28, 2025

April 28, 2025

The png_set_swap() explicitly checks for 16 bit depth, and bit_depth field is set by png_write_info(). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7917

3 2

[PATCH 0/5] MR7910: windowscodecs/converter: Add 24bppBGR -> 128bppRGBAFloat conversion path.
by Nikolay Sivov (＠nsivov) April 28, 2025

April 28, 2025

Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7910

3 6

[PATCH 0/1] MR7923: shell32: Change type of KNOWNFOLDERID FOLDERID_UserProgramFiles from...
by Louis Lenders (＠xe) April 27, 2025

April 27, 2025

shell32: Change type of KNOWNFOLDERID FOLDERID_UserProgramFiles from CSIDL_Type_Disallowed to CSIDL_Type_User. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=58169 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7923

2 1

[PATCH 0/1] MR7916: kernel32: Fixed MulDiv's inconsistencies with Windows.
by Shavit Yosef (＠shavitush) April 27, 2025

April 27, 2025

`MulDiv` doesn't match the correct Windows behavior for the function, which actually differs in some edge cases for 32-bit and 64-bit. Test case: ```c MulDiv(1, 0x80000000, 0x80000000); // WINE - both 32-bit and 64-bit: 0 // Windows (32-bit/kernel32.dll): 2 // Windows (64-bit/kernelbase.dll): -1 ``` On Windows 32-bit, this would return 2. On 64-bit, it'd return -1. However, on WINE, this call would return 0 for both 32-bit and 64-bit. This discrepancy was previously abused by threat actors to detect if the application runs under WINE. --- On Windows, `MulDiv` has a different implementation for 32-bit (from `kernel32.dll`) and 64-bit (leveraged to `kernelbase.dll` from `kernel32.dll`). After adjusting the 64-bit code to be an exact replica of the original code from `kernelbase.dll`, it'd still return the incorrect value due to Microsoft's code having undefined behavior with negation. It could be avoided with `-O0` but I opted into rewriting it with the undefined behavior to be, well.. defined. To clarify, the original code from Microsoft has a bug (logically, at least). But WINE is supposed to have matching behavior so we can run Windows programs as intended. The 32-bit version of `MulDiv` in Windows has a more complicated bug; I've documented it in the code. --- This is my first contribution to WINE - I cannot find any `CONTRIBUTING.md` or similar file with guidelines so I did my best to follow the convention of other merge requests. I've added unit testing for `MulDiv` to ensure there won't be a regression in the future. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7916

4 8

[PATCH v17 0/3] MR7542: ws2_32: Implement WSALookupService for Bluetooth device discovery.
by Vibhav Pant (＠vibhavp) April 27, 2025

April 27, 2025

This MR adds an initial implementation of the winsock `WSALookupsService*` methods for performing Bluetooth device discovery (`LUP_CONTAINERS`). -- v17: ws2_32: Implement WSALookupServiceNext for Bluetooth device discovery. ws2_32: Implement WSALookupServiceBegin for Bluetooth device discovery. https://gitlab.winehq.org/wine/wine/-/merge_requests/7542

2 3