July 2024 - Wine-GitLab - list.winehq.org

[PATCH v3 0/6] MR6042: d3d: Miscellaneous tests/fixes.
by Connor McAdams (＠cmcadams) July 17, 2024

July 17, 2024

The first two patches here address an issue I encountered when using Wine's d3dx9 with DXVK in the game Command and Conquer 3. The game calls `D3DXLoadSurfaceFromMemory` on a multisampled render target. This ends up going through the `lock_surface()` helper in d3dx9, which eventually calls `UpdateSurface()` in `unlock_surface()` to copy the staging surface back to the multisampled render target. This succeeds on wined3d even though it shouldn't, and fails on DXVK like it should. The second two patches address an issue reported in Empires: Dawn of the Modern World, where the game expects a device reset to reset the scene state. -- v3: wined3d: Clear scene state on device state reset. d3d8/tests: Add a test for device reset after beginning a scene. d3d9/tests: Add a test for device reset after beginning a scene. https://gitlab.winehq.org/wine/wine/-/merge_requests/6042

4 10

[PATCH 0/2] MR6098: odbc32 fixes transact
by Alistair Leslie-Hughes (＠alesliehughes) July 17, 2024

July 17, 2024

Wine doesn't handle Pooling as yet and in my scenario it's set to 0 (No Pooling), so not an issue but requires a successful return value. SQLTransact incorrectly assumed that if either handle was NULL an error occurred. I'm seeing a call where the Environment handle is NULL, even unixODBC handles this scenario. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6098

3 3

[PATCH v2 0/7] MR6036: win32u: Use the thread input and desktop shared memory for reading keyboard state.
by Rémi Bernon July 17, 2024

July 17, 2024

-- v2: win32u: Use the thread input shared memory for GetKeyState. server: Expose the thread input keystate lock through shared memory. server: Make the get_key_state request key code mandatory. win32u: Use the desktop shared memory in get_async_keyboard_state. win32u: Use the thread input shared memory in GetKeyboardState. win32u: Introduce a new NtUserGetAsyncKeyboardState call. server: Expose the thread input keystate through shared memory. https://gitlab.winehq.org/wine/wine/-/merge_requests/6036

2 10

[PATCH 0/1] MR6105: winex11: Reset empty window shape even without a window surface.
by Rémi Bernon July 17, 2024

July 17, 2024

And invalidate the window surface shape if any, so that it gets updated again in update_surface_region if the window is actually shaped. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=56972 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6105

1 1

[PATCH 0/1] MR6104: light.msstyles: Use slightly darker color for GrayText to make text more readable.
by Dmitry Timoshkov (＠dmitry) July 17, 2024

July 17, 2024

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/6104

4 3

[PATCH v2 0/2] MR6026: winemac: Use the new surface shape helper to update window shape.
by Rémi Bernon July 17, 2024

July 17, 2024

-- v2: winemac: Use window surface shape for window shape region. winemac: Use window surface shape for color key transparency. https://gitlab.winehq.org/wine/wine/-/merge_requests/6026

1 3

[PATCH v2 0/2] MR6084: opengl32: Serialize debug callbacks message string.
by Rémi Bernon July 17, 2024

July 17, 2024

-- v2: opengl32: Serialize debug callbacks message string. opengl32: Use integer types in debug callbacks parameter structs. https://gitlab.winehq.org/wine/wine/-/merge_requests/6084

1 3

[PATCH v2 0/3] MR6083: winevulkan: Serialize debug callbacks parameter structures.
by Rémi Bernon July 17, 2024

July 17, 2024

-- v2: winevulkan: Serialize debug callbacks parameter structures. winevulkan: Use integer types in debug callbacks parameter structs. winevulkan: Fix incorrect 32->64 conversion of debug callbacks. https://gitlab.winehq.org/wine/wine/-/merge_requests/6083

1 4

[PATCH 0/1] MR6102: odbc32: Handle NULL attribute in SQLColAttributeW.
by Daniel Lehman (＠dlehman25) July 17, 2024

July 17, 2024

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/6102

2 1

Re: [PATCH v2 0/1] MR5979: Draft: mfreadwrite: Use GPU copy when available.
by Brendan McGrath (＠redmcg) July 17, 2024

July 17, 2024

Marking this as draft for now, as I need to revisit it at some point. If removing the copy is possible, that would be even better than optimizing it. In which case the final solution may look very different to the one here (resulting in the closure of this MR). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/5979#note_76372

1 0