April 2026 - Wine-GitLab - list.winehq.org

[PATCH v2 0/6] MR10550: winex11: Keyboard layouts rework, part 1
by Matteo Bruni (＠Mystral) April 16, 2026

April 16, 2026

The idea is to improve the detection of the available keyboard layouts, normally via Xkb group names + xkbregistry, and then propagate the current layout to win32u and the PE / win32 side state. Along the way also start using `KBDTABLES` and the related newer win32u machinery that was introduced for winewayland. I'm taking a lot of inspiration (and also actual patches, or pieces of code) from the winewayland driver and Rémi's draft MR !2122. The main difference from that attempt is that I'm going to keep around `X11DRV_KEYBOARD_DetectLayout()` and use it as a fallback detection mechanism (e.g. for vncserver), but still move winex11 to `KBDTABLES` and such. You can find the current (still WIP, especially for later patches) branch at https://gitlab.winehq.org/Mystral/wine/-/commits/hl-kbl?ref_type=heads This first MR does very little, just testing the waters... -- v2: winex11: Decode current layout from Xkb group name. winex11: Update an obsolete comment about the Xkb group index. winex11: Look up keysyms for the currently selected Xkb group in X11DRV_KEYBOARD_DetectLayout(). winex11: Factor out mapping keycodes to scancodes / keysyms into a separate function. winex11: Factor out Xkb initialization. winex11: Get Xkb group names. https://gitlab.winehq.org/wine/wine/-/merge_requests/10550

3 13

Re: [PATCH v2 0/4] MR10654: winegstreamer: Fix wg_parser/native video stride mismatch.
by Brendan McGrath (＠redmcg) April 16, 2026

April 16, 2026

> There really is no way to get pitch info from an `IMFMediaBuffer`, right? Am I missing something? You are right. When working with an `IMFMediaBuffer` that doesn't implement `IMF2DBuffer`, the stride is suppose to come from the `MF_MT_DEFAULT_STRIDE` attribute within the media type. Otherwise you can fallback to using `MFGetStrideForBitmapInfoHeader`. It's actually pretty well documented here: https://learn.microsoft.com/en-us/windows/win32/medfound/uncompressed-video… I also found an MS example using a YV12 image here: https://learn.microsoft.com/en-us/windows/win32/medfound/image-stride It seems a stride value here is solely for the `Y` plane and they just halve it for the `U` and `V` planes. Ideally we would use `IMF2DBuffer2::Lock2DSize` everywhere, as when using `IMFMediaBuffer`, we have to perform an expensive `memcpy` to and from the 2D buffer to a temporary 1D buffer. `Lock2DSize` also has `MF2DBuffer_LockFlags` which allows you to specify if your intent is to read and/or write from the buffer. This is useful, for example, if the buffer is backed by a DXGI surface, as it means we can skip a transfer to/from the GPU (depending on if you are just reading or just writing respectively). FWIW, commit d8a271c053bb23171355854d7af155b3bda7e8a9 was written to fix River City Girls on the DMO path (which goes via `wg_transform`). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10654#note_136446

1 0

[PATCH 0/3] MR10671: wintab32: Create the internal window on demand and make it message-only.
by Tim Clem (＠tclem) April 16, 2026

April 16, 2026

Split from !10627. If the window isn't message-only it comes back from EnumWindows, and since it responds 0 to WM_QUERYENDSESSION, it prevented macdrv from quitting via Cocoa. I don't see a reason why it can't be message-only. Also it seems risky to register a class and create a window from PROCESS_ATTACH; see !9822. Adding @bshanks since he looked at !10627. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10671

3 4

[PATCH 0/8] MR10679: msxml3: Reduce duplication in IXMLDOMCharacterData methods.
by Nikolay Sivov (＠nsivov) April 16, 2026

April 16, 2026

Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10679

2 8

[PATCH v3 0/1] MR10627: winemac.drv, wintab32: Some fixes for Cocoa-initiated quits
by Tim Clem (＠tclem) April 16, 2026

April 16, 2026

The EA app was refusing to exit when quit via Cocoa for two reasons: - the wintab32 window was responding 0 to WM_QUERYENDSESSION - the app itself closes a window in response to WM_QUERYENDSESSION, and we weren't expecting the corresponding WM_ENDSESSION to fail -- v3: winemac.drv: Handle windows becoming invalid between WM_QUERYENDSESSION and WM_ENDSESSION. https://gitlab.winehq.org/wine/wine/-/merge_requests/10627

3 3

[PATCH v16 0/8] MR10515: vbscript: Bind local variables, class properties, and loop variables at compile time.
by Francis De Brabandere (＠francisdb) April 15, 2026

April 15, 2026

Resolve identifiers to direct indices during compilation instead of doing runtime `wcsicmp` string scans in `lookup_identifier()`. This adds new opcodes that access local variables, function arguments, class properties, and loop variables by index, eliminating the linear string comparison overhead for the most common cases. **New opcodes:** - `OP_local` / `OP_assign_local` / `OP_set_local` - read/write local vars and args by index - `OP_local_prop` / `OP_assign_local_prop` / `OP_set_local_prop` - read/write class properties by index - `OP_step_local` / `OP_incc_local` - For-loop step/increment with indexed counter - `OP_enumnext_local` - For-Each iteration with indexed loop variable This mirrors jscript's `OP_local` / `bind_local()` optimization (`dlls/jscript/compile.c`) and addresses the FIXME in `lookup_identifier()` noting that class property access should be bound at compile time. Names that cannot be resolved at compile time (globals in Execute/ExecuteGlobal code, dynamic vars, host objects) continue to use the existing string-based path. The last commit also adds a fast path for SAFEARRAY iteration that bypasses the `IEnumVARIANT` COM vtable dispatch and copies elements directly from the array data, eliminating one intermediate `VariantCopy` per iteration. ### Local variable / argument / class property access (1M iterations, 10 reads per iteration) | Benchmark | Windows (VM) | Wine master | Wine (7 commits) | **Speedup** | vs Windows | |-----------|-------------|-------------|-------------------|---------|------------| | Local vars | 156 ms | 413 ms | 234 ms | **1.8x** | 1.5x slower | | Arguments | 171 ms | 357 ms | 231 ms | **1.5x** | 1.4x slower | | Class props | 265 ms | 393 ms | 226 ms | **1.7x** | 0.9x (faster!) | ### For-loop variants (10M iterations) | Benchmark | Windows (VM) | Wine master | Wine (7 commits) | **Speedup** | vs Windows | |-----------|-------------|-------------|-------------------|---------|------------| | Empty For | 117 ms | 370 ms | 270 ms | **1.4x** | 2.3x slower | | For + assign | 203 ms | 613 ms | 430 ms | **1.4x** | 2.1x slower | | Do-While manual | 437 ms | 646 ms | 513 ms | **1.3x** | 1.2x slower | | For Step 2 | 62 ms | 370 ms | 271 ms | **1.4x** | 4.4x slower | | Nested For | 125 ms | 387 ms | 268 ms | **1.4x** | 2.1x slower | | For local Sub | 109 ms | 294 ms | 251 ms | **1.2x** | 2.3x slower | | For R8 counter | 203 ms | 2,812 ms | 2,529 ms | **1.1x** | 12x slower | | For-Each array | 7 ms | 286 ms | 62 ms | **4.6x** | 8.9x slower | The remaining 4-13x gap vs Windows in For-loops is from `VarAdd`/`VarCmp` going through full `VariantChangeTypeEx` machinery (including locale alloc/free per call). A separate MR ([!10528](https://gitlab.winehq.org/wine/wine/-/merge_requests/10529)) adds I2/I4 fast paths in oleaut32 that bring integer loops within 1.2-1.9x of Windows when combined with this branch. -- v16: vbscript: Fast-path For-Each iteration over SAFEARRAY. vbscript: Bind For-Each loop variable at compile time. vbscript: Bind global-scope Dim variables at compile time. vbscript: Bind class properties at compile time. vbscript: Bind For-loop variables at compile time. vbscript: Bind local variable assignments at compile time. vbscript: Factor out assign_local_var from assign_ident. vbscript: Bind local variables and arguments at compile time. https://gitlab.winehq.org/wine/wine/-/merge_requests/10515

2 8

Re: [PATCH v2 0/4] MR10654: winegstreamer: Fix wg_parser/native video stride mismatch.
by Elizabeth Figura (＠zfigura) April 15, 2026

April 15, 2026

> Yeah I wanted to test with a 161 pixel width video, but h264 requires the width to be a multiple of 2. Maybe I should try a different codec or maybe raw video? I'm confused, wasn't the point here to fix raw video? > But OTOH how would UV sub-sampling work with an alignment of 1 byte? I feel it would be pretty safe to assume the minimum would be 2 bytes. IIRC it just doesn't actually subsample on the last column, but it's possible that native just doesn't handle odd width correctly or encoders refuse to create it. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10654#note_136419

1 0

[PATCH v8 0/7] MR10604: setupapi, ntoskrnl: Implement CM_Get_Parent with device tree properties
by Jan Robert Gerigk (＠RgSg86) April 15, 2026

April 15, 2026

Wine Bug: https://bugs.winehq.org/show_bug.cgi?id=59604 Implement CM_Get_Parent by storing device tree relationship properties when ntoskrnl enumerates child devices via IRP_MN_QUERY_DEVICE_RELATIONS: - DEVPKEY_Device_Parent on each child device - DEVPKEY_Device_Children on the parent device - DEVPKEY_Device_Siblings on each child device CM_Get_Parent reads DEVPKEY_Device_Parent through SetupDiGetDevicePropertyW. Also fix SETUPDI_EnumerateMatchingInterfaces to remove devices without interfaces from DIGCF_DEVICEINTERFACE enumeration results. --\> no more phantom devices Based on discussion with Rémi Bernon (Bug 59604) Tested on a fresh Wine 11.6 setup with KNX ETS 6.4.1: A USB HID license dongle (VID 2A07, PID 0102) is now correctly identified through CM_Get_Parent device tree navigation. -- v8: ntoskrnl/tests: Add DEVPKEY_Device_Parent test. setupapi/tests: Add CM_Get_Parent tests. cfgmgr32: Implement CM_Get_Parent using DEVPKEY_Device_Parent property. ntoskrnl: Store DEVPKEY_Device_Siblings on each child after bus enumeration. ntoskrnl: Store DEVPKEY_Device_Children on parent after bus enumeration. ntoskrnl: Store DEVPKEY_Device_Parent when enumerating child devices. https://gitlab.winehq.org/wine/wine/-/merge_requests/10604

3 9

[PATCH v5 0/1] MR10674: vbscript: Coerce VT_EMPTY operands before Var* calls.
by Francis De Brabandere (＠francisdb) April 15, 2026

April 15, 2026

Native VBScript treats Empty as VT_I4 in And/Or/Xor/Eqv/Imp/Not and as VT_I2 in Mod. Pre-coerce in the interp_* handlers so the result types match and VarMod's VT_EMPTY→VT_I4 shortcut is bypassed. props to @jsm174 for finding this issue: https://github.com/vpinball/libwinevbs/commit/7e877fafe34483623ca33d26ab76f… -- v5: vbscript: Coerce VT_EMPTY operands before Var* calls. https://gitlab.winehq.org/wine/wine/-/merge_requests/10674

2 1

Re: [PATCH v7 0/2] MR10463: winemac: Replace OSAtomic functions with builtin __atomic counterparts. - approved
by Alexandre Julliard (＠julliard) April 15, 2026

April 15, 2026

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10463

1 0