- Wine-GitLab - list.winehq.org

[PATCH v3 0/1] MR10670: kernelbase: Add messsage string for ERROR_NOT_A_REPARSE_POINT
by Antoine Leresche (＠A2va) April 16, 2026

April 16, 2026

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=59449 -- v3: kernelbase: Add messsage string for ERROR_NOT_A_REPARSE_POINT https://gitlab.winehq.org/wine/wine/-/merge_requests/10670

2 1

[PATCH 0/5] MR10681: opengl32: Misc make_opengl cleanups.
by Rémi Bernon (＠rbernon) April 16, 2026

April 16, 2026

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/10681

3 6

[PATCH v2 0/3] MR10637: vbscript: Return specific errors for Do keyword, End Class, and property arg count
by Francis De Brabandere (＠francisdb) April 16, 2026

April 16, 2026

Return proper error codes instead of generic 1002 (syntax error): - 1028: "Expected 'While', 'Until' or end of statement" for `Do For`, `Do If`, etc. - 1047: "Expected 'Class'" for mismatched `End` keyword inside class body - 1051: "Number of arguments must be consistent across properties specification" for Property Get/Let/Set arg count mismatch -- v2: vbscript: Return error 1051 for inconsistent property argument counts. vbscript: Return error 1047 for wrong 'End' keyword inside class. vbscript: Return error 1028 for invalid keyword after 'Do'. https://gitlab.winehq.org/wine/wine/-/merge_requests/10637

3 4

[PATCH v2 0/1] MR10635: vbscript: Return proper error for wrong number of arguments.
by Francis De Brabandere (＠francisdb) April 16, 2026

April 16, 2026

Return VBSE_FUNC_ARITY_MISMATCH (450) instead of E_FAIL when a function or sub is called with the wrong number of arguments. -- v2: vbscript: Return proper error for wrong number of arguments. https://gitlab.winehq.org/wine/wine/-/merge_requests/10635

3 2

[PATCH v6 0/1] MR10592: vbscript: Add call depth limit to prevent stack overflow on infinite recursion.
by Francis De Brabandere (＠francisdb) April 16, 2026

April 16, 2026

Track call depth in script_ctx_t and return error 28 "Out of stack space" when it exceeds 572, matching Windows behavior. Previously Wine would crash with a native stack overflow. -- v6: vbscript: Add call depth limit to prevent stack overflow on infinite recursion. https://gitlab.winehq.org/wine/wine/-/merge_requests/10592

3 2

[PATCH v7 0/2] MR10546: vbscript: Use indexed lookup for global functions and variables.
by Francis De Brabandere (＠francisdb) April 16, 2026

April 16, 2026

Replace O(N) linear scans through `global_funcs[]` and `global_vars[]` with O(log N) tree lookups for global function and variable resolution. **Global functions:** Scripts that use `ExecuteGlobal` to dynamically generate functions (e.g. Visual Pinball's GLF framework generates ~1000 functions at startup) hit pathological O(N) scan behavior on every `GetRef` call and `GetDispID` lookup. With this change, `GetRef` dispatch over 1000 functions drops from 503ms to 58ms (**8.7x**), reaching parity with Windows (62ms). | Benchmark (1000 dynamic functions) | Before | After | Speedup | Windows | |-----------|--------|-------|---------|---------| | GetRef varied x100K | 503 ms | 58 ms | **8.7x** | 62 ms | | GetRef loop x200K | 292 ms | 82 ms | **3.6x** | 125 ms | | GetRef cached x200K | 35 ms | 35 ms | 1.0x | 85 ms | | Direct call x200K | 35 ms | 46 ms | ~1.0x | 31 ms | **Global variables:** Profiling a real-world 75K-line Visual Pinball table script with 3000+ global variables (registered via multiple `ExecuteGlobal` calls) shows `lookup_identifier`'s linear scan over `global_vars[]` consuming ~55% of total CPU time. Compile-time binding (`bind_local`) does not help here because it only resolves variables declared within the same compilation unit, variables from prior `ExecuteGlobal` calls remain unbound at runtime. Depends on !10544 for the fast `vbs_wcsicmp` used in the tree comparators. -- v7: vbscript: Use indexed lookup for global variables. vbscript: Use indexed lookup for global functions. https://gitlab.winehq.org/wine/wine/-/merge_requests/10546

4 4

[PATCH v4 0/2] MR10603: win32u/freetype: Fix tmExternalLeading for CFF fonts.
by Arie Miller (＠Arimil) April 16, 2026

April 16, 2026

This fixes font scaling issues with CFF fonts this was discovered from games utilizing DxLib. https://github.com/yumetodo/DxLib/tree/master https://dxlib.xsrv.jp/index.html I've ran tests on Windows and to confirm that this change is consistent with a number of fonts. Before: ![image.png](/uploads/a398d5543df8561f1a6fbe9edded3b45/image.png){width="801" height="600"} After (this matches the windows visual from testing in a VM): ![image.png](/uploads/cf4e4c8753b3482dd24211b519a5fe23/image.png){width="767" height="572"} Related Proton Issue: https://github.com/ValveSoftware/Proton/issues/7299 Full Disclosure: An LLM was used to assist in finding this bug, the original fix written by the LLM wrote a loader binary that changed the pointers for `GetTextMetricsA` and `GetTextMetricsW` to point to a custom implementation that always set `tmExternalLeading`. The LLM determined that `GDI returns 0 when (winAscent + winDescent) > (Ascender - Descender)`, after testing on WIndows I found this to be incorrect, the actual difference is that CFF fonts always set `tmExternalLeading` to 0. The test case for this fix was LLM generated but reviewed and tested on both Windows and Linux by me. Test case: ``` WINEPREFIX=/tmp/wine-font winetricks allfonts WINEPREFIX=/tmp/wine-font ./loader/wine ./dlls/gdi32/tests/i386-windows/gdi32_test.exe ``` Patched: ``` font.c:4266: test_CFF_external_leading: tested 51 CFF fonts (0 failures) ``` Unpatched Wine: ``` font.c:4266: test_CFF_external_leading: tested 51 CFF fonts (51 failures) ``` Windows 11: ``` font.c:4266: test_CFF_external_leading: tested 48 CFF fonts (0 failures) ``` -- v4: gdi32: Add test for tmExternalLeading for CFF fonts. win32u/freetype: Fix tmExternalLeading for CFF fonts. https://gitlab.winehq.org/wine/wine/-/merge_requests/10603

3 6

[PATCH 0/4] MR10683: wineandroid: startup and JNI fixes (part of !10569)
by Twaik Yont (＠twaik) April 16, 2026

April 16, 2026

This MR contains a subset of changes extracted from MR !10569 (first 4 commits after merge-base). Included changes: 1. Fix 64-bit crash in JNI setCursor call * Pass NULL instead of integer 0 for jobject argument in CallVoidMethod * Prevents invalid jobject usage and JNI abort on 64-bit platforms 2. Lower targetSdkVersion to 28 * Avoid Android 10 W^X restrictions (API 29+) affecting exec and memory mappings * Add Java 1.8 compile options required by the adjusted SDK level 3. Fix Wine loader path and library lookup * Update WINELOADER to match current Wine layout * Extend LD_LIBRARY_PATH to support loading libraries directly from APK * Restore correct startup on modern Android/Wine setups 4. Move environment setup to Java layer and simplify wine_init * Replace JNI environment passing with setenv() in WineActivity * Simplify native initialization and remove environment array handling * Preserve LD_LIBRARY_PATH update and WINEDEBUGLOG behavior Reason: This MR is split out from !10569 to simplify review by isolating Android startup and stability changes into a smaller, focused set. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10683

3 5

[PATCH v2 0/1] MR10633: kernelbase: Ignore GENERIC_WRITE in CreateFile on CDs on Windows 9x.
by Alex Henrie (＠alexhenrie) April 16, 2026

April 16, 2026

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=59636 -- v2: kernelbase: Ignore GENERIC_WRITE in CreateFile on CDs on Windows 9x. https://gitlab.winehq.org/wine/wine/-/merge_requests/10633

3 2

[PATCH v2 0/8] MR10645: mf/topology_loader: Insert optional nodes after each branch is connected.
by Conor McCarthy (＠cmccarthy) April 16, 2026

April 16, 2026

New tests show each optional node is connected after the nodes on either side are connected. This is an issue in the game Psycho Sleuth (at least in the demo) because it uses a transform which must have its input types enumerated before it will work correctly. This occurs when a decoder is inserted before the transform, but an optional node lies immediately upstream of the transform, and if optional nodes are not handled correctly, the decoder is initially connected to the optional instead of being inserted later. A couple of other issues revealed while making new tests are fixed here too. -- v2: mf/topology_loader: Try to insert optional nodes after resolving the surrounding branch. mf/topology_loader: Introduce a helper to get the connection semantic flags. https://gitlab.winehq.org/wine/wine/-/merge_requests/10645

3 15