- Wine-GitLab - list.winehq.org

[PATCH 0/1] MR9725: dwrite: Do not modify file path when \??\unix path is used.
by Nikolay Sivov (＠nsivov) Dec. 8, 2025

Dec. 8, 2025

It's possible to get system font files paths in unix-path format, if there is no drive configured that allows access to corresponding trees. When this happens registry font list starts using unix paths, and those can't be modified the same way regular win32 paths could be. This fixes a regression caused by 2a2cc34db71. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=58811 Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9725

2 1

[PATCH v3 0/1] MR9671: include: Add bti jc for __ASM_DEFINE_FUNC on ARM64
by André Zwing Dec. 8, 2025

Dec. 8, 2025

Branch Target Identification (BTI) was introduced with ARMv8.5 and is enabled by newer distributions. It's a NOP for older hardware, so make our code compatible to it. See also: https://github.com/AndreRH/hangover/issues/192 Note: I wasn't able to test this on Hardware myself! -- v3: ntdll/unix: Add bti c for call dispatchers https://gitlab.winehq.org/wine/wine/-/merge_requests/9671

3 5

[PATCH 0/1] MR9724: winex11: Retrieve the whole window when creating client window for other process.
by Rémi Bernon Dec. 8, 2025

Dec. 8, 2025

As soon as x11drv window data has been created, X11DRV_get_whole_window will return it instead of looking up the whole window from the window property. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=59091 Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=59094 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9724

1 1

[PATCH 0/1] MR9723: opengl32: Return the buffers when invalid draw/read buffer are used.
by Rémi Bernon Dec. 8, 2025

Dec. 8, 2025

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=59100 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9723

1 1

[PATCH v11 0/5] MR9632: sxs: avoid uninitialised len_name length
by Hans Chen (＠hans.chen) Dec. 8, 2025

Dec. 8, 2025

The `redirect_data->clrdata_len` [is unset in /dlls/ntdll/actctx.c#L4411-4412](https://gitlab.winehq.org/wine/wine/-/blob/master/dlls/ntdll/actctx.c?ref_type=heads#L4411-4412) when using registration-free COM in C# NET code. Program.cs ```C# //... Guid guid = Guid.Parse("COM clsid"); Type type = Type.GetTypeFromCLSID(guid, true); var instance = Activator.CreateInstance(type)); //... ``` App.manifest ```xml <assembly> <file name="ExampleCOM.dll"> <comClass description="ExampleCOM" progid="ExampleCOM" clsid="{COM clsid}" threadingModel="Free"/> </file> </assembly> ``` The above C# code gets this error on its second line. ```Shell System.AccessViolationException: Attempted to read or write protected memory. This is often an indication that other memory is corrupt. ``` This is because `len_name` is uninitialized when calling `memcpy(ret_strings, ptr_name, len_name);` in `SxsLookupClrGuid`, which is fixed in this MR. -- v11: rollback changes in tests https://gitlab.winehq.org/wine/wine/-/merge_requests/9632

2 5

[PATCH 0/1] MR9686: winhttp: Implement WINHTTP_AUTOPROXY_ALLOW_STATIC
by Tobias Gruetzmacher (＠TobiX) Dec. 8, 2025

Dec. 8, 2025

This seems to be a convenience feature to return the static/default proxy config when that is configured. This is a first stab at the "correct" implementation, passes the test and seems right to me, but might be crap. Found in [PyManager](https://github.com/python/pymanager/blob/fa2de8c8820f0b2521cc4e3…, which at least fails at a different place with this change. :D -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9686

3 4

[PATCH 0/4] MR9721: win32u, winex11: Use the D3D present rect for fullscreen exclusive windows.
by Rémi Bernon Dec. 8, 2025

Dec. 8, 2025

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=58844 Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=58443 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9721

1 4

[PATCH 0/2] MR9711: msado15: Remove left over free in get_accessor (ASan).
by Bernhard Übelacker Dec. 8, 2025

Dec. 8, 2025

The first patch removes an unconditional `free( key )`, which holds here a pointer to the stack variable `tmp`. I assume this is a left over, because the next line outside the block is again a `free( key )`, this time with condition. The second patch adds a return in case the `malloc` of `elem` failed. Are both of my assumptions right here? Followup to f2ea57b7b7. CC: @piotr [ASan report in gitlab](https://gitlab.winehq.org/bernhardu/wine/-/jobs/215621#L6107) <details> <summary>ASan details</summary> ``` (lldb) bt * thread #1, stop reason = breakpoint 1.4 * frame #0: 0x79e8b843 libclang_rt.asan_dynamic-i386.dll`ReportError at asan_report.cpp:227:29 [inlined] frame #1: 0x79e8b843 libclang_rt.asan_dynamic-i386.dll`::ReportFreeNotMalloced() at asan_report.cpp:269:13 frame #2: 0x79e68da4 libclang_rt.asan_dynamic-i386.dll`::Deallocate() at asan_allocator.cpp:0 frame #3: 0x79e66cb8 libclang_rt.asan_dynamic-i386.dll`__ZN6__asan9asan_freeEPvPN11__sanitizer18BufferedStackTraceE at asan_allocator.cpp:1017:12 frame #4: 0x79e8520c libclang_rt.asan_dynamic-i386.dll`::free() at asan_malloc_win.cpp:72:10 frame #5: 0x6e9ab979 msado15.dll`get_accessor(recordset=0x03403970, fields=0x0059f8f0, hacc=0x0059f940) at recordset.c:2280 frame #6: 0x6e9a11e9 msado15.dll`_recordset_AddNew(a)36(iface=<unavailable>, field_list=<unavailable>, values=<unavailable>) at recordset.c:2326 frame #7: 0x0041107a msado15_test.exe`_Recordset_AddNew(This=<unavailable>, field_list=<unavailable>, values=<unavailable>) at msado15_backcompat.h:5779 frame #8: 0x00405c01 msado15_test.exe`test_ADORecordsetConstruction(exact_scroll=<unavailable>) at msado15.c:2061 frame #9: 0x00401038 msado15_test.exe`func_msado15 at msado15.c:3221 frame #10: 0x0041a662 msado15_test.exe`run_test(name="msado15") at test.h:780 frame #11: 0x0041a17c msado15_test.exe`main(argc=<unavailable>, argv=<unavailable>) at test.h:0 frame #12: 0x0041be56 msado15_test.exe`mainCRTStartup at crt_main.c:62 frame #13: 0x7bd33690 kernel32.dll`BaseThreadInitThunk + 16 frame #14: 0x7be667ab ntdll.dll`_call_thread_func_wrapper + 23 frame #15: 0x7be6711b ntdll.dll`call_thread_func(entry=(msado15_test.exe`mainCRTStartup at crt_main.c:45), arg=0x7ffd1000) at signal_i386.c:503 (lldb) frame select 5 frame #5: 0x6e9ab979 msado15.dll`get_accessor(recordset=0x03403970, fields=0x0059f8f0, hacc=0x0059f940) at recordset.c:2280 2277 memcpy( elem->key.data, key->data, key->len * sizeof(key->data[0]) ); 2278 rb_put( &recordset->hacc_cache, key, &elem->entry ); 2279 } -> 2280 free( key ); 2281 } 2282 if ((BYTE *)key != tmp) free( key ); 2283 (lldb) print key (hacc_cache_key *) 0x0059f6c0 (lldb) print (char*)tmp (char *) 0x0059f6c0 "" (lldb) ================================================================= ==msado15_test.exe==1092==ERROR: AddressSanitizer: attempting free on address which was not malloc()-ed: 0x0059f6c0 in thread T0 #0 0x79e851bb in free /home/runner/work/llvm-mingw/llvm-mingw/llvm-project/compiler-rt\lib/asan/asan_malloc_win.cpp:71:3 #1 0x6e9ab978 in get_accessor .../wine/dlls/msado15/recordset.c:2280 #2 0x6e9a11e8 in recordset_AddNew .../wine/dlls/msado15/recordset.c:2326 #3 0x00411079 in _Recordset_AddNew .../obj\include\msado15_backcompat.h:5779 #4 0x00405c00 in test_ADORecordsetConstruction .../wine/dlls/msado15/tests/msado15.c:2061 #5 0x00401037 in func_msado15 .../wine/dlls/msado15/tests/msado15.c:3221 #6 0x0041a661 in run_test .../wine/include/wine/test.h:780 #7 0x0041a17b in main .../wine/include/wine/test.h #8 0x0041be55 in mainCRTStartup .../wine/dlls/msvcrt/crt_main.c:62 #9 0x7bd3368f in BaseThreadInitThunk (C:\windows\system32\kernel32.dll+0x7b82368f) #10 0x7be667aa in call_thread_func_wrapper (C:\windows\system32\ntdll.dll+0x7bc467aa) #11 0x7be6711a in call_thread_func .../wine/dlls/ntdll/signal_i386.c:503 Address 0x0059f6c0 is located in stack of thread T0 at offset 32 in frame #0 0x6e9aac2f in get_accessor .../wine/dlls/msado15/recordset.c:2179 This frame has 2 object(s): [16, 20) 'refcount' (line 2182) [32, 288) 'tmp' (line 2183) <== Memory access at offset 32 is inside this variable HINT: this may be a false positive if your program uses some custom stack unwind mechanism, swapcontext or vfork (longjmp, SEH and C++ exceptions *are* supported) SUMMARY: AddressSanitizer: bad-free .../wine/dlls/msado15/recordset.c:2280 in get_accessor ``` </details> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/9711

2 3

[PATCH 0/1] MR9720: msado15: Ignore IRowset:GetProperties return value in rsconstruction_put_Rowset (Coverity).
by Piotr Caban (＠piotr) Dec. 8, 2025

Dec. 8, 2025

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/9720

2 1

Re: [PATCH v3 0/1] MR8701: devenum: Add DevicePath field for a video input device.
by Haoyang Chen (＠chenhaoyang) Dec. 8, 2025

Dec. 8, 2025

On Mon Dec 8 01:49:57 2025 +0000, Bernhard Übelacker wrote: > I submitted MR !9707 which copies above skip when no device is found > from qcap to devenum, hope that is ok. Hi Bernhard Übelacker. Thank you very much for your feedback and the suggested fix! This test requires a webcam to function properly. I have reviewed your code [!9707](https://gitlab.winehq.org/wine/wine/-/merge_requests/9707) and indeed need to skip it. I really appreciate you taking the time to look into this and provide such a clear solution. Thanks again! -- https://gitlab.winehq.org/wine/wine/-/merge_requests/8701#note_125121

1 0