- Wine-GitLab - list.winehq.org

Re: [PATCH v2 0/1] MR10387: vbscript: Invoke default property when calling object variable with empty parens. - approved
by Alexandre Julliard (＠julliard) March 24, 2026

March 24, 2026

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10387

1 0

Re: [PATCH v8 0/1] MR10396: vbscript: Implement IActiveScriptError::GetSourceLineText. - approved
by Alexandre Julliard (＠julliard) March 24, 2026

March 24, 2026

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10396

1 0

Re: [PATCH v2 0/1] MR10433: ntdll: Do not keep next handler pointer outside of lock in call_vectored_handlers(). - approved
by Alexandre Julliard (＠julliard) March 24, 2026

March 24, 2026

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10433

1 0

[PATCH v2 0/4] MR8688: server: Start to expose Unix symlinks as NT reparse points.
by Elizabeth Figura (＠zfigura) March 24, 2026

March 24, 2026

A Wine user has requested the ability to interact with Unix symlinks from Win32 code. While the generic functionality of NT reparse points is still not implemented upstream, it will probably involve a completely different code path than this, so there does not seem to me to be any point in waiting for that to be done before submitting these patches. This does not implement all the relevant functionality. FILE_OPEN_REPARSE_POINT is now respected, but returns an fd with no fd->unix_fd. Thus far only file disposition (i.e. deletion), renaming and hardlinking are implemented, and all other operations return the no_fd_status, for which STATUS_REPARSE_POINT_NOT_RESOLVED is arbitrarily chosen. This is probably going to be controversial. As controversial patches tend to get deferred forever, I'm going to at least try to pose some questions related to concept and implementation which may be easier to answer. * Do we want to expose Unix symlinks at all? This was requested by a Wine user, and it does seem a strong argument that Wine, in general, supports some measure of integration with the host, including exposing custom interfaces for programs written to take advantage of it. We also, currently, already expose Unix *directory* symlinks via NtQueryAttributesFile(). We don't currently expose such file symlinks (and this series does not change that particular point, but it would be done by follow-up patches). In a sense we are expanding existing partial support. On the other hand, Unix symlinks which are completely opaque can be useful, and some downstream distributions (CrossOver and Proton) currently rely on them to reduce disk space usage while keeping the application unaware of a symlink (even if it were to use FILE_OPEN_REPARSE_POINT. Not that any applications are currently known that would break if this were to change, and in any case they could simply revert these patches.) * Implementation-wise, is having unix_fd == -1 a non-starter? This can be true of other real fds in some specific cases... * It's also true that this current implementation suffers from some degree of TOCTOU problems: because we only store the name and not the FD anymore, the actual underlying object at that path can change. This would not be true on Windows. Is this a problem? * One alternate approach would be to use O_PATH (Linux, FreeBSD) and O_SYMLINK (Mac). I don't see any equivalent flag for NetBSD or Solaris, so we would probably just not be able to expose this functionality there. Note that while the documentation seems to imply that O_PATH only stores the path, manual testing shows that (at least on Linux) a fd opened with O_PATH will remain pointing at the same *inode* even if it is moved or deleted. So this would avoid the aforementioned TOCTOU problem. -- v2: server: Allow renaming and hardlinking Unix symlinks. server: Allow setting a disposition on Unix symlinks. server: Allow opening Unix symlinks with FILE_OPEN_REPARSE_POINT. server: Separate a open_unix_fd() helper. https://gitlab.winehq.org/wine/wine/-/merge_requests/8688

3 7

Re: [PATCH v2 0/1] MR10415: mshtml: Doc typo fix - approved
by Alexandre Julliard (＠julliard) March 24, 2026

March 24, 2026

This merge request was approved by Alexandre Julliard. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10415

1 0

[PATCH 0/4] MR10421: winegstreamer: Improve quartz MPEG-1 video seeking.
by Akihiro Sagawa (＠sgwaki) March 24, 2026

March 24, 2026

This MR improves quartz video seeking particularly for MPEG-1 video. The quartz decoder now flushes the wg_transform buffers on seek. Furthermore, the GStreamer plugin, avdec_mpeg2video, now receives additional information equivalent to an MPEG-1 sequence header, allowing seeking even when such headers are not regularly present. If you're looking for a test video, I recommend the Kirby video available at [1]. It contains only a single sequence header, which makes it suitable for testing. [1] https://www.nintendo.co.jp/ngc/gkyj/movies/htp.mpg -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10421

4 8

[PATCH v14 0/1] MR10339: quartz/tests: Add tests for IMediaSeeking time formats in avisplit.
by José Luis Jiménez López March 24, 2026

March 24, 2026

Added comprehensive tests for IMediaSeeking_ConvertTimeFormat. Replaced magic numbers with descriptive constants. Verified on Windows 10. Signed-off-by: José Luis Jiménez López josejl1987(a)gmail.com -- v14: quartz/tests/avisplit: Add tests for IMediaSeeking time formats. https://gitlab.winehq.org/wine/wine/-/merge_requests/10339

3 4

[PATCH v19 0/2] MR10295: Fix ImageGetDigestStream
by Trent Waddington (＠trent.waddington) March 24, 2026

March 24, 2026

ImageGetDigestStream is used to hash the "important" parts of PE files. After rigorous experimentation we discovered these changes were sufficient to get identical hashes (and thereby signatures) on Windows and Wine. -- v19: imagehlp: Fix ImageGetDigestStream Add test case for image with rsrc/reloc sections https://gitlab.winehq.org/wine/wine/-/merge_requests/10295

2 4

[PATCH 0/4] MR10417: winebth.sys: Handle BlueZ crashes and restarts, only make calls to BlueZ if it is available.
by Vibhav Pant (＠vibhavp) March 24, 2026

March 24, 2026

* Only start initial object enumeration if GetNameOwner succeeds for `org.bluez` (i.e, BlueZ is up). * Listen for the `NameOwnerChanged` signal on the system bus, this happens when: * BlueZ shuts down/crashes: In this case, we emit a `BLUETOOTH_WATCHER_EVENT_TYPE_SERVICE_DOWN`, which removes all added radios/devices/services from the PE driver. * BlueZ starts, here we perform an initial enumeration of objects, and register the authentication agent with BlueZ. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10417

3 5

[PATCH v4 0/5] MR10208: winebth.sys part 15: Add support for reading/caching GATT characteristic values
by Vibhav Pant (＠vibhavp) March 24, 2026

March 24, 2026

Characteristic values represent data that can be read from or written to. The `winebluetooth_gatt_characteristic_value` struct gets used to send the buffer from Unix to PE land, and has the following structure: ```c #define WINEBLUETOOTH_CHARACTERISTIC_INLINE_SIZE 256 struct winebluetooth_gatt_characteristic_value { UINT32 size; union { BYTE buf[WINEBLUETOOTH_CHARACTERISTIC_INLINE_SIZE]; UINT_PTR handle; }; }; ``` ~~For values \<= 256 bytes, the value gets stored inline in `buf`, and copying is trivial. Otherwise, `handle` stores the DBus message containing the value and the `bluetooth_gatt_characteristic_value_move` unixlib function performs a copy and frees the associated DBus message.~~ -- v4: bluetoothapis/tests: Add tests for BluetoothGATTGetCharacteristicValue. bluetoothapis: Implement BluetoothGATTGetCharacteristicValue. winebth.sys: Implement IOCTL_WINEBTH_GATT_SERVICE_READ_CHARACTERISITIC_VALUE. winebth.sys: Update the cached GATT characteristic value on receiving a PropertiesChanged signal for a GATT characteristic value from BlueZ. winebth.sys: Store GATT characteristic value cached by the Unix Bluetooth service on the PE driver. https://gitlab.winehq.org/wine/wine/-/merge_requests/10208

3 6

[PATCH 0/1] MR6138: crypt32: When importing certificates to Root store set 'done' flag before entering semaphore.
by Dmitry Timoshkov (＠dmitry) March 24, 2026

March 24, 2026

CSP is loaded for each imported certificate, and if CSP accesses Root store it deadlocks. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6138

5 11

Re: [PATCH v4 0/1] MR10404: Draft: vbscript: Reject arguments on Class_Initialize and Class_Terminate.
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

waiting for !10386 to set this one ready -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10404#note_133580

1 0

Re: [PATCH v5 0/1] MR10405: Draft: vbscript: Return proper error for undefined variables with Option Explicit.
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

waiting for !10386 to set this one ready -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10405#note_133578

1 0

[PATCH v3 0/1] MR10416: oleaut32: Handle reference tables when processing SLTG records
by Wolfgang Hartl (＠wolfo.dev) March 24, 2026

March 24, 2026

https://bugs.winehq.org/show_bug.cgi?id=59549 When processing SLTG record structures, references to other structures were not handled correctly. That causes an error "err:typelib:sltg_get_typelib_ref Unable to find reference". I added the handling of references to the SLTG_ProcessRecord() function. as they were already in other functions available. Testing: 1) With MS Money 2000: before the bug fix several error were logged, with the fix applied the errors were gone. 2) I created a small IDL with a record type and a second nested record type that referenced the first record, compiled it to an SLTG-tlb using widl. Then I dumped the tlb before and after the fix with a small C program. The output before the fix showed the error message and a corrupted structure in the dump. After the fix, the error was not shown anymore and the nested record was fully shown. Please see details with logs in the bug. -- v3: oleaut32: Handle reference tables when processing SLTG records https://gitlab.winehq.org/wine/wine/-/merge_requests/10416

2 1

[PATCH v3 0/5] MR10405: Draft: vbscript: Return proper error for undefined variables with Option Explicit.
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

When Option Explicit is set and an undefined variable is used, return VBS error 500 (Variable is undefined) with the variable name in the description, matching Windows behavior. Previously Wine returned E_FAIL with no description. -- v3: vbscript: Return proper error for undefined variables with Option Explicit. vbscript: Improve some error codes in the parser. vbscript: Improve some error codes in the lexer. vbscript: Improve some error codes in the compiler. vbscript/tests: Add error code tests for compile errors. https://gitlab.winehq.org/wine/wine/-/merge_requests/10405

2 5

[PATCH v12 0/4] MR10386: vbscript: Return proper error codes for compile errors.
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

Add error_code field to parse error tests, using negative values for codes that Wine does not yet return correctly (todo_wine), matching the existing pattern for error_char. Return MAKE_VBSERROR with proper error codes instead of E_FAIL for name redefinition, invalid exit statements, and invalid parenthesized call statements. -- v12: vbscript: Improve some error codes in the parser. vbscript: Improve some error codes in the lexer. vbscript: Improve some error codes in the compiler. vbscript/tests: Add error code tests for compile errors. https://gitlab.winehq.org/wine/wine/-/merge_requests/10386

2 5

[PATCH v4 0/1] MR10362: vbscript: Support chained call syntax like dict.Keys()(i).
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

Treat '(' after tEMPTYBRACKETS as an argument list opener in the lexer, so that expressions like GetArr()(0) or dict.Keys()(i) parse correctly. Also dereference VT_BYREF|VT_VARIANT results in vcall so chained indexing on call results works at runtime. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=58051 This includes some of the work done in !10308, preferably that one is merged first. -- v4: vbscript: Support chained call syntax like dict.Keys()(i). https://gitlab.winehq.org/wine/wine/-/merge_requests/10362

2 3

[PATCH v3 0/1] MR10384: vbscript: Allow hex literals with excess leading zeros.
by Francis De Brabandere (＠francisdb) March 24, 2026

March 24, 2026

Windows allows any number of leading zeros in hex literals (e.g. &H000000031 = 49). Wine rejected literals with more than 8 hex digits regardless of value. Fix by skipping leading zeros before counting significant digits, so that only actual 32-bit overflow is rejected. -- v3: vbscript: Allow hex literals with excess leading zeros. https://gitlab.winehq.org/wine/wine/-/merge_requests/10384

2 2

[PATCH 0/1] MR10437: ntoskrnl.exe/tests: Fix stack use-after-free of OVERLAPPED.
by Yuxuan Shui (＠yshui) March 24, 2026

March 24, 2026

The ioctls started in `test_cancel_thread` complete only after the function has already returned, thus the OVERLAPPED objects needed have become invalid. Make them global variables. This should be fine since we only start the next test thread after we have completed the previously started IO operations. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10437

3 3

[PATCH v3 0/2] MR10438: configure: fail early on invalid Gradle/SDK
by Twaik Yont (＠twaik) March 24, 2026

March 24, 2026

This change is motivated by the long-standing issue that most Linux distributions ship an outdated Gradle, which is often too old for modern Android builds. In typical Android projects this is solved by using the Gradle wrapper, which ensures a compatible Gradle version is downloaded and used automatically. However, Wine cannot rely on the Gradle wrapper for this purpose, and integrating automatic downloads into the build system is not acceptable, since the build process must not fetch external resources. As a result, Gradle and the Android SDK effectively become external prerequisites that must be provided by the user or by higher-level build scripts. Previously, missing or outdated dependencies would only surface later during the build, often resulting in confusing or hard-to-diagnose errors. This change makes these requirements explicit and enforces them at configure time: - `gradle` is detected, its version is checked, and the resolved path is cached. - `ANDROID_HOME` is required and validated against the expected SDK layout. - A minimal sanity check ensures that a usable `build-tools` installation is present. This allows the build to fail early with clear error messages if the environment is not properly set up, instead of failing deep inside the Gradle build. Additionally, `GRADLE` and `ANDROID_HOME` are cached by configure, so they do not need to be re-specified when invoking make or even present in environment. -- v3: wineandroid: print hint for full APK build when using dummy target https://gitlab.winehq.org/wine/wine/-/merge_requests/10438

2 3

Re: [PATCH v2 0/1] MR10352: cabinet: Add stubs for compressor/decompressor
by Chris Kadar (＠nspitko) March 24, 2026

March 24, 2026

On Tue Mar 24 11:40:24 2026 +0000, Nikolay Sivov wrote: > Could you check which algorithm this game is using? We have at least > some MSZIP code scattered around, and maybe this could be consolidated > at some point (not in this MR obviously). It's using algorithm 4, which appears to be "XPRESS compression algorithm with Huffman encoding". -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10352#note_133550

1 0

[PATCH 0/1] MR10438: configure: fail early on invalid Gradle/SDK
by Twaik Yont (＠twaik) March 24, 2026

March 24, 2026

This change is motivated by the long-standing issue that most Linux distributions ship an outdated Gradle, which is often too old for modern Android builds. In typical Android projects this is solved by using the Gradle wrapper, which ensures a compatible Gradle version is downloaded and used automatically. However, Wine cannot rely on the Gradle wrapper for this purpose, and integrating automatic downloads into the build system is not acceptable, since the build process must not fetch external resources. As a result, Gradle and the Android SDK effectively become external prerequisites that must be provided by the user or by higher-level build scripts. Previously, missing or outdated dependencies would only surface later during the build, often resulting in confusing or hard-to-diagnose errors. This change makes these requirements explicit and enforces them at configure time: - `gradle` is detected, its version is checked, and the resolved path is cached. - `ANDROID_HOME` is required and validated against the expected SDK layout. - A minimal sanity check ensures that a usable `build-tools` installation is present. This allows the build to fail early with clear error messages if the environment is not properly set up, instead of failing deep inside the Gradle build. Additionally, `GRADLE` and `ANDROID_HOME` are cached by configure, so they do not need to be re-specified when invoking make or even present in environment. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10438

2 1

[PATCH v2 0/7] MR10009: mf/topology_loader: Refactor the topology loader and fix a few test failures.
by Conor McCarthy (＠cmccarthy) March 24, 2026

March 24, 2026

This changes the functionality relatively little, but the refactoring will allow greater changes in a followup MR which will fix more tests. -- v2: mf/topology_loader: Load the branch up connection method only for source nodes. mf/topology_loader: Use the branch connection context for down nodes. mf/topology_loader: Set the branch upstream node media type if not a source node. mf/topology_loader: Introduce a branch connection context. mf/topology_loader: Delete unused struct connect_context. mf/topology_loader: Do not create a new branch list containing downstream nodes. mf: Trace additional topology details. https://gitlab.winehq.org/wine/wine/-/merge_requests/10009

4 12

[PATCH 0/6] MR10364: dinput: Add support for generating consistent guidInstance values for devices.
by Connor McAdams (＠cmcadams) March 24, 2026

March 24, 2026

This MR adds support for creating `guidInstance` values for joystick devices in a way that matches native behavior. Our current method of creating a `guidInstance` value uses the device's rawinput handle, which changes each time a controller is plugged in. Fixes a bug in Rocket League where a controller plug/unplug cycle results in a new controller being detected each time. The process is like this: - Controller is connected, the game gets its `guidInstance` value from `EnumDevices()`, stores it and associates it with player 1. - Controller is disconnected. - The same controller is reconnected, the game calls `EnumDevices()`, and finds a new `guidInstance` value. - It then assumes this is a new controller, creates a new player, and behaves as though player 1's controller is unplugged. This MR implements `guidInstance` in a way that matches native, and ensures that the `guidInstance` value remains consistent. Doing things this way also makes it easier to add support for `DIPROP_JOYSTICKID` and `GUID_Joystick` in a future MR. I've implemented those in a branch [here](https://gitlab.winehq.org/cmcadams/wine/-/tree/WIP/dinput-instance-gu… using this system. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10364

3 15

Re: [PATCH v2 0/14] MR10349: opengl32: Implement wglGetProcAddress on the client side.
by Rémi Bernon March 24, 2026

March 24, 2026

> Specs for most of them are available, so we know they don't add new commands. They are just not a part of the official registry. `GL_ATI_texture_compression_3dc` has no spec. `GL_APPLE_packed_pixels` had a unofficial one but it's now difficult to find, all the tokens it defines are also already defined in OpenGL 1.2, so I don't think exposing it is going to make any difference. The same reasoning as above wrt platform extensions applies for it as well. > It is part of the extension strings exposed by Mesa. In any case, we could filter it if there is an actual benefit, but I do not see why we would need to. What exactly looks wrong to you? GLSL extensions aren't supposed to be exposed there, unless they have a matching GL extension to add tokens / commands or change GL behavior. This one doesn't do any of that on the GL side, and I believe it is not meant to be exposed in GL_EXTENSIONS. This leaves `GL_EXT_texture_edge_clamp` / `GL_EXT_texture_rectangle` / `GL_MESA_texture_signed_rgba` with unofficial specs and which we can expose, yes. And then there are more spec-less extensions that are exposed by native drivers. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10349#note_133537

1 0