- wine-gitlab - winehq.org

Re: [PATCH v2 0/7] MR6911: ntdll: Add WaitCompletionPacket.
by Zhiyi Zhang (＠zhiyi) 16 Dec '24

16 Dec '24

> NtAssociateWaitCompletionPacket() initiates an asynchronous wait on the object handle. When the wait is done (the object is signaled), it queues a completion to the associated I/O completion port. Make sense. I will revisit this MR after the code freeze. > Maybe, in the particular case of React Native, it only uses WaitCompletionPacket on manual reset events or other objects with no-op satisfied()? FWIW I couldn't find any reference to NtAssociateWaitCompletionPacket in https://github.com/facebook/react-native. I am testing a React Native sample application. If I remember correctly, the NtAssociateWaitCompletionPacket() is from some Microsoft DLL, coremessagingxp.dll. I didn't look into the type of object it waits on. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6911#note_90563

1 0

[PATCH v6 0/4] MR6904: ntdll.dll: Update NtQueryDirectoryFile to align with current Windows behaviour
by Eugene McArdle (＠eugenemcardle) 16 Dec '24

16 Dec '24

This PR updates the behaviour of `NtQueryDirectoryFile`, bringing it in line with current Windows behaviour. The need for this update was discovered when attempting to build the Unreal Engine with MSVC under Wine. In certain cases conditional include statements do not behave as expected, due to MSVC depending on undocumented behaviour of `NtQueryDirectoryFile`. We ran tests on multiple versions of Windows, and discovered that the behaviour has changed since the original Wine implementation, but the documentation has not. The source code for our test tool, and a set of results can be found [here](https://github.com/TensorWorks/NtQueryDirectoryFile-Test). As of Windows 8, calling `NtQueryDirectoryFile` with a re-used handle, a new mask, and setting the `RestartScan` flag to True, causes the cached results to be erased and a new scan to be performed with the updated mask. Currently, Wine performs as did earlier versions of Windows, where the changed mask is ignored, and the cache is reused. This can cause `NtQueryDirectoryFile` under Wine to falsely report that files exist, when they do not. This PR corrects this behaviour, invalidating the cache when required. Implementing this exposed further undocumented behaviour of `NtQueryDirectoryFile`, where a search for a non-existent file will return either `STATUS_NO_MORE_FILES` or `STATUS_NO_SUCH_FILE`, depending on whether or not the handle had been previously used regardless of the value of `RestartScan`. This was reflected in a `winetest` which allowed for the response to be either `STATUS_SUCCESS` or `STATUS_NO_MORE_FILES`. This test has been updated to only allow current Windows behaviour, and `NtQueryDirectoryFile` will return either `STATUS_NO_MORE_FILES` or `STATUS_NO_SUCH_FILE` as appropriate. This patch also adds unit tests for the new behaviour of `NtQueryDirectoryFile`. These tests pass when running `winetest` under Windows, and under Wine with these changes in place, but they will fail under older versions of Wine. -- v6: ntdll: Simplify check that unexpected files were not found https://gitlab.winehq.org/wine/wine/-/merge_requests/6904

4 8

Re: [PATCH v2 0/2] MR6990: win32u, server: Fix some window surface shape sometimes not being updated.
by darkblaze69 (＠darkblaze69) 15 Dec '24

15 Dec '24

Unfortunately this (ce7b293c5b29111d256f5416bb44a6c77d6ef673) broke some dialogs: https://bugs.winehq.org/show_bug.cgi?id=57551 -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6990#note_90561

1 0

[PATCH v2 0/1] MR7023: kernelbase: Workaround for SetCurrentDirectoryW misuse in Vanguard: Saga of Heroes
by Lois Gomez (＠loisgomez) 15 Dec '24

15 Dec '24

Context: Vanguard Saga of Heroes is an MMORPG that was closed years ago by SOE. However the game is still alive in the emulator scene at https://vgoemulator.net/ The game seems to have relied on an old behaviour of the function SetCurrentDirectoryW( LPCWSTR dir ) which currently crashes the game. This patch was a solution discovered by the community at https://vgoemulator.net/phpBB3/viewtopic.php?t=5563 It was previously reported to wine bugzilla in this bug report. https://bugs.winehq.org/show_bug.cgi?id=34285 Summary of changes: Updates the function SetCurrentDirectoryW( LPCWSTR dir ) to remove any trailing "." Previously the game will set the current directory as "bin\." and subsequently crash when accessing "bin\.Caches". It will now access "bin\Caches" and run successfully Notes: This is my first contribution to the project so I might be making a lot of incorrect assumptions on how this might affect other binaries. Please do let me know if there is a better way to implement a change like this in the codebase. Thanks for your help -- v2: kernelbase: Workaound for SetCurrentDirectoryW misuse https://gitlab.winehq.org/wine/wine/-/merge_requests/7023

4 4

[PATCH 0/2] MR7024: Disable native titlebar while lacking enough space for SM_CYCAPTION.
by Jactry Zeng (＠jactry) 15 Dec '24

15 Dec '24

On Windows, applications can disable titlebar by extending the client rect to a bigger rect which will cover the area of the titlebar from WM_NCCALCSIZE: https://learn.microsoft.com/en-us/windows/win32/dwm/customframe#removing-th… Our current x11 and Mac driver doesn't handle this case. Attached is a test case for reproducing the issue. [hide_caption.c](/uploads/46a725e32569a6a0a929a0f183cf9ed8/hide_caption.c) -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7024

3 3

Re: [PATCH v2 0/1] MR7017: ntdll: Set the correct character map name for Shift JIS
by Akihiro Sagawa (＠sgwaki) 15 Dec '24

15 Dec '24

From my point of view, SJIS is correct. Here is an example of SJIS locale usage written by [Red Hat](https://access.redhat.com/solutions/727043) and [HP-UX](https://web.archive.org/web/20061206132454/http://docs.hp.com/en/UXL… (see Appendix A). Historically, until around 2000, the Japanese UNIX environment mainly used an encoding different from Shift-JIS, called EUC-JP (Extended Unix Code for Japanese). Therefore, with the exception of HP-UX, SJIS has rarely been used in locale settings. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7017#note_90546

1 0

[PATCH 0/1] MR7020: include: Add methods arguments annotations for DirectWrite types.
by Nikolay Sivov (＠nsivov) 14 Dec '24

14 Dec '24

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=57476 Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com> -- https://gitlab.winehq.org/wine/wine/-/merge_requests/7020

3 2

[PATCH 0/1] MR7019: gdiplus: GdipPathAddRectangle should close the path.
by Esme Povirk (＠madewokherd) 14 Dec '24

14 Dec '24

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/7019

2 1

Re: [PATCH v2 0/7] MR6911: ntdll: Add WaitCompletionPacket.
by Jinoh Kang (＠iamahuman) 14 Dec '24

14 Dec '24

On Sat Dec 14 13:05:29 2024 +0000, Jinoh Kang wrote: > > The problem is that there might be no threads that initiate any wait. > You're probably talking about GetQueuedCompletionStatus(). Sure, no > threads might be waiting on IOCP. By "initiate the wait" I mean the wait > on the (arbitrary) target object handle, not the IOCP. > NtAssociateWaitCompletionPacket() initiates an *asynchronous* wait on > the object handle. When the wait is done (the object is signaled), it > queues a completion to the associated I/O completion port. Note that this issue is currently masked by another omission in this MR: WaitCompletionPacket doesn't call obj->ops->satisfied(). Waiting on mutexes, semaphores, and auto-reset events have the side effect of acquiring or resetting their state. Your MR, in its current state, does _not_ let this side effect take place, rendering them useless. Maybe, in the particular case of React Native, it only uses WaitCompletionPacket on manual reset events or other objects with no-op satisfied()? FWIW I couldn't find any reference to NtAssociateWaitCompletionPacket in https://github.com/facebook/react-native. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6911#note_90537

1 0

Re: [PATCH v2 0/7] MR6911: ntdll: Add WaitCompletionPacket.
by Jinoh Kang (＠iamahuman) 14 Dec '24

14 Dec '24

> The problem is that there might be no threads that initiate any wait. You're probably thinking about synchronous wait. By "initiate the wait" I mean the thread that called NtAssociateWaitCompletionPacket(). Not all waits are synchronous (like WaitForSingleObject). NtAssociateWaitCompletionPacket() initiates an *asynchronous* wait. When the wait is done (the object is signaled), it queues a completion to the associated I/O completion port. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/6911#note_90531

1 0

2025

2024

2023

2022

wine-gitlab