- Wine-GitLab - list.winehq.org

[PATCH 0/5] MR8225: PDB rewrite #15.
by eric pouech (＠epo) June 6, 2025

June 6, 2025

This MR does: - improve startup time when loading a large PDB file (a chromium based PDB file (larger than 4G) now loads in 3,5s) (at the begining of this rewrite, loading the very same PDB file ended with a reboot triggered by the Linux OOM after 30 minutes of heavy swap) - finally untie the new PDB backend for the old one (they are now considered as two independant debug formats, so that we can eventually drop the old one at some point) -- https://gitlab.winehq.org/wine/wine/-/merge_requests/8225

2 5

[PATCH 0/1] MR8224: hid/tests: Fix stack use-after-return in test_read_device.
by Yuxuan Shui (＠yshui) June 6, 2025

June 6, 2025

The read loop could exit while a async read request is still in flight, which could complete after test_read_device has returned. This results in a stack use-after-return of `overlapped`. * * * Finally had something caught by the unix side AddressSanitizer, PE ASan wouldn't have caught this. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/8224

3 2

[PATCH v2 0/2] MR7024: Disable native titlebar while lacking enough space for SM_CYCAPTION.
by Jactry Zeng (＠jactry) June 6, 2025

June 6, 2025

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) -- v2: winemac.drv: Disable native titlebar while lacking enough space for SM_CYCAPTION. winex11.drv: Disable native titlebar while lacking enough space for SM_CYCAPTION. https://gitlab.winehq.org/wine/wine/-/merge_requests/7024

2 2

[PATCH v2 0/1] MR8205: comctl32: Make sure line_widths is big enough.
by Yuxuan Shui (＠yshui) June 6, 2025

June 6, 2025

When button_count is 0, we would allocate 0 bytes for line_widths, but later line_count would be 1 so we would still try to access line_widths[0] which is out-of-bound. -- v2: comctl32: Make sure line_widths is big enough. https://gitlab.winehq.org/wine/wine/-/merge_requests/8205

4 6

[PATCH 0/5] MR8222: widl: Add rows for the struct type.
by Hans Leidekker (＠hans) June 6, 2025

June 6, 2025

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/8222

3 7

[PATCH v5 0/5] MR8206: windows.devices.bluetooth: Add stubs for (I)BluetoothDeviceStatics.
by Vibhav Pant (＠vibhavp) June 6, 2025

June 6, 2025

-- v5: windows.devices.bluetooth: Add stub for IBluetoothDeviceStatics. windows.devices.bluetooth/tests: Add tests for BluetoothDeviceStatics. include: Add definitions for Windows.Devices.Bluetooth.BluetoothDevice. include: Add windows.devices.bluetooth.rfcomm.idl. https://gitlab.winehq.org/wine/wine/-/merge_requests/8206

3 14

[PATCH 0/1] MR8221: urlmon: Skip fragment part when checking filenames for file protocol.
by Gabriel Ivăncescu June 6, 2025

June 6, 2025

I have no idea what the previous test was trying to do, but it was clearly wrong and misleading. First, it placed the fragment for no reason in file_name_buf because it then replaced it with a NUL, making you believe it's part of it but when it really was not. Then, it opened the original index.html to write XXX to it, so the fragment filename never actually existed in the first place, and was never tested. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/8221

2 2

[PATCH 0/2] MR8196: winewayland: Ignore text-input "done" events that don't modify state.
by Alexandros Frantzis (＠afrantzis) June 6, 2025

June 6, 2025

Compositors send a "done" event after every text-input commit, even if the reported state (preedit etc) hasn't changed. Acting on such events is at best wasteful, but can additionally lead to incorrect IME related effects (e.g., deleting the currently selected text), so ignore them, similarly to what Qt and GTK do. The first commit performs some state related cleanups. -- This solves an IME problem I have been experiencing with compositors that support text-input-v3 (e.g., kwin). 1. Run notepad 2. Write some text 3. Try to select text Expected result: text is selected Actual result: text is deleted What happens is that while the user is selecting text the driver gets a series of `SetIMECompositionRect` callback which cause a series of text-input-v3 `set_cursor_rectangle`-`commit` requests. These `commit`s elicit `done` events from the compositor with no state changes, which we interpret as "clear the composition string" (and by extension the active selection), because that's what such states normally means, and forward them to windows IME. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/8196

4 4

[PATCH v2 0/4] MR8220: comdlg32: Don't use a shadow devmode structure in PrintDlg.
by Jacob Czekalla (＠JacobCzekalla) June 6, 2025

June 6, 2025

If the "Properties" button is clicked in the printdlg it opens the print properties dlg. The problem is the values set in the properties dlg don't get passed through correctly to the final printing stage. This prevents properties like duplex and page size from working. -- v2: comdlg32: Don't use a shadow devmode structure in PrintDlgA. comdlg32/tests: Add tests for changing devmode properties in the hook procedure for PrintDlgA. comdlg32: Don't use a shadow devmode structure in PrintDlgW. comdlg32/tests: Add tests for changing devmode properties in the hook procedure for PrintDlgW. https://gitlab.winehq.org/wine/wine/-/merge_requests/8220

3 5

[PATCH v3 0/1] MR8145: ldap: Avoid code duplication between sasl_client_start() and sasl_client_step().
by Dmitry Timoshkov (＠dmitry) June 6, 2025

June 6, 2025

-- v3: ldap: Avoid code duplication between sasl_client_start() and sasl_client_step(). https://gitlab.winehq.org/wine/wine/-/merge_requests/8145

3 2