- Wine-GitLab - list.winehq.org

[PATCH v9 0/4] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Joerg Rueppel (＠jrueppel) May 29, 2026

May 29, 2026

Space Empires V interprets data returned from (IDirectXFileData_GetData()-4) as refcount, reads and writes to it and makes allocation decisions based on it. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=49722 -- v9: Apply 1 suggestion(s) to 1 file(s) https://gitlab.winehq.org/wine/wine/-/merge_requests/10278

2 4

[PATCH v7 0/2] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Joerg Rueppel (＠jrueppel) May 29, 2026

May 29, 2026

Space Empires V interprets data returned from (IDirectXFileData_GetData()-4) as refcount, reads and writes to it and makes allocation decisions based on it. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=49722 -- v7: Apply 1 suggestion(s) to 1 file(s) https://gitlab.winehq.org/wine/wine/-/merge_requests/10278

2 2

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

I left a bunch of small comments but, in general, now this seems pretty good to me! -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141675

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

I don't think we want the `trace()`s upstream. Adding a similar `ok()` for the child data seems useful though. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141674

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

```suggestion:-0+0 buf->cur_pos_data += sizeof(DWORD); ``` -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141673

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

```suggestion:-0+0 if (!check_buffer(buf, sizeof(DWORD))) ``` -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141672

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

I'd use something like `0x55555555` as filler, it's more conspicuous should it pop up in exceptions or register values. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141671

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

I would put it in a more generic way (there might well be more games depending on this implementation detail even without Delphi involvement) before explaining the particular example this is fixing. Maybe something like: ```suggestion:-3+0 /* Applications have been known to depend on the presence of padding before * the object data. * Specifically, Space Empires V reinterprets the returned object as a * Delphi dynamic array and updates the value at (DWORD *)ptr - 1 for * refcounting and allocation decisions. Must be >= 2. */ ``` -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141670

1 0

Re: [PATCH v6 0/1] MR10278: d3dxof: Pad with magic value before object data in parse_object().
by Matteo Bruni (＠Mystral) May 29, 2026

May 29, 2026

Matteo Bruni (@Mystral) commented about dlls/d3dxof/tests/d3dxof.c: > "3;;;, 0;;;, 1;;;, 2;;;,\n" > "3;;;, 1;;;, 2;;;, 3;;;,\n" > "3;;;, 3;;;, 1;;;, 2;;;,\n" > +/* child object for heap alloc test */ This comment now needs to be updated. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/10278#note_141669

1 0

[PATCH 0/1] MR11017: msi: Catch all exceptions in the custom DLL procedure.
by Dmitry Timoshkov (＠dmitry) May 29, 2026

May 29, 2026

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/11017

3 8