feat(options): Implement game options for texture filter mode, anisotropy and MSAA selection

[Mauller]

Merge By Rebase

Closes: #2474
Closes: #2375

This PR implements options for allowing the enablement of MSAA and the selection of different texture filtering modes and setting the anisotropic filtering level.

The following options are added for anti aliasing

AntiAliasing = 0, 2, 4, 8
The numbers relate to the MSAA sample level and Zero disables MSAA.

TextureFilter = None, Point, Bilinear, Trilinear, Anisotropic
Self explanatory, just choose one and see how the game looks

AnisotropyLevel = 2, 4, 8, 16
The anisotropy level only comes into play when Anisotropic filtering is set. 2 is also the minimum value.

---

Generals is broken untill the changes are replicated due to changes in texture filter init.

• Replicate in generals

[greptile-apps]

Greptile Summary

This PR implements three new game options — anti-aliasing (MSAA level), texture filter mode, and anisotropic filtering level — wiring them from Options.ini through OptionPreferences/GlobalData into the WW3D rendering pipeline. Previous review concerns about the <= vs > clamping bug in Set_Anisotropy_level() and the stray semicolon in OptionsMenu.cpp have been addressed.

Key changes:

• MultiSampleModeEnum values are now explicitly set to their sample counts (0, 2, 4, 8) so the INI stores human-readable numbers rather than combo-box positions.
• _Init_Filters() signature extended to accept AnisotropicFilterMode, removing the hardcoded 2X default.
• W3DDisplay::init() reads back the GPU-accepted MSAA level post device-creation, gracefully handling GPUs that don't support the requested sample count.
• AliasingMode enum moved from a local definition inside OptionsMenuInit into OptionPreferences for shared access.
• Two minor style issues remain: a no-op == guard in getAntiAliasing() that should use <= for consistency, and always-true val >= 0 conditions in the texture filter / anisotropy save blocks.

Confidence Score: 4/5

• Safe to merge after addressing two minor style nits; the previously flagged critical clamping bug has been correctly fixed.
• The core logic — enum value assignment, INI serialisation/deserialisation, GPU read-back after device creation, and anisotropy clamping — is all correct. Previous P0/P1 concerns from the thread are resolved. The two remaining comments are P2 style issues (a no-op == guard and always-true val >= 0 conditions) that do not affect runtime behaviour.
• Core/GameEngine/Source/Common/OptionPreferences.cpp (no-op guard), GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Menus/OptionsMenu.cpp (always-true conditions)

Important Files Changed

Filename	Overview
Core/GameEngine/Source/Common/OptionPreferences.cpp	Adds getAntiAliasing(), getTextureFilterMode(), getTextureAnisotropyLevel() — logic is sound except for an inconsistent == guard in getAntiAliasing() that should be <=.
Core/Libraries/Source/WWVegas/WW3D2/texturefilter.cpp	_Init_Filters now takes an explicit anisotropy_level argument and forwards it to _Set_Max_Anisotropy instead of hardcoding 2X. Correct.
GeneralsMD/Code/Libraries/Source/WWVegas/WW3D2/ww3d.cpp	Adds Set_Anisotropy_level() with correct ≤-based clamping and a new AnisotropyLevel static; Set_Texture_Filter() now passes the stored anisotropy level through. No issues.
GeneralsMD/Code/Libraries/Source/WWVegas/WW3D2/ww3d.h	MultiSampleModeEnum values are now explicitly assigned (0, 2, 4, 8) to match the INI values; new Set/Get_Anisotropy_level API added. Clean.
GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Menus/OptionsMenu.cpp	Anti-aliasing saving/loading now correctly converts between combo-box position and MSAA sample count. Two always-true val >= 0 guards in the texture filter / anisotropy blocks.
GeneralsMD/Code/GameEngineDevice/Source/W3DDevice/GameClient/W3DDisplay.cpp	Applies MSAA mode from global data before device creation, then reads back the GPU-accepted mode afterwards and applies texture filter + anisotropy level. Post-init read-back pattern is sound.

Sequence Diagram

sequenceDiagram
    participant INI as Options.ini
    participant OP as OptionPreferences
    participant GD as GlobalData
    participant W3DD as W3DDisplay::init()
    participant WW3D as WW3D
    participant TF as TextureFilterClass

    INI->>OP: load("Options.ini")
    OP->>GD: getAntiAliasing() → m_antiAliasLevel
    OP->>GD: getTextureFilterMode() → m_textureFilteringMode
    OP->>GD: getTextureAnisotropyLevel() → m_textureAnisotropyLevel

    GD->>W3DD: m_antiAliasLevel
    W3DD->>WW3D: Set_MSAA_Mode(m_antiAliasLevel)
    W3DD->>WW3D: Set_Render_Device(...)
    WW3D-->>W3DD: Get_MSAA_Mode() [GPU-clamped]
    W3DD->>GD: m_antiAliasLevel ← GPU-accepted value

    W3DD->>WW3D: Set_Texture_Filter(m_textureFilteringMode)
    WW3D->>TF: _Init_Filters(TextureFilter, AnisotropyLevel)
    W3DD->>WW3D: Set_Anisotropy_level(m_textureAnisotropyLevel)
    WW3D->>TF: _Set_Max_Anisotropy(level)

Loading

Prompt To Fix All With AI

This is a comment left during a code review.
Path: Core/GameEngine/Source/Common/OptionPreferences.cpp
Line: 79-80

Comment:
**No-op branch, inconsistent with the `<=` pattern**

The first `if` branch is a no-op — it reads `if (level == MULTISAMPLE_MODE_NONE) level = MULTISAMPLE_MODE_NONE`, which assigns the same value it just tested for. It works correctly only because `UnsignedInt` can never be less than zero, but it is inconsistent with every other range-clamp in the file (including `getTextureAnisotropyLevel()` and `Set_Anisotropy_level()`), all of which use `<=`. Prefer `<=` here for clarity and uniformity:

```suggestion
	if (level <= WW3D::MultiSampleModeEnum::MULTISAMPLE_MODE_NONE)
		level = WW3D::MultiSampleModeEnum::MULTISAMPLE_MODE_NONE;
```

How can I resolve this? If you propose a fix, please make it concise.

---

This is a comment left during a code review.
Path: GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Menus/OptionsMenu.cpp
Line: 586

Comment:
**`val >= 0` is an always-true condition**

`val` is declared as `Int`, and both `getTextureFilterMode()` and `getTextureAnisotropyLevel()` always return a small, bounded `UnsignedInt` (0–4 and 2/4/8/16 respectively), so this guard is never false. The same always-true pattern appears at line 616 in the anisotropy block. If the intent is simply to guard against a null `TheGlobalData`, the condition should reflect that directly:

```suggestion
	if (TheGlobalData)
```

The same applies to the matching check at line 616.

How can I resolve this? If you propose a fix, please make it concise.

_{Last reviewed commit: "feat(options): Imple..."}

[Mauller]

Generals is currently broken till i replicate across due to changes in initaialising the texture filtering