[ZH] Fix compilation using MinGW #547

zzambers · 2025-03-30T02:31:19Z

Fixes code to be buildable using MinGW. Contains only fixes to game code. Does add support of MinGW to build system (cmake files) or any CI testing yet.

Partially fixes: #486

GeneralsMD/Code/Libraries/Source/WWVegas/WWDebug/wwprofile.h

Mauller · 2025-03-30T06:54:20Z

There are VC6 breaking changes in this due to the use of c++11 features.

If those are required then this cannot be used yet.

Marked do not merge for now.

zzambers · 2025-03-30T10:42:51Z

There are VC6 breaking changes in this due to the use of c++11 features.

If those are required then this cannot be used yet.

Marked do not merge for now.

Thanks, this is definitely not yet ready for merge :)

Yes, I was afraid that enum Name : int syntax is probably not supported by VC6, but tried it anyway. So I'll probably need to go for macros and #ifdefs to fix forward enum declarations. :( (untyped forward declarations of enums are VC extension and typed ones are since c++11 and not supported by VC6...)

Maybe macro like IENUM(Name), INT_ENUM(Name) or ENUM_INT(Name) can be crated...

xezon · 2025-03-30T14:16:45Z

Maybe macro like IENUM(Name), INT_ENUM(Name) or ENUM_INT(Name) can be crated...

Perhaps create a univeral macro

#if __cplusplus >= 201103L
    #define CPP_11(code) code
#else
    #define CPP_11(code)
#endif

Used as

enum EnumName CPP_11(: int)
{
}

Then this macro can be used for different things concerning c++11 features.

Or make it enum specific with

#if __cplusplus >= 201103L
    #define ENUM_UNDERLYING(type) : type
#else
    #define ENUM_UNDERLYING(type)
#endif

zzambers · 2025-03-30T15:20:44Z

Maybe macro like IENUM(Name), INT_ENUM(Name) or ENUM_INT(Name) can be crated...

Perhaps create a univeral macro
#if __cplusplus >= 201103L
    #define CPP_11(code) code
#else
    #define CPP_11(code)
#endif

Interesting idea with universal CPP_11 macro, I quite like it. :)

zzambers · 2025-03-30T18:08:19Z

@xezon I addapted approch with CPP_11 macro you suggested. thanks

Finally made it compile on VC6, I'll clean it up later for easier review (rebasing, combining related commits together, better commit messages etc..) I expect asm replacements to be merged to main first, I'll then rebase my changes on top.

If you prefer, I can also make fix of enums (forward declarations) separate PR. (as it is quite a big change set)

GeneralsMD/Code/GameEngine/Include/Common/ActionManager.h

GeneralsMD/Code/Libraries/Source/WWVegas/WWLib/RANDOM.H

xezon · 2025-04-09T19:44:47Z

GeneralsMD/Code/Libraries/Source/EABrowserEngine/BrowserEngine.idl

 // 
 // typelib filename: <could not determine filename>

+import "oaidl.idl";


This file was already moved to Core folder so you will need to replicate this there.

xezon · 2025-04-09T19:44:55Z

GeneralsMD/Code/Libraries/Source/EABrowserEngine/BrowserEngine.idl

 // 
 // typelib filename: <could not determine filename>

+import "oaidl.idl";


The IDL file was generated by a tool. Perhaps add a fat comment here that this has been added by hand.

Error by widl is:

BrowserEngine.idl:31:44: error: type 'IDispatch' not found in global namespace interface IFEBrowserEngine2 : IDispatch { ^

IDispatch is in oaidl.idl in wine headers.

Same with MinGW headers. (IDispatch is in oaidl.idl)

xezon · 2025-04-09T20:03:03Z

Dependencies/Utility/Utility/CppMacros.h

 #pragma once

+#include <cassert>
+#define UNIMPLEMEMTED_ERROR(msg) do { \


UNIMPLEMENTED_ERROR

Oh, I had to read it several times to spot the typo :) I'll fix that.

xezon · 2025-04-09T20:04:35Z

GeneralsMD/Code/Main/WinMain.cpp

 #include "GameClient/IMEManager.h"
 #include "Win32Device/GameClient/Win32Mouse.h"
 #include "Win32Device/Common/Win32GameEngine.h"
+#include "WWLib/trim.h"


Typing error in commit title:

"[ZH] Removal of duplicit strtrim"

"duplicate"

Ah, that's probably not an English world, sorry. :) However, I think, this has already been addressed by other PR already in meantime...

xezon · 2025-04-09T20:09:54Z

GeneralsMD/Code/Main/WinMain.cpp

How do you plan to replicate these changes if we decide to merge this before we move things to Core?

Right now the amount of single commits is massive. Replicating this equally to Generals would mean 64 commits for this effort.

Perhaps combine some of the smaller commits into new commits, and also add Generals changes to each Zero Hour change commit once we are happy with Zero Hour? This could perhaps get the commit count down to 24 or so.

Each of the commits needs to be able to compile on their own.

Originally I was just planning to do ZH as whole, followed by backport to GEN. But I see, there is quite a lot of activity in this project, causing code to diverge quickly. Also sometimes other efforts (like fixes of warnings) address same things, as this PR.

So now, I am considering breaking this into multiple smaller PRs, (which would probably be [GEN][ZH]). I have seen others following similar approach for their PRs. I would not like to end up in endless rebase loop. :)

I'll probably do bigger change sets as separate PRs (referencing discussions here), hopefully this one will then have more manageable size...

The Enum one should be a PR of its own considering how many fines it hits.

Just taking that one out will cleanup a lot of this.

EDIT: NVm i just saw that you did exactly that.

I second this. Let us get the enum fixes in one single Pull Request, without anything else. A lot of things are going into main line right now, so you want to keep the amount of things that happen inside the Pull Request small, so it can get reviewed swiftly.

GeneralsMD/Code/Main/WinMain.cpp

Mauller · 2025-04-12T13:51:40Z

At level 4 warnings, the untyped enumerations actually create nearly 10k worth of warnings, a lot of it due to the forward declarations in headers.

zzambers · 2025-04-12T19:33:33Z

@xezon @Mauller @OmniBlade @DevGeniusCode Thank you for your reviews.

As I have written higher I'll try to separate parts of this into new PRs, to keep this manageble size. (Possibly using this PR for of smaller changes.) I'll address raised issues there, linking this PR with reviews.

I'll post link to new derived PRs here

zzambers · 2025-04-12T19:34:10Z

Fixes to forward enum declarations: #665

zzambers · 2025-04-13T13:50:43Z

Rest of asm made VC only: #670

zzambers · 2025-04-13T14:36:58Z

Fixes to intrin_compat.h: #671

zzambers · 2025-04-13T15:43:34Z

COM related fixes: #672

Caball009 · 2025-06-10T04:31:54Z

A number of changes are also a part of the PR for clang-cl compilation: #888

xezon · 2025-06-23T20:39:01Z

This change is severely outdated by now.

Mauller · 2025-07-15T08:49:55Z

Is this still having work done to it?

With recent changes needed for clang build support, this PR may be significantly simplified when rebased.

Skyaero42 · 2025-11-08T19:16:28Z

@zzambers Is this something you still want to work on?

Prevent Windows headers from defining min/max macros that conflict with C++ code. Minimal change following PR TheSuperHackers#547 approach - only adds guard, keeps existing min/max macro definitions with their #ifndef guards. Co-authored-by: openhands <openhands@all-hands.dev>

Following PR TheSuperHackers#547 minimal approach: - Add #ifndef guard around NOMINMAX to prevent redefinition warnings - Wrap min/max template functions with _MIN_MAX_TEMPLATES_DEFINED_ guard to prevent conflicts when included multiple times This works with BaseTypeCore.h's NOMINMAX + existing #ifndef guards to resolve all min/max macro/template conflicts. Co-authored-by: openhands <openhands@all-hands.dev>

…re.h - Change min(x,y) to MIN(x,y) macro definition - Change max(x,y) to MAX(x,y) macro definition - Remove NOMINMAX definition (no longer needed with uppercase macros) - Prevents conflicts with C++ template functions from STL - Part of MinGW compatibility improvements following PR TheSuperHackers#547 approach

- Add conditional compilation for MinGW to use STL's min/max functions - Preserve MSVC custom template implementations - Use 'using std::min; using std::max;' for MinGW compatibility - Prevents multiple definition errors when linking with MinGW - Aligns with PR TheSuperHackers#547 approach for MinGW compatibility - Maintains VC6 and MSVC compatibility with existing custom templates

Prevent Windows headers from defining min/max macros that conflict with C++ code. Minimal change following PR TheSuperHackers#547 approach - only adds guard, keeps existing min/max macro definitions with their #ifndef guards.

Following PR TheSuperHackers#547 minimal approach: - Add #ifndef guard around NOMINMAX to prevent redefinition warnings - Wrap min/max template functions with _MIN_MAX_TEMPLATES_DEFINED_ guard to prevent conflicts when included multiple times This works with BaseTypeCore.h's NOMINMAX + existing #ifndef guards to resolve all min/max macro/template conflicts.

…re.h - Change min(x,y) to MIN(x,y) macro definition - Change max(x,y) to MAX(x,y) macro definition - Remove NOMINMAX definition (no longer needed with uppercase macros) - Prevents conflicts with C++ template functions from STL - Part of MinGW compatibility improvements following PR TheSuperHackers#547 approach

- Add conditional compilation for MinGW to use STL's min/max functions - Preserve MSVC custom template implementations - Use 'using std::min; using std::max;' for MinGW compatibility - Prevents multiple definition errors when linking with MinGW - Aligns with PR TheSuperHackers#547 approach for MinGW compatibility - Maintains VC6 and MSVC compatibility with existing custom templates

MinGW-w64 (GCC) does not allow implicit conversion from function pointers to void*. Added explicit (void*) casts to all function pointer entries in the FunctionLexicon lookup tables for both Generals and GeneralsMD. This follows the same approach used in PR TheSuperHackers#547 for MinGW compatibility while maintaining VC6 compatibility with minimal code changes.

Include stddef.h to ensure size_t is properly defined for MinGW builds. This prevents compilation errors where size_t is used but not declared. This fix is based on PR TheSuperHackers#547 by zzambers which addresses MinGW compilation issues throughout the codebase.

zzambers marked this pull request as draft March 30, 2025 02:32

DevGeniusCode reviewed Mar 30, 2025

View reviewed changes

GeneralsMD/Code/Libraries/Source/WWVegas/WWDebug/wwprofile.h Outdated Show resolved Hide resolved

Mauller added 🚫Do Not Merge Build Anything related to building, compiling ZH Relates to Zero Hour labels Mar 30, 2025

xezon added this to the Code foundation build up milestone Mar 31, 2025

zzambers mentioned this pull request Mar 31, 2025

[GEN][ZH] Add compilation for Linux #522

Closed

17 tasks

zzambers added 18 commits April 2, 2025 22:01

[ZH] Basic preprocessor fixes

6b5f516

[ZH] Replacements of _int64 with __int64

176f1ce

[ZH] Introdution of CPP_11 macro for code requiring c++11 or higher

5ec2ac7

[ZH] Forward enum declacrations portability fix

43c49b2

[ZH] Fix of missing forward declarations

d89f254

[ZH] Fix of missing includes

3c05bd8

[ZH] Fix of accesses to parent class members

1a37f94

[ZH] Fixes of template specializations

3d4c893

[ZH] Removal of static keyword for enums

32c2234

[ZH] Fixed boolean redefinition

eaf937e

[ZH] Fix of non-static declarations with static definitions

f5e72d9

[ZH] Fix of function pointer catsts to void*

862b45d

[ZH] Fix of direct constructor calls

337a01d

[ZH] Fix of unsigned values in signed arrays

d8fb831

[ZH] COM related fixes

be8d67b

[ZH] Fixes of test main methods

11928ee

[ZH] Fix to do stack trace only when support is enabled

95e8064

[ZH] Fix of unsigned short casts

ed1a5dd

xezon reviewed Apr 9, 2025

View reviewed changes

xezon reviewed Apr 10, 2025

View reviewed changes

GeneralsMD/Code/Main/WinMain.cpp Show resolved Hide resolved

xezon added Major Severity: Minor < Major < Critical < Blocker and removed 🚫Do Not Merge labels Apr 10, 2025

zzambers mentioned this pull request Apr 12, 2025

[GEN][ZH] Add underlying types to enum declarations for GCC build #665

Merged

zzambers mentioned this pull request Apr 13, 2025

[GEN][ZH] Replacements for rest of asm code #670

Open

zzambers mentioned this pull request Apr 13, 2025

Fix intrin_compat.h to work for MinGW #671

Merged

zzambers mentioned this pull request Apr 13, 2025

[GEN][ZH] Fix COM related compile errors for MinGW #672

Merged

Skaronator mentioned this pull request Jul 18, 2025

[GITHUB][ZH] Validate replays for mismatches in CI #1239

Closed

JohnsterID mentioned this pull request Jan 6, 2026

build: Add MinGW-w64 32-bit (i686) cross-compilation support #2067

Draft

[ZH] Fix compilation using MinGW #547

Are you sure you want to change the base?

[ZH] Fix compilation using MinGW #547

Uh oh!

Conversation

zzambers commented Mar 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Mauller commented Mar 30, 2025

Uh oh!

zzambers commented Mar 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

xezon commented Mar 30, 2025

Uh oh!

zzambers commented Mar 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zzambers commented Mar 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zzambers Apr 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Mauller Apr 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Mauller commented Apr 12, 2025

Uh oh!

zzambers commented Apr 12, 2025

Uh oh!

zzambers commented Apr 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zzambers commented Apr 13, 2025

Uh oh!

zzambers commented Apr 13, 2025

Uh oh!

zzambers commented Apr 13, 2025

Uh oh!

Caball009 commented Jun 10, 2025

Uh oh!

xezon commented Jun 23, 2025

Uh oh!

Mauller commented Jul 15, 2025

Uh oh!

Skyaero42 commented Nov 8, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

zzambers commented Mar 30, 2025 •

edited

Loading

zzambers commented Mar 30, 2025 •

edited

Loading

zzambers commented Mar 30, 2025 •

edited

Loading

zzambers commented Mar 30, 2025 •

edited

Loading

zzambers Apr 12, 2025 •

edited

Loading

Mauller Apr 12, 2025 •

edited

Loading

zzambers commented Apr 12, 2025 •

edited

Loading