Fix bugzilla issue 24882 - COM class is allocated using GC not malloc

[rikkimax]

Has spec PR: dlang/dlang.org#3928

Fixes it by introducing the trait isCOMClass

[dlang-bot]

Thanks for your pull request and interest in making D better, @rikkimax! We are looking forward to reviewing it, and you should be hearing from a maintainer soon.
Please verify that your PR follows this checklist:

• My PR is fully covered with tests (you can see the coverage diff by visiting the details link of the codecov check)
• My PR is as minimal as possible (smaller, focused PRs are easier to review than big ones)
• I have provided a detailed rationale explaining my changes
• New or modified functions have Ddoc comments (with Params: and Returns:)

Please see CONTRIBUTING.md for more information.

---

If you have addressed all reviews or aren't sure how to proceed, don't hesitate to ping us with a simple comment.

Bugzilla references

Auto-close	Bugzilla	Severity	Description
✓	24882	regression	COM class is allocated using GC not malloc

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "stable + dmd#17095"

[dkorpel]

How do you check that flag? If this works as an alternative, why is __traits(isCOMClass) introduced?

[rikkimax]

Here is the original code from the hook:

extern (C) Object _d_newclass(const ClassInfo ci) @weak
{
    import core.stdc.stdlib;
    import core.exception : onOutOfMemoryError;
    void* p;
    auto init = ci.initializer;

    debug(PRINTF) printf("_d_newclass(ci = %p, %s)\n", ci, cast(char *)ci.name);
    if (ci.m_flags & TypeInfo_Class.ClassFlags.isCOMclass)

The templated hook that has been fixed to use the new trait:

T _d_newclassT(T)() @trusted
if (is(T == class))
{
   import core.internal.traits : hasIndirections;
   import core.exception : onOutOfMemoryError;
   import core.memory : pureMalloc;
   import core.memory : GC;

   alias BlkAttr = GC.BlkAttr;

   auto init = __traits(initSymbol, T);
   void* p;

   static if (__traits(isCOMClass, T))

While both do the same thing new a class, one uses purely CT available information, the other RT only information.

[rikkimax]

Either we revert back to TypeInfo for this hook, or we go in the direction of some sort of CT available information.

Of course we could derive it by looking for a IUnknown interface, but that isn't an easy thing to do, let alone perform on every single class new instantiation when there is a flag in the compiler easily readable with the information we need.

[dkorpel]

The new trait is good, I just think the last line of the changelog should be removed or clarified. The current wording implies TypeInfo_Class can be used during compilation to check the same thing.

[rikkimax]

Done.

[dkorpel]

Something like "The trait is equivalent to the runtime check ClassInfo.m_flags & TypeInfo_Class.ClassFlags.isCOMclass ".

[rikkimax]

I don't want to encourage people to use anything TypeInfo-related, it's stated so that people know what existing code they can replace with the trait.

[adamdruppe]

What's the problem with static if(is(T : IUnknown)) again?

[rikkimax]

What's the problem with static if(is(T : IUnknown)) again?

A COM class is any class derived from an interface called IUnknown.

That would only work with a specific IUnknown interface in druntime.