{"id":108192,"date":"2023-05-16T07:00:00","date_gmt":"2023-05-16T14:00:00","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/oldnewthing\/?p=108192"},"modified":"2023-05-15T19:58:12","modified_gmt":"2023-05-16T02:58:12","slug":"20230516-00","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/oldnewthing\/20230516-00\/?p=108192","title":{"rendered":"Summary of the duck-typing requirements of C++ COM wrappers"},"content":{"rendered":"

Here’s a summary of the tables we spent the past few days building. Points of interest are highlighted.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
\u00a0<\/th>\n_com_ptr_t<\/code><\/th>\nMFC IPTR<\/code><\/th>\nATL CComPtr<\/code><\/th>\nWRL ComPtr<\/code><\/th>\nwil::com_ptr<\/code><\/th>\nC++\/WinRT
\nwinrt::com_ptr<\/code><\/th>\n<\/tr>\n
Default construction<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Construct from raw pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nN\/A<\/td>\n<\/tr>\n
Copy construction<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Destruction<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Attach and detach<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Assign to same-type raw pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass (copy_from<\/code>)<\/td>\n<\/tr>\n
Assign to same-type smart pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Fetch the wrapped pointer<\/td>\nGetInterfacePtr()<\/code><\/td>\nGetInterfacePtr()<\/code><\/td>\np<\/code>, or
\nimplicit conversion<\/td>\n		Get()<\/code><\/td>\nget()<\/code><\/td>\nget()<\/code><\/td>\n<\/tr>\n
Access the wrapped object<\/td>\n-><\/code><\/td>\n-><\/code><\/td>\n-><\/code><\/p>\n
(suboptimal)<\/div>\n<\/td>\n
-><\/code><\/td>\n-><\/code><\/td>\n-><\/code><\/td>\n<\/tr>\n
Receive pointer via &<\/code><\/td>\nrelease old<\/td>\nrelease old<\/td>\nmust be empty<\/td>\nrelease old<\/td>\nrelease old<\/td>\nN\/A<\/td>\n<\/tr>\n
Release and receive pointer<\/td>\n&<\/code><\/td>\n&<\/code><\/td>\n&<\/code><\/td>\nReleaseAndGetAddressOf()<\/code><\/td>\nput()<\/code><\/td>\nput()<\/code><\/td>\n<\/tr>\n
Preserve and receive pointer<\/td>\nN\/A<\/td>\nN\/A<\/td>\n&p<\/code><\/td>\nGetAddressOf()<\/code><\/td>\naddressof()<\/code><\/td>\nN\/A<\/td>\n<\/tr>\n
Return to empty state<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Comparison<\/td>\nFail<\/td>\nN\/A<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Accidental bypass<\/td>\nFail<\/td>\nFail<\/td>\nPass<\/td>\nFail<\/td>\nFail<\/td>\nFail<\/td>\n<\/tr>\n
Construct from other-type raw pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Construct from other-type smart pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Assign from other-type raw pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
Assign from other-type smart pointer<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\nPass<\/td>\n<\/tr>\n
T<\/code> can be final<\/code><\/td>\nYes<\/td>\nYes<\/td>\nNo<\/td>\nYes<\/td>\nYes<\/td>\nYes<\/td>\n<\/tr>\n
T::Release<\/code> must return ULONG<\/code><\/td>\nNo<\/td>\nNo<\/td>\nYes<\/td>\nYes<\/td>\nYes<\/td>\nNo<\/td>\n<\/tr>\n
T::Release<\/code> must be __stdcall<\/code><\/td>\nNo<\/td>\nNo<\/td>\nYes<\/td>\nNo<\/td>\nNo<\/td>\nNo<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

For the most part, all of the smart pointer wrappers can wrap pointers to non-COM objects, provided they have AddRef<\/code> and Release<\/code> methods which increment and decrement the object reference count.<\/p>\n

ATL is the only one that protects against accidental bypass by using the “coloring” technique. However, the cost is that the wrapped object may not be final<\/code>. Furthermore, its implementation of “coloring” is suboptimal if the underlying object does not derive from IUnknown<\/code>. (If it does, then the “coloring” has no overhead.)<\/p>\n

The semantics of the &<\/code> address-of operator vary greatly, a topic I had discussed some time ago. ATL is the only one whose &<\/code> operator does not release the pointer before producing the address. C++\/WinRT does not overload the &<\/code> operator at all.<\/p>\n

_com_ptr_t<\/code>, IPTR<\/code>, and C++\/WinRT do not provide a way to access the wrapped pointer’s address without freeing it.<\/p>\n

Comparison operators break only in _com_ptr_t<\/code>. (Comparison operators never worked with IPTR<\/code>, so there was nothing to break.)<\/p>\n

C++\/WinRT does not support constructing directly from a raw pointer. This is annoying because the case we are most likely to use it for a non-IUnknown-derived class is specifically to construct directly from this<\/code>:<\/p>\n

\/\/ Sadly doesn't work\r\nRegisterCallback([strongThis = winrt::com_ptr(this)] { ... });\r\n<\/pre>\n
ATL, WRL, and wil require that the Release<\/code> method return the new reference count as a ULONG<\/code>. ATL even requires that the Release<\/code> method be __stdcall<\/code>.<\/p>\n
The MFC IPTR<\/code> macro is not really used any more, but I listed it in the table for completeness. If you try to use it in a modern compiler, you will get warnings about its use of exception specifiers, which were deprecated in C++17.<\/p>\n
Glancing through the table, it seems that the least weird wrapper classes for wrapping types that aren’t COM interfaces are WRL::ComPtr<\/code> and wil::com_ptr<\/code>. You do have to make sure that your Release()<\/code> method returns the new reference count, but this is likely something you already have readily available, so returning it is not that big of an obstacle.<\/p>\n","protected":false},"excerpt":{"rendered":"
Wrapping up and comparing.<\/p>\n","protected":false},"author":1069,"featured_media":111744,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[25],"class_list":["post-108192","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-oldnewthing","tag-code"],"acf":[],"blog_post_summary":"
Wrapping up and comparing.<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/108192","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/users\/1069"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/comments?post=108192"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/108192\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media\/111744"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media?parent=108192"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/categories?post=108192"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/tags?post=108192"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}