René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"MSVC 5274 warning suppression for _Py_ALIGN_AS should be local","articleBody":"# Bug report\n\nIn https://github.com/python/cpython/pull/133966#issuecomment-2888485497 we added a `Python.h`-wide suppression for MSVC warning [5274](https://learn.microsoft.com/en-us/cpp/overview/cpp-conformance-improvements?view=msvc-170#application-of-_alignas-on-a-structured-type-in-c) “Application of _Alignas on a structured type in C”.\n\nI think it should be instead applied locally, after a bit of thought.\n\nAFAICS, 5274 warns that structs compiled with MSVC versions older than 17.9 might end up with a different layout (ABI) than ones compiled with 17.9+. That can cause trouble for extensions.\n\nIn `PyASCIIObject`, `_Py_ALIGN_AS` is applied to an anonymous struct. I think we can assume that no one is pulling `state`'s type out with something like `typeof(my_PyASCIIObject.state)` – and if they do, getting an 4B-aligned struct is very unlikely to cause trouble in practice.\nBut, that is the kind of analysis one should do when a 5274 warning pops up.\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-135184\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/encukou","@type":"Person","name":"encukou"},"datePublished":"2025-06-05T14:40:23.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":1},"url":"https://github.com/135183/cpython/issues/135183"}