Can someone review #70193 (Limit maximum alignment in #[repr(align)] to 4K)? It keeps getting bounced from reviewer to reviewer and somehow ended up on my list...
@Amanieu says that
I believe the latest consensus is that we should emit a warning, with the intention of changing it to a hard error later.
Is that decision documented somewhere? The last I saw is @nikomatsakis asking some questions.
I believe it was discussed in the lang team meeting.
I guess I could reassign it to Niko for the moment. Sorry for the "reviewer juggling" :sweat_smile:
I can confirm that that was the decision. We did, however, want the limit to come from the target definition, so that targets can easily define a larger limit.