StructureRareData::m_replacementWatchpointSets should not be a pointer t o a pointer
Created attachment 428705 [details] Patch
Comment on attachment 428705 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=428705&action=review r=me > Source/JavaScriptCore/runtime/StructureRareData.h:130 > typedef HashMap<PropertyOffset, RefPtr<WatchpointSet>, WTF::IntHash<PropertyOffset>, WTF::UnsignedWithZeroKeyHashTraits<PropertyOffset>> PropertyWatchpointMap; nit: maybe add `static_assert(sizeof(PropertyWatchpointMap) == sizeof(void*));` after this line?
Created attachment 428888 [details] Patch for landing
Comment on attachment 428888 [details] Patch for landing CQ+
Committed r277620 (237835@main): <https://commits.webkit.org/237835@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 428888 [details].
<rdar://problem/78130993>
Maybe HashMap should make itself allergic to allocation on the heap. It does WTF_MAKE_FAST_ALLOCATED, but it should be WTF_MAKE_DONT_DO_IT_DONT_PUT_ME_IN_A_HEAP_BLOCK_ALL_BY_MYSELF. Same for HashTableConstIterator, HashTableIterator, HashTable, HashSet, and ListHashSet. All of these are so small it is silly to put them on the heap, I think.
Working on this in bug 225984.