You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
sfink: Yes, a [GCHashMap<JS::Heap<JSString*>>] seems wrong to me if any of the string keys could be nursery strings. DefaultHasher<JSString*> just uses the pointer
one option would be to hash and match on the string contents. We have a couple of instances of that in the tree. One for string deduplication, one for counting duplicate strings for about:memory. tcampbell: what version did we turn on nursery strings?
this probably needs a good note and example in the examples repo.. sfink: [...] I was surprised that we make so little of strings as keys; there's not much prior art in the tree.
it looks nursery strings were Fx67, from bug 1442481
We should have an example of these sorts of hash maps using MovableCellHasher.
The text was updated successfully, but these errors were encountered:
Conversation starting here: https://chat.mozilla.org/#/room/#spidermonkey:mozilla.org/$YXBhZxFLHDHPc9pRp6Jz2-Vm9EDbzec0Rbeig_fPtJM
Especially:
We should have an example of these sorts of hash maps using MovableCellHasher.
The text was updated successfully, but these errors were encountered: