[Lldb-commits] [lldb] Lazy deference underlying object for shared/weak/unique_ptr synthetic… (PR #67069)

[via lldb-commits]

jimingham wrote:

How right you are.  

I was thinking this was the ValueObject updated stuff, but it's the SyntheticChildrenFrontEnd::Update.  If I'm reading the code in ValueObjectSyntheticFilter.cpp aright, the return for that actually means the opposite of what you remembered.  If you return `false` from the SyntheticChildrenFrontEnd::Update, that means the VOSynthetic that uses this FrontEnd has to delete all its cached children, and recreate them when asked by GetChildAtIndex.

So for instance, in the case where a synthetic child front end just hands out a ValueObjectChild in the tree of the underlying ValueObject, then the child that gets cached in the VOSynthetic::m_synthetic_children_cache  knows how to update itself from its real parent, it doesn't need help from the ValueObjectSynthetic's FrontEndProvider to get the correct value.  So there's no point removing it from the cache and making a new child every time we stop, the one in the cache knows how to keep its value up to date.

But in the case where the children are, for instance, ValueObjectConstResults calculated by some means by the FrontEnd, to return `true` from `Update`, you would have to know that the FrontEnd at this stop point would produce the same VOConstResult as it did when you last stopped.  Or you would have to know how to manually update each of the cached children.  It would be possible for Update to do either of those things, but in general it seems like that's too much trouble and we really never try to figure that out. 

I am pretty sure the "Dereferenced" child of a ValueObject can update itself from the VO that generated it, in which case, the two formatters in this patch could have returned true and saved the work of recreating the child VO as the dereference.  But I haven't tested that theory, and I doubt just making up a new VO is worth the effort.  The only downside is if you keep throwing the children away at every stop, the synthetic children have no hope of reporting "IsChanged" correctly.

Anyway, not relevant to this patch.  I think returning false was wrong before, but that's just a small bit more work, so it's not worth worrying over.

https://github.com/llvm/llvm-project/pull/67069