Diogo Netto
Diogo Netto
> If you have a 1gb Vector{Uint8} having to copy all that data to get a string will be pretty slow (and double your memory requirement). If I'm not misunderstanding...
> Right, so if you don't copy the memory, the reason we set maxsize to 0 is to prevent the original array from mutating the string afterward. I meant to...
In that case, to avoid changing the behavior of resetting the array, I just added a call (`jl_gc_count_freed `) to track the number of freed bytes when we copy such...
Sorry about the delay. Not sure if I can see how to use this pointer tagging for chunks? Could you elaborate?
This seems like a clear improvement in GC time for the `objarray.jl` benchmark. For the other ones seems mostly neutral given the noise margins (may possibly be a very small...
Sorry about the delay. Have been OOO these days and not checking GH notifications very frequently. We didn't assess the performance impact of this change at RAI yet, so I'd...
Have benchmarked it with the new extended suite from GCBenchmarks. Seems like a win for the multithreaded benchmarks. master ``` category = "big_arrays" bench = "issue-52937.jl" [ Info: Setting GC...
We fixed in our fork. Closing it since there won't be 1.9 releases (probably) where this code is used.
> Are you able to make those small cleanups soon? It would be nice to get it in before https://github.com/JuliaLang/julia/pull/54788 causes conflicts Yes. Plan to take a look at this...