Revert #26418, remove noinline annotation from fill! #31626
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This used to be necessary to avoid a strange edge case in the compiler, but it is no longer necessary -- and can now in fact cause other performance snags. Using the test case from [the original discourse post that prompted #26418](https://discourse.julialang.org/t/performance-degradation-of-fill-in-latest-julia-0-7-dev/9648): ```julia julia> @Btime fill(1.0,5,5); 49.335 ns (1 allocation: 288 bytes) julia> @Btime fill(0.0,5,5); 52.773 ns (1 allocation: 288 bytes) julia> @Btime fill(0.0,5,5); 46.724 ns (1 allocation: 288 bytes) julia> @Btime fill(1.0,5,5); 42.202 ns (1 allocation: 288 bytes) ``` Even more compelling is the case for a larger array where LLVM can exploit some sort of wider/simdier implementation for zeros when this gets inlined thanks to constant propagation: ```julia julia> A = Array{Float64}(undef, 1000, 1000); julia> @Btime fill!($A,0.0); 345.103 μs (0 allocations: 0 bytes) julia> @Btime fill!($A,1.0); 458.976 μs (0 allocations: 0 bytes) ``` Ref https://discourse.julialang.org/t/performance-of-filling-an-array/22788
|
@nanosoldier |
KristofferC
approved these changes
Apr 5, 2019
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @ararslan |
|
I think this caused the doctests to start failing. |
KristofferC
pushed a commit
that referenced
this pull request
Apr 11, 2020
This used to be necessary to avoid a strange edge case in the compiler, but it is no longer necessary -- and can now in fact cause other performance snags. Using the test case from [the original discourse post that prompted #26418](https://discourse.julialang.org/t/performance-degradation-of-fill-in-latest-julia-0-7-dev/9648): ```julia julia> @Btime fill(1.0,5,5); 49.335 ns (1 allocation: 288 bytes) julia> @Btime fill(0.0,5,5); 52.773 ns (1 allocation: 288 bytes) julia> @Btime fill(0.0,5,5); 46.724 ns (1 allocation: 288 bytes) julia> @Btime fill(1.0,5,5); 42.202 ns (1 allocation: 288 bytes) ``` Even more compelling is the case for a larger array where LLVM can exploit some sort of wider/simdier implementation for zeros when this gets inlined thanks to constant propagation: ```julia julia> A = Array{Float64}(undef, 1000, 1000); julia> @Btime fill!($A,0.0); 345.103 μs (0 allocations: 0 bytes) julia> @Btime fill!($A,1.0); 458.976 μs (0 allocations: 0 bytes) ``` Ref https://discourse.julialang.org/t/performance-of-filling-an-array/22788
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This used to be necessary to avoid a strange edge case in the compiler, but it is no longer necessary -- and can now in fact cause other performance snags.
Using the test case from the original discourse post that prompted #26418:
Even more compelling is the case for a larger array where LLVM can exploit some sort of wider/simdier implementation for zeros when this gets inlined thanks to constant propagation:
Ref https://discourse.julialang.org/t/performance-of-filling-an-array/22788