Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on poptask #50802

Merged
merged 1 commit into from
Aug 10, 2023
Merged

Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on poptask #50802

merged 1 commit into from
Aug 10, 2023

Conversation

andrebsguedes
Copy link
Contributor

Currently poptask is allocating every time it calls into jl_task_get_next due to StickyWorkqueue (IntrusiveLinkedListSynchronized) being immutable and requiring an allocation to be used as Any on the ccall.

The allocations can be seen in the following snippet:

function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)

Which after the change becomes:

julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)

@jpsamaroo jpsamaroo added the multithreading Base.Threads and related functionality label Aug 5, 2023
@NHDaly NHDaly added the performance Must go faster label Aug 7, 2023
@NHDaly NHDaly mentioned this pull request Aug 9, 2023
Closed
@kpamnany kpamnany requested a review from vtjnash August 9, 2023 15:55
@tveldhui
Copy link

tveldhui commented Aug 9, 2023

Oh nice!! I had seen allocations here but was mystified since it was an IntrusiveLinkedList.

@kpamnany kpamnany merged commit 2f03072 into JuliaLang:master Aug 10, 2023
3 checks passed
@kpamnany
Copy link
Contributor

@KristofferC: can we backport to 1.10?

@KristofferC KristofferC added the backport 1.10 Change should be backported to the 1.10 release label Aug 10, 2023
KristofferC pushed a commit that referenced this pull request Aug 10, 2023
@andrebsguedes @KristofferC
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
185dd81 
…poptask (#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```

(cherry picked from commit 2f03072)
kpamnany pushed a commit to RelationalAI/julia that referenced this pull request Aug 16, 2023
@andrebsguedes @kpamnany
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
c842f64 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
@IanButterworth IanButterworth mentioned this pull request Aug 19, 2023
Merged
58 tasks
kpamnany pushed a commit to RelationalAI/julia that referenced this pull request Aug 21, 2023
@andrebsguedes @kpamnany
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
29e2849 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
quinnj pushed a commit to RelationalAI/julia that referenced this pull request Sep 5, 2023
@andrebsguedes @quinnj
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
3ba6f48 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Sep 28, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
c3435b8 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 1, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
3d148b4 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
KristofferC added a commit that referenced this pull request Oct 2, 2023
@KristofferC
[release-1.10] Backports for Julia 1.10.0-x (#50971)
e6378ea 
Backported PRs:
- [x] #48625 <!-- add replace(io, str, patterns...) -->
- [x] #48387 <!-- Improve documentation of sort-related functions -->
- [x] #48363 <!-- Revise sort.md and docstrings in sort.jl -->
- [x] #48977 <!-- Update SparseArrays.jl stdlib for SuiteSparse 7 -->
- [x] #50719 <!-- fix `CyclePadding(::DataType)` -->
- [x] #50694 <!-- inference: permit recursive type traits -->
- [x] #50860 <!-- Add `Base.get_extension` to docs/API -->
- [x] #50594 <!-- Disallow non-index Integer types in isassigned -->
- [x] #50802 <!-- Makes IntrusiveLinkedListSynchronized mutable to avoid
allocation on poptask -->
- [x] #50858 <!-- Add a `threadpool` parameter to `Channel` constructor
-->
- [x] #50874 <!-- Restrict COFF to a single thread when symbol count is
high -->
- [x] #50822 <!-- Add default method for setmodifiers! -->
- [x] #50730 <!-- Fix integer overflow in `isapprox` -->
- [x] #50850 <!-- Remove weird Rational dispatch and add pi functions to
list -->
- [x] #50809 <!-- Limit type-printing in MethodError -->
- [x] #50915 <!-- Add note the `Task` about sticky bit -->
- [x] #50929 <!-- when widening tuple types in tmerge, only widen the
complex parts -->
- [x] #50928 <!-- Bump JuliaSyntax to 0.4.6 -->
- [x] #50959 <!-- Update libssh2 patches -->
- [x] #50823 <!-- Make ranges more robust with unsigned indexes. -->
- [x] #48542 <!-- Add docs on task-specific buffering using
multithreading -->
- [x] #50912 <!-- Separate foreign threads into a :foreign threadpool
-->
- [x] #51010 <!-- Add ORIGIN to SuiteSparse rpath on Linux/FreeBSD -->
- [x] #50753 <!-- cat: remove unused promote_eltype methods that confuse
inference -->
- [x] #51027 <!-- Implement realloc accounting correctly -->
- [x] #51019 <!-- fix a case of potentially use of undefined variable
when handling error in distributed message processing -->
- [x] #51039 <!-- Revert "Optimize findall(f, ::AbstractArray{Bool})
(#42202)" -->
- [x] #51036 <!-- add missing invoke edge for nospecialize targets -->
- [x] #51042 <!-- inference: fix return_type_tfunc modeling of concrete
functions -->
- [x] #51114 <!-- Workaround upstream FreeBSD issue #272992 -->
- [x] #50892 <!-- Add `JL_DLLIMPORT` to `small_typeof` declaration -->
- [x] #51154 <!-- broadcast: use recursion rather than ntuple to map
over a tuple -->
- [x] #51153 <!-- fix debug typo in "add missing invoke edge for
nospecialize targets (#51036)" -->
- [x] #51222 <!-- Check again if the tty is open inside the IO lock -->
- [x] #51236 <!-- Add lock around uv_unref during init -->
- [x] #51243 <!-- GMP: Gracefully handle more overflows. -->
- [x] #51254 <!-- Ryu: make sure adding zeros does not overwrite
trailing dot -->
- [x] #51175 <!-- shorten stale_age for cachefile lock -->
- [x] #51300 <!-- fix method definition error for bad vararg -->
- [x] #51307 <!-- fix force-throw ctrl-C on Windows -->
- [x] #51303 <!-- ensure revising structs is safe -->
- [x] #51393 
- [x] #51403 

Need manual backport:
- [x] #51009 <!-- fix #50562, regression in `in` of tuple of Symbols -->
- [x] #51053 <!-- Bump Statistics stdlib -->
- [x] #51013 <!-- fix #50709, type bound error due to free typevar in
sparam env -->
- [x] #51305 <!-- reduce test time for rounding and floatfuncs -->

Contains multiple commits, manual intervention needed:
- [ ] #50663 <!-- Fix Expr(:loopinfo) codegen -->
- [ ] #51035 <!-- refactor GC scanning code to reflect jl_binding_t are
now first class -->
- [ ] #51092 <!-- inference: fix bad effects for recursion -->
- [x] #51247 <!-- Check if malloc has succeeded before incrementing gc
stats -->
- [x] #51294 <!-- use LibGit2_jll for LibGit2 library -->

Non-merged PRs with backport label:
- [ ] #51132 <!-- Handle `AbstractQ` in concatenation -->
- [x] #51029 <!-- testdefs: make sure that if a test set changes the
active project, they change it back when they're done -->
- [ ] #50919 <!-- Code loading: do the "skipping mtime check for stdlib"
check regardless of the value of `ispath(f)` -->
- [ ] #50824 <!-- Add some aliasing warnings to docstrings for mutating
functions -->
- [x] #50385 <!-- Precompile pidlocks: add to NEWS and docs -->
- [ ] #49805 <!-- Limit TimeType subtraction to AbstractDateTime -->
@KristofferC KristofferC removed the backport 1.10 Change should be backported to the 1.10 release label Oct 2, 2023
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 12, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
f0b3ffe 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 14, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
75c3a18 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 17, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
e5d16ae 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 18, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
45a6a9b 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
kpamnany pushed a commit to RelationalAI/julia that referenced this pull request Oct 18, 2023
@andrebsguedes @kpamnany
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
f73e723 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
kpamnany pushed a commit to RelationalAI/julia that referenced this pull request Oct 19, 2023
@andrebsguedes @kpamnany
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
ed4168e 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 20, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
747fb2f 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
kpamnany pushed a commit to RelationalAI/julia that referenced this pull request Oct 21, 2023
@andrebsguedes @kpamnany
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
444170f 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Oct 23, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
32162ae 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 1, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
5899d7f 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 2, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
bc01fa8 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 7, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
c3f8770 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 10, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
de1666e 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 14, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
4003623 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
DelveCI pushed a commit to RelationalAI/julia that referenced this pull request Nov 15, 2023
@andrebsguedes
Makes IntrusiveLinkedListSynchronized mutable to avoid allocation on …
b006695 
…poptask (JuliaLang#50802)

Currently `poptask` is allocating every time it calls into
`jl_task_get_next` due to `StickyWorkqueue`
(`IntrusiveLinkedListSynchronized`) being immutable and requiring an
allocation to be used as `Any` on the `ccall`.

The allocations can be seen in the following snippet:
```
function work()
    done = 0.0
    l = Threads.SpinLock()
    Threads.@sync for _ in 1:(Threads.nthreads() / 2)
        Threads.@Spawn begin
            v = 1.0
            for i in 1:(1<<33)
                v *= 1.0001
                if i % (1 << 13) == 0
                    yield()
                end
            end
            Base.@lock_nofail l done += v
        end
    end
    return done
end

julia> @time work()
 15.758794 seconds (5.03 M allocations: 153.523 MiB, 0.35% gc time)
```

Which after the change becomes:
```
julia> @time work()
 15.907513 seconds (67 allocations: 4.719 KiB)
```
@mofeing mofeing mentioned this pull request Oct 27, 2024
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@vtjnash vtjnash vtjnash approved these changes

Assignees
No one assigned
Labels
multithreading Base.Threads and related functionality performance Must go faster
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
7 participants
@andrebsguedes @tveldhui @kpamnany @vtjnash @KristofferC @NHDaly @jpsamaroo