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.

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

Fix rewriter for version-gated expressions #535

Merged
merged 1 commit into from
Feb 4, 2025

Conversation

christiangnrd
Copy link
Contributor

No description provided.

@christiangnrd christiangnrd merged commit 924a130 into JuliaGPU:main Feb 4, 2025
5 of 7 checks passed
@christiangnrd christiangnrd deleted the fixrewriter branch February 4, 2025 19:19
Copy link

codecov bot commented Feb 4, 2025

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 75.57%. Comparing base (52d7056) to head (769bc84).
Report is 412 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff             @@
##             main     #535      +/-   ##
==========================================
+ Coverage   71.04%   75.57%   +4.53%     
==========================================
  Files          36       57      +21     
  Lines        1143     2731    +1588     
==========================================
+ Hits          812     2064    +1252     
- Misses        331      667     +336     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Copy link
Contributor

@github-actions github-actions bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Metal Benchmarks

Benchmark suite Current: 769bc84 Previous: ca092c8 Ratio
private array/construct 27670.083333333336 ns 24829.916666666664 ns 1.11
private array/broadcast 467334 ns 458500 ns 1.02
private array/random/randn/Float32 796333 ns 798750 ns 1.00
private array/random/randn!/Float32 620875 ns 615041.5 ns 1.01
private array/random/rand!/Int64 567875 ns 563000 ns 1.01
private array/random/rand!/Float32 587083 ns 598021 ns 0.98
private array/random/rand/Int64 787625 ns 774083 ns 1.02
private array/random/rand/Float32 614958 ns 611583 ns 1.01
private array/copyto!/gpu_to_gpu 660500 ns 654250 ns 1.01
private array/copyto!/cpu_to_gpu 638937.5 ns 624208 ns 1.02
private array/copyto!/gpu_to_cpu 817083 ns 817708 ns 1.00
private array/accumulate/1d 1333625 ns 1329333 ns 1.00
private array/accumulate/2d 1390791 ns 1398375 ns 0.99
private array/iteration/findall/int 2066291 ns 2103583.5 ns 0.98
private array/iteration/findall/bool 1835875 ns 1824375 ns 1.01
private array/iteration/findfirst/int 1710958 ns 1688792 ns 1.01
private array/iteration/findfirst/bool 1662750 ns 1643000 ns 1.01
private array/iteration/scalar 3898125 ns 3772458 ns 1.03
private array/iteration/logical 3167187 ns 3187749.5 ns 0.99
private array/iteration/findmin/1d 1755833 ns 1760708 ns 1.00
private array/iteration/findmin/2d 1350417 ns 1344437.5 ns 1.00
private array/reductions/reduce/1d 1030750 ns 1031583 ns 1.00
private array/reductions/reduce/2d 658083.5 ns 654750 ns 1.01
private array/reductions/mapreduce/1d 1043083 ns 1033875 ns 1.01
private array/reductions/mapreduce/2d 658833 ns 659000 ns 1.00
private array/permutedims/4d 2541583 ns 2503500 ns 1.02
private array/permutedims/2d 1013146 ns 1028750 ns 0.98
private array/permutedims/3d 1582042 ns 1580708 ns 1.00
private array/copy 576500 ns 590270.5 ns 0.98
latency/precompile 8890282000 ns 8811389416 ns 1.01
latency/ttfp 3608043458 ns 3608628500 ns 1.00
latency/import 1236656084 ns 1231898292 ns 1.00
integration/metaldevrt 685750 ns 713792 ns 0.96
integration/byval/slices=1 1617625 ns 1617854.5 ns 1.00
integration/byval/slices=3 9300999.5 ns 9687812.5 ns 0.96
integration/byval/reference 1585438 ns 1589625 ns 1.00
integration/byval/slices=2 2692146 ns 2675542 ns 1.01
kernel/indexing 458500 ns 470792 ns 0.97
kernel/indexing_checked 457958 ns 463208 ns 0.99
kernel/launch 10152.666666666666 ns 9527.666666666666 ns 1.07
metal/synchronization/stream 14375 ns 15125 ns 0.95
metal/synchronization/context 15083 ns 14834 ns 1.02
shared array/construct 26701.416666666664 ns 24604.166666666668 ns 1.09
shared array/broadcast 460667 ns 461166 ns 1.00
shared array/random/randn/Float32 825292 ns 738958.5 ns 1.12
shared array/random/randn!/Float32 626541 ns 633292 ns 0.99
shared array/random/rand!/Int64 574958.5 ns 561625 ns 1.02
shared array/random/rand!/Float32 590833 ns 600416 ns 0.98
shared array/random/rand/Int64 798104 ns 778375 ns 1.03
shared array/random/rand/Float32 610875 ns 616000 ns 0.99
shared array/copyto!/gpu_to_gpu 85583 ns 79250 ns 1.08
shared array/copyto!/cpu_to_gpu 82875 ns 82084 ns 1.01
shared array/copyto!/gpu_to_cpu 83458 ns 82750 ns 1.01
shared array/accumulate/1d 1353624.5 ns 1335833 ns 1.01
shared array/accumulate/2d 1384583.5 ns 1388833 ns 1.00
shared array/iteration/findall/int 1805750 ns 1871833 ns 0.96
shared array/iteration/findall/bool 1595208 ns 1569500 ns 1.02
shared array/iteration/findfirst/int 1392209 ns 1396916 ns 1.00
shared array/iteration/findfirst/bool 1365083.5 ns 1367500 ns 1.00
shared array/iteration/scalar 155916 ns 154834 ns 1.01
shared array/iteration/logical 2976709 ns 2987020.5 ns 1.00
shared array/iteration/findmin/1d 1474896 ns 1477062.5 ns 1.00
shared array/iteration/findmin/2d 1365209 ns 1364708 ns 1.00
shared array/reductions/reduce/1d 723854 ns 731750 ns 0.99
shared array/reductions/reduce/2d 651792 ns 666250 ns 0.98
shared array/reductions/mapreduce/1d 740791 ns 736667 ns 1.01
shared array/reductions/mapreduce/2d 655208 ns 672459 ns 0.97
shared array/permutedims/4d 2453041 ns 2493333 ns 0.98
shared array/permutedims/2d 1014500 ns 1024646 ns 0.99
shared array/permutedims/3d 1567583 ns 1576667 ns 0.99
shared array/copy 244645.5 ns 244000 ns 1.00

This comment was automatically generated by workflow using github-action-benchmark.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant