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

Some cudadrv tests #2684

Merged
merged 2 commits into from
Mar 10, 2025
Merged

Some cudadrv tests #2684

merged 2 commits into from
Mar 10, 2025

Conversation

kshyatt
Copy link
Contributor

@kshyatt kshyatt commented Mar 8, 2025

Also added some more @testset blocks instead of let

@kshyatt kshyatt added the tests Adds or changes tests. label Mar 8, 2025
@kshyatt kshyatt requested a review from maleadt March 8, 2025 23:47
@github-actions GitHub Actions
Copy link
Contributor

github-actions bot commented Mar 8, 2025 •
edited
Loading

Your PR requires formatting changes to meet the project's style guidelines.
Please consider running Runic (git runic master) to apply these changes.

Click here to view the suggested changes. 
diff --git a/src/texture.jl b/src/texture.jl
index fc1590423..bbdca57db 100644
--- a/src/texture.jl
+++ b/src/texture.jl
@@ -97,7 +97,7 @@ end
 
 # idempotency
 CuTextureArray{T,N}(xs::CuTextureArray{T,N}) where {T,N} = xs
-CuTextureArray(xs::CuTextureArray{T,N}) where {T,N} = xs
+CuTextureArray(xs::CuTextureArray{T, N}) where {T, N} = xs
 
 CuTextureArray(A::AbstractArray{T,N}) where {T,N} = CuTextureArray{T,N}(A)
 
diff --git a/test/core/cudadrv.jl b/test/core/cudadrv.jl
index 2372cb568..7772cf9cb 100644
--- a/test/core/cudadrv.jl
+++ b/test/core/cudadrv.jl
@@ -435,15 +435,15 @@ nb = sizeof(data)
 typed_pointer(buf::Union{CUDA.DeviceMemory, CUDA.UnifiedMemory}, T) = convert(CuPtr{T}, buf)
 typed_pointer(buf::CUDA.HostMemory, T)                              = convert(Ptr{T},   buf)
 
-@testset "showing" begin
-    for (Ty, str) in zip([CUDA.DeviceMemory, CUDA.HostMemory, CUDA.UnifiedMemory], ("DeviceMemory", "HostMemory", "UnifiedMemory"))
-        dummy = CUDA.alloc(Ty, 0)
-        @test startswith(sprint(show, dummy), str)
-        CUDA.free(dummy)
+    @testset "showing" begin
+        for (Ty, str) in zip([CUDA.DeviceMemory, CUDA.HostMemory, CUDA.UnifiedMemory], ("DeviceMemory", "HostMemory", "UnifiedMemory"))
+            dummy = CUDA.alloc(Ty, 0)
+            @test startswith(sprint(show, dummy), str)
+            CUDA.free(dummy)
+        end
     end
-end
 
-@testset "allocations and copies, src $srcTy dst $dstTy" for srcTy in [CUDA.DeviceMemory, CUDA.HostMemory, CUDA.UnifiedMemory],
+    @testset "allocations and copies, src $srcTy dst $dstTy" for srcTy in [CUDA.DeviceMemory, CUDA.HostMemory, CUDA.UnifiedMemory],
     dstTy in [CUDA.DeviceMemory, CUDA.HostMemory, CUDA.UnifiedMemory]
 
     dummy = CUDA.alloc(srcTy, 0)
@@ -479,7 +479,7 @@ end
 
     # test device with context in which pointer was allocated.
     @test device(typed_pointer(src, T)) == device()
-    @test context(typed_pointer(src, T)) == context()
+        @test context(typed_pointer(src, T)) == context()
     if !memory_pools_supported(device())
         # NVIDIA bug #3319609
         @test context(typed_pointer(src, T)) == context()
@@ -503,7 +503,7 @@ end
     CUDA.free(dst)
 end
 
-@testset "pointer attributes" begin
+    @testset "pointer attributes" begin
     src = CUDA.alloc(CUDA.DeviceMemory, nb)
 
     attribute!(typed_pointer(src, T), CUDA.POINTER_ATTRIBUTE_SYNC_MEMOPS, 0)
@@ -511,7 +511,7 @@ end
     CUDA.free(src)
 end
 
-@testset "asynchronous operations" begin
+    @testset "asynchronous operations" begin
     src = CUDA.alloc(CUDA.DeviceMemory, nb)
 
     unsafe_copyto!(typed_pointer(src, T), pointer(data), N; async=true)
@@ -521,7 +521,7 @@ end
     CUDA.free(src)
 end
 
-@testset "pinned memory" begin
+    @testset "pinned memory" begin
     # create a pinned and mapped buffer
     src = CUDA.alloc(CUDA.HostMemory, nb, CUDA.MEMHOSTALLOC_DEVICEMAP)
 
@@ -553,16 +553,16 @@ if attribute(device(), CUDA.DEVICE_ATTRIBUTE_HOST_REGISTER_SUPPORTED) != 0
 
     CUDA.unregister(src)
 
-    # with a RefValue
-    src = Ref{T}(T(42))
-    CUDA.pin(src)
-    cpu_ptr = Base.unsafe_convert(Ptr{T}, src)
-    ref = Array{T}(undef, 1)
-    unsafe_copyto!(pointer(ref), cpu_ptr, 1)
-    @test ref == [T(42)]
+        # with a RefValue
+        src = Ref{T}(T(42))
+        CUDA.pin(src)
+        cpu_ptr = Base.unsafe_convert(Ptr{T}, src)
+        ref = Array{T}(undef, 1)
+        unsafe_copyto!(pointer(ref), cpu_ptr, 1)
+        @test ref == [T(42)]
 end
 
-@testset "unified memory" begin
+    @testset "unified memory" begin
     src = CUDA.alloc(CUDA.UnifiedMemory, nb)
 
     @test_throws BoundsError CUDA.prefetch(src, 2*nb; device=CUDA.DEVICE_CPU)
@@ -583,7 +583,7 @@ end
     CUDA.free(src)
 end
 
-@testset "3d memcpy" begin
+    @testset "3d memcpy" begin
     # TODO: use cuMemAllocPitch (and put pitch in buffer?) to actually get benefit from this
 
     data = collect(reshape(1:27, 3, 3, 3))

@codecov Codecov
Copy link

codecov bot commented Mar 9, 2025 •
edited
Loading

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.86%. Comparing base (6bf72dd) to head (dbb2215).

Additional details and impacted files 
@@            Coverage Diff             @@
##           master    #2684      +/-   ##
==========================================
+ Coverage   82.57%   82.86%   +0.28%     
==========================================
  Files         153      153              
  Lines       13606    13606              
==========================================
+ Hits        11235    11274      +39     
+ Misses       2371     2332      -39

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

🚀 New features to boost your workflow:
  • ❄ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

github-actions[bot]
github-actions bot reviewed Mar 9, 2025
View reviewed changes
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.

CUDA.jl Benchmarks

Benchmark suite Current: dbb2215 Previous: 6bf72dd Ratio
latency/precompile 46183641505 ns 46450283323 ns 0.99
latency/ttfp 7001382825 ns 7028014177 ns 1.00
latency/import 3651822156 ns 3667348189 ns 1.00
integration/volumerhs 9616637.5 ns 9625836 ns 1.00
integration/byval/slices=1 146680.5 ns 146875 ns 1.00
integration/byval/slices=3 425045 ns 424860 ns 1.00
integration/byval/reference 144900 ns 144959 ns 1.00
integration/byval/slices=2 285974 ns 285961 ns 1.00
integration/cudadevrt 103233 ns 103332 ns 1.00
kernel/indexing 13949 ns 14061 ns 0.99
kernel/indexing_checked 14499 ns 14775 ns 0.98
kernel/occupancy 637.1046511627907 ns 656.859649122807 ns 0.97
kernel/launch 2004.2 ns 2090.2 ns 0.96
kernel/rand 14598 ns 16643 ns 0.88
array/reverse/1d 19363 ns 19602 ns 0.99
array/reverse/2d 24535 ns 24553 ns 1.00
array/reverse/1d_inplace 10719 ns 11159 ns 0.96
array/reverse/2d_inplace 12480 ns 13010 ns 0.96
array/copy 21175 ns 20672 ns 1.02
array/iteration/findall/int 157906 ns 157732 ns 1.00
array/iteration/findall/bool 138756 ns 138853.5 ns 1.00
array/iteration/findfirst/int 152794 ns 153651.5 ns 0.99
array/iteration/findfirst/bool 154428 ns 153926 ns 1.00
array/iteration/scalar 70496 ns 72447 ns 0.97
array/iteration/logical 213423 ns 206640.5 ns 1.03
array/iteration/findmin/1d 41018 ns 40619 ns 1.01
array/iteration/findmin/2d 93431 ns 93219 ns 1.00
array/reductions/reduce/1d 35842 ns 34826 ns 1.03
array/reductions/reduce/2d 40715 ns 50621 ns 0.80
array/reductions/mapreduce/1d 32909 ns 32476 ns 1.01
array/reductions/mapreduce/2d 41241 ns 50769 ns 0.81
array/broadcast 20512 ns 20427 ns 1.00
array/copyto!/gpu_to_gpu 13698 ns 11886 ns 1.15
array/copyto!/cpu_to_gpu 208479.5 ns 207751.5 ns 1.00
array/copyto!/gpu_to_cpu 243344 ns 245794 ns 0.99
array/accumulate/1d 108244 ns 109015 ns 0.99
array/accumulate/2d 80382 ns 79626 ns 1.01
array/construct 1302.7 ns 1306 ns 1.00
array/random/randn/Float32 43238 ns 43298.5 ns 1.00
array/random/randn!/Float32 26422 ns 26052 ns 1.01
array/random/rand!/Int64 26958 ns 26998 ns 1.00
array/random/rand!/Float32 8688.5 ns 8602.333333333334 ns 1.01
array/random/rand/Int64 29725 ns 29780 ns 1.00
array/random/rand/Float32 13025 ns 12942 ns 1.01
array/permutedims/4d 61583 ns 60894 ns 1.01
array/permutedims/2d 55519.5 ns 55115 ns 1.01
array/permutedims/3d 55841.5 ns 55898 ns 1.00
array/sorting/1d 2775794 ns 2776458 ns 1.00
array/sorting/by 3367253 ns 3369147.5 ns 1.00
array/sorting/2d 1084250 ns 1084406 ns 1.00
cuda/synchronization/stream/auto 1004 ns 1025.7 ns 0.98
cuda/synchronization/stream/nonblocking 6261.8 ns 6461.2 ns 0.97
cuda/synchronization/stream/blocking 776.5849056603773 ns 789.3663366336634 ns 0.98
cuda/synchronization/context/auto 1160.1 ns 1164 ns 1.00
cuda/synchronization/context/nonblocking 6588 ns 6604.6 ns 1.00
cuda/synchronization/context/blocking 903.7234042553191 ns 889.4285714285714 ns 1.02

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

@maleadt maleadt merged commit 9455b65 into master Mar 10, 2025
2 of 3 checks passed
@maleadt maleadt deleted the ksh/cudadrv branch March 10, 2025 05:52
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@github-actions github-actions[bot] github-actions[bot] left review comments

@maleadt maleadt maleadt approved these changes

Assignees
No one assigned
Labels
tests Adds or changes tests.
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@kshyatt @maleadt