Skip to content

Don't load the profiler on arm64#7169

Merged
andrewlock merged 4 commits into
masterfrom
andrew/dont-load-the-profiler
Jul 4, 2025
Merged

Don't load the profiler on arm64#7169
andrewlock merged 4 commits into
masterfrom
andrew/dont-load-the-profiler

Conversation

@andrewlock

@andrewlock andrewlock commented Jul 2, 2025 •
edited
Loading

Copy link
Copy Markdown
Member

Summary of changes

Reason for change

#7153 ran into what we think is glibc dlclose issue (i.e. #7125). Given the profiler always bails out on arm64 as it's not yet supported, this is the easiest way to ensure that we don't load (and therefore, more importantly, we don't unload the profiler on arm64).

Implementation details

Comment out the profiler from the loader.conf for arm64. That ensures we don't try to load it on arm64, while making it easy to reenable for testing etc.

Test coverage

Cherry picked the #7153 smoke tests, to confirm it resolves the issues

Other details

Supersedes #7153 given that it enables the tests to verify the fix

@andrewlock andrewlock added area:shared-components area:profiler Issues related to the continous-profiler labels Jul 2, 2025
@dd-trace-dotnet-ci-bot

dd-trace-dotnet-ci-bot Bot commented Jul 2, 2025 •
edited
Loading

Copy link
Copy Markdown

Execution-Time Benchmarks Report ⏱️

Execution-time results for samples comparing the following branches/commits:

Execution-time benchmarks measure the whole time it takes to execute a program. And are intended to measure the one-off costs. Cases where the execution time results for the PR are worse than latest master results are shown in red. The following thresholds were used for comparing the execution times:

  • Welch test with statistical test for significance of 5%
  • Only results indicating a difference greater than 5% and 5 ms are considered.

Note that these results are based on a single point-in-time result for each branch. For full results, see the dashboard.

Graphs show the p99 interval based on the mean and StdDev of the test run, as well as the mean value of the run (shown as a diamond below the graph).

gantt
    title Execution time (ms) FakeDbCommand (.NET Framework 4.8) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (75ms)  : 73, 78
     .   : milestone, 75,
    master - mean (74ms)  : 73, 76
     .   : milestone, 74,

    section Baseline
    This PR (7169) - mean (71ms)  : 68, 75
     .   : milestone, 71,
    master - mean (71ms)  : 67, 75
     .   : milestone, 71,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (1,032ms)  : 1014, 1050
     .   : milestone, 1032,
    master - mean (1,034ms)  : 993, 1075
     .   : milestone, 1034,
Loading 
gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (108ms)  : 106, 111
     .   : milestone, 108,
    master - mean (108ms)  : 106, 109
     .   : milestone, 108,

    section Baseline
    This PR (7169) - mean (108ms)  : 104, 112
     .   : milestone, 108,
    master - mean (106ms)  : 104, 109
     .   : milestone, 106,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (718ms)  : 699, 737
     .   : milestone, 718,
    master - mean (712ms)  : 692, 732
     .   : milestone, 712,
Loading 
gantt
    title Execution time (ms) FakeDbCommand (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (96ms)  : 93, 99
     .   : milestone, 96,
    master - mean (95ms)  : 93, 97
     .   : milestone, 95,

    section Baseline
    This PR (7169) - mean (95ms)  : 93, 98
     .   : milestone, 95,
    master - mean (94ms)  : 92, 97
     .   : milestone, 94,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (679ms)  : 660, 697
     .   : milestone, 679,
    master - mean (674ms)  : 654, 693
     .   : milestone, 674,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.8) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (192ms)  : 190, 195
     .   : milestone, 192,
    master - mean (192ms)  : 190, 194
     .   : milestone, 192,

    section Baseline
    This PR (7169) - mean (188ms)  : 185, 192
     .   : milestone, 188,
    master - mean (188ms)  : 185, 192
     .   : milestone, 188,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (1,115ms)  : 1086, 1143
     .   : milestone, 1115,
    master - mean (1,121ms)  : 1090, 1151
     .   : milestone, 1121,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (269ms)  : 266, 273
     .   : milestone, 269,
    master - mean (270ms)  : 267, 273
     .   : milestone, 270,

    section Baseline
    This PR (7169) - mean (269ms)  : 264, 274
     .   : milestone, 269,
    master - mean (268ms)  : 264, 272
     .   : milestone, 268,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (882ms)  : 858, 906
     .   : milestone, 882,
    master - mean (883ms)  : 858, 907
     .   : milestone, 883,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (263ms)  : 260, 266
     .   : milestone, 263,
    master - mean (262ms)  : 259, 265
     .   : milestone, 262,

    section Baseline
    This PR (7169) - mean (263ms)  : 260, 267
     .   : milestone, 263,
    master - mean (262ms)  : 259, 266
     .   : milestone, 262,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (875ms)  : 836, 914
     .   : milestone, 875,
    master - mean (878ms)  : 839, 916
     .   : milestone, 878,
Loading 
gantt
    title Execution time (ms) HttpMessageHandler (.NET 8) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Bailout
    This PR (7169) - mean (262ms)  : 259, 265
     .   : milestone, 262,
    master - mean (261ms)  : 258, 263
     .   : milestone, 261,

    section Baseline
    This PR (7169) - mean (261ms)  : 257, 264
     .   : milestone, 261,
    master - mean (261ms)  : 258, 264
     .   : milestone, 261,

    section CallTarget+Inlining+NGEN
    This PR (7169) - mean (779ms)  : 760, 797
     .   : milestone, 779,
    master - mean (782ms)  : 760, 804
     .   : milestone, 782,
Loading

@pr-commenter

pr-commenter Bot commented Jul 2, 2025 •
edited
Loading

Copy link
Copy Markdown

Benchmarks

Benchmarks Report for benchmark platform 🐌

Benchmarks for #7169 compared to master:

  • 1 benchmarks are slower, with geometric mean 1.245
  • 43 benchmarks have fewer allocations
  • 5 benchmarks have more allocations

The following thresholds were used for comparing the benchmark speeds:

  • Mann–Whitney U test with statistical test for significance of 5%
  • Only results indicating a difference greater than 10% and 0.3 ns are considered.

Allocation changes below 0.5% are ignored.

Benchmark details

Benchmarks.Trace.ActivityBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.ActivityBenchmark.StartStopWithChild‑net472 6.09 KB 6.05 KB -39 B -0.64%
Benchmarks.Trace.ActivityBenchmark.StartStopWithChild‑netcoreapp3.1 5.75 KB 5.7 KB -54 B -0.94%
Benchmarks.Trace.ActivityBenchmark.StartStopWithChild‑net6.0 5.58 KB 5.51 KB -70 B -1.26%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master StartStopWithChild net6.0 11.2μs 59.7ns 322ns 0 0 0 5.58 KB
master StartStopWithChild netcoreapp3.1 14.1μs 67.5ns 286ns 0 0 0 5.75 KB
master StartStopWithChild net472 22.1μs 119ns 651ns 1.03 0.411 0.103 6.09 KB
#7169 StartStopWithChild net6.0 10.5μs 57.2ns 319ns 0 0 0 5.51 KB
#7169 StartStopWithChild netcoreapp3.1 14.4μs 45.7ns 177ns 0 0 0 5.7 KB
#7169 StartStopWithChild net472 21.7μs 118ns 680ns 0.992 0.331 0.11 6.05 KB
Benchmarks.Trace.AgentWriterBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.AgentWriterBenchmark.WriteAndFlushEnrichedTraces‑net472 3.33 KB 3.31 KB -23 B -0.69%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master WriteAndFlushEnrichedTraces net6.0 927μs 26.6ns 103ns 0 0 0 2.71 KB
master WriteAndFlushEnrichedTraces netcoreapp3.1 1.04ms 404ns 1.56μs 0 0 0 2.7 KB
master WriteAndFlushEnrichedTraces net472 1.25ms 357ns 1.38μs 0 0 0 3.33 KB
#7169 WriteAndFlushEnrichedTraces net6.0 933μs 127ns 441ns 0 0 0 2.71 KB
#7169 WriteAndFlushEnrichedTraces netcoreapp3.1 1.04ms 121ns 435ns 0 0 0 2.7 KB
#7169 WriteAndFlushEnrichedTraces net472 1.23ms 739ns 2.86μs 0 0 0 3.31 KB
Benchmarks.Trace.Asm.AppSecBodyBenchmark - Same speed ✔️ More allocations ⚠️

More allocations ⚠️ in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleSimpleBody‑net472 236.35 KB 239.66 KB 3.3 KB 1.40%
Benchmarks.Trace.Asm.AppSecBodyBenchmark.AllCycleMoreComplexBody‑net472 239.87 KB 243.17 KB 3.3 KB 1.38%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master AllCycleSimpleBody net6.0 331μs 1.76μs 8.42μs 0 0 0 197.06 KB
master AllCycleSimpleBody netcoreapp3.1 510μs 1.45μs 5.61μs 0 0 0 204.77 KB
master AllCycleSimpleBody net472 436μs 119ns 460ns 36.6 2.16 0 236.35 KB
master AllCycleMoreComplexBody net6.0 338μs 1.76μs 8.8μs 0 0 0 200.56 KB
master AllCycleMoreComplexBody netcoreapp3.1 495μs 987ns 3.56μs 0 0 0 208.18 KB
master AllCycleMoreComplexBody net472 446μs 106ns 412ns 36.6 2.16 0 239.87 KB
master ObjectExtractorSimpleBody net6.0 311ns 1.77ns 12.3ns 0 0 0 280 B
master ObjectExtractorSimpleBody netcoreapp3.1 409ns 1.93ns 8.2ns 0 0 0 272 B
master ObjectExtractorSimpleBody net472 303ns 0.175ns 0.676ns 0.0442 0 0 281 B
master ObjectExtractorMoreComplexBody net6.0 6.52μs 29.5ns 110ns 0 0 0 3.78 KB
master ObjectExtractorMoreComplexBody netcoreapp3.1 7.76μs 36.2ns 140ns 0 0 0 3.69 KB
master ObjectExtractorMoreComplexBody net472 6.66μs 0.89ns 3.33ns 0.599 0 0 3.8 KB
#7169 AllCycleSimpleBody net6.0 333μs 121ns 470ns 0 0 0 197.61 KB
#7169 AllCycleSimpleBody netcoreapp3.1 510μs 1.13μs 4.23μs 0 0 0 205.37 KB
#7169 AllCycleSimpleBody net472 442μs 343ns 1.33μs 36.6 2.16 0 239.66 KB
#7169 AllCycleMoreComplexBody net6.0 338μs 1.57μs 6.08μs 0 0 0 201.11 KB
#7169 AllCycleMoreComplexBody netcoreapp3.1 501μs 2.2μs 8.51μs 0 0 0 208.78 KB
#7169 AllCycleMoreComplexBody net472 452μs 102ns 396ns 37.9 2.23 0 243.17 KB
#7169 ObjectExtractorSimpleBody net6.0 318ns 1.74ns 10.3ns 0 0 0 280 B
#7169 ObjectExtractorSimpleBody netcoreapp3.1 401ns 2.1ns 10.1ns 0 0 0 272 B
#7169 ObjectExtractorSimpleBody net472 303ns 0.0447ns 0.173ns 0.0443 0 0 281 B
#7169 ObjectExtractorMoreComplexBody net6.0 6.34μs 29.9ns 119ns 0 0 0 3.78 KB
#7169 ObjectExtractorMoreComplexBody netcoreapp3.1 7.7μs 30.4ns 118ns 0 0 0 3.69 KB
#7169 ObjectExtractorMoreComplexBody net472 7.2μs 1.59ns 5.96ns 0.576 0 0 3.8 KB
Benchmarks.Trace.Asm.AppSecEncoderBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeLegacyArgs‑net6.0 2.16 KB 2.15 KB -11 B -0.51%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EncodeArgs net6.0 73.3μs 263ns 984ns 0 0 0 32.41 KB
master EncodeArgs netcoreapp3.1 95.7μs 36.2ns 130ns 0 0 0 32.4 KB
master EncodeArgs net472 107μs 17.6ns 65.8ns 4.82 0 0 32.51 KB
master EncodeLegacyArgs net6.0 143μs 122ns 472ns 0 0 0 2.16 KB
master EncodeLegacyArgs netcoreapp3.1 197μs 42.5ns 147ns 0 0 0 2.14 KB
master EncodeLegacyArgs net472 261μs 66.8ns 250ns 0 0 0 2.16 KB
#7169 EncodeArgs net6.0 77.8μs 344ns 1.33μs 0 0 0 32.4 KB
#7169 EncodeArgs netcoreapp3.1 97μs 322ns 1.25μs 0 0 0 32.4 KB
#7169 EncodeArgs net472 111μs 8.93ns 33.4ns 4.97 0 0 32.51 KB
#7169 EncodeLegacyArgs net6.0 148μs 41.4ns 149ns 0 0 0 2.15 KB
#7169 EncodeLegacyArgs netcoreapp3.1 196μs 72.9ns 282ns 0 0 0 2.14 KB
#7169 EncodeLegacyArgs net472 261μs 42.9ns 166ns 0 0 0 2.15 KB
Benchmarks.Trace.Asm.AppSecWafBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master RunWafRealisticBenchmark net6.0 271μs 133ns 499ns 0 0 0 4.55 KB
master RunWafRealisticBenchmark netcoreapp3.1 294μs 264ns 989ns 0 0 0 4.48 KB
master RunWafRealisticBenchmark net472 307μs 38.1ns 147ns 0 0 0 4.66 KB
master RunWafRealisticBenchmarkWithAttack net6.0 181μs 80.1ns 300ns 0 0 0 2.24 KB
master RunWafRealisticBenchmarkWithAttack netcoreapp3.1 198μs 104ns 404ns 0 0 0 2.22 KB
master RunWafRealisticBenchmarkWithAttack net472 207μs 45.3ns 169ns 0 0 0 2.28 KB
#7169 RunWafRealisticBenchmark net6.0 279μs 91.7ns 355ns 0 0 0 4.55 KB
#7169 RunWafRealisticBenchmark netcoreapp3.1 294μs 168ns 630ns 0 0 0 4.48 KB
#7169 RunWafRealisticBenchmark net472 308μs 46.4ns 173ns 0 0 0 4.66 KB
#7169 RunWafRealisticBenchmarkWithAttack net6.0 181μs 45.4ns 170ns 0 0 0 2.24 KB
#7169 RunWafRealisticBenchmarkWithAttack netcoreapp3.1 195μs 59.2ns 229ns 0 0 0 2.22 KB
#7169 RunWafRealisticBenchmarkWithAttack net472 208μs 30.4ns 118ns 0 0 0 2.28 KB
Benchmarks.Trace.AspNetCoreBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendRequest net6.0 61.1μs 29ns 104ns 0 0 0 14.53 KB
master SendRequest netcoreapp3.1 69.9μs 118ns 440ns 0 0 0 17.42 KB
master SendRequest net472 0.0171ns 0.00147ns 0.00551ns 0 0 0 0 b
#7169 SendRequest net6.0 60.9μs 57.8ns 216ns 0 0 0 14.52 KB
#7169 SendRequest netcoreapp3.1 72.6μs 84ns 325ns 0 0 0 17.42 KB
#7169 SendRequest net472 0.00803ns 0.00133ns 0.00514ns 0 0 0 0 b
Benchmarks.Trace.CharSliceBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
#7169 OriginalCharSlice net6.0 1.92ms 5.85μs 21.9μs 0 0 0 640.01 KB
#7169 OriginalCharSlice netcoreapp3.1 2.11ms 1.04μs 4.04μs 0 0 0 640 KB
#7169 OriginalCharSlice net472 2.65ms 149ns 558ns 100 0 0 641.95 KB
#7169 OptimizedCharSlice net6.0 1.37ms 525ns 1.89μs 0 0 0 4 B
#7169 OptimizedCharSlice netcoreapp3.1 1.71ms 323ns 1.21μs 0 0 0 1 B
#7169 OptimizedCharSlice net472 2.02ms 566ns 2.19μs 0 0 0 73 B
#7169 OptimizedCharSliceWithPool net6.0 894μs 94ns 364ns 0 0 0 5 B
#7169 OptimizedCharSliceWithPool netcoreapp3.1 936μs 58.1ns 209ns 0 0 0 0 b
#7169 OptimizedCharSliceWithPool net472 1.21ms 96.9ns 375ns 0 0 0 47 B
Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark - Same speed ✔️ More allocations ⚠️

More allocations ⚠️ in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces‑net472 55.75 KB 56.66 KB 910 B 1.63%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master WriteAndFlushEnrichedTraces net6.0 645μs 512ns 1.98μs 0 0 0 41.73 KB
master WriteAndFlushEnrichedTraces netcoreapp3.1 631μs 1.67μs 6.67μs 0 0 0 41.89 KB
master WriteAndFlushEnrichedTraces net472 923μs 1.93μs 7.2μs 4.46 0 0 55.75 KB
#7169 WriteAndFlushEnrichedTraces net6.0 704μs 1.07μs 3.7μs 0 0 0 41.91 KB
#7169 WriteAndFlushEnrichedTraces netcoreapp3.1 657μs 3.38μs 22.9μs 0 0 0 41.94 KB
#7169 WriteAndFlushEnrichedTraces net472 870μs 3.97μs 18.6μs 8.33 0 0 56.66 KB
Benchmarks.Trace.DbCommandBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery‑net6.0 1.03 KB 1.02 KB -8 B -0.78%
Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery‑netcoreapp3.1 1.02 KB 1.02 KB -8 B -0.78%
Benchmarks.Trace.DbCommandBenchmark.ExecuteNonQuery‑net472 995 B 987 B -8 B -0.80%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master ExecuteNonQuery net6.0 1.99μs 10.2ns 46.8ns 0 0 0 1.03 KB
master ExecuteNonQuery netcoreapp3.1 2.53μs 3.43ns 13.3ns 0 0 0 1.02 KB
master ExecuteNonQuery net472 2.7μs 2.41ns 9.34ns 0.147 0.0134 0 995 B
#7169 ExecuteNonQuery net6.0 1.89μs 8.78ns 35.1ns 0 0 0 1.02 KB
#7169 ExecuteNonQuery netcoreapp3.1 2.59μs 11.4ns 44ns 0 0 0 1.02 KB
#7169 ExecuteNonQuery net472 2.83μs 2.08ns 7.48ns 0.156 0.0142 0 987 B
Benchmarks.Trace.ElasticsearchBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync‑net472 1.11 KB 1.1 KB -8 B -0.72%
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync‑netcoreapp3.1 1.09 KB 1.08 KB -8 B -0.74%
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch‑net472 1.05 KB 1.04 KB -8 B -0.76%
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch‑net6.0 1.04 KB 1.03 KB -8 B -0.77%
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch‑netcoreapp3.1 1.04 KB 1.03 KB -8 B -0.77%
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearchAsync‑net6.0 1.02 KB 1.01 KB -8 B -0.79%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master CallElasticsearch net6.0 1.83μs 5.85ns 22.7ns 0 0 0 1.04 KB
master CallElasticsearch netcoreapp3.1 2.3μs 11.8ns 56.6ns 0 0 0 1.04 KB
master CallElasticsearch net472 3.55μs 3.05ns 11.8ns 0.159 0 0 1.05 KB
master CallElasticsearchAsync net6.0 1.81μs 3.89ns 14.5ns 0 0 0 1.02 KB
master CallElasticsearchAsync netcoreapp3.1 2.35μs 7.26ns 28.1ns 0 0 0 1.09 KB
master CallElasticsearchAsync net472 3.84μs 3.19ns 12.4ns 0.169 0 0 1.11 KB
#7169 CallElasticsearch net6.0 1.83μs 3.22ns 12.5ns 0 0 0 1.03 KB
#7169 CallElasticsearch netcoreapp3.1 2.29μs 11.3ns 49ns 0 0 0 1.03 KB
#7169 CallElasticsearch net472 3.72μs 2.47ns 9.24ns 0.151 0 0 1.04 KB
#7169 CallElasticsearchAsync net6.0 1.9μs 8.74ns 32.7ns 0 0 0 1.01 KB
#7169 CallElasticsearchAsync netcoreapp3.1 2.42μs 7.85ns 30.4ns 0 0 0 1.08 KB
#7169 CallElasticsearchAsync net472 3.71μs 3.79ns 14.7ns 0.166 0 0 1.1 KB
Benchmarks.Trace.GraphQLBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync‑net6.0 960 B 952 B -8 B -0.83%
Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync‑netcoreapp3.1 960 B 952 B -8 B -0.83%
Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync‑net472 923 B 915 B -8 B -0.87%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master ExecuteAsync net6.0 1.78μs 9.27ns 43.5ns 0 0 0 960 B
master ExecuteAsync netcoreapp3.1 2.29μs 7.45ns 25.8ns 0 0 0 960 B
master ExecuteAsync net472 2.58μs 1.78ns 6.88ns 0.143 0 0 923 B
#7169 ExecuteAsync net6.0 1.85μs 2.56ns 9.91ns 0 0 0 952 B
#7169 ExecuteAsync netcoreapp3.1 2.39μs 8.76ns 32.8ns 0 0 0 952 B
#7169 ExecuteAsync net472 2.54μs 1.72ns 6.44ns 0.139 0 0 915 B
Benchmarks.Trace.HttpClientBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendAsync net6.0 7.1μs 10.4ns 39ns 0 0 0 2.37 KB
master SendAsync netcoreapp3.1 8.69μs 14.1ns 54.6ns 0 0 0 2.9 KB
master SendAsync net472 12.5μs 10.3ns 38.5ns 0.498 0 0 3.19 KB
#7169 SendAsync net6.0 6.64μs 4.82ns 18.7ns 0 0 0 2.36 KB
#7169 SendAsync netcoreapp3.1 8.62μs 24.6ns 95.4ns 0 0 0 2.9 KB
#7169 SendAsync net472 12.8μs 10.1ns 37.8ns 0.447 0 0 3.18 KB
Benchmarks.Trace.Iast.StringAspectsBenchmark - Slower ⚠️ More allocations ⚠️

Slower ⚠️ in #7169

Benchmark diff/base Base Median (ns) Diff Median (ns) Modality
Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark‑netcoreapp3.1 1.245 413,700.00 514,950.00

More allocations ⚠️ in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark‑net6.0 259.96 KB 286.45 KB 26.49 KB 10.19%
Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatBenchmark‑netcoreapp3.1 42.64 KB 43.37 KB 728 B 1.71%

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatBenchmark‑net6.0 43.83 KB 43.48 KB -352 B -0.80%
Benchmarks.Trace.Iast.StringAspectsBenchmark.StringConcatAspectBenchmark‑net472 286.72 KB 278.53 KB -8.19 KB -2.86%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master StringConcatBenchmark net6.0 44.6μs 212ns 1.6μs 0 0 0 43.83 KB
master StringConcatBenchmark netcoreapp3.1 47.2μs 224ns 838ns 0 0 0 42.64 KB
master StringConcatBenchmark net472 56.8μs 259ns 968ns 0 0 0 57.34 KB
master StringConcatAspectBenchmark net6.0 458μs 1.08μs 3.89μs 0 0 0 259.96 KB
master StringConcatAspectBenchmark netcoreapp3.1 447μs 6.44μs 63.7μs 0 0 0 274.93 KB
master StringConcatAspectBenchmark net472 410μs 2.07μs 9.27μs 0 0 0 286.72 KB
#7169 StringConcatBenchmark net6.0 45.3μs 243ns 2.14μs 0 0 0 43.48 KB
#7169 StringConcatBenchmark netcoreapp3.1 50.6μs 289ns 1.9μs 0 0 0 43.37 KB
#7169 StringConcatBenchmark net472 56.7μs 152ns 548ns 0 0 0 57.34 KB
#7169 StringConcatAspectBenchmark net6.0 495μs 2.15μs 7.45μs 0 0 0 286.45 KB
#7169 StringConcatAspectBenchmark netcoreapp3.1 514μs 2.26μs 8.47μs 0 0 0 275.1 KB
#7169 StringConcatAspectBenchmark net472 404μs 2.14μs 10.9μs 0 0 0 278.53 KB
Benchmarks.Trace.ILoggerBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.ILoggerBenchmark.EnrichedLog‑net6.0 1.76 KB 1.7 KB -56 B -3.18%
Benchmarks.Trace.ILoggerBenchmark.EnrichedLog‑netcoreapp3.1 1.76 KB 1.7 KB -56 B -3.18%
Benchmarks.Trace.ILoggerBenchmark.EnrichedLog‑net472 1.69 KB 1.64 KB -56 B -3.31%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 2.63μs 2.6ns 10.1ns 0 0 0 1.76 KB
master EnrichedLog netcoreapp3.1 3.44μs 4.38ns 17ns 0 0 0 1.76 KB
master EnrichedLog net472 4.05μs 4.83ns 18.7ns 0.265 0 0 1.69 KB
#7169 EnrichedLog net6.0 2.59μs 2.87ns 11.1ns 0 0 0 1.7 KB
#7169 EnrichedLog netcoreapp3.1 3.53μs 14ns 52.4ns 0 0 0 1.7 KB
#7169 EnrichedLog net472 3.97μs 2.9ns 11.2ns 0.258 0 0 1.64 KB
Benchmarks.Trace.Log4netBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.Log4netBenchmark.EnrichedLog‑net6.0 4.37 KB 4.31 KB -56 B -1.28%
Benchmarks.Trace.Log4netBenchmark.EnrichedLog‑netcoreapp3.1 4.37 KB 4.31 KB -56 B -1.28%
Benchmarks.Trace.Log4netBenchmark.EnrichedLog‑net472 4.57 KB 4.51 KB -60 B -1.31%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 123μs 145ns 563ns 0 0 0 4.37 KB
master EnrichedLog netcoreapp3.1 126μs 330ns 1.24μs 0 0 0 4.37 KB
master EnrichedLog net472 167μs 161ns 603ns 0 0 0 4.57 KB
#7169 EnrichedLog net6.0 123μs 180ns 650ns 0 0 0 4.31 KB
#7169 EnrichedLog netcoreapp3.1 127μs 52.9ns 191ns 0 0 0 4.31 KB
#7169 EnrichedLog net472 168μs 146ns 527ns 0 0 0 4.51 KB
Benchmarks.Trace.NLogBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.NLogBenchmark.EnrichedLog‑net6.0 2.32 KB 2.26 KB -56 B -2.41%
Benchmarks.Trace.NLogBenchmark.EnrichedLog‑netcoreapp3.1 2.32 KB 2.26 KB -56 B -2.41%
Benchmarks.Trace.NLogBenchmark.EnrichedLog‑net472 2.14 KB 2.08 KB -56 B -2.62%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 4.94μs 19.3ns 74.6ns 0 0 0 2.32 KB
master EnrichedLog netcoreapp3.1 6.77μs 22.2ns 79.9ns 0 0 0 2.32 KB
master EnrichedLog net472 7.45μs 8.04ns 31.1ns 0.335 0 0 2.14 KB
#7169 EnrichedLog net6.0 4.95μs 22.9ns 91.6ns 0 0 0 2.26 KB
#7169 EnrichedLog netcoreapp3.1 7μs 26.1ns 101ns 0 0 0 2.26 KB
#7169 EnrichedLog net472 7.42μs 8.53ns 33ns 0.298 0 0 2.08 KB
Benchmarks.Trace.RedisBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.RedisBenchmark.SendReceive‑net472 1.21 KB 1.2 KB -8 B -0.66%
Benchmarks.Trace.RedisBenchmark.SendReceive‑net6.0 1.21 KB 1.2 KB -8 B -0.66%
Benchmarks.Trace.RedisBenchmark.SendReceive‑netcoreapp3.1 1.21 KB 1.2 KB -8 B -0.66%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master SendReceive net6.0 2.04μs 10.5ns 49.4ns 0 0 0 1.21 KB
master SendReceive netcoreapp3.1 2.53μs 11.8ns 47.4ns 0 0 0 1.21 KB
master SendReceive net472 3.28μs 2.71ns 10.5ns 0.178 0 0 1.21 KB
#7169 SendReceive net6.0 2μs 10.5ns 50.5ns 0 0 0 1.2 KB
#7169 SendReceive netcoreapp3.1 2.6μs 12.2ns 47.1ns 0 0 0 1.2 KB
#7169 SendReceive net472 3.38μs 3.23ns 12.5ns 0.185 0 0 1.2 KB
Benchmarks.Trace.SerilogBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.SerilogBenchmark.EnrichedLog‑net472 2.08 KB 2.03 KB -56 B -2.69%
Benchmarks.Trace.SerilogBenchmark.EnrichedLog‑netcoreapp3.1 1.69 KB 1.63 KB -56 B -3.32%
Benchmarks.Trace.SerilogBenchmark.EnrichedLog‑net6.0 1.64 KB 1.58 KB -56 B -3.41%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master EnrichedLog net6.0 4.18μs 0.92ns 3.44ns 0 0 0 1.64 KB
master EnrichedLog netcoreapp3.1 5.64μs 23ns 89.3ns 0 0 0 1.69 KB
master EnrichedLog net472 6.67μs 7.28ns 27.3ns 0.298 0 0 2.08 KB
#7169 EnrichedLog net6.0 4.17μs 9.21ns 33.2ns 0 0 0 1.58 KB
#7169 EnrichedLog netcoreapp3.1 5.33μs 20.4ns 78.9ns 0 0 0 1.63 KB
#7169 EnrichedLog net472 6.35μs 5.79ns 22.4ns 0.316 0 0 2.03 KB
Benchmarks.Trace.SpanBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.SpanBenchmark.StartFinishScope‑net6.0 704 B 696 B -8 B -1.14%
Benchmarks.Trace.SpanBenchmark.StartFinishScope‑netcoreapp3.1 704 B 696 B -8 B -1.14%
Benchmarks.Trace.SpanBenchmark.StartFinishScope‑net472 666 B 658 B -8 B -1.20%
Benchmarks.Trace.SpanBenchmark.StartFinishSpan‑net472 586 B 578 B -8 B -1.37%
Benchmarks.Trace.SpanBenchmark.StartFinishSpan‑net6.0 584 B 576 B -8 B -1.37%
Benchmarks.Trace.SpanBenchmark.StartFinishSpan‑netcoreapp3.1 584 B 576 B -8 B -1.37%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master StartFinishSpan net6.0 751ns 3.97ns 22.4ns 0 0 0 584 B
master StartFinishSpan netcoreapp3.1 953ns 4.42ns 17.7ns 0 0 0 584 B
master StartFinishSpan net472 914ns 0.809ns 3.13ns 0.0912 0 0 586 B
master StartFinishScope net6.0 918ns 0.479ns 1.73ns 0 0 0 704 B
master StartFinishScope netcoreapp3.1 1.15μs 6.24ns 34.2ns 0 0 0 704 B
master StartFinishScope net472 1.09μs 0.174ns 0.652ns 0.104 0 0 666 B
#7169 StartFinishSpan net6.0 763ns 0.311ns 1.2ns 0 0 0 576 B
#7169 StartFinishSpan netcoreapp3.1 954ns 5.2ns 27.5ns 0 0 0 576 B
#7169 StartFinishSpan net472 903ns 0.247ns 0.926ns 0.0908 0 0 578 B
#7169 StartFinishScope net6.0 904ns 0.208ns 0.777ns 0 0 0 696 B
#7169 StartFinishScope netcoreapp3.1 1.17μs 6.33ns 35.3ns 0 0 0 696 B
#7169 StartFinishScope net472 1.11μs 0.317ns 1.23ns 0.1 0 0 658 B
Benchmarks.Trace.TraceAnnotationsBenchmark - Same speed ✔️ Fewer allocations 🎉

Fewer allocations 🎉 in #7169

Benchmark Base Allocated Diff Allocated Change Change %
Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin‑net6.0 704 B 696 B -8 B -1.14%
Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin‑netcoreapp3.1 704 B 696 B -8 B -1.14%
Benchmarks.Trace.TraceAnnotationsBenchmark.RunOnMethodBegin‑net472 666 B 658 B -8 B -1.20%

Raw results

Branch Method Toolchain Mean StdError StdDev Gen 0 Gen 1 Gen 2 Allocated
master RunOnMethodBegin net6.0 1.02μs 5.36ns 26.8ns 0 0 0 704 B
master RunOnMethodBegin netcoreapp3.1 1.38μs 2.28ns 8.82ns 0 0 0 704 B
master RunOnMethodBegin net472 1.36μs 0.141ns 0.544ns 0.102 0 0 666 B
#7169 RunOnMethodBegin net6.0 1.07μs 0.273ns 1.06ns 0 0 0 696 B
#7169 RunOnMethodBegin netcoreapp3.1 1.41μs 5.66ns 21.9ns 0 0 0 696 B
#7169 RunOnMethodBegin net472 1.47μs 0.916ns 3.43ns 0.103 0 0 658 B

@andrewlock andrewlock force-pushed the andrew/dont-load-the-profiler branch from 7c98999 to 07e1860 Compare July 2, 2025 14:39
@andrewlock andrewlock marked this pull request as ready for review July 3, 2025 11:11
@andrewlock andrewlock requested a review from a team as a code owner July 3, 2025 11:11
gleocadie
gleocadie approved these changes Jul 3, 2025
View reviewed changes

@gleocadie gleocadie left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

👍

@andrewlock andrewlock merged commit df69689 into master Jul 4, 2025
153 of 155 checks passed
@andrewlock andrewlock deleted the andrew/dont-load-the-profiler branch July 4, 2025 11:13
@github-actions github-actions Bot added this to the vNext-v3 milestone Jul 4, 2025
andrewlock added a commit that referenced this pull request Jul 7, 2025
@andrewlock @e-n-0
Don't load the profiler on arm64 (#7169)
cea9aa1 
## Summary of changes

- Don't try to load the profiler on Linux arm64 at all
- Cherry pick #7153 to see if it resolves the issues

## Reason for change

#7153 ran into what we think is glibc `dlclose` issue (i.e. #7125).
Given the profiler always bails out on arm64 as it's not yet supported,
this is the easiest way to ensure that we don't load (and therefore,
more importantly, we don't _unload_ the profiler on arm64).

## Implementation details

Comment out the profiler from the loader.conf for arm64. That ensures we
don't _try_ to load it on arm64, while making it easy to reenable for
testing etc.

## Test coverage

Cherry picked the #7153 smoke tests, to confirm it resolves the issues

## Other details

Supersedes #7153 given that it enables the tests to verify the fix

---------

Co-authored-by: Flavien Darche <11708575+e-n-0@users.noreply.github.com>
chojomok pushed a commit that referenced this pull request Jul 15, 2025
@andrewlock @e-n-0 @chojomok
Don't load the profiler on arm64 (#7169)
cd0a978 
## Summary of changes

- Don't try to load the profiler on Linux arm64 at all
- Cherry pick #7153 to see if it resolves the issues

## Reason for change

#7153 ran into what we think is glibc `dlclose` issue (i.e. #7125).
Given the profiler always bails out on arm64 as it's not yet supported,
this is the easiest way to ensure that we don't load (and therefore,
more importantly, we don't _unload_ the profiler on arm64).

## Implementation details

Comment out the profiler from the loader.conf for arm64. That ensures we
don't _try_ to load it on arm64, while making it easy to reenable for
testing etc.

## Test coverage

Cherry picked the #7153 smoke tests, to confirm it resolves the issues

## Other details

Supersedes #7153 given that it enables the tests to verify the fix

---------

Co-authored-by: Flavien Darche <11708575+e-n-0@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@gleocadie gleocadie gleocadie approved these changes

Assignees

No one assigned

Labels

area:profiler Issues related to the continous-profiler area:shared-components

Projects

None yet

Milestone

3.21.0

Development

Successfully merging this pull request may close these issues.

3 participants

@andrewlock @gleocadie @e-n-0