Skip to content

[Tracing] Instrument static methods defined in non-generic value types#7920

Merged
zacharycmontoya merged 9 commits intomasterfrom
zach.montoya/instrument-static-valuetype
Feb 10, 2026
Merged

[Tracing] Instrument static methods defined in non-generic value types#7920
zacharycmontoya merged 9 commits intomasterfrom
zach.montoya/instrument-static-valuetype

Conversation

@zacharycmontoya
Copy link
Contributor

@zacharycmontoya zacharycmontoya commented Dec 9, 2025 •
edited by lucaspimentel
Loading

Summary of changes

Adds the capability to instrument static methods defined in non-generic value types

Reason for change

We will need to instrument OpenTelemetry Baggage, which utilizes static methods in a Baggage struct

Implementation details

  • When instrumenting a method for CallTarget instrumentation, optionally update the LocalVarSig with a local variable of the owning type if the following conditions are met:
    1. The method is static
    2. The owning type of the method is a value type
  • Update TracerMethodRewriter::Rewrite such that when it is loading the object instance to put on the stack for a static method, for non-generic value types it does the following:
    • ldloca.s [localIndex] (Loads the variable addresses)
    • initobj [valueType] (Initializes, or re-initializes, an empty struct)
    • ldloc.s [localIndex] (Loads the struct onto the stack)

Test coverage

The CallTargetNativeTest is updated with the following cases to demonstrate the new functionality, especially as it compares to reference types:

  • New ArgumentsGenericStatic types demonstrate the successful instrumentation of static methods in generic reference types
  • New ArgumentsStaticStruct types demonstrate the successful instrumentation of static methods in value types
    • New ArgumentsGenericStaticStruct types demonstrate the unsuccessful instrumentation of static methods in generic value types

Other details

Note: This does not implement the functionality for the live debugger IL rewriting, though it does update the method call so that everything compiles.

@zacharycmontoya zacharycmontoya mentioned this pull request Dec 9, 2025
Open
@zacharycmontoya zacharycmontoya force-pushed the zach.montoya/instrument-static-valuetype branch from e7ef3d6 to 5d57d03 Compare January 29, 2026 23:05
@dd-trace-dotnet-ci-bot
Copy link

dd-trace-dotnet-ci-bot bot commented Jan 29, 2026 •
edited
Loading

Execution-Time Benchmarks Report ⏱️

Execution-time results for samples comparing This PR (7920) and master.

⚠️ Potential regressions detected

FakeDbCommand

Metric Master (Mean ± 95% CI) Current (Mean ± 95% CI) Change Status
.NET Framework 4.8 - Baseline
duration68.30 ± (68.33 - 68.58) ms78.56 ± (78.58 - 78.99) ms+15.0%❌⬆️
.NET Framework 4.8 - Bailout
duration72.05 ± (71.98 - 72.20) ms82.23 ± (82.10 - 82.51) ms+14.1%❌⬆️
.NET Framework 4.8 - CallTarget+Inlining+NGEN
duration1025.15 ± (1027.07 - 1033.25) ms1110.89 ± (1111.19 - 1119.24) ms+8.4%❌⬆️

HttpMessageHandler

Metric Master (Mean ± 95% CI) Current (Mean ± 95% CI) Change Status
.NET Framework 4.8 - Baseline
duration193.79 ± (193.64 - 194.38) ms207.19 ± (207.70 - 209.75) ms+6.9%❌⬆️
.NET Framework 4.8 - Bailout
duration197.11 ± (197.02 - 197.59) ms212.35 ± (212.62 - 214.58) ms+7.7%❌⬆️
Full Metrics Comparison

FakeDbCommand

Metric Master (Mean ± 95% CI) Current (Mean ± 95% CI) Change Status
.NET Framework 4.8 - Baseline
duration68.30 ± (68.33 - 68.58) ms78.56 ± (78.58 - 78.99) ms+15.0%❌⬆️
.NET Framework 4.8 - Bailout
duration72.05 ± (71.98 - 72.20) ms82.23 ± (82.10 - 82.51) ms+14.1%❌⬆️
.NET Framework 4.8 - CallTarget+Inlining+NGEN
duration1025.15 ± (1027.07 - 1033.25) ms1110.89 ± (1111.19 - 1119.24) ms+8.4%❌⬆️
.NET Core 3.1 - Baseline
process.internal_duration_ms22.46 ± (22.44 - 22.48) ms24.77 ± (24.70 - 24.83) ms+10.3%✅⬆️
process.time_to_main_ms86.82 ± (86.67 - 86.96) ms101.53 ± (101.29 - 101.77) ms+16.9%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed15.51 ± (15.51 - 15.52) MB15.50 ± (15.50 - 15.50) MB-0.1%✅
runtime.dotnet.threads.count12 ± (12 - 12)12 ± (12 - 12)+0.0%✅
.NET Core 3.1 - Bailout
process.internal_duration_ms22.40 ± (22.37 - 22.43) ms24.81 ± (24.74 - 24.88) ms+10.7%✅⬆️
process.time_to_main_ms87.84 ± (87.73 - 87.95) ms103.63 ± (103.36 - 103.90) ms+18.0%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed15.55 ± (15.55 - 15.56) MB15.54 ± (15.54 - 15.55) MB-0.1%✅
runtime.dotnet.threads.count13 ± (13 - 13)13 ± (13 - 13)+0.0%✅
.NET Core 3.1 - CallTarget+Inlining+NGEN
process.internal_duration_ms255.49 ± (251.86 - 259.12) ms291.56 ± (289.33 - 293.79) ms+14.1%✅⬆️
process.time_to_main_ms503.66 ± (503.12 - 504.20) ms560.59 ± (559.77 - 561.41) ms+11.3%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed53.23 ± (53.20 - 53.25) MB53.22 ± (53.20 - 53.24) MB-0.0%✅
runtime.dotnet.threads.count28 ± (28 - 28)28 ± (28 - 28)+0.9%✅⬆️
.NET 6 - Baseline
process.internal_duration_ms21.04 ± (21.01 - 21.06) ms23.07 ± (23.01 - 23.13) ms+9.7%✅⬆️
process.time_to_main_ms75.06 ± (74.92 - 75.21) ms87.47 ± (87.23 - 87.71) ms+16.5%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed15.23 ± (15.23 - 15.23) MB15.22 ± (15.21 - 15.22) MB-0.1%✅
runtime.dotnet.threads.count10 ± (10 - 10)10 ± (10 - 10)+0.0%✅
.NET 6 - Bailout
process.internal_duration_ms20.87 ± (20.84 - 20.90) ms23.13 ± (23.08 - 23.19) ms+10.8%✅⬆️
process.time_to_main_ms75.77 ± (75.68 - 75.86) ms89.40 ± (89.15 - 89.65) ms+18.0%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed15.36 ± (15.36 - 15.37) MB15.34 ± (15.34 - 15.35) MB-0.1%✅
runtime.dotnet.threads.count11 ± (11 - 11)11 ± (11 - 11)+0.0%✅
.NET 6 - CallTarget+Inlining+NGEN
process.internal_duration_ms253.80 ± (252.89 - 254.71) ms277.07 ± (276.10 - 278.03) ms+9.2%✅⬆️
process.time_to_main_ms481.12 ± (480.50 - 481.75) ms529.75 ± (528.95 - 530.55) ms+10.1%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed53.98 ± (53.95 - 54.01) MB53.89 ± (53.87 - 53.91) MB-0.2%✅
runtime.dotnet.threads.count28 ± (28 - 28)28 ± (28 - 28)+0.2%✅⬆️
.NET 8 - Baseline
process.internal_duration_ms19.24 ± (19.21 - 19.26) ms21.19 ± (21.13 - 21.25) ms+10.2%✅⬆️
process.time_to_main_ms73.87 ± (73.76 - 73.98) ms86.42 ± (86.23 - 86.61) ms+17.0%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed12.27 ± (12.27 - 12.28) MB12.26 ± (12.25 - 12.26) MB-0.1%✅
runtime.dotnet.threads.count10 ± (10 - 10)10 ± (10 - 10)+0.0%✅
.NET 8 - Bailout
process.internal_duration_ms19.20 ± (19.17 - 19.23) ms20.93 ± (20.86 - 21.00) ms+9.0%✅⬆️
process.time_to_main_ms75.08 ± (75.00 - 75.16) ms86.74 ± (86.52 - 86.95) ms+15.5%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed12.34 ± (12.33 - 12.35) MB12.33 ± (12.32 - 12.34) MB-0.1%✅
runtime.dotnet.threads.count11 ± (11 - 11)11 ± (11 - 11)+0.0%✅
.NET 8 - CallTarget+Inlining+NGEN
process.internal_duration_ms181.20 ± (180.26 - 182.14) ms202.81 ± (201.96 - 203.66) ms+11.9%✅⬆️
process.time_to_main_ms461.65 ± (461.02 - 462.29) ms507.63 ± (506.65 - 508.60) ms+10.0%✅⬆️
runtime.dotnet.exceptions.count0 ± (0 - 0)0 ± (0 - 0)+0.0%✅
runtime.dotnet.mem.committed41.23 ± (41.20 - 41.25) MB41.69 ± (41.65 - 41.72) MB+1.1%✅⬆️
runtime.dotnet.threads.count27 ± (27 - 27)27 ± (27 - 27)+0.2%✅⬆️

HttpMessageHandler

Metric Master (Mean ± 95% CI) Current (Mean ± 95% CI) Change Status
.NET Framework 4.8 - Baseline
duration193.79 ± (193.64 - 194.38) ms207.19 ± (207.70 - 209.75) ms+6.9%❌⬆️
.NET Framework 4.8 - Bailout
duration197.11 ± (197.02 - 197.59) ms212.35 ± (212.62 - 214.58) ms+7.7%❌⬆️
.NET Framework 4.8 - CallTarget+Inlining+NGEN
duration1145.42 ± (1145.20 - 1150.97) ms1203.32 ± (1204.17 - 1211.83) ms+5.1%✅⬆️
.NET Core 3.1 - Baseline
process.internal_duration_ms193.39 ± (192.93 - 193.85) ms205.03 ± (204.36 - 205.71) ms+6.0%✅⬆️
process.time_to_main_ms89.41 ± (89.12 - 89.70) ms94.39 ± (94.03 - 94.74) ms+5.6%✅⬆️
runtime.dotnet.exceptions.count3 ± (3 - 3)3 ± (3 - 3)+0.0%✅
runtime.dotnet.mem.committed20.85 ± (20.83 - 20.88) MB20.66 ± (20.64 - 20.67) MB-0.9%✅
runtime.dotnet.threads.count20 ± (20 - 20)20 ± (20 - 20)+1.5%✅⬆️
.NET Core 3.1 - Bailout
process.internal_duration_ms191.60 ± (191.32 - 191.88) ms202.47 ± (201.91 - 203.04) ms+5.7%✅⬆️
process.time_to_main_ms90.28 ± (90.12 - 90.45) ms95.36 ± (95.07 - 95.66) ms+5.6%✅⬆️
runtime.dotnet.exceptions.count3 ± (3 - 3)3 ± (3 - 3)+0.0%✅
runtime.dotnet.mem.committed20.78 ± (20.76 - 20.81) MB20.62 ± (20.60 - 20.64) MB-0.8%✅
runtime.dotnet.threads.count21 ± (21 - 21)21 ± (21 - 21)+0.9%✅⬆️
.NET Core 3.1 - CallTarget+Inlining+NGEN
process.internal_duration_ms443.88 ± (441.58 - 446.18) ms460.18 ± (457.71 - 462.65) ms+3.7%✅⬆️
process.time_to_main_ms509.20 ± (508.55 - 509.84) ms536.96 ± (536.04 - 537.89) ms+5.5%✅⬆️
runtime.dotnet.exceptions.count3 ± (3 - 3)3 ± (3 - 3)+0.0%✅
runtime.dotnet.mem.committed63.62 ± (63.51 - 63.72) MB63.05 ± (62.93 - 63.18) MB-0.9%✅
runtime.dotnet.threads.count29 ± (29 - 29)29 ± (29 - 29)+0.1%✅⬆️
.NET 6 - Baseline
process.internal_duration_ms196.37 ± (196.07 - 196.67) ms208.49 ± (207.81 - 209.17) ms+6.2%✅⬆️
process.time_to_main_ms77.04 ± (76.81 - 77.26) ms81.51 ± (81.22 - 81.81) ms+5.8%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed20.96 ± (20.93 - 20.99) MB20.78 ± (20.76 - 20.79) MB-0.9%✅
runtime.dotnet.threads.count19 ± (19 - 19)20 ± (20 - 20)+0.9%✅⬆️
.NET 6 - Bailout
process.internal_duration_ms195.41 ± (195.11 - 195.70) ms207.72 ± (207.04 - 208.39) ms+6.3%✅⬆️
process.time_to_main_ms78.00 ± (77.86 - 78.14) ms82.30 ± (81.97 - 82.63) ms+5.5%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed21.02 ± (20.99 - 21.05) MB20.90 ± (20.88 - 20.92) MB-0.6%✅
runtime.dotnet.threads.count20 ± (20 - 21)21 ± (20 - 21)+0.5%✅⬆️
.NET 6 - CallTarget+Inlining+NGEN
process.internal_duration_ms463.92 ± (462.32 - 465.51) ms481.41 ± (479.47 - 483.34) ms+3.8%✅⬆️
process.time_to_main_ms487.27 ± (486.58 - 487.96) ms513.97 ± (512.83 - 515.10) ms+5.5%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed63.13 ± (63.02 - 63.24) MB62.59 ± (62.48 - 62.70) MB-0.9%✅
runtime.dotnet.threads.count30 ± (30 - 30)30 ± (30 - 30)+0.6%✅⬆️
.NET 8 - Baseline
process.internal_duration_ms194.33 ± (194.00 - 194.66) ms206.59 ± (205.97 - 207.21) ms+6.3%✅⬆️
process.time_to_main_ms77.18 ± (76.98 - 77.38) ms81.38 ± (81.06 - 81.70) ms+5.4%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed16.32 ± (16.30 - 16.34) MB16.22 ± (16.21 - 16.24) MB-0.6%✅
runtime.dotnet.threads.count19 ± (19 - 19)19 ± (19 - 19)+1.0%✅⬆️
.NET 8 - Bailout
process.internal_duration_ms193.13 ± (192.82 - 193.45) ms207.27 ± (206.58 - 207.96) ms+7.3%✅⬆️
process.time_to_main_ms77.91 ± (77.76 - 78.06) ms82.72 ± (82.40 - 83.04) ms+6.2%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed16.42 ± (16.39 - 16.45) MB16.29 ± (16.28 - 16.30) MB-0.8%✅
runtime.dotnet.threads.count20 ± (20 - 20)20 ± (20 - 20)+1.2%✅⬆️
.NET 8 - CallTarget+Inlining+NGEN
process.internal_duration_ms371.48 ± (370.24 - 372.72) ms430.93 ± (423.69 - 438.18) ms+16.0%✅⬆️
process.time_to_main_ms468.84 ± (468.24 - 469.44) ms492.24 ± (491.22 - 493.26) ms+5.0%✅⬆️
runtime.dotnet.exceptions.count4 ± (4 - 4)4 ± (4 - 4)+0.0%✅
runtime.dotnet.mem.committed52.98 ± (52.94 - 53.01) MB54.92 ± (54.80 - 55.04) MB+3.7%✅⬆️
runtime.dotnet.threads.count29 ± (29 - 29)29 ± (29 - 29)+0.3%✅⬆️
Comparison explanation

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 highlighted 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).

Duration charts
FakeDbCommand (.NET Framework 4.8) 
gantt
    title Execution time (ms) FakeDbCommand (.NET Framework 4.8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (79ms)  : 76, 82
    master - mean (68ms)  : 67, 70

    section Bailout
    This PR (7920) - mean (82ms)  : crit, 80, 85
    master - mean (72ms)  : 71, 73

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (1,115ms)  : crit, 1055, 1176
    master - mean (1,030ms)  : 986, 1074
Loading
FakeDbCommand (.NET Core 3.1) 
gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (134ms)  : 130, 138
    master - mean (115ms)  : 113, 118

    section Bailout
    This PR (7920) - mean (137ms)  : crit, 132, 141
    master - mean (116ms)  : 114, 118

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (889ms)  : crit, 839, 940
    master - mean (797ms)  : 742, 852
Loading
FakeDbCommand (.NET 6) 
gantt
    title Execution time (ms) FakeDbCommand (.NET 6)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (118ms)  : 114, 122
    master - mean (101ms)  : 99, 104

    section Bailout
    This PR (7920) - mean (120ms)  : crit, 115, 125
    master - mean (102ms)  : 101, 103

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (845ms)  : crit, 818, 873
    master - mean (776ms)  : 755, 797
Loading
FakeDbCommand (.NET 8) 
gantt
    title Execution time (ms) FakeDbCommand (.NET 8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (116ms)  : 113, 120
    master - mean (100ms)  : 97, 102

    section Bailout
    This PR (7920) - mean (116ms)  : crit, 113, 119
    master - mean (101ms)  : 99, 102

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (753ms)  : crit, 728, 778
    master - mean (686ms)  : 673, 698
Loading
HttpMessageHandler (.NET Framework 4.8) 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (209ms)  : 193, 224
    master - mean (194ms)  : 190, 198

    section Bailout
    This PR (7920) - mean (214ms)  : crit, 199, 228
    master - mean (197ms)  : 195, 200

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (1,208ms)  : 1154, 1262
    master - mean (1,148ms)  : 1107, 1189
Loading
HttpMessageHandler (.NET Core 3.1) 
gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (310ms)  : 294, 326
    master - mean (292ms)  : 284, 300

    section Bailout
    This PR (7920) - mean (309ms)  : crit, 294, 323
    master - mean (291ms)  : 287, 295

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (1,038ms)  : 996, 1080
    master - mean (991ms)  : 953, 1029
Loading
HttpMessageHandler (.NET 6) 
gantt
    title Execution time (ms) HttpMessageHandler (.NET 6)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (301ms)  : 287, 314
    master - mean (282ms)  : 278, 287

    section Bailout
    This PR (7920) - mean (300ms)  : crit, 289, 311
    master - mean (282ms)  : 279, 286

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (1,033ms)  : 989, 1076
    master - mean (988ms)  : 950, 1026
Loading
HttpMessageHandler (.NET 8) 
gantt
    title Execution time (ms) HttpMessageHandler (.NET 8)
    dateFormat  x
    axisFormat %Q
    todayMarker off
    section Baseline
    This PR (7920) - mean (301ms)  : 285, 316
    master - mean (282ms)  : 277, 287

    section Bailout
    This PR (7920) - mean (303ms)  : crit, 286, 319
    master - mean (281ms)  : 277, 286

    section CallTarget+Inlining+NGEN
    This PR (7920) - mean (968ms)  : crit, 852, 1083
    master - mean (873ms)  : 849, 896
Loading

@pr-commenter
Copy link

pr-commenter bot commented Jan 29, 2026 •
edited
Loading

Benchmarks

Benchmark execution time: 2026-02-06 17:36:29

Comparing candidate commit b58ce81 in PR branch zach.montoya/instrument-static-valuetype with baseline commit c0dad86 in branch master.

Found 11 performance improvements and 4 performance regressions! Performance is the same for 155 metrics, 22 unstable metrics.

scenario:Benchmarks.Trace.ActivityBenchmark.StartStopWithChild netcoreapp3.1

  • 🟥 throughput [-8880.295op/s; -6810.032op/s] or [-8.986%; -6.891%]

scenario:Benchmarks.Trace.AgentWriterBenchmark.WriteAndFlushEnrichedTraces net6.0

  • 🟥 execution_time [+99.995ms; +100.168ms] or [+98.908%; +99.079%]

scenario:Benchmarks.Trace.Asm.AppSecBodyBenchmark.ObjectExtractorSimpleBody net6.0

  • 🟥 execution_time [+12.009ms; +18.311ms] or [+5.999%; +9.147%]

scenario:Benchmarks.Trace.Asm.AppSecEncoderBenchmark.EncodeLegacyArgs netcoreapp3.1

  • 🟩 execution_time [-24.967ms; -23.908ms] or [-12.239%; -11.720%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces net472

  • 🟩 execution_time [-33.120ms; -28.913ms] or [-15.024%; -13.115%]
  • 🟩 throughput [+155.960op/s; +176.232op/s] or [+15.344%; +17.339%]

scenario:Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark.WriteAndFlushEnrichedTraces netcoreapp3.1

  • 🟩 execution_time [-60.705ms; -56.914ms] or [-27.668%; -25.940%]

scenario:Benchmarks.Trace.CharSliceBenchmark.OriginalCharSlice net6.0

  • 🟩 execution_time [-310.421µs; -218.912µs] or [-14.012%; -9.881%]
  • 🟩 throughput [+51.678op/s; +72.573op/s] or [+11.449%; +16.078%]

scenario:Benchmarks.Trace.GraphQLBenchmark.ExecuteAsync net6.0

  • 🟩 throughput [+48686.525op/s; +59174.393op/s] or [+9.698%; +11.787%]

scenario:Benchmarks.Trace.ILoggerBenchmark.EnrichedLog net6.0

  • 🟩 execution_time [-16.531ms; -11.350ms] or [-7.728%; -5.306%]

scenario:Benchmarks.Trace.Log4netBenchmark.EnrichedLog netcoreapp3.1

  • 🟩 execution_time [-27.192ms; -25.757ms] or [-13.528%; -12.814%]

scenario:Benchmarks.Trace.RedisBenchmark.SendReceive net6.0

  • 🟩 throughput [+25197.643op/s; +28169.556op/s] or [+5.014%; +5.606%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishScope netcoreapp3.1

  • 🟥 execution_time [+15.018ms; +19.301ms] or [+7.588%; +9.752%]

scenario:Benchmarks.Trace.SpanBenchmark.StartFinishTwoScopes net6.0

  • 🟩 execution_time [-17.245ms; -15.335ms] or [-7.918%; -7.041%]

@zacharycmontoya
Implement support for instrumenting static methods in value types in …
ab368d3 
…the tracer. This skips instrumentation for the live debugger product, and it needs to be tested against structs that are generic instantiations.
@zacharycmontoya
Set up the With0Arguments classes in CallTargetNativeTest to test tha…
accc336 
…t static methods in reference types and value types can be instrumented. Also add tests for static methods inside generic reference type and generic value types -- this only succeeds for generic reference types.
@zacharycmontoya
Implement testing of static methods in structs for With1Arguments tes…
da37999 
…t types
@zacharycmontoya
Add tests for instrumenting methods with 2-9 arguments
cbfdd00
@zacharycmontoya zacharycmontoya force-pushed the zach.montoya/instrument-static-valuetype branch from 5d57d03 to cbfdd00 Compare January 30, 2026 22:40
@zacharycmontoya zacharycmontoya marked this pull request as ready for review January 30, 2026 23:53
@zacharycmontoya zacharycmontoya requested review from a team as code owners January 30, 2026 23:53
chatgpt-codex-connector[bot]
chatgpt-codex-connector bot reviewed Jan 31, 2026
View reviewed changes
Copy link

@chatgpt-codex-connector chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

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

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: cbfdd004c0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

tracer/src/Datadog.Tracer.Native/tracer_method_rewriter.cpp Outdated
Comment on lines +210 to +214
if (caller->type.valueType)
{
// Static methods in a ValueType can't be instrumented.
// In the future this can be supported by adding a local for the valuetype and initialize it to the default
// value. After the signature modification we need to emit the following IL to initialize and load into the
// stack.
// ldloca.s [localIndex]
// initobj [valueType]
// ldloc.s [localIndex]
Logger::Warn("*** CallTarget_RewriterCallback(): Static methods in a ValueType cannot be instrumented. ");
return S_FALSE;
reWriterWrapper.LoadLocalAddress(staticValueTypeIndex);
if (caller->type.type_spec != mdTypeSpecNil)
{
Copy link

@chatgpt-codex-connector chatgpt-codex-connector bot Jan 31, 2026

Choose a reason for hiding this comment

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

P2 Badge Avoid passing value-type instance when CallTarget falls back to object

For static methods on value types nested in generic parents, GetCurrentTypeRef falls back to mdTokenNil, so WriteBeginMethod/EndMethod specialize CallTarget with TTarget=object. This branch now always emits initobj + ldloc for any value type, so the stack holds an unboxed struct while the calltarget method spec expects object, which can produce invalid IL/JIT failures when instrumenting Outer<T>.Inner.StaticMethod. Consider checking the generic-parent case (same condition used in GetCurrentTypeRef) and using LoadNull or boxing instead of pushing the value type.

Useful? React with 👍 / 👎.

Copy link
Contributor Author

@zacharycmontoya zacharycmontoya Feb 3, 2026

Choose a reason for hiding this comment

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

This was useful as I didn't cover this edge case. Although we don't instrument this case today, in our CallTargetNativeTest program we can enable this scenario. With the latest set of commits, we correctly handle this and load a null value as the CallTarget "instance" so we can in fact instrument this scenario.

@zacharycmontoya
Add support for instrumenting static methods inside value types, wher…
919ab1a 
…e the value type is enclosed in a generic parent type.

We can easily support this because in such cases we cannot obtain the actual type token for struct, so we default to loading a null value on the stack as the "instance" for the CallTarget instrumentation, just as we would for instrumenting the static method of a reference type.
@zacharycmontoya zacharycmontoya added the area:native-library Automatic instrumentation native C++ code (Datadog.Trace.ClrProfiler.Native) label Feb 3, 2026
tonyredondo
tonyredondo approved these changes Feb 6, 2026
View reviewed changes
Copy link
Member

@tonyredondo tonyredondo left a comment

Choose a reason for hiding this comment

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

LGTM

andrewlock
andrewlock approved these changes Feb 6, 2026
View reviewed changes
Copy link
Member

@andrewlock andrewlock left a comment

Choose a reason for hiding this comment

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

Nice, thanks! Just tiny nits/suggestions!

zacharycmontoya and others added 4 commits February 6, 2026 08:30
@zacharycmontoya @andrewlock
Apply suggestion from @andrewlock
c80005f 
Co-authored-by: Andrew Lock <andrew.lock@datadoghq.com>
@zacharycmontoya @andrewlock
Apply suggestion from @andrewlock
d9dd260 
Co-authored-by: Andrew Lock <andrew.lock@datadoghq.com>
@zacharycmontoya @andrewlock
Apply suggestion from @andrewlock
9acf025 
Co-authored-by: Andrew Lock <andrew.lock@datadoghq.com>
@zacharycmontoya
PR Feedback: Add code comment in CallTargetNativeTest code to more cl…
b58ce81 
…early call out the instrumentation scenarios that don't work today
@lucaspimentel lucaspimentel changed the title [Tracing] Instrument static methods defined in non-generic value type classes [Tracing] Instrument static methods defined in non-generic value types Feb 9, 2026
@zacharycmontoya zacharycmontoya merged commit 9509d91 into master Feb 10, 2026
145 checks passed
@zacharycmontoya zacharycmontoya deleted the zach.montoya/instrument-static-valuetype branch February 10, 2026 15:23
@github-actions github-actions bot added this to the vNext-v3 milestone Feb 10, 2026
@andrewlock andrewlock added the area:automatic-instrumentation Automatic instrumentation managed C# code (Datadog.Trace.ClrProfiler.Managed) label Feb 23, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@chatgpt-codex-connector chatgpt-codex-connector[bot] chatgpt-codex-connector[bot] left review comments

@tonyredondo tonyredondo tonyredondo approved these changes

@dudikeleti dudikeleti dudikeleti approved these changes

@andrewlock andrewlock andrewlock approved these changes

Assignees

No one assigned

Labels

area:automatic-instrumentation Automatic instrumentation managed C# code (Datadog.Trace.ClrProfiler.Managed) area:native-library Automatic instrumentation native C++ code (Datadog.Trace.ClrProfiler.Native)

Projects

None yet

Milestone

3.38.0

Development

Successfully merging this pull request may close these issues.

4 participants

@zacharycmontoya @tonyredondo @dudikeleti @andrewlock