Verify version of glibc targeted in native tracer/profiler by andrewlock · Pull Request #6104 · DataDog/dd-trace-dotnet

andrewlock · 2024-10-01T11:18:29Z

Summary of changes

~~Bail out of tracing if glibc < 2.17 on x64 or < 2.23 on arm64~~
Add validation checks to the native libraries to ensure we don't exceed these

Reason for change

The native tracer and profiler are compiled against glibc/musl and as such have minimum versions they require - the musl version is generally not an issue, but glibc can be. The required glibc version is generally controlled by the system on which it's built. For our native build we have

x64 - CentOS 7 with glibc 2.17
arm64 - Debian 10 with glibc 2.28

If we run on a glibc version that is too low, then we don't fail to load because we use RTLD_LAZY lazy binding (instead of RTLD_NOW):

Perform lazy binding. Resolve symbols only as the code that references them is executed. If the symbol is never referenced, then it is never resolved. (Lazy
binding is performed only for function references; references to variables are always immediately bound when the shared object is loaded.) Since glibc 2.1.1,
this flag is overridden by the effect of the LD_BIND_NOW environment variable.

Ultimately, this means that we might load but crash later when a symbol that is not available is required.

To guard against that, we could add checks for the glibc version into the native loader, and bail out if we detect a low glibc version.

Unfortunately, things get tricky with arm64, where we technically require 2.28. .NET 5-8 only require 2.23, and the popular Amazon Linux 2 distro only requires 2.26. So if we were to block <2.28 that could be a significant problem.

The reason this isn't a problem today is that while the profiler requires 2.28, it's not currently supported on arm64. The native tracer, meanwhile, currently only requires 2.18.

The longer term solution is to fix our min glibc dependency to match .NET's 2.23 requirement, but that requires a significant investment, likely using a rootfs build technique like the .NET runtime does or something like https://github.com/wheybags/glibc_version_header.

In the meantime, we will bail out on the following:

x64 - glibc < 2.17
arm64 - glibc < 2.23. Note not 2.28

This will be fine as long as

The native tracer doesn't require symbols from > 2.23.
The continuous profiler doesn't enable arm64.

This PR enforces the former condition by verifying the required glibc version in the native binary. We cannot enforce the latter currently, so will need to bear this in mind when we enable arm64 profiling. Given that those thresholds are also the minimum for .NET, we will not verify them in the native loader directly, instead we will rely on tests to assert we don't exceed 2.23 in the native tracer.

Implementation details

The glibc check in the native loader uses the same technique as the SSI injector in https://github.com/DataDog/auto_inject/pull/433 (which is compatible with musl, and even busybox) to retrieve the glibc version.

To retrieve the actual glibc symbols used by the native library, it's effectively running the following command:

nm --with-symbol-versions -D <TRACERFILE> | grep GLIBC | sed 's/^.*GLIBC_//' | sort -u -t "." -k1,1nr -k2,2nr -k3,3nr | head -1

Which is obviously easier to follow in C# 😉 The nm command returns a list of symbols, similar to this:

U __iswctype_l@GLIBC_2.17
U __lxstat@GLIBC_2.17
U __newlocale@GLIBC_2.17

We just grab the suffix and find the highest: that's our real minimum glibc version.

One complication is that the version of nm in centos7 is too old, so it doesn't work 🙄 So we have to perform the check on the debian distro.

Test coverage

I've tested this all manually, and I'll run against the full smoke tests to be sure. But it's kind of hard to test in practice, given we're basically limiting our self to .NET's glibc version now.

Other details

Given that final statement, a reasonable question is "why bother then"? Which is reasonable - I started down this route precisely because we actually require 2.28. So we can always park this if we want.

EDIT: Yeah, decided to drop the check in the native loader in the end

andrewlock · 2024-10-01T11:49:31Z

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 4.6.2) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (70ms)  : 68, 73
     .   : milestone, 70,
    master - mean (71ms)  : 68, 74
     .   : milestone, 71,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (1,107ms)  : 1077, 1137
     .   : milestone, 1107,
    master - mean (1,114ms)  : 1089, 1139
     .   : milestone, 1114,

gantt
    title Execution time (ms) FakeDbCommand (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (109ms)  : 106, 112
     .   : milestone, 109,
    master - mean (110ms)  : 107, 113
     .   : milestone, 110,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (775ms)  : 760, 790
     .   : milestone, 775,
    master - mean (776ms)  : 762, 790
     .   : milestone, 776,

gantt
    title Execution time (ms) FakeDbCommand (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (94ms)  : 90, 97
     .   : milestone, 94,
    master - mean (93ms)  : 90, 95
     .   : milestone, 93,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (729ms)  : 709, 748
     .   : milestone, 729,
    master - mean (730ms)  : 717, 744
     .   : milestone, 730,

gantt
    title Execution time (ms) HttpMessageHandler (.NET Framework 4.6.2) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (191ms)  : 187, 195
     .   : milestone, 191,
    master - mean (190ms)  : 187, 194
     .   : milestone, 190,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (1,215ms)  : 1184, 1246
     .   : milestone, 1215,
    master - mean (1,199ms)  : 1180, 1218
     .   : milestone, 1199,

gantt
    title Execution time (ms) HttpMessageHandler (.NET Core 3.1) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (277ms)  : 274, 281
     .   : milestone, 277,
    master - mean (275ms)  : 270, 279
     .   : milestone, 275,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (945ms)  : 925, 965
     .   : milestone, 945,
    master - mean (939ms)  : 921, 957
     .   : milestone, 939,

gantt
    title Execution time (ms) HttpMessageHandler (.NET 6) 
    dateFormat  X
    axisFormat %s
    todayMarker off
    section Baseline
    This PR (6104) - mean (267ms)  : 258, 277
     .   : milestone, 267,
    master - mean (264ms)  : 260, 267
     .   : milestone, 264,

    section CallTarget+Inlining+NGEN
    This PR (6104) - mean (933ms)  : 905, 961
     .   : milestone, 933,
    master - mean (960ms)  : 851, 1070
     .   : milestone, 960,

datadog-ddstaging · 2024-10-01T12:14:02Z

Datadog Report

Branch report: andrew/add-extra-libc-checks
Commit report: ae01b79
Test service: dd-trace-dotnet

✅ 0 Failed, 369629 Passed, 2357 Skipped, 16h 55m 14.95s Total Time

andrewlock · 2024-10-01T12:16:54Z

Throughput/Crank Report ⚡

Throughput results for AspNetCoreSimpleController comparing the following branches/commits:

Cases where throughput results for the PR are worse than latest master (5% drop or greater), results are shown in red.

Note that these results are based on a single point-in-time result for each branch. For full results, see one of the many, many dashboards!

gantt
    title Throughput Linux x64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6104) (11.203M)   : 0, 11203149
    master (11.072M)   : 0, 11071593
    benchmarks/2.9.0 (11.081M)   : 0, 11080577

    section Automatic
    This PR (6104) (7.334M)   : 0, 7333864
    master (7.300M)   : 0, 7300144
    benchmarks/2.9.0 (7.732M)   : 0, 7732233

    section Trace stats
    master (7.747M)   : 0, 7746563

    section Manual
    master (11.039M)   : 0, 11038752

    section Manual + Automatic
    This PR (6104) (6.784M)   : 0, 6784483
    master (6.779M)   : 0, 6779324

    section DD_TRACE_ENABLED=0
    master (10.270M)   : 0, 10269736

gantt
    title Throughput Linux arm64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6104) (9.584M)   : 0, 9583994
    master (9.551M)   : 0, 9550662
    benchmarks/2.9.0 (9.798M)   : 0, 9798067

    section Automatic
    This PR (6104) (6.557M)   : 0, 6557286
    master (6.526M)   : 0, 6526183

    section Trace stats
    master (6.825M)   : 0, 6824918

    section Manual
    master (9.511M)   : 0, 9510563

    section Manual + Automatic
    This PR (6104) (4.745M)   : crit ,0, 4745497
    master (6.042M)   : 0, 6042167

    section DD_TRACE_ENABLED=0
    master (8.958M)   : 0, 8958244

gantt
    title Throughput Windows x64 (Total requests) 
    dateFormat  X
    axisFormat %s
    section Baseline
    This PR (6104) (10.100M)   : 0, 10099991
    master (10.136M)   : 0, 10135995
    benchmarks/2.9.0 (10.067M)   : 0, 10067315

    section Automatic
    This PR (6104) (6.794M)   : 0, 6793546
    master (6.765M)   : 0, 6765109
    benchmarks/2.9.0 (7.552M)   : 0, 7552193

    section Trace stats
    master (7.398M)   : 0, 7398168

    section Manual
    master (10.039M)   : 0, 10039465

    section Manual + Automatic
    This PR (6104) (6.138M)   : 0, 6138373
    master (6.241M)   : 0, 6241272

    section DD_TRACE_ENABLED=0
    master (9.564M)   : 0, 9564233

andrewlock · 2024-10-01T12:20:56Z

Benchmarks Report for tracer 🐌

Benchmarks for #6104 compared to master:

1 benchmarks are faster, with geometric mean 1.151
1 benchmarks are slower, with geometric mean 1.181
All benchmarks have the same 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 ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Gen 2	Allocated
master	`StartStopWithChild`	net6.0	8μs	45ns	305ns	0.0166	0.00829	0	5.42 KB
master	`StartStopWithChild`	netcoreapp3.1	10μs	53ns	265ns	0.0234	0.00935	0	5.61 KB
master	`StartStopWithChild`	net472	16.2μs	71.7ns	278ns	1.03	0.314	0.105	6.07 KB
#6104	`StartStopWithChild`	net6.0	7.64μs	41.4ns	262ns	0.0193	0.00773	0	5.43 KB
#6104	`StartStopWithChild`	netcoreapp3.1	9.98μs	48.8ns	195ns	0.0262	0.0105	0	5.61 KB
#6104	`StartStopWithChild`	net472	16.3μs	65.3ns	253ns	1.01	0.287	0.0956	6.07 KB

Benchmarks.Trace.AgentWriterBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`WriteAndFlushEnrichedTraces`	net6.0	494μs	221ns	796ns	0	2.7 KB
master	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	650μs	394ns	1.52μs	0	2.7 KB
master	`WriteAndFlushEnrichedTraces`	net472	831μs	806ns	3.12μs	0.417	3.3 KB
#6104	`WriteAndFlushEnrichedTraces`	net6.0	483μs	392ns	1.52μs	0	2.7 KB
#6104	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	645μs	324ns	1.21μs	0	2.7 KB
#6104	`WriteAndFlushEnrichedTraces`	net472	829μs	391ns	1.41μs	0.411	3.3 KB

Benchmarks.Trace.AspNetCoreBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`SendRequest`	net6.0	199μs	1.13μs	9.82μs	0.207	18.45 KB
master	`SendRequest`	netcoreapp3.1	220μs	1.26μs	10.1μs	0.219	20.61 KB
master	`SendRequest`	net472	0.00129ns	0.000559ns	0.00209ns	0	0 b
#6104	`SendRequest`	net6.0	192μs	1.08μs	7.23μs	0.154	18.45 KB
#6104	`SendRequest`	netcoreapp3.1	215μs	1.15μs	8.75μs	0.203	20.61 KB
#6104	`SendRequest`	net472	0.00135ns	0.000641ns	0.00248ns	0	0 b

Benchmarks.Trace.CIVisibilityProtocolWriterBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Gen 2	Allocated
master	`WriteAndFlushEnrichedTraces`	net6.0	558μs	2.08μs	7.79μs	0.556	0	0	41.68 KB
master	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	682μs	2.28μs	8.53μs	0.359	0	0	41.72 KB
master	`WriteAndFlushEnrichedTraces`	net472	859μs	4.32μs	20.2μs	8.45	2.53	0.422	53.33 KB
#6104	`WriteAndFlushEnrichedTraces`	net6.0	561μs	3.01μs	15.3μs	0.563	0	0	41.63 KB
#6104	`WriteAndFlushEnrichedTraces`	netcoreapp3.1	670μs	3.02μs	11.7μs	0.331	0	0	41.74 KB
#6104	`WriteAndFlushEnrichedTraces`	net472	842μs	3.91μs	15.6μs	8.33	2.5	0.417	53.33 KB

Benchmarks.Trace.DbCommandBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`ExecuteNonQuery`	net6.0	1.26μs	1.86ns	6.71ns	0.0145	1.02 KB
master	`ExecuteNonQuery`	netcoreapp3.1	1.68μs	1.08ns	4.2ns	0.0135	1.02 KB
master	`ExecuteNonQuery`	net472	2.11μs	2.14ns	8.27ns	0.157	987 B
#6104	`ExecuteNonQuery`	net6.0	1.23μs	1.69ns	6.53ns	0.0141	1.02 KB
#6104	`ExecuteNonQuery`	netcoreapp3.1	1.83μs	1.72ns	6.43ns	0.0138	1.02 KB
#6104	`ExecuteNonQuery`	net472	2.08μs	2.64ns	9.86ns	0.157	987 B

Benchmarks.Trace.ElasticsearchBenchmark - Faster 🎉 Same allocations ✔️

Faster 🎉 in #6104

Benchmark	base/diff	Base Median (ns)	Diff Median (ns)	Modality
Benchmarks.Trace.ElasticsearchBenchmark.CallElasticsearch‑net6.0	1.151	1,314.85	1,142.80

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`CallElasticsearch`	net6.0	1.32μs	0.378ns	1.42ns	0.0139	976 B
master	`CallElasticsearch`	netcoreapp3.1	1.55μs	0.732ns	2.74ns	0.0131	976 B
master	`CallElasticsearch`	net472	2.51μs	1.37ns	5.31ns	0.157	995 B
master	`CallElasticsearchAsync`	net6.0	1.26μs	0.764ns	2.86ns	0.0132	952 B
master	`CallElasticsearchAsync`	netcoreapp3.1	1.62μs	0.796ns	3.08ns	0.0139	1.02 KB
master	`CallElasticsearchAsync`	net472	2.58μs	0.729ns	2.73ns	0.167	1.05 KB
#6104	`CallElasticsearch`	net6.0	1.14μs	0.65ns	2.43ns	0.0136	976 B
#6104	`CallElasticsearch`	netcoreapp3.1	1.51μs	1.67ns	6.24ns	0.0134	976 B
#6104	`CallElasticsearch`	net472	2.44μs	1.64ns	6.37ns	0.157	995 B
#6104	`CallElasticsearchAsync`	net6.0	1.3μs	0.635ns	2.38ns	0.0131	952 B
#6104	`CallElasticsearchAsync`	netcoreapp3.1	1.54μs	1.33ns	4.97ns	0.0138	1.02 KB
#6104	`CallElasticsearchAsync`	net472	2.6μs	1.95ns	7.56ns	0.167	1.05 KB

Benchmarks.Trace.GraphQLBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`ExecuteAsync`	net6.0	1.33μs	0.71ns	2.75ns	0.0135	952 B
master	`ExecuteAsync`	netcoreapp3.1	1.61μs	0.699ns	2.71ns	0.0128	952 B
master	`ExecuteAsync`	net472	1.8μs	0.943ns	3.4ns	0.145	915 B
#6104	`ExecuteAsync`	net6.0	1.24μs	0.774ns	3ns	0.0135	952 B
#6104	`ExecuteAsync`	netcoreapp3.1	1.69μs	1.61ns	6.03ns	0.0126	952 B
#6104	`ExecuteAsync`	net472	1.8μs	0.43ns	1.61ns	0.145	915 B

Benchmarks.Trace.HttpClientBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`SendAsync`	net6.0	4.37μs	2.01ns	7.78ns	0.0305	2.22 KB
master	`SendAsync`	netcoreapp3.1	5.15μs	3.54ns	13.7ns	0.036	2.76 KB
master	`SendAsync`	net472	7.73μs	2.64ns	10.2ns	0.498	3.15 KB
#6104	`SendAsync`	net6.0	4.1μs	1.88ns	7.03ns	0.0308	2.22 KB
#6104	`SendAsync`	netcoreapp3.1	5.06μs	2.28ns	8.2ns	0.0352	2.76 KB
#6104	`SendAsync`	net472	7.94μs	2.17ns	8.4ns	0.496	3.15 KB

Benchmarks.Trace.ILoggerBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	1.52μs	0.98ns	3.8ns	0.0233	1.64 KB
master	`EnrichedLog`	netcoreapp3.1	2.2μs	0.563ns	2.11ns	0.0221	1.64 KB
master	`EnrichedLog`	net472	2.62μs	1.16ns	4.49ns	0.249	1.57 KB
#6104	`EnrichedLog`	net6.0	1.46μs	0.915ns	3.54ns	0.0233	1.64 KB
#6104	`EnrichedLog`	netcoreapp3.1	2.37μs	2.4ns	8.97ns	0.0223	1.64 KB
#6104	`EnrichedLog`	net472	2.56μs	1.03ns	3.72ns	0.249	1.57 KB

Benchmarks.Trace.Log4netBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Allocated
master	`EnrichedLog`	net6.0	115μs	188ns	728ns	0	0	4.28 KB
master	`EnrichedLog`	netcoreapp3.1	118μs	92.9ns	335ns	0	0	4.28 KB
master	`EnrichedLog`	net472	147μs	165ns	641ns	0.664	0.221	4.46 KB
#6104	`EnrichedLog`	net6.0	115μs	112ns	436ns	0	0	4.28 KB
#6104	`EnrichedLog`	netcoreapp3.1	120μs	187ns	726ns	0.0598	0	4.28 KB
#6104	`EnrichedLog`	net472	147μs	67ns	242ns	0.658	0.219	4.46 KB

Benchmarks.Trace.NLogBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	2.97μs	0.945ns	3.66ns	0.0304	2.2 KB
master	`EnrichedLog`	netcoreapp3.1	4.18μs	1.67ns	6.46ns	0.0294	2.2 KB
master	`EnrichedLog`	net472	4.86μs	1.22ns	4.58ns	0.32	2.02 KB
#6104	`EnrichedLog`	net6.0	3.1μs	1.21ns	4.69ns	0.0315	2.2 KB
#6104	`EnrichedLog`	netcoreapp3.1	4.12μs	1.1ns	4.1ns	0.0289	2.2 KB
#6104	`EnrichedLog`	net472	4.88μs	2.05ns	7.67ns	0.319	2.02 KB

Benchmarks.Trace.RedisBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Gen 1	Allocated
master	`SendReceive`	net6.0	1.3μs	0.631ns	2.36ns	0.0163	0	1.14 KB
master	`SendReceive`	netcoreapp3.1	1.78μs	0.815ns	3.16ns	0.0154	0	1.14 KB
master	`SendReceive`	net472	2.14μs	1.5ns	5.8ns	0.183	0	1.16 KB
#6104	`SendReceive`	net6.0	1.34μs	0.873ns	3.38ns	0.0161	0	1.14 KB
#6104	`SendReceive`	netcoreapp3.1	1.73μs	0.902ns	3.37ns	0.015	0	1.14 KB
#6104	`SendReceive`	net472	2.02μs	0.527ns	1.97ns	0.183	0.00101	1.16 KB

Benchmarks.Trace.SerilogBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`EnrichedLog`	net6.0	2.84μs	0.863ns	3.34ns	0.0227	1.6 KB
master	`EnrichedLog`	netcoreapp3.1	3.79μs	12ns	46.3ns	0.0225	1.65 KB
master	`EnrichedLog`	net472	4.38μs	1.55ns	6.01ns	0.322	2.04 KB
#6104	`EnrichedLog`	net6.0	2.68μs	1.16ns	4.33ns	0.0227	1.6 KB
#6104	`EnrichedLog`	netcoreapp3.1	3.88μs	12.3ns	47.7ns	0.0212	1.65 KB
#6104	`EnrichedLog`	net472	4.3μs	1.11ns	4.16ns	0.324	2.04 KB

Benchmarks.Trace.SpanBenchmark - Slower ⚠️ Same allocations ✔️

Slower ⚠️ in #6104

Benchmark	diff/base	Base Median (ns)	Diff Median (ns)	Modality
Benchmarks.Trace.SpanBenchmark.StartFinishScope‑netcoreapp3.1	1.181	668.45	789.37

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`StartFinishSpan`	net6.0	403ns	0.134ns	0.484ns	0.00809	576 B
master	`StartFinishSpan`	netcoreapp3.1	633ns	0.473ns	1.77ns	0.00791	576 B
master	`StartFinishSpan`	net472	690ns	0.27ns	0.936ns	0.0916	578 B
master	`StartFinishScope`	net6.0	484ns	0.31ns	1.2ns	0.00989	696 B
master	`StartFinishScope`	netcoreapp3.1	668ns	0.711ns	2.76ns	0.00939	696 B
master	`StartFinishScope`	net472	857ns	0.218ns	0.816ns	0.105	658 B
#6104	`StartFinishSpan`	net6.0	396ns	0.198ns	0.766ns	0.00807	576 B
#6104	`StartFinishSpan`	netcoreapp3.1	613ns	1.79ns	6.91ns	0.00801	576 B
#6104	`StartFinishSpan`	net472	675ns	0.549ns	2.13ns	0.0917	578 B
#6104	`StartFinishScope`	net6.0	475ns	0.244ns	0.945ns	0.00979	696 B
#6104	`StartFinishScope`	netcoreapp3.1	789ns	0.39ns	1.51ns	0.00937	696 B
#6104	`StartFinishScope`	net472	846ns	0.546ns	2.04ns	0.104	658 B

Benchmarks.Trace.TraceAnnotationsBenchmark - Same speed ✔️ Same allocations ✔️

Raw results

Branch	Method	Toolchain	Mean	StdError	StdDev	Gen 0	Allocated
master	`RunOnMethodBegin`	net6.0	664ns	0.288ns	1.12ns	0.00968	696 B
master	`RunOnMethodBegin`	netcoreapp3.1	874ns	0.535ns	2ns	0.00922	696 B
master	`RunOnMethodBegin`	net472	1.11μs	0.391ns	1.51ns	0.104	658 B
#6104	`RunOnMethodBegin`	net6.0	649ns	0.363ns	1.41ns	0.0097	696 B
#6104	`RunOnMethodBegin`	netcoreapp3.1	958ns	0.676ns	2.62ns	0.00922	696 B
#6104	`RunOnMethodBegin`	net472	1.12μs	0.546ns	2.04ns	0.104	658 B

dromanol

LGTM

…ufficient" This reverts commit 2980c8b.

andrewlock added 3 commits October 1, 2024 11:20

nit: fix duplicate names in profiler steps

fd0e225

Add check for libc version available and bail out if it's insufficient

2980c8b

Add validation of glibc compatibility to profiler and tracer

9bc8610

andrewlock added the area:native-library Automatic instrumentation native C++ code (Datadog.Trace.ClrProfiler.Native) label Oct 1, 2024

andrewlock requested a review from a team as a code owner October 1, 2024 11:18

dromanol approved these changes Oct 1, 2024

View reviewed changes

gleocadie approved these changes Oct 2, 2024

View reviewed changes

kevingosse approved these changes Oct 4, 2024

View reviewed changes

Revert "Add check for libc version available and bail out if it's ins…

ae01b79

…ufficient" This reverts commit 2980c8b.

andrewlock changed the title ~~Add explicit bailing out of the tracer on low glibc version~~ Verify and assert version of glibc targeted in native tracer/profiler Oct 4, 2024

andrewlock merged commit 4e124ee into master Oct 4, 2024

andrewlock deleted the andrew/add-extra-libc-checks branch October 4, 2024 13:23

github-actions Bot added this to the vNext-v3 milestone Oct 4, 2024

andrewlock changed the title ~~Verify and assert version of glibc targeted in native tracer/profiler~~ Verify version of glibc targeted in native tracer/profiler Oct 23, 2024

andrewlock added area:tests unit tests, integration tests and removed area:native-library Automatic instrumentation native C++ code (Datadog.Trace.ClrProfiler.Native) labels Oct 23, 2024

Uh oh!

Conversation

andrewlock commented Oct 1, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary of changes

Reason for change

Implementation details

Test coverage

Other details

Uh oh!

andrewlock commented Oct 1, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Execution-Time Benchmarks Report ⏱️

Uh oh!

datadog-ddstaging Bot commented Oct 1, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Datadog Report

Uh oh!

andrewlock commented Oct 1, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Throughput/Crank Report ⚡

Uh oh!

andrewlock commented Oct 1, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmarks Report for tracer 🐌

Benchmark details

Raw results

Raw results

Raw results

Raw results

Raw results

Faster 🎉 in #6104

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Raw results

Slower ⚠️ in #6104

Raw results

Raw results

Uh oh!

dromanol left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

andrewlock commented Oct 1, 2024 •

edited

Loading

andrewlock commented Oct 1, 2024 •

edited

Loading

datadog-ddstaging Bot commented Oct 1, 2024 •

edited

Loading

andrewlock commented Oct 1, 2024 •

edited

Loading

andrewlock commented Oct 1, 2024 •

edited

Loading