Repository navigation
Commit 599145d
[PERF] Retire default Mono performance runs and builds on main (#134602)
<!-- -->
> [!IMPORTANT]
> This is dependent on dotnet/performance#5322. Those shared controls
must land before this change because the current runtime callers pass
the new `runMono` parameter.
## Summary
Retire default Mono performance runs and unnecessary performance builds
from runtime main while preserving CoreCLR and NativeAOT workloads.
- Disable Mono desktop and WASM runs in the current performance callers,
including the public WASM PR sanity pipeline.
- Remove the Mono-only slow nightly schedule while retaining private
CoreCLR Linux and Windows ARM64 runs.
- Default the five Mono build selectors to false. Keep registration and
upload selections unchanged, and route WASM builds through the
runtime-owned wrapper exactly once.
- Make desktop and WASM build wrappers CoreCLR-only by default. Retain
guarded manual Mono builders and source-only gapfill Mono opt-ins.
- Preserve the CoreCLR WASM flavor forwarded by the existing
platform/global-build adapter when preparing artifacts. Preserve
`BrowserWasmCoreCLR` and host tool packs; publish workload-transfer
artifacts only for an explicit manual Mono build.
Older runtime branches are intentionally unchanged. The shared
performance controls default to true for backward compatibility; only
current runtime main callers opt out. General runtime product and CI
Mono jobs, manual performance tools, and guarded builder implementations
remain available.
Related: #134397 and #134337. This change
does not close either issue.
## Validation
Based on runtime main `586bc324f6d3aa3c11aac1d1fde8a2eea072b91e`, pulled
with `git pull --ff-only upstream main`.
- Parsed all seven changed YAML files and ran `git diff --check`.
- Checked 64 Mono/WASM build-selector combinations for
registration/upload symmetry and unique WASM producers.
- Checked 120 normal, PR, manual, scheduled, and project caller
configurations, preserving non-Mono consumers.
- Checked 32 gapfill source/cached/manual combinations, with extra Mono
producers restricted to explicit source-build opt-ins.
- Compared CoreCLR wrapper definitions and default/manual WASM payloads
with the baseline, accounting for the conditional transfer change. The
strengthened local check traces existing parameter forwarding, verifies
effective CoreCLR flavor, and rejects duplicate post-build arguments.
- Cross-checked the shared performance caller/producer/artifact contract
with the companion change.
- Checked boolean and string-valued Mono flags, including effective
preparation scripts, `BrowserWasm` publication, and workload-transfer
artifacts.
Azure Pipelines service-side previews of public pipeline 156 and
internal performance pipeline 702 exposed a duplicate `runtimeFlavor`
parameter in the initial PR head: the global-build adapter already
forwards this parameter, so repeating it in the nested helper arguments
is invalid. The public preview of the runtime baseline with the
performance PR source successfully expanded, confirming this was
introduced by the initial patch. Corrective commit
`7c40ead5642cf32385907884d2b3c16524ec0f42` removes only the redundant
nested argument; the strengthened local check fails before the
correction and passes afterward.
The internal slow pipeline 1012 preview succeeded at the initial PR
sources with `onlySanityCheck: true` and `runPrivateJobs: true`,
producing the monitor, Linux/Windows ARM64 CoreCLR builds, and two
corresponding sanity runs, with no Mono jobs.
Subsequent public/internal performance previews compiled but exposed a
second issue: the legacy mapping-style `runMono` value was tested for
truthiness, admitting the nonempty string `false`. Commit
`61d7faf36fa2204b1480e0061e49238ceca88daf` uses `eq(parameters.runMono,
true)`; the local artifact/script check fails before this correction and
passes afterward for both boolean and string inputs.
### End-to-end sanity test runs (2026-09-24)
All three runs used both PR sources together, with resolved repository
versions verified as runtime
**`61d7faf36fa2204b1480e0061e49238ceca88daf`** (this PR) and performance
**`3ac9d6f5c73175e677c7211d2ea44d0f8496c16b`**
(dotnet/performance#5322).
| Pipeline | Test run | Result |
|---|---|---|
| dotnet-runtime-perf |
[3086818](https://dev.azure.com/dnceng/internal/_build/results?buildId=3086818)
| Failed only on the pre-existing SVE sanity-filter issue; 46/47 Helix
work items passed. |
| dotnet-runtime-perf-slow |
[3086817](https://dev.azure.com/dnceng/internal/_build/results?buildId=3086817)
| Passed: Linux and Windows ARM64 CoreCLR builds and both Helix sanity
work items. |
| runtime-wasm-perf |
[1611277](https://dev.azure.com/dnceng-public/public/_build/results?buildId=1611277)
| Passed: CoreCLR WASM build and both V8 benchmark jobs, including R2R.
|
Final-head service-side expansion succeeded for all three pipelines,
with no Mono build/run jobs or Mono-only workload transfers. All
benchmark commands used `--only-sanity`; the slow run also enabled
`runPrivateJobs` so the CoreCLR ARM64 jobs actually executed. The
internal runs used `parkerbibus/mono-runtime-retirement` and
`parkerbibus/mono-pipeline-controls`; public WASM used the two GitHub PR
head refs.
The standard run's only failed work item,
`arm64.sve.net11.0.Partition0`, built successfully but combined
`--anyCategories SVE` with `--filter System.Tests.Perf_*`, excluding all
SVE benchmark classes under `SveBenchmarks`. The later missing-artifact
exception was secondary. This conflicting argument generation was
reproduced from both the PR and unchanged [upstream-base harness
source](https://github.com/dotnet/performance/blob/1f0e0f68889a0996f30e265846446549a801800e/scripts/run_performance_job.py#L516-L632),
not by running a separate baseline pipeline. [Failure
console](https://helix.dot.net/api/2019-06-17/jobs/1c31a264-58b7-40c0-a7a6-4eeef60eb586/workitems/arm64.sve.net11.0.Partition0/console).
All other builds and work items passed; SVE validation remains outside
this change and was not rerun.
## Known limitations and scope
- Existing gapfill forwarding of undeclared desktop `viperMicroNoR2R`
and `downloadSpecificBuild` parameters remains unchanged. Existing Mono
AOT cached-artifact handling also remains unchanged. Full live gapfill
validity is not claimed.
- The unrelated CoreCLR R2R-interpreter x64 upload versus ARM64 producer
mismatch remains untouched.
- No release-branch changes, runtime-version selectors, live pipeline
service changes, merges, or ready-for-review actions are included.
> [!NOTE]
> This description and the implementation were generated with GitHub
Copilot.
---------
Co-authored-by: Copilot App <[email protected]>1 parent c0c1c33 commit 599145d
7 files changed
Lines changed: 58 additions & 65 deletions
File tree
- eng/pipelines/performance
- templates
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
31 | 31 | | |
32 | 32 | | |
33 | 33 | | |
34 | | - | |
| 34 | + | |
35 | 35 | | |
36 | | - | |
| 36 | + | |
37 | 37 | | |
38 | 38 | | |
39 | 39 | | |
40 | | - | |
| 40 | + | |
41 | 41 | | |
42 | 42 | | |
43 | 43 | | |
44 | | - | |
| 44 | + | |
45 | 45 | | |
46 | 46 | | |
47 | 47 | | |
48 | | - | |
| 48 | + | |
49 | 49 | | |
50 | 50 | | |
51 | 51 | | |
52 | | - | |
| 52 | + | |
53 | 53 | | |
54 | 54 | | |
55 | 55 | | |
| |||
152 | 152 | | |
153 | 153 | | |
154 | 154 | | |
155 | | - | |
156 | | - | |
157 | 155 | | |
158 | 156 | | |
159 | 157 | | |
| |||
166 | 164 | | |
167 | 165 | | |
168 | 166 | | |
169 | | - | |
170 | | - | |
171 | 167 | | |
172 | 168 | | |
173 | 169 | | |
174 | 170 | | |
175 | 171 | | |
| 172 | + | |
| 173 | + | |
| 174 | + | |
| 175 | + | |
| 176 | + | |
| 177 | + | |
176 | 178 | | |
177 | 179 | | |
178 | 180 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
32 | 32 | | |
33 | 33 | | |
34 | 34 | | |
35 | | - | |
| 35 | + | |
36 | 36 | | |
37 | 37 | | |
38 | 38 | | |
39 | | - | |
| 39 | + | |
40 | 40 | | |
41 | 41 | | |
42 | 42 | | |
43 | | - | |
| 43 | + | |
44 | 44 | | |
45 | 45 | | |
46 | 46 | | |
| |||
79 | 79 | | |
80 | 80 | | |
81 | 81 | | |
| 82 | + | |
82 | 83 | | |
83 | 84 | | |
84 | 85 | | |
| |||
90 | 91 | | |
91 | 92 | | |
92 | 93 | | |
| 94 | + | |
| 95 | + | |
| 96 | + | |
| 97 | + | |
| 98 | + | |
| 99 | + | |
| 100 | + | |
93 | 101 | | |
94 | 102 | | |
95 | 103 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
2 | 2 | | |
3 | 3 | | |
4 | 4 | | |
5 | | - | |
6 | | - | |
7 | | - | |
8 | 5 | | |
9 | 6 | | |
10 | 7 | | |
| |||
38 | 35 | | |
39 | 36 | | |
40 | 37 | | |
41 | | - | |
42 | | - | |
43 | | - | |
44 | | - | |
45 | | - | |
46 | | - | |
47 | | - | |
48 | | - | |
49 | 38 | | |
50 | 39 | | |
51 | 40 | | |
| |||
65 | 54 | | |
66 | 55 | | |
67 | 56 | | |
68 | | - | |
69 | | - | |
| 57 | + | |
70 | 58 | | |
71 | 59 | | |
72 | 60 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
32 | 32 | | |
33 | 33 | | |
34 | 34 | | |
35 | | - | |
36 | | - | |
37 | | - | |
38 | | - | |
39 | | - | |
40 | | - | |
41 | | - | |
42 | 35 | | |
43 | 36 | | |
44 | 37 | | |
| |||
67 | 60 | | |
68 | 61 | | |
69 | 62 | | |
70 | | - | |
71 | | - | |
| 63 | + | |
72 | 64 | | |
73 | 65 | | |
74 | 66 | | |
| |||
82 | 74 | | |
83 | 75 | | |
84 | 76 | | |
| 77 | + | |
| 78 | + | |
| 79 | + | |
| 80 | + | |
| 81 | + | |
| 82 | + | |
85 | 83 | | |
86 | 84 | | |
87 | 85 | | |
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
38 | 38 | | |
39 | 39 | | |
40 | 40 | | |
| 41 | + | |
41 | 42 | | |
42 | 43 | | |
43 | 44 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
11 | 11 | | |
12 | 12 | | |
13 | 13 | | |
14 | | - | |
15 | | - | |
16 | | - | |
17 | | - | |
18 | | - | |
19 | | - | |
20 | | - | |
| 14 | + | |
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
| 1 | + | |
| 2 | + | |
| 3 | + | |
1 | 4 | | |
2 | | - | |
3 | | - | |
4 | | - | |
5 | | - | |
6 | | - | |
7 | | - | |
8 | | - | |
9 | | - | |
10 | | - | |
11 | | - | |
12 | | - | |
13 | | - | |
14 | | - | |
15 | | - | |
16 | | - | |
17 | | - | |
18 | | - | |
19 | | - | |
20 | | - | |
21 | | - | |
22 | | - | |
23 | | - | |
| 5 | + | |
| 6 | + | |
| 7 | + | |
| 8 | + | |
| 9 | + | |
| 10 | + | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + | |
| 19 | + | |
| 20 | + | |
| 21 | + | |
| 22 | + | |
| 23 | + | |
| 24 | + | |
| 25 | + | |
24 | 26 | | |
25 | 27 | | |
26 | 28 | | |
| |||
42 | 44 | | |
43 | 45 | | |
44 | 46 | | |
45 | | - | |
| 47 | + | |
0 commit comments