ci: add profiling diff workflow

[camchenry]

No description provided.

[camchenry]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• ci: add profiling diff workflow #336 👈 (View in Graphite)
• perf: prefer-reduce-type-parameter: only generate fixes when requested #335
• perf: no-unnecessary-type-assertion: only do scanning when fixes are generated #334
• perf: no-unnecessary-type-arguments: only compute remove range when creating fixes #333
• perf: no-unnecessary-boolean-literal-compare: only build fixes when needed #332
• perf: no-duplicate-type-constituents: only build fixes when needed #324
• perf: no-confusing-void-expression: only generate fixes and suggestions when fixing #321
• perf: no-array-delete: only get token ranges when fixing #320
• feat!: make fixes/suggestions opt-in #317
• refactor: allow suggestions/fixes to be generated lazily #313
• main

---

How to use the Graphite Merge Queue

Add the label 0-merge to this PR to add it to the merge queue.

You must have a Graphite account in order to use the merge queue. Sign up using this link.

An organization admin has enabled the Graphite Merge Queue in this repository.

Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.

This stack of pull requests is managed by Graphite. Learn more about stacking.

[github-actions]

Benchmark Results

Benchmark Mode: Headless (4525 files, 16 rules)

CPU Profile Comparison

Base Branch (10-27-perf_prefer-reduce-type-parameter_only_generate_fixes_when_requested)

File: tsgolint
Build ID: 70bcc09a74094d751414d4a1e23891232f918d7e
Type: cpu
Time: 2025-10-28 15:24:51 UTC
Duration: 22.09s, Total samples = 68.38s (309.55%)
Showing nodes accounting for 46.54s, 68.06% of 68.38s total
Dropped 2153 nodes (cum <= 0.34s)
      flat  flat%   sum%        cum   cum%
     4.36s  6.38%  6.38%      4.36s  6.38%  runtime.memclrNoHeapPointers
     2.97s  4.34% 10.72%      2.97s  4.34%  internal/runtime/maps.h2 (inline)
     2.54s  3.71% 14.43%      7.97s 11.66%  runtime.mapaccess1_fast64
     2.37s  3.47% 17.90%      3.07s  4.49%  runtime.findObject
     2.24s  3.28% 21.18%      3.82s  5.59%  runtime.tryDeferToSpanScan
     2.03s  2.97% 24.14%      2.03s  2.97%  github.com/microsoft/typescript-go/internal/tspath.hasRelativePathSegment
     1.20s  1.75% 25.90%      5.46s  7.98%  runtime.scanobject
     1.18s  1.73% 27.63%      1.18s  1.73%  internal/runtime/maps.h1 (inline)
     0.99s  1.45% 29.07%      0.99s  1.45%  internal/runtime/maps.ctrlGroup.matchH2 (inline)
     0.97s  1.42% 30.49%      0.97s  1.42%  aeshashbody
     0.85s  1.24% 31.73%      2.16s  3.16%  runtime.mapaccess1_faststr
     0.84s  1.23% 32.96%      0.84s  1.23%  runtime.spanClass.sizeclass (inline)

PR Branch (10-27-ci_add_profiling_diff_workflow)

File: tsgolint
Build ID: b092de7f4e30efb5a41b7cf6993532692d6bc1db
Type: cpu
Time: 2025-10-28 15:21:08 UTC
Duration: 21.56s, Total samples = 67.14s (311.44%)
Showing nodes accounting for 45.85s, 68.29% of 67.14s total
Dropped 2163 nodes (cum <= 0.34s)
      flat  flat%   sum%        cum   cum%
     3.96s  5.90%  5.90%      3.96s  5.90%  runtime.memclrNoHeapPointers
     2.83s  4.22% 10.11%      2.83s  4.22%  internal/runtime/maps.h2 (inline)
     2.53s  3.77% 13.88%      8.04s 11.97%  runtime.mapaccess1_fast64
     2.29s  3.41% 17.29%      2.88s  4.29%  runtime.findObject
     2.20s  3.28% 20.57%      3.49s  5.20%  runtime.tryDeferToSpanScan
     1.89s  2.82% 23.38%      1.89s  2.82%  github.com/microsoft/typescript-go/internal/tspath.hasRelativePathSegment
     1.12s  1.67% 25.05%      1.12s  1.67%  aeshashbody
     1.11s  1.65% 26.71%      1.11s  1.65%  internal/runtime/maps.h1 (inline)
     1.06s  1.58% 28.28%      1.06s  1.58%  internal/runtime/maps.ctrlGroup.matchH2 (inline)
     0.97s  1.44% 29.73%      4.79s  7.13%  runtime.scanobject
     0.82s  1.22% 30.95%      0.82s  1.22%  runtime.memhash64
     0.66s  0.98% 31.93%      0.66s  0.98%  internal/runtime/maps.(*ctrlGroup).setEmpty (inline)

Differential Profile

Comparing PR against base (positive values = PR is slower):

File: tsgolint
Build ID: b092de7f4e30efb5a41b7cf6993532692d6bc1db
Type: cpu
Time: 2025-10-28 15:21:08 UTC
Duration: 43.65s, Total samples = 135.52s (310.48%)
Showing nodes accounting for -1.44s, 1.06% of 135.52s total
Dropped 88 nodes (cum <= 0.68s)
      flat  flat%   sum%        cum   cum%
    -0.40s   0.3%   0.3%     -0.40s   0.3%  runtime.memclrNoHeapPointers
    -0.27s   0.2%  0.49%     -0.25s  0.18%  runtime.mapaccess1_faststr
    -0.25s  0.18%  0.68%     -0.25s  0.18%  runtime.spanClass.sizeclass (inline)
    -0.23s  0.17%  0.85%     -0.67s  0.49%  runtime.scanobject
     0.17s  0.13%  0.72%      0.24s  0.18%  github.com/microsoft/typescript-go/internal/checker.(*Checker).getReducedType
     0.15s  0.11%  0.61%      0.15s  0.11%  aeshashbody
    -0.14s   0.1%  0.72%     -0.14s   0.1%  github.com/microsoft/typescript-go/internal/tspath.hasRelativePathSegment
    -0.14s   0.1%  0.82%     -0.14s   0.1%  internal/runtime/maps.h2 (inline)
     0.14s   0.1%  0.72%      0.14s   0.1%  internal/runtime/syscall.Syscall6
     0.13s 0.096%  0.62%     -0.34s  0.25%  gcWriteBarrier
    -0.13s 0.096%  0.72%     -0.10s 0.074%  github.com/microsoft/typescript-go/internal/ast.GetSourceFileOfNode (inline)
    -0.13s 0.096%  0.81%     -0.13s 0.096%  github.com/microsoft/typescript-go/internal/ast.GetThisContainer
    -0.13s 0.096%  0.91%     -0.13s 0.096%  internal/runtime/maps.(*Map).directoryAt (inline)
     0.13s 0.096%  0.81%      0.13s 0.096%  internal/runtime/maps.(*groupsReference).group (inline)
    -0.13s 0.096%  0.91%     -0.13s 0.096%  runtime.typePointers.nextFast (inline)
    -0.12s 0.089%     1%     -0.12s 0.089%  runtime.releasem (inline)
    -0.12s 0.089%  1.08%     -0.13s 0.096%  runtime.spanOf (inline)

View CPU Profiles

Note: Download the benchmark artifacts to analyze CPU profiles locally with pprof.

# Interactive analysis
go tool pprof -http=:8080 -base base/cpu.prof pr/cpu.prof

---

📊 Download the benchmark artifacts to view flame graphs and detailed profiles.