Skip to content

fix(v8): use jemalloc with pointer compression - #372

Open
ldm0 wants to merge 1 commit into
mainfrom
recap
Open

fix(v8): use jemalloc with pointer compression#372
ldm0 wants to merge 1 commit into
mainfrom
recap

Conversation

@ldm0

@ldm0 ldm0 commented Sep 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Member

Use jemalloc for performance.

@github-actions

github-actions Bot commented Sep 4, 2026

Copy link
Copy Markdown

Spider Bench A/B

✅ All benchmark browser service runs completed; results are informational.

Public HEAD: 184 / 240 rows; 38 / 48 sites produced rows; 10 unexpected empty sites.

✅ Deterministic fixture contract is clean.

Common ancestor 8c34e035d3b2 → HEAD 944fea9b6cef; benchmark order: head-first; workflow: success.

Public 48-site run · informational

Metric Base HEAD Δ
Run status success success —
Extracted rows / contract 189 / 240 184 / 240 -5 (-2.65%)
Contract fill 78.75% 76.67% -2.08 pp (-2.64%)
Sites with rows 39 / 48 38 / 48 -1 (-2.56%)
Sites meeting row contract 37 / 48 36 / 48 -1 (-2.70%)
Unexpected empty sites 9 10 +1 (+11.11%)
Partial-row sites 2 2 0 (0.00%)
Site outcomes extracted: 38; HTTP error + rows: 1; empty: 6; HTTP error + empty: 1; snapshot error + empty: 2 extracted: 36; HTTP error + rows: 2; empty: 7; HTTP error + empty: 1; snapshot error + empty: 2 —
Observed duration 67.41 s 61.11 s -6.30 s (-9.34%)
Estimated CPU 49.48 s 31.58 s -17.90 s (-36.17%)
Peak PSS 227.69 MiB 274.79 MiB +47.09 MiB (+20.68%)
Peak RSS 231.23 MiB 278.23 MiB +47.00 MiB (+20.33%)
Peak threads 27 27 0 (0.00%)
Mismatched sites 0 0 0
Timeout with items 0 0 0
Category rows Base HEAD Δ
news 35 / 40 35 / 40 0
stocks 20 / 40 20 / 40 0
tech 35 / 40 35 / 40 0
sports 32 / 40 27 / 40 -5
games 37 / 40 37 / 40 0
life 30 / 40 30 / 40 0

Public-site content, timing, and memory are noisy. This single A/B run reports evidence only; site outcome counts explain missing rows without treating them as a deterministic regression.

CPU and memory timelines

These are bounded 40-point views of the same complete process-tree samples used by the HTML report. CPU uses 100% per occupied logical core; memory lines are RSS first and PSS second when complete PSS samples are available.

Base

xychart-beta
    title "Base CPU"
    x-axis "Elapsed seconds" [2.037, 4.04, 5.543, 6.043, 7.546, 9.549, 11.55, 13.052, 15.055, 17.057, 19.06, 20.562, 21.064, 22.566, 24.568, 26.573, 28.074, 30.075, 32.075, 34.077, 35.579, 37.581, 39.582, 41.585, 43.088, 45.09, 47.093, 49.096, 50.599, 52.601, 54.603, 56.606, 58.108, 60.111, 62.113, 64.115, 65.616, 66.116]
    y-axis "CPU percent" 0 --> 500
    line [83.88, 139.76, 207.44, 283.69, 2, 141.61, 189.85, 55.97, 103.91, 89.92, 51.97, 3.99, 0, 0, 0, 0, 2, 8, 88, 125.95, 89.94, 121.79, 103.94, 113.9, 19.95, 85.82, 135.66, 17.98, 0, 0, 1.99, 0, 0, 113.89, 165.89, 113.85, 146, 161.97]
Loading 
xychart-beta
    title "Base memory: RSS then PSS"
    x-axis "Elapsed seconds" [0.035, 2.037, 4.04, 5.543, 6.043, 7.546, 9.549, 11.55, 13.052, 15.055, 17.057, 19.06, 20.562, 21.064, 22.566, 24.568, 26.573, 28.074, 30.075, 32.075, 34.077, 35.579, 37.581, 39.582, 41.585, 43.088, 45.09, 47.093, 49.096, 50.599, 52.601, 54.603, 56.606, 58.108, 60.111, 62.113, 64.115, 65.616, 66.116, 67.407]
    y-axis "Memory MiB" 0 --> 500
    line [11.89, 108.52, 140.99, 167.59, 176.57, 161.06, 178.07, 186.95, 161.65, 174.61, 160.16, 155.73, 154.43, 142.19, 143.42, 142.07, 142.07, 142.07, 140.82, 142.82, 170.92, 171.03, 204.62, 170.46, 191.96, 157.57, 160.44, 215.7, 172.5, 164.59, 163.62, 164.06, 165.05, 166.6, 184.62, 183.35, 185.79, 228.1, 231.23, 0]
    line [8.69, 105.03, 137.45, 164.06, 173.04, 157.52, 174.09, 183.43, 157.99, 171.06, 156.61, 152.18, 150.88, 138.64, 139.87, 138.51, 138.51, 138.51, 137.27, 139.27, 168.05, 167.47, 197.97, 166.42, 188.46, 154.08, 156.89, 212.3, 168.94, 161.04, 160.06, 160.51, 161.49, 163.04, 181.08, 180.95, 182.24, 224.6, 227.69, 0]
Loading

HEAD

xychart-beta
    title "HEAD CPU"
    x-axis "Elapsed seconds" [1.034, 1.534, 2.034, 3.537, 5.54, 7.042, 9.044, 10.547, 11.048, 12.549, 14.552, 16.556, 18.057, 20.06, 21.563, 23.563, 25.564, 27.064, 29.065, 31.068, 32.569, 34.57, 36.072, 38.074, 40.078, 41.58, 43.582, 45.084, 45.584, 47.086, 49.088, 50.589, 52.593, 54.595, 56.097, 58.1, 59.603, 61.104]
    y-axis "CPU percent" 0 --> 200
    line [0, 0, 0, 3.99, 105.74, 103.84, 129.97, 163.58, 67.92, 63.97, 73.95, 23.96, 50, 0, 2, 0, 2, 0, 2, 33.96, 107.92, 105.91, 41.89, 105.93, 67.82, 7.98, 0, 121.72, 101.93, 83.94, 0, 2, 9.97, 125.81, 73.96, 133.77, 19.97, 27.99]
Loading 
xychart-beta
    title "HEAD memory: RSS"
    x-axis "Elapsed seconds" [0.034, 1.034, 1.534, 2.034, 3.537, 5.54, 7.042, 9.044, 10.547, 11.048, 12.549, 14.552, 16.556, 18.057, 20.06, 21.563, 23.563, 25.564, 27.064, 29.065, 31.068, 32.569, 34.57, 36.072, 38.074, 40.078, 41.58, 43.582, 45.084, 45.584, 47.086, 49.088, 50.589, 52.593, 54.595, 56.097, 58.1, 59.603, 61.104, 61.109]
    y-axis "Memory MiB" 0 --> 500
    line [15.32, 56.97, 57.06, 57.06, 57.74, 124.15, 175.92, 201.49, 220.98, 203.92, 205.19, 180.14, 160.57, 166.14, 150.41, 149.87, 149.53, 149.14, 149.31, 151.12, 167.08, 177.04, 234.8, 181.81, 186.61, 175.5, 166.34, 158.49, 277.18, 278.23, 210.39, 171.77, 171.15, 173.88, 203.66, 185.84, 248.43, 182.48, 0, 0]
Loading

Deterministic fixture · required contract

The fixture exercises 15 routes: 8 are expected to emit rows and 7 intentionally exercise empty, loading, or timeout behavior.

Metric Base HEAD Δ
Run status success success —
Extracted rows / contract 40 / 40 40 / 40 0 (0.00%)
Contract fill 100.00% 100.00% 0.00 pp (0.00%)
Sites with rows 8 / 15 8 / 15 0 (0.00%)
Sites meeting row contract 15 / 15 15 / 15 0 (0.00%)
Unexpected empty sites 0 0 0
Partial-row sites 0 0 0
Site outcomes extracted: 7; HTTP error + rows: 1; empty: 3; timeout + empty: 4 extracted: 7; HTTP error + rows: 1; empty: 3; timeout + empty: 4 —
Observed duration 19.89 s 19.88 s -0.02 s (-0.08%)
Estimated CPU 0.43 s 0.36 s -0.07 s (-16.01%)
Peak PSS 57.73 MiB 62.74 MiB +5.01 MiB (+8.68%)
Peak RSS 61.03 MiB 65.98 MiB +4.95 MiB (+8.12%)
Peak threads 18 18 0 (0.00%)
Mismatched sites 0 0 0
Timeout with items 0 0 0

Full HTML, JSON, CSV, logs, and page snapshots: workflow run and full spider-bench-results artifact.

The public-web A/B is informational. The exact HEAD fixture contract runs as a separate required CI check; timing and public-site content deltas remain non-blocking.

@ldm0 ldm0 changed the title fix(v8): decouple pointer compression from allocator policy fix(v8): use jemalloc with pointer compression Sep 4, 2026
@github-actions

github-actions Bot commented Sep 4, 2026

Copy link
Copy Markdown

CI Regression Report

Source CI run · source state at render: in_progress · artifacts: 5/5

Check Status Signal
Release regression ✅ HEAD/base failures 0/0; raw binary +2.815151%
Frontend differential ✅ 1,020/1,020 cases matched; 0 issues
Agent episodes ✅ 8/8 Moli episodes passed; 0 failures
Runtime/CDP contracts ✅ 26 contract cases; 0 failures
CDP smoke ✅ 42/42 groups passed; 447 scenarios
Release regression — ✅ HEAD/base failures 0/0; raw binary +2.815151%

Package and image size

Metric Base HEAD Delta Delta %
Raw binary 149.08 MiB 153.28 MiB +4.20 MiB +2.815151%
Stripped binary 103.35 MiB 103.57 MiB +229.14 KiB +0.216514%
gzip binary 50.14 MiB 51.85 MiB +1.71 MiB +3.408023%
Rootfs 155.98 MiB 160.18 MiB +4.20 MiB +2.692211%
gzip rootfs 58.96 MiB 60.82 MiB +1.86 MiB +3.149313%

Startup latency and PSS

Case Base p50 HEAD p50 Delta Base PSS HEAD PSS Delta
Serve ready 27.07 ms 27.14 ms +0.08 ms 4.84 MiB 4.89 MiB +53.50 KiB
CDP first page 50.34 ms 51.22 ms +0.88 ms 4.13 MiB 4.89 MiB +772.50 KiB
CDP warm page 13.75 ms 13.66 ms −0.09 ms 48.69 MiB 54.38 MiB +5.68 MiB
CLI about:blank 13.14 ms 14.78 ms +1.64 ms 4.15 MiB 4.87 MiB +733.00 KiB
CLI local JS 18.09 ms 16.95 ms −1.14 ms 2.77 MiB 7.80 MiB +5.03 MiB

Concurrency matrix

Concurrency Base p50 HEAD p50 Delta Base PSS HEAD PSS Delta HEAD failures Unstable B/H
1 16.17 ms 16.35 ms +0.18 ms 784.00 KiB 798.00 KiB +14.00 KiB 0 0/0
5 19.86 ms 19.18 ms −0.68 ms 12.22 MiB 29.92 MiB +17.70 MiB 0 0/0
10 19.10 ms 18.83 ms −0.27 ms 26.12 MiB 17.19 MiB −8.94 MiB 0 0/0
25 19.90 ms 18.79 ms −1.10 ms 13.62 MiB 17.07 MiB +3.46 MiB 0 0/0
100 18.81 ms 18.97 ms +0.16 ms 20.03 MiB 30.60 MiB +10.57 MiB 0 0/0
Frontend differential — ✅ 1,020/1,020 cases matched; 0 issues
Match DOM mismatch Diagnostic mismatch Moli error Reference error Infrastructure error Mismatched frames Duration
1,020 0 0 0 0 0 0 6m 4.73s
Agent episodes · Moli vs Chromium — ✅ 8/8 Moli episodes passed; 0 failures
Fast contract Episodes Passed Assertions Ready p50 Episode p50
Moli 8 8 47/47 3.67 ms 23.56 ms
Chromium 8 8 47/47 251.03 ms 79.62 ms
Operation p50 Chromium Moli Delta Delta %
navigate 40.98 ms 6.25 ms −34.73 ms −84.76%
observe 0.81 ms 0.77 ms −0.04 ms −5.35%
fill 3.24 ms 3.09 ms −0.15 ms −4.70%
click 2.58 ms 3.26 ms +0.68 ms +26.46%
Idle-resource episode Peak RSS Peak PSS Average CPU Peak CPU Peak processes Sampler
Moli 72.43 MiB 69.75 MiB 0.65% 10.00% 1 ✅
Chromium 1287.46 MiB 350.57 MiB 1.81% 90.15% 13 ⚠️ PSS partial
Runtime and CDP session contracts — ✅ 26 contract cases; 0 failures
Suite Cases Failures Median case p50 Median case p95 Median peak PSS p50
Synthetic fetch 13 0 16.57 ms 17.02 ms 649.00 KiB
Long-lived CDP session 13 0 17.80 ms 18.43 ms —
CDP smoke — ✅ 42/42 groups passed; 447 scenarios

Workers: 4 · cumulative group time: 123.74 s · failed groups: 0

Slowest group Scenarios Duration Status
chromium-cdp 56 13.72 s ✅
multi-page 45 10.60 s ✅
dom-input 26 10.27 s ✅
layout-screenshot 5 7.73 s ✅
core 35 6.91 s ✅

All artifact fields are parsed by the trusted default-branch renderer; missing or invalid inputs remain visible as unavailable.

@github-actions

github-actions Bot commented Sep 4, 2026

Copy link
Copy Markdown

Sequential Navigation Soak A/B

✅ HEAD completed the 200-navigation resilience and memory observation.

One browser process, one target, and one CDP session navigate CSDN → SegmentFault → Huaban → example.com repeatedly. This run issued 200 Page.navigate commands per binary.

Common ancestor 8c34e035d3 → HEAD 944fea9b6c; benchmark order: head-first; workflow: success.

Session resilience

Metric Base HEAD
Attempted / planned 200 / 200 200 / 200
Direct observable pass 198 / 200 197 / 200
Failures / recovered 2 / 2 3 / 3
Recovery failures 0 0
Lifecycle order violations 0 0
Network order violations 0 0
Wall time 346.9 s 381.5 s

A public-page failure is reported separately from an unrecoverable session. The soak requires all navigation attempts, zero failed recovery, zero lifecycle/network ordering violations, and complete resource evidence.

Process-tree memory

Metric Base HEAD Δ
Boundary samples 200 200 0
Boundary RSS samples 200 200 0
Boundary PSS samples 200 200 0
Periodic peak RSS 232.1 MiB 264.7 MiB +32.6 MiB
Periodic peak PSS 229.3 MiB 261.8 MiB +32.5 MiB
RSS · first 10 avg 171.2 MiB 182.6 MiB +11.3 MiB
RSS · last 10 avg 161.7 MiB 177.4 MiB +15.7 MiB
RSS · last − first window -9.5 MiB -5.2 MiB +4.3 MiB
RSS · warm slope / 100 nav +1.0 MiB -0.7 MiB -1.8 MiB
PSS · first 10 avg 168.3 MiB 179.0 MiB +10.7 MiB
PSS · last 10 avg 159.9 MiB 174.5 MiB +14.5 MiB
PSS · last − first window -8.4 MiB -4.6 MiB +3.8 MiB
PSS · warm slope / 100 nav +1.3 MiB -0.4 MiB -1.7 MiB
Peak file descriptors 60 60 0
Peak threads 31 31 0

HEAD memory by 50-navigation quarter

HEAD range Avg RSS Final RSS Peak RSS Avg PSS Final PSS Peak FD
1–50 181.0 MiB 167.0 MiB 225.9 MiB 177.6 MiB 164.0 MiB 53
51–100 177.3 MiB 134.4 MiB 207.9 MiB 174.3 MiB 131.5 MiB 50
101–150 183.6 MiB 170.0 MiB 257.3 MiB 180.8 MiB 167.1 MiB 50
151–200 178.2 MiB 137.9 MiB 206.5 MiB 175.6 MiB 135.0 MiB 52

Raw reports and per-navigation boundary samples: workflow run and full sequential-navigation-soak-results artifact.

Public-site timing and memory are observational. Use the A/B deltas and trend shape as evidence, not as a deterministic performance threshold.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@ldm0