Commit abae1c0
fix: make perf benchmark script more reliable
- Fix lsof() bug: .toString (property ref) -> .toString() (method call)
and wrap in try/catch for non-zero exit codes
- Replace fixed sleep(5000) + broken lsof check with waitForServer()
that polls with HTTP fetch until servers actually respond
- Store startVitePreview child processes and add .catch() error handlers
Co-authored-by: NullVoxPopuli <199018+NullVoxPopuli@users.noreply.github.com>
Agent-Logs-Url: https://github.com/emberjs/ember.js/sessions/abc30b6b-e3d1-44c5-988e-648096650173
chore: remove unnecessary comments from benchmark scripts
Co-authored-by: NullVoxPopuli <199018+NullVoxPopuli@users.noreply.github.com>
Agent-Logs-Url: https://github.com/emberjs/ember.js/sessions/b3fd4afb-3714-4a3d-b7d6-acb56ba1a9dc
fix: default to non-headless on macOS to fix bench trace capture
Chrome's --headless=new mode on macOS (Chrome 147+) silently fails to
capture performance.mark() events in traces. Tracerbench then fails
immediately with 'Could not find mark renderStart in trace'.
Non-headless mode (visible browser windows) works correctly on macOS.
Default to non-headless on macOS, keep headless on Linux for CI.
Users can still force headless with --headless, or force non-headless
with --no-headless (now a no-op on macOS since it's already the default).
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
fix: make headless work on macOS via tracerbench patches
Root cause: Chrome 147 (headless=new) on macOS emits an initial blank
navigationStart (documentLoaderURL: '') before the real page navigation.
Tracerbench's findPhases() locked onto the wrong navigationId, so all
performance.mark() events failed the ID check and 'renderStart' was never
found in the trace.
Fix in extract-navigation-sample.js: when a second navigationStart with a
non-empty documentLoaderURL is encountered, update navigationStartArgs so
subsequent markers match the correct navigationId.
Additional improvements to the tracerbench patch:
- inject-mark-observer.js: remove self===top guard and add buffered:true to
PerformanceObserver so pre-existing marks are captured reliably
- run-trace.js: use traceConfig.includedCategories (modern CDP API) with
recordAsMuchAsPossible to avoid trace buffer overflow
Also reverts the earlier macOS non-headless workaround in benchmark.mjs —
headless mode now works on macOS too.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
fix: correct hunk line count in tracerbench patch
After manually editing the run-trace.js comment from 3 lines to 2,
the @@ hunk header still said +10,13 instead of +10,12, breaking
pnpm install with 'hunk header integrity check failed'.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Add lockfile1 parent 0531e62 commit abae1c0
File tree
5 files changed
+108
-23
lines changed- bin
- benchmark
- patches
5 files changed
+108
-23
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
17 | 17 | | |
18 | 18 | | |
19 | 19 | | |
20 | | - | |
21 | 20 | | |
22 | 21 | | |
23 | 22 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
6 | 6 | | |
7 | 7 | | |
8 | 8 | | |
9 | | - | |
| 9 | + | |
10 | 10 | | |
11 | 11 | | |
12 | 12 | | |
| |||
112 | 112 | | |
113 | 113 | | |
114 | 114 | | |
115 | | - | |
116 | | - | |
| 115 | + | |
| 116 | + | |
117 | 117 | | |
118 | 118 | | |
119 | 119 | | |
120 | 120 | | |
| 121 | + | |
| 122 | + | |
| 123 | + | |
| 124 | + | |
| 125 | + | |
| 126 | + | |
| 127 | + | |
121 | 128 | | |
122 | 129 | | |
123 | 130 | | |
| |||
132 | 139 | | |
133 | 140 | | |
134 | 141 | | |
135 | | - | |
136 | | - | |
137 | | - | |
138 | | - | |
139 | | - | |
140 | | - | |
141 | | - | |
142 | | - | |
143 | | - | |
144 | | - | |
145 | | - | |
146 | | - | |
147 | | - | |
148 | | - | |
| 142 | + | |
| 143 | + | |
| 144 | + | |
| 145 | + | |
| 146 | + | |
| 147 | + | |
149 | 148 | | |
150 | 149 | | |
151 | 150 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
23 | 23 | | |
24 | 24 | | |
25 | 25 | | |
26 | | - | |
27 | | - | |
| 26 | + | |
| 27 | + | |
| 28 | + | |
| 29 | + | |
| 30 | + | |
| 31 | + | |
| 32 | + | |
| 33 | + | |
| 34 | + | |
| 35 | + | |
| 36 | + | |
| 37 | + | |
| 38 | + | |
| 39 | + | |
| 40 | + | |
| 41 | + | |
| 42 | + | |
| 43 | + | |
| 44 | + | |
| 45 | + | |
| 46 | + | |
| 47 | + | |
| 48 | + | |
| 49 | + | |
| 50 | + | |
| 51 | + | |
| 52 | + | |
| 53 | + | |
| 54 | + | |
| 55 | + | |
| 56 | + | |
28 | 57 | | |
29 | 58 | | |
30 | 59 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
58 | 58 | | |
59 | 59 | | |
60 | 60 | | |
| 61 | + | |
| 62 | + | |
| 63 | + | |
| 64 | + | |
| 65 | + | |
| 66 | + | |
| 67 | + | |
| 68 | + | |
| 69 | + | |
| 70 | + | |
| 71 | + | |
| 72 | + | |
| 73 | + | |
| 74 | + | |
| 75 | + | |
| 76 | + | |
| 77 | + | |
| 78 | + | |
| 79 | + | |
| 80 | + | |
| 81 | + | |
| 82 | + | |
| 83 | + | |
| 84 | + | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
| 88 | + | |
| 89 | + | |
| 90 | + | |
| 91 | + | |
| 92 | + | |
| 93 | + | |
| 94 | + | |
| 95 | + | |
| 96 | + | |
| 97 | + | |
| 98 | + | |
| 99 | + | |
| 100 | + | |
| 101 | + | |
| 102 | + | |
| 103 | + | |
| 104 | + | |
| 105 | + | |
| 106 | + | |
| 107 | + | |
| 108 | + | |
| 109 | + | |
| 110 | + | |
| 111 | + | |
| 112 | + | |
| 113 | + | |
| 114 | + | |
| 115 | + | |
| 116 | + | |
| 117 | + | |
| 118 | + | |
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.
0 commit comments