Open Bug 1808200 Opened 2 years ago Updated 2 years ago

DAMP performance regression on debugger open (10-20%)

Categories

(DevTools :: Debugger, defect)

defect

Tracking

(firefox-esr102 unaffected, firefox108 unaffected, firefox109 unaffected, firefox110 fix-optional)

Tracking Status
firefox-esr102 --- unaffected
firefox108 --- unaffected
firefox109 --- unaffected
firefox110 --- fix-optional

People

(Reporter: jdescottes, Unassigned, NeedInfo)

References

(Regression)

Details

(Keywords: regression)

== Change summary for alert #36538 (as of Thu, 22 Dec 2022 19:08:13 GMT) ==

Regressions:

Ratio Test Platform Options Absolute values (old vs new)
23% damp simple.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender 567.36 -> 696.62
23% damp console.objectexpand linux1804-64-shippable-qr e10s fission stylo webrender 325.03 -> 399.01
22% damp console.objectexpand linux1804-64-shippable-qr e10s fission stylo webrender-sw 325.26 -> 395.22
21% damp simple.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender-sw 587.61 -> 710.47
20% damp cold.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender-sw 708.39 -> 850.53
20% damp custom.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender 711.67 -> 854.40
18% damp custom.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender-sw 731.00 -> 859.42
16% damp cold.jsdebugger.open.DAMP macosx1015-64-shippable-qr e10s fission stylo webrender 704.59 -> 816.61
13% damp simple.jsdebugger.open.DAMP windows10-64-shippable-qr e10s fission stylo webrender 478.58 -> 541.91
13% damp custom.jsdebugger.open.DAMP windows10-64-shippable-qr e10s fission stylo webrender 650.81 -> 736.72
... ... ... ... ...
5% damp complicated.jsdebugger.open.DAMP windows10-64-shippable-qr e10s fission stylo webrender 854.51 -> 893.02
4% damp browser-toolbox.inspector-ready.DAMP windows10-64-shippable-qr e10s fission stylo webrender-sw 440.06 -> 458.19
4% damp complicated.jsdebugger.open.DAMP windows10-64-shippable-qr e10s fission stylo webrender-sw 860.81 -> 892.85
4% damp browser-toolbox.inspector-ready.DAMP windows10-64-shippable-qr e10s fission stylo webrender 441.99 -> 458.27
3% damp console.openwithcache.open.DAMP linux1804-64-shippable-qr e10s fission stylo webrender 356.98 -> 367.50

Improvements:

Ratio Test Platform Options Absolute values (old vs new)
22% damp browser-toolbox.start-process.DAMP linux1804-64-shippable-qr e10s fission stylo webrender 60.03 -> 46.99
13% damp complicated.jsdebugger.close.DAMP linux1804-64-shippable-qr e10s fission stylo webrender 31.59 -> 27.54
11% damp custom.netmonitor.close.DAMP windows10-64-shippable-qr e10s fission stylo webrender-sw 25.40 -> 22.54
11% damp custom.netmonitor.close.DAMP linux1804-64-shippable-qr e10s fission stylo webrender-sw 20.29 -> 18.06
11% damp complicated.jsdebugger.close.DAMP windows10-64-shippable-qr e10s fission stylo webrender 31.65 -> 28.19
... ... ... ... ...
3% damp simple.netmonitor.requestsFinished.DAMP windows10-64-shippable-qr e10s fission stylo webrender 104.08 -> 100.73

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=36538

Set release status flags based on info from the regressing bug 1805378

:ochameau, since you are the author of the regressor, bug 1805378, could you take a look? Also, could you set the severity field?

For more information, please visit auto_nag documentation.

Flags: needinfo?(poirot.alex)
Keywords: regression

Some comparison for individual changeset as that bug was having lots of various changes:
https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&newRevision=addcfdf843fc6a669cc6e08be45377318f0c4aaf&originalSignature=4081483&newSignature=4081483&framework=12&originalRevision=023928942e60fbb65a3a35dd3c4716a4b51e7856&page=1&showOnlyImportant=1
Nothing

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=addcfdf843fc6a669cc6e08be45377318f0c4aaf&newRevision=b653577e1560b1292fa2c64b3c76444c54887bff&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
Nothing

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=b653577e1560b1292fa2c64b3c76444c54887bff&newRevision=7c21fad7bd94b8a99d14780968574c089d9203a9&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
Nothing

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=7c21fad7bd94b8a99d14780968574c089d9203a9&newRevision=02fa60a78162798d40e5fa2ef3b10206a64a3bb8&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
Nothing

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=02fa60a78162798d40e5fa2ef3b10206a64a3bb8&newRevision=1455d117b1ddf49cb900c7de719b68a3f7a35113&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
https://phabricator.services.mozilla.com/D164547
Too many to report, but:
64% improvement on source-map-loader.init
5-15% improvement on a couple of reload/close tests
15% improvement on browser-toolbox.start-process
6-9% regression on debugger opening tests

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=1455d117b1ddf49cb900c7de719b68a3f7a35113&newRevision=b67b765d57b5403a49cee80df8ad912ed8b7911f&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
https://phabricator.services.mozilla.com/D164692
4% regression on cold.netmonitor.open and simple.netmonitor.close

https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=b67b765d57b5403a49cee80df8ad912ed8b7911f&newRevision=459e7c077d46471cc0d52888d345b01580dcda59&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
https://phabricator.services.mozilla.com/D164694
10% improvement on source-map-loader.getOriginalLocation
200% regression on source-map-loader.init

And a comparision without all patches vs with all the patches:
https://treeherder.mozilla.org/perfherder/comparesubtest?originalProject=try&newProject=try&originalRevision=023928942e60fbb65a3a35dd3c4716a4b51e7856&&newRevision=459e7c077d46471cc0d52888d345b01580dcda59&originalSignature=4081483&newSignature=4081483&framework=12&page=1&showOnlyImportant=1
The regression is smaller on linux, or at least in these try pushes.
It comes only from D164547 and is about 5-9% regression on debugger opening tests

The regression is smaller on linux, or at least in these try pushes.

Correct, I think the regression is smaller on linux. Macos was the one which regressed the most.

You need to log in before you can comment on or make changes to this bug.