1902249 - 0.13% installer size (OSX) regression on Wed June 5 2024

Status: RESOLVED WONTFIX

(Testing :: Performance, defect, P5)

Description

[Florin Bilt]

Perfherder has detected a build_metrics performance regression from push cd09151a1c46cba9d56b45684eebc6465a5e869c. As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

Ratio	Test	Platform	Options	Absolute values (old vs new)
0.13%	installer size	osx-aarch64-shippable	aarch64 nightly	92,055,786.08 -> 92,175,068.50

Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests. Please follow our guide to handling regression bugs and let us know your plans within 3 business days, or the patch(es) may be backed out in accordance with our regression policy.

You can run these tests on try with ./mach try perf --alert 632

For more information on performance sheriffing please see our FAQ.

Comment 1

[Liz Henry (:lizzard)]

Emilio, can you take a look to see if this may have come from the patch in bug 1898214? Thanks!

Comment 2

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1898214

Comment 3

[Emilio Cobos Álvarez [:emilio]]

I don't think it can come from there? It doesn't add that much code. Did we see a progression on the back out and subsequent regression? Since the patch linked in comment 0 was backed out.

Comment 4

[Florin Bilt]

(In reply to Emilio Cobos Álvarez (:emilio) from comment #3)

I don't think it can come from there? It doesn't add that much code. Did we see a progression on the back out and subsequent regression? Since the patch linked in comment 0 was backed out.

Thank you for your help. I found the real culprit. I have asked for assistance from the responsible parties for this bugs : Bug 1898214, Bug 1900504 and Bug 1883396.

Comment 5

[Florin Bilt]

Can you please investigate this?

Comment 6

[Alex Franchuk [:afranchuk]]

Per https://bugzilla.mozilla.org/show_bug.cgi?id=1900504#c8, we won't be backing the change out. However I can try to confirm the size change.

Comment 7

[Alex Franchuk [:afranchuk]]

If I've selected the correct comparison here, it doesn't look like 1900504 is the cause of the change: https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=1d43d839bef510db024ecaaa7a13f7def8f9bb9b&newProject=try&newRevision=c6d46c631144931543225110e65ff756d42c4c45&framework=2&page=1&filter=installer+size

Context: I ran two build-macosx64-aarch64-shippable/opt jobs for before and after the patch, and compared the installer size in build_metrics.

Comment 8

[Dave Hunt [:davehunt] [he/him] ⌚BST]

The data seems too noisy to determine the culprit here, but it does look like we have a slow regression on the installer size. In May we were <92.2 and now we're clearly above. See this graph. @sergesanspaille are you able to help here?

Comment 9

[[:sergesanspaille]]

I'll have a look, possibly by comparing libxul's binary between now and two months ago, that should give a reasonable hint of what's happening here

Comment 10

[Kash Shampur [:kshampur] ⌚EST]

hi sguelton as there has been no activity in the past year we are closing this as wontfix. if you feel otherwise please feel free to reopen