1894906 - 2x performance regression when rendering PDFs with smask

Status: NEW

(Core :: Graphics: Canvas2D, defect, P3)

Description

[Marco Castelluccio [:marco]]

STR:

1. Open https://mozilla.github.io/pdf.js/web/viewer.html
2. Load https://bugzilla.mozilla.org/attachment.cgi?id=9372077 through the menu of the PDF viewer (button in the top right corner)

Before bug 1878884, the rendering takes ~5-6 seconds for me. After bug 1878884, it takes ~10-12 seconds.

Comment 1

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

:lsalzman, since you are the author of the regressor, bug 1878884, could you take a look? Also, could you set the severity field?

For more information, please visit BugBot documentation.

Comment 2

[Marco Castelluccio [:marco]]

Profile from today: https://share.firefox.dev/3Qx28Q2.

Profile from January: https://share.firefox.dev/3xWGI8k.

Comment 3

[Marco Castelluccio [:marco]]

Note for performance triage: these kinds of PDFs are pretty common, see the number of duplicates and see also in bug 1874013.

Comment 4

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

Set release status flags based on info from the regressing bug 1878884

Comment 5

[Liz Henry (:lizzard) (relman/hg->git project)]

Is this a duplicate of bug 1895496?

Comment 6

[Marco Castelluccio [:marco]]

(In reply to Liz Henry (:lizzard) (relman/hg->git project) from comment #5)

Is this a duplicate of bug 1895496?

Bug 1895496 is about incorrect rendering with specific PDFs using specific transformations. It is a bug in the PDF viewer.

This bug is instead in Gecko, and it is about the performance regression introduced by bug 1878884 (affecting the same PDF from bug 1895496, but also other kinds of PDFs).

Comment 7

[Bas Schouten (:bas.schouten)]

The Performance Impact Calculator has determined this bug's performance impact to be high. If you'd like to request re-triage, you can reset the Performance Impact flag to "?" or needinfo the triage sheriff.

Platforms: Windows
Page load impact: Severe
[x] Able to reproduce locally

Comment 8

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

The severity field for this bug is set to S3. However, the Performance Impact field flags this bug as having a high impact on the performance.
:lsalzman, could you consider increasing the severity of this performance-impacting bug? Alternatively, if you think the performance impact is lower than previously assessed, could you request a re-triage from the performance team by setting the Performance Impact flag to ??

For more information, please visit BugBot documentation.

Comment 9

[Lee Salzman [:lsalzman]]

This may have been fixed in bug 1907755.

Comment 10

[Marco Castelluccio [:marco]]

It's better now, but a build before bug 1878884 is still better for me than a build from today.

Comment 11

[Frank Doty [:fdoty]]

Revisiting in triage to determine if this is still a High impact based on bug 1878884

Comment 12

[Bas Schouten (:bas.schouten)]

The Performance Impact Calculator has determined this bug's performance impact to be medium. If you'd like to request re-triage, you can reset the Performance Impact flag to "?" or needinfo the triage sheriff.

Platforms: Windows
Page load impact: Some
[x] Able to reproduce locally