1678643 - Very bad browser jank due to CRLite

Status: RESOLVED DUPLICATE of bug 1677399

(Core :: Security: PSM, defect)

Description

[Denis Palmeiro [:denispal]]

I am seeing very bad jank on the whole browser which basically makes the whole browser unusable, and this seems to be due to CRLite holding up the parent process for long periods of time.

See profile here: https://share.firefox.dev/2KsahGc

The jank is about 30s-1min long then it feels fine for a minute, then it happens again and this seems to repeat.

Comment 1

[Dana Keeler (she/her) [:keeler]]

Are you experiencing this with the latest nightly? Also, what is the value of services.settings.poll_interval?

Comment 2

[Denis Palmeiro [:denispal]]

(In reply to Dana Keeler (she/her) (use needinfo) (:keeler for reviews) from comment #1)

Are you experiencing this with the latest nightly? Also, what is the value of services.settings.poll_interval?

Yes, nightly 2020-11-20. services.settings.poll_interval is set to 86400. After restarting the browser, I haven't experienced any further issues. I'll keep my eye out to see if it resurfaces in the next few days again and try to capture more threads this time.

Comment 3

[Dana Keeler (she/her) [:keeler]]

Thanks! The threads we'd be interested in are the ones created by NS_CreateBackgroundTaskQueue.

Comment 4

[BugBot [:suhaib / :marco]]

The severity field is not set for this bug.
:keeler, could you have a look please?

For more information, please visit auto_nag documentation.

Comment 5

[Dana Keeler (she/her) [:keeler]]

We recently landed two fixes that should have improved things. Can you try to reproduce this again on the latest Nightly? Thanks!

Comment 6

[Dana Keeler (she/her) [:keeler]]

(see comment 5)

Comment 7

[Denis Palmeiro [:denispal]]

I haven't encountered this since I opened this bug, so it does seem to be resolved. Thanks!