1622025 - Fission-friendly RDM is really, really slow

Status: RESOLVED FIXED

(DevTools :: Responsive Design Mode, defect, P3)

Description

[Sören Hentzschel]

Attachment: screencast

The new Fission-friendly is really slow, please see the attached screencast. First I open the RDM, then I close it. In Both cases you can see the poor performance. I don't think that fixing bug 1621937 will be enough because even with bug 1621937 fixed it will still be much slower than the old RDM.

The screencast was captured on a MacBook Pro Late 2018 with 2,6 GHz 6-Core Intel Core i7 and 16 GiB RAM, so it should be powerful enough…

Comment 1

[Micah [:mtigley] (she/her)]

Thank you for reporting! Indeed, bug 1621937 only solves that particular UX issue. I'll have this block the dt-rdm-fission meta bug to make sure we investigate this further.

Comment 2

[Chris Peterson [:cpeterson]]

Tracking Fission DevTools bugs for Fission Nightly (M6) milestone

Comment 3

[Brad Werth [:bradwerth]]

Even though I can't get a particularly slow start, I will attempt to run a profile on startup and see what jumps out.

Comment 4

[Brad Werth [:bradwerth]]

It's possible we could get better performance if we blocked reflow/layout until the RDM panel was in the right spot (not clipped) and the Touch Simulation setting was set. I'll see if we have options to do that.

Comment 5

[Brad Werth [:bradwerth]]

Emilio noted that we could set display:none on the RDM content to turn off reflow until we have everything ready. visibility:hidden would turn off draw, but that seems unlikely to be the slow part, if there is a slow part.

Comment 6

[Firefox Bug Husbandry Bot]

Because this bug's Severity has not been changed from the default since it was filed, and it's Priority is P3 (Backlog,) indicating it has been triaged, the bug's Severity is being updated to S3 (normal.)

Comment 7

[Sören Hentzschel]

FYI: This is much better now and I can't see a serious performance problem anymore, at least on my device. There were some optimizations mentioned in the previous comments, maybe these are still useful, otherwise feel free to close the ticket if you feel it's good enough.

Comment 8

[Chris Peterson [:cpeterson]]

(In reply to Sören Hentzschel from comment #7)

FYI: This is much better now and I can't see a serious performance problem anymore, at least on my device. There were some optimizations mentioned in the previous comments, maybe these are still useful, otherwise feel free to close the ticket if you feel it's good enough.

Micah, given Sören's report here, can we close this bug report about RDM Fission performance?

Comment 9

[Micah [:mtigley] (she/her)]

(In reply to Sören Hentzschel from comment #7)

FYI: This is much better now and I can't see a serious performance problem anymore, at least on my device. There were some optimizations mentioned in the previous comments, maybe these are still useful, otherwise feel free to close the ticket if you feel it's good enough.

Thanks for updating here! I haven't been able to observe significant performance improvements testing out the suggestions in Comment 5. Let's close this issue then.

Comment 10

[Brad Werth [:bradwerth]]

I think that part of the issue is one of perception. In my testing, Firefox has a lot of pop-in effects when transitioning in/out of RDM. Chrome Device Mode has none. This is interesting because normal Chrome rendering of a webpage might still have some pop-in, but that doesn't occur when transitioning in and out of their Device Mode. So perhaps Chrome is gating the first paint until all the Device Mode reflow has finished, which is different from their heuristic for normal browsing.